Full Terms amp Conditions of access and use can be found athttpwwwtandfonlinecomactionjournalInformationjournalCode=hmpe20

Download by [Luc Pelletier] Date 25 October 2016 At 1130

Measurement in Physical Education and Exercise Science

ISSN 1091-367X (Print) 1532-7841 (Online) Journal homepage httpwwwtandfonlinecomloihmpe20

The Validity of the Interpersonal BehaviorsQuestionnaire (IBQ) in Sport

Meredith Rocchi Luc Pelletier amp Philippe Desmarais

To cite this article Meredith Rocchi Luc Pelletier amp Philippe Desmarais (2016) The Validity ofthe Interpersonal Behaviors Questionnaire (IBQ) in Sport Measurement in Physical Educationand Exercise Science DOI 1010801091367X20161242488

To link to this article httpdxdoiorg1010801091367X20161242488

Published online 25 Oct 2016

Submit your article to this journal

View related articles

View Crossmark data

The Validity of the Interpersonal Behaviors Questionnaire (IBQ) in SportMeredith Rocchi Luc Pelletier and Philippe Desmarais

School of Psychology University of Ottawa Ottawa Ontario Canada

ABSTRACTAccording to Self-Determination Theory (SDT) basic psychological needs will be influenced byother individualsrsquo interpersonal behaviors The objective of the present research is to extend thevalidity of the Interpersonal Behaviors Questionnaire (IBQ and IBQ-Self) to the sport context Themeasure was designed to assess perceptions of interpersonal behaviors of others (IBQ) or self-reports of interpersonal behaviors (IBQ-Self) in the context of SDT This measure consists of 24items and six subscales looking at autonomy-supportive autonomy-thwarting (controlling) com-petence-supportive competence-thwarting relatedness-supportive and relatedness-thwartinginterpersonal behaviors In Study 1 athletes were asked to report on their perceptions of theircoachesrsquo interpersonal behaviors (IBQ) In Study 2 coaches were asked to report on their inter-personal behaviors when they coach their athletes (IBQ-Self) The results supported that the scalehad a strong factor structure internal consistency and validity Overall the results supported theIBQ and IBQ-Self are valid measures of interpersonal behaviors in sport

KEYWORDSathlete developmentinterpersonal behaviorself-determination theorysport

Over the years sport research has moved beyond sim-ply examining athletesrsquo performances and has shiftedfocus to the psychological factors that determinewhether athletes have a successful or unsuccessfulexperience in sport (eg Vallerand 2001) One aspectof athletesrsquo psychological experiences that have receiveda lot of empirical attention is their motivation forparticipating in their sport Self-Determination Theory(SDT Deci amp Ryan 1985) is a leading motivationtheory that has helped guide a significant amount ofresearch in sport and has helped explain how the sportcontext and athletesrsquo psychological experiences interactand influence the reasons athletes participate in sportand enjoy what they do (Hagger amp Chatzisarantis2007)

SDT posits that coachesrsquo interpersonal behaviorstyles play an essential role in determining athletesrsquoexperiences in sport through the extent to which thesebehaviors either support or thwart their athletesrsquo psy-chological needs (Deci amp Ryan 1985) As such researchhas begun to examine athletesrsquo perceptions of thesebehaviors (Mageau amp Vallerand 2003) as well as thefactors that influence coachesrsquo reported interpersonalbehaviors when they coach (Rocchi Pelletier ampCouture 2013) One important limitation to thisresearch is that there is currently no valid measure ofperceptions of interpersonal behaviors or self-reportsof interpersonal behaviors for all six interpersonal

behavior styles identified by SDT Therefore the objec-tive of the present studies is to extend the validity of theInterpersonal Behaviors Questionnaire (IBQ RocchiPelletier Cheung Baxter amp Beaudry 2017) to thesport context

SDT and sport

According to SDT motivation orientations for sportdiffer in their quality based on the degree to whichthe reasons for practicing sport have been internalizedand integrated into an individualrsquos sense of selfPrevious research has supported that when the reasonsfor practicing their sport are more internalized anindividual experiences autonomous motivation wherethey participate because they value andor enjoy it Thisis shown to result in positive outcomes for athletes suchas greater interest in sport better concentration moreenjoyment increased sport satisfaction and improvedcompetitive results (Mageau amp Vallerand 2003) Whenthe reasons for practicing sport are less internalizedthey experience controlled motivation towards theirsport and participate because of external or internalpressures which has been shown to lead to more nega-tive outcomes like burnout sport anxiety or dropout(Li Wang Pyun amp Kee 2013)

In order to encourage this process of internalizationan individual requires the support and satisfaction of

CONTACT Meredith Rocchi MeredithRocchiuottawaca School of Psychology University of Ottawa 136 Jean Jacques Lussier Ottawa Ontario K1N6N5 Canada

MEASUREMENT IN PHYSICAL EDUCATION AND EXERCISE SCIENCEhttpdxdoiorg1010801091367X20161242488

copy 2016 Taylor amp Francis

the three basic psychological needs (autonomy compe-tence and relatedness Deci amp Ryan 2002) In sportthe need for autonomy represents the need for indivi-duals to act in line with their own interests and valueswhile practicing their sport Competence requiresopportunities for athletes to increase the level of chal-lenge in their sport and to develop increased skill mas-tery Finally the need for relatedness refers to needing asupportive network and strong interpersonal connec-tions with other people involved in sport (Vallerand2001) Research has shown that when these psycholo-gical needs are satisfied for athletes it leads to anincrease in sport motivation quality and they experi-ence positive outcomes in sport adversely when theseneeds are frustrated it leads to a decrease in sportmotivation quality and promotes negative outcomes(ie Gould Dieffenbach amp Moffett 2002)

The sport context has an impact on the extent towhich athletesrsquo psychological needs are satisfied or fru-strated (Deci Shwartz Sheinman amp Ryan 1981) Thecontext includes both the structure of the sport (ielevel of competition) as well as the people within it(ie coaches) Focusing on the people only SDT pos-tulates other peoplersquos behavior either positively or nega-tively influences athlete need satisfaction andfrustration Specifically when people in the sport con-text engage in need-supportive interpersonal behaviorsit will promote the satisfaction of the basic psychologi-cal needs for athletes (Deci amp Ryan 1985)Alternatively when people engage in (or are perceivedto engage in) need-thwarting interpersonal behaviorsathletes will experience need frustration (Sheldon ampFilak 2008) As a result depending on whether indivi-duals within the sport context act in ways that supportor thwart athletesrsquo psychological needs they can act topromote or undermine the quality of athletesrsquomotivation

Need-supportive and need-thwarting interpersonalbehaviors

SDT postulates there are six different types of inter-personal behaviors autonomy-supportive (AS) compe-tence-supportive (CS) relatedness-supportive (RS)autonomy-thwarting (AT also called controlling) com-petence-thwarting (CT) and relatedness-thwarting(RT eg Williams Whip Jackson amp Dimmok 2013)Looking at the need-supportive behaviors first ASbehaviors include providing athletes with rationale achoice and acknowledging their perspectives (Mageauet al 2015) CS behaviors include acknowledging ath-letesrsquo improvements believing they are capable ofachieving their goals and success and providing

athletes with positive feedback (Sheldon amp Filak2008) Lastly RS behaviors include being warm withathletes having an interest in their activities showing agenuine liking for them and providing them with sup-port and care (Jones Armour amp Potrac 2004) As forthe need-thwarting behaviors AT behaviors includeusing intimidating language with athletes makingdemands and incorporating rewards (BartholomewNtoumanis amp Thorgenson-Ntoumani 2009) CT beha-viors consist of discouraging athletes from trying diffi-cult tasks sending them the message they areincompetent doubting their capacity to improve withintheir sport and emphasizing their faults (Sheldon ampFilak 2008) Finally RT behaviors include being distantwith athletes not listening to them not being availableand excluding them from activities or opportunities(Sheldon amp Filak 2008)

Limitations of the existing research

Although significant research has been conducted toexplore the role of interpersonal behaviors in under-standing sport outcomes for athletes there are somelimitations First SDT stipulates supporting all threepsychological needs beyond just autonomy shouldlead to an increase in need satisfaction in athletesand a subsequent increase in athletesrsquo autonomousmotivation for sport and other outcomes (egPomerantz Cheung amp Qin 2012) To date how-ever most research has focused on autonomy exclu-sively (eg Moreau amp Mageau 2013) and only onestudy has examined the relationship between ath-letesrsquo perceptions of their coachesrsquo AS CS and RSinterpersonal behaviors concurrently (Amorose ampAnderson-Butcher 2007) Related to this first lim-itation it is also essential to measure both suppor-tive and thwarting behaviors concurrently since theabsence of supportive behaviors cannot automati-cally imply the presence of thwarting behavior(Sheldon 2011) When considering the role ofneed-thwarting behaviors in sport recent researchhas begun exploring the influence of athletesrsquo per-ceptions of AT behaviors (Bartholomew NtoumanisRyan Bosch amp Thorgenson-Ntoumani 2011) how-ever similar to the need-supportive behaviors therole of autonomy has received most of the empiricalattention so far Overall no studies have exploredhow a coachrsquos use of all six types of interpersonalbehaviors influence athletesrsquo need satisfaction andfrustration in sport

Next looking at coaches research in coaching shouldfocus on examining the factors that predict coachesrsquo use ofall three types of need-supportive and need-thwarting

2 M ROCCHI ET AL

behaviors (ie all six interpersonal behaviors) No studiesto date have examined the factors that predict coachesrsquoASbehaviors as well as CS and RS behaviors (eg PelletierSeacuteguin-Leacutevesque amp Legault 2002 Taylor Ntoumanis ampStandage 2008) Additionally since coaches may simul-taneously engage in need-thwarting behaviors and theabsence of need-support does not necessarily mean thecoach is using need-thwarting styles (Sheldon amp Filak2008) Coaching research should also examine the factorsthat predict all three types of need-thwarting behavior(Bartholomew Ntoumanis amp Thorgenson-Ntoumani2010) Again similar to need-supportive behaviors ATbehaviors have received the most attention (StebbingsTaylor Spray amp Ntoumanis 2012) and research alsoneeds to examine the antecedents of CT and RT inter-personal behavior in coaches

One reason for these limitations in the existingresearch is that there is currently no validated measureavailable for examining perceptions or self-reports of allsix types of interpersonal behaviors according to SDT inthe sport context There are existing measures that look atathletesrsquo perceptions of some dimensions like AS (ieConroy amp Coatsworth 2007) or AS AT and RS (Smithet al 2015) or coachesrsquo self-reports of some dimensionslike AS AT and RS (Smith et al 2016) There arehowever no measures assessing CS CT or RT behaviorsand also no measures focusing on athletesrsquo perceptions ofall six need-supportive and need-thwarting interpersonalbehaviors or coachesrsquo self-reports of their use of all sixtypes of behaviors while coaching Outside of sport thereis one measure the IBQ (Rocchi et al 2017) that hasbeen validated as both a measure of perceptions of othersrsquointerpersonal behaviors (IBQ) and as a self-report of anindividualrsquos own behaviors (IBQ-Self) for all six types ofinterpersonal behaviors according to SDT In order toaddress the limitations related to the lack of research insport motivation there is a need to develop and validatean instrument designed to assess both perceptions andself-reports of all six types of need-supportive and need-thwarting interpersonal behaviors in the sport context

Present research

The purpose of the present studies is to validate theIBQ and IBQ-Self (Rocchi et al 2017) in a sport set-ting The original scale consisted of 24 items measuringsix subscales that represent AS AT CS CT RS and RTinterpersonal behaviors The structure of the scale wasdetermined through three studies using undergraduatestudents The purpose was to create a scale measuringperceptions of othersrsquo interpersonal behaviors as wellas a self-report of onersquos own behavior that could beused across multiple contexts As such the items were

developed by a pool of experts with advanced knowl-edge of SDT using a general stem (ldquoThe people in myliferdquomdashIBQ ldquoWhen I am with the people that are impor-tant to me IrdquomdashIBQ-Self) to ensure they were notbound to a specific interpersonal relationship or con-text In Study 1 (N = 572) the structure of the IBQ wasdetermined through a series of confirmatory factoranalyses and the results supported that the scale hada strong factor structure good internal consistencystrong convergent and divergent validity and the sub-scales correlated with other outcomesmdashas would beexpected according to SDT The results also supportedthat perceptions of need-supportive interpersonal beha-viors were related to increases in reported generalautonomy competence and relatedness satisfactionwhile perceptions of need-thwarting interpersonalbehaviors were related to decreases in the general satis-faction of all three needs These results were replicatedin Study 2 (N = 372) with a new samplemdashprovidingadditional support for the validity of the IBQ In Study3 (N = 736) the factor structure of the IBQ-Self wastested and demonstrated that it also has a strongstructure good internal consistency acceptable validityand reliability and the subscales correlated with out-comes as would be expected For the IBQ-Self it wasfound that when individuals reported they engaged inneed-supportive behaviors this was also associatedwith increased general need satisfaction and decreasedneed frustration while individuals who reportedincreased need-thwarting behaviors saw the opposite

Since the items were developed with the intention tomake the scale applicable to multiple social contextsand social relationships (ie education workplacesport etc) the objective of the present studies is tovalidate the scale as a measure of perceptions of inter-personal behaviors (IBQ) as well as a self-report ofinterpersonal behaviors (IBQ-Self) in sport In Study1 the IBQ will be validated with a sample of athletesreporting on their perceptions of their coachesrsquo inter-personal behaviors In Study 2 the IBQ-Self will bevalidated with a sample of coaches reporting on theirown interpersonal behaviors in their interactions withtheir athletes For both studies the factor structurevalidity reliability and correlations with outcomeswill be examined Overall it is anticipated these studieswill extend the validity of both the IBQ and theIBQ-Self and will demonstrate a strong fit in sport

Study 1

The objective of this study was to extend the validity ofthe IBQ by confirming the scale structure held in asport context Specifically this study will confirm the

MEASUREMENT IN PHYSICAL EDUCATION AND EXERCISE SCIENCE 3

IBQ is an appropriate measure for assessing athletesrsquoperceptions of their coachesrsquo interpersonal behaviors Itis hypothesized the structure of the scale will remainconsistent and the subscales will relate to the outcomesin the same ways as the original validation studies(Rocchi et al 2017) for athletes when they report ontheir coachesrsquo behaviors

Method

Participants

The sample was composed of 239 full-time undergrad-uate student-athletes (Nmale = 130 Nfemale = 109) withan average age of 2015 years (SD = 316) who wereenrolled in first-year courses These students had beencompeting in their sport for an average of 437 years(SD = 345) had been working with their current coachfor an average of 259 years (SD = 308) and trained anaverage of 867 hr per week (SD = 328) with them Themajority of the athletes played hockey (N = 48 16)soccer (N = 39 13) basketball (N = 15 5) volleyball(N = 16 5) or cross country running (N = 12 4)The remaining athletes (N = 109 47) came from avariety of sport backgrounds including badmintonbaseball boxing equestrianism figure skating footballgolf road cycling rowing rugby swimming tennisand weightlifting

Procedures

The athletes were selected from a research participationpool and received credit towards their final course gradefor their participation and participated in an onlinesurvey about their sport experience Participation wasvoluntary and participants gave their informed consentbefore beginning the study To be eligible for this studythe participants had to be actively training in their sportat the time of data collection and have been workingwith their current coach for at least 1 year

Materials

The following measures were completed through anonline questionnaire

Coach interpersonal behaviorsParticipants completed the 24-item IBQ scale measur-ing perceived interpersonal behaviors of their coachusing the stem ldquoMy coach rdquo (Rocchi et al 2017)The athletes indicated their agreement with each state-ment using a 7-point scale ranging from 1 (do not agreeat all) to 7 (completely agree) The scale consists of six

subscales assessing coachesrsquo use of AS AT CS CT RSand RT interpersonal behaviors in their interactionsThe original validation studies supported that the scalehad sufficient internal reliability (α gt 74)

Need satisfactionParticipants responded to the positive items from theBasic Need Satisfaction at Work Scale (Deci et al 2001)to assess the extent their three basic psychologicalneeds were met while practicing their sport The scaleconsists of three subscales measuring participantsrsquoautonomy (three items) competence (three items)and relatedness satisfaction (four items) A confirma-tory factor analysis (CFA) using Mplus Version 6(Muthen amp Muthen 2010 Los Angeles CA this soft-ware was used for all analyses) was conducted on thescale to confirm the structure The fit indices suggestthe model has an acceptable fit (Satorra-Bentler [SB]scaled χ2(41) = 6450 p lt 001 Root Mean Square Errorof Approximation [RMSEA] = 06 [90 ConfidenceInterval (CI) 04 08] Standardized Root MeanSquare Residual [SRMR] = 05 Comparative Fit Index[CFI] = 94 Tucker-Lewis Index [TLI] = 92) Theinternal consistency estimates for the three subscaleswere within acceptable range (autonomy α = 72 com-petence α = 70 relatedness α = 83) and a mean scorewas calculated for each subscale

Need frustrationTo assess psychological need frustration in sport partici-pants also completed the Psychological Need ThwartingScale in Sport (Bartholomew et al 2010) The scale alsoconsists of three subscales measuring participantsrsquo auton-omy competence and relatedness frustration with fouritems each A confirmatory factor analysis was conductedto confirm the structure and the results supported themodel had an excellent fit (SBχ2(32) = 2268 p lt 001RMSEA = 00 [90 CI 00 05] SRMR = 03CFI = 100 TLI = 100) The internal consistency for thethree subscales were within acceptable range (α gt 76)and a mean score was calculated for each subscale torepresent participantsrsquo frustration of each need

Athlete motivationParticipants also completed the Revised SportMotivation Scale (Pelletier Rocchi Vallerand Deci ampRyan 2013) to assess their reasons for participating insport The 6-factor scale is comprised of 18 items (threeper subscale) measuring sport motivation according toeach of the six types of behavioral regulation accordingto SDT A confirmatory factor analysis supported thatthe model had a good fit (SBχ2(120) = 19492 p lt 001RMSEA = 06 [90 CI 04 07] SRMR = 03

4 M ROCCHI ET AL

CFI = 95 TLI = 93) The Cronbachrsquos alpha was calcu-lated for each subscale and revealed they were abovethe acceptable limit (α gt 77) except for the introjectedsubscale which had a lower internal consistency(α = 60) A mean score was calculated for autonomous(mean of intrinsic integrated and identified regula-tion) and controlled motivation (mean of introjectedexternal and amotivated regulation)

Analyses

First the data was cleaned and screened for univariateand multivariate outliers Next a confirmatory factoranalysis was performed on the IBQ to confirm thestructure of the scale held for a sample of athletesreporting on their coachesrsquo behaviors The model wasestimated using maximum likelihood robust (MLR)estimator (Muthen amp Muthen 2010) The fit of themodel was assessed according to Hu and Bentlerrsquos(1999) recommendation of using a scaled chi-square(χ2) and the SRMR as absolute fit indices the TLI asa relative fit index and finally the RMSEA and the CFIas noncentrality-based indices For the SRMR andRMSEA values below 08 indicated adequate modelfit and values below 06 indicated excellent fit for theCFI and TLI values above 90 represented good fit andvalues above 95 indicated excellent fit (HooperCoughlan amp Mullen 2008) Next invariance testswere conducted to confirm the scale performed equallyfor both male and female athletes Then internal con-sistency was examined using Cronbachrsquos alpha for eachsubscale and the discriminant and convergent validitywere examined using the average variance extracted(AVE) and the average shared square variance (ASV)Finally a series of outcome correlation analyses wereconducted to confirm the subscales related to the out-comes as expected according to SDT

Results

First the scoring distributions of the 24 IBQ items wereexamined for univariate normality and results sug-gested that similar to the original validation studies(Rocchi et al 2017) it had not been achieved for allvariables (skewness range minus1015 to 954 kurtosisrange minus655 to 541) Next missing data patternswere examined and it was revealed 24 participants(11) were missing between 1 and 3 observations ontheir reports of the IBQ Since this represented less than1 of the overall sample the missing data was esti-mated using full information ML (FIML) Then theunivariate and multivariate outlier analyses revealedtwo multivariate outliers and they were removed from

the subsequent analyses Finally the composite scoreswere calculated for the coachesrsquo interpersonal behaviors(see Table 1 for descriptive statistics) as well as forathletesrsquo need satisfaction need frustration and sportmotivation (see Table 2 for descriptive statistics)

Scale structure

The structure of the IBQ was tested through a CFA usingthe MLR estimator The results supported the factorialmodel had excellent fit (SBχ2(237) = 29623 p lt 001CFI = 95 TLI = 95 RMSEA = 05 [90 CI 04 06]SRMR= 05) and the standardized factor loadings for eachsubscale were larger than 60 (see Table 3) Next sinceboth groups achieved the minimum recommended sam-ple size of 100 a series of invariance tests were conductedto determine the scale performed equally for both men(N = 105) and women (N = 131) athletes (MacCallumWidaman Zhang amp Hong 1999 Meade 2005) In thefirst step the baseline models were tested for each genderto confirm the structure of the IBQ held for men (SBχ2

(237) = 31776 p lt 001 CFI = 90 TLI = 90 RMSEA = 07

Table 1 Study 1 and 2 Final items with standardized factorloadings (FL) and standard errors (SE)

Study 1 Study 2

Items FL SE FL SE

STEM ldquoMy coach rdquo (Study 1) ldquoWhen I am with my athletes rdquo (Study 2)ASGives me the freedom to make my ownchoices

791 043 712 041

Supports my decisions 872 032 722 036Supports the choices that I make for myself 828 045 681 031Encourages me to make my own decisions 813 045 731 035

ATPressures me to do things their way 699 051 722 030Imposes their opinions on me 794 045 741 029Pressures me to adopt certain behaviors 756 048 693 039Limits my choices 756 051 687 103

CSEncourages me to improve my skills 693 066 658 045Provides valuable feedback 739 052 720 041Acknowledges my ability to achieve my goals 811 043 765 032Tells me that I can accomplish things 859 035 748 033

CTPoints out that I will likely fail 739 054 761 024Sends me the message that I amincompetent

830 039 652 044

Doubts my capacity to improve 741 051 674 030Questions my ability to overcome challenges 819 049 748 037

RSIs interested in what I do 804 054 715 050Takes the time to get to know me 830 039 755 032Honestly enjoy spending time with me 741 051 734 045Relates to me 819 049 633 033

RTDoes not comfort me when I am feeling low 804 054 443 028Is distant when we spend time together 862 028 529 031Does not connect with me 832 031 656 046Does not care about me 855 027 618 055

Note Study 1 N = 237 Study 2 N = 240 The verb tenses and pronouns of theitems weremodified in the IBQ-Self to reflect the new stem For example ldquoGivethem the freedom to make their own choicesrdquo instead of ldquoGives me thefreedom to make my own choicesrdquo or ldquoDo not comfort them when they arefeeling lowrdquo instead of ldquoDoes not comfort me when i am feeling lowrdquo

MEASUREMENT IN PHYSICAL EDUCATION AND EXERCISE SCIENCE 5

[90 CI 06 08] SRMR = 06) and women (SBχ2

(237) = 33495 p lt 001 CFI = 92 TLI = 91RMSEA = 07 [90 CI 06 08] SRMR = 06) Thenconfigural invariance was established by testing a CFAmodel with both groups where no constraints were placedon the parameters The results suggested the factor struc-ture of the IBQ was the same for men and women athletes(SBχ2(474) = 67240 p lt 001 CFI = 91 TLI = 90RMSEA = 07 [90 CI 06 08] SRMR = 06) Nextmetric invariance was established by constraining thefactor indicators to be equal for groups and the resultsfound the data fit the model well (SBχ2(492) = 68120p lt 001 CFI = 91 TLI = 90 RMSEA = 07 [90CI 06 08] SRMR = 07) and the change in chi-squarebetween the constrained and unconstrained model wasnot significant (SBΔχ2(18) = 1199 p gt 05) Finally scalarinvariance was established by constraining the interceptsto be equal for both groups Again the results supportedthe model had good fit (SBχ2(510) = 70916 p lt 001

CFI = 91 TLI = 90 RMSEA = 07 [90 CI 06 08]SRMR = 07) and the chi-square difference test confirmedthe model fit remained stable (SBΔχ2(18) = 2761 p gt 05)Overall these results support men and women athleteswere invariant with regards to the factor structure andcomposite variables can be created for both groups

Validity and reliability

The AVE and ASV were examined to assess conver-gent and divergent validity (see Table 1) All sub-scales met the requirements for convergent validity asthe AVEs were above 5 (Hair Black Babin ampAnderson 2010) Discriminate validity was alsoachieved since all ASV values were smaller thantheir respective AVEs (Hair et al 2010) The factorcorrelations and internal reliabilities for each subscalesupport each subscale has achieved acceptable inter-nal consistency (see Table 1)

Table 2 Study 1 and 2 Factor correlations variance means standard deviations and internal consistencySubscales 1 2 3 4 5 6 AVE ASV M SD

Study 1 (IBQmdashGeneral Stem)1 AS ndash89 68 28 503 1392 AT minus39 ndash84 57 15 323 1493 CS 70 minus30 ndash87 61 18 554 1284 CT minus34 69 minus47 ndash87 62 18 137 1375 RS 71 minus32 74 minus36 ndash90 64 19 498 1446 RT minus40 60 minus37 68 minus50 ndash87 70 17 241 141

Study 2 (IBQ-SelfmdashAthlete Stem)1 AS ndash71 51 07 601 0832 AT minus36 ndash70 51 05 211 2113 CS 40 minus14 ndash73 53 09 566 0594 CT minus13 42 04 ndash75 52 05 140 0765 RS 44 minus18 65 minus05 ndash72 50 11 600 0656 RT minus17 16 minus27 11 minus27 ndash40 33 03 141 073

Note Study 1 N = 237 Study 2 N = 240 α = diagonals AVE = average variance extracted ASV = average shared square variancep lt 05 p lt 001

Table 3 Study 1 and 2 Descriptive statistics and correlations with outcome variables for the IBQ (Study1) and the IBQ-Self (Study 2)

IBQ Subscales

Need Satisfaction Need Frustration Motivation

Soc DesAut Comp Rel Aut Comp Rel Aut Ctl

Study 1AS 58 48 46 minus28 minus32 minus24 40 minus08 mdashAT minus33 minus20 minus27 63 56 54 minus08 45 mdashCS 43 52 48 minus30 minus40 minus34 39 minus17 mdashCT minus30 minus23 minus33 64 66 63 minus14 51 mdashRS 52 50 49 minus26 minus32 minus25 37 minus06 mdashRT minus44 minus29 minus39 62 58 64 minus25 41 mdashMean 519 570 582 304 286 254 550 326 mdashStandard Deviation 114 104 103 114 154 103 116 112 mdash

Study 2AS 20 23 23 minus02 minus02 minus01 25 02 10AT minus11 05 02 23 30 24 11 36 06CS 31 48 34 minus18 minus11 minus06 28 minus01 10CT 01 06 minus03 07 28 26 06 minus30 08RS 36 53 37 minus16 minus12 minus12 29 05 06RT minus04 minus06 minus02 01 22 35 12 minus19 11Mean 611 504 132 173 160 167 505 229 453Standard Deviation 092 079 088 106 100 078 120 097 257

Note Study 1 N = 237 Study 2 N = 240 Soc Des = social desirability For Study 2 correlations are while controlling for social desirability except for thecorrelations between the subscales of the IBQ and social desirability Need satisfaction and frustration are related to sport or coaching Motivation is forparticipating in sport or coaching

p lt 05 p lt 001

6 M ROCCHI ET AL

Outcome correlations

The IBQ subscales were correlated with need satisfaction insport need frustration in sport and sport motivation aspredicted (see Table 3) Specifically the need-supportivesubscales of the IBQ correlated positively with need satis-faction and negatively with need frustration The need-thwarting subscales of the IBQ correlated negatively withneed satisfaction and positively with need frustrationLooking at athletesrsquomotivation for sport the need-suppor-tive subscales correlated positively with autonomous moti-vation and the need-thwarting subscales correlated withcontrolled motivation For CT it was related to an increasein athletesrsquo autonomous motivation and a decrease in theircontrolled motivation For both perceptions of CT and RTcoach behaviors this was related to decreases in autono-mous motivation for athletes as well as an increase incontrolled motivation

Discussion

Overall the results of Study 1 support the psychometricproperties of the IBQ in a sample of athletes reportingon their coachesrsquo interpersonal behaviors and the scaleperformed equally for male and female athletes This isthe first set of results to support the IBQ can be used toassess perceptions of interpersonal behaviors in thesport context In this sample the structure of the scaleheld as well as the reliabilities and outcome correla-tions supporting the scale performed as would beexpected similar to the original validation studies(Rocchi et al 2017) Although there are relativelyhigh relationships between the need-supportive andneed-thwarting subscales that are in line with theresults of the original validation studies the relation-ships are not high enough to suggest multicollinearity(lt 80) may be an issue (Field 2009) The results of thisstudy support the validity of the IBQ to assess athletesrsquoperceptions of coachesrsquo behaviors Specifically theseresults are in line with what would be expected accord-ing to SDT as athletes who report their coaches useneed-supportive interpersonal behaviors reportedincreased need satisfaction and autonomous motivationin sport (Mageau amp Vallerand 2003) while athleteswho reported their coaches use need-thwarting inter-personal behaviors reported need frustration and con-trolled motivation in sport (Bartholomew et al 2009)

Study 2

The objective of Study 2 is to evaluate the psychometricproperties of the IBQ self-report version within the

sport context Specifically this will be validated usinga sample of sport coaches who are reporting on theirbehaviors when they interact with their athletes It isanticipated the factor structure of the IBQ-Self will holdfor this sample in this context and it will correlate withoutcomes as would be expected according to SDT

Method

Participants

The sample was composed of 240 coaches (Nmale = 170Nfemale = 66 Nmissing = 4) with an average age of4701 years (SD = 1031) The large majority of thecoaches had a college (N = 53 22) or universityeducation (N = 168 70) They were either basketball(N = 111 46) or track and field coaches (N = 12954) and they had an average of 1750 years of coach-ing experience (SD = 1283) The majority (N = 12251) identified as a head coach or as an assistant coach(N = 60 25) and indicated their athletes were very(N = 94 39) or extremely (N = 61 25) competitivecompared to other athletes of their age

Procedures

Coaches were recruited through their respective pro-vincial sporting associations (basketball or athletics) toparticipate in an online survey Each organization sentan email to all coaches inviting them to participate inan online study The coachesrsquo participation was volun-tary and they gave their informed consent before parti-cipating To be eligible the coaches had to be registeredwith a provincial sport organization at the time of datacollection and actively coaching

Materials

The following measures were completed by the coachesthrough an online questionnaire

My interpersonal behaviorsThe coaches completed the IBQ-Self to assess the extentto which they believed they engaged in AS AT CS CTRS and RT interpersonal behaviors with their athletes(Rocchi et al 2017) Coaches were presented each itemusing the stem ldquoWhen I am with my athletes I rdquo andindicated their agreement with each statement using a7-point scale ranging from 1 (do not agree at all) to 7(completely agree) The original validation studies sup-ported the scale had sufficient internal reliabil-ity (α gt 80)

MEASUREMENT IN PHYSICAL EDUCATION AND EXERCISE SCIENCE 7

Need satisfactionThe coaches responded to the positive items from theBasic Need Satisfaction at Work Scale adapted to thesport context (Deci et al 2001) to assess the extenttheir three basic psychological needs were met whilecoaching Like in Study 1 the stems were modified toask the coaches about their need satisfaction whilecoaching instead of at work The fit indices suggestthe model has an acceptable fit except for the TLI(SBχ2(41) = 3836 p lt 001 RMSEA = 06 [90 CI03 08] SRMR = 05 CFI = 90 TLI = 89) howeversince the other fit indices were acceptable the scale wasused The internal consistency estimates were withinacceptable range (autonomy α = 83 competenceα = 71 relatedness α = 74) and a mean score wascalculated for each subscale

Need frustrationThe coaches also completed the Psychological NeedThwarting Scale in Sport (Bartholomew et al 2010) toassess the extent to which their needs were frustratedwhile coaching like in Study 1 A confirmatory factoranalysis confirmed the structure of the scale for coachesand the results supported that the model had an excel-lent fit (SBχ2(32) = 2600 p lt 001 RMSEA = 04 [90CI 00 07] SRMR = 06 CFI = 96 TLI = 94) Theinternal consistency for the three subscales were abovethe minimum (α gt 74) and a mean score was calcu-lated for each subscale to represent the coachesrsquo frus-tration of each need

Coach motivationThe coaches completed the Coach MotivationQuestionnaire (CMQ McLean Mallet amp Newcombe2012) to assess their reasons for participating in sportThe CMQ is a 6-factor scale comprised of 22 itemsmeasuring sport motivation according to each of thesix types of behavioral regulation according to SDT Aconfirmatory factor confirmed the structure of the scalein this sample (SBχ2(194) = 33657 p lt 001RMSEA = 06 [90 CI 04 07] SRMR = 07CFI = 91 TLI = 90) The Cronbachrsquos alphas werecalculated for each subscale and revealed they wereabove the acceptable limit (α gt 72) except for theintrojected subscale which had a lower internal consis-tency (α = 68) Like in Study 1 a mean score wascalculated for autonomous motivation and controlledmotivation

Social desirabilityFinally the coaches also completed the short form ofthe Marlowe-Crowne Social Desirability Scale(Reynolds 1982) to control for whether they were

responding to survey items based on providing favor-able responses The short form of this measure consistsof 10 items where participants are asked to indicatewhether these statements are true or false for themThe sum of all responses considered to be sociallydesirable is calculated to provide an overall measureof social desirability

Analyses

Using the same procedures as the first study the datawas cleaned and then the scale structure was examinedThen the validity and reliability of the subscales wereexamined Finally outcome correlations were con-ducted to examine the relationship between coachneed satisfaction frustration and motivation whilecontrolling for social desirability

Results

Similar to the first study the descriptive statistics analyseson the 24 IBQ-Self items suggested the variables did notall have a normal distribution (skewness range minus1325 to912 kurtosis range minus620 to 821) Data screening ana-lyses did not reveal any multivariate outliers as such theentire sample was retained for the full analyses Themissing data analyses suggested 15 participants weremissing one or two observations on the IBQ-Self how-ever since these missing observations made up less than5 of the total observations the data was imputed usingthe same methods as Study 1 The composite scores forthe coachesrsquo reported use of need-supportive andneed-thwarting behaviors as well as their need satisfac-tion and frustration in coaching their motivation towardscoaching and their likelihood of responding in sociallydesirable ways was calculated (see Tables 1 and 2 for thedescriptive statistics)

Scale structure

The results supported that the factorial model had goodfit (SBχ2(237) = 30304 p lt 001 CFI = 93 TLI = 92RMSEA = 06 [90 CI 03 05] SRMR = 04) Thestandardized factor loadings for each subscale werelarger than 46 (see Table 1) and the factor correlationsas well as the internal reliabilities for each subscale(Cronbachrsquos alpha) are in Table 1 Since there wereonly 66 female coaches in this sample gender invar-iance tests were not performed in this study Nextconvergent and divergent validity was examined foreach of the subscales (see Table 2) and the resultssupported that the scale met the thresholds for bothacross all subscales except for the RT subscale Internal

8 M ROCCHI ET AL

consistency analyses revealed all of the alphas were overthe minimum criteria again except for the RT subscale(α = 40) which was nowhere near the baseline criteriaThe internal consistency for that subscale was calcu-lated using all of the potential combinations of items tosee if one specific item was causing the issue and theresults found there was no combination of the fouritems that promoted a better internal consistency Asa result the reliability for the four items was retained

Outcome correlations

The IBQ-Self subscales were correlated with need satis-faction need frustration autonomous and controlledmotivation while controlling for social desirability(see Table 2) to confirm the relationships were in theexpected directions The results found that the supportsubscales of the IBQ-Self correlated moderately andpositively with need satisfaction and autonomous moti-vation while they had weak or negative relationshipswith need frustration and controlled motivation Theopposite was found with the need-thwarting subscaleswhere there were moderate positive relationships withneed frustration and controlled motivation and weaknegative or no relationships with need satisfaction andautonomous motivation Overall the results were simi-lar to those in Study 1 except the strength of therelationships was weaker

Discussion

The results of this study support the factor structureand validity of the IBQ-Self as a questionnaire forcoaches about their use of interpersonal behaviorsaccording to SDT in their interactions with theirathletes Unfortunately the reliability for the RT sub-scale was weak Since some coaches and athletes maychoose to engage in the relationship when theychoose to participate in the sport (ie select athletefor a team) and others do not (ie everyone trainsduring the same session) it is not totally surprisingthat coaches care about their athletes and this madethe results somewhat ambiguous Perhaps in anothercontext where there is either less control over beingin a relationship with the other people such as afamily member or a supervisor at work or morecontrol like a best friend or partner that subscaleinternal consistency would improve Overall theseresults support that coaches who report using moreneed-supportive behaviors also report higher needsatisfaction and autonomous motivation for theircoaching while coaches who report more need-thwarting behaviors report more need frustration

and controlled motivation for coachingmdashall aftercontrolling for social desirability The results of thisstudy support that generally the IBQ-Self is a validmeasure of coachesrsquo own reports of their interperso-nal behaviors when interacting with their athletes

General discussion

The purpose of these two studies was to extend thevalidity of the IBQ by testing the structure of the scalein a sport setting The results of Study 1 supported thatthe IBQ scale factor structure holds in cases whereathletes report on their perceptions of their coachesrsquointerpersonal behaviors The scale also demonstratedadequate validity and reliability and overall the inter-correlations between the IBQ subscales as well as thecorrelations between the IBQ subscales and the otherrelated factors provided support for the fit of the scalewithin the existing SDT measures and literature Interms of the relationships between the IBQ subscalesthe results supported there were moderately high posi-tive and negative correlations between the subscalesAdditionally the subscales also correlated with needsatisfaction need frustration and motivation in thedirections that would be expected according to SDTIn the case of some relationships the outcomes hadsimilar correlations with all three need-supportivebehaviors and all three need-thwarting behaviors butthis is not unexpected since the satisfaction or frustra-tion of the three basic needs are often related (Deci ampRyan 1985)

The results of Study 2 supported that the IBQ-Selfstructure held in a sample of coachesmdashreporting ontheir behaviors with their athletes Similar to Study 1the scale also demonstrated adequate validity and relia-bility Overall the inter-correlations between theIBQ-Self subscales and the correlations between theIBQ-Self subscales and other factors showed a similarpattern to Study 1mdashsupporting that the scale relates toother factors as would be expected

Limitations

These results support the applicability of the scale toa sport context however there are some limitationsFirst the structure of the scales was tested using astudent sample of athletes as well as basketball andtrack and field coaches The IBQ should be testedwith a more varied sample of athletes to ensure thescale is a valid measure of perceptions of coaches forother groups such as older athletes and recreationalathletes Furthermore the IBQ-Self should be vali-dated with coaches from a larger variety of sports

MEASUREMENT IN PHYSICAL EDUCATION AND EXERCISE SCIENCE 9

In this sample coaches were coaching team (basket-ball) and individual sports (athletics) however coa-ches can work with all sorts of different athletes atvarying levels of competition and the scalersquos validityshould be extended to include these coaches too Asecond limitation is that the IBQ and IBQ-Self onlyfocused on two people within the sport context ath-letes and coaches There are a number of differentsocial agents in the sport context that impact athletes(or coaches) including the athletesrsquo parents the otherathletes sport administrators etc and the validity ofthe scale should also be tested to examine perceptionsof their behaviors or their own reports of their beha-viors Finally since the IBQ and IBQ-Self were vali-dated using coachesrsquo and athletesrsquo self-reports theirreports should be triangulated by a third party (egvideo recording and coding of interpersonal beha-viors) to confirm reports on these measures corre-spond to their real world behaviors

Future directions

The IBQ and the IBQ-Self promote a number ofinitiatives for future research in sport motivationFirst research should continue to examine the psy-chometric properties of the IBQ For example ath-letes or coaches should be examined at multiple timepoints in order to establish the scalersquos testndashretestreliability and coachesrsquo scores should be corroboratedwith athletesrsquo scores in order to provide support forthe construct validity of the scale Next the IBQ canbe used to explore the role of all six types of need-supportive and need-thwarting interpersonal beha-viors in relation to how they impact athletesrsquo psycho-logical need satisfaction and frustration This willhelp extend the existing research in sport to movebeyond the influential role of AS and AT and focuson CS CT RS and RT interpersonal behaviors aswell Next the IBQ-Self can be used to identify andunderstand the antecedents of all six types of inter-personal behaviors according to SDT Specifically thescale should be used to explore the factors that influ-ence coachesrsquo behaviors with their athletes Finallythe IBQ and IBQ-Self should be used to explorehow coachesrsquo behaviors change over the course of agiven season and how this relates to athlete psycho-logical needs and their motivation for sport

Overall the IBQ and IBQ-Self are new instrumentsfor assessing perceptions or self-reports of all six typesof need-supportive and need-thwarting interpersonalbehaviors for coaches and athletes It is hoped theIBQ and the IBQ-Self will promote research directed

at assessing the behavior of others in sport and how itimpacts need satisfaction need frustration and motiva-tion for athletes

References

Amorose J amp Anderson-Butcher D (2007) Autonomy-supportive coaching and self-determined motivation inhigh school and college athletes A test of self-determina-tion theory Psychology of Sport and Exercise 8 654ndash670doi101016jpsychsport200611003

Bartholomew K Ntoumanis N Ryan R Bosch J ampThoslashgersen-Ntoumani C (2011) Self-determination the-ory and diminished functioning The role of interpersonalcontrol and psychological need thwarting Personality andSocial Psychology Bulletin 37 1459ndash1473 doi1011770146167211413125

Bartholomew K Ntoumanis N amp Thoslashgersen-Ntoumani C(2009) A review of controlling motivational strategies from aself-determination theory perspective Implications for sportscoaches International Review of Sport and ExercisePsychology 2 215ndash233 doi10108017509840903235330

Bartholomew K Ntoumanis N amp Thoslashrgenson-NtoumaniC (2010) The controlling interpersonal style in a coachingcontext Development and initial validation of a psycho-metric scale Journal of Sport amp Exercise Psychology 32193ndash216 Retrieved from httpwwwncbinlmnihgovpubmed20479478

Conroy D amp Coatsworth J (2007) Assessing autonomy-sup-portive coaching strategies in youth sport Psychology of Sportand Exercise 8 671ndash684 doi101016jpsychsport200612001

Deci E amp Ryan R (1985) Intrinsic motivation and self-determination in human behaviour In E Deci amp R Ryan(Eds) Handbook of self-determination research RochesterNY University of Rochester Press

Deci E amp Ryan R (2002) Handbook of self-determinationresearch Rochester NY The University of Rochester Press

Deci E Ryan R Gagneacute M Leone D Usunov J ampKornazheva B (2001) Need satisfaction motivation andwell-being in the work organizations of a former easternbloc country A cross-cultural study of self-determinationPersonality and Social Psychology Bulletin 27 930ndash942doi1011770146167201278002

Deci E Schwartz A Sheinman L amp Ryan R (1981) Aninstrument to assess adultsrsquo orientations toward control ver-sus autonomy with children Reflections on intrinsic motiva-tion and perceived competence Journal of EducationalPsychology 73 642ndash650 doi1010370022-0663735642

Field A (2009) Discovering statistics using SPSS Sage Publi-cations

Gould D Dieffenbach K amp Moffett A (2002) Psychologicalcharacteristics and their development in Olympic cham-pions Journal of Applied Sports Psychology 14 172ndash204doi10108010413200290103482

Hagger M amp Chatzisarantis N (2007) Intrinsic motivationand self-determination in exercise and sport Leeds UKHuman Kinetics Europe Ltd

Hair J Black W Babin B amp Anderson R (2010)Multivariate data analysis (7th ed) Upper Saddle RiverNJ Prentice-Hall Inc

10 M ROCCHI ET AL

the three basic psychological needs (autonomy compe-tence and relatedness Deci amp Ryan 2002) In sportthe need for autonomy represents the need for indivi-duals to act in line with their own interests and valueswhile practicing their sport Competence requiresopportunities for athletes to increase the level of chal-lenge in their sport and to develop increased skill mas-tery Finally the need for relatedness refers to needing asupportive network and strong interpersonal connec-tions with other people involved in sport (Vallerand2001) Research has shown that when these psycholo-gical needs are satisfied for athletes it leads to anincrease in sport motivation quality and they experi-ence positive outcomes in sport adversely when theseneeds are frustrated it leads to a decrease in sportmotivation quality and promotes negative outcomes(ie Gould Dieffenbach amp Moffett 2002)

The sport context has an impact on the extent towhich athletesrsquo psychological needs are satisfied or fru-strated (Deci Shwartz Sheinman amp Ryan 1981) Thecontext includes both the structure of the sport (ielevel of competition) as well as the people within it(ie coaches) Focusing on the people only SDT pos-tulates other peoplersquos behavior either positively or nega-tively influences athlete need satisfaction andfrustration Specifically when people in the sport con-text engage in need-supportive interpersonal behaviorsit will promote the satisfaction of the basic psychologi-cal needs for athletes (Deci amp Ryan 1985)Alternatively when people engage in (or are perceivedto engage in) need-thwarting interpersonal behaviorsathletes will experience need frustration (Sheldon ampFilak 2008) As a result depending on whether indivi-duals within the sport context act in ways that supportor thwart athletesrsquo psychological needs they can act topromote or undermine the quality of athletesrsquomotivation

Need-supportive and need-thwarting interpersonalbehaviors

SDT postulates there are six different types of inter-personal behaviors autonomy-supportive (AS) compe-tence-supportive (CS) relatedness-supportive (RS)autonomy-thwarting (AT also called controlling) com-petence-thwarting (CT) and relatedness-thwarting(RT eg Williams Whip Jackson amp Dimmok 2013)Looking at the need-supportive behaviors first ASbehaviors include providing athletes with rationale achoice and acknowledging their perspectives (Mageauet al 2015) CS behaviors include acknowledging ath-letesrsquo improvements believing they are capable ofachieving their goals and success and providing

athletes with positive feedback (Sheldon amp Filak2008) Lastly RS behaviors include being warm withathletes having an interest in their activities showing agenuine liking for them and providing them with sup-port and care (Jones Armour amp Potrac 2004) As forthe need-thwarting behaviors AT behaviors includeusing intimidating language with athletes makingdemands and incorporating rewards (BartholomewNtoumanis amp Thorgenson-Ntoumani 2009) CT beha-viors consist of discouraging athletes from trying diffi-cult tasks sending them the message they areincompetent doubting their capacity to improve withintheir sport and emphasizing their faults (Sheldon ampFilak 2008) Finally RT behaviors include being distantwith athletes not listening to them not being availableand excluding them from activities or opportunities(Sheldon amp Filak 2008)

Limitations of the existing research

Although significant research has been conducted toexplore the role of interpersonal behaviors in under-standing sport outcomes for athletes there are somelimitations First SDT stipulates supporting all threepsychological needs beyond just autonomy shouldlead to an increase in need satisfaction in athletesand a subsequent increase in athletesrsquo autonomousmotivation for sport and other outcomes (egPomerantz Cheung amp Qin 2012) To date how-ever most research has focused on autonomy exclu-sively (eg Moreau amp Mageau 2013) and only onestudy has examined the relationship between ath-letesrsquo perceptions of their coachesrsquo AS CS and RSinterpersonal behaviors concurrently (Amorose ampAnderson-Butcher 2007) Related to this first lim-itation it is also essential to measure both suppor-tive and thwarting behaviors concurrently since theabsence of supportive behaviors cannot automati-cally imply the presence of thwarting behavior(Sheldon 2011) When considering the role ofneed-thwarting behaviors in sport recent researchhas begun exploring the influence of athletesrsquo per-ceptions of AT behaviors (Bartholomew NtoumanisRyan Bosch amp Thorgenson-Ntoumani 2011) how-ever similar to the need-supportive behaviors therole of autonomy has received most of the empiricalattention so far Overall no studies have exploredhow a coachrsquos use of all six types of interpersonalbehaviors influence athletesrsquo need satisfaction andfrustration in sport

Next looking at coaches research in coaching shouldfocus on examining the factors that predict coachesrsquo use ofall three types of need-supportive and need-thwarting

2 M ROCCHI ET AL

behaviors (ie all six interpersonal behaviors) No studiesto date have examined the factors that predict coachesrsquoASbehaviors as well as CS and RS behaviors (eg PelletierSeacuteguin-Leacutevesque amp Legault 2002 Taylor Ntoumanis ampStandage 2008) Additionally since coaches may simul-taneously engage in need-thwarting behaviors and theabsence of need-support does not necessarily mean thecoach is using need-thwarting styles (Sheldon amp Filak2008) Coaching research should also examine the factorsthat predict all three types of need-thwarting behavior(Bartholomew Ntoumanis amp Thorgenson-Ntoumani2010) Again similar to need-supportive behaviors ATbehaviors have received the most attention (StebbingsTaylor Spray amp Ntoumanis 2012) and research alsoneeds to examine the antecedents of CT and RT inter-personal behavior in coaches

One reason for these limitations in the existingresearch is that there is currently no validated measureavailable for examining perceptions or self-reports of allsix types of interpersonal behaviors according to SDT inthe sport context There are existing measures that look atathletesrsquo perceptions of some dimensions like AS (ieConroy amp Coatsworth 2007) or AS AT and RS (Smithet al 2015) or coachesrsquo self-reports of some dimensionslike AS AT and RS (Smith et al 2016) There arehowever no measures assessing CS CT or RT behaviorsand also no measures focusing on athletesrsquo perceptions ofall six need-supportive and need-thwarting interpersonalbehaviors or coachesrsquo self-reports of their use of all sixtypes of behaviors while coaching Outside of sport thereis one measure the IBQ (Rocchi et al 2017) that hasbeen validated as both a measure of perceptions of othersrsquointerpersonal behaviors (IBQ) and as a self-report of anindividualrsquos own behaviors (IBQ-Self) for all six types ofinterpersonal behaviors according to SDT In order toaddress the limitations related to the lack of research insport motivation there is a need to develop and validatean instrument designed to assess both perceptions andself-reports of all six types of need-supportive and need-thwarting interpersonal behaviors in the sport context

Present research

The purpose of the present studies is to validate theIBQ and IBQ-Self (Rocchi et al 2017) in a sport set-ting The original scale consisted of 24 items measuringsix subscales that represent AS AT CS CT RS and RTinterpersonal behaviors The structure of the scale wasdetermined through three studies using undergraduatestudents The purpose was to create a scale measuringperceptions of othersrsquo interpersonal behaviors as wellas a self-report of onersquos own behavior that could beused across multiple contexts As such the items were

developed by a pool of experts with advanced knowl-edge of SDT using a general stem (ldquoThe people in myliferdquomdashIBQ ldquoWhen I am with the people that are impor-tant to me IrdquomdashIBQ-Self) to ensure they were notbound to a specific interpersonal relationship or con-text In Study 1 (N = 572) the structure of the IBQ wasdetermined through a series of confirmatory factoranalyses and the results supported that the scale hada strong factor structure good internal consistencystrong convergent and divergent validity and the sub-scales correlated with other outcomesmdashas would beexpected according to SDT The results also supportedthat perceptions of need-supportive interpersonal beha-viors were related to increases in reported generalautonomy competence and relatedness satisfactionwhile perceptions of need-thwarting interpersonalbehaviors were related to decreases in the general satis-faction of all three needs These results were replicatedin Study 2 (N = 372) with a new samplemdashprovidingadditional support for the validity of the IBQ In Study3 (N = 736) the factor structure of the IBQ-Self wastested and demonstrated that it also has a strongstructure good internal consistency acceptable validityand reliability and the subscales correlated with out-comes as would be expected For the IBQ-Self it wasfound that when individuals reported they engaged inneed-supportive behaviors this was also associatedwith increased general need satisfaction and decreasedneed frustration while individuals who reportedincreased need-thwarting behaviors saw the opposite

Since the items were developed with the intention tomake the scale applicable to multiple social contextsand social relationships (ie education workplacesport etc) the objective of the present studies is tovalidate the scale as a measure of perceptions of inter-personal behaviors (IBQ) as well as a self-report ofinterpersonal behaviors (IBQ-Self) in sport In Study1 the IBQ will be validated with a sample of athletesreporting on their perceptions of their coachesrsquo inter-personal behaviors In Study 2 the IBQ-Self will bevalidated with a sample of coaches reporting on theirown interpersonal behaviors in their interactions withtheir athletes For both studies the factor structurevalidity reliability and correlations with outcomeswill be examined Overall it is anticipated these studieswill extend the validity of both the IBQ and theIBQ-Self and will demonstrate a strong fit in sport

Study 1

The objective of this study was to extend the validity ofthe IBQ by confirming the scale structure held in asport context Specifically this study will confirm the

MEASUREMENT IN PHYSICAL EDUCATION AND EXERCISE SCIENCE 3

IBQ is an appropriate measure for assessing athletesrsquoperceptions of their coachesrsquo interpersonal behaviors Itis hypothesized the structure of the scale will remainconsistent and the subscales will relate to the outcomesin the same ways as the original validation studies(Rocchi et al 2017) for athletes when they report ontheir coachesrsquo behaviors

Method

Participants

The sample was composed of 239 full-time undergrad-uate student-athletes (Nmale = 130 Nfemale = 109) withan average age of 2015 years (SD = 316) who wereenrolled in first-year courses These students had beencompeting in their sport for an average of 437 years(SD = 345) had been working with their current coachfor an average of 259 years (SD = 308) and trained anaverage of 867 hr per week (SD = 328) with them Themajority of the athletes played hockey (N = 48 16)soccer (N = 39 13) basketball (N = 15 5) volleyball(N = 16 5) or cross country running (N = 12 4)The remaining athletes (N = 109 47) came from avariety of sport backgrounds including badmintonbaseball boxing equestrianism figure skating footballgolf road cycling rowing rugby swimming tennisand weightlifting

Procedures

The athletes were selected from a research participationpool and received credit towards their final course gradefor their participation and participated in an onlinesurvey about their sport experience Participation wasvoluntary and participants gave their informed consentbefore beginning the study To be eligible for this studythe participants had to be actively training in their sportat the time of data collection and have been workingwith their current coach for at least 1 year

Materials

The following measures were completed through anonline questionnaire

Coach interpersonal behaviorsParticipants completed the 24-item IBQ scale measur-ing perceived interpersonal behaviors of their coachusing the stem ldquoMy coach rdquo (Rocchi et al 2017)The athletes indicated their agreement with each state-ment using a 7-point scale ranging from 1 (do not agreeat all) to 7 (completely agree) The scale consists of six

subscales assessing coachesrsquo use of AS AT CS CT RSand RT interpersonal behaviors in their interactionsThe original validation studies supported that the scalehad sufficient internal reliability (α gt 74)

Need satisfactionParticipants responded to the positive items from theBasic Need Satisfaction at Work Scale (Deci et al 2001)to assess the extent their three basic psychologicalneeds were met while practicing their sport The scaleconsists of three subscales measuring participantsrsquoautonomy (three items) competence (three items)and relatedness satisfaction (four items) A confirma-tory factor analysis (CFA) using Mplus Version 6(Muthen amp Muthen 2010 Los Angeles CA this soft-ware was used for all analyses) was conducted on thescale to confirm the structure The fit indices suggestthe model has an acceptable fit (Satorra-Bentler [SB]scaled χ2(41) = 6450 p lt 001 Root Mean Square Errorof Approximation [RMSEA] = 06 [90 ConfidenceInterval (CI) 04 08] Standardized Root MeanSquare Residual [SRMR] = 05 Comparative Fit Index[CFI] = 94 Tucker-Lewis Index [TLI] = 92) Theinternal consistency estimates for the three subscaleswere within acceptable range (autonomy α = 72 com-petence α = 70 relatedness α = 83) and a mean scorewas calculated for each subscale

Need frustrationTo assess psychological need frustration in sport partici-pants also completed the Psychological Need ThwartingScale in Sport (Bartholomew et al 2010) The scale alsoconsists of three subscales measuring participantsrsquo auton-omy competence and relatedness frustration with fouritems each A confirmatory factor analysis was conductedto confirm the structure and the results supported themodel had an excellent fit (SBχ2(32) = 2268 p lt 001RMSEA = 00 [90 CI 00 05] SRMR = 03CFI = 100 TLI = 100) The internal consistency for thethree subscales were within acceptable range (α gt 76)and a mean score was calculated for each subscale torepresent participantsrsquo frustration of each need

Athlete motivationParticipants also completed the Revised SportMotivation Scale (Pelletier Rocchi Vallerand Deci ampRyan 2013) to assess their reasons for participating insport The 6-factor scale is comprised of 18 items (threeper subscale) measuring sport motivation according toeach of the six types of behavioral regulation accordingto SDT A confirmatory factor analysis supported thatthe model had a good fit (SBχ2(120) = 19492 p lt 001RMSEA = 06 [90 CI 04 07] SRMR = 03

4 M ROCCHI ET AL

CFI = 95 TLI = 93) The Cronbachrsquos alpha was calcu-lated for each subscale and revealed they were abovethe acceptable limit (α gt 77) except for the introjectedsubscale which had a lower internal consistency(α = 60) A mean score was calculated for autonomous(mean of intrinsic integrated and identified regula-tion) and controlled motivation (mean of introjectedexternal and amotivated regulation)

Analyses

First the data was cleaned and screened for univariateand multivariate outliers Next a confirmatory factoranalysis was performed on the IBQ to confirm thestructure of the scale held for a sample of athletesreporting on their coachesrsquo behaviors The model wasestimated using maximum likelihood robust (MLR)estimator (Muthen amp Muthen 2010) The fit of themodel was assessed according to Hu and Bentlerrsquos(1999) recommendation of using a scaled chi-square(χ2) and the SRMR as absolute fit indices the TLI asa relative fit index and finally the RMSEA and the CFIas noncentrality-based indices For the SRMR andRMSEA values below 08 indicated adequate modelfit and values below 06 indicated excellent fit for theCFI and TLI values above 90 represented good fit andvalues above 95 indicated excellent fit (HooperCoughlan amp Mullen 2008) Next invariance testswere conducted to confirm the scale performed equallyfor both male and female athletes Then internal con-sistency was examined using Cronbachrsquos alpha for eachsubscale and the discriminant and convergent validitywere examined using the average variance extracted(AVE) and the average shared square variance (ASV)Finally a series of outcome correlation analyses wereconducted to confirm the subscales related to the out-comes as expected according to SDT

Results

First the scoring distributions of the 24 IBQ items wereexamined for univariate normality and results sug-gested that similar to the original validation studies(Rocchi et al 2017) it had not been achieved for allvariables (skewness range minus1015 to 954 kurtosisrange minus655 to 541) Next missing data patternswere examined and it was revealed 24 participants(11) were missing between 1 and 3 observations ontheir reports of the IBQ Since this represented less than1 of the overall sample the missing data was esti-mated using full information ML (FIML) Then theunivariate and multivariate outlier analyses revealedtwo multivariate outliers and they were removed from

the subsequent analyses Finally the composite scoreswere calculated for the coachesrsquo interpersonal behaviors(see Table 1 for descriptive statistics) as well as forathletesrsquo need satisfaction need frustration and sportmotivation (see Table 2 for descriptive statistics)

Scale structure

The structure of the IBQ was tested through a CFA usingthe MLR estimator The results supported the factorialmodel had excellent fit (SBχ2(237) = 29623 p lt 001CFI = 95 TLI = 95 RMSEA = 05 [90 CI 04 06]SRMR= 05) and the standardized factor loadings for eachsubscale were larger than 60 (see Table 3) Next sinceboth groups achieved the minimum recommended sam-ple size of 100 a series of invariance tests were conductedto determine the scale performed equally for both men(N = 105) and women (N = 131) athletes (MacCallumWidaman Zhang amp Hong 1999 Meade 2005) In thefirst step the baseline models were tested for each genderto confirm the structure of the IBQ held for men (SBχ2

(237) = 31776 p lt 001 CFI = 90 TLI = 90 RMSEA = 07

Table 1 Study 1 and 2 Final items with standardized factorloadings (FL) and standard errors (SE)

Study 1 Study 2

Items FL SE FL SE

STEM ldquoMy coach rdquo (Study 1) ldquoWhen I am with my athletes rdquo (Study 2)ASGives me the freedom to make my ownchoices

791 043 712 041

Supports my decisions 872 032 722 036Supports the choices that I make for myself 828 045 681 031Encourages me to make my own decisions 813 045 731 035

ATPressures me to do things their way 699 051 722 030Imposes their opinions on me 794 045 741 029Pressures me to adopt certain behaviors 756 048 693 039Limits my choices 756 051 687 103

CSEncourages me to improve my skills 693 066 658 045Provides valuable feedback 739 052 720 041Acknowledges my ability to achieve my goals 811 043 765 032Tells me that I can accomplish things 859 035 748 033

CTPoints out that I will likely fail 739 054 761 024Sends me the message that I amincompetent

830 039 652 044

Doubts my capacity to improve 741 051 674 030Questions my ability to overcome challenges 819 049 748 037

RSIs interested in what I do 804 054 715 050Takes the time to get to know me 830 039 755 032Honestly enjoy spending time with me 741 051 734 045Relates to me 819 049 633 033

RTDoes not comfort me when I am feeling low 804 054 443 028Is distant when we spend time together 862 028 529 031Does not connect with me 832 031 656 046Does not care about me 855 027 618 055

Note Study 1 N = 237 Study 2 N = 240 The verb tenses and pronouns of theitems weremodified in the IBQ-Self to reflect the new stem For example ldquoGivethem the freedom to make their own choicesrdquo instead of ldquoGives me thefreedom to make my own choicesrdquo or ldquoDo not comfort them when they arefeeling lowrdquo instead of ldquoDoes not comfort me when i am feeling lowrdquo

MEASUREMENT IN PHYSICAL EDUCATION AND EXERCISE SCIENCE 5

[90 CI 06 08] SRMR = 06) and women (SBχ2

(237) = 33495 p lt 001 CFI = 92 TLI = 91RMSEA = 07 [90 CI 06 08] SRMR = 06) Thenconfigural invariance was established by testing a CFAmodel with both groups where no constraints were placedon the parameters The results suggested the factor struc-ture of the IBQ was the same for men and women athletes(SBχ2(474) = 67240 p lt 001 CFI = 91 TLI = 90RMSEA = 07 [90 CI 06 08] SRMR = 06) Nextmetric invariance was established by constraining thefactor indicators to be equal for groups and the resultsfound the data fit the model well (SBχ2(492) = 68120p lt 001 CFI = 91 TLI = 90 RMSEA = 07 [90CI 06 08] SRMR = 07) and the change in chi-squarebetween the constrained and unconstrained model wasnot significant (SBΔχ2(18) = 1199 p gt 05) Finally scalarinvariance was established by constraining the interceptsto be equal for both groups Again the results supportedthe model had good fit (SBχ2(510) = 70916 p lt 001

CFI = 91 TLI = 90 RMSEA = 07 [90 CI 06 08]SRMR = 07) and the chi-square difference test confirmedthe model fit remained stable (SBΔχ2(18) = 2761 p gt 05)Overall these results support men and women athleteswere invariant with regards to the factor structure andcomposite variables can be created for both groups

Validity and reliability

The AVE and ASV were examined to assess conver-gent and divergent validity (see Table 1) All sub-scales met the requirements for convergent validity asthe AVEs were above 5 (Hair Black Babin ampAnderson 2010) Discriminate validity was alsoachieved since all ASV values were smaller thantheir respective AVEs (Hair et al 2010) The factorcorrelations and internal reliabilities for each subscalesupport each subscale has achieved acceptable inter-nal consistency (see Table 1)

Table 2 Study 1 and 2 Factor correlations variance means standard deviations and internal consistencySubscales 1 2 3 4 5 6 AVE ASV M SD

Study 1 (IBQmdashGeneral Stem)1 AS ndash89 68 28 503 1392 AT minus39 ndash84 57 15 323 1493 CS 70 minus30 ndash87 61 18 554 1284 CT minus34 69 minus47 ndash87 62 18 137 1375 RS 71 minus32 74 minus36 ndash90 64 19 498 1446 RT minus40 60 minus37 68 minus50 ndash87 70 17 241 141

Study 2 (IBQ-SelfmdashAthlete Stem)1 AS ndash71 51 07 601 0832 AT minus36 ndash70 51 05 211 2113 CS 40 minus14 ndash73 53 09 566 0594 CT minus13 42 04 ndash75 52 05 140 0765 RS 44 minus18 65 minus05 ndash72 50 11 600 0656 RT minus17 16 minus27 11 minus27 ndash40 33 03 141 073

Note Study 1 N = 237 Study 2 N = 240 α = diagonals AVE = average variance extracted ASV = average shared square variancep lt 05 p lt 001

Table 3 Study 1 and 2 Descriptive statistics and correlations with outcome variables for the IBQ (Study1) and the IBQ-Self (Study 2)

IBQ Subscales

Need Satisfaction Need Frustration Motivation

Soc DesAut Comp Rel Aut Comp Rel Aut Ctl

Study 1AS 58 48 46 minus28 minus32 minus24 40 minus08 mdashAT minus33 minus20 minus27 63 56 54 minus08 45 mdashCS 43 52 48 minus30 minus40 minus34 39 minus17 mdashCT minus30 minus23 minus33 64 66 63 minus14 51 mdashRS 52 50 49 minus26 minus32 minus25 37 minus06 mdashRT minus44 minus29 minus39 62 58 64 minus25 41 mdashMean 519 570 582 304 286 254 550 326 mdashStandard Deviation 114 104 103 114 154 103 116 112 mdash

Study 2AS 20 23 23 minus02 minus02 minus01 25 02 10AT minus11 05 02 23 30 24 11 36 06CS 31 48 34 minus18 minus11 minus06 28 minus01 10CT 01 06 minus03 07 28 26 06 minus30 08RS 36 53 37 minus16 minus12 minus12 29 05 06RT minus04 minus06 minus02 01 22 35 12 minus19 11Mean 611 504 132 173 160 167 505 229 453Standard Deviation 092 079 088 106 100 078 120 097 257

Note Study 1 N = 237 Study 2 N = 240 Soc Des = social desirability For Study 2 correlations are while controlling for social desirability except for thecorrelations between the subscales of the IBQ and social desirability Need satisfaction and frustration are related to sport or coaching Motivation is forparticipating in sport or coaching

p lt 05 p lt 001

6 M ROCCHI ET AL

Outcome correlations

The IBQ subscales were correlated with need satisfaction insport need frustration in sport and sport motivation aspredicted (see Table 3) Specifically the need-supportivesubscales of the IBQ correlated positively with need satis-faction and negatively with need frustration The need-thwarting subscales of the IBQ correlated negatively withneed satisfaction and positively with need frustrationLooking at athletesrsquomotivation for sport the need-suppor-tive subscales correlated positively with autonomous moti-vation and the need-thwarting subscales correlated withcontrolled motivation For CT it was related to an increasein athletesrsquo autonomous motivation and a decrease in theircontrolled motivation For both perceptions of CT and RTcoach behaviors this was related to decreases in autono-mous motivation for athletes as well as an increase incontrolled motivation

Discussion

Overall the results of Study 1 support the psychometricproperties of the IBQ in a sample of athletes reportingon their coachesrsquo interpersonal behaviors and the scaleperformed equally for male and female athletes This isthe first set of results to support the IBQ can be used toassess perceptions of interpersonal behaviors in thesport context In this sample the structure of the scaleheld as well as the reliabilities and outcome correla-tions supporting the scale performed as would beexpected similar to the original validation studies(Rocchi et al 2017) Although there are relativelyhigh relationships between the need-supportive andneed-thwarting subscales that are in line with theresults of the original validation studies the relation-ships are not high enough to suggest multicollinearity(lt 80) may be an issue (Field 2009) The results of thisstudy support the validity of the IBQ to assess athletesrsquoperceptions of coachesrsquo behaviors Specifically theseresults are in line with what would be expected accord-ing to SDT as athletes who report their coaches useneed-supportive interpersonal behaviors reportedincreased need satisfaction and autonomous motivationin sport (Mageau amp Vallerand 2003) while athleteswho reported their coaches use need-thwarting inter-personal behaviors reported need frustration and con-trolled motivation in sport (Bartholomew et al 2009)

Study 2

The objective of Study 2 is to evaluate the psychometricproperties of the IBQ self-report version within the

sport context Specifically this will be validated usinga sample of sport coaches who are reporting on theirbehaviors when they interact with their athletes It isanticipated the factor structure of the IBQ-Self will holdfor this sample in this context and it will correlate withoutcomes as would be expected according to SDT

Method

Participants

The sample was composed of 240 coaches (Nmale = 170Nfemale = 66 Nmissing = 4) with an average age of4701 years (SD = 1031) The large majority of thecoaches had a college (N = 53 22) or universityeducation (N = 168 70) They were either basketball(N = 111 46) or track and field coaches (N = 12954) and they had an average of 1750 years of coach-ing experience (SD = 1283) The majority (N = 12251) identified as a head coach or as an assistant coach(N = 60 25) and indicated their athletes were very(N = 94 39) or extremely (N = 61 25) competitivecompared to other athletes of their age

Procedures

Coaches were recruited through their respective pro-vincial sporting associations (basketball or athletics) toparticipate in an online survey Each organization sentan email to all coaches inviting them to participate inan online study The coachesrsquo participation was volun-tary and they gave their informed consent before parti-cipating To be eligible the coaches had to be registeredwith a provincial sport organization at the time of datacollection and actively coaching

Materials

The following measures were completed by the coachesthrough an online questionnaire

My interpersonal behaviorsThe coaches completed the IBQ-Self to assess the extentto which they believed they engaged in AS AT CS CTRS and RT interpersonal behaviors with their athletes(Rocchi et al 2017) Coaches were presented each itemusing the stem ldquoWhen I am with my athletes I rdquo andindicated their agreement with each statement using a7-point scale ranging from 1 (do not agree at all) to 7(completely agree) The original validation studies sup-ported the scale had sufficient internal reliabil-ity (α gt 80)

MEASUREMENT IN PHYSICAL EDUCATION AND EXERCISE SCIENCE 7

Need satisfactionThe coaches responded to the positive items from theBasic Need Satisfaction at Work Scale adapted to thesport context (Deci et al 2001) to assess the extenttheir three basic psychological needs were met whilecoaching Like in Study 1 the stems were modified toask the coaches about their need satisfaction whilecoaching instead of at work The fit indices suggestthe model has an acceptable fit except for the TLI(SBχ2(41) = 3836 p lt 001 RMSEA = 06 [90 CI03 08] SRMR = 05 CFI = 90 TLI = 89) howeversince the other fit indices were acceptable the scale wasused The internal consistency estimates were withinacceptable range (autonomy α = 83 competenceα = 71 relatedness α = 74) and a mean score wascalculated for each subscale

Need frustrationThe coaches also completed the Psychological NeedThwarting Scale in Sport (Bartholomew et al 2010) toassess the extent to which their needs were frustratedwhile coaching like in Study 1 A confirmatory factoranalysis confirmed the structure of the scale for coachesand the results supported that the model had an excel-lent fit (SBχ2(32) = 2600 p lt 001 RMSEA = 04 [90CI 00 07] SRMR = 06 CFI = 96 TLI = 94) Theinternal consistency for the three subscales were abovethe minimum (α gt 74) and a mean score was calcu-lated for each subscale to represent the coachesrsquo frus-tration of each need

Coach motivationThe coaches completed the Coach MotivationQuestionnaire (CMQ McLean Mallet amp Newcombe2012) to assess their reasons for participating in sportThe CMQ is a 6-factor scale comprised of 22 itemsmeasuring sport motivation according to each of thesix types of behavioral regulation according to SDT Aconfirmatory factor confirmed the structure of the scalein this sample (SBχ2(194) = 33657 p lt 001RMSEA = 06 [90 CI 04 07] SRMR = 07CFI = 91 TLI = 90) The Cronbachrsquos alphas werecalculated for each subscale and revealed they wereabove the acceptable limit (α gt 72) except for theintrojected subscale which had a lower internal consis-tency (α = 68) Like in Study 1 a mean score wascalculated for autonomous motivation and controlledmotivation

Social desirabilityFinally the coaches also completed the short form ofthe Marlowe-Crowne Social Desirability Scale(Reynolds 1982) to control for whether they were

responding to survey items based on providing favor-able responses The short form of this measure consistsof 10 items where participants are asked to indicatewhether these statements are true or false for themThe sum of all responses considered to be sociallydesirable is calculated to provide an overall measureof social desirability

Analyses

Using the same procedures as the first study the datawas cleaned and then the scale structure was examinedThen the validity and reliability of the subscales wereexamined Finally outcome correlations were con-ducted to examine the relationship between coachneed satisfaction frustration and motivation whilecontrolling for social desirability

Results

Similar to the first study the descriptive statistics analyseson the 24 IBQ-Self items suggested the variables did notall have a normal distribution (skewness range minus1325 to912 kurtosis range minus620 to 821) Data screening ana-lyses did not reveal any multivariate outliers as such theentire sample was retained for the full analyses Themissing data analyses suggested 15 participants weremissing one or two observations on the IBQ-Self how-ever since these missing observations made up less than5 of the total observations the data was imputed usingthe same methods as Study 1 The composite scores forthe coachesrsquo reported use of need-supportive andneed-thwarting behaviors as well as their need satisfac-tion and frustration in coaching their motivation towardscoaching and their likelihood of responding in sociallydesirable ways was calculated (see Tables 1 and 2 for thedescriptive statistics)

Scale structure

The results supported that the factorial model had goodfit (SBχ2(237) = 30304 p lt 001 CFI = 93 TLI = 92RMSEA = 06 [90 CI 03 05] SRMR = 04) Thestandardized factor loadings for each subscale werelarger than 46 (see Table 1) and the factor correlationsas well as the internal reliabilities for each subscale(Cronbachrsquos alpha) are in Table 1 Since there wereonly 66 female coaches in this sample gender invar-iance tests were not performed in this study Nextconvergent and divergent validity was examined foreach of the subscales (see Table 2) and the resultssupported that the scale met the thresholds for bothacross all subscales except for the RT subscale Internal

8 M ROCCHI ET AL

consistency analyses revealed all of the alphas were overthe minimum criteria again except for the RT subscale(α = 40) which was nowhere near the baseline criteriaThe internal consistency for that subscale was calcu-lated using all of the potential combinations of items tosee if one specific item was causing the issue and theresults found there was no combination of the fouritems that promoted a better internal consistency Asa result the reliability for the four items was retained

Outcome correlations

The IBQ-Self subscales were correlated with need satis-faction need frustration autonomous and controlledmotivation while controlling for social desirability(see Table 2) to confirm the relationships were in theexpected directions The results found that the supportsubscales of the IBQ-Self correlated moderately andpositively with need satisfaction and autonomous moti-vation while they had weak or negative relationshipswith need frustration and controlled motivation Theopposite was found with the need-thwarting subscaleswhere there were moderate positive relationships withneed frustration and controlled motivation and weaknegative or no relationships with need satisfaction andautonomous motivation Overall the results were simi-lar to those in Study 1 except the strength of therelationships was weaker

Discussion

The results of this study support the factor structureand validity of the IBQ-Self as a questionnaire forcoaches about their use of interpersonal behaviorsaccording to SDT in their interactions with theirathletes Unfortunately the reliability for the RT sub-scale was weak Since some coaches and athletes maychoose to engage in the relationship when theychoose to participate in the sport (ie select athletefor a team) and others do not (ie everyone trainsduring the same session) it is not totally surprisingthat coaches care about their athletes and this madethe results somewhat ambiguous Perhaps in anothercontext where there is either less control over beingin a relationship with the other people such as afamily member or a supervisor at work or morecontrol like a best friend or partner that subscaleinternal consistency would improve Overall theseresults support that coaches who report using moreneed-supportive behaviors also report higher needsatisfaction and autonomous motivation for theircoaching while coaches who report more need-thwarting behaviors report more need frustration

and controlled motivation for coachingmdashall aftercontrolling for social desirability The results of thisstudy support that generally the IBQ-Self is a validmeasure of coachesrsquo own reports of their interperso-nal behaviors when interacting with their athletes

General discussion

The purpose of these two studies was to extend thevalidity of the IBQ by testing the structure of the scalein a sport setting The results of Study 1 supported thatthe IBQ scale factor structure holds in cases whereathletes report on their perceptions of their coachesrsquointerpersonal behaviors The scale also demonstratedadequate validity and reliability and overall the inter-correlations between the IBQ subscales as well as thecorrelations between the IBQ subscales and the otherrelated factors provided support for the fit of the scalewithin the existing SDT measures and literature Interms of the relationships between the IBQ subscalesthe results supported there were moderately high posi-tive and negative correlations between the subscalesAdditionally the subscales also correlated with needsatisfaction need frustration and motivation in thedirections that would be expected according to SDTIn the case of some relationships the outcomes hadsimilar correlations with all three need-supportivebehaviors and all three need-thwarting behaviors butthis is not unexpected since the satisfaction or frustra-tion of the three basic needs are often related (Deci ampRyan 1985)

The results of Study 2 supported that the IBQ-Selfstructure held in a sample of coachesmdashreporting ontheir behaviors with their athletes Similar to Study 1the scale also demonstrated adequate validity and relia-bility Overall the inter-correlations between theIBQ-Self subscales and the correlations between theIBQ-Self subscales and other factors showed a similarpattern to Study 1mdashsupporting that the scale relates toother factors as would be expected

Limitations

These results support the applicability of the scale toa sport context however there are some limitationsFirst the structure of the scales was tested using astudent sample of athletes as well as basketball andtrack and field coaches The IBQ should be testedwith a more varied sample of athletes to ensure thescale is a valid measure of perceptions of coaches forother groups such as older athletes and recreationalathletes Furthermore the IBQ-Self should be vali-dated with coaches from a larger variety of sports

MEASUREMENT IN PHYSICAL EDUCATION AND EXERCISE SCIENCE 9

In this sample coaches were coaching team (basket-ball) and individual sports (athletics) however coa-ches can work with all sorts of different athletes atvarying levels of competition and the scalersquos validityshould be extended to include these coaches too Asecond limitation is that the IBQ and IBQ-Self onlyfocused on two people within the sport context ath-letes and coaches There are a number of differentsocial agents in the sport context that impact athletes(or coaches) including the athletesrsquo parents the otherathletes sport administrators etc and the validity ofthe scale should also be tested to examine perceptionsof their behaviors or their own reports of their beha-viors Finally since the IBQ and IBQ-Self were vali-dated using coachesrsquo and athletesrsquo self-reports theirreports should be triangulated by a third party (egvideo recording and coding of interpersonal beha-viors) to confirm reports on these measures corre-spond to their real world behaviors

Future directions

The IBQ and the IBQ-Self promote a number ofinitiatives for future research in sport motivationFirst research should continue to examine the psy-chometric properties of the IBQ For example ath-letes or coaches should be examined at multiple timepoints in order to establish the scalersquos testndashretestreliability and coachesrsquo scores should be corroboratedwith athletesrsquo scores in order to provide support forthe construct validity of the scale Next the IBQ canbe used to explore the role of all six types of need-supportive and need-thwarting interpersonal beha-viors in relation to how they impact athletesrsquo psycho-logical need satisfaction and frustration This willhelp extend the existing research in sport to movebeyond the influential role of AS and AT and focuson CS CT RS and RT interpersonal behaviors aswell Next the IBQ-Self can be used to identify andunderstand the antecedents of all six types of inter-personal behaviors according to SDT Specifically thescale should be used to explore the factors that influ-ence coachesrsquo behaviors with their athletes Finallythe IBQ and IBQ-Self should be used to explorehow coachesrsquo behaviors change over the course of agiven season and how this relates to athlete psycho-logical needs and their motivation for sport

Overall the IBQ and IBQ-Self are new instrumentsfor assessing perceptions or self-reports of all six typesof need-supportive and need-thwarting interpersonalbehaviors for coaches and athletes It is hoped theIBQ and the IBQ-Self will promote research directed

at assessing the behavior of others in sport and how itimpacts need satisfaction need frustration and motiva-tion for athletes

References

Amorose J amp Anderson-Butcher D (2007) Autonomy-supportive coaching and self-determined motivation inhigh school and college athletes A test of self-determina-tion theory Psychology of Sport and Exercise 8 654ndash670doi101016jpsychsport200611003

Bartholomew K Ntoumanis N Ryan R Bosch J ampThoslashgersen-Ntoumani C (2011) Self-determination the-ory and diminished functioning The role of interpersonalcontrol and psychological need thwarting Personality andSocial Psychology Bulletin 37 1459ndash1473 doi1011770146167211413125

Bartholomew K Ntoumanis N amp Thoslashgersen-Ntoumani C(2009) A review of controlling motivational strategies from aself-determination theory perspective Implications for sportscoaches International Review of Sport and ExercisePsychology 2 215ndash233 doi10108017509840903235330

Bartholomew K Ntoumanis N amp Thoslashrgenson-NtoumaniC (2010) The controlling interpersonal style in a coachingcontext Development and initial validation of a psycho-metric scale Journal of Sport amp Exercise Psychology 32193ndash216 Retrieved from httpwwwncbinlmnihgovpubmed20479478

Conroy D amp Coatsworth J (2007) Assessing autonomy-sup-portive coaching strategies in youth sport Psychology of Sportand Exercise 8 671ndash684 doi101016jpsychsport200612001

Deci E amp Ryan R (1985) Intrinsic motivation and self-determination in human behaviour In E Deci amp R Ryan(Eds) Handbook of self-determination research RochesterNY University of Rochester Press

Deci E amp Ryan R (2002) Handbook of self-determinationresearch Rochester NY The University of Rochester Press

Deci E Ryan R Gagneacute M Leone D Usunov J ampKornazheva B (2001) Need satisfaction motivation andwell-being in the work organizations of a former easternbloc country A cross-cultural study of self-determinationPersonality and Social Psychology Bulletin 27 930ndash942doi1011770146167201278002

Deci E Schwartz A Sheinman L amp Ryan R (1981) Aninstrument to assess adultsrsquo orientations toward control ver-sus autonomy with children Reflections on intrinsic motiva-tion and perceived competence Journal of EducationalPsychology 73 642ndash650 doi1010370022-0663735642

Field A (2009) Discovering statistics using SPSS Sage Publi-cations

Gould D Dieffenbach K amp Moffett A (2002) Psychologicalcharacteristics and their development in Olympic cham-pions Journal of Applied Sports Psychology 14 172ndash204doi10108010413200290103482

Hagger M amp Chatzisarantis N (2007) Intrinsic motivationand self-determination in exercise and sport Leeds UKHuman Kinetics Europe Ltd

Hair J Black W Babin B amp Anderson R (2010)Multivariate data analysis (7th ed) Upper Saddle RiverNJ Prentice-Hall Inc

10 M ROCCHI ET AL

behaviors (ie all six interpersonal behaviors) No studiesto date have examined the factors that predict coachesrsquoASbehaviors as well as CS and RS behaviors (eg PelletierSeacuteguin-Leacutevesque amp Legault 2002 Taylor Ntoumanis ampStandage 2008) Additionally since coaches may simul-taneously engage in need-thwarting behaviors and theabsence of need-support does not necessarily mean thecoach is using need-thwarting styles (Sheldon amp Filak2008) Coaching research should also examine the factorsthat predict all three types of need-thwarting behavior(Bartholomew Ntoumanis amp Thorgenson-Ntoumani2010) Again similar to need-supportive behaviors ATbehaviors have received the most attention (StebbingsTaylor Spray amp Ntoumanis 2012) and research alsoneeds to examine the antecedents of CT and RT inter-personal behavior in coaches

One reason for these limitations in the existingresearch is that there is currently no validated measureavailable for examining perceptions or self-reports of allsix types of interpersonal behaviors according to SDT inthe sport context There are existing measures that look atathletesrsquo perceptions of some dimensions like AS (ieConroy amp Coatsworth 2007) or AS AT and RS (Smithet al 2015) or coachesrsquo self-reports of some dimensionslike AS AT and RS (Smith et al 2016) There arehowever no measures assessing CS CT or RT behaviorsand also no measures focusing on athletesrsquo perceptions ofall six need-supportive and need-thwarting interpersonalbehaviors or coachesrsquo self-reports of their use of all sixtypes of behaviors while coaching Outside of sport thereis one measure the IBQ (Rocchi et al 2017) that hasbeen validated as both a measure of perceptions of othersrsquointerpersonal behaviors (IBQ) and as a self-report of anindividualrsquos own behaviors (IBQ-Self) for all six types ofinterpersonal behaviors according to SDT In order toaddress the limitations related to the lack of research insport motivation there is a need to develop and validatean instrument designed to assess both perceptions andself-reports of all six types of need-supportive and need-thwarting interpersonal behaviors in the sport context

Present research

The purpose of the present studies is to validate theIBQ and IBQ-Self (Rocchi et al 2017) in a sport set-ting The original scale consisted of 24 items measuringsix subscales that represent AS AT CS CT RS and RTinterpersonal behaviors The structure of the scale wasdetermined through three studies using undergraduatestudents The purpose was to create a scale measuringperceptions of othersrsquo interpersonal behaviors as wellas a self-report of onersquos own behavior that could beused across multiple contexts As such the items were

developed by a pool of experts with advanced knowl-edge of SDT using a general stem (ldquoThe people in myliferdquomdashIBQ ldquoWhen I am with the people that are impor-tant to me IrdquomdashIBQ-Self) to ensure they were notbound to a specific interpersonal relationship or con-text In Study 1 (N = 572) the structure of the IBQ wasdetermined through a series of confirmatory factoranalyses and the results supported that the scale hada strong factor structure good internal consistencystrong convergent and divergent validity and the sub-scales correlated with other outcomesmdashas would beexpected according to SDT The results also supportedthat perceptions of need-supportive interpersonal beha-viors were related to increases in reported generalautonomy competence and relatedness satisfactionwhile perceptions of need-thwarting interpersonalbehaviors were related to decreases in the general satis-faction of all three needs These results were replicatedin Study 2 (N = 372) with a new samplemdashprovidingadditional support for the validity of the IBQ In Study3 (N = 736) the factor structure of the IBQ-Self wastested and demonstrated that it also has a strongstructure good internal consistency acceptable validityand reliability and the subscales correlated with out-comes as would be expected For the IBQ-Self it wasfound that when individuals reported they engaged inneed-supportive behaviors this was also associatedwith increased general need satisfaction and decreasedneed frustration while individuals who reportedincreased need-thwarting behaviors saw the opposite

Since the items were developed with the intention tomake the scale applicable to multiple social contextsand social relationships (ie education workplacesport etc) the objective of the present studies is tovalidate the scale as a measure of perceptions of inter-personal behaviors (IBQ) as well as a self-report ofinterpersonal behaviors (IBQ-Self) in sport In Study1 the IBQ will be validated with a sample of athletesreporting on their perceptions of their coachesrsquo inter-personal behaviors In Study 2 the IBQ-Self will bevalidated with a sample of coaches reporting on theirown interpersonal behaviors in their interactions withtheir athletes For both studies the factor structurevalidity reliability and correlations with outcomeswill be examined Overall it is anticipated these studieswill extend the validity of both the IBQ and theIBQ-Self and will demonstrate a strong fit in sport

Study 1

The objective of this study was to extend the validity ofthe IBQ by confirming the scale structure held in asport context Specifically this study will confirm the

MEASUREMENT IN PHYSICAL EDUCATION AND EXERCISE SCIENCE 3

IBQ is an appropriate measure for assessing athletesrsquoperceptions of their coachesrsquo interpersonal behaviors Itis hypothesized the structure of the scale will remainconsistent and the subscales will relate to the outcomesin the same ways as the original validation studies(Rocchi et al 2017) for athletes when they report ontheir coachesrsquo behaviors

Method

Participants

The sample was composed of 239 full-time undergrad-uate student-athletes (Nmale = 130 Nfemale = 109) withan average age of 2015 years (SD = 316) who wereenrolled in first-year courses These students had beencompeting in their sport for an average of 437 years(SD = 345) had been working with their current coachfor an average of 259 years (SD = 308) and trained anaverage of 867 hr per week (SD = 328) with them Themajority of the athletes played hockey (N = 48 16)soccer (N = 39 13) basketball (N = 15 5) volleyball(N = 16 5) or cross country running (N = 12 4)The remaining athletes (N = 109 47) came from avariety of sport backgrounds including badmintonbaseball boxing equestrianism figure skating footballgolf road cycling rowing rugby swimming tennisand weightlifting

Procedures

The athletes were selected from a research participationpool and received credit towards their final course gradefor their participation and participated in an onlinesurvey about their sport experience Participation wasvoluntary and participants gave their informed consentbefore beginning the study To be eligible for this studythe participants had to be actively training in their sportat the time of data collection and have been workingwith their current coach for at least 1 year

Materials

The following measures were completed through anonline questionnaire

Coach interpersonal behaviorsParticipants completed the 24-item IBQ scale measur-ing perceived interpersonal behaviors of their coachusing the stem ldquoMy coach rdquo (Rocchi et al 2017)The athletes indicated their agreement with each state-ment using a 7-point scale ranging from 1 (do not agreeat all) to 7 (completely agree) The scale consists of six

subscales assessing coachesrsquo use of AS AT CS CT RSand RT interpersonal behaviors in their interactionsThe original validation studies supported that the scalehad sufficient internal reliability (α gt 74)

Need satisfactionParticipants responded to the positive items from theBasic Need Satisfaction at Work Scale (Deci et al 2001)to assess the extent their three basic psychologicalneeds were met while practicing their sport The scaleconsists of three subscales measuring participantsrsquoautonomy (three items) competence (three items)and relatedness satisfaction (four items) A confirma-tory factor analysis (CFA) using Mplus Version 6(Muthen amp Muthen 2010 Los Angeles CA this soft-ware was used for all analyses) was conducted on thescale to confirm the structure The fit indices suggestthe model has an acceptable fit (Satorra-Bentler [SB]scaled χ2(41) = 6450 p lt 001 Root Mean Square Errorof Approximation [RMSEA] = 06 [90 ConfidenceInterval (CI) 04 08] Standardized Root MeanSquare Residual [SRMR] = 05 Comparative Fit Index[CFI] = 94 Tucker-Lewis Index [TLI] = 92) Theinternal consistency estimates for the three subscaleswere within acceptable range (autonomy α = 72 com-petence α = 70 relatedness α = 83) and a mean scorewas calculated for each subscale

Need frustrationTo assess psychological need frustration in sport partici-pants also completed the Psychological Need ThwartingScale in Sport (Bartholomew et al 2010) The scale alsoconsists of three subscales measuring participantsrsquo auton-omy competence and relatedness frustration with fouritems each A confirmatory factor analysis was conductedto confirm the structure and the results supported themodel had an excellent fit (SBχ2(32) = 2268 p lt 001RMSEA = 00 [90 CI 00 05] SRMR = 03CFI = 100 TLI = 100) The internal consistency for thethree subscales were within acceptable range (α gt 76)and a mean score was calculated for each subscale torepresent participantsrsquo frustration of each need

Athlete motivationParticipants also completed the Revised SportMotivation Scale (Pelletier Rocchi Vallerand Deci ampRyan 2013) to assess their reasons for participating insport The 6-factor scale is comprised of 18 items (threeper subscale) measuring sport motivation according toeach of the six types of behavioral regulation accordingto SDT A confirmatory factor analysis supported thatthe model had a good fit (SBχ2(120) = 19492 p lt 001RMSEA = 06 [90 CI 04 07] SRMR = 03

4 M ROCCHI ET AL

CFI = 95 TLI = 93) The Cronbachrsquos alpha was calcu-lated for each subscale and revealed they were abovethe acceptable limit (α gt 77) except for the introjectedsubscale which had a lower internal consistency(α = 60) A mean score was calculated for autonomous(mean of intrinsic integrated and identified regula-tion) and controlled motivation (mean of introjectedexternal and amotivated regulation)

Analyses

First the data was cleaned and screened for univariateand multivariate outliers Next a confirmatory factoranalysis was performed on the IBQ to confirm thestructure of the scale held for a sample of athletesreporting on their coachesrsquo behaviors The model wasestimated using maximum likelihood robust (MLR)estimator (Muthen amp Muthen 2010) The fit of themodel was assessed according to Hu and Bentlerrsquos(1999) recommendation of using a scaled chi-square(χ2) and the SRMR as absolute fit indices the TLI asa relative fit index and finally the RMSEA and the CFIas noncentrality-based indices For the SRMR andRMSEA values below 08 indicated adequate modelfit and values below 06 indicated excellent fit for theCFI and TLI values above 90 represented good fit andvalues above 95 indicated excellent fit (HooperCoughlan amp Mullen 2008) Next invariance testswere conducted to confirm the scale performed equallyfor both male and female athletes Then internal con-sistency was examined using Cronbachrsquos alpha for eachsubscale and the discriminant and convergent validitywere examined using the average variance extracted(AVE) and the average shared square variance (ASV)Finally a series of outcome correlation analyses wereconducted to confirm the subscales related to the out-comes as expected according to SDT

Results

First the scoring distributions of the 24 IBQ items wereexamined for univariate normality and results sug-gested that similar to the original validation studies(Rocchi et al 2017) it had not been achieved for allvariables (skewness range minus1015 to 954 kurtosisrange minus655 to 541) Next missing data patternswere examined and it was revealed 24 participants(11) were missing between 1 and 3 observations ontheir reports of the IBQ Since this represented less than1 of the overall sample the missing data was esti-mated using full information ML (FIML) Then theunivariate and multivariate outlier analyses revealedtwo multivariate outliers and they were removed from

the subsequent analyses Finally the composite scoreswere calculated for the coachesrsquo interpersonal behaviors(see Table 1 for descriptive statistics) as well as forathletesrsquo need satisfaction need frustration and sportmotivation (see Table 2 for descriptive statistics)

Scale structure

The structure of the IBQ was tested through a CFA usingthe MLR estimator The results supported the factorialmodel had excellent fit (SBχ2(237) = 29623 p lt 001CFI = 95 TLI = 95 RMSEA = 05 [90 CI 04 06]SRMR= 05) and the standardized factor loadings for eachsubscale were larger than 60 (see Table 3) Next sinceboth groups achieved the minimum recommended sam-ple size of 100 a series of invariance tests were conductedto determine the scale performed equally for both men(N = 105) and women (N = 131) athletes (MacCallumWidaman Zhang amp Hong 1999 Meade 2005) In thefirst step the baseline models were tested for each genderto confirm the structure of the IBQ held for men (SBχ2

(237) = 31776 p lt 001 CFI = 90 TLI = 90 RMSEA = 07

Table 1 Study 1 and 2 Final items with standardized factorloadings (FL) and standard errors (SE)

Study 1 Study 2

Items FL SE FL SE

STEM ldquoMy coach rdquo (Study 1) ldquoWhen I am with my athletes rdquo (Study 2)ASGives me the freedom to make my ownchoices

791 043 712 041

Supports my decisions 872 032 722 036Supports the choices that I make for myself 828 045 681 031Encourages me to make my own decisions 813 045 731 035

ATPressures me to do things their way 699 051 722 030Imposes their opinions on me 794 045 741 029Pressures me to adopt certain behaviors 756 048 693 039Limits my choices 756 051 687 103

CSEncourages me to improve my skills 693 066 658 045Provides valuable feedback 739 052 720 041Acknowledges my ability to achieve my goals 811 043 765 032Tells me that I can accomplish things 859 035 748 033

CTPoints out that I will likely fail 739 054 761 024Sends me the message that I amincompetent

830 039 652 044

Doubts my capacity to improve 741 051 674 030Questions my ability to overcome challenges 819 049 748 037

RSIs interested in what I do 804 054 715 050Takes the time to get to know me 830 039 755 032Honestly enjoy spending time with me 741 051 734 045Relates to me 819 049 633 033

RTDoes not comfort me when I am feeling low 804 054 443 028Is distant when we spend time together 862 028 529 031Does not connect with me 832 031 656 046Does not care about me 855 027 618 055

Note Study 1 N = 237 Study 2 N = 240 The verb tenses and pronouns of theitems weremodified in the IBQ-Self to reflect the new stem For example ldquoGivethem the freedom to make their own choicesrdquo instead of ldquoGives me thefreedom to make my own choicesrdquo or ldquoDo not comfort them when they arefeeling lowrdquo instead of ldquoDoes not comfort me when i am feeling lowrdquo

MEASUREMENT IN PHYSICAL EDUCATION AND EXERCISE SCIENCE 5

[90 CI 06 08] SRMR = 06) and women (SBχ2

(237) = 33495 p lt 001 CFI = 92 TLI = 91RMSEA = 07 [90 CI 06 08] SRMR = 06) Thenconfigural invariance was established by testing a CFAmodel with both groups where no constraints were placedon the parameters The results suggested the factor struc-ture of the IBQ was the same for men and women athletes(SBχ2(474) = 67240 p lt 001 CFI = 91 TLI = 90RMSEA = 07 [90 CI 06 08] SRMR = 06) Nextmetric invariance was established by constraining thefactor indicators to be equal for groups and the resultsfound the data fit the model well (SBχ2(492) = 68120p lt 001 CFI = 91 TLI = 90 RMSEA = 07 [90CI 06 08] SRMR = 07) and the change in chi-squarebetween the constrained and unconstrained model wasnot significant (SBΔχ2(18) = 1199 p gt 05) Finally scalarinvariance was established by constraining the interceptsto be equal for both groups Again the results supportedthe model had good fit (SBχ2(510) = 70916 p lt 001

CFI = 91 TLI = 90 RMSEA = 07 [90 CI 06 08]SRMR = 07) and the chi-square difference test confirmedthe model fit remained stable (SBΔχ2(18) = 2761 p gt 05)Overall these results support men and women athleteswere invariant with regards to the factor structure andcomposite variables can be created for both groups

Validity and reliability

The AVE and ASV were examined to assess conver-gent and divergent validity (see Table 1) All sub-scales met the requirements for convergent validity asthe AVEs were above 5 (Hair Black Babin ampAnderson 2010) Discriminate validity was alsoachieved since all ASV values were smaller thantheir respective AVEs (Hair et al 2010) The factorcorrelations and internal reliabilities for each subscalesupport each subscale has achieved acceptable inter-nal consistency (see Table 1)

Table 2 Study 1 and 2 Factor correlations variance means standard deviations and internal consistencySubscales 1 2 3 4 5 6 AVE ASV M SD

Study 1 (IBQmdashGeneral Stem)1 AS ndash89 68 28 503 1392 AT minus39 ndash84 57 15 323 1493 CS 70 minus30 ndash87 61 18 554 1284 CT minus34 69 minus47 ndash87 62 18 137 1375 RS 71 minus32 74 minus36 ndash90 64 19 498 1446 RT minus40 60 minus37 68 minus50 ndash87 70 17 241 141

Study 2 (IBQ-SelfmdashAthlete Stem)1 AS ndash71 51 07 601 0832 AT minus36 ndash70 51 05 211 2113 CS 40 minus14 ndash73 53 09 566 0594 CT minus13 42 04 ndash75 52 05 140 0765 RS 44 minus18 65 minus05 ndash72 50 11 600 0656 RT minus17 16 minus27 11 minus27 ndash40 33 03 141 073

Note Study 1 N = 237 Study 2 N = 240 α = diagonals AVE = average variance extracted ASV = average shared square variancep lt 05 p lt 001

Table 3 Study 1 and 2 Descriptive statistics and correlations with outcome variables for the IBQ (Study1) and the IBQ-Self (Study 2)

IBQ Subscales

Need Satisfaction Need Frustration Motivation

Soc DesAut Comp Rel Aut Comp Rel Aut Ctl

Study 1AS 58 48 46 minus28 minus32 minus24 40 minus08 mdashAT minus33 minus20 minus27 63 56 54 minus08 45 mdashCS 43 52 48 minus30 minus40 minus34 39 minus17 mdashCT minus30 minus23 minus33 64 66 63 minus14 51 mdashRS 52 50 49 minus26 minus32 minus25 37 minus06 mdashRT minus44 minus29 minus39 62 58 64 minus25 41 mdashMean 519 570 582 304 286 254 550 326 mdashStandard Deviation 114 104 103 114 154 103 116 112 mdash

Study 2AS 20 23 23 minus02 minus02 minus01 25 02 10AT minus11 05 02 23 30 24 11 36 06CS 31 48 34 minus18 minus11 minus06 28 minus01 10CT 01 06 minus03 07 28 26 06 minus30 08RS 36 53 37 minus16 minus12 minus12 29 05 06RT minus04 minus06 minus02 01 22 35 12 minus19 11Mean 611 504 132 173 160 167 505 229 453Standard Deviation 092 079 088 106 100 078 120 097 257

Note Study 1 N = 237 Study 2 N = 240 Soc Des = social desirability For Study 2 correlations are while controlling for social desirability except for thecorrelations between the subscales of the IBQ and social desirability Need satisfaction and frustration are related to sport or coaching Motivation is forparticipating in sport or coaching

p lt 05 p lt 001

6 M ROCCHI ET AL

Outcome correlations

The IBQ subscales were correlated with need satisfaction insport need frustration in sport and sport motivation aspredicted (see Table 3) Specifically the need-supportivesubscales of the IBQ correlated positively with need satis-faction and negatively with need frustration The need-thwarting subscales of the IBQ correlated negatively withneed satisfaction and positively with need frustrationLooking at athletesrsquomotivation for sport the need-suppor-tive subscales correlated positively with autonomous moti-vation and the need-thwarting subscales correlated withcontrolled motivation For CT it was related to an increasein athletesrsquo autonomous motivation and a decrease in theircontrolled motivation For both perceptions of CT and RTcoach behaviors this was related to decreases in autono-mous motivation for athletes as well as an increase incontrolled motivation

Discussion

Overall the results of Study 1 support the psychometricproperties of the IBQ in a sample of athletes reportingon their coachesrsquo interpersonal behaviors and the scaleperformed equally for male and female athletes This isthe first set of results to support the IBQ can be used toassess perceptions of interpersonal behaviors in thesport context In this sample the structure of the scaleheld as well as the reliabilities and outcome correla-tions supporting the scale performed as would beexpected similar to the original validation studies(Rocchi et al 2017) Although there are relativelyhigh relationships between the need-supportive andneed-thwarting subscales that are in line with theresults of the original validation studies the relation-ships are not high enough to suggest multicollinearity(lt 80) may be an issue (Field 2009) The results of thisstudy support the validity of the IBQ to assess athletesrsquoperceptions of coachesrsquo behaviors Specifically theseresults are in line with what would be expected accord-ing to SDT as athletes who report their coaches useneed-supportive interpersonal behaviors reportedincreased need satisfaction and autonomous motivationin sport (Mageau amp Vallerand 2003) while athleteswho reported their coaches use need-thwarting inter-personal behaviors reported need frustration and con-trolled motivation in sport (Bartholomew et al 2009)

Study 2

The objective of Study 2 is to evaluate the psychometricproperties of the IBQ self-report version within the

sport context Specifically this will be validated usinga sample of sport coaches who are reporting on theirbehaviors when they interact with their athletes It isanticipated the factor structure of the IBQ-Self will holdfor this sample in this context and it will correlate withoutcomes as would be expected according to SDT

Method

Participants

The sample was composed of 240 coaches (Nmale = 170Nfemale = 66 Nmissing = 4) with an average age of4701 years (SD = 1031) The large majority of thecoaches had a college (N = 53 22) or universityeducation (N = 168 70) They were either basketball(N = 111 46) or track and field coaches (N = 12954) and they had an average of 1750 years of coach-ing experience (SD = 1283) The majority (N = 12251) identified as a head coach or as an assistant coach(N = 60 25) and indicated their athletes were very(N = 94 39) or extremely (N = 61 25) competitivecompared to other athletes of their age

Procedures

Coaches were recruited through their respective pro-vincial sporting associations (basketball or athletics) toparticipate in an online survey Each organization sentan email to all coaches inviting them to participate inan online study The coachesrsquo participation was volun-tary and they gave their informed consent before parti-cipating To be eligible the coaches had to be registeredwith a provincial sport organization at the time of datacollection and actively coaching

Materials

The following measures were completed by the coachesthrough an online questionnaire

My interpersonal behaviorsThe coaches completed the IBQ-Self to assess the extentto which they believed they engaged in AS AT CS CTRS and RT interpersonal behaviors with their athletes(Rocchi et al 2017) Coaches were presented each itemusing the stem ldquoWhen I am with my athletes I rdquo andindicated their agreement with each statement using a7-point scale ranging from 1 (do not agree at all) to 7(completely agree) The original validation studies sup-ported the scale had sufficient internal reliabil-ity (α gt 80)

MEASUREMENT IN PHYSICAL EDUCATION AND EXERCISE SCIENCE 7

Need satisfactionThe coaches responded to the positive items from theBasic Need Satisfaction at Work Scale adapted to thesport context (Deci et al 2001) to assess the extenttheir three basic psychological needs were met whilecoaching Like in Study 1 the stems were modified toask the coaches about their need satisfaction whilecoaching instead of at work The fit indices suggestthe model has an acceptable fit except for the TLI(SBχ2(41) = 3836 p lt 001 RMSEA = 06 [90 CI03 08] SRMR = 05 CFI = 90 TLI = 89) howeversince the other fit indices were acceptable the scale wasused The internal consistency estimates were withinacceptable range (autonomy α = 83 competenceα = 71 relatedness α = 74) and a mean score wascalculated for each subscale

Need frustrationThe coaches also completed the Psychological NeedThwarting Scale in Sport (Bartholomew et al 2010) toassess the extent to which their needs were frustratedwhile coaching like in Study 1 A confirmatory factoranalysis confirmed the structure of the scale for coachesand the results supported that the model had an excel-lent fit (SBχ2(32) = 2600 p lt 001 RMSEA = 04 [90CI 00 07] SRMR = 06 CFI = 96 TLI = 94) Theinternal consistency for the three subscales were abovethe minimum (α gt 74) and a mean score was calcu-lated for each subscale to represent the coachesrsquo frus-tration of each need

Coach motivationThe coaches completed the Coach MotivationQuestionnaire (CMQ McLean Mallet amp Newcombe2012) to assess their reasons for participating in sportThe CMQ is a 6-factor scale comprised of 22 itemsmeasuring sport motivation according to each of thesix types of behavioral regulation according to SDT Aconfirmatory factor confirmed the structure of the scalein this sample (SBχ2(194) = 33657 p lt 001RMSEA = 06 [90 CI 04 07] SRMR = 07CFI = 91 TLI = 90) The Cronbachrsquos alphas werecalculated for each subscale and revealed they wereabove the acceptable limit (α gt 72) except for theintrojected subscale which had a lower internal consis-tency (α = 68) Like in Study 1 a mean score wascalculated for autonomous motivation and controlledmotivation

Social desirabilityFinally the coaches also completed the short form ofthe Marlowe-Crowne Social Desirability Scale(Reynolds 1982) to control for whether they were

responding to survey items based on providing favor-able responses The short form of this measure consistsof 10 items where participants are asked to indicatewhether these statements are true or false for themThe sum of all responses considered to be sociallydesirable is calculated to provide an overall measureof social desirability

Analyses

Using the same procedures as the first study the datawas cleaned and then the scale structure was examinedThen the validity and reliability of the subscales wereexamined Finally outcome correlations were con-ducted to examine the relationship between coachneed satisfaction frustration and motivation whilecontrolling for social desirability

Results

Similar to the first study the descriptive statistics analyseson the 24 IBQ-Self items suggested the variables did notall have a normal distribution (skewness range minus1325 to912 kurtosis range minus620 to 821) Data screening ana-lyses did not reveal any multivariate outliers as such theentire sample was retained for the full analyses Themissing data analyses suggested 15 participants weremissing one or two observations on the IBQ-Self how-ever since these missing observations made up less than5 of the total observations the data was imputed usingthe same methods as Study 1 The composite scores forthe coachesrsquo reported use of need-supportive andneed-thwarting behaviors as well as their need satisfac-tion and frustration in coaching their motivation towardscoaching and their likelihood of responding in sociallydesirable ways was calculated (see Tables 1 and 2 for thedescriptive statistics)

Scale structure

The results supported that the factorial model had goodfit (SBχ2(237) = 30304 p lt 001 CFI = 93 TLI = 92RMSEA = 06 [90 CI 03 05] SRMR = 04) Thestandardized factor loadings for each subscale werelarger than 46 (see Table 1) and the factor correlationsas well as the internal reliabilities for each subscale(Cronbachrsquos alpha) are in Table 1 Since there wereonly 66 female coaches in this sample gender invar-iance tests were not performed in this study Nextconvergent and divergent validity was examined foreach of the subscales (see Table 2) and the resultssupported that the scale met the thresholds for bothacross all subscales except for the RT subscale Internal

8 M ROCCHI ET AL

consistency analyses revealed all of the alphas were overthe minimum criteria again except for the RT subscale(α = 40) which was nowhere near the baseline criteriaThe internal consistency for that subscale was calcu-lated using all of the potential combinations of items tosee if one specific item was causing the issue and theresults found there was no combination of the fouritems that promoted a better internal consistency Asa result the reliability for the four items was retained

Outcome correlations

The IBQ-Self subscales were correlated with need satis-faction need frustration autonomous and controlledmotivation while controlling for social desirability(see Table 2) to confirm the relationships were in theexpected directions The results found that the supportsubscales of the IBQ-Self correlated moderately andpositively with need satisfaction and autonomous moti-vation while they had weak or negative relationshipswith need frustration and controlled motivation Theopposite was found with the need-thwarting subscaleswhere there were moderate positive relationships withneed frustration and controlled motivation and weaknegative or no relationships with need satisfaction andautonomous motivation Overall the results were simi-lar to those in Study 1 except the strength of therelationships was weaker

Discussion

The results of this study support the factor structureand validity of the IBQ-Self as a questionnaire forcoaches about their use of interpersonal behaviorsaccording to SDT in their interactions with theirathletes Unfortunately the reliability for the RT sub-scale was weak Since some coaches and athletes maychoose to engage in the relationship when theychoose to participate in the sport (ie select athletefor a team) and others do not (ie everyone trainsduring the same session) it is not totally surprisingthat coaches care about their athletes and this madethe results somewhat ambiguous Perhaps in anothercontext where there is either less control over beingin a relationship with the other people such as afamily member or a supervisor at work or morecontrol like a best friend or partner that subscaleinternal consistency would improve Overall theseresults support that coaches who report using moreneed-supportive behaviors also report higher needsatisfaction and autonomous motivation for theircoaching while coaches who report more need-thwarting behaviors report more need frustration

and controlled motivation for coachingmdashall aftercontrolling for social desirability The results of thisstudy support that generally the IBQ-Self is a validmeasure of coachesrsquo own reports of their interperso-nal behaviors when interacting with their athletes

General discussion

The purpose of these two studies was to extend thevalidity of the IBQ by testing the structure of the scalein a sport setting The results of Study 1 supported thatthe IBQ scale factor structure holds in cases whereathletes report on their perceptions of their coachesrsquointerpersonal behaviors The scale also demonstratedadequate validity and reliability and overall the inter-correlations between the IBQ subscales as well as thecorrelations between the IBQ subscales and the otherrelated factors provided support for the fit of the scalewithin the existing SDT measures and literature Interms of the relationships between the IBQ subscalesthe results supported there were moderately high posi-tive and negative correlations between the subscalesAdditionally the subscales also correlated with needsatisfaction need frustration and motivation in thedirections that would be expected according to SDTIn the case of some relationships the outcomes hadsimilar correlations with all three need-supportivebehaviors and all three need-thwarting behaviors butthis is not unexpected since the satisfaction or frustra-tion of the three basic needs are often related (Deci ampRyan 1985)

The results of Study 2 supported that the IBQ-Selfstructure held in a sample of coachesmdashreporting ontheir behaviors with their athletes Similar to Study 1the scale also demonstrated adequate validity and relia-bility Overall the inter-correlations between theIBQ-Self subscales and the correlations between theIBQ-Self subscales and other factors showed a similarpattern to Study 1mdashsupporting that the scale relates toother factors as would be expected

Limitations

These results support the applicability of the scale toa sport context however there are some limitationsFirst the structure of the scales was tested using astudent sample of athletes as well as basketball andtrack and field coaches The IBQ should be testedwith a more varied sample of athletes to ensure thescale is a valid measure of perceptions of coaches forother groups such as older athletes and recreationalathletes Furthermore the IBQ-Self should be vali-dated with coaches from a larger variety of sports

MEASUREMENT IN PHYSICAL EDUCATION AND EXERCISE SCIENCE 9

In this sample coaches were coaching team (basket-ball) and individual sports (athletics) however coa-ches can work with all sorts of different athletes atvarying levels of competition and the scalersquos validityshould be extended to include these coaches too Asecond limitation is that the IBQ and IBQ-Self onlyfocused on two people within the sport context ath-letes and coaches There are a number of differentsocial agents in the sport context that impact athletes(or coaches) including the athletesrsquo parents the otherathletes sport administrators etc and the validity ofthe scale should also be tested to examine perceptionsof their behaviors or their own reports of their beha-viors Finally since the IBQ and IBQ-Self were vali-dated using coachesrsquo and athletesrsquo self-reports theirreports should be triangulated by a third party (egvideo recording and coding of interpersonal beha-viors) to confirm reports on these measures corre-spond to their real world behaviors

Future directions

The IBQ and the IBQ-Self promote a number ofinitiatives for future research in sport motivationFirst research should continue to examine the psy-chometric properties of the IBQ For example ath-letes or coaches should be examined at multiple timepoints in order to establish the scalersquos testndashretestreliability and coachesrsquo scores should be corroboratedwith athletesrsquo scores in order to provide support forthe construct validity of the scale Next the IBQ canbe used to explore the role of all six types of need-supportive and need-thwarting interpersonal beha-viors in relation to how they impact athletesrsquo psycho-logical need satisfaction and frustration This willhelp extend the existing research in sport to movebeyond the influential role of AS and AT and focuson CS CT RS and RT interpersonal behaviors aswell Next the IBQ-Self can be used to identify andunderstand the antecedents of all six types of inter-personal behaviors according to SDT Specifically thescale should be used to explore the factors that influ-ence coachesrsquo behaviors with their athletes Finallythe IBQ and IBQ-Self should be used to explorehow coachesrsquo behaviors change over the course of agiven season and how this relates to athlete psycho-logical needs and their motivation for sport

Overall the IBQ and IBQ-Self are new instrumentsfor assessing perceptions or self-reports of all six typesof need-supportive and need-thwarting interpersonalbehaviors for coaches and athletes It is hoped theIBQ and the IBQ-Self will promote research directed

at assessing the behavior of others in sport and how itimpacts need satisfaction need frustration and motiva-tion for athletes

References

Amorose J amp Anderson-Butcher D (2007) Autonomy-supportive coaching and self-determined motivation inhigh school and college athletes A test of self-determina-tion theory Psychology of Sport and Exercise 8 654ndash670doi101016jpsychsport200611003

Bartholomew K Ntoumanis N Ryan R Bosch J ampThoslashgersen-Ntoumani C (2011) Self-determination the-ory and diminished functioning The role of interpersonalcontrol and psychological need thwarting Personality andSocial Psychology Bulletin 37 1459ndash1473 doi1011770146167211413125

Bartholomew K Ntoumanis N amp Thoslashgersen-Ntoumani C(2009) A review of controlling motivational strategies from aself-determination theory perspective Implications for sportscoaches International Review of Sport and ExercisePsychology 2 215ndash233 doi10108017509840903235330

Bartholomew K Ntoumanis N amp Thoslashrgenson-NtoumaniC (2010) The controlling interpersonal style in a coachingcontext Development and initial validation of a psycho-metric scale Journal of Sport amp Exercise Psychology 32193ndash216 Retrieved from httpwwwncbinlmnihgovpubmed20479478

Conroy D amp Coatsworth J (2007) Assessing autonomy-sup-portive coaching strategies in youth sport Psychology of Sportand Exercise 8 671ndash684 doi101016jpsychsport200612001

Deci E amp Ryan R (1985) Intrinsic motivation and self-determination in human behaviour In E Deci amp R Ryan(Eds) Handbook of self-determination research RochesterNY University of Rochester Press

Deci E amp Ryan R (2002) Handbook of self-determinationresearch Rochester NY The University of Rochester Press

Deci E Ryan R Gagneacute M Leone D Usunov J ampKornazheva B (2001) Need satisfaction motivation andwell-being in the work organizations of a former easternbloc country A cross-cultural study of self-determinationPersonality and Social Psychology Bulletin 27 930ndash942doi1011770146167201278002

Deci E Schwartz A Sheinman L amp Ryan R (1981) Aninstrument to assess adultsrsquo orientations toward control ver-sus autonomy with children Reflections on intrinsic motiva-tion and perceived competence Journal of EducationalPsychology 73 642ndash650 doi1010370022-0663735642

Field A (2009) Discovering statistics using SPSS Sage Publi-cations

Gould D Dieffenbach K amp Moffett A (2002) Psychologicalcharacteristics and their development in Olympic cham-pions Journal of Applied Sports Psychology 14 172ndash204doi10108010413200290103482

Hagger M amp Chatzisarantis N (2007) Intrinsic motivationand self-determination in exercise and sport Leeds UKHuman Kinetics Europe Ltd

Hair J Black W Babin B amp Anderson R (2010)Multivariate data analysis (7th ed) Upper Saddle RiverNJ Prentice-Hall Inc

10 M ROCCHI ET AL

IBQ is an appropriate measure for assessing athletesrsquoperceptions of their coachesrsquo interpersonal behaviors Itis hypothesized the structure of the scale will remainconsistent and the subscales will relate to the outcomesin the same ways as the original validation studies(Rocchi et al 2017) for athletes when they report ontheir coachesrsquo behaviors

Method

Participants

The sample was composed of 239 full-time undergrad-uate student-athletes (Nmale = 130 Nfemale = 109) withan average age of 2015 years (SD = 316) who wereenrolled in first-year courses These students had beencompeting in their sport for an average of 437 years(SD = 345) had been working with their current coachfor an average of 259 years (SD = 308) and trained anaverage of 867 hr per week (SD = 328) with them Themajority of the athletes played hockey (N = 48 16)soccer (N = 39 13) basketball (N = 15 5) volleyball(N = 16 5) or cross country running (N = 12 4)The remaining athletes (N = 109 47) came from avariety of sport backgrounds including badmintonbaseball boxing equestrianism figure skating footballgolf road cycling rowing rugby swimming tennisand weightlifting

Procedures

The athletes were selected from a research participationpool and received credit towards their final course gradefor their participation and participated in an onlinesurvey about their sport experience Participation wasvoluntary and participants gave their informed consentbefore beginning the study To be eligible for this studythe participants had to be actively training in their sportat the time of data collection and have been workingwith their current coach for at least 1 year

Materials

The following measures were completed through anonline questionnaire

Coach interpersonal behaviorsParticipants completed the 24-item IBQ scale measur-ing perceived interpersonal behaviors of their coachusing the stem ldquoMy coach rdquo (Rocchi et al 2017)The athletes indicated their agreement with each state-ment using a 7-point scale ranging from 1 (do not agreeat all) to 7 (completely agree) The scale consists of six

subscales assessing coachesrsquo use of AS AT CS CT RSand RT interpersonal behaviors in their interactionsThe original validation studies supported that the scalehad sufficient internal reliability (α gt 74)

Need satisfactionParticipants responded to the positive items from theBasic Need Satisfaction at Work Scale (Deci et al 2001)to assess the extent their three basic psychologicalneeds were met while practicing their sport The scaleconsists of three subscales measuring participantsrsquoautonomy (three items) competence (three items)and relatedness satisfaction (four items) A confirma-tory factor analysis (CFA) using Mplus Version 6(Muthen amp Muthen 2010 Los Angeles CA this soft-ware was used for all analyses) was conducted on thescale to confirm the structure The fit indices suggestthe model has an acceptable fit (Satorra-Bentler [SB]scaled χ2(41) = 6450 p lt 001 Root Mean Square Errorof Approximation [RMSEA] = 06 [90 ConfidenceInterval (CI) 04 08] Standardized Root MeanSquare Residual [SRMR] = 05 Comparative Fit Index[CFI] = 94 Tucker-Lewis Index [TLI] = 92) Theinternal consistency estimates for the three subscaleswere within acceptable range (autonomy α = 72 com-petence α = 70 relatedness α = 83) and a mean scorewas calculated for each subscale

Need frustrationTo assess psychological need frustration in sport partici-pants also completed the Psychological Need ThwartingScale in Sport (Bartholomew et al 2010) The scale alsoconsists of three subscales measuring participantsrsquo auton-omy competence and relatedness frustration with fouritems each A confirmatory factor analysis was conductedto confirm the structure and the results supported themodel had an excellent fit (SBχ2(32) = 2268 p lt 001RMSEA = 00 [90 CI 00 05] SRMR = 03CFI = 100 TLI = 100) The internal consistency for thethree subscales were within acceptable range (α gt 76)and a mean score was calculated for each subscale torepresent participantsrsquo frustration of each need

Athlete motivationParticipants also completed the Revised SportMotivation Scale (Pelletier Rocchi Vallerand Deci ampRyan 2013) to assess their reasons for participating insport The 6-factor scale is comprised of 18 items (threeper subscale) measuring sport motivation according toeach of the six types of behavioral regulation accordingto SDT A confirmatory factor analysis supported thatthe model had a good fit (SBχ2(120) = 19492 p lt 001RMSEA = 06 [90 CI 04 07] SRMR = 03

4 M ROCCHI ET AL

CFI = 95 TLI = 93) The Cronbachrsquos alpha was calcu-lated for each subscale and revealed they were abovethe acceptable limit (α gt 77) except for the introjectedsubscale which had a lower internal consistency(α = 60) A mean score was calculated for autonomous(mean of intrinsic integrated and identified regula-tion) and controlled motivation (mean of introjectedexternal and amotivated regulation)

Analyses

First the data was cleaned and screened for univariateand multivariate outliers Next a confirmatory factoranalysis was performed on the IBQ to confirm thestructure of the scale held for a sample of athletesreporting on their coachesrsquo behaviors The model wasestimated using maximum likelihood robust (MLR)estimator (Muthen amp Muthen 2010) The fit of themodel was assessed according to Hu and Bentlerrsquos(1999) recommendation of using a scaled chi-square(χ2) and the SRMR as absolute fit indices the TLI asa relative fit index and finally the RMSEA and the CFIas noncentrality-based indices For the SRMR andRMSEA values below 08 indicated adequate modelfit and values below 06 indicated excellent fit for theCFI and TLI values above 90 represented good fit andvalues above 95 indicated excellent fit (HooperCoughlan amp Mullen 2008) Next invariance testswere conducted to confirm the scale performed equallyfor both male and female athletes Then internal con-sistency was examined using Cronbachrsquos alpha for eachsubscale and the discriminant and convergent validitywere examined using the average variance extracted(AVE) and the average shared square variance (ASV)Finally a series of outcome correlation analyses wereconducted to confirm the subscales related to the out-comes as expected according to SDT

Results

First the scoring distributions of the 24 IBQ items wereexamined for univariate normality and results sug-gested that similar to the original validation studies(Rocchi et al 2017) it had not been achieved for allvariables (skewness range minus1015 to 954 kurtosisrange minus655 to 541) Next missing data patternswere examined and it was revealed 24 participants(11) were missing between 1 and 3 observations ontheir reports of the IBQ Since this represented less than1 of the overall sample the missing data was esti-mated using full information ML (FIML) Then theunivariate and multivariate outlier analyses revealedtwo multivariate outliers and they were removed from

the subsequent analyses Finally the composite scoreswere calculated for the coachesrsquo interpersonal behaviors(see Table 1 for descriptive statistics) as well as forathletesrsquo need satisfaction need frustration and sportmotivation (see Table 2 for descriptive statistics)

Scale structure

The structure of the IBQ was tested through a CFA usingthe MLR estimator The results supported the factorialmodel had excellent fit (SBχ2(237) = 29623 p lt 001CFI = 95 TLI = 95 RMSEA = 05 [90 CI 04 06]SRMR= 05) and the standardized factor loadings for eachsubscale were larger than 60 (see Table 3) Next sinceboth groups achieved the minimum recommended sam-ple size of 100 a series of invariance tests were conductedto determine the scale performed equally for both men(N = 105) and women (N = 131) athletes (MacCallumWidaman Zhang amp Hong 1999 Meade 2005) In thefirst step the baseline models were tested for each genderto confirm the structure of the IBQ held for men (SBχ2

(237) = 31776 p lt 001 CFI = 90 TLI = 90 RMSEA = 07

Table 1 Study 1 and 2 Final items with standardized factorloadings (FL) and standard errors (SE)

Study 1 Study 2

Items FL SE FL SE

STEM ldquoMy coach rdquo (Study 1) ldquoWhen I am with my athletes rdquo (Study 2)ASGives me the freedom to make my ownchoices

791 043 712 041

Supports my decisions 872 032 722 036Supports the choices that I make for myself 828 045 681 031Encourages me to make my own decisions 813 045 731 035

ATPressures me to do things their way 699 051 722 030Imposes their opinions on me 794 045 741 029Pressures me to adopt certain behaviors 756 048 693 039Limits my choices 756 051 687 103

CSEncourages me to improve my skills 693 066 658 045Provides valuable feedback 739 052 720 041Acknowledges my ability to achieve my goals 811 043 765 032Tells me that I can accomplish things 859 035 748 033

CTPoints out that I will likely fail 739 054 761 024Sends me the message that I amincompetent

830 039 652 044

Doubts my capacity to improve 741 051 674 030Questions my ability to overcome challenges 819 049 748 037

RSIs interested in what I do 804 054 715 050Takes the time to get to know me 830 039 755 032Honestly enjoy spending time with me 741 051 734 045Relates to me 819 049 633 033

RTDoes not comfort me when I am feeling low 804 054 443 028Is distant when we spend time together 862 028 529 031Does not connect with me 832 031 656 046Does not care about me 855 027 618 055

Note Study 1 N = 237 Study 2 N = 240 The verb tenses and pronouns of theitems weremodified in the IBQ-Self to reflect the new stem For example ldquoGivethem the freedom to make their own choicesrdquo instead of ldquoGives me thefreedom to make my own choicesrdquo or ldquoDo not comfort them when they arefeeling lowrdquo instead of ldquoDoes not comfort me when i am feeling lowrdquo

MEASUREMENT IN PHYSICAL EDUCATION AND EXERCISE SCIENCE 5

[90 CI 06 08] SRMR = 06) and women (SBχ2

(237) = 33495 p lt 001 CFI = 92 TLI = 91RMSEA = 07 [90 CI 06 08] SRMR = 06) Thenconfigural invariance was established by testing a CFAmodel with both groups where no constraints were placedon the parameters The results suggested the factor struc-ture of the IBQ was the same for men and women athletes(SBχ2(474) = 67240 p lt 001 CFI = 91 TLI = 90RMSEA = 07 [90 CI 06 08] SRMR = 06) Nextmetric invariance was established by constraining thefactor indicators to be equal for groups and the resultsfound the data fit the model well (SBχ2(492) = 68120p lt 001 CFI = 91 TLI = 90 RMSEA = 07 [90CI 06 08] SRMR = 07) and the change in chi-squarebetween the constrained and unconstrained model wasnot significant (SBΔχ2(18) = 1199 p gt 05) Finally scalarinvariance was established by constraining the interceptsto be equal for both groups Again the results supportedthe model had good fit (SBχ2(510) = 70916 p lt 001

CFI = 91 TLI = 90 RMSEA = 07 [90 CI 06 08]SRMR = 07) and the chi-square difference test confirmedthe model fit remained stable (SBΔχ2(18) = 2761 p gt 05)Overall these results support men and women athleteswere invariant with regards to the factor structure andcomposite variables can be created for both groups

Validity and reliability

The AVE and ASV were examined to assess conver-gent and divergent validity (see Table 1) All sub-scales met the requirements for convergent validity asthe AVEs were above 5 (Hair Black Babin ampAnderson 2010) Discriminate validity was alsoachieved since all ASV values were smaller thantheir respective AVEs (Hair et al 2010) The factorcorrelations and internal reliabilities for each subscalesupport each subscale has achieved acceptable inter-nal consistency (see Table 1)

Table 2 Study 1 and 2 Factor correlations variance means standard deviations and internal consistencySubscales 1 2 3 4 5 6 AVE ASV M SD

Study 1 (IBQmdashGeneral Stem)1 AS ndash89 68 28 503 1392 AT minus39 ndash84 57 15 323 1493 CS 70 minus30 ndash87 61 18 554 1284 CT minus34 69 minus47 ndash87 62 18 137 1375 RS 71 minus32 74 minus36 ndash90 64 19 498 1446 RT minus40 60 minus37 68 minus50 ndash87 70 17 241 141

Study 2 (IBQ-SelfmdashAthlete Stem)1 AS ndash71 51 07 601 0832 AT minus36 ndash70 51 05 211 2113 CS 40 minus14 ndash73 53 09 566 0594 CT minus13 42 04 ndash75 52 05 140 0765 RS 44 minus18 65 minus05 ndash72 50 11 600 0656 RT minus17 16 minus27 11 minus27 ndash40 33 03 141 073

Note Study 1 N = 237 Study 2 N = 240 α = diagonals AVE = average variance extracted ASV = average shared square variancep lt 05 p lt 001

Table 3 Study 1 and 2 Descriptive statistics and correlations with outcome variables for the IBQ (Study1) and the IBQ-Self (Study 2)

IBQ Subscales

Need Satisfaction Need Frustration Motivation

Soc DesAut Comp Rel Aut Comp Rel Aut Ctl

Study 1AS 58 48 46 minus28 minus32 minus24 40 minus08 mdashAT minus33 minus20 minus27 63 56 54 minus08 45 mdashCS 43 52 48 minus30 minus40 minus34 39 minus17 mdashCT minus30 minus23 minus33 64 66 63 minus14 51 mdashRS 52 50 49 minus26 minus32 minus25 37 minus06 mdashRT minus44 minus29 minus39 62 58 64 minus25 41 mdashMean 519 570 582 304 286 254 550 326 mdashStandard Deviation 114 104 103 114 154 103 116 112 mdash

Study 2AS 20 23 23 minus02 minus02 minus01 25 02 10AT minus11 05 02 23 30 24 11 36 06CS 31 48 34 minus18 minus11 minus06 28 minus01 10CT 01 06 minus03 07 28 26 06 minus30 08RS 36 53 37 minus16 minus12 minus12 29 05 06RT minus04 minus06 minus02 01 22 35 12 minus19 11Mean 611 504 132 173 160 167 505 229 453Standard Deviation 092 079 088 106 100 078 120 097 257

Note Study 1 N = 237 Study 2 N = 240 Soc Des = social desirability For Study 2 correlations are while controlling for social desirability except for thecorrelations between the subscales of the IBQ and social desirability Need satisfaction and frustration are related to sport or coaching Motivation is forparticipating in sport or coaching

p lt 05 p lt 001

6 M ROCCHI ET AL

Outcome correlations

The IBQ subscales were correlated with need satisfaction insport need frustration in sport and sport motivation aspredicted (see Table 3) Specifically the need-supportivesubscales of the IBQ correlated positively with need satis-faction and negatively with need frustration The need-thwarting subscales of the IBQ correlated negatively withneed satisfaction and positively with need frustrationLooking at athletesrsquomotivation for sport the need-suppor-tive subscales correlated positively with autonomous moti-vation and the need-thwarting subscales correlated withcontrolled motivation For CT it was related to an increasein athletesrsquo autonomous motivation and a decrease in theircontrolled motivation For both perceptions of CT and RTcoach behaviors this was related to decreases in autono-mous motivation for athletes as well as an increase incontrolled motivation

Discussion

Overall the results of Study 1 support the psychometricproperties of the IBQ in a sample of athletes reportingon their coachesrsquo interpersonal behaviors and the scaleperformed equally for male and female athletes This isthe first set of results to support the IBQ can be used toassess perceptions of interpersonal behaviors in thesport context In this sample the structure of the scaleheld as well as the reliabilities and outcome correla-tions supporting the scale performed as would beexpected similar to the original validation studies(Rocchi et al 2017) Although there are relativelyhigh relationships between the need-supportive andneed-thwarting subscales that are in line with theresults of the original validation studies the relation-ships are not high enough to suggest multicollinearity(lt 80) may be an issue (Field 2009) The results of thisstudy support the validity of the IBQ to assess athletesrsquoperceptions of coachesrsquo behaviors Specifically theseresults are in line with what would be expected accord-ing to SDT as athletes who report their coaches useneed-supportive interpersonal behaviors reportedincreased need satisfaction and autonomous motivationin sport (Mageau amp Vallerand 2003) while athleteswho reported their coaches use need-thwarting inter-personal behaviors reported need frustration and con-trolled motivation in sport (Bartholomew et al 2009)

Study 2

The objective of Study 2 is to evaluate the psychometricproperties of the IBQ self-report version within the

sport context Specifically this will be validated usinga sample of sport coaches who are reporting on theirbehaviors when they interact with their athletes It isanticipated the factor structure of the IBQ-Self will holdfor this sample in this context and it will correlate withoutcomes as would be expected according to SDT

Method

Participants

The sample was composed of 240 coaches (Nmale = 170Nfemale = 66 Nmissing = 4) with an average age of4701 years (SD = 1031) The large majority of thecoaches had a college (N = 53 22) or universityeducation (N = 168 70) They were either basketball(N = 111 46) or track and field coaches (N = 12954) and they had an average of 1750 years of coach-ing experience (SD = 1283) The majority (N = 12251) identified as a head coach or as an assistant coach(N = 60 25) and indicated their athletes were very(N = 94 39) or extremely (N = 61 25) competitivecompared to other athletes of their age

Procedures

Coaches were recruited through their respective pro-vincial sporting associations (basketball or athletics) toparticipate in an online survey Each organization sentan email to all coaches inviting them to participate inan online study The coachesrsquo participation was volun-tary and they gave their informed consent before parti-cipating To be eligible the coaches had to be registeredwith a provincial sport organization at the time of datacollection and actively coaching

Materials

The following measures were completed by the coachesthrough an online questionnaire

My interpersonal behaviorsThe coaches completed the IBQ-Self to assess the extentto which they believed they engaged in AS AT CS CTRS and RT interpersonal behaviors with their athletes(Rocchi et al 2017) Coaches were presented each itemusing the stem ldquoWhen I am with my athletes I rdquo andindicated their agreement with each statement using a7-point scale ranging from 1 (do not agree at all) to 7(completely agree) The original validation studies sup-ported the scale had sufficient internal reliabil-ity (α gt 80)

MEASUREMENT IN PHYSICAL EDUCATION AND EXERCISE SCIENCE 7

Need satisfactionThe coaches responded to the positive items from theBasic Need Satisfaction at Work Scale adapted to thesport context (Deci et al 2001) to assess the extenttheir three basic psychological needs were met whilecoaching Like in Study 1 the stems were modified toask the coaches about their need satisfaction whilecoaching instead of at work The fit indices suggestthe model has an acceptable fit except for the TLI(SBχ2(41) = 3836 p lt 001 RMSEA = 06 [90 CI03 08] SRMR = 05 CFI = 90 TLI = 89) howeversince the other fit indices were acceptable the scale wasused The internal consistency estimates were withinacceptable range (autonomy α = 83 competenceα = 71 relatedness α = 74) and a mean score wascalculated for each subscale

Need frustrationThe coaches also completed the Psychological NeedThwarting Scale in Sport (Bartholomew et al 2010) toassess the extent to which their needs were frustratedwhile coaching like in Study 1 A confirmatory factoranalysis confirmed the structure of the scale for coachesand the results supported that the model had an excel-lent fit (SBχ2(32) = 2600 p lt 001 RMSEA = 04 [90CI 00 07] SRMR = 06 CFI = 96 TLI = 94) Theinternal consistency for the three subscales were abovethe minimum (α gt 74) and a mean score was calcu-lated for each subscale to represent the coachesrsquo frus-tration of each need

Coach motivationThe coaches completed the Coach MotivationQuestionnaire (CMQ McLean Mallet amp Newcombe2012) to assess their reasons for participating in sportThe CMQ is a 6-factor scale comprised of 22 itemsmeasuring sport motivation according to each of thesix types of behavioral regulation according to SDT Aconfirmatory factor confirmed the structure of the scalein this sample (SBχ2(194) = 33657 p lt 001RMSEA = 06 [90 CI 04 07] SRMR = 07CFI = 91 TLI = 90) The Cronbachrsquos alphas werecalculated for each subscale and revealed they wereabove the acceptable limit (α gt 72) except for theintrojected subscale which had a lower internal consis-tency (α = 68) Like in Study 1 a mean score wascalculated for autonomous motivation and controlledmotivation

Social desirabilityFinally the coaches also completed the short form ofthe Marlowe-Crowne Social Desirability Scale(Reynolds 1982) to control for whether they were

responding to survey items based on providing favor-able responses The short form of this measure consistsof 10 items where participants are asked to indicatewhether these statements are true or false for themThe sum of all responses considered to be sociallydesirable is calculated to provide an overall measureof social desirability

Analyses

Using the same procedures as the first study the datawas cleaned and then the scale structure was examinedThen the validity and reliability of the subscales wereexamined Finally outcome correlations were con-ducted to examine the relationship between coachneed satisfaction frustration and motivation whilecontrolling for social desirability

Results

Similar to the first study the descriptive statistics analyseson the 24 IBQ-Self items suggested the variables did notall have a normal distribution (skewness range minus1325 to912 kurtosis range minus620 to 821) Data screening ana-lyses did not reveal any multivariate outliers as such theentire sample was retained for the full analyses Themissing data analyses suggested 15 participants weremissing one or two observations on the IBQ-Self how-ever since these missing observations made up less than5 of the total observations the data was imputed usingthe same methods as Study 1 The composite scores forthe coachesrsquo reported use of need-supportive andneed-thwarting behaviors as well as their need satisfac-tion and frustration in coaching their motivation towardscoaching and their likelihood of responding in sociallydesirable ways was calculated (see Tables 1 and 2 for thedescriptive statistics)

Scale structure

The results supported that the factorial model had goodfit (SBχ2(237) = 30304 p lt 001 CFI = 93 TLI = 92RMSEA = 06 [90 CI 03 05] SRMR = 04) Thestandardized factor loadings for each subscale werelarger than 46 (see Table 1) and the factor correlationsas well as the internal reliabilities for each subscale(Cronbachrsquos alpha) are in Table 1 Since there wereonly 66 female coaches in this sample gender invar-iance tests were not performed in this study Nextconvergent and divergent validity was examined foreach of the subscales (see Table 2) and the resultssupported that the scale met the thresholds for bothacross all subscales except for the RT subscale Internal

8 M ROCCHI ET AL

consistency analyses revealed all of the alphas were overthe minimum criteria again except for the RT subscale(α = 40) which was nowhere near the baseline criteriaThe internal consistency for that subscale was calcu-lated using all of the potential combinations of items tosee if one specific item was causing the issue and theresults found there was no combination of the fouritems that promoted a better internal consistency Asa result the reliability for the four items was retained

Outcome correlations

The IBQ-Self subscales were correlated with need satis-faction need frustration autonomous and controlledmotivation while controlling for social desirability(see Table 2) to confirm the relationships were in theexpected directions The results found that the supportsubscales of the IBQ-Self correlated moderately andpositively with need satisfaction and autonomous moti-vation while they had weak or negative relationshipswith need frustration and controlled motivation Theopposite was found with the need-thwarting subscaleswhere there were moderate positive relationships withneed frustration and controlled motivation and weaknegative or no relationships with need satisfaction andautonomous motivation Overall the results were simi-lar to those in Study 1 except the strength of therelationships was weaker

Discussion

The results of this study support the factor structureand validity of the IBQ-Self as a questionnaire forcoaches about their use of interpersonal behaviorsaccording to SDT in their interactions with theirathletes Unfortunately the reliability for the RT sub-scale was weak Since some coaches and athletes maychoose to engage in the relationship when theychoose to participate in the sport (ie select athletefor a team) and others do not (ie everyone trainsduring the same session) it is not totally surprisingthat coaches care about their athletes and this madethe results somewhat ambiguous Perhaps in anothercontext where there is either less control over beingin a relationship with the other people such as afamily member or a supervisor at work or morecontrol like a best friend or partner that subscaleinternal consistency would improve Overall theseresults support that coaches who report using moreneed-supportive behaviors also report higher needsatisfaction and autonomous motivation for theircoaching while coaches who report more need-thwarting behaviors report more need frustration

and controlled motivation for coachingmdashall aftercontrolling for social desirability The results of thisstudy support that generally the IBQ-Self is a validmeasure of coachesrsquo own reports of their interperso-nal behaviors when interacting with their athletes

General discussion

The purpose of these two studies was to extend thevalidity of the IBQ by testing the structure of the scalein a sport setting The results of Study 1 supported thatthe IBQ scale factor structure holds in cases whereathletes report on their perceptions of their coachesrsquointerpersonal behaviors The scale also demonstratedadequate validity and reliability and overall the inter-correlations between the IBQ subscales as well as thecorrelations between the IBQ subscales and the otherrelated factors provided support for the fit of the scalewithin the existing SDT measures and literature Interms of the relationships between the IBQ subscalesthe results supported there were moderately high posi-tive and negative correlations between the subscalesAdditionally the subscales also correlated with needsatisfaction need frustration and motivation in thedirections that would be expected according to SDTIn the case of some relationships the outcomes hadsimilar correlations with all three need-supportivebehaviors and all three need-thwarting behaviors butthis is not unexpected since the satisfaction or frustra-tion of the three basic needs are often related (Deci ampRyan 1985)

The results of Study 2 supported that the IBQ-Selfstructure held in a sample of coachesmdashreporting ontheir behaviors with their athletes Similar to Study 1the scale also demonstrated adequate validity and relia-bility Overall the inter-correlations between theIBQ-Self subscales and the correlations between theIBQ-Self subscales and other factors showed a similarpattern to Study 1mdashsupporting that the scale relates toother factors as would be expected

Limitations

These results support the applicability of the scale toa sport context however there are some limitationsFirst the structure of the scales was tested using astudent sample of athletes as well as basketball andtrack and field coaches The IBQ should be testedwith a more varied sample of athletes to ensure thescale is a valid measure of perceptions of coaches forother groups such as older athletes and recreationalathletes Furthermore the IBQ-Self should be vali-dated with coaches from a larger variety of sports

MEASUREMENT IN PHYSICAL EDUCATION AND EXERCISE SCIENCE 9

In this sample coaches were coaching team (basket-ball) and individual sports (athletics) however coa-ches can work with all sorts of different athletes atvarying levels of competition and the scalersquos validityshould be extended to include these coaches too Asecond limitation is that the IBQ and IBQ-Self onlyfocused on two people within the sport context ath-letes and coaches There are a number of differentsocial agents in the sport context that impact athletes(or coaches) including the athletesrsquo parents the otherathletes sport administrators etc and the validity ofthe scale should also be tested to examine perceptionsof their behaviors or their own reports of their beha-viors Finally since the IBQ and IBQ-Self were vali-dated using coachesrsquo and athletesrsquo self-reports theirreports should be triangulated by a third party (egvideo recording and coding of interpersonal beha-viors) to confirm reports on these measures corre-spond to their real world behaviors

Future directions

The IBQ and the IBQ-Self promote a number ofinitiatives for future research in sport motivationFirst research should continue to examine the psy-chometric properties of the IBQ For example ath-letes or coaches should be examined at multiple timepoints in order to establish the scalersquos testndashretestreliability and coachesrsquo scores should be corroboratedwith athletesrsquo scores in order to provide support forthe construct validity of the scale Next the IBQ canbe used to explore the role of all six types of need-supportive and need-thwarting interpersonal beha-viors in relation to how they impact athletesrsquo psycho-logical need satisfaction and frustration This willhelp extend the existing research in sport to movebeyond the influential role of AS and AT and focuson CS CT RS and RT interpersonal behaviors aswell Next the IBQ-Self can be used to identify andunderstand the antecedents of all six types of inter-personal behaviors according to SDT Specifically thescale should be used to explore the factors that influ-ence coachesrsquo behaviors with their athletes Finallythe IBQ and IBQ-Self should be used to explorehow coachesrsquo behaviors change over the course of agiven season and how this relates to athlete psycho-logical needs and their motivation for sport

Overall the IBQ and IBQ-Self are new instrumentsfor assessing perceptions or self-reports of all six typesof need-supportive and need-thwarting interpersonalbehaviors for coaches and athletes It is hoped theIBQ and the IBQ-Self will promote research directed

at assessing the behavior of others in sport and how itimpacts need satisfaction need frustration and motiva-tion for athletes

References

Amorose J amp Anderson-Butcher D (2007) Autonomy-supportive coaching and self-determined motivation inhigh school and college athletes A test of self-determina-tion theory Psychology of Sport and Exercise 8 654ndash670doi101016jpsychsport200611003

Bartholomew K Ntoumanis N Ryan R Bosch J ampThoslashgersen-Ntoumani C (2011) Self-determination the-ory and diminished functioning The role of interpersonalcontrol and psychological need thwarting Personality andSocial Psychology Bulletin 37 1459ndash1473 doi1011770146167211413125

Bartholomew K Ntoumanis N amp Thoslashgersen-Ntoumani C(2009) A review of controlling motivational strategies from aself-determination theory perspective Implications for sportscoaches International Review of Sport and ExercisePsychology 2 215ndash233 doi10108017509840903235330

Bartholomew K Ntoumanis N amp Thoslashrgenson-NtoumaniC (2010) The controlling interpersonal style in a coachingcontext Development and initial validation of a psycho-metric scale Journal of Sport amp Exercise Psychology 32193ndash216 Retrieved from httpwwwncbinlmnihgovpubmed20479478

Conroy D amp Coatsworth J (2007) Assessing autonomy-sup-portive coaching strategies in youth sport Psychology of Sportand Exercise 8 671ndash684 doi101016jpsychsport200612001

Deci E amp Ryan R (1985) Intrinsic motivation and self-determination in human behaviour In E Deci amp R Ryan(Eds) Handbook of self-determination research RochesterNY University of Rochester Press

Deci E amp Ryan R (2002) Handbook of self-determinationresearch Rochester NY The University of Rochester Press

Deci E Ryan R Gagneacute M Leone D Usunov J ampKornazheva B (2001) Need satisfaction motivation andwell-being in the work organizations of a former easternbloc country A cross-cultural study of self-determinationPersonality and Social Psychology Bulletin 27 930ndash942doi1011770146167201278002

Deci E Schwartz A Sheinman L amp Ryan R (1981) Aninstrument to assess adultsrsquo orientations toward control ver-sus autonomy with children Reflections on intrinsic motiva-tion and perceived competence Journal of EducationalPsychology 73 642ndash650 doi1010370022-0663735642

Field A (2009) Discovering statistics using SPSS Sage Publi-cations

Gould D Dieffenbach K amp Moffett A (2002) Psychologicalcharacteristics and their development in Olympic cham-pions Journal of Applied Sports Psychology 14 172ndash204doi10108010413200290103482

Hagger M amp Chatzisarantis N (2007) Intrinsic motivationand self-determination in exercise and sport Leeds UKHuman Kinetics Europe Ltd

Hair J Black W Babin B amp Anderson R (2010)Multivariate data analysis (7th ed) Upper Saddle RiverNJ Prentice-Hall Inc

10 M ROCCHI ET AL

CFI = 95 TLI = 93) The Cronbachrsquos alpha was calcu-lated for each subscale and revealed they were abovethe acceptable limit (α gt 77) except for the introjectedsubscale which had a lower internal consistency(α = 60) A mean score was calculated for autonomous(mean of intrinsic integrated and identified regula-tion) and controlled motivation (mean of introjectedexternal and amotivated regulation)

Analyses

First the data was cleaned and screened for univariateand multivariate outliers Next a confirmatory factoranalysis was performed on the IBQ to confirm thestructure of the scale held for a sample of athletesreporting on their coachesrsquo behaviors The model wasestimated using maximum likelihood robust (MLR)estimator (Muthen amp Muthen 2010) The fit of themodel was assessed according to Hu and Bentlerrsquos(1999) recommendation of using a scaled chi-square(χ2) and the SRMR as absolute fit indices the TLI asa relative fit index and finally the RMSEA and the CFIas noncentrality-based indices For the SRMR andRMSEA values below 08 indicated adequate modelfit and values below 06 indicated excellent fit for theCFI and TLI values above 90 represented good fit andvalues above 95 indicated excellent fit (HooperCoughlan amp Mullen 2008) Next invariance testswere conducted to confirm the scale performed equallyfor both male and female athletes Then internal con-sistency was examined using Cronbachrsquos alpha for eachsubscale and the discriminant and convergent validitywere examined using the average variance extracted(AVE) and the average shared square variance (ASV)Finally a series of outcome correlation analyses wereconducted to confirm the subscales related to the out-comes as expected according to SDT

Results

First the scoring distributions of the 24 IBQ items wereexamined for univariate normality and results sug-gested that similar to the original validation studies(Rocchi et al 2017) it had not been achieved for allvariables (skewness range minus1015 to 954 kurtosisrange minus655 to 541) Next missing data patternswere examined and it was revealed 24 participants(11) were missing between 1 and 3 observations ontheir reports of the IBQ Since this represented less than1 of the overall sample the missing data was esti-mated using full information ML (FIML) Then theunivariate and multivariate outlier analyses revealedtwo multivariate outliers and they were removed from

the subsequent analyses Finally the composite scoreswere calculated for the coachesrsquo interpersonal behaviors(see Table 1 for descriptive statistics) as well as forathletesrsquo need satisfaction need frustration and sportmotivation (see Table 2 for descriptive statistics)

Scale structure

The structure of the IBQ was tested through a CFA usingthe MLR estimator The results supported the factorialmodel had excellent fit (SBχ2(237) = 29623 p lt 001CFI = 95 TLI = 95 RMSEA = 05 [90 CI 04 06]SRMR= 05) and the standardized factor loadings for eachsubscale were larger than 60 (see Table 3) Next sinceboth groups achieved the minimum recommended sam-ple size of 100 a series of invariance tests were conductedto determine the scale performed equally for both men(N = 105) and women (N = 131) athletes (MacCallumWidaman Zhang amp Hong 1999 Meade 2005) In thefirst step the baseline models were tested for each genderto confirm the structure of the IBQ held for men (SBχ2

(237) = 31776 p lt 001 CFI = 90 TLI = 90 RMSEA = 07

Table 1 Study 1 and 2 Final items with standardized factorloadings (FL) and standard errors (SE)

Study 1 Study 2

Items FL SE FL SE

STEM ldquoMy coach rdquo (Study 1) ldquoWhen I am with my athletes rdquo (Study 2)ASGives me the freedom to make my ownchoices

791 043 712 041

Supports my decisions 872 032 722 036Supports the choices that I make for myself 828 045 681 031Encourages me to make my own decisions 813 045 731 035

ATPressures me to do things their way 699 051 722 030Imposes their opinions on me 794 045 741 029Pressures me to adopt certain behaviors 756 048 693 039Limits my choices 756 051 687 103

CSEncourages me to improve my skills 693 066 658 045Provides valuable feedback 739 052 720 041Acknowledges my ability to achieve my goals 811 043 765 032Tells me that I can accomplish things 859 035 748 033

CTPoints out that I will likely fail 739 054 761 024Sends me the message that I amincompetent

830 039 652 044

Doubts my capacity to improve 741 051 674 030Questions my ability to overcome challenges 819 049 748 037

RSIs interested in what I do 804 054 715 050Takes the time to get to know me 830 039 755 032Honestly enjoy spending time with me 741 051 734 045Relates to me 819 049 633 033

RTDoes not comfort me when I am feeling low 804 054 443 028Is distant when we spend time together 862 028 529 031Does not connect with me 832 031 656 046Does not care about me 855 027 618 055

Note Study 1 N = 237 Study 2 N = 240 The verb tenses and pronouns of theitems weremodified in the IBQ-Self to reflect the new stem For example ldquoGivethem the freedom to make their own choicesrdquo instead of ldquoGives me thefreedom to make my own choicesrdquo or ldquoDo not comfort them when they arefeeling lowrdquo instead of ldquoDoes not comfort me when i am feeling lowrdquo

MEASUREMENT IN PHYSICAL EDUCATION AND EXERCISE SCIENCE 5

[90 CI 06 08] SRMR = 06) and women (SBχ2

(237) = 33495 p lt 001 CFI = 92 TLI = 91RMSEA = 07 [90 CI 06 08] SRMR = 06) Thenconfigural invariance was established by testing a CFAmodel with both groups where no constraints were placedon the parameters The results suggested the factor struc-ture of the IBQ was the same for men and women athletes(SBχ2(474) = 67240 p lt 001 CFI = 91 TLI = 90RMSEA = 07 [90 CI 06 08] SRMR = 06) Nextmetric invariance was established by constraining thefactor indicators to be equal for groups and the resultsfound the data fit the model well (SBχ2(492) = 68120p lt 001 CFI = 91 TLI = 90 RMSEA = 07 [90CI 06 08] SRMR = 07) and the change in chi-squarebetween the constrained and unconstrained model wasnot significant (SBΔχ2(18) = 1199 p gt 05) Finally scalarinvariance was established by constraining the interceptsto be equal for both groups Again the results supportedthe model had good fit (SBχ2(510) = 70916 p lt 001

CFI = 91 TLI = 90 RMSEA = 07 [90 CI 06 08]SRMR = 07) and the chi-square difference test confirmedthe model fit remained stable (SBΔχ2(18) = 2761 p gt 05)Overall these results support men and women athleteswere invariant with regards to the factor structure andcomposite variables can be created for both groups

Validity and reliability

The AVE and ASV were examined to assess conver-gent and divergent validity (see Table 1) All sub-scales met the requirements for convergent validity asthe AVEs were above 5 (Hair Black Babin ampAnderson 2010) Discriminate validity was alsoachieved since all ASV values were smaller thantheir respective AVEs (Hair et al 2010) The factorcorrelations and internal reliabilities for each subscalesupport each subscale has achieved acceptable inter-nal consistency (see Table 1)

Table 2 Study 1 and 2 Factor correlations variance means standard deviations and internal consistencySubscales 1 2 3 4 5 6 AVE ASV M SD

Study 1 (IBQmdashGeneral Stem)1 AS ndash89 68 28 503 1392 AT minus39 ndash84 57 15 323 1493 CS 70 minus30 ndash87 61 18 554 1284 CT minus34 69 minus47 ndash87 62 18 137 1375 RS 71 minus32 74 minus36 ndash90 64 19 498 1446 RT minus40 60 minus37 68 minus50 ndash87 70 17 241 141

Study 2 (IBQ-SelfmdashAthlete Stem)1 AS ndash71 51 07 601 0832 AT minus36 ndash70 51 05 211 2113 CS 40 minus14 ndash73 53 09 566 0594 CT minus13 42 04 ndash75 52 05 140 0765 RS 44 minus18 65 minus05 ndash72 50 11 600 0656 RT minus17 16 minus27 11 minus27 ndash40 33 03 141 073

Note Study 1 N = 237 Study 2 N = 240 α = diagonals AVE = average variance extracted ASV = average shared square variancep lt 05 p lt 001

Table 3 Study 1 and 2 Descriptive statistics and correlations with outcome variables for the IBQ (Study1) and the IBQ-Self (Study 2)

IBQ Subscales

Need Satisfaction Need Frustration Motivation

Soc DesAut Comp Rel Aut Comp Rel Aut Ctl

Study 1AS 58 48 46 minus28 minus32 minus24 40 minus08 mdashAT minus33 minus20 minus27 63 56 54 minus08 45 mdashCS 43 52 48 minus30 minus40 minus34 39 minus17 mdashCT minus30 minus23 minus33 64 66 63 minus14 51 mdashRS 52 50 49 minus26 minus32 minus25 37 minus06 mdashRT minus44 minus29 minus39 62 58 64 minus25 41 mdashMean 519 570 582 304 286 254 550 326 mdashStandard Deviation 114 104 103 114 154 103 116 112 mdash

Study 2AS 20 23 23 minus02 minus02 minus01 25 02 10AT minus11 05 02 23 30 24 11 36 06CS 31 48 34 minus18 minus11 minus06 28 minus01 10CT 01 06 minus03 07 28 26 06 minus30 08RS 36 53 37 minus16 minus12 minus12 29 05 06RT minus04 minus06 minus02 01 22 35 12 minus19 11Mean 611 504 132 173 160 167 505 229 453Standard Deviation 092 079 088 106 100 078 120 097 257

Note Study 1 N = 237 Study 2 N = 240 Soc Des = social desirability For Study 2 correlations are while controlling for social desirability except for thecorrelations between the subscales of the IBQ and social desirability Need satisfaction and frustration are related to sport or coaching Motivation is forparticipating in sport or coaching

p lt 05 p lt 001

6 M ROCCHI ET AL

Outcome correlations

The IBQ subscales were correlated with need satisfaction insport need frustration in sport and sport motivation aspredicted (see Table 3) Specifically the need-supportivesubscales of the IBQ correlated positively with need satis-faction and negatively with need frustration The need-thwarting subscales of the IBQ correlated negatively withneed satisfaction and positively with need frustrationLooking at athletesrsquomotivation for sport the need-suppor-tive subscales correlated positively with autonomous moti-vation and the need-thwarting subscales correlated withcontrolled motivation For CT it was related to an increasein athletesrsquo autonomous motivation and a decrease in theircontrolled motivation For both perceptions of CT and RTcoach behaviors this was related to decreases in autono-mous motivation for athletes as well as an increase incontrolled motivation

Discussion

Overall the results of Study 1 support the psychometricproperties of the IBQ in a sample of athletes reportingon their coachesrsquo interpersonal behaviors and the scaleperformed equally for male and female athletes This isthe first set of results to support the IBQ can be used toassess perceptions of interpersonal behaviors in thesport context In this sample the structure of the scaleheld as well as the reliabilities and outcome correla-tions supporting the scale performed as would beexpected similar to the original validation studies(Rocchi et al 2017) Although there are relativelyhigh relationships between the need-supportive andneed-thwarting subscales that are in line with theresults of the original validation studies the relation-ships are not high enough to suggest multicollinearity(lt 80) may be an issue (Field 2009) The results of thisstudy support the validity of the IBQ to assess athletesrsquoperceptions of coachesrsquo behaviors Specifically theseresults are in line with what would be expected accord-ing to SDT as athletes who report their coaches useneed-supportive interpersonal behaviors reportedincreased need satisfaction and autonomous motivationin sport (Mageau amp Vallerand 2003) while athleteswho reported their coaches use need-thwarting inter-personal behaviors reported need frustration and con-trolled motivation in sport (Bartholomew et al 2009)

Study 2

The objective of Study 2 is to evaluate the psychometricproperties of the IBQ self-report version within the

sport context Specifically this will be validated usinga sample of sport coaches who are reporting on theirbehaviors when they interact with their athletes It isanticipated the factor structure of the IBQ-Self will holdfor this sample in this context and it will correlate withoutcomes as would be expected according to SDT

Method

Participants

The sample was composed of 240 coaches (Nmale = 170Nfemale = 66 Nmissing = 4) with an average age of4701 years (SD = 1031) The large majority of thecoaches had a college (N = 53 22) or universityeducation (N = 168 70) They were either basketball(N = 111 46) or track and field coaches (N = 12954) and they had an average of 1750 years of coach-ing experience (SD = 1283) The majority (N = 12251) identified as a head coach or as an assistant coach(N = 60 25) and indicated their athletes were very(N = 94 39) or extremely (N = 61 25) competitivecompared to other athletes of their age

Procedures

Coaches were recruited through their respective pro-vincial sporting associations (basketball or athletics) toparticipate in an online survey Each organization sentan email to all coaches inviting them to participate inan online study The coachesrsquo participation was volun-tary and they gave their informed consent before parti-cipating To be eligible the coaches had to be registeredwith a provincial sport organization at the time of datacollection and actively coaching

Materials

The following measures were completed by the coachesthrough an online questionnaire

My interpersonal behaviorsThe coaches completed the IBQ-Self to assess the extentto which they believed they engaged in AS AT CS CTRS and RT interpersonal behaviors with their athletes(Rocchi et al 2017) Coaches were presented each itemusing the stem ldquoWhen I am with my athletes I rdquo andindicated their agreement with each statement using a7-point scale ranging from 1 (do not agree at all) to 7(completely agree) The original validation studies sup-ported the scale had sufficient internal reliabil-ity (α gt 80)

MEASUREMENT IN PHYSICAL EDUCATION AND EXERCISE SCIENCE 7

Need satisfactionThe coaches responded to the positive items from theBasic Need Satisfaction at Work Scale adapted to thesport context (Deci et al 2001) to assess the extenttheir three basic psychological needs were met whilecoaching Like in Study 1 the stems were modified toask the coaches about their need satisfaction whilecoaching instead of at work The fit indices suggestthe model has an acceptable fit except for the TLI(SBχ2(41) = 3836 p lt 001 RMSEA = 06 [90 CI03 08] SRMR = 05 CFI = 90 TLI = 89) howeversince the other fit indices were acceptable the scale wasused The internal consistency estimates were withinacceptable range (autonomy α = 83 competenceα = 71 relatedness α = 74) and a mean score wascalculated for each subscale

Need frustrationThe coaches also completed the Psychological NeedThwarting Scale in Sport (Bartholomew et al 2010) toassess the extent to which their needs were frustratedwhile coaching like in Study 1 A confirmatory factoranalysis confirmed the structure of the scale for coachesand the results supported that the model had an excel-lent fit (SBχ2(32) = 2600 p lt 001 RMSEA = 04 [90CI 00 07] SRMR = 06 CFI = 96 TLI = 94) Theinternal consistency for the three subscales were abovethe minimum (α gt 74) and a mean score was calcu-lated for each subscale to represent the coachesrsquo frus-tration of each need

Coach motivationThe coaches completed the Coach MotivationQuestionnaire (CMQ McLean Mallet amp Newcombe2012) to assess their reasons for participating in sportThe CMQ is a 6-factor scale comprised of 22 itemsmeasuring sport motivation according to each of thesix types of behavioral regulation according to SDT Aconfirmatory factor confirmed the structure of the scalein this sample (SBχ2(194) = 33657 p lt 001RMSEA = 06 [90 CI 04 07] SRMR = 07CFI = 91 TLI = 90) The Cronbachrsquos alphas werecalculated for each subscale and revealed they wereabove the acceptable limit (α gt 72) except for theintrojected subscale which had a lower internal consis-tency (α = 68) Like in Study 1 a mean score wascalculated for autonomous motivation and controlledmotivation

Social desirabilityFinally the coaches also completed the short form ofthe Marlowe-Crowne Social Desirability Scale(Reynolds 1982) to control for whether they were

responding to survey items based on providing favor-able responses The short form of this measure consistsof 10 items where participants are asked to indicatewhether these statements are true or false for themThe sum of all responses considered to be sociallydesirable is calculated to provide an overall measureof social desirability

Analyses

Using the same procedures as the first study the datawas cleaned and then the scale structure was examinedThen the validity and reliability of the subscales wereexamined Finally outcome correlations were con-ducted to examine the relationship between coachneed satisfaction frustration and motivation whilecontrolling for social desirability

Results

Similar to the first study the descriptive statistics analyseson the 24 IBQ-Self items suggested the variables did notall have a normal distribution (skewness range minus1325 to912 kurtosis range minus620 to 821) Data screening ana-lyses did not reveal any multivariate outliers as such theentire sample was retained for the full analyses Themissing data analyses suggested 15 participants weremissing one or two observations on the IBQ-Self how-ever since these missing observations made up less than5 of the total observations the data was imputed usingthe same methods as Study 1 The composite scores forthe coachesrsquo reported use of need-supportive andneed-thwarting behaviors as well as their need satisfac-tion and frustration in coaching their motivation towardscoaching and their likelihood of responding in sociallydesirable ways was calculated (see Tables 1 and 2 for thedescriptive statistics)

Scale structure

The results supported that the factorial model had goodfit (SBχ2(237) = 30304 p lt 001 CFI = 93 TLI = 92RMSEA = 06 [90 CI 03 05] SRMR = 04) Thestandardized factor loadings for each subscale werelarger than 46 (see Table 1) and the factor correlationsas well as the internal reliabilities for each subscale(Cronbachrsquos alpha) are in Table 1 Since there wereonly 66 female coaches in this sample gender invar-iance tests were not performed in this study Nextconvergent and divergent validity was examined foreach of the subscales (see Table 2) and the resultssupported that the scale met the thresholds for bothacross all subscales except for the RT subscale Internal

8 M ROCCHI ET AL

consistency analyses revealed all of the alphas were overthe minimum criteria again except for the RT subscale(α = 40) which was nowhere near the baseline criteriaThe internal consistency for that subscale was calcu-lated using all of the potential combinations of items tosee if one specific item was causing the issue and theresults found there was no combination of the fouritems that promoted a better internal consistency Asa result the reliability for the four items was retained

Outcome correlations

The IBQ-Self subscales were correlated with need satis-faction need frustration autonomous and controlledmotivation while controlling for social desirability(see Table 2) to confirm the relationships were in theexpected directions The results found that the supportsubscales of the IBQ-Self correlated moderately andpositively with need satisfaction and autonomous moti-vation while they had weak or negative relationshipswith need frustration and controlled motivation Theopposite was found with the need-thwarting subscaleswhere there were moderate positive relationships withneed frustration and controlled motivation and weaknegative or no relationships with need satisfaction andautonomous motivation Overall the results were simi-lar to those in Study 1 except the strength of therelationships was weaker

Discussion

The results of this study support the factor structureand validity of the IBQ-Self as a questionnaire forcoaches about their use of interpersonal behaviorsaccording to SDT in their interactions with theirathletes Unfortunately the reliability for the RT sub-scale was weak Since some coaches and athletes maychoose to engage in the relationship when theychoose to participate in the sport (ie select athletefor a team) and others do not (ie everyone trainsduring the same session) it is not totally surprisingthat coaches care about their athletes and this madethe results somewhat ambiguous Perhaps in anothercontext where there is either less control over beingin a relationship with the other people such as afamily member or a supervisor at work or morecontrol like a best friend or partner that subscaleinternal consistency would improve Overall theseresults support that coaches who report using moreneed-supportive behaviors also report higher needsatisfaction and autonomous motivation for theircoaching while coaches who report more need-thwarting behaviors report more need frustration

and controlled motivation for coachingmdashall aftercontrolling for social desirability The results of thisstudy support that generally the IBQ-Self is a validmeasure of coachesrsquo own reports of their interperso-nal behaviors when interacting with their athletes

General discussion

The purpose of these two studies was to extend thevalidity of the IBQ by testing the structure of the scalein a sport setting The results of Study 1 supported thatthe IBQ scale factor structure holds in cases whereathletes report on their perceptions of their coachesrsquointerpersonal behaviors The scale also demonstratedadequate validity and reliability and overall the inter-correlations between the IBQ subscales as well as thecorrelations between the IBQ subscales and the otherrelated factors provided support for the fit of the scalewithin the existing SDT measures and literature Interms of the relationships between the IBQ subscalesthe results supported there were moderately high posi-tive and negative correlations between the subscalesAdditionally the subscales also correlated with needsatisfaction need frustration and motivation in thedirections that would be expected according to SDTIn the case of some relationships the outcomes hadsimilar correlations with all three need-supportivebehaviors and all three need-thwarting behaviors butthis is not unexpected since the satisfaction or frustra-tion of the three basic needs are often related (Deci ampRyan 1985)

The results of Study 2 supported that the IBQ-Selfstructure held in a sample of coachesmdashreporting ontheir behaviors with their athletes Similar to Study 1the scale also demonstrated adequate validity and relia-bility Overall the inter-correlations between theIBQ-Self subscales and the correlations between theIBQ-Self subscales and other factors showed a similarpattern to Study 1mdashsupporting that the scale relates toother factors as would be expected

Limitations

These results support the applicability of the scale toa sport context however there are some limitationsFirst the structure of the scales was tested using astudent sample of athletes as well as basketball andtrack and field coaches The IBQ should be testedwith a more varied sample of athletes to ensure thescale is a valid measure of perceptions of coaches forother groups such as older athletes and recreationalathletes Furthermore the IBQ-Self should be vali-dated with coaches from a larger variety of sports

MEASUREMENT IN PHYSICAL EDUCATION AND EXERCISE SCIENCE 9

In this sample coaches were coaching team (basket-ball) and individual sports (athletics) however coa-ches can work with all sorts of different athletes atvarying levels of competition and the scalersquos validityshould be extended to include these coaches too Asecond limitation is that the IBQ and IBQ-Self onlyfocused on two people within the sport context ath-letes and coaches There are a number of differentsocial agents in the sport context that impact athletes(or coaches) including the athletesrsquo parents the otherathletes sport administrators etc and the validity ofthe scale should also be tested to examine perceptionsof their behaviors or their own reports of their beha-viors Finally since the IBQ and IBQ-Self were vali-dated using coachesrsquo and athletesrsquo self-reports theirreports should be triangulated by a third party (egvideo recording and coding of interpersonal beha-viors) to confirm reports on these measures corre-spond to their real world behaviors

Future directions

The IBQ and the IBQ-Self promote a number ofinitiatives for future research in sport motivationFirst research should continue to examine the psy-chometric properties of the IBQ For example ath-letes or coaches should be examined at multiple timepoints in order to establish the scalersquos testndashretestreliability and coachesrsquo scores should be corroboratedwith athletesrsquo scores in order to provide support forthe construct validity of the scale Next the IBQ canbe used to explore the role of all six types of need-supportive and need-thwarting interpersonal beha-viors in relation to how they impact athletesrsquo psycho-logical need satisfaction and frustration This willhelp extend the existing research in sport to movebeyond the influential role of AS and AT and focuson CS CT RS and RT interpersonal behaviors aswell Next the IBQ-Self can be used to identify andunderstand the antecedents of all six types of inter-personal behaviors according to SDT Specifically thescale should be used to explore the factors that influ-ence coachesrsquo behaviors with their athletes Finallythe IBQ and IBQ-Self should be used to explorehow coachesrsquo behaviors change over the course of agiven season and how this relates to athlete psycho-logical needs and their motivation for sport

Overall the IBQ and IBQ-Self are new instrumentsfor assessing perceptions or self-reports of all six typesof need-supportive and need-thwarting interpersonalbehaviors for coaches and athletes It is hoped theIBQ and the IBQ-Self will promote research directed

at assessing the behavior of others in sport and how itimpacts need satisfaction need frustration and motiva-tion for athletes

References

Amorose J amp Anderson-Butcher D (2007) Autonomy-supportive coaching and self-determined motivation inhigh school and college athletes A test of self-determina-tion theory Psychology of Sport and Exercise 8 654ndash670doi101016jpsychsport200611003

Bartholomew K Ntoumanis N Ryan R Bosch J ampThoslashgersen-Ntoumani C (2011) Self-determination the-ory and diminished functioning The role of interpersonalcontrol and psychological need thwarting Personality andSocial Psychology Bulletin 37 1459ndash1473 doi1011770146167211413125

Bartholomew K Ntoumanis N amp Thoslashgersen-Ntoumani C(2009) A review of controlling motivational strategies from aself-determination theory perspective Implications for sportscoaches International Review of Sport and ExercisePsychology 2 215ndash233 doi10108017509840903235330

Bartholomew K Ntoumanis N amp Thoslashrgenson-NtoumaniC (2010) The controlling interpersonal style in a coachingcontext Development and initial validation of a psycho-metric scale Journal of Sport amp Exercise Psychology 32193ndash216 Retrieved from httpwwwncbinlmnihgovpubmed20479478

Conroy D amp Coatsworth J (2007) Assessing autonomy-sup-portive coaching strategies in youth sport Psychology of Sportand Exercise 8 671ndash684 doi101016jpsychsport200612001

Deci E amp Ryan R (1985) Intrinsic motivation and self-determination in human behaviour In E Deci amp R Ryan(Eds) Handbook of self-determination research RochesterNY University of Rochester Press

Deci E amp Ryan R (2002) Handbook of self-determinationresearch Rochester NY The University of Rochester Press

Deci E Ryan R Gagneacute M Leone D Usunov J ampKornazheva B (2001) Need satisfaction motivation andwell-being in the work organizations of a former easternbloc country A cross-cultural study of self-determinationPersonality and Social Psychology Bulletin 27 930ndash942doi1011770146167201278002

Deci E Schwartz A Sheinman L amp Ryan R (1981) Aninstrument to assess adultsrsquo orientations toward control ver-sus autonomy with children Reflections on intrinsic motiva-tion and perceived competence Journal of EducationalPsychology 73 642ndash650 doi1010370022-0663735642

Field A (2009) Discovering statistics using SPSS Sage Publi-cations

Gould D Dieffenbach K amp Moffett A (2002) Psychologicalcharacteristics and their development in Olympic cham-pions Journal of Applied Sports Psychology 14 172ndash204doi10108010413200290103482

Hagger M amp Chatzisarantis N (2007) Intrinsic motivationand self-determination in exercise and sport Leeds UKHuman Kinetics Europe Ltd

Hair J Black W Babin B amp Anderson R (2010)Multivariate data analysis (7th ed) Upper Saddle RiverNJ Prentice-Hall Inc

10 M ROCCHI ET AL

[90 CI 06 08] SRMR = 06) and women (SBχ2

(237) = 33495 p lt 001 CFI = 92 TLI = 91RMSEA = 07 [90 CI 06 08] SRMR = 06) Thenconfigural invariance was established by testing a CFAmodel with both groups where no constraints were placedon the parameters The results suggested the factor struc-ture of the IBQ was the same for men and women athletes(SBχ2(474) = 67240 p lt 001 CFI = 91 TLI = 90RMSEA = 07 [90 CI 06 08] SRMR = 06) Nextmetric invariance was established by constraining thefactor indicators to be equal for groups and the resultsfound the data fit the model well (SBχ2(492) = 68120p lt 001 CFI = 91 TLI = 90 RMSEA = 07 [90CI 06 08] SRMR = 07) and the change in chi-squarebetween the constrained and unconstrained model wasnot significant (SBΔχ2(18) = 1199 p gt 05) Finally scalarinvariance was established by constraining the interceptsto be equal for both groups Again the results supportedthe model had good fit (SBχ2(510) = 70916 p lt 001

CFI = 91 TLI = 90 RMSEA = 07 [90 CI 06 08]SRMR = 07) and the chi-square difference test confirmedthe model fit remained stable (SBΔχ2(18) = 2761 p gt 05)Overall these results support men and women athleteswere invariant with regards to the factor structure andcomposite variables can be created for both groups

Validity and reliability

The AVE and ASV were examined to assess conver-gent and divergent validity (see Table 1) All sub-scales met the requirements for convergent validity asthe AVEs were above 5 (Hair Black Babin ampAnderson 2010) Discriminate validity was alsoachieved since all ASV values were smaller thantheir respective AVEs (Hair et al 2010) The factorcorrelations and internal reliabilities for each subscalesupport each subscale has achieved acceptable inter-nal consistency (see Table 1)

Table 2 Study 1 and 2 Factor correlations variance means standard deviations and internal consistencySubscales 1 2 3 4 5 6 AVE ASV M SD

Study 1 (IBQmdashGeneral Stem)1 AS ndash89 68 28 503 1392 AT minus39 ndash84 57 15 323 1493 CS 70 minus30 ndash87 61 18 554 1284 CT minus34 69 minus47 ndash87 62 18 137 1375 RS 71 minus32 74 minus36 ndash90 64 19 498 1446 RT minus40 60 minus37 68 minus50 ndash87 70 17 241 141

Study 2 (IBQ-SelfmdashAthlete Stem)1 AS ndash71 51 07 601 0832 AT minus36 ndash70 51 05 211 2113 CS 40 minus14 ndash73 53 09 566 0594 CT minus13 42 04 ndash75 52 05 140 0765 RS 44 minus18 65 minus05 ndash72 50 11 600 0656 RT minus17 16 minus27 11 minus27 ndash40 33 03 141 073

Note Study 1 N = 237 Study 2 N = 240 α = diagonals AVE = average variance extracted ASV = average shared square variancep lt 05 p lt 001

Table 3 Study 1 and 2 Descriptive statistics and correlations with outcome variables for the IBQ (Study1) and the IBQ-Self (Study 2)

IBQ Subscales

Need Satisfaction Need Frustration Motivation

Soc DesAut Comp Rel Aut Comp Rel Aut Ctl

Study 1AS 58 48 46 minus28 minus32 minus24 40 minus08 mdashAT minus33 minus20 minus27 63 56 54 minus08 45 mdashCS 43 52 48 minus30 minus40 minus34 39 minus17 mdashCT minus30 minus23 minus33 64 66 63 minus14 51 mdashRS 52 50 49 minus26 minus32 minus25 37 minus06 mdashRT minus44 minus29 minus39 62 58 64 minus25 41 mdashMean 519 570 582 304 286 254 550 326 mdashStandard Deviation 114 104 103 114 154 103 116 112 mdash

Study 2AS 20 23 23 minus02 minus02 minus01 25 02 10AT minus11 05 02 23 30 24 11 36 06CS 31 48 34 minus18 minus11 minus06 28 minus01 10CT 01 06 minus03 07 28 26 06 minus30 08RS 36 53 37 minus16 minus12 minus12 29 05 06RT minus04 minus06 minus02 01 22 35 12 minus19 11Mean 611 504 132 173 160 167 505 229 453Standard Deviation 092 079 088 106 100 078 120 097 257

Note Study 1 N = 237 Study 2 N = 240 Soc Des = social desirability For Study 2 correlations are while controlling for social desirability except for thecorrelations between the subscales of the IBQ and social desirability Need satisfaction and frustration are related to sport or coaching Motivation is forparticipating in sport or coaching

p lt 05 p lt 001

6 M ROCCHI ET AL

Outcome correlations

The IBQ subscales were correlated with need satisfaction insport need frustration in sport and sport motivation aspredicted (see Table 3) Specifically the need-supportivesubscales of the IBQ correlated positively with need satis-faction and negatively with need frustration The need-thwarting subscales of the IBQ correlated negatively withneed satisfaction and positively with need frustrationLooking at athletesrsquomotivation for sport the need-suppor-tive subscales correlated positively with autonomous moti-vation and the need-thwarting subscales correlated withcontrolled motivation For CT it was related to an increasein athletesrsquo autonomous motivation and a decrease in theircontrolled motivation For both perceptions of CT and RTcoach behaviors this was related to decreases in autono-mous motivation for athletes as well as an increase incontrolled motivation

Discussion

Overall the results of Study 1 support the psychometricproperties of the IBQ in a sample of athletes reportingon their coachesrsquo interpersonal behaviors and the scaleperformed equally for male and female athletes This isthe first set of results to support the IBQ can be used toassess perceptions of interpersonal behaviors in thesport context In this sample the structure of the scaleheld as well as the reliabilities and outcome correla-tions supporting the scale performed as would beexpected similar to the original validation studies(Rocchi et al 2017) Although there are relativelyhigh relationships between the need-supportive andneed-thwarting subscales that are in line with theresults of the original validation studies the relation-ships are not high enough to suggest multicollinearity(lt 80) may be an issue (Field 2009) The results of thisstudy support the validity of the IBQ to assess athletesrsquoperceptions of coachesrsquo behaviors Specifically theseresults are in line with what would be expected accord-ing to SDT as athletes who report their coaches useneed-supportive interpersonal behaviors reportedincreased need satisfaction and autonomous motivationin sport (Mageau amp Vallerand 2003) while athleteswho reported their coaches use need-thwarting inter-personal behaviors reported need frustration and con-trolled motivation in sport (Bartholomew et al 2009)

Study 2

The objective of Study 2 is to evaluate the psychometricproperties of the IBQ self-report version within the

sport context Specifically this will be validated usinga sample of sport coaches who are reporting on theirbehaviors when they interact with their athletes It isanticipated the factor structure of the IBQ-Self will holdfor this sample in this context and it will correlate withoutcomes as would be expected according to SDT

Method

Participants

The sample was composed of 240 coaches (Nmale = 170Nfemale = 66 Nmissing = 4) with an average age of4701 years (SD = 1031) The large majority of thecoaches had a college (N = 53 22) or universityeducation (N = 168 70) They were either basketball(N = 111 46) or track and field coaches (N = 12954) and they had an average of 1750 years of coach-ing experience (SD = 1283) The majority (N = 12251) identified as a head coach or as an assistant coach(N = 60 25) and indicated their athletes were very(N = 94 39) or extremely (N = 61 25) competitivecompared to other athletes of their age

Procedures

Coaches were recruited through their respective pro-vincial sporting associations (basketball or athletics) toparticipate in an online survey Each organization sentan email to all coaches inviting them to participate inan online study The coachesrsquo participation was volun-tary and they gave their informed consent before parti-cipating To be eligible the coaches had to be registeredwith a provincial sport organization at the time of datacollection and actively coaching

Materials

The following measures were completed by the coachesthrough an online questionnaire

My interpersonal behaviorsThe coaches completed the IBQ-Self to assess the extentto which they believed they engaged in AS AT CS CTRS and RT interpersonal behaviors with their athletes(Rocchi et al 2017) Coaches were presented each itemusing the stem ldquoWhen I am with my athletes I rdquo andindicated their agreement with each statement using a7-point scale ranging from 1 (do not agree at all) to 7(completely agree) The original validation studies sup-ported the scale had sufficient internal reliabil-ity (α gt 80)

MEASUREMENT IN PHYSICAL EDUCATION AND EXERCISE SCIENCE 7

Need satisfactionThe coaches responded to the positive items from theBasic Need Satisfaction at Work Scale adapted to thesport context (Deci et al 2001) to assess the extenttheir three basic psychological needs were met whilecoaching Like in Study 1 the stems were modified toask the coaches about their need satisfaction whilecoaching instead of at work The fit indices suggestthe model has an acceptable fit except for the TLI(SBχ2(41) = 3836 p lt 001 RMSEA = 06 [90 CI03 08] SRMR = 05 CFI = 90 TLI = 89) howeversince the other fit indices were acceptable the scale wasused The internal consistency estimates were withinacceptable range (autonomy α = 83 competenceα = 71 relatedness α = 74) and a mean score wascalculated for each subscale

Need frustrationThe coaches also completed the Psychological NeedThwarting Scale in Sport (Bartholomew et al 2010) toassess the extent to which their needs were frustratedwhile coaching like in Study 1 A confirmatory factoranalysis confirmed the structure of the scale for coachesand the results supported that the model had an excel-lent fit (SBχ2(32) = 2600 p lt 001 RMSEA = 04 [90CI 00 07] SRMR = 06 CFI = 96 TLI = 94) Theinternal consistency for the three subscales were abovethe minimum (α gt 74) and a mean score was calcu-lated for each subscale to represent the coachesrsquo frus-tration of each need

Coach motivationThe coaches completed the Coach MotivationQuestionnaire (CMQ McLean Mallet amp Newcombe2012) to assess their reasons for participating in sportThe CMQ is a 6-factor scale comprised of 22 itemsmeasuring sport motivation according to each of thesix types of behavioral regulation according to SDT Aconfirmatory factor confirmed the structure of the scalein this sample (SBχ2(194) = 33657 p lt 001RMSEA = 06 [90 CI 04 07] SRMR = 07CFI = 91 TLI = 90) The Cronbachrsquos alphas werecalculated for each subscale and revealed they wereabove the acceptable limit (α gt 72) except for theintrojected subscale which had a lower internal consis-tency (α = 68) Like in Study 1 a mean score wascalculated for autonomous motivation and controlledmotivation

Social desirabilityFinally the coaches also completed the short form ofthe Marlowe-Crowne Social Desirability Scale(Reynolds 1982) to control for whether they were

responding to survey items based on providing favor-able responses The short form of this measure consistsof 10 items where participants are asked to indicatewhether these statements are true or false for themThe sum of all responses considered to be sociallydesirable is calculated to provide an overall measureof social desirability

Analyses

Using the same procedures as the first study the datawas cleaned and then the scale structure was examinedThen the validity and reliability of the subscales wereexamined Finally outcome correlations were con-ducted to examine the relationship between coachneed satisfaction frustration and motivation whilecontrolling for social desirability

Results

Similar to the first study the descriptive statistics analyseson the 24 IBQ-Self items suggested the variables did notall have a normal distribution (skewness range minus1325 to912 kurtosis range minus620 to 821) Data screening ana-lyses did not reveal any multivariate outliers as such theentire sample was retained for the full analyses Themissing data analyses suggested 15 participants weremissing one or two observations on the IBQ-Self how-ever since these missing observations made up less than5 of the total observations the data was imputed usingthe same methods as Study 1 The composite scores forthe coachesrsquo reported use of need-supportive andneed-thwarting behaviors as well as their need satisfac-tion and frustration in coaching their motivation towardscoaching and their likelihood of responding in sociallydesirable ways was calculated (see Tables 1 and 2 for thedescriptive statistics)

Scale structure

The results supported that the factorial model had goodfit (SBχ2(237) = 30304 p lt 001 CFI = 93 TLI = 92RMSEA = 06 [90 CI 03 05] SRMR = 04) Thestandardized factor loadings for each subscale werelarger than 46 (see Table 1) and the factor correlationsas well as the internal reliabilities for each subscale(Cronbachrsquos alpha) are in Table 1 Since there wereonly 66 female coaches in this sample gender invar-iance tests were not performed in this study Nextconvergent and divergent validity was examined foreach of the subscales (see Table 2) and the resultssupported that the scale met the thresholds for bothacross all subscales except for the RT subscale Internal

8 M ROCCHI ET AL

consistency analyses revealed all of the alphas were overthe minimum criteria again except for the RT subscale(α = 40) which was nowhere near the baseline criteriaThe internal consistency for that subscale was calcu-lated using all of the potential combinations of items tosee if one specific item was causing the issue and theresults found there was no combination of the fouritems that promoted a better internal consistency Asa result the reliability for the four items was retained

Outcome correlations

The IBQ-Self subscales were correlated with need satis-faction need frustration autonomous and controlledmotivation while controlling for social desirability(see Table 2) to confirm the relationships were in theexpected directions The results found that the supportsubscales of the IBQ-Self correlated moderately andpositively with need satisfaction and autonomous moti-vation while they had weak or negative relationshipswith need frustration and controlled motivation Theopposite was found with the need-thwarting subscaleswhere there were moderate positive relationships withneed frustration and controlled motivation and weaknegative or no relationships with need satisfaction andautonomous motivation Overall the results were simi-lar to those in Study 1 except the strength of therelationships was weaker

Discussion

The results of this study support the factor structureand validity of the IBQ-Self as a questionnaire forcoaches about their use of interpersonal behaviorsaccording to SDT in their interactions with theirathletes Unfortunately the reliability for the RT sub-scale was weak Since some coaches and athletes maychoose to engage in the relationship when theychoose to participate in the sport (ie select athletefor a team) and others do not (ie everyone trainsduring the same session) it is not totally surprisingthat coaches care about their athletes and this madethe results somewhat ambiguous Perhaps in anothercontext where there is either less control over beingin a relationship with the other people such as afamily member or a supervisor at work or morecontrol like a best friend or partner that subscaleinternal consistency would improve Overall theseresults support that coaches who report using moreneed-supportive behaviors also report higher needsatisfaction and autonomous motivation for theircoaching while coaches who report more need-thwarting behaviors report more need frustration

and controlled motivation for coachingmdashall aftercontrolling for social desirability The results of thisstudy support that generally the IBQ-Self is a validmeasure of coachesrsquo own reports of their interperso-nal behaviors when interacting with their athletes

General discussion

The purpose of these two studies was to extend thevalidity of the IBQ by testing the structure of the scalein a sport setting The results of Study 1 supported thatthe IBQ scale factor structure holds in cases whereathletes report on their perceptions of their coachesrsquointerpersonal behaviors The scale also demonstratedadequate validity and reliability and overall the inter-correlations between the IBQ subscales as well as thecorrelations between the IBQ subscales and the otherrelated factors provided support for the fit of the scalewithin the existing SDT measures and literature Interms of the relationships between the IBQ subscalesthe results supported there were moderately high posi-tive and negative correlations between the subscalesAdditionally the subscales also correlated with needsatisfaction need frustration and motivation in thedirections that would be expected according to SDTIn the case of some relationships the outcomes hadsimilar correlations with all three need-supportivebehaviors and all three need-thwarting behaviors butthis is not unexpected since the satisfaction or frustra-tion of the three basic needs are often related (Deci ampRyan 1985)

The results of Study 2 supported that the IBQ-Selfstructure held in a sample of coachesmdashreporting ontheir behaviors with their athletes Similar to Study 1the scale also demonstrated adequate validity and relia-bility Overall the inter-correlations between theIBQ-Self subscales and the correlations between theIBQ-Self subscales and other factors showed a similarpattern to Study 1mdashsupporting that the scale relates toother factors as would be expected

Limitations

These results support the applicability of the scale toa sport context however there are some limitationsFirst the structure of the scales was tested using astudent sample of athletes as well as basketball andtrack and field coaches The IBQ should be testedwith a more varied sample of athletes to ensure thescale is a valid measure of perceptions of coaches forother groups such as older athletes and recreationalathletes Furthermore the IBQ-Self should be vali-dated with coaches from a larger variety of sports

MEASUREMENT IN PHYSICAL EDUCATION AND EXERCISE SCIENCE 9

In this sample coaches were coaching team (basket-ball) and individual sports (athletics) however coa-ches can work with all sorts of different athletes atvarying levels of competition and the scalersquos validityshould be extended to include these coaches too Asecond limitation is that the IBQ and IBQ-Self onlyfocused on two people within the sport context ath-letes and coaches There are a number of differentsocial agents in the sport context that impact athletes(or coaches) including the athletesrsquo parents the otherathletes sport administrators etc and the validity ofthe scale should also be tested to examine perceptionsof their behaviors or their own reports of their beha-viors Finally since the IBQ and IBQ-Self were vali-dated using coachesrsquo and athletesrsquo self-reports theirreports should be triangulated by a third party (egvideo recording and coding of interpersonal beha-viors) to confirm reports on these measures corre-spond to their real world behaviors

Future directions

The IBQ and the IBQ-Self promote a number ofinitiatives for future research in sport motivationFirst research should continue to examine the psy-chometric properties of the IBQ For example ath-letes or coaches should be examined at multiple timepoints in order to establish the scalersquos testndashretestreliability and coachesrsquo scores should be corroboratedwith athletesrsquo scores in order to provide support forthe construct validity of the scale Next the IBQ canbe used to explore the role of all six types of need-supportive and need-thwarting interpersonal beha-viors in relation to how they impact athletesrsquo psycho-logical need satisfaction and frustration This willhelp extend the existing research in sport to movebeyond the influential role of AS and AT and focuson CS CT RS and RT interpersonal behaviors aswell Next the IBQ-Self can be used to identify andunderstand the antecedents of all six types of inter-personal behaviors according to SDT Specifically thescale should be used to explore the factors that influ-ence coachesrsquo behaviors with their athletes Finallythe IBQ and IBQ-Self should be used to explorehow coachesrsquo behaviors change over the course of agiven season and how this relates to athlete psycho-logical needs and their motivation for sport

Overall the IBQ and IBQ-Self are new instrumentsfor assessing perceptions or self-reports of all six typesof need-supportive and need-thwarting interpersonalbehaviors for coaches and athletes It is hoped theIBQ and the IBQ-Self will promote research directed

at assessing the behavior of others in sport and how itimpacts need satisfaction need frustration and motiva-tion for athletes

References

Amorose J amp Anderson-Butcher D (2007) Autonomy-supportive coaching and self-determined motivation inhigh school and college athletes A test of self-determina-tion theory Psychology of Sport and Exercise 8 654ndash670doi101016jpsychsport200611003

Bartholomew K Ntoumanis N Ryan R Bosch J ampThoslashgersen-Ntoumani C (2011) Self-determination the-ory and diminished functioning The role of interpersonalcontrol and psychological need thwarting Personality andSocial Psychology Bulletin 37 1459ndash1473 doi1011770146167211413125

Bartholomew K Ntoumanis N amp Thoslashgersen-Ntoumani C(2009) A review of controlling motivational strategies from aself-determination theory perspective Implications for sportscoaches International Review of Sport and ExercisePsychology 2 215ndash233 doi10108017509840903235330

Bartholomew K Ntoumanis N amp Thoslashrgenson-NtoumaniC (2010) The controlling interpersonal style in a coachingcontext Development and initial validation of a psycho-metric scale Journal of Sport amp Exercise Psychology 32193ndash216 Retrieved from httpwwwncbinlmnihgovpubmed20479478

Conroy D amp Coatsworth J (2007) Assessing autonomy-sup-portive coaching strategies in youth sport Psychology of Sportand Exercise 8 671ndash684 doi101016jpsychsport200612001

Deci E amp Ryan R (1985) Intrinsic motivation and self-determination in human behaviour In E Deci amp R Ryan(Eds) Handbook of self-determination research RochesterNY University of Rochester Press

Deci E amp Ryan R (2002) Handbook of self-determinationresearch Rochester NY The University of Rochester Press

Deci E Ryan R Gagneacute M Leone D Usunov J ampKornazheva B (2001) Need satisfaction motivation andwell-being in the work organizations of a former easternbloc country A cross-cultural study of self-determinationPersonality and Social Psychology Bulletin 27 930ndash942doi1011770146167201278002

Deci E Schwartz A Sheinman L amp Ryan R (1981) Aninstrument to assess adultsrsquo orientations toward control ver-sus autonomy with children Reflections on intrinsic motiva-tion and perceived competence Journal of EducationalPsychology 73 642ndash650 doi1010370022-0663735642

Field A (2009) Discovering statistics using SPSS Sage Publi-cations

Gould D Dieffenbach K amp Moffett A (2002) Psychologicalcharacteristics and their development in Olympic cham-pions Journal of Applied Sports Psychology 14 172ndash204doi10108010413200290103482

Hagger M amp Chatzisarantis N (2007) Intrinsic motivationand self-determination in exercise and sport Leeds UKHuman Kinetics Europe Ltd

Hair J Black W Babin B amp Anderson R (2010)Multivariate data analysis (7th ed) Upper Saddle RiverNJ Prentice-Hall Inc

10 M ROCCHI ET AL

Outcome correlations

The IBQ subscales were correlated with need satisfaction insport need frustration in sport and sport motivation aspredicted (see Table 3) Specifically the need-supportivesubscales of the IBQ correlated positively with need satis-faction and negatively with need frustration The need-thwarting subscales of the IBQ correlated negatively withneed satisfaction and positively with need frustrationLooking at athletesrsquomotivation for sport the need-suppor-tive subscales correlated positively with autonomous moti-vation and the need-thwarting subscales correlated withcontrolled motivation For CT it was related to an increasein athletesrsquo autonomous motivation and a decrease in theircontrolled motivation For both perceptions of CT and RTcoach behaviors this was related to decreases in autono-mous motivation for athletes as well as an increase incontrolled motivation

Discussion

Overall the results of Study 1 support the psychometricproperties of the IBQ in a sample of athletes reportingon their coachesrsquo interpersonal behaviors and the scaleperformed equally for male and female athletes This isthe first set of results to support the IBQ can be used toassess perceptions of interpersonal behaviors in thesport context In this sample the structure of the scaleheld as well as the reliabilities and outcome correla-tions supporting the scale performed as would beexpected similar to the original validation studies(Rocchi et al 2017) Although there are relativelyhigh relationships between the need-supportive andneed-thwarting subscales that are in line with theresults of the original validation studies the relation-ships are not high enough to suggest multicollinearity(lt 80) may be an issue (Field 2009) The results of thisstudy support the validity of the IBQ to assess athletesrsquoperceptions of coachesrsquo behaviors Specifically theseresults are in line with what would be expected accord-ing to SDT as athletes who report their coaches useneed-supportive interpersonal behaviors reportedincreased need satisfaction and autonomous motivationin sport (Mageau amp Vallerand 2003) while athleteswho reported their coaches use need-thwarting inter-personal behaviors reported need frustration and con-trolled motivation in sport (Bartholomew et al 2009)

Study 2

The objective of Study 2 is to evaluate the psychometricproperties of the IBQ self-report version within the

sport context Specifically this will be validated usinga sample of sport coaches who are reporting on theirbehaviors when they interact with their athletes It isanticipated the factor structure of the IBQ-Self will holdfor this sample in this context and it will correlate withoutcomes as would be expected according to SDT

Method

Participants

The sample was composed of 240 coaches (Nmale = 170Nfemale = 66 Nmissing = 4) with an average age of4701 years (SD = 1031) The large majority of thecoaches had a college (N = 53 22) or universityeducation (N = 168 70) They were either basketball(N = 111 46) or track and field coaches (N = 12954) and they had an average of 1750 years of coach-ing experience (SD = 1283) The majority (N = 12251) identified as a head coach or as an assistant coach(N = 60 25) and indicated their athletes were very(N = 94 39) or extremely (N = 61 25) competitivecompared to other athletes of their age

Procedures

Coaches were recruited through their respective pro-vincial sporting associations (basketball or athletics) toparticipate in an online survey Each organization sentan email to all coaches inviting them to participate inan online study The coachesrsquo participation was volun-tary and they gave their informed consent before parti-cipating To be eligible the coaches had to be registeredwith a provincial sport organization at the time of datacollection and actively coaching

Materials

The following measures were completed by the coachesthrough an online questionnaire

My interpersonal behaviorsThe coaches completed the IBQ-Self to assess the extentto which they believed they engaged in AS AT CS CTRS and RT interpersonal behaviors with their athletes(Rocchi et al 2017) Coaches were presented each itemusing the stem ldquoWhen I am with my athletes I rdquo andindicated their agreement with each statement using a7-point scale ranging from 1 (do not agree at all) to 7(completely agree) The original validation studies sup-ported the scale had sufficient internal reliabil-ity (α gt 80)

MEASUREMENT IN PHYSICAL EDUCATION AND EXERCISE SCIENCE 7

Need satisfactionThe coaches responded to the positive items from theBasic Need Satisfaction at Work Scale adapted to thesport context (Deci et al 2001) to assess the extenttheir three basic psychological needs were met whilecoaching Like in Study 1 the stems were modified toask the coaches about their need satisfaction whilecoaching instead of at work The fit indices suggestthe model has an acceptable fit except for the TLI(SBχ2(41) = 3836 p lt 001 RMSEA = 06 [90 CI03 08] SRMR = 05 CFI = 90 TLI = 89) howeversince the other fit indices were acceptable the scale wasused The internal consistency estimates were withinacceptable range (autonomy α = 83 competenceα = 71 relatedness α = 74) and a mean score wascalculated for each subscale

Need frustrationThe coaches also completed the Psychological NeedThwarting Scale in Sport (Bartholomew et al 2010) toassess the extent to which their needs were frustratedwhile coaching like in Study 1 A confirmatory factoranalysis confirmed the structure of the scale for coachesand the results supported that the model had an excel-lent fit (SBχ2(32) = 2600 p lt 001 RMSEA = 04 [90CI 00 07] SRMR = 06 CFI = 96 TLI = 94) Theinternal consistency for the three subscales were abovethe minimum (α gt 74) and a mean score was calcu-lated for each subscale to represent the coachesrsquo frus-tration of each need

Coach motivationThe coaches completed the Coach MotivationQuestionnaire (CMQ McLean Mallet amp Newcombe2012) to assess their reasons for participating in sportThe CMQ is a 6-factor scale comprised of 22 itemsmeasuring sport motivation according to each of thesix types of behavioral regulation according to SDT Aconfirmatory factor confirmed the structure of the scalein this sample (SBχ2(194) = 33657 p lt 001RMSEA = 06 [90 CI 04 07] SRMR = 07CFI = 91 TLI = 90) The Cronbachrsquos alphas werecalculated for each subscale and revealed they wereabove the acceptable limit (α gt 72) except for theintrojected subscale which had a lower internal consis-tency (α = 68) Like in Study 1 a mean score wascalculated for autonomous motivation and controlledmotivation

Social desirabilityFinally the coaches also completed the short form ofthe Marlowe-Crowne Social Desirability Scale(Reynolds 1982) to control for whether they were

responding to survey items based on providing favor-able responses The short form of this measure consistsof 10 items where participants are asked to indicatewhether these statements are true or false for themThe sum of all responses considered to be sociallydesirable is calculated to provide an overall measureof social desirability

Analyses

Using the same procedures as the first study the datawas cleaned and then the scale structure was examinedThen the validity and reliability of the subscales wereexamined Finally outcome correlations were con-ducted to examine the relationship between coachneed satisfaction frustration and motivation whilecontrolling for social desirability

Results

Similar to the first study the descriptive statistics analyseson the 24 IBQ-Self items suggested the variables did notall have a normal distribution (skewness range minus1325 to912 kurtosis range minus620 to 821) Data screening ana-lyses did not reveal any multivariate outliers as such theentire sample was retained for the full analyses Themissing data analyses suggested 15 participants weremissing one or two observations on the IBQ-Self how-ever since these missing observations made up less than5 of the total observations the data was imputed usingthe same methods as Study 1 The composite scores forthe coachesrsquo reported use of need-supportive andneed-thwarting behaviors as well as their need satisfac-tion and frustration in coaching their motivation towardscoaching and their likelihood of responding in sociallydesirable ways was calculated (see Tables 1 and 2 for thedescriptive statistics)

Scale structure

The results supported that the factorial model had goodfit (SBχ2(237) = 30304 p lt 001 CFI = 93 TLI = 92RMSEA = 06 [90 CI 03 05] SRMR = 04) Thestandardized factor loadings for each subscale werelarger than 46 (see Table 1) and the factor correlationsas well as the internal reliabilities for each subscale(Cronbachrsquos alpha) are in Table 1 Since there wereonly 66 female coaches in this sample gender invar-iance tests were not performed in this study Nextconvergent and divergent validity was examined foreach of the subscales (see Table 2) and the resultssupported that the scale met the thresholds for bothacross all subscales except for the RT subscale Internal

8 M ROCCHI ET AL

consistency analyses revealed all of the alphas were overthe minimum criteria again except for the RT subscale(α = 40) which was nowhere near the baseline criteriaThe internal consistency for that subscale was calcu-lated using all of the potential combinations of items tosee if one specific item was causing the issue and theresults found there was no combination of the fouritems that promoted a better internal consistency Asa result the reliability for the four items was retained

Outcome correlations

The IBQ-Self subscales were correlated with need satis-faction need frustration autonomous and controlledmotivation while controlling for social desirability(see Table 2) to confirm the relationships were in theexpected directions The results found that the supportsubscales of the IBQ-Self correlated moderately andpositively with need satisfaction and autonomous moti-vation while they had weak or negative relationshipswith need frustration and controlled motivation Theopposite was found with the need-thwarting subscaleswhere there were moderate positive relationships withneed frustration and controlled motivation and weaknegative or no relationships with need satisfaction andautonomous motivation Overall the results were simi-lar to those in Study 1 except the strength of therelationships was weaker

Discussion

The results of this study support the factor structureand validity of the IBQ-Self as a questionnaire forcoaches about their use of interpersonal behaviorsaccording to SDT in their interactions with theirathletes Unfortunately the reliability for the RT sub-scale was weak Since some coaches and athletes maychoose to engage in the relationship when theychoose to participate in the sport (ie select athletefor a team) and others do not (ie everyone trainsduring the same session) it is not totally surprisingthat coaches care about their athletes and this madethe results somewhat ambiguous Perhaps in anothercontext where there is either less control over beingin a relationship with the other people such as afamily member or a supervisor at work or morecontrol like a best friend or partner that subscaleinternal consistency would improve Overall theseresults support that coaches who report using moreneed-supportive behaviors also report higher needsatisfaction and autonomous motivation for theircoaching while coaches who report more need-thwarting behaviors report more need frustration

and controlled motivation for coachingmdashall aftercontrolling for social desirability The results of thisstudy support that generally the IBQ-Self is a validmeasure of coachesrsquo own reports of their interperso-nal behaviors when interacting with their athletes

General discussion

The purpose of these two studies was to extend thevalidity of the IBQ by testing the structure of the scalein a sport setting The results of Study 1 supported thatthe IBQ scale factor structure holds in cases whereathletes report on their perceptions of their coachesrsquointerpersonal behaviors The scale also demonstratedadequate validity and reliability and overall the inter-correlations between the IBQ subscales as well as thecorrelations between the IBQ subscales and the otherrelated factors provided support for the fit of the scalewithin the existing SDT measures and literature Interms of the relationships between the IBQ subscalesthe results supported there were moderately high posi-tive and negative correlations between the subscalesAdditionally the subscales also correlated with needsatisfaction need frustration and motivation in thedirections that would be expected according to SDTIn the case of some relationships the outcomes hadsimilar correlations with all three need-supportivebehaviors and all three need-thwarting behaviors butthis is not unexpected since the satisfaction or frustra-tion of the three basic needs are often related (Deci ampRyan 1985)

The results of Study 2 supported that the IBQ-Selfstructure held in a sample of coachesmdashreporting ontheir behaviors with their athletes Similar to Study 1the scale also demonstrated adequate validity and relia-bility Overall the inter-correlations between theIBQ-Self subscales and the correlations between theIBQ-Self subscales and other factors showed a similarpattern to Study 1mdashsupporting that the scale relates toother factors as would be expected

Limitations

These results support the applicability of the scale toa sport context however there are some limitationsFirst the structure of the scales was tested using astudent sample of athletes as well as basketball andtrack and field coaches The IBQ should be testedwith a more varied sample of athletes to ensure thescale is a valid measure of perceptions of coaches forother groups such as older athletes and recreationalathletes Furthermore the IBQ-Self should be vali-dated with coaches from a larger variety of sports

MEASUREMENT IN PHYSICAL EDUCATION AND EXERCISE SCIENCE 9

In this sample coaches were coaching team (basket-ball) and individual sports (athletics) however coa-ches can work with all sorts of different athletes atvarying levels of competition and the scalersquos validityshould be extended to include these coaches too Asecond limitation is that the IBQ and IBQ-Self onlyfocused on two people within the sport context ath-letes and coaches There are a number of differentsocial agents in the sport context that impact athletes(or coaches) including the athletesrsquo parents the otherathletes sport administrators etc and the validity ofthe scale should also be tested to examine perceptionsof their behaviors or their own reports of their beha-viors Finally since the IBQ and IBQ-Self were vali-dated using coachesrsquo and athletesrsquo self-reports theirreports should be triangulated by a third party (egvideo recording and coding of interpersonal beha-viors) to confirm reports on these measures corre-spond to their real world behaviors

Future directions

The IBQ and the IBQ-Self promote a number ofinitiatives for future research in sport motivationFirst research should continue to examine the psy-chometric properties of the IBQ For example ath-letes or coaches should be examined at multiple timepoints in order to establish the scalersquos testndashretestreliability and coachesrsquo scores should be corroboratedwith athletesrsquo scores in order to provide support forthe construct validity of the scale Next the IBQ canbe used to explore the role of all six types of need-supportive and need-thwarting interpersonal beha-viors in relation to how they impact athletesrsquo psycho-logical need satisfaction and frustration This willhelp extend the existing research in sport to movebeyond the influential role of AS and AT and focuson CS CT RS and RT interpersonal behaviors aswell Next the IBQ-Self can be used to identify andunderstand the antecedents of all six types of inter-personal behaviors according to SDT Specifically thescale should be used to explore the factors that influ-ence coachesrsquo behaviors with their athletes Finallythe IBQ and IBQ-Self should be used to explorehow coachesrsquo behaviors change over the course of agiven season and how this relates to athlete psycho-logical needs and their motivation for sport

Overall the IBQ and IBQ-Self are new instrumentsfor assessing perceptions or self-reports of all six typesof need-supportive and need-thwarting interpersonalbehaviors for coaches and athletes It is hoped theIBQ and the IBQ-Self will promote research directed

at assessing the behavior of others in sport and how itimpacts need satisfaction need frustration and motiva-tion for athletes

References

Amorose J amp Anderson-Butcher D (2007) Autonomy-supportive coaching and self-determined motivation inhigh school and college athletes A test of self-determina-tion theory Psychology of Sport and Exercise 8 654ndash670doi101016jpsychsport200611003

Bartholomew K Ntoumanis N Ryan R Bosch J ampThoslashgersen-Ntoumani C (2011) Self-determination the-ory and diminished functioning The role of interpersonalcontrol and psychological need thwarting Personality andSocial Psychology Bulletin 37 1459ndash1473 doi1011770146167211413125

Bartholomew K Ntoumanis N amp Thoslashgersen-Ntoumani C(2009) A review of controlling motivational strategies from aself-determination theory perspective Implications for sportscoaches International Review of Sport and ExercisePsychology 2 215ndash233 doi10108017509840903235330

Bartholomew K Ntoumanis N amp Thoslashrgenson-NtoumaniC (2010) The controlling interpersonal style in a coachingcontext Development and initial validation of a psycho-metric scale Journal of Sport amp Exercise Psychology 32193ndash216 Retrieved from httpwwwncbinlmnihgovpubmed20479478

Conroy D amp Coatsworth J (2007) Assessing autonomy-sup-portive coaching strategies in youth sport Psychology of Sportand Exercise 8 671ndash684 doi101016jpsychsport200612001

Deci E amp Ryan R (1985) Intrinsic motivation and self-determination in human behaviour In E Deci amp R Ryan(Eds) Handbook of self-determination research RochesterNY University of Rochester Press

Deci E amp Ryan R (2002) Handbook of self-determinationresearch Rochester NY The University of Rochester Press

Deci E Ryan R Gagneacute M Leone D Usunov J ampKornazheva B (2001) Need satisfaction motivation andwell-being in the work organizations of a former easternbloc country A cross-cultural study of self-determinationPersonality and Social Psychology Bulletin 27 930ndash942doi1011770146167201278002

Deci E Schwartz A Sheinman L amp Ryan R (1981) Aninstrument to assess adultsrsquo orientations toward control ver-sus autonomy with children Reflections on intrinsic motiva-tion and perceived competence Journal of EducationalPsychology 73 642ndash650 doi1010370022-0663735642

Field A (2009) Discovering statistics using SPSS Sage Publi-cations

Gould D Dieffenbach K amp Moffett A (2002) Psychologicalcharacteristics and their development in Olympic cham-pions Journal of Applied Sports Psychology 14 172ndash204doi10108010413200290103482

Hagger M amp Chatzisarantis N (2007) Intrinsic motivationand self-determination in exercise and sport Leeds UKHuman Kinetics Europe Ltd

Hair J Black W Babin B amp Anderson R (2010)Multivariate data analysis (7th ed) Upper Saddle RiverNJ Prentice-Hall Inc

10 M ROCCHI ET AL

Need satisfactionThe coaches responded to the positive items from theBasic Need Satisfaction at Work Scale adapted to thesport context (Deci et al 2001) to assess the extenttheir three basic psychological needs were met whilecoaching Like in Study 1 the stems were modified toask the coaches about their need satisfaction whilecoaching instead of at work The fit indices suggestthe model has an acceptable fit except for the TLI(SBχ2(41) = 3836 p lt 001 RMSEA = 06 [90 CI03 08] SRMR = 05 CFI = 90 TLI = 89) howeversince the other fit indices were acceptable the scale wasused The internal consistency estimates were withinacceptable range (autonomy α = 83 competenceα = 71 relatedness α = 74) and a mean score wascalculated for each subscale

Need frustrationThe coaches also completed the Psychological NeedThwarting Scale in Sport (Bartholomew et al 2010) toassess the extent to which their needs were frustratedwhile coaching like in Study 1 A confirmatory factoranalysis confirmed the structure of the scale for coachesand the results supported that the model had an excel-lent fit (SBχ2(32) = 2600 p lt 001 RMSEA = 04 [90CI 00 07] SRMR = 06 CFI = 96 TLI = 94) Theinternal consistency for the three subscales were abovethe minimum (α gt 74) and a mean score was calcu-lated for each subscale to represent the coachesrsquo frus-tration of each need

Coach motivationThe coaches completed the Coach MotivationQuestionnaire (CMQ McLean Mallet amp Newcombe2012) to assess their reasons for participating in sportThe CMQ is a 6-factor scale comprised of 22 itemsmeasuring sport motivation according to each of thesix types of behavioral regulation according to SDT Aconfirmatory factor confirmed the structure of the scalein this sample (SBχ2(194) = 33657 p lt 001RMSEA = 06 [90 CI 04 07] SRMR = 07CFI = 91 TLI = 90) The Cronbachrsquos alphas werecalculated for each subscale and revealed they wereabove the acceptable limit (α gt 72) except for theintrojected subscale which had a lower internal consis-tency (α = 68) Like in Study 1 a mean score wascalculated for autonomous motivation and controlledmotivation

Social desirabilityFinally the coaches also completed the short form ofthe Marlowe-Crowne Social Desirability Scale(Reynolds 1982) to control for whether they were

responding to survey items based on providing favor-able responses The short form of this measure consistsof 10 items where participants are asked to indicatewhether these statements are true or false for themThe sum of all responses considered to be sociallydesirable is calculated to provide an overall measureof social desirability

Analyses

Using the same procedures as the first study the datawas cleaned and then the scale structure was examinedThen the validity and reliability of the subscales wereexamined Finally outcome correlations were con-ducted to examine the relationship between coachneed satisfaction frustration and motivation whilecontrolling for social desirability

Results

Similar to the first study the descriptive statistics analyseson the 24 IBQ-Self items suggested the variables did notall have a normal distribution (skewness range minus1325 to912 kurtosis range minus620 to 821) Data screening ana-lyses did not reveal any multivariate outliers as such theentire sample was retained for the full analyses Themissing data analyses suggested 15 participants weremissing one or two observations on the IBQ-Self how-ever since these missing observations made up less than5 of the total observations the data was imputed usingthe same methods as Study 1 The composite scores forthe coachesrsquo reported use of need-supportive andneed-thwarting behaviors as well as their need satisfac-tion and frustration in coaching their motivation towardscoaching and their likelihood of responding in sociallydesirable ways was calculated (see Tables 1 and 2 for thedescriptive statistics)

Scale structure

The results supported that the factorial model had goodfit (SBχ2(237) = 30304 p lt 001 CFI = 93 TLI = 92RMSEA = 06 [90 CI 03 05] SRMR = 04) Thestandardized factor loadings for each subscale werelarger than 46 (see Table 1) and the factor correlationsas well as the internal reliabilities for each subscale(Cronbachrsquos alpha) are in Table 1 Since there wereonly 66 female coaches in this sample gender invar-iance tests were not performed in this study Nextconvergent and divergent validity was examined foreach of the subscales (see Table 2) and the resultssupported that the scale met the thresholds for bothacross all subscales except for the RT subscale Internal

8 M ROCCHI ET AL

consistency analyses revealed all of the alphas were overthe minimum criteria again except for the RT subscale(α = 40) which was nowhere near the baseline criteriaThe internal consistency for that subscale was calcu-lated using all of the potential combinations of items tosee if one specific item was causing the issue and theresults found there was no combination of the fouritems that promoted a better internal consistency Asa result the reliability for the four items was retained

Outcome correlations

The IBQ-Self subscales were correlated with need satis-faction need frustration autonomous and controlledmotivation while controlling for social desirability(see Table 2) to confirm the relationships were in theexpected directions The results found that the supportsubscales of the IBQ-Self correlated moderately andpositively with need satisfaction and autonomous moti-vation while they had weak or negative relationshipswith need frustration and controlled motivation Theopposite was found with the need-thwarting subscaleswhere there were moderate positive relationships withneed frustration and controlled motivation and weaknegative or no relationships with need satisfaction andautonomous motivation Overall the results were simi-lar to those in Study 1 except the strength of therelationships was weaker

Discussion

The results of this study support the factor structureand validity of the IBQ-Self as a questionnaire forcoaches about their use of interpersonal behaviorsaccording to SDT in their interactions with theirathletes Unfortunately the reliability for the RT sub-scale was weak Since some coaches and athletes maychoose to engage in the relationship when theychoose to participate in the sport (ie select athletefor a team) and others do not (ie everyone trainsduring the same session) it is not totally surprisingthat coaches care about their athletes and this madethe results somewhat ambiguous Perhaps in anothercontext where there is either less control over beingin a relationship with the other people such as afamily member or a supervisor at work or morecontrol like a best friend or partner that subscaleinternal consistency would improve Overall theseresults support that coaches who report using moreneed-supportive behaviors also report higher needsatisfaction and autonomous motivation for theircoaching while coaches who report more need-thwarting behaviors report more need frustration

and controlled motivation for coachingmdashall aftercontrolling for social desirability The results of thisstudy support that generally the IBQ-Self is a validmeasure of coachesrsquo own reports of their interperso-nal behaviors when interacting with their athletes

General discussion

The purpose of these two studies was to extend thevalidity of the IBQ by testing the structure of the scalein a sport setting The results of Study 1 supported thatthe IBQ scale factor structure holds in cases whereathletes report on their perceptions of their coachesrsquointerpersonal behaviors The scale also demonstratedadequate validity and reliability and overall the inter-correlations between the IBQ subscales as well as thecorrelations between the IBQ subscales and the otherrelated factors provided support for the fit of the scalewithin the existing SDT measures and literature Interms of the relationships between the IBQ subscalesthe results supported there were moderately high posi-tive and negative correlations between the subscalesAdditionally the subscales also correlated with needsatisfaction need frustration and motivation in thedirections that would be expected according to SDTIn the case of some relationships the outcomes hadsimilar correlations with all three need-supportivebehaviors and all three need-thwarting behaviors butthis is not unexpected since the satisfaction or frustra-tion of the three basic needs are often related (Deci ampRyan 1985)

The results of Study 2 supported that the IBQ-Selfstructure held in a sample of coachesmdashreporting ontheir behaviors with their athletes Similar to Study 1the scale also demonstrated adequate validity and relia-bility Overall the inter-correlations between theIBQ-Self subscales and the correlations between theIBQ-Self subscales and other factors showed a similarpattern to Study 1mdashsupporting that the scale relates toother factors as would be expected

Limitations

These results support the applicability of the scale toa sport context however there are some limitationsFirst the structure of the scales was tested using astudent sample of athletes as well as basketball andtrack and field coaches The IBQ should be testedwith a more varied sample of athletes to ensure thescale is a valid measure of perceptions of coaches forother groups such as older athletes and recreationalathletes Furthermore the IBQ-Self should be vali-dated with coaches from a larger variety of sports

MEASUREMENT IN PHYSICAL EDUCATION AND EXERCISE SCIENCE 9

In this sample coaches were coaching team (basket-ball) and individual sports (athletics) however coa-ches can work with all sorts of different athletes atvarying levels of competition and the scalersquos validityshould be extended to include these coaches too Asecond limitation is that the IBQ and IBQ-Self onlyfocused on two people within the sport context ath-letes and coaches There are a number of differentsocial agents in the sport context that impact athletes(or coaches) including the athletesrsquo parents the otherathletes sport administrators etc and the validity ofthe scale should also be tested to examine perceptionsof their behaviors or their own reports of their beha-viors Finally since the IBQ and IBQ-Self were vali-dated using coachesrsquo and athletesrsquo self-reports theirreports should be triangulated by a third party (egvideo recording and coding of interpersonal beha-viors) to confirm reports on these measures corre-spond to their real world behaviors

Future directions

The IBQ and the IBQ-Self promote a number ofinitiatives for future research in sport motivationFirst research should continue to examine the psy-chometric properties of the IBQ For example ath-letes or coaches should be examined at multiple timepoints in order to establish the scalersquos testndashretestreliability and coachesrsquo scores should be corroboratedwith athletesrsquo scores in order to provide support forthe construct validity of the scale Next the IBQ canbe used to explore the role of all six types of need-supportive and need-thwarting interpersonal beha-viors in relation to how they impact athletesrsquo psycho-logical need satisfaction and frustration This willhelp extend the existing research in sport to movebeyond the influential role of AS and AT and focuson CS CT RS and RT interpersonal behaviors aswell Next the IBQ-Self can be used to identify andunderstand the antecedents of all six types of inter-personal behaviors according to SDT Specifically thescale should be used to explore the factors that influ-ence coachesrsquo behaviors with their athletes Finallythe IBQ and IBQ-Self should be used to explorehow coachesrsquo behaviors change over the course of agiven season and how this relates to athlete psycho-logical needs and their motivation for sport

Overall the IBQ and IBQ-Self are new instrumentsfor assessing perceptions or self-reports of all six typesof need-supportive and need-thwarting interpersonalbehaviors for coaches and athletes It is hoped theIBQ and the IBQ-Self will promote research directed

at assessing the behavior of others in sport and how itimpacts need satisfaction need frustration and motiva-tion for athletes

References

Amorose J amp Anderson-Butcher D (2007) Autonomy-supportive coaching and self-determined motivation inhigh school and college athletes A test of self-determina-tion theory Psychology of Sport and Exercise 8 654ndash670doi101016jpsychsport200611003

Bartholomew K Ntoumanis N Ryan R Bosch J ampThoslashgersen-Ntoumani C (2011) Self-determination the-ory and diminished functioning The role of interpersonalcontrol and psychological need thwarting Personality andSocial Psychology Bulletin 37 1459ndash1473 doi1011770146167211413125

Bartholomew K Ntoumanis N amp Thoslashgersen-Ntoumani C(2009) A review of controlling motivational strategies from aself-determination theory perspective Implications for sportscoaches International Review of Sport and ExercisePsychology 2 215ndash233 doi10108017509840903235330

Bartholomew K Ntoumanis N amp Thoslashrgenson-NtoumaniC (2010) The controlling interpersonal style in a coachingcontext Development and initial validation of a psycho-metric scale Journal of Sport amp Exercise Psychology 32193ndash216 Retrieved from httpwwwncbinlmnihgovpubmed20479478

Conroy D amp Coatsworth J (2007) Assessing autonomy-sup-portive coaching strategies in youth sport Psychology of Sportand Exercise 8 671ndash684 doi101016jpsychsport200612001

Deci E amp Ryan R (1985) Intrinsic motivation and self-determination in human behaviour In E Deci amp R Ryan(Eds) Handbook of self-determination research RochesterNY University of Rochester Press

Deci E amp Ryan R (2002) Handbook of self-determinationresearch Rochester NY The University of Rochester Press

Deci E Ryan R Gagneacute M Leone D Usunov J ampKornazheva B (2001) Need satisfaction motivation andwell-being in the work organizations of a former easternbloc country A cross-cultural study of self-determinationPersonality and Social Psychology Bulletin 27 930ndash942doi1011770146167201278002

Deci E Schwartz A Sheinman L amp Ryan R (1981) Aninstrument to assess adultsrsquo orientations toward control ver-sus autonomy with children Reflections on intrinsic motiva-tion and perceived competence Journal of EducationalPsychology 73 642ndash650 doi1010370022-0663735642

Field A (2009) Discovering statistics using SPSS Sage Publi-cations

Gould D Dieffenbach K amp Moffett A (2002) Psychologicalcharacteristics and their development in Olympic cham-pions Journal of Applied Sports Psychology 14 172ndash204doi10108010413200290103482

Hagger M amp Chatzisarantis N (2007) Intrinsic motivationand self-determination in exercise and sport Leeds UKHuman Kinetics Europe Ltd

Hair J Black W Babin B amp Anderson R (2010)Multivariate data analysis (7th ed) Upper Saddle RiverNJ Prentice-Hall Inc

10 M ROCCHI ET AL

consistency analyses revealed all of the alphas were overthe minimum criteria again except for the RT subscale(α = 40) which was nowhere near the baseline criteriaThe internal consistency for that subscale was calcu-lated using all of the potential combinations of items tosee if one specific item was causing the issue and theresults found there was no combination of the fouritems that promoted a better internal consistency Asa result the reliability for the four items was retained

Outcome correlations

The IBQ-Self subscales were correlated with need satis-faction need frustration autonomous and controlledmotivation while controlling for social desirability(see Table 2) to confirm the relationships were in theexpected directions The results found that the supportsubscales of the IBQ-Self correlated moderately andpositively with need satisfaction and autonomous moti-vation while they had weak or negative relationshipswith need frustration and controlled motivation Theopposite was found with the need-thwarting subscaleswhere there were moderate positive relationships withneed frustration and controlled motivation and weaknegative or no relationships with need satisfaction andautonomous motivation Overall the results were simi-lar to those in Study 1 except the strength of therelationships was weaker

Discussion

The results of this study support the factor structureand validity of the IBQ-Self as a questionnaire forcoaches about their use of interpersonal behaviorsaccording to SDT in their interactions with theirathletes Unfortunately the reliability for the RT sub-scale was weak Since some coaches and athletes maychoose to engage in the relationship when theychoose to participate in the sport (ie select athletefor a team) and others do not (ie everyone trainsduring the same session) it is not totally surprisingthat coaches care about their athletes and this madethe results somewhat ambiguous Perhaps in anothercontext where there is either less control over beingin a relationship with the other people such as afamily member or a supervisor at work or morecontrol like a best friend or partner that subscaleinternal consistency would improve Overall theseresults support that coaches who report using moreneed-supportive behaviors also report higher needsatisfaction and autonomous motivation for theircoaching while coaches who report more need-thwarting behaviors report more need frustration

and controlled motivation for coachingmdashall aftercontrolling for social desirability The results of thisstudy support that generally the IBQ-Self is a validmeasure of coachesrsquo own reports of their interperso-nal behaviors when interacting with their athletes

General discussion

The purpose of these two studies was to extend thevalidity of the IBQ by testing the structure of the scalein a sport setting The results of Study 1 supported thatthe IBQ scale factor structure holds in cases whereathletes report on their perceptions of their coachesrsquointerpersonal behaviors The scale also demonstratedadequate validity and reliability and overall the inter-correlations between the IBQ subscales as well as thecorrelations between the IBQ subscales and the otherrelated factors provided support for the fit of the scalewithin the existing SDT measures and literature Interms of the relationships between the IBQ subscalesthe results supported there were moderately high posi-tive and negative correlations between the subscalesAdditionally the subscales also correlated with needsatisfaction need frustration and motivation in thedirections that would be expected according to SDTIn the case of some relationships the outcomes hadsimilar correlations with all three need-supportivebehaviors and all three need-thwarting behaviors butthis is not unexpected since the satisfaction or frustra-tion of the three basic needs are often related (Deci ampRyan 1985)

The results of Study 2 supported that the IBQ-Selfstructure held in a sample of coachesmdashreporting ontheir behaviors with their athletes Similar to Study 1the scale also demonstrated adequate validity and relia-bility Overall the inter-correlations between theIBQ-Self subscales and the correlations between theIBQ-Self subscales and other factors showed a similarpattern to Study 1mdashsupporting that the scale relates toother factors as would be expected

Limitations

These results support the applicability of the scale toa sport context however there are some limitationsFirst the structure of the scales was tested using astudent sample of athletes as well as basketball andtrack and field coaches The IBQ should be testedwith a more varied sample of athletes to ensure thescale is a valid measure of perceptions of coaches forother groups such as older athletes and recreationalathletes Furthermore the IBQ-Self should be vali-dated with coaches from a larger variety of sports

MEASUREMENT IN PHYSICAL EDUCATION AND EXERCISE SCIENCE 9

In this sample coaches were coaching team (basket-ball) and individual sports (athletics) however coa-ches can work with all sorts of different athletes atvarying levels of competition and the scalersquos validityshould be extended to include these coaches too Asecond limitation is that the IBQ and IBQ-Self onlyfocused on two people within the sport context ath-letes and coaches There are a number of differentsocial agents in the sport context that impact athletes(or coaches) including the athletesrsquo parents the otherathletes sport administrators etc and the validity ofthe scale should also be tested to examine perceptionsof their behaviors or their own reports of their beha-viors Finally since the IBQ and IBQ-Self were vali-dated using coachesrsquo and athletesrsquo self-reports theirreports should be triangulated by a third party (egvideo recording and coding of interpersonal beha-viors) to confirm reports on these measures corre-spond to their real world behaviors

Future directions

The IBQ and the IBQ-Self promote a number ofinitiatives for future research in sport motivationFirst research should continue to examine the psy-chometric properties of the IBQ For example ath-letes or coaches should be examined at multiple timepoints in order to establish the scalersquos testndashretestreliability and coachesrsquo scores should be corroboratedwith athletesrsquo scores in order to provide support forthe construct validity of the scale Next the IBQ canbe used to explore the role of all six types of need-supportive and need-thwarting interpersonal beha-viors in relation to how they impact athletesrsquo psycho-logical need satisfaction and frustration This willhelp extend the existing research in sport to movebeyond the influential role of AS and AT and focuson CS CT RS and RT interpersonal behaviors aswell Next the IBQ-Self can be used to identify andunderstand the antecedents of all six types of inter-personal behaviors according to SDT Specifically thescale should be used to explore the factors that influ-ence coachesrsquo behaviors with their athletes Finallythe IBQ and IBQ-Self should be used to explorehow coachesrsquo behaviors change over the course of agiven season and how this relates to athlete psycho-logical needs and their motivation for sport

Overall the IBQ and IBQ-Self are new instrumentsfor assessing perceptions or self-reports of all six typesof need-supportive and need-thwarting interpersonalbehaviors for coaches and athletes It is hoped theIBQ and the IBQ-Self will promote research directed

at assessing the behavior of others in sport and how itimpacts need satisfaction need frustration and motiva-tion for athletes

References

Amorose J amp Anderson-Butcher D (2007) Autonomy-supportive coaching and self-determined motivation inhigh school and college athletes A test of self-determina-tion theory Psychology of Sport and Exercise 8 654ndash670doi101016jpsychsport200611003

Bartholomew K Ntoumanis N Ryan R Bosch J ampThoslashgersen-Ntoumani C (2011) Self-determination the-ory and diminished functioning The role of interpersonalcontrol and psychological need thwarting Personality andSocial Psychology Bulletin 37 1459ndash1473 doi1011770146167211413125

Bartholomew K Ntoumanis N amp Thoslashgersen-Ntoumani C(2009) A review of controlling motivational strategies from aself-determination theory perspective Implications for sportscoaches International Review of Sport and ExercisePsychology 2 215ndash233 doi10108017509840903235330

Bartholomew K Ntoumanis N amp Thoslashrgenson-NtoumaniC (2010) The controlling interpersonal style in a coachingcontext Development and initial validation of a psycho-metric scale Journal of Sport amp Exercise Psychology 32193ndash216 Retrieved from httpwwwncbinlmnihgovpubmed20479478

Conroy D amp Coatsworth J (2007) Assessing autonomy-sup-portive coaching strategies in youth sport Psychology of Sportand Exercise 8 671ndash684 doi101016jpsychsport200612001

Deci E amp Ryan R (1985) Intrinsic motivation and self-determination in human behaviour In E Deci amp R Ryan(Eds) Handbook of self-determination research RochesterNY University of Rochester Press

Deci E amp Ryan R (2002) Handbook of self-determinationresearch Rochester NY The University of Rochester Press

Deci E Ryan R Gagneacute M Leone D Usunov J ampKornazheva B (2001) Need satisfaction motivation andwell-being in the work organizations of a former easternbloc country A cross-cultural study of self-determinationPersonality and Social Psychology Bulletin 27 930ndash942doi1011770146167201278002

Deci E Schwartz A Sheinman L amp Ryan R (1981) Aninstrument to assess adultsrsquo orientations toward control ver-sus autonomy with children Reflections on intrinsic motiva-tion and perceived competence Journal of EducationalPsychology 73 642ndash650 doi1010370022-0663735642

Field A (2009) Discovering statistics using SPSS Sage Publi-cations

Gould D Dieffenbach K amp Moffett A (2002) Psychologicalcharacteristics and their development in Olympic cham-pions Journal of Applied Sports Psychology 14 172ndash204doi10108010413200290103482

Hagger M amp Chatzisarantis N (2007) Intrinsic motivationand self-determination in exercise and sport Leeds UKHuman Kinetics Europe Ltd

Hair J Black W Babin B amp Anderson R (2010)Multivariate data analysis (7th ed) Upper Saddle RiverNJ Prentice-Hall Inc

10 M ROCCHI ET AL

In this sample coaches were coaching team (basket-ball) and individual sports (athletics) however coa-ches can work with all sorts of different athletes atvarying levels of competition and the scalersquos validityshould be extended to include these coaches too Asecond limitation is that the IBQ and IBQ-Self onlyfocused on two people within the sport context ath-letes and coaches There are a number of differentsocial agents in the sport context that impact athletes(or coaches) including the athletesrsquo parents the otherathletes sport administrators etc and the validity ofthe scale should also be tested to examine perceptionsof their behaviors or their own reports of their beha-viors Finally since the IBQ and IBQ-Self were vali-dated using coachesrsquo and athletesrsquo self-reports theirreports should be triangulated by a third party (egvideo recording and coding of interpersonal beha-viors) to confirm reports on these measures corre-spond to their real world behaviors

Future directions

The IBQ and the IBQ-Self promote a number ofinitiatives for future research in sport motivationFirst research should continue to examine the psy-chometric properties of the IBQ For example ath-letes or coaches should be examined at multiple timepoints in order to establish the scalersquos testndashretestreliability and coachesrsquo scores should be corroboratedwith athletesrsquo scores in order to provide support forthe construct validity of the scale Next the IBQ canbe used to explore the role of all six types of need-supportive and need-thwarting interpersonal beha-viors in relation to how they impact athletesrsquo psycho-logical need satisfaction and frustration This willhelp extend the existing research in sport to movebeyond the influential role of AS and AT and focuson CS CT RS and RT interpersonal behaviors aswell Next the IBQ-Self can be used to identify andunderstand the antecedents of all six types of inter-personal behaviors according to SDT Specifically thescale should be used to explore the factors that influ-ence coachesrsquo behaviors with their athletes Finallythe IBQ and IBQ-Self should be used to explorehow coachesrsquo behaviors change over the course of agiven season and how this relates to athlete psycho-logical needs and their motivation for sport

Overall the IBQ and IBQ-Self are new instrumentsfor assessing perceptions or self-reports of all six typesof need-supportive and need-thwarting interpersonalbehaviors for coaches and athletes It is hoped theIBQ and the IBQ-Self will promote research directed

at assessing the behavior of others in sport and how itimpacts need satisfaction need frustration and motiva-tion for athletes

References

Amorose J amp Anderson-Butcher D (2007) Autonomy-supportive coaching and self-determined motivation inhigh school and college athletes A test of self-determina-tion theory Psychology of Sport and Exercise 8 654ndash670doi101016jpsychsport200611003

Bartholomew K Ntoumanis N Ryan R Bosch J ampThoslashgersen-Ntoumani C (2011) Self-determination the-ory and diminished functioning The role of interpersonalcontrol and psychological need thwarting Personality andSocial Psychology Bulletin 37 1459ndash1473 doi1011770146167211413125

Bartholomew K Ntoumanis N amp Thoslashgersen-Ntoumani C(2009) A review of controlling motivational strategies from aself-determination theory perspective Implications for sportscoaches International Review of Sport and ExercisePsychology 2 215ndash233 doi10108017509840903235330

Bartholomew K Ntoumanis N amp Thoslashrgenson-NtoumaniC (2010) The controlling interpersonal style in a coachingcontext Development and initial validation of a psycho-metric scale Journal of Sport amp Exercise Psychology 32193ndash216 Retrieved from httpwwwncbinlmnihgovpubmed20479478

Conroy D amp Coatsworth J (2007) Assessing autonomy-sup-portive coaching strategies in youth sport Psychology of Sportand Exercise 8 671ndash684 doi101016jpsychsport200612001

Deci E amp Ryan R (1985) Intrinsic motivation and self-determination in human behaviour In E Deci amp R Ryan(Eds) Handbook of self-determination research RochesterNY University of Rochester Press

Deci E amp Ryan R (2002) Handbook of self-determinationresearch Rochester NY The University of Rochester Press

Deci E Ryan R Gagneacute M Leone D Usunov J ampKornazheva B (2001) Need satisfaction motivation andwell-being in the work organizations of a former easternbloc country A cross-cultural study of self-determinationPersonality and Social Psychology Bulletin 27 930ndash942doi1011770146167201278002

Deci E Schwartz A Sheinman L amp Ryan R (1981) Aninstrument to assess adultsrsquo orientations toward control ver-sus autonomy with children Reflections on intrinsic motiva-tion and perceived competence Journal of EducationalPsychology 73 642ndash650 doi1010370022-0663735642

Field A (2009) Discovering statistics using SPSS Sage Publi-cations

Gould D Dieffenbach K amp Moffett A (2002) Psychologicalcharacteristics and their development in Olympic cham-pions Journal of Applied Sports Psychology 14 172ndash204doi10108010413200290103482

Hagger M amp Chatzisarantis N (2007) Intrinsic motivationand self-determination in exercise and sport Leeds UKHuman Kinetics Europe Ltd

Hair J Black W Babin B amp Anderson R (2010)Multivariate data analysis (7th ed) Upper Saddle RiverNJ Prentice-Hall Inc

10 M ROCCHI ET AL

Hooper D Coughlan J amp Mullen M R (2008) Structuralequation modelling Guidelines for determining model fitThe Electronic Journal of Business Research Methods 6 53ndash60 Retrieved from wwwejbrmcomissuedownloadhtmlidArticle=183

Hu L amp Bentler P (1999) Cutoff criteria for fit indexes incovariance structure analysis Conventional criteria versusnew alternatives Structural Equation Modeling 6 1ndash55doi10108010705519909540118

Jones R Armour K amp Potrac P (2004) Sports coachingcultures From practice to theory London UK Routledge

Li C Wang C Pyun D amp Kee Y (2013) Burnout and itsrelations with basic psychological needs and motivationamong athletes A systematic review and meta-analysisPsychology of Sport and Exercise 14 692ndash700 doi101016jpsychsport201304009

MacCallum R Widaman K Zhang S amp Hong S (1999)Sample size in factor analysis Psychological Methods 484ndash99 doi1010371082-989X4184

Mageau G Ranger F Joussemet M Koestner RMoreau E amp Forest J (2015) Validation of the per-ceived parental autonomy support scale (P-PASS)Canadian Journal of Behavioural Science 47 251ndash262doi101037a0039325

Mageau G amp Vallerand R (2003) The coach-athlete rela-tionship A motivational model Journal of Sports Sciences21 883ndash904 doi1010800264041031000140374

McLean K Mallet C amp Newcombe P (2012) Assessingcoach motivation The development of the CoachMotivation Questionnaire (CMQ) Journal of Sport ampExercise Psychology 34 184ndash207 doi101123jsep342184

Meade A W (2005 April) Sample size and tests of measure-ment invariance Los Angeles CA Society for Industrialand Organizational Psychology Retrieved from httpwww4ncsuedu~awmeadeLinksPapersNampCFA(SIOP05)pdf

Moreau E amp Mageau G (2013) Consequences and correlatesof autonomy support in various life domains PsychologieFranccedilaise 58 195ndash227 doi101016jpsfr201303003

Muthen L amp Muthen B (2010) Mplus userrsquos guideStatistical analysis with latent variables (6th ed) LosAngeles CA Authors

Pelletier L Rocchi M Vallerand R Deci E amp Ryan R(2013) Validation of the revised sport motivation scale(SMS-II) Psychology of Sport and Exercise 14 329ndash341doi101016jpsychsport201212002

Pelletier L Seacuteguin-Leacutevesque C amp Legault L (2002) Pressurefrom above and pressure from below as determinants of tea-chersrsquo motivation and teaching behaviors Journal ofEducational Psychology 94 186 doi1010370022-0663941186

Pomerantz E Cheung C amp Qin L (2012) Relatednessbetween children and parents Implications for motivation

In R Ryan (Ed) The Oxford handbook of motivation (pp335ndash349) New York NY Oxford University Press

Reynolds W (1982) Development of reliable and valid shortforms of the Marlowe-Crowne social desirability scaleJournal of Clinical Psychology 38 119ndash125 doi1010021097-4679

Rocchi M Pelletier L Cheung S Baxter D amp Beaudry S(2017) Assessing need-supportive and need-thwartinginterpersonal behaviours The Interpersonal BehavioursQuestionnaire (IBQ) Personality and IndividualDifferences 104 423ndash433 doi101016jpaid201608034

Rocchi M Pelletier L amp Couture A (2013) Determinantsof coach motivation and autonomy supportive coachingbehaviours Psychology of Sport and Exercise 14 852ndash859doi101016jpsychsport201307002

Sheldon K (2011) Integrating behavioural-motive andexperiential-requirement perspectives on psychologicalneeds A two process model Psychological Review 118552ndash569 doi101037a0024758

Sheldon K amp Filak V (2008) Manipulating autonomy compe-tence and relatedness support in a game-learning contextNewevidence that all three needs matter British Journal of SocialPsychology 47 257ndash283 doi101348014466607χ238797

Smith N Tessier D Tzioumakis Y Fabra P Quested EAppleton P Duda J L (2016) The relationship betweenobserved and perceived assessments of the coach-createdmotivational environment and links to athlete motivationPsychology of Sport and Exercise 23 51ndash63 doi101016jpsychsport201511001

Smith N Tessier D Tzioumakis Y Quested E Appleton PSarrazin P Duda J L (2015) Development and valida-tion of the multidimensional motivational climate observa-tion system Journal of Sport amp Exercise Psychology 37 4ndash22doi101123jsep2014-0059

Stebbings J Taylor M Spray C amp Ntoumanis N(2012) Antecedents of perceived coach interpersonalbehaviors The coaching environment and coach psycho-logical well- and ill-being Journal of Sport amp ExercisePsychology 34 481ndash502 Retrieved from httpjournalshumankineticscomjsep

Taylor I Ntoumanis N amp Standage M (2008) A self-determination theory approach to understanding theantecedents of teachersrsquo motivational strategies in phy-sical education Journal of Sport and Exercise Psychology30 75ndash94 doi101123jsep30175

Vallerand R (2001) A hierarchical model of intrinsic andextrinsic motivation in sport and exercise In G C Roberts(Ed) Advances in motivation in sport and exercise (pp263ndash319) Champaign IL Human Kinetics

Williams N Whipp P Jackson B amp Dimmock J (2013)Relatedness support and the retention of young femalegolfers Journal of Applied Sport Psychology 25 412ndash430doi101080104132002012749311

MEASUREMENT IN PHYSICAL EDUCATION AND EXERCISE SCIENCE 11