2006-Wilson Et Al. - Integrated Regulation and Exercise (JABR)

79

Journal of Applied Biobehavioral Research 2006 11 2 pp 79-104Copyright copy 2006 by Bellwether Publishing Ltd All rights reserved

ldquoItrsquos Who I Am Reallyrdquo The Importance of Integrated Regulation in Exercise Contexts1

PHILIP M WILSON2

Department of Physical Education and KinesiologyBrock University

St Catharines Ontario Canada

WENDY M RODGERS CHRISTINA C LOITZ AND GIULIA SCIMEFaculty of Physical Education and Recreation

University of AlbertaEdmonton Alberta Canada

The purpose of this series of studies was to evaluate a measure of integrated regulationspecific to exercise contexts in line with Self-Determination Theory (SDT Deci amp Ryan1985 2002) To address this purpose three studies were conducted to test select psycho-metric and theoretical properties of four integrated regulation items created for use withinthe Behavioural Regulation in Exercise Questionnaire (BREQ) Confirmatory factor anal-yses conducted in Studies 1 and 2 supported the inclusion of integrated regulation withinthe expanded BREQ measurement model Simultaneous multiple regression analyses(SMRAs) conducted in Study 2 indicated that greater need satisfaction promoted endorse-ment of autonomous exercise motives including integrated regulation Finally SMRAconducted in Study 3 revealed that integrated regulation contributed to the prediction ofexercise behavior and physical self-worth Collectively the results of this investigationsuggest that the new integrated regulation items can be used in conjunction with theBREQ without compromising validity and support Deci and Ryanrsquos (1985 2002) asser-tions regarding the importance of autonomous extrinsic motives including integrated reg-ulation in exercise domains

1The first author was supported by Grant 410-2005-1485 from the Social Sciences and Humani-ties Research Council of Canada (SSHRC) during manuscript preparation Funding for the datacollected in this study was provided through an SSHRC grant awarded to the second author Theauthors thank the staff of Jeanrsquos Marines (Ontario Canada) who provided assistance during the data-collection phase of Study 2 and the participants who gave freely of their time to be involved in eachof these studies The authors gratefully acknowledge the assistance of Shawn N Fraser and Terra CMurray (both at the University of Alberta) at different phases throughout the projectrsquos data manage-ment and the comments of an anonymous reviewer who offered useful and thought-provokingsuggestions on an earlier draft of the manuscript

2Correspondence concerning this article should be addressed to Philip M Wilson Department ofPhysical Education and Kinesiology 500 Glenridge Avenue Brock University St CatharinesOntario L2S 2A1 Canada E-mail phwilsonbrockuca

80 WILSON ET AL

Despite the health benefits associated with regular physical activity (BouchardBlair amp Haskell 2006) population health data indicate that most adults remaininsufficiently active to offset chronic diseases or to promote quality of life (BiddleFox amp Boutcher 2000 Craig amp Cameron 2004) Considering these participationtrends a greater focus has been placed on understanding why people engage inphysical activity using theoretical frameworks that elucidate the processes shapinghealth behaviors (Biddle et al 2000) A theoretical approach holding considerableappeal for understanding multiple health behaviors including physical activity isself-determination theory (SDT Deci amp Ryan 1985 2002)

Deci and Ryan (1985 2002) proposed within SDT that motivation residesalong an internalization continuum reflecting the degree to which the behaviorhas been integrated with the self A personrsquos relative position along the internal-ization continuum is fostered by the extent to which social contexts satisfy basicpsychological needs for competence autonomy and relatedness (Ryan 1995)The distal ends of the internalization continuum are anchored by amotivation andintrinsic regulation Amotivation concerns a lack of intention to act and repre-sents a state akin to learned helplessness (Deci amp Ryan 2002) Converselyintrinsic regulation is the most autonomous form of motivation in which partici-pation is regulated by fun interest or the behaviorrsquos self-rewarding nature (Deciamp Ryan 2002) Previous research has indicated that amotivation is associatedwith a reduction in effort and importance ascribed to exercise (Wilson RodgersFraser amp Murray 2004) and greater dropout from sport (Pelletier FortierVallerand amp Briegravere 2001) In contrast intrinsic regulation is associated withgreater intention to continue exercising (Wilson et al 2004) persistent sportbehavior (Pelletier et al 2002) and enhanced physical self-worth (Wilson ampRodgers 2002)

Deci and Ryan (1985 2002) proposed four types of extrinsic motivationwithin SDT that vary considerably in terms of their integration with the selfExternal regulation and introjected regulation epitomize controlling internaliza-tions that motivate behavior via a desire to appease others avoid negative feel-ings or maintain conditional self-worth (Deci amp Ryan 2002) Greater reliance oncontrolling motives has been linked with dropout from sport (Pelletier et al1995) and lower physical self-worth in female exercisers (Wilson amp Rodgers2002) Identified regulation and integrated regulation represent more autono-mous extrinsic motives proposed by SDT (Deci amp Ryan 1985 2002) Identifiedregulation represents the lower limit of autonomous motivation in which partici-pation is regulated by goal values or the importance of behavioral outcomes(Deci amp Ryan 2002) Integrated regulation represents the most autonomousform of extrinsic motivation occurring when congruence exists between behav-ioral regulation and ldquopersonally endorsed values goals and needs that arealready part of the selfrdquo (Deci amp Ryan 2002 p 18) Previous research hasindicated that identified regulation is associated with more frequent exercise

INTEGRATED REGULATION IN EXERCISE CONTEXTS 81

participation in university students (Wilson et al 2004) and persistence behaviorin athletes (Pelletier et al 2001)

A limitation of physical activity research applying SDT to study motivationalissues concerns the shortage of research focusing on integrated regulation Previ-ous research examining motivation in sport settings has indicated that integratedreasons for sport involvement were not relevant to university-based athletes(Pelletier et al 1995) Nevertheless research by Pelletier and colleagues hasindicated that integrated regulation is associated with healthier eating patterns(Pelletier Dion Slovinec-DrsquoAngelo amp Reid 2004) and predicts environmen-tally responsible behaviors (Green-Demers Pelletier amp Meacutenard 1997) Collec-tively these studies indicate that integrated regulation may be an importantmotivational influence in promoting health behaviors and it is surprising thatlimited research has examined this aspect of SDTrsquos motivational continuum inexercise contexts

A reason why research examining integrated regulation has been limited inexercise settings concerns the use of the Behavioural Regulation in ExerciseQuestionnaire (BREQ Mullan Markland amp Ingeldew 1997) as the instrumentof choice for SDT-based investigations In their original article Mullan et al sup-ported a four-factor measurement model capturing external introjected identi-fied and intrinsic regulations for exercise underpinning BREQ responses fromdiverse samples of sport-center attendees and workers in Great Britain Subse-quent research by Markland and Tobin (2004) has supported the construct valid-ity of scores from the BREQ-2 which includes a subscale to assess amotivationtoward exercise behavior Additional research in North American samples hassupported the structural validity of BREQ (Wilson Rodgers amp Fraser 2002) andBREQ-2 (Wilson amp Rodgers 2004) scores and the ability of BREQ scores todifferentiate between physically active and inactive groups (Landry amp Solomon2004) Overall the BREQrsquos construct validity evidence is impressive howeverone limitation of the instrument is the absence of an integrated regulation sub-scale that represents the most self-determined form of extrinsic motivation inSDTrsquos framework (Deci amp Ryan 1985 2002)3

3Li (1999) developed the Exercise Motivation Scale (EMS) which measures amotivation extrinsicmotivation according to SDT (external introjected identified and integrated regulation) and intrinsicmotivation to learn experience sensations and accomplish based on adaptations of Vallerandrsquos (2001)hierarchical model A new set of integrated-regulation items were created for these studies rather thanadopting the EMS items for various reasons First Li reported global statistical results in his paper thatmake it difficult to discern clear psychometric support for the EMS or the integrated-regulation itemsSecond the format of the EMS items does not fit comfortably with either the BREQ or the BREQ-2 andwould require modification for use with either instrument Finally Lirsquos data indicate that the correla-tions between extrinsic motives assessed with the EMS do not conform to the ldquoquasi-simplex patternrdquo(Deci amp Ryan 2002 p 18) advocated within SDT given that integrated regulation is correlated morestrongly with introjected regulation (oslash = 46) than with identified regulation (oslash = 40)

82 WILSON ET AL

Considering the utility of SDT for understanding motivational processes inphysical activity (Frederick-Recascino 2002) and the evidence supporting theconstruct validity of BREQ scores (Landry amp Solomon 2004 Mullan et al1997 Wilson et al 2002 2004) it seems reasonable to suggest that furtherexploration of the role of integrated regulation in exercise may be worthwhile forat least two reasons First it is conceivable that while some individuals partici-pate in exercise for intrinsic reasons others may initiate or sustain exercisebehavior for extrinsic reasons that according to SDT include participatingbecause the behavior is congruent with their self-identity and thereby regulatedfor integrated reasons (Deci amp Ryan 1985 2002) Second research examiningthe feasibility of extending the BREQrsquos (Mullan et al 1997) item pool to includean assessment of integrated regulation seems appropriate given that the instru-ment was created initially to capture SDTrsquos motivational continuum in exerciseand previously was modified to include items that were not part of the BREQrsquosoriginal development (Markland amp Tobin 2004) Consequently the overall pur-pose of this investigation is to evaluate select measurement properties of a pre-liminary set of items designed to measure integrated regulation in exercise and tosit comfortably within the BREQ (Mullan et al 1997)

Study 1

The purpose of Study 1 is to evaluate the structural and convergentdivergentvalidity of scores derived from the BREQ (Mullan et al 1997) measurementmodel including the new integrated regulation items

Method

Participants

Participants were 61 male (Mage = 1954 years SD = 175 MBMI = 2373 kgm2 SDBMI = 361 685 lt 2499 kgm2) and 146 female (Mage = 1918 yearsSD = 152 MBMI = 2188 kgm2 SDBMI = 392 854 lt 2499 kgm) undergrad-uate psychology students enrolled at a large Canadian university Considerablevariability in physical activity behavior was evident in the sample data (malesMMETS = 5982 SDMETS = 4493 females MMETS = 7058 SDMETS = 6974446 of the overall sample engaging in 3 or more strenuous exercise sessionsweek for past 7 days)4

4Estimates were derived from global responses to the Godin Leisure Time Exercise Question-naire (Godin amp Shepherd 1985) that is described in Study 3


Measures

Behavioural Regulation in Exercise Questionnaire (BREQ Mullan et al1997) Participants completed the 15-item BREQ as an index of SDTrsquos exerciseregulations The BREQ contains four subscales measuring external introjectedidentified and intrinsic exercise regulation Following the stem ldquoWhy do youexerciserdquo participants responded to each item on a 5-point scale ranging from 0(not true for me) to 4 (very true for me) Previous research has supported theBREQrsquos factor structure and subscale reliability (Cronbachrsquos αs gt 70 Mullan etal 1997) and the ability of BREQ scores to distinguish physically active frominactive groups (Landry amp Solomon 2004)5

Integrated regulation Participants completed four items designed to assessintegrated regulation in line with SDT (Deci amp Ryan 1985 2002) and to fit com-fortably with the BREQ (Mullan et al 1997) The integrated regulation itemswere created based on theoretical considerations (Deci amp Ryan 1985 2002) andadaptations of items from other instruments (Pelletier et al 2004)

Exercise Motivation Scale (EMS Li 1999) Participants completed four EMSitems assessing integrated regulation The EMS contains eight subscales assess-ing amotivation extrinsic motivation and intrinsic motivation in accordancewith Vallerandrsquos (2001) hierarchical model Following a stem (ldquoWhy are you cur-rently participating in this exercise activityrdquo) participants responded to eachitem on a 6-point scale ranging from 1 (strongly disagree) to 6 (strongly agree)Li provided evidence of reliability (α gt 75 7-day testndashretest rs gt 78) and crite-rion validity linking integrated regulation with greater competence locus ofcontrol and social support in university students We included the EMS items toassess convergentndashdivergent validity of the scores from the new integrated regu-lation items with an existing measure of integrated regulation for exercise

Procedure

Data were collected in small groups (n lt 25 in each instance) during prear-ranged times Following arrival in a designated classroom each student wasinformed about the nature of the study given an opportunity to ask questions andprovided written informed consent prior to questionnaire distribution The sameresearchers were responsible for all data collection in Study 1 and used standardinstructions to reduce potential bias associated with multiple test administrators

5We chose to use the BREQ in this investigation as opposed to the BREQ-2 given our interest incurrent exercisers who theoretically could be motivated for various reasons While the BREQ-2 is con-sistent with SDT (Deci amp Ryan 2002) we felt that the inclusion of an amotivation subscale is likelymore relevant to sedentary populations or exercise initiates such as the sample used by Markland andTobin (2004) that consisted of patients referred by physicians to exercise on prescription programs

84 WILSON ET AL

Analyses

Data analysis proceeded in sequential stages First the distributional proper-ties of each variable were examined to determine their conformity with statisticalassumptions Second the suitability of the unidimensional integrated regulationmeasurement model and a multidimensional exercise regulation measurementmodel (BREQ items plus integrated regulation items Mullan et al 1997) weretested using confirmatory factor analytic (CFA) procedures with AMOS 50(Arbuckle 1997) Finally descriptive statistics reliability estimates (Cronbachrsquosα Cronbach 1951) and bivariate correlations were computed Conventionalstandards were specified in the measurement model analyses including correlat-ing latent factors loading manifest items exclusively on target latent factorsconstraining uniqueness values to zero and fixing a single item loading to unityto define the scale of each factor

Results

Preliminary Data Screening and Selection of an Estimator

Inspection of the data indicates that less than 68 of the data were missingon any one variable6 no out-of-range responses were observed and univariatedistributions approximated normality although multivariate kurtosis was evident(Table 1) Therefore maximum likelihood (ML) estimation procedures wereused with the incremental fit index (IFI) comparative fit index (CFI) and rootmean square error of approximation (RMSEA) to evaluate global model fit(West Finch amp Curran 1995) Although values indicative of satisfactory modelfit in hypothesis testing CFAs remain ambiguous (Hu amp Bentler 1999 MarshHau amp Wen 2004) it generally is accepted that a plausible model maximizesCFI and IFI values (gt90) and minimizes RMSEA (values lt 10)

CFA of the Exercise Motivation Measurement Models

The results of the CFA conducted to evaluate the fit of the integrated regula-tion items to a unidimensional measurement model indicate that the proposedmeasurement model differed from the reference model χ2(2 N = 207) = 3092p lt 01 However a desirable pattern of model fit estimates was observed (CFI =96 IFI = 96 RMSEA = 27 90 CI = 19ndash35) along with strong standardized

6Missing values were replaced using a mean imputation procedure This procedure involvedaveraging the scored items per construct for each participant and imputing the resultant value per caseprior to further data analysis The majority of the missing data was evident in the EMS items withless than 20 nonresponse error observed on other variables


Table 1

Distributional Properties of Manifest BREQ and Integrated Regulation Items Study 1

Latent subscales M SDSkew-ness

Kurto-sis FL SE

BREQmdashexternal regulationBecause other people say I should 065 093 094 -010 78 13Because others say I should 086 098 145 148 74 07Because others will not be pleased with

me 051 083 182 299 61 06I feel under pressure from others 126 125 064 -077 83 07

BREQmdashintrojected regulationI feel guilty when I donrsquot exercise 183 131 018 -114 66 07I feel ashamed when I miss exercise 139 119 058 -054 81 14I feel a failure when I havenrsquot exercised 188 116 000 -090 84 16

BREQmdashidentified regulationI value the benefits of exercise 346 078 -176 381 60 20Itrsquos important to me to exercise regularly 339 078 -134 121 86 10Itrsquos important to make an effort to exercise 346 086 -183 412 87 11I get restless if I donrsquot exercise regularly 209 130 -011 -120 65 09

Integrated regulationI exercise because it is consistent with life

goals 257 115 -067 -015 78 04I consider exercise to be part of my

identity 210 143 -016 -127 90 06I consider exercise a fundamental part of

who I am 205 143 -010 -128 94 07I consider exercise consistent with my

values 229 128 -035 -093 84 07BREQmdashintrinsic regulation

I exercise because itrsquos fun 265 116 -069 -032 92 05I enjoy my exercise sessions 285 110 -078 -025 84 05I find exercise a pleasurable activity 294 103 -087 031 86 05I get pleasuresatisfaction from exercise 319 097 -119 091 89 06

Note BREQ = Behavioural Regulation in Exercise Questionnaire FL = standardized parameterloading from CFA of the full measurement model SE = standard error from the CFA of the fullmeasurement model Mardiarsquos coefficient integrated regulation measurement model = 669Mardiarsquos coefficient BREQ and integrated regulation measurement model = 6559

86 WILSON ET AL

parameter loadings on the target factor and minimal evidence of over- or under-estimation of fitted correlations in the distribution of standardized residuals (0z gt |15|) Subsequent analyses of the expanded BREQ measurement modelincluding the integrated regulation items indicate that the model differed from thereference model χ2(142 N = 207) = 35751 p lt 01 Notwithstanding this obser-vation satisfactory estimates of global model fit (CFI = 92 IFI = 92 RMSEA =09 90 CI = 08ndash10) and parameter loadings were observed (Table 1) alongwith a distribution of standardized residuals showing little discrepancy betweenthe observed and implied covariance matrices (924 z lt |20| 0 z gt |30|)7

Descriptive Statistics Reliability Estimates and Bivariate Relationships

Descriptive statistics (Table 2) indicate that participants endorsed moreautonomous than controlled motives for exercise Reliability estimates ranged

Table 2

Descriptive Statistics Reliability Estimates and Bivariate Correlations From Variable Scores Study 1

Variable M SD 1 2 3 4 5 61 BREQmdashexternal regulation 082 082 822 BREQmdashintrojected regulation 128 078 36 813 BREQmdashidentified regulation 309 074 01 46 784 Integrated regulation 225 119 -03 37 70 925 BREQmdashintrinsic regulation 449 098 -18 24 67 65 936 EMSmdashintegrated regulation 449 098 -04 45 67 76 65 85

Note BREQ = Behavioural Regulation in Exercise Questionnaire EMS = ExerciseMotivation Scale Pearson correlations are reported in the lower triangle of thematrix Internal consistency estimates (Cronbachrsquos coefficient α) are reported on thediagonal Correlation matrix is based on pairwise comparisons and sample size isconsistent across each element in the matrix All rs gt |10| are significant at p lt 05two-tailed All rs gt |15| are significant at p lt 01 two-tailed

7A separate CFA was conducted using the four-factor BREQ measurement model proposed byMullan et al (1997) with the addition of a latent factor representing the EMS integrated regulationitems developed by Li (1999) Although the fit of this alternative measurement model was compara-ble with the data reported in Study 1 χ2(142 N = 207) = 35278 p lt 01 (CFI = 91 IFI = 91RMSEA = 09 90 CI = 07ndash10) the observed correlation matrix among the latent factors did notdemonstrate a quasi-simplex pattern given that introjected regulation was correlated more stronglywith integrated regulation (oslash = 51) than with identified regulation (oslash = 48)


from 78 to 93 in this sample An examination of the bivariate correlations(Table 2) reveals that adjacent constructs on SDTrsquos continuum were correlatedmost positively with one another and scores on the new integrated regulationsubscale were correlated most strongly with EMS integrated regulation subscalescores

Discussion

The overall purpose of Study 1 was to establish the structural validity of anexpanded BREQ measurement model including the new integrated regulationitems and to examine the degree of convergence between two measures of inte-grated regulation for exercise The results suggest that the integrated regulationitems appear to be factorially distinct from other BREQ constructs yet not mutu-ally exclusive which is consistent with SDTrsquos notion of a regulatory continuumMoreover the pattern of correlations suggests reasonable support for the conver-gence of integrated regulation scores with an existing measure of the sameconstruct and initial evidence of divergence from other motives assessed by theBREQ It is noted however that magnitude of the discrepancy between thevalidity coefficients is most pronounced when comparing controlled to autono-mous regulations

Study 2

The primary purpose of Study 2 is to evaluate the expanded BREQ measure-ment in an independent sample of exercisers The second purpose is to extend theconstruct validity evidence of the integrated regulation item scores by examiningtheir relationships with a portion of SDTrsquos nomological network A nomologicalnetwork represents an interconnected system of laws that comprise a theory(Cronbach amp Meehl 1955) One nomological network that SDT proposesincludes the satisfaction of competence autonomy and relatedness needs whichfoster the internalization process resulting in autonomous motives that are wellintegrated with the self (Deci amp Ryan 2002)

Method

Participants

Participants (N = 132 953 female Mage = 475 years SD = 823) weretraining to complete a marathon with a running club in central Canada Bodymass index (BMI) values approximated the healthy range for this age cohort(MBMI = 2434 kgm2 SD = 341 kgm2) However 297 of the sample reportedBMI values exceeding 2600 kgm2 A total of 620 of the sample were married

88 WILSON ET AL

or equivalent and 265 were single Most participants were employed full-time(758) 370 held bachelorrsquos degrees while another 340 held postgraduatedegrees or professional designations (eg LLB) Participants indicated that theyhad been involved with the running club for approximately 6 months (M = 541months range = 1ndash24 months)

Measures

Exercise motivation Participants completed the same version of the BREQand integrated regulation items that were used in Study 1

Psychological need satisfaction Participants completed three items designedto assess perceived competence (ldquofeeling competent and capable in the exercisesI attemptrdquo) autonomy (ldquofeeling autonomous and choiceful in the exercises Idordquo) and relatedness (ldquofeeling related and connected to the people I exercisewithrdquo) Following the stem ldquoTo what extent do you typically have these experi-ences when you exercise rdquo participants responded to each item on a 7-pointscale ranging from 1 (very little) to 7 (very much) While such indexes remaincontroversial (Crocker amp Algina 1986) single items exhibiting normal distribu-tions and representing the focal construct can be as useful as their multi-itemcounterparts (Gardner Cummings Dunham amp Pierce 1998) Previous researchhas linked greater scores on these items with autonomous exercise motives(Wilson et al 2002) and adjustment (Sheldon amp Elliot 1999) Considering thatthe items were developed for testing SDT-based hypotheses and exhibited nomajor distributional concerns in this sample (Table 3) their use in this studyseems justified given the absence of a suitable multi-item measure8

Procedure

Administrators who were involved with the running group distributed ques-tionnaires including an information letter and informed consent form to mem-bers of the group Participants were provided with a stamped envelope to returnthe questionnaire and a 50 response rate was observed

Analyses

Data analyses proceeded in sequential stages First assessment of the unidi-mensional and multidimensional exercise motivation measurement models were

8At the time of data collection the Psychological Need Satisfaction Exercise scale (PNSEWilson et al 2006) was not developed or available for use The PNSE is an instrument that wasdesigned specifically to measure perceptions of competence autonomy and relatedness experiencedin exercise contexts in line with SDT (Deci amp Ryan 1985 2002)


Table 3


Latent constructs M SDSkew-ness

Kurto-sis FL SE

BREQmdashexternal regulationBecause other people say I should 044 088 227 485 76 11Because others say I should 021 057 336 1205 78 11Because others will not be pleased with

me 018 052 354 1397 84 06I feel under pressure from others 052 096 222 464 93 07

BREQmdashintrojected regulationI feel guilty when I donrsquot exercise 158 120 052 -056 66 21I feel ashamed when I miss exercise 072 109 155 159 77 16I feel a failure when I havenrsquot exercised 152 134 045 -098 65 16



goals 316 107 -136 132 75 06I consider exercise to be part of my




I exercise because itrsquos fun 261 126 -066 -054 89 07I enjoy my exercise sessions 318 105 -115 039 81 07I find exercise a pleasurable activity 303 107 -109 054 82 09I get pleasuresatisfaction from exercise 333 095 -161 224 89 07

Note BREQ = Behavioural Regulation in Exercise Questionnaire Skew = univariate skewnessKurt = univariate kurtosis FL = standardized parameter loading from CFA Mardiarsquos coefficent(integrated regulation measurement model) = 1105 Mardiarsquos coeffecient (BREQ plus integratedregulation measurement model) = 9410

90 WILSON ET AL

evaluated consistent with Study 1 Model specification and identification proce-dures followed the conventional standards that were used in Study 1 Seconddescriptive statistics and bivariate correlations were computed for all study vari-ables Finally a series of simultaneous multiple regression analyses (SMRAs)were computed to determine the contribution of perceived competence auton-omy and relatedness to exercise regulations

Results


No out-of-range responses were observed in the sample data and mean impu-tation was used to replace the small amount (lt223) of nonresponse errorevident in the sample No grave concerns were evident in the datarsquos univariatedistributions (Table 3) although the external and identified regulation items wereleptokurtic and notable multivariate kurtosis was present ML estimation proce-dures were used in conjunction with the same fit indexes that were reported inStudy 1 given the distributions observed in this sample The results of the CFAsindicated that both the integrated regulation only χ2(2 N = 132) = 2637 p lt 01(CFI = 94 IFI = 94 RMSEA = 28 90 CI = 18ndash39) and the expandedBREQ χ2(142 N = 132) = 25382 p lt 01 (CFI = 93 IFI = 93 RMSEA = 0990 CI = 07ndash09) measurement models provided an adequate account of thesample data A pattern of strong parameter loadings on target latent factors (Table3) was observed along with a distribution of standardized residuals in both theintegrated regulation only (8333 z lt |10| 0 z gt |20|) and expanded BREQ(9707 z lt |20| 0 z gt |30|) measurement models which suggested no majordiscrepancies between the implied and observed covariance matrices


Descriptive statistics indicate that participants felt a strong sense of perceivedcompetence autonomy and relatedness in exercise and reported a more autono-mous (rather than controlled) motivational profile (Table 4) Reliability estimates(Table 4) ranged from 70 to 91 in this sample The bivariate correlations (Table4) indicate moderate to strong relationships between psychological need satisfac-tion scores scores on adjacent regulations were correlated more positively withone another than distal regulations and autonomous compared with controllingmotives were linked with greater perceptions of psychological need satisfactionin exercise


Regression Analyses Predicting Exercise Motivation From Perceived Psychological Need Satisfaction

The contribution of perceived competence autonomy and relatedness to pre-dicting SDTrsquos exercise regulations was examined using SMRA There are fiveseparate regression equations that were conducted using the BREQ and integratedregulation scores as criterion variables and perceived psychological need satisfac-tion scores as predictors The a priori probability value for each F test was set at01 to account for the number of models estimated (ie p = 055) The varianceinflation factor (043ndash088) and tolerance values (114ndash234) suggest the presenceof collinearity However when the condition index was high (ie gt100) only thevariance proportion values for competence and autonomy exceeded the 050threshold (Pedhazur 1997) Consequently structure coefficients (rs) were used tointerpret the results of the SMRA given their lack of distortion from collinearitywithin the sample data (Courville amp Thompson 2001)

Table 4


Variable M SD 1 2 3 4 5 6 7 81 PNSmdashcompetence 561 127 ndash2 PNSmdashautonomy 588 113 75 ndash3 PNSmdashrelatedness 561 125 34 31 ndash4 BREQmdashexternal

regulation 034 062 -26 -17 00 835 BREQmdashintrojected

regulation 127 098 -06 12 13 33 756 BREQmdashidentified

regulation 315 080 41 44 16 -18 18 707 Integrated regulation 269 113 48 55 29 -20 16 65 898 BREQmdashintrinsic

regulation 304 095 54 52 25 -19 -02 63 62 91

Note PNS = Psychological Need Satisfaction (items used by Wilson et al 2002)BREQ = Behavioural Regulation in Exercise Questionnaire Internal consistency reli-ability estimates (Cronbachrsquos coefficient α) are placed along the diagonal for allconstructs measured using multi-item subscales Correlation matrix is based on pair-wise comparisons with equivalent sample sizes across each element in the matrix Allrs gt |20| are significant at p lt 01 two-tailed

92 WILSON ET AL

The results of the SMRA (Table 5) reveal that perceived psychological needsatisfaction exerted stronger effects on autonomous (adjusted R2 values rangedfrom 19 to 31) than controlling (adjusted R2 value = 06) motives Moreover thedirection of the rs correlations indicates that perceived competence predicted lesscontrolling exercise regulations and greater intrinsic regulation whereas perceivedautonomy was the dominant predictor of identified and integrated regulations Per-ceived relatedness predicted both autonomous and controlling exercise motiveswhile perceived autonomy and competence were the strongest predictors ofintrojected regulation

Discussion

The purpose of Study 2 was to extend the construct validity evidence of theexpanded BREQ measurement model by examining the structural validity ofitem scores and relationships with aspects of one nomological network drawnfrom SDT (Deci amp Ryan 1985 2002) The results of the measurement modelanalyses support the tenability of the expanded BREQ measurement modelincluding the integrated regulation items developed in Study 1 in a communitysample of exercisers The regression analyses suggest that greater perceived psy-chological need satisfaction was associated positively with autonomous motivesincluding the new integrated regulation items in a manner consistent with SDTConsidering the fact that one argument stemming from SDT is that perceivedpsychological need satisfaction fosters the internalization of motives with the self(Ryan 1995) it is encouraging to observe scores on the new integrated regula-tion items corresponding to this theoretical premise

Study 3

The primary purpose of Study 3 is to examine the criterion validity of theintegrated regulation items developed across Studies 1 and 2 by examining thecontribution of extrinsic motives outlined within SDT to the prediction of physi-cal self-worth and exercise participation (Deci amp Ryan 1985 2002) Exercisebehavior was included because motives are theorized to determine behavior(Vallerand 2001) and previous research has indicated that autonomous extrinsicmotives predict more frequent exercise participation (Wilson et al 2004) Physi-cal self-worth was chosen to reflect the degree of positive evaluations felt by theself and to represent an index of emotional adjustment and well-being thataccording to Deci and Ryan (2002) should be associated more positively withmore autonomous motives The secondary purpose is to evaluate the stability ofintegrated regulation scores over a 2-week period This time frame was chosen toreduce the likelihood of true change confounding score stability (Pedhazur ampSchmelkin 1991) when analyzing the motivational data


Table 5

Simultaneous Multiple Regression Analyses Predicting Exercise Regulations Study 2

Criterion variablePredictor variables F(3 129)

AdjustedR2 B SE B β t rs

BREQmdashexternal regulationPNSmdashcompetence -16 06 -34 -254 -90PNSmdashautonomy 03 07 05 035 -58PNSmdashrelatedness 371 06 06 05 13 139 10

BREQmdashintrojectedPNSmdashcompetence -29 10 -38 -284 -21PNSmdashautonomy 31 12 36 273 41PNSmdashrelatedness 375 06 12 07 15 158 44

BREQmdashidentifiedPNSmdashcompetence 10 07 18 142 89PNSmdashautonomy 21 08 31 255 97PNSmdashrelatedness 1085 19 01 05 01 009 35

Integrated regulationPNSmdashcompetence 12 10 13 116 85PNSmdashautonomy 42 11 41 365 98PNSmdashrelatedness 1951 31 11 07 12 155 52

BREQmdashintrinsic regulationPNSmdashcompetence 25 09 34 296 95PNSmdashautonomy 21 10 25 218 91PNSmdashrelatedness 1943 31 05 06 06 076 44

Note PNS = Psychological Need Satisfaction BREQ = Behavioural Regulation inExercise Questionnaire Multiple R for BREQmdashexternal regulation = 289 multipleR for BREQmdashintrojected regulation = 291 multiple R for BREQmdashidentified regu-lation = 459 multiple R for integrated regulation = 569 multiple R for BREQmdashintrinsic regulation = 569 rs = structure coefficients (Courville amp Thompson 2001Thompson amp Borello 1985)p lt 01 two-tailed

94 WILSON ET AL

Method

Participants

A total of 89 female (Mage = 1935 years SD = 255 MBMI = 2164 kgm2SD = 259 906 lt 2499 kgm2) and 50 male (Mage = 2006 years SD = 382MBMI = 2367 kgm2 SD = 315 720 lt 2499 kgm2) undergraduate psychol-ogy students participated in exchange for course credit (3 participants did notprovide their gender) Substantial variability in physical activity behavior wasevident (female MMETS = 4738 SDMETS = 2809 male MMETS = 3632SDMETS = 2474 480 of the total sample engaged in gt 3 strenuous exercisesessions during the past 7 days)

Measures

Exercise motivation Participants completed 11 items assessing extrinsicmotives from the BREQ plus the 4 integrated regulation items used in Studies 1and 2

Exercise behavior Participants completed a modified version of the GodinLeisure Time Exercise Questionnaire (GLTEQ Godin amp Shepherd 1985) Thisinstrument assesses the frequency of mild moderate and strenuous exercisecompleted for 20 min or more per session over 7 days A global exercise scoreexpressed in METS (a unit representing the metabolic equivalent of physicalactivity in multiples of resting oxygen consumption) can be calculated using anequation proposed by Godin and Shepherd

[(mild times 3) + (moderate times 5) + (strenuous times 9)] (1)

Previous research has indicated that the GLTEQ (Godin amp Shepherd 1985) isunderstandable responsive to exercise behavior change and correlates in theexpected direction with other physical activity and fitness indexes (JacobsAinsworth Hartman amp Leon 1993) The global score (GLTEQ-METS) wasused to determine the current physical activity level of the participants in METSaccording to Godin and Shepard Participant responses to each GLTEQ item wereweighted by their corresponding MET value using Godin and Shepherdrsquos for-mula and aggregated into an omnibus exercise behavior score

Physical self-worth Participants completed six items from the physical self-concept subscale of the Physical Self-Description Questionnaire (PSDQ-PSCMarsh Richards Johnson Roche amp Tremayne 1994) The PSDQ-PSC providesa global evaluation of the degree of positive feelings a person holds about his orher physical self (Marsh 1996 Marsh et al 1994) Research supports the struc-tural validity of PSDQ-PSC scores and links higher PSDQ-PSC scores in

Σ


meaningful ways with other salient self-perceptions and physical activity behav-iors (Marsh 1996 Marsh et al 1994) Participants responded to each item on a6-point Likert scale ranging from 1 (false) to 6 (true) Consistent with previousrecommendations (Marsh et al 1994) consecutive PSDQ-PSC items were aver-aged to form item parcels (two itemsparcel) that were summed to form a totalPSDQ-PSC score

Procedure

Data were collected using the same procedures outlined in Study 1

Analyses

Data analyses followed sequential stages First descriptive statistics reliabilityestimates and bivariate correlations were computed Second SMRA estimated thecontribution of each extrinsic motive to the prediction of PSDQ-PSC and GLTEQ-METS scores Structure coefficients (rs) evaluated the contribution of the predictorset (ie extrinsic motives) to the criterion variables (ie physical self-worth andexercise behavior) following Courville and Thompsonrsquos (2001) recommendations

Results

Preliminary Analyses

No out-of-range scores were noted in the sample data and minimal departurefrom univariate normality was evident (skewness values ranged from 016 to089 kurtosis values ranged from 119 to 263) No consistent pattern was evi-dent in the missing data (27 across all study variables) Therefore missingvalues were estimated by averaging the scored items per construct for each par-ticipant and imputing the resultant value per case

Main Analyses

Descriptive statistics reveal that participants endorsed identified regulation toa greater extent than integrated introjected and external regulations for exercise(Table 6) Moreover participants reported high perceptions of their physical self-worth and low to moderate physical activity behavior Internal consistency reli-ability estimates ranged from 83 to 96 (Table 6) Bivariate correlations (Table 6)indicate that scores from adjacent extrinsic motives exhibited stronger relation-ships than distal regulations identified and integrated regulations were thestrongest correlates of physical self-worth and exercise behavior and higherintrojected regulation scores were linked positively with GLTEQ-METS scores

96 WILSON ET AL

Variance inflation (037 to 095) and tolerance (106 to 269) values observedimplied collinearity in the data However no two variance proportion factorsexceeded Pedhazurrsquos (1997) recommended threshold (050) when the conditionindex was high (gt10) The results of the SMRA (Table 7) reveal that integratedregulation contributed positively to predicting both PSDQ-PSC and GLTEQ-METS scores irrespective of the contributions from other extrinsic motives Anexamination of the rs reveals that integrated regulation is the strongest predictorof both PSDQ-PSC and GLTEQ-METS scores Of additional interest the vari-ance accounted for in each SMRA corresponds with Cohenrsquos (1992) criteriaindicative of medium effect sizes (15 lt R2 lt 35 Cohen 1992)

TestndashRetest Reliability Analysis

Intraclass correlation coefficients (ρ) were calculated using the two-waymixed-effect model to examine stability of BREQ and integrated regulation scoresover a 2-week period The results of this analysis suggest minimal fluctuation inscore stability (ρexternal = 78 ρintrojected = 78 ρidentified = 83 ρintegrated = 84ρintrinsic = 82 95 confidence intervals range from 70 to 88 respectively)

Table 6


Variable M SD 1 2 3 4 5 6BREQmdashexternal

regulation 062 082 85BREQmdashintrojected

regulation 184 097 23 85BREQmdashidentified

regulation 273 093 09 61 83Integrated regulation 176 124 14 54 75 93GLTEQndashMETS 4026 2641 -05 39 53 53 ndashPSDQndashPSC 424 113 -21 08 41 43 21 96

Note BREQ = Behavioural Regulation in Exercise Questionnaire GLTEQ = GodinLeisure Time Exercise Questionnaire PSDQndashPSC = Physical Self-Description Ques-tionnairendashPhysical Self-Concept Pearson correlations are placed in the lower triangleof the matrix Internal consistency estimates (Cronbachrsquos coefficient α) are placedalong the diagonal Correlation matrix is based on pairwise comparisons and samplesize is consistent across each element in the matrix All rs gt |10| are significant at p lt05 two-tailed All rs gt |20| are significant at p lt 01 two-tailed


Using Vincentrsquos (1995) guidelines for the behavioral sciences (ie acceptable ρsranging from 70 to 80) the BREQ and integrated regulation scores appear to bestable in order and magnitude across test periods

Discussion

The primary purpose of Study 3 was to examine the contribution of integratedregulation to the prediction of motivational consequences while considering theeffects of other extrinsic motives specified by SDT (Deci amp Ryan 1985 2002)The secondary purpose of Study 3 was to provide data attesting to the stability of

Table 7

Simultaneous Multiple Regression Analyses Predicting Motivational Consequences Study 3

Criterion variablePredictor variables

F(4 136)

Adjusted R2 B SE B β t rs

GLTEQndashMETS 1208BREQmdashexternal

regulation -452 245 -14 -184 -16BREQmdashintrojected

regulation 154 259 06 053 60BREQmdashidentified

regulation 641 354 22 181 90Integrated regulation 25 644 248 30 259 92PSDQndashPSC 1495BREQmdashexternal


regulation -031 010 -27 -302 14BREQmdashidentified

regulation 039 013 34 309 73Integrated regulation 32 032 009 37 369 77

Note BREQ = Behavioural Regulation in Exercise Questionnaire GLTEQndashMETS =Godin Leisure Time Exercise Questionnaire omnibus score PSDQndashPSC = PhysicalSelf-Description QuestionnairendashPhysical Self-Concept Multiple R for GLTEQndashMETSequation = 517 multiple R for PSDQndashPSC = 579 rs = structure coefficients (Courvilleamp Thompson 2001)p lt 05 p lt 01

98 WILSON ET AL

integrated regulation and BREQ scores over a 2-week interval The results of theSMRA indicate that integrated regulation has some predictive capacity for under-standing both exercise behavior and physical self-worth as motivational conse-quences beyond the effects of SDTrsquos other extrinsic motives The results of Study3 also support the stability of BREQ and integrated regulation test scores over ashort time interval This is encouraging given that although motivation as mea-sured by the expanded BREQ measurement model is theorized to be dynamic itis unlikely to change in the absence of intervention over short time frames

General Discussion

The purpose of this investigation was to evaluate the construct validity ofscores derived from a set of integrated regulation items designed to sit comfort-ably within the BREQ The results of Studies 1 and 2 suggest that including theintegrated regulation items in the expanded BREQ measurement model did notcompromise the instrumentrsquos theoretical fidelity or the structural validity ofBREQ scores Additional support for the convergentdivergent validity of theintegrated regulation items was evident in Study 1 based on observed correlationswith scores from Lirsquos (1999) EMS-Integrated Regulation subscale as well as evi-dence of criterion validity in Studies 2 and 3 that linked scores from theintegrated regulation items with indexes of need satisfaction exercise behaviorand physical self-worth that comprise a nomological network derived from SDT(Deci amp Ryan 1985 2002)

Corroborating the construct validity data for the integrated regulation itemsevidence of internal consistency reliability in each study and stability across a 2-week period in Study 3 reveals that scores derived from the integrated regulationsubscale and BREQ exhibit no obvious reliability concerns in these samplesCollectively this investigation suggests that the new integrated regulation itemsdisplay congeneric measurement properties and hold promise as an extension ofthe original BREQ to measure the full range of SDTrsquos extrinsic motives (Deci ampRyan 1985 2002)

This investigation provides initial evidence for the construct validity of theexpanded BREQ measurement model In addition to model fit estimates suggest-ing that the implied measurement model accounted for the observed data theresults of both CFAs suggest that including the integrated regulation items didnot compromise the theoretical underpinnings of the BREQ A major propositionput forth by Deci and Ryan (1985 2002) concerns the presence of a quasi-simplex pattern of associations whereby scores on measures of adjacent motivesproposed along SDTrsquos internalization continuum are associated more positivelythan distal motives

Given that theory and measurement are linked inextricably (Messick 1995)it is particularly encouraging that the pattern of observed relationships in the


expanded BREQ measurement model is consistent with SDT (Deci amp Ryan1985 2002) Notwithstanding this observation the CFAs indicate that two origi-nal BREQ items capturing external regulation (ldquoI exercise because others will notbe pleased with me if I donrsquotrdquo) and identified regulation (ldquoI get restless when Idonrsquot exercise regularlyrdquo) exhibited a pattern of standardized residuals exceeding|20| in each sample and the observed phi coefficients indicate considerable over-lap between latent factors in the expanded BREQ measurement model especiallywith respect to autonomous motives Given that previous research in Canadian(Wilson et al 2002) and British (Mullan et al 1997) samples has reported simi-lar findings with the BREQ future research may consider examining the degreeof item-content relevance and representation inherent in items comprising theexpanded BREQ measurement model (Dunn Bouffard amp Rogers 1999)

Consistent with Messickrsquos (1995) assertion that construct validity is an ongo-ing endeavor the current study extends psychometric evidence of the expandedBREQ measurement model in two ways First we examined and supported struc-tural validity of scores derived from the expanded BREQ measurement model inboth university- and community-based exercisers Second this investigation pro-vides initial evidence of internal consistency reliability and temporal stability ofscores compiled from participant responses to the expanded set of BREQ itemsincluding integrated regulation Evidence of stability across short time periods isconsidered important in initial stages of item development and evaluation whentrue change in the variables of interest (ie BREQ subscale scores) is not antici-pated on the basis of substantive theory (Pedhazur amp Schmelkin 1991)

Notwithstanding these observations the process of construct validation isongoing (Messick 1995) The composition of our samples in this investigationwas predominantly young active females which limits the external validity ofour data Future research may consider extending this work by testing theexpanded BREQ measurement model in different samples in which initiating andmaintaining exercise behavior is an important issue (eg older adults peoplewith chronic disease)

The pattern of relationships observed between subscale scores from theexpanded BREQ measurement model indexes of need satisfaction and motiva-tional consequences offer convergent evidence of construct validity Accordingto Deci and Ryan (1985 2002) contexts that support basic psychological needsfacilitate the internalization of motives into more autonomous forms which inturn promote enduring behavior and well-being The data in Study 2 indicate thatperceptions of competence and autonomy in particular demonstrated markedlystronger relationships with autonomous than controlled motives This is in linewith SDT given that relatedness is considered to be the catalyst instigating theprocess of internalizing behavioral regulation and therefore is unlikely to beassociated with integrated and intrinsic regulations that already have been assim-ilated with the self (Deci amp Ryan 2002) Although the results of Study 2 are

100 WILSON ET AL

informative and consistent with theory (Deci amp Ryan 1985 2002) the cross-sectional nature of the design prevents firm conclusions regarding the causalnature of the need-satisfactionmotivation relationship that future studies usingtrue experimental or prospective longitudinal designs may wish to explore

The predictive analyses conducted in Study 3 highlight an interesting patternof relationships between motives varying in their degree of internalization andselect consequences in the form of exercise behavior and physical self-worth Aninspection of the data presented in Table 7 makes it clear that integrated regulationscores exert the strongest predictive influence on each motivational consequencewhile controlling for the contributions of other extrinsic motives According toDeci and Ryan (1985 2002) autonomous regulations underpinned by greater psy-chological need satisfaction nurture more positive and enduring consequences in agiven domain compared with their controlling counterparts

Our investigation supports this assertion and extends the literature in twoways First this study expands Lirsquos (1999) work by providing empirical supportlinking integrated exercise regulations with exercise behavior and physical self-worth in a manner consistent with SDT (Deci amp Ryan 1985 2002) Second thisstudy complements previous research (Landry amp Solomon 2004 Pelletier et al2004 Wilson et al 2004) by providing additional support for the argument thatautonomous motives irrespective of their intrinsic or extrinsic orientation areinfluential in the regulation of important health behaviors (Deci amp Ryan 19852002) Nevertheless these data should be interpreted cautiously prior to replica-tion with more objective exercise behavior indexes as suggested by Pelletier etal (2004) to alleviate concerns regarding potential contamination of the criterionvalidity coefficients from common methods variance (Pedhazur amp Schemlkin1991)

A number of important observations are evident and noteworthy in thepresent investigation that warrant further elaboration to advance the study ofSDT in exercise contexts From a theoretical perspective the measurement andstatistical treatment of the data provide insights into the differential role playedby each extrinsic motive specified by SDTrsquos regulatory continuum (Deci amp Ryan1985 2002) This supports arguments made by Koestner and Losier (2002) ques-tioning the use of global motivational scores created by weighting then summingdiscrete points along SDTrsquos regulatory continuum given that such approachesmask important distinctions between sources of extrinsic motivation

In addition to extending the evidence base informing the validity of SDTrsquosmotivational continuum in exercise the present investigation also offers impor-tant practical information for health professionals who are interested in encourag-ing exercise participation and psychological well-being as part of an overallhealth-promotion program Results from this investigation suggest that it is thequality not the intensity of motivation that is the critical ingredient associatedwith frequent exercise participation and positive physical self-perceptions that


are considered an indicator of emotional adjustment and well-being (Biddle et al2000)

Health professionals should avoid structuring programs that promote feelingsof isolation from others in early stages of behavioral adoption that likely under-mine perceived relatedness and resist imposing unrealistic deadlines or goals onparticipation that likely undermine perceptions of competence and autonomyAlternatively health professionals who listen with empathy encourage self-initiated choices and confer positive feedback in a genuine manner will likelysupport the satisfaction of basic psychological needs that promote integrated reg-ulation that appears to be an influential process to consider in terms of bothexercise behavior and physical self-worth

In summary the purpose of the current investigation was to evaluate selectmeasurement properties associated with a set of preliminary integrated regulationitems developed using SDT (Deci amp Ryan 1985 2002) as a guiding frameworkThe results of three studies provide evidence supporting the construct validity ofthe new integrated regulation items when used alone and in conjunction with anexpanded BREQ measurement model The results also suggest that scores on theintegrated regulation items appear to be stable over a 2-week interval and con-verge in a theoretically expected direction with an existing measure of integratedregulation (Li 1999) perceptions of psychological need satisfaction and formsof extrinsic and intrinsic motivation aligning SDTrsquos regulatory continuum (Deciamp Ryan 1985 2002)

This investigation offers additional theoretical evidence for the multidimen-sional nature of extrinsic motivation and provides insight into the function ofmotives that vary in the level of self-determination facilitating their developmentby demonstrating links between extrinsic motives and select consequences Onthe basis of this study it appears that integrated regulation can be measured inexercise contexts as a related yet distinct form of extrinsic motivation within theexpanded BREQ measurement model Future research employing this instrumentin exercise settings appears to be justified

References

Arbuckle J L (1997) AMOS Version 36 [computer software] Chicago ILSmallwaters

Biddle S J H Fox K R amp Boutcher S H (2000) Physical activity and psy-chological well-being New York NY Routledge

Bouchard C Blair S N amp Haskell W L (2006) Physical activity and healthChampaign IL Human Kinetics

Cohen J (1992) A power primer Psychological Bulletin 112 1155ndash1159

102 WILSON ET AL

Courville S amp Thompson B (2001) Use of structure coefficients in publishedmultiple regression articles α is not enough Educational and PsychologicalMeasurement 61 229ndash248

Craig C L amp Cameron C (2004) Increasing physical activity Assessingtrends from 1998ndash2003 Ottawa Ontario Canada Canadian Fitness and Life-style Research Institute

Crocker L amp Algina J (1986) Introduction to classical and modern test theoryBelmont CA Wadsworth

Cronbach L J (1951) Coefficient alpha and the internal structure of tests Psy-chometrika 16 297ndash234

Cronbach L J amp Meehl P E (1955) Construct validity in psychological testsPsychological Bulletin 52 281ndash302

Deci E L amp Ryan R M (1985) Intrinsic motivation and self-determination inhuman behavior New York NY Plenum

Deci E L amp Ryan R M (2002) Handbook of self-determination researchRochester NY University of Rochester Press

Dunn J G H Bouffard M amp Rogers W T (1999) Assessing item contentrelevance in sport psychology scale-construction research Issues and recom-mendations Measurement in Physical Education and Exercise Science 315ndash36

Frederick-Recascino C M (2002) Self-determination theory and participantmotivation research in the sport and exercise domain In E L Deci amp R MRyan (Eds) Handbook of self-determination research (pp 278ndash294)Rochester NY University of Rochester Press

Gardner D G Cummings L L Dunham R B amp Pierce J L (1998) Single-item versus multiple-item measurement scales An empirical example Educa-tional and Psychological Measurement 58 898ndash906

Godin G amp Shepherd R (1985) A simple method to assess exercise behaviourin the community Canadian Journal of Applied Sport Science 10 141ndash146

Green-Demers I Pelletier L G amp Menard S (1997) The impact of behav-ioural difficulty on the saliency of the association between self-determinedmotivation and environmental behaviours Canadian Journal of BehaviouralScience 29 157ndash166

Hu L amp Bentler P M (1999) Cut-off criteria for fit indexes in covariancestructure analysis Conventional criteria versus new alternatives StructuralEquation Modeling 6 1ndash55

Jacobs D R Jr Ainsworth B E Hartman T J amp Leon A S (1993) Asimultaneous evaluation of 10 commonly used physical activity question-naires Medicine and Science in Sports and Exercise 25 81ndash91

Koestner R amp Losier G F (2002) Distinguishing three ways of being highlymotivated A closer look at introjection identification and intrinsic motivation


In E L Deci amp R M Ryan (Eds) Handbook of self-determination research(pp 101ndash123) Rochester NY University of Rochester Press

Landry J B amp Solomon M A (2004) African American womenrsquos self-determination across the stages of change for exercise Journal of Sport andExercise Psychology 26 457ndash469

Li F (1999) The Exercise Motivation Scale Its multifaceted structure and con-struct validity Journal of Applied Sport Psychology 11 97ndash115

Markland D amp Tobin V (2004) A modification to the Behavioural Regulationin Exercise Questionnaire to include an assessment of amotivation Journal ofSport and Exercise Psychology 26 191ndash196

Marsh H W (1996) Physical Self-Description Questionnaire Stability and dis-criminant validity Research Quarterly for Exercise and Sport 67 249ndash262

Marsh H W Hau K T amp Wen Z (2004) In search of golden rules Commenton hypothesis-testing approaches to setting cut-off values for fit indexes anddangers in overgeneralizing Hu and Bentlerrsquos (1999) findings StructuralEquation Modeling A Multidisciplinary Journal 11 320ndash341

Marsh H W Richards G E Johnson S Roche L amp Tremayne P (1994)Physical Self-Description Questionnaire Psychometric properties and multi-traitndashmultimethod analysis of relations to existing instruments Journal ofSport and Exercise Psychology 16 270ndash305

Messick S (1995) Validity of psychological assessment Validation of infer-ences from personsrsquo responses and performances as scientific inquiry intoscore meaning American Psychologist 50 741ndash749

Mullan E Markland D amp Ingledew D K (1997) A graded conceptualizationof self-determination in the regulation of exercise behavior Development of ameasure using confirmatory factor analysis procedures Personality and Indi-vidual Differences 23 745ndash752

Pedhazur E J (1997) Multiple regression in behavioral research Explanationand prediction Orlando FL Harcourt-Brace

Pedhazur E J amp Schmelkin L P (1991) Measurement design and analysisAn integrated approach Hillsdale NJ Lawrence Erlbaum

Pelletier L G Dion S C Slovinec-DrsquoAngelo M amp Reid R (2004) Why doyou regulate what you eat Relationships between forms of regulation eatingbehaviors sustained dietary behavior change and psychological adjustmentMotivation and Emotion 28 245ndash277

Pelletier L G Fortier M S Vallerand R J amp Briegravere N M (2001) Associa-tions among perceived autonomy support forms of self-regulation and per-sistence A prospective study Motivation and Emotion 25 279ndash306

Pelletier L G Fortier M S Vallerand R J Tuson K M Briegravere N M ampBlais M (1995) Toward a new measure of intrinsic motivation extrinsicmotivation and amotivation in sports The Sport Motivation Scale (SMS)Journal of Sport and Exercise Psychology 17 35ndash53

104 WILSON ET AL

Ryan R M (1995) Psychological needs and the facilitation of integrative pro-cesses Journal of Personality 63 397ndash428

Sheldon K M amp Elliot A J (1999) Goal striving need satisfaction and psy-chological well-being The self-concordance model Journal of Personalityand Social Psychology 76 482ndash497

Thompson B amp Borello G M (1985) The importance of structure coefficientsin regression research Educational amp Psychological Measurement 45 203ndash209

Vallerand R J (2001) A hierarchical model of intrinsic and extrinsic motivationin sport and exercise In G C Roberts (Ed) Advances in motivation in sportand exercise (pp 263ndash319) Champaign IL Human Kinetics

Vincent J (1995) Statistics in kinesiology (2nd ed) Champaign IL HumanKinetics

West S G Finch J F amp Curran P J (1995) Structural equation models withnonnormal variables Problems and remedies In R H Hoyle (Ed) Struc-tural equation modeling Concepts issues and applications (pp 56ndash75)Thousand Oaks CA Sage

Wilson P M amp Rodgers W M (2002) The relationship between exercisemotives and physical self-esteem in female exercise participants An applica-tion of self-determination theory Journal of Applied Biobehavioral Research7 30ndash43

Wilson P M amp Rodgers W M (2004) The relationship between perceivedautonomy support exercise regulations and behavioral intentions in womenPsychology of Sport and Exercise 5 229ndash242

Wilson P M Rodgers W M amp Fraser S N (2002) Examining the psycho-metric properties of the Behavioural Regulation in Exercise QuestionnaireMeasurement in Physical Education and Exercise Science 6 1ndash21

Wilson P M Rodgers W M Fraser S N amp Murray T C (2004) Relationshipsbetween exercise regulations and motivational consequences in university stu-dents Research Quarterly for Exercise and Sport 75 81ndash91

Wilson P M Rogers W T Rodgers W M amp Wild C T (2006) The psycho-logical need satisfaction in exercise scale Journal of Sport amp Exercise Psy-chology 28 231ndash251

80 WILSON ET AL

Despite the health benefits associated with regular physical activity (BouchardBlair amp Haskell 2006) population health data indicate that most adults remaininsufficiently active to offset chronic diseases or to promote quality of life (BiddleFox amp Boutcher 2000 Craig amp Cameron 2004) Considering these participationtrends a greater focus has been placed on understanding why people engage inphysical activity using theoretical frameworks that elucidate the processes shapinghealth behaviors (Biddle et al 2000) A theoretical approach holding considerableappeal for understanding multiple health behaviors including physical activity isself-determination theory (SDT Deci amp Ryan 1985 2002)

Deci and Ryan (1985 2002) proposed within SDT that motivation residesalong an internalization continuum reflecting the degree to which the behaviorhas been integrated with the self A personrsquos relative position along the internal-ization continuum is fostered by the extent to which social contexts satisfy basicpsychological needs for competence autonomy and relatedness (Ryan 1995)The distal ends of the internalization continuum are anchored by amotivation andintrinsic regulation Amotivation concerns a lack of intention to act and repre-sents a state akin to learned helplessness (Deci amp Ryan 2002) Converselyintrinsic regulation is the most autonomous form of motivation in which partici-pation is regulated by fun interest or the behaviorrsquos self-rewarding nature (Deciamp Ryan 2002) Previous research has indicated that amotivation is associatedwith a reduction in effort and importance ascribed to exercise (Wilson RodgersFraser amp Murray 2004) and greater dropout from sport (Pelletier FortierVallerand amp Briegravere 2001) In contrast intrinsic regulation is associated withgreater intention to continue exercising (Wilson et al 2004) persistent sportbehavior (Pelletier et al 2002) and enhanced physical self-worth (Wilson ampRodgers 2002)

Deci and Ryan (1985 2002) proposed four types of extrinsic motivationwithin SDT that vary considerably in terms of their integration with the selfExternal regulation and introjected regulation epitomize controlling internaliza-tions that motivate behavior via a desire to appease others avoid negative feel-ings or maintain conditional self-worth (Deci amp Ryan 2002) Greater reliance oncontrolling motives has been linked with dropout from sport (Pelletier et al1995) and lower physical self-worth in female exercisers (Wilson amp Rodgers2002) Identified regulation and integrated regulation represent more autono-mous extrinsic motives proposed by SDT (Deci amp Ryan 1985 2002) Identifiedregulation represents the lower limit of autonomous motivation in which partici-pation is regulated by goal values or the importance of behavioral outcomes(Deci amp Ryan 2002) Integrated regulation represents the most autonomousform of extrinsic motivation occurring when congruence exists between behav-ioral regulation and ldquopersonally endorsed values goals and needs that arealready part of the selfrdquo (Deci amp Ryan 2002 p 18) Previous research hasindicated that identified regulation is associated with more frequent exercise






82 WILSON ET AL


Study 1


Method

Participants




Measures




Procedure



84 WILSON ET AL

Analyses


Results







Table 1



Kurto-sis FL SE












86 WILSON ET AL




Table 2







Discussion


Study 2


Method

Participants


88 WILSON ET AL


Measures



Procedure


Analyses




Table 3



Kurto-sis FL SE












90 WILSON ET AL


Results








Table 4






regulation 304 095 54 52 25 -19 -02 63 62 91


92 WILSON ET AL


Discussion


Study 3



Table 5










94 WILSON ET AL

Method

Participants


Measures






Σ



Procedure


Analyses


Results



Main Analyses


96 WILSON ET AL




Table 6









Discussion


Table 7



F(4 136)










98 WILSON ET AL


General Discussion










100 WILSON ET AL











References





102 WILSON ET AL































104 WILSON ET AL

















82 WILSON ET AL


Study 1


Method

Participants




Measures




Procedure



84 WILSON ET AL

Analyses


Results







Table 1



Kurto-sis FL SE












86 WILSON ET AL




Table 2







Discussion


Study 2


Method

Participants


88 WILSON ET AL


Measures



Procedure


Analyses




Table 3



Kurto-sis FL SE












90 WILSON ET AL


Results








Table 4






regulation 304 095 54 52 25 -19 -02 63 62 91


92 WILSON ET AL


Discussion


Study 3



Table 5










94 WILSON ET AL

Method

Participants


Measures






Σ



Procedure


Analyses


Results



Main Analyses


96 WILSON ET AL




Table 6









Discussion


Table 7



F(4 136)










98 WILSON ET AL


General Discussion










100 WILSON ET AL











References





102 WILSON ET AL































104 WILSON ET AL













Measures




Procedure



84 WILSON ET AL

Analyses


Results







Table 1



Kurto-sis FL SE












86 WILSON ET AL




Table 2







Discussion


Study 2


Method

Participants


88 WILSON ET AL


Measures



Procedure


Analyses




Table 3



Kurto-sis FL SE












90 WILSON ET AL


Results








Table 4






regulation 304 095 54 52 25 -19 -02 63 62 91


92 WILSON ET AL


Discussion


Study 3



Table 5










94 WILSON ET AL

Method

Participants


Measures






Σ



Procedure


Analyses


Results



Main Analyses


96 WILSON ET AL




Table 6









Discussion


Table 7



F(4 136)










98 WILSON ET AL


General Discussion










100 WILSON ET AL











References





102 WILSON ET AL































104 WILSON ET AL












84 WILSON ET AL

Analyses


Results







Table 1



Kurto-sis FL SE












86 WILSON ET AL




Table 2







Discussion


Study 2


Method

Participants


88 WILSON ET AL


Measures



Procedure


Analyses




Table 3



Kurto-sis FL SE












90 WILSON ET AL


Results








Table 4






regulation 304 095 54 52 25 -19 -02 63 62 91


92 WILSON ET AL


Discussion


Study 3



Table 5










94 WILSON ET AL

Method

Participants


Measures






Σ



Procedure


Analyses


Results



Main Analyses


96 WILSON ET AL




Table 6









Discussion


Table 7



F(4 136)










98 WILSON ET AL


General Discussion










100 WILSON ET AL











References





102 WILSON ET AL































104 WILSON ET AL













Table 1



Kurto-sis FL SE












86 WILSON ET AL




Table 2







Discussion


Study 2


Method

Participants


88 WILSON ET AL


Measures



Procedure


Analyses




Table 3



Kurto-sis FL SE












90 WILSON ET AL


Results








Table 4






regulation 304 095 54 52 25 -19 -02 63 62 91


92 WILSON ET AL


Discussion


Study 3



Table 5










94 WILSON ET AL

Method

Participants


Measures






Σ



Procedure


Analyses


Results



Main Analyses


96 WILSON ET AL




Table 6









Discussion


Table 7



F(4 136)










98 WILSON ET AL


General Discussion










100 WILSON ET AL











References





102 WILSON ET AL































104 WILSON ET AL












86 WILSON ET AL




Table 2







Discussion


Study 2


Method

Participants


88 WILSON ET AL


Measures



Procedure


Analyses




Table 3



Kurto-sis FL SE












90 WILSON ET AL


Results








Table 4






regulation 304 095 54 52 25 -19 -02 63 62 91


92 WILSON ET AL


Discussion


Study 3



Table 5










94 WILSON ET AL

Method

Participants


Measures






Σ



Procedure


Analyses


Results



Main Analyses


96 WILSON ET AL




Table 6









Discussion


Table 7



F(4 136)










98 WILSON ET AL


General Discussion










100 WILSON ET AL











References





102 WILSON ET AL































104 WILSON ET AL














Discussion


Study 2


Method

Participants


88 WILSON ET AL


Measures



Procedure


Analyses




Table 3



Kurto-sis FL SE












90 WILSON ET AL


Results








Table 4






regulation 304 095 54 52 25 -19 -02 63 62 91


92 WILSON ET AL


Discussion


Study 3



Table 5










94 WILSON ET AL

Method

Participants


Measures






Σ



Procedure


Analyses


Results



Main Analyses


96 WILSON ET AL




Table 6









Discussion


Table 7



F(4 136)










98 WILSON ET AL


General Discussion










100 WILSON ET AL











References





102 WILSON ET AL































104 WILSON ET AL












88 WILSON ET AL


Measures



Procedure


Analyses




Table 3



Kurto-sis FL SE












90 WILSON ET AL


Results








Table 4






regulation 304 095 54 52 25 -19 -02 63 62 91


92 WILSON ET AL


Discussion


Study 3



Table 5










94 WILSON ET AL

Method

Participants


Measures






Σ



Procedure


Analyses


Results



Main Analyses


96 WILSON ET AL




Table 6









Discussion


Table 7



F(4 136)










98 WILSON ET AL


General Discussion










100 WILSON ET AL











References





102 WILSON ET AL































104 WILSON ET AL













Table 3



Kurto-sis FL SE












90 WILSON ET AL


Results








Table 4






regulation 304 095 54 52 25 -19 -02 63 62 91


92 WILSON ET AL


Discussion


Study 3



Table 5










94 WILSON ET AL

Method

Participants


Measures






Σ



Procedure


Analyses


Results



Main Analyses


96 WILSON ET AL




Table 6









Discussion


Table 7



F(4 136)










98 WILSON ET AL


General Discussion










100 WILSON ET AL











References





102 WILSON ET AL































104 WILSON ET AL












90 WILSON ET AL


Results








Table 4






regulation 304 095 54 52 25 -19 -02 63 62 91


92 WILSON ET AL


Discussion


Study 3



Table 5










94 WILSON ET AL

Method

Participants


Measures






Σ



Procedure


Analyses


Results



Main Analyses


96 WILSON ET AL




Table 6









Discussion


Table 7



F(4 136)










98 WILSON ET AL


General Discussion










100 WILSON ET AL











References





102 WILSON ET AL































104 WILSON ET AL















Table 4






regulation 304 095 54 52 25 -19 -02 63 62 91


92 WILSON ET AL


Discussion


Study 3



Table 5










94 WILSON ET AL

Method

Participants


Measures






Σ



Procedure


Analyses


Results



Main Analyses


96 WILSON ET AL




Table 6









Discussion


Table 7



F(4 136)










98 WILSON ET AL


General Discussion










100 WILSON ET AL











References





102 WILSON ET AL































104 WILSON ET AL












92 WILSON ET AL


Discussion


Study 3



Table 5










94 WILSON ET AL

Method

Participants


Measures






Σ



Procedure


Analyses


Results



Main Analyses


96 WILSON ET AL




Table 6









Discussion


Table 7



F(4 136)










98 WILSON ET AL


General Discussion










100 WILSON ET AL











References





102 WILSON ET AL































104 WILSON ET AL













Table 5










94 WILSON ET AL

Method

Participants


Measures






Σ



Procedure


Analyses


Results



Main Analyses


96 WILSON ET AL




Table 6









Discussion


Table 7



F(4 136)










98 WILSON ET AL


General Discussion










100 WILSON ET AL











References





102 WILSON ET AL































104 WILSON ET AL












94 WILSON ET AL

Method

Participants


Measures






Σ



Procedure


Analyses


Results



Main Analyses


96 WILSON ET AL




Table 6









Discussion


Table 7



F(4 136)










98 WILSON ET AL


General Discussion










100 WILSON ET AL











References





102 WILSON ET AL































104 WILSON ET AL














Procedure


Analyses


Results



Main Analyses


96 WILSON ET AL




Table 6









Discussion


Table 7



F(4 136)










98 WILSON ET AL


General Discussion










100 WILSON ET AL











References





102 WILSON ET AL































104 WILSON ET AL












96 WILSON ET AL




Table 6









Discussion


Table 7



F(4 136)










98 WILSON ET AL


General Discussion










100 WILSON ET AL











References





102 WILSON ET AL































104 WILSON ET AL














Discussion


Table 7



F(4 136)










98 WILSON ET AL


General Discussion










100 WILSON ET AL











References





102 WILSON ET AL































104 WILSON ET AL












98 WILSON ET AL


General Discussion










100 WILSON ET AL











References





102 WILSON ET AL































104 WILSON ET AL

















100 WILSON ET AL











References





102 WILSON ET AL































104 WILSON ET AL

















References





102 WILSON ET AL































104 WILSON ET AL












102 WILSON ET AL































104 WILSON ET AL



























104 WILSON ET AL












2006-Wilson Et Al. - Integrated Regulation and Exercise (JABR)

Documents

Transcript of 2006-Wilson Et Al. - Integrated Regulation and Exercise (JABR)