Kristof and Brown

PERSONNEL PSYCHOLOGY2000,53

PERCEIVED APPLICANT FIT: DISTINGUISHINGBETWEEN RECRUITERS' PERCEPTIONS OFPERSON-JOB AND PERSON-ORGANIZATION FIT

AMY L. KRISTOF-BROWNDepartment of Management and Organizations

University of Iowa

Two studies were conducted to assess whether recruiters form dis-tinguishable perceptions of applicant person-job (P-J) and person-organization (P-O) fit. The first study used repertory grid method-ology with actual recruiters and mock applicants to demonstrate thatknowledge, skills, and abilities are relied on more frequently to assessP-J fit, and values and personality traits more often to assess P-O fit.Study 2, which involved actual recruiters making decisions on appli-cants in a field setting, supported P-J and P-O fit perceptions as 2 dis-cemable factors. Study 2 also found that both types of perceived fitoffer unique prediction of hiring recommendations. Tkken together,these results present compelling evidence that recruiters discriminatebetween applicants' P-J and P-O fit during early interviews.

In studying organizational selection practices, researchers have com-monly identified two forms of fit that may be important to hiring de-cisions: (a) person-job (P-J) fit, or the match between an applicantand the requirements of a specific job; and (b) person-organization (P-O) fit, or the match between an applicant and broader organizationalattributes (Judge & Ferris, 1992; Rynes & Gerhart, 1990), P-J fit istypically operationalized as the match between employees knowledge,skills, and abilities (KSAs) and job demands (e.g., Caldwell & O'Reilly,1990; O'Reilly, Chatman, & Caldwell, 1991); whereas P-O fit has most

This study was funded by a grant from the SHRM Foundation. The interpretations,conclusions, and recommendations, however, are those of the author, and do not neces-sarily represent those of the Foundation. This study received the 1998 Ralph AlexanderDissertation Award from the Human Resources Division of the Academy of Management.

I thank the three anonymous reviewers and John Hollenbeck at Personnel Psychologyfor their assistance with crafting this manuscript. My deepest appreciation also to TimJudge, Sara Rynes, Murray Barrick, Suzanne Masterson, and Ken Brown for their intel-lectual and emotional support throughout the publication process. Thanks also to my dis-sertation committee who supported this study from its inception: Cindy Stevens, Ed Locke,Ben Schneider, Susan TJiylor, and Rhonda Reger; and to all of the doctoral students at theUniversity of Maryland and Michigan State University who helped with interviewing andcoding for this study.

Correspondence and requests for reprints should be addressed to Amy L. Kristof-Brown, Department of Management and Organizations, Henry B. Tippie College of Busi-ness, University of Iowa, Iowa City, IA, 52242; [email protected].

COPYRIGHT © 2000 PERSONNEL PSYCHOLOGY, INC.

643

644 PERSONNEL PSYCHOLOGY

frequently been studied as individual-organizational value congruence(Adkins, Russel, & Werbel, 1994; Cable & Judge, 1997; Chatman, 1989;Posner, 1992). To date, however, there has been no compelling empir-ical evidence demonstrating that, in actual practice, recruiters differ-entiate between these two concepts. In fact, the discriminant validitybetween the two types of fit has been questioned (Werbel & Gilliland,1999). Moreover, no research has shown whether the two types of fitperceptions provide unique value in predicting recruiting outcomes.

The current manuscript describes two studies designed to addressthese issues. The first study investigates whether recruiters' perceptionsof P-J and P-O fit are infiuenced by unique types of applicant charac-teristics. Finding different antecedents for each fit perception wouldprovide initial evidence of their distinctiveness to recruiters. In the sec-ond study, the factor composition of a perceived fit measure is examinedto determine if two discriminable factors (P-J and P-O fit) comprise re-cruiters' fit perceptions. Further, the relationships between perceivedP-O and P-J fit and hiring recommendations are tested to determine ifeach type of fit contributes uniquely to selection outcomes. Tkken to-gether, these studies provide a comprehensive examination of the va-lidity and utility of discriminating between perceived P-J and P-O fit inselection research.

Fit in Selection

The traditional focus in selection was on P-J fit, or hiring individu-als with particular skill sets to fill vacant positions (Cascio, 1991; Guion,1987). This focus began with the Army's use of cognitive ability testsfor soldier selection in World War I and prescribed a pattern for selec-tion research for over 75 years (Snow & Snell, 1993). However, as thebusiness world became increasingly complex and dynamic, many com-panies expressed a rising interest in fiexible staffing. To meet this need,scholars recommended broadening the focus to include P-O fit, as wellas P-J fit (Bowen, Ledford, & Nathan, 1991; Bretz, Rynes, & Gerhart,1993; Rynes & Gerhart, 1990). Unlike P-J fit, the characteristics oftenassociated with P-O fit are values and personality traits, and the focus ison fit with broad organizational attributes, rather than job-specific tasks.Current thinking supports the notion that both P-J and P-O fit are im-portant, and should be sought after during recruiting (Bowen et al., 1991;Judge & Ferris, 1992; Rynes & Gerhart, 1990).

There are many benefits to matching applicants at both the job andorganization level. Early writings on realistic job previews (RJPs) pre-sented a "matching model" in which realistic job and company infor-mation was presented to aid applicants in self-selecting into desirable

AMY L. KRISTOF-BROWN 645

positions (Phillips, 1998; Wanous, 1978, 1980). Schneider's attraction-selection-attrition model (1987) built on these ideas and emphasized thatfit assessments made by recruiters, as well as applicants, would reduceemployee turnover. Whether the process is lead by the applicant or re-cruiter, hiring for multiple types of fit has been shown to improve em-ployee attitudes and job performance, provide tighter control over newhires, and enhance organizational image (Judge & Ferris, 1992; Kristof,1996; O'Reilly et al., 1991).

Despite these benefits, there is a lack of empirical evidence support-ing the distinctiveness of perceived P-J and P-O fit to recruiters. Be-cause it is common for recruiters to spend less than 30 minutes with jobapplicants during early screening interviews, they have a limited amountof information on which to base their evaluations of fit. Werbel andGilliland (1999) suggest that this is likely to lead to extensive isomor-phism in the applicant characteristics used to assess various types of fit.Supporting this point, Bretz et al. (1993) reported that recruiters usedapplicant characteristics such as work experience and skills when de-scribing an applicant with a good P-O fit, even though these character-istics are typically considered within the domain of P-J fit. Thus, despiterecommendations for recruiters to consider both P-J and P-O fit in se-lection, it remains to be empirically demonstrated that recruiters candifferentiate between them.

Antecedents of Perceived Fit

One indication of whether recruiters can distinguish between P-J andP-O fit is if they assess each type of fit using different applicant charac-teristics. In a comprehensive review of the P-J fit literature, Edwards(1991) defined P-J fit as existing when an individual has the skills tomeet the demands of a specific job (demands-abilities fit), or when thejob meets the needs of the individual (needs-supplies fit). Although ap-plicants are undoubtedly concerned with finding a job that meets theirneeds, recruiters are most concerned with hiring employees who havethe requisite KSAs. This means that from the perspective of recruiters,demands-abilities fit is most relevant to filling a particular position (An-derson & Ostroff, 1997; Werbel & Gililland, 1999). Therefore, appli-cants' KSAs are expected to be highly salient to recruiters when assessingP-J fit.

However, results by Bretz et al. (1993) and Rynes and Gerhart (1990)suggest that KSAs may also have an important infiuence on recruiters'perceptions of how well an applicant fits in their company. Specifically,they report that recruiters cite characteristics such as GPA, job-relatedcoursework, work experience, articulateness, and interpersonal skills as


the basis for their judgments of P-O fit. Cable and Judge (1997) alsofound that GPA had a weak but significant relationship with recruiters'P-O fit perceptions.

Although KSAs are typically associated with P-J fit, recent atten-tion to contextual performance—that which extends beyond the bound-aries of any one job (Borman & Motowidlo, 1993)—may be establishingorganization-wide needs for particular KSAs. However, because contex-tual performance is more strongly infiuenced by personality traits andvalues (Motowidlo, Borman, & Schmit, 1997; Motowidlo & Van Scot-ter, 1994), it is unlikely that recruiters will make extensive use of ICSAsto judge P-O fit. KSAs have a stronger link with task-related perfor-mance (Motowidlo et al., 1997; Motowidlo & Van Scotter, 1994), and assuch, are more likely to be used by recruiters to judge P-J fit.

Hypothesis 1: Recruiters will use applicant KSAs more frequently to assessapplicants' P-J fit than P-O fit.

Edwards' definition of P-J fit emphasizes that fit occurs when a per-son meets the demands of a job, or a job fulfills the needs of a person.Muchinslgr and Monahan (1987) describe this type of fit as complemen-tary, which occurs when one entity completes another. Most studiesof P-O fit, however, have emphasized a different underlying relation-ship—one of fit based on similarity (Kristof, 1996). This type of fit isreferred to as supplementary fit, and it occurs when two entities (i.e.,person and organization) share similar characteristics, and because ofthat similarity are compatible.

Individual-organization value congruence is the most commonly stud-ied form of supplementary fit, and there is reason to believe that re-cruiters consider it when assessing P-O fit. Chatman (1989) suggestedthat value similarity is an important component of P-O fit because val-ues are "fundamental and enduring aspects of individuals and organiza-tions" (p. 339). As such, they guide both individual and organizationalbehaviors and decision-making. Recruiters should be more likely to rec-ommend hiring applicants who share their organizations' values becausethe extent to which employees share values determines the strengthof an organization's culture (Chatman, 1989; Schein, 1990), and cul-ture strength has been linked with organizational performance (Deal &Kennedy, 1982; Schein, 1990). At a more personal level, when peoplehave congruent value systems, they tend to view external stimuli and be-have in similar ways, which leads to easier communication and coordi-nation (Meglino & Ravlin, 1998). Judge & Ferris (1992) suggest thatrecruiters may use themselves as benchmarks to assess P-O fit. There-fore, if an applicant and recruiter appear to share the same values, the


recruiter is likely to judge the applicant as having a good fit. Adkins etal. (1994) found support for this relationship by showing that congru-ence between applicant and recruiter values predicted recruiters' per-ceptions of the applicant's P-O fit. In addition. Cable and Judge (1997)reported that perceived individual-organizational value congruence wasa stronger predictor of P-O fit perceptions than any demographic or hu-man capital variables. Thus, there is good evidence to suggest that re-cruiters consider applicants' values when evaluating their P-O fit.

Values are less likely to be related to recruiter perceptions of appli-cant P-J fit. Certain occupational values can be identified, such as thedesire to help people for a medical career or a belief in the sanctity ofjustice for a career in law. However, by the time a person has reachedthe step of applying for a job, their values are likely to have guided theminto selecting compatible professions (Dawis, 1991). For example, it isnot expected that hospital recruiters searching for doctors would see ap-plicants who vary considerably on the value of helping others. What may,however, distinguish an applicant is whether he or she has values rele-vant to a specific medical organization, such as a value for teamwork atthe Mayo Clinics (Roberts, 1999). Therefore, recruiters are unlikely touse values as a standard for P-J fit, but they are likely to use them toassess fit with their particular organization.

Hypothesis 2: Recruiters will use applicant values more frequently to as-sess applicants' P-O fit than P-J fit.

In addition to KSAs and values, research also suggests that person-ality traits are relevant to P-O fit (Bowen et al., 1991). Personality traitsare "stylistic consistencies in a person's social behavior" (Hogan, 1991,p. 875), so people with similar personalities are expected to interact andbehave in similar ways. The similarity-attraction paradigm suggests thatsimilarity leads to liking and increased attraction between individuals(Byrne, 1971). Therefore, recruiters have better interactions with, andare more attracted to, applicants with personalities similar to themselvesand others in their organization (Dipboye, 1992). Moreover, Schnei-der's (1987) attraction-selection-attrition framework suggests that re-cruiters prefer applicants whose dispositional characteristics are com-patible with organizational refiections of the founder's personality (e.g.,organizational goals, processes, structures, and culture). Once hired,people with compatible personalities are likely to enjoy higher satisfac-tion, reduced stress, and increased career success and tenure (e.g., Bretz& Judge, 1994; Ivancevich & Matteson, 1984). Because of increased at-traction and the benefits of personality compatibility, it is expected thatrecruiters will use personality traits to judge applicants' P-O fit.


Alternatively, several studies have shown that personality traits mayalso be an important determinant of fit with specific jobs (e.g., Jack-son, Peacock, & Smith, 1980; O'Reilly, 1977; Paunonen, Jackson, &Oberman, 1987). Personality traits fi-equently fall into the category of"other" when a job analysis isolates the KSAOs (knowledge, skills, abil-ities, and other things) that are necessary for satisfactory performanceon the job (Harvey, 1991). A growing body of evidence suggests thatcertain traits, such as conscientiousness and emotional stability, have ameaningful relationship with job performance (e.g., Barrick & Mount,1991). In addition, research on Holland's (1985) RIASEC typology ofvocational choice finds clear evidence that people with certain person-ality traits will be more satisfied and successful in particular vocations(Assouline & Meir, 1987; Spokane, 1985). Therefore, recruiters mayalso use personality information to make judgments of P-J fit. However,as in the case of values, because the selection into particular vocationstakes place prior to interviewing for specific jobs, recruiters are likely tointerview applicants with a restricted range of job-relevant personalitytraits. Therefore, despite the relevance of job-specific personality traits,it is predicted that recruiters will use applicants' personality traits morefi-equently to assess P-O fit than P-J fit.

Hypothesis 3: Recruiters will use applicant personality traits more fre-quently to assess applicants' P-O fit than P-J fit.

Although the preceding discussion centered around broad categoriesof characteristics (KSAs, values, personality traits), each category con-tains numerous specific characteristics that recruiters might use to judgeP-J or P-O fit. An interesting question arises concerning the extent thatrecruiters agree about which characteristics indicate P-J and P-O fit.Rynes and Gerhart (1990) suggested that applicant characteristics canbe partitioned into three categories: (a) idiosyncratic preferences, heldby individual recruiters; (b) organization-specific preferences, held bysome organizations but not others; and (c) general or universal prefer-ences, held by all organizations. Although evidence suggests that inter-viewers use characteristics from all three categories (Adkins et al., 1994;Bretz et al., 1993; Cable & Judge, 1997; Rynes & Gerhart, 1990), lit-tle is known about their relative frequency. Therefore, an exploratoryinvestigation into the extent of agreement among recruiters on specificcharacteristics used to assess P-J and P-O fit was conducted.


Factor Structure of Perceived Fit

For recruiters' perceptions of P-J and P-O fit to be unique, theyshould not only have different predictors, but they should also be sta-tistically distinct. O'Reilly et al. (1991) offer evidence that in a sampleof existing employees P-J and P-O fit are not strongly related (r = .16).However, their study assessed actual, rather than perceived fit (whereactual P-J fit was assessed using a KSA inventory and actual P-O fit wasassessed as the difference between employees' and the organization'svalue profiles), which may have contributed to the weak relationship be-tween the two constructs. No published study has examined whethertheir results also hold for recruiters' perceptions of the two fit constructs.

Because of the anticipated overlap in characteristics recruiters useto assess P-J and P-O fit, it is likely that perceived P-J and P-O fit willcovary more strongly in a recruiting setting than in a sample of existingemployees. In fact, due to the limited time that recruiters have to inter-act with applicants, it is questionable whether they perceive P-J and P-Ofit as two distinct concepts. Because halo bias infiuences recruiters to as-sign consistent ratings across dimensions for a single applicant (Cooper,1981), it is possible that recruiters do not discriminate between the twotypes of fit. Dipboye (1992) summarized, "the striking aspect of postin-terview judgments is the extent to which they are dominated by globalimpressions of the applicant's qualifications with little attention to vari-ations among different dimensions." (p. 119). Thus, it is possible thatan assessment of overall applicant fit dominates recruiters' perceptions,rather than distinct evaluations of applicants' fit with the job and the or-ganization. If this is the case, then there is little reason to investigateboth P-J and P-O fit in recruitment research, or to encourage recruitersto consider both when making hiring recommendations.

Despite these concerns, it is expected that recruiters are able to dif-ferentiate between P-J and P-O fit for two reasons. First, each refers toa distinct aspect of the environment—job versus organization. Regard-less of the firm in which it resides, the job of an entry-level consultantrequires a common set of KSAs. Likewise, an organization is character-ized by a set of core values, which transcend any one particular job titleand infiuence the behavior of all employees. Therefore, the job and or-ganization are conceptually distinct elements of the work environment;recruiters should be able to discriminate between them. Moreover, eachtype of fit may be infiuenced by a unique set of applicant characteristics(i.e., P-J fit by KSAs, and P-O fit by values). For these reasons, it ispredicted that a 2-factor model (representing P-J and P-O fit) will bet-ter fit the data of recruiters' fit perceptions than will a 1-factor model(representing overall applicant fit).


Hypothesis 4: With respect to recruiters' perceptions of fit, a 2-factormodel, representing P-J and P-O fit, will better fit the data than will a1-factor model, representing overall fit.

Perceived Fit and Interview Outcomes

A final way to determine the distinctiveness of perceived P-J and P-Ofit is to show that each makes unique contributions to explaining inter-view outcomes. Cable and Judge (1997) reported that perceived P-O fitpredicted hiring recommendations above and beyond objective qualifi-cations, such as GPA or work experience, and demographics. However,subjective evaluations typically have a stronger influence than objectivequalifications on interview outcomes (Gilmore & Ferris, 1989). There-fore, a stronger test of whether each type of fit explains unique variancein hiring recommendations is to compare the prediction of perceived P-O fit with that of perceived P-J fit.

As previously noted, perceptions of P-J and P-O fit are likely to berelated in the interview context. Adkins et al. (1994) encountered a simi-lar situation with perceptions of P-O fit and general employability, whichthey found correlated at r = .81. Despite this strong relationship, theywere able to show that both perceptions explained unique variance in in-vitations for a second interview. In light of Adkins et al.'s (1994) results,and because perceivedP-O and P-J fit describe conceptually distinct applicant evaluations, itis predicted that both will predict unique variance in hiring recommen-dations.

Hypothesis 5: Both perceived P-O and perceived P-J fit will explain uniquevariance in recruiters' hiring recommendations.

Study One Method

Participants

Thirty-one recruiters from the mid-Atlantic divisions of four con-sulting organizations participated in this study. These included eightrecruiters in information systems (IS) consulting and seven in finan-cial consulting firom Company A (650 people local; 67,000 worldwide),^

^The two consulting groups in Company A (IS and Financial) were coded separatelybecause there is little interaction between consultants in the two groups. Therefore, thetotal number of organizational units considered in later analyses was flve.


consulting firom Company D (2,200 IS consultants; 10,000 consultantsworldwide). All participants were fulltime consultants who recruited anaverage of 3.3 (SD = 6.6) days a year, and received very limited recruit-ing training (M = 1.7 hours, SD = 3.5 hours).

Within the constraint that only people in the office on the data col-lection dates could participate, all recruiters were invited to take part inthe study. Participation rates in Companies B and C were over 50% ofall recruiters, and in Company D, 75% of the senior IS recruiters wererepresented. Respondents fi^om Company A comprised less than 25%of their recruiters, so post hoc analyses were conducted to determineif participants were a representative sample. An analysis of variancebetween participants and nonparticipants indicated no demographic orexperience-related differences.

Procedure

Overview. This study used a technique called the repertory grid(Kelly, 1955) to solicit recruiters' schemas of the applicant characteris-tics associated with perceptions of good P-J and P-O fit. This techniquehas been used commonly in clinical settings and has recently gained pop-ularity in studies of organizational phenomena (Anderson, 1990; Brook,1986; Reger, 1990; Smith, 1986). The repertory grid methodology waschosen because it meets specifications outlined by Bretz et al. (1993) forhow to study recruiters' fit perceptions. Bretz et al. (1993) first advisedavoiding researcher-generated rating scales (e.g., commonly acceptedvalues or personality measures), because of the demand characteristicsthat they can induce. Instead they recommended allowing recruiters toarticulate their own conceptualizations of fit. The second specificationis that these articulations should be generated in the context of concretesituations or stimuli. This is necessary because people often have diffi-culty recalling decision criteria from abstract situations. The repertorygrid technique meets both of these requirements.

Operationally, the repertory grid involves a semistructured interviewin which respondents are asked to compare members of a standard setof stimuli. For each comparison respondents are asked to indicate thecharacteristics used to differentiate between the stimuli. For this studythe process involved asking recruiters to compare a standard set of appli-cants, and describe the characteristics that distinguished those with goodP-J and P-O fit from the others. The technique proceeded as follows:

Step 1. Participants watched a videotape containing abbreviatedmock interviews with seven MBA students, and reviewed their corre-sponding resumes. At the conclusion of each interview segment, partic-


ipants were provided time to take notes on the applicant they had justviewed.

Step 2. After the video of all seven applicants had been viewed, re-searchers met one-on-one with participants to ask them questions fromthe repertory grid protocol (see Appendix). Each participant was pre-sented with the numbers of three randomly selected applicants to deter-mine which had the best P-O fit and which had the best P-J fit. For ex-ample, the researcher might ask "Comparing applicants #4, #5, and #2,which of these people do you think is the best fit with your company?"After responses to that question were noted, the researcher would ask"Comparing these same three applicants, which is the best fit with thejob of an entry level consultant?" To aid recall of the applicants, partic-ipants were encouraged to refer back to their notes and the applicants'resumes. In addition, to promote visual recall they were provided witha photograph of each applicant. Tb ensure systematic bias did not resultfrom the question order, half the participants were asked about P-O fitfirst, and the other half about P-J fit first. Analysis of variance indicatedno effect of question ordering on the total number or type of applicantcharacteristics generated.

Step 3. After identifying the best fitting applicant in each category(P-J and P-O) the research asked each participant, "Can you please de-scribe the applicant characteristics that led you to make that choice?"They provided responses such as "This applicant has the most years ofwork experience," or "This applicant seems to like variety." When par-ticipants noted multiple characteristics, each was recorded for complete-ness. Finally, the researcher asked, "Please describe the opposite ofthose characteristics, to give me a clear sense of what you would con-sider not a good fit." (i.e., no work experience; enjoys routine).

Step 4. After comparing the first triad of applicants, the researcherthen presented a second triad of applicants and repeated the questions.This questioning strategy continued through a series of 10 applicanttriads. To reduce the possibility that particular applicant comparisonscould bias the results, four different triad sequences were used. Eachsequence was generated by randomly drawing three applicant numbers,replacing the numbers, and drawing three new numbers. This processcontinued until an applicant's number had been selected four times, atwhich point it was discarded to avoid overrepresentation. Analyses ofvariance suggested that the particular series of applicants did not signif-icantly infiuence the number or type of characteristics reported.


Materials

The repertory grid technique necessitates that participants base theirresponses on a common set of stimuli (Kelly, 1955). Therefore, video-taped interviews and resumes of seven MBA students were viewed byall recruiters. Tb bolster external validity, the videos contained actualfootage firom mock interviews conducted with another consulting com-pany. The materials are described in detail below.

Videos. Independent of this research project, mock interviews with 47MBA students were conducted by a nonparticipating consulting organi-zation and the MBA placement office of a large, public eastern univer-sity. After contacting a random sample of half of the students, 11 video-taped interviews were obtained (others had been lost or taped over).Each interview was coded by the researcher into "turns at talk" (one re-cruiter question and one applicant response) according to the method-ology used in Stevens and Kristof (1995). To ensure that comparable in-formation was presented across interviews, only those turns at talk con-taining similar recruiter questions were retained. This editing processresulted in 5-7 minute video clips for each student applicant.

Due to time restrictions set by the participating organizations, only 7of the 11 interview segments could be used. Tb ensure they representeda wide range of applicants, interviews from the seven most diverse stu-dents were selected. Diversity was evaluated in three ways. First, nineindependent coders viewed the interview segments and rated each ontraits Rynes and Gerhart (1990) and Bretz et al. (1993) had previouslyassociated with P-O fit (e.g., warmth, work experience, work ethic, in-novation). Second, the applicants' scores on Goldberg's (1992) measureof the Big Five personality traits (Costa & McCrae, 1988) and their val-ues, as assessed by the Comparative Emphasis Scale (CES; Ravlin &Meglino, 1987), were reviewed. All students had completed these mea-sures as part of participation in an unrelated research project. Finally,the demographic variables of gender, nationality, and MBA concentra-tion were also considered. Based on these criteria, four applicants mak-ing "redundant" impressions were eliminated. The remaining applicantswere 57% female, of three nationalities, had seven different MBA con-centrations, represented each of the dominant values on the CES, andhad a score on at least one of the Big Five personality traits that was onestandard deviation either above or below the MBA average.

Resumes. Based on the university's MBA resume book, one finance-related and one IS-related resume was aeated for each student. As withthe videos, maximum variability was desirable on the corresponding re-sumes. Therefore, two applicants were randomly selected to be por-trayed as highly experienced in the relevant field (IS or finance), three


as moderately experienced, and two as experienced in unrelated fields(marketing and transportation). GPAs ranged from 3.3 to 4.0, which rep-resented the high and low points for that year's MBA students applyingfor consulting positions. Honors and activities ranged from none to fourof the following: academic honors, religious or minority-group involve-ment, leadership positions, and volunteer work. These activities werebased on the student's actual resumes, so their content matched whatwas discussed during the interview.

Anafyses

Coding is typically the first step in analyzing qualitative data. Thecoding process for this study can be compared to simple content analysis(Krippendorf, 1980), with the basic intent being to categorize applicantcharacteristics into the codes of KSAs, values, personality traits, andothers. After the codes have been assigned, the data can be retrieved andanalyzed in "quasi-quantitative ways," including measuring the relativeincidence of different codes for perceived P-J fit and P-O fit. This code-and-retrieve process is one of the most commonly used in qualitativedata analysis (Coffey & Atkinson, 1996).

The repertory grid interviews generated a list of 119 characteris-tics associated with either, or both, perceived P-J fit or P-O fit. Toensure that each of these characteristics represented a unique appli-cant attribute, five independent raters were asked to group together anycharacteristics they judged to be sufficiently similar (i.e., outgoing andextraverted). Groupings were only created if the characteristics werejudged to be similar by at least three of the five raters. All other charac-teristics were left as distinct—a conservative method for assessing sim-ilarity (Bretz et al., 1993). This process resulted in a reduced list of 62unique applicant characteristics (see Table 1). The fact that these stim-ulus materials generated 62 distinct fit-related characteristics, as com-pared to 45 generated using actual applicants in Bretz et al. (1993), sug-gests that the videos and resumes allowed recruiters to view a realisticrange of applicant characteristics.

Three additional coders, blind to the study's hypotheses, were thentrained to place each of the 62 characteristics into the categories of KSA,value, personality trait, or other attribute. Miles and Huberman (1994)recommend this strategy of creating the categories prior to reading thedata when the research questions are already established. To promotereliable coding, the coders were trained using established definitions andexamples for each category. Because distinguishing between categorieswas anticipated to be difficult in some cases (e.g., someone who statesthat they value hard work may also be described as conscientious), the


TABLE 1

Content Categorizations of Perceived P-J and P-O Fit Characteristics

Category 1: Knowledge, skills, and abilitiesConsulting skills and experiencel&chnical/computer skills and experienceQuantitative skills and experienceProblem-solving skills and experienceAnalysis skills and experienceProcess focused skills and experienceClient service experienceManagerial skills and experienceLeadership skills and experiencel&am skills and experienceBroad, general work skills and experienceHigh level of work experienceBreadth of life experiencesCommunication skills; Listening skillsPeople/human interpersonal skillsOrganizational skills; Time managementExperience in Big 6 consulting firmExperience with small companiesRuns own business; Entrepreneurial experienceNiche skills or majorRelevant major (IS or Finance)Smart; High GPA; SharpCPASubstantive comments on experienceRealistic/knowledgeable view of industry

Category 2: PersonalityBig picture focus; Results-orientedSees the gray areas; DeepPolitical savvy; Smooth talker; DiplomaticExtroverted; Dynamic; EnthusiasticLikable; Friendly; Personable; WarmSuccess oriented; Strives for excellence; Ikkes pride in workAmbitious; Aggressive; Likes challengesConfident; Can sell him or herselfRealistic view of selfProfessional; Mature; PoisedPrepared for interviewFocused; Goal oriented; Has 5-year plan


TABLE 1, continued

Adaptable; Flexible; VersatileSelf-motivated; Ikkes initiativeTfeam player; HelpfulPositive attitudeIndependent; Doesn't need hand-holdingEasy to label; PredictableDirect; StraightforwardSincere; HonestInquisitive; CuriousCreativePlannerLogicalThorough; Detail-oriented; PerfectionistTlikes responsibility; Follow-throughRelaxed; Easy goingPersonally fulfilled; Comfortable with self; Well-adjustedDown-to-earth; Not flashyAll American; Joe 6-pack

Category 3: ValuesInterested in work and extra effort; Good work ethicInterested in continuous learningWants feedback; Wants to improve selfLikes pressure and deadlinesLikes variety and change, not routine or rules

Category 4: OthersGoals align with ours (consulting)Young

coders were instructed to use the definitions provided and be consistent.^Perfect consensus existed on 89% of the categorizations; the remaining11% were agreed upon by at least two of the three coders.

•̂ KSAs were categorized using Harvey's definitions (1991, pp. 75-76): Knowledge—"abody of information applied directly to the performance of a function... Skills—learned,observable, psychomotor acts... [Abilities]—psychological characteristics (e.g., numericalability, spatial ability) considered to vary or differ across individuals... relatively stableand enduring personal characteristics that defy efforts at change (e.g., general intelli-gence. . . ) ." Values were defined as: "An enduring belief that a specific mode of conductor end-state of existence is personally or socially preferable to an opposite or conversemode of conduct or end-state of existence (Rokeach, 1973, p. 5), and "beliefs, attitudes.


Once eacb characteristic had been categorized as a KSA, value, orpersonality trait, the coders analyzed the responses from each repertorygrid interview to generate frequency data for each interview. This in-cluded the total number and type of specific characteristics reported byeach recruiter. Intercoder correlations ranged from .69-.96 on the to-tal number of characteristics per interview, and average pairwise Kappas(Cohen, 1960), calculated on the reported presence or absence of a char-acteristic in each interview, were .78 for P-J fit and .73 for P-O fit. Thesenumbers indicate that there was a reasonably high level of agreementbetween coders on the characteristics in each interview. A characteris-tic was included in the quantitative analyses only if two or more codersagreed that it was present.

Study One Results

Antecedents of Perceived Fit

Hypotheses 1 proposed that recruiters would rely more heavily onKSAs to judge an applicant's P-J fit than P-O fit. Hypotheses 2 and 3predicted that values and personality would be used more frequently toassess applicants' P-O fit than P-J fit. To test these hypotheses the to-tal number of KSAs, values, and personality traits reported by each re-cruiter as indicators of P-J and P-O fit was calculated using the informa-tion provided by the coders (see Tkble 2). Because the same recruitersprovided data on both types of fit, a paired-samples t-test was used tocompare the mean values for P-J and P-O fit.

Results indicate that 100% of recruiters mentioned KSAs as indica-tors of both P-J and P-O fit. However, as predicted, the mean number ofKSAs reported as indicators of P-J fit (M = 5.77 per interview) was sig-nificantly higher than the mean number reported for P-O fit (M = 3.16),d = 1.06; t(30) = 5.92, p < .01. Values were mentioned as indicators ofP-O fit by 65% of recruiters, and as indicators of P-J fit by 39%. As hy-pothesized, the mean number of values reported for P-O fit (M = 1.10)was significantly higher than the mean number of values mentioned forP-J fit (M = .58), d = -.39; t(30) = -2.19, p < .05). A similar re-sult was found for the mean number of personality traits (M = 5.19 forP-O fit, M = 2.13 for P-J fit, d = -1.02; t(30) = -5.67, p < .01), with100% of recruiters using them to assess P-O fit versus 81% for P-J fit.

needs, interests, preferences, standards or criteria and the conception of the desirable"(Dawis, 1991, p. 837). Personality traits were defined by Hogan (1991, p. 875): "a person'ssocial reputation and inner nature," including traits or "stylistic consistencies in a person'ssocial behavior." Anything that didn't fall into these categories was, by default, placed inthe "other" category.


a

I I

00 00 t^

00 f^

" 4

l l

7 5 |

m.S so,

11

2 '-5a §-2

si2

Pii

1 5


Effect sizes indicate that the mean differences are strongest for KSAsand personality traits, with differences approximating one standard de-viation. The difference for values is smaller, but this may be due theoverall low base rate of values reported by recruiters. Together, theseresults support Hypotheses 1-3, indicating that recruiters rely differen-tially on various types of applicant characteristics to assess P-J fit andP-O fit.

It should be noted that although comparisons can be made betweenthe number of KSAs, values, and personality traits used to judge onetype of fit versus the other, comparisons should not be made betweentypes of characteristics within each fit category. For example, althoughthe relative number of KSAs used to judge P-O versus P-J fit can becompared, it is not appropriate to contrast the number of KSAs versusvalues used to judge P-O fit. This is because of the varying base rates foreach category in the recruiter-generated list of characteristics. Havingsaid this, the high frequency with which personality traits were usedto assess both P-J and P-O fit is interesting to note. It appears thatrecruiters made extensive judgments about applicant personality, evenfrom the short videotaped interview.

Exploratory analyses into the extent of recruiter agreement on thespecific characteristic associated with P-J and P-O fit were also con-ducted. Based on Rynes and Gerhart's (1990) definitions, a characteris-tic was categorized as: (a) organization-specific if it was endorsed by amajority of recruiters in a minority of the firms; (b) a universal indicator,if endorsed by a majority of recruiters in a majority of the organizations;and (c) an idiosyncratic preference if not agreed upon by a majority of re-cruiters in any of the firms. Of the characteristics reported as indicatorsof P-J fit, 74% were idiosyncratic preferences, 22% were organization-specific, and only 4% were universal. Percentages were similar for char-acteristics associated with P-O fit: 73% idiosyncratic, 23% organization-specific, and 3% universal.

Study Two Method

Participants

Participants included 46 recruiters^ fi^om Company C and CompanyA, who had not participated in Study 1. Their time spent each year

^Because some recruiters evaluated more than one applicant, the possibility that sys-tematic differences across recruiters influenced the results was considered. An analysis ofvariance using recruiter as the grouping variable resulted in a significant F-ratio for per-ceived P-O fit, F(46, 79) = 1.89, p < .05. Post hoc comparisons indicated one recruiter,who evaluated one applicant, reported aberrantly low scores. This recruiter was removedfrom all analyses, resulting in 46 recruiters and a total N = 79.


recruiting (M = 2.5 days; SD = 3.5) and in recruiting-focused training(M = 4.0 hours, SD = 8.4) was similar to that of the participants inStudy 1.

Procedure

After conducting actual 30-minute screening interviews, the partici-pants were asked to evaluate at least one "successful" and one "unsuc-cessful" applicant. This procedure is similar to one used by Adkins et al.(1994), Bretz et al. (1993), and Cable and Judge (1997). Recruiters eval-uated a mean of 1.7 applicants, making 79 applicants the total sampleavailable for analysis. The applicants represented 35% of those inter-viewed by Company C. No estimate could be obtained on the percent-age from Company A. Of the interviews, 75% were conducted on-siteand 25% on-campus. Because these recruiters were authorized to eval-uate the applicants, but not make final selection decisions, their hiringrecommendations were the dependent variable.

Measures

Perceived applicant P-Oflt. Four items based on measures by Adkinset al. (1994) and Cable and Judge (1997) were used to indicate recruiters'perceptions of applicants' P-O fit (To what degree does this applicant fitwith your organization?; To what extent is this applicant similar to other[insert company] employees?; To what extent will other employees thinkthis candidate fits well in your organization?; How confident are youthat this applicant would be compatible with your organization?). Itemresponses were obtained on a 5-point Likert scale, with (1 = not at all)to (5 =completefy) anchors. The scale had an alpha reliability of .96.

Perceived applicant P-J fit. Three items were included to measurerecruiters' perceptions of applicants' P-J fit (To what extent does thisapplicant fit the demands of the job?. To what extent will other employ-ees think this candidate is qualified to do this job?. How confident areyou that this applicant is qualified for this job?). Responses were notedon the 5-point scale described above and the three items had an alphareliability of .94.

Hiring recommendations. A 4-item scale (a = .96) was used to assessrecruiters' hiring recommendations. Similar to the measure used byCable and Judge (1997), recruiters were asked to report the likelihoodthat they would recommend a particular applicant be hired and thatthe applicant would actually be hired. In addition they were asked torate the confidence with which they would recommend that applicantfor employment and their confidence that the applicant would succeed


TABLE 3

Means, Standard Deviations, and Correlations Between Perceived Fitand Hiring Recommendations

Variable Af SD 1 2 3

1. Perceived P-J Fit2. Perceived P-O Fit3. Hiring recommendations

3.83.83.6

.82

.811.0

(.94).72*.86*

(.96).79* (96)

•n = 79 p<.05

in their organization. All responses were given on a 5-point Likert scaleranging from (1 =not at all) to (5 =extremefy).

Study Two Results

The means, standard deviations, and correlations between perceivedP-J fit, P-O fit, and hiring recommendations appear in "Eible 3. Relia-bilities for each scale are located along the diagonal. As expected, butin contrast to results by O'Reilly et al. (1991), perceived P-J and P-Ofit were highly correlated (r = .72). Perceived P-J fit was slightly morecorrelated with hiring recommendations than P-O fit.

Factor Structure of Perceived Fit

Confirmatory factor analysis (CFA) was conducted using EQS (Bent-ler, 1985), to determine whether a 2-factor structure (P-J and P-O fit) or1-factor (overall fit) structure better fit the data. The results show thatthe data fit the 2-factor model with covarying factors fit the data well (x^[13, AT = 79] = 31.09, p < .01; NNFI = .954; CFI = .972). Goodness offit indices were substantially lower for the 1 factor-model, obtained bytreating all seven items as indicators of an overall fit perception (x^ [14,AT = 79] = 130.52; hfNFI = .727; CFI = .818). To evaluate Hypothesis4, a x^-difference test was conducted to determine whether the x^ valuefor the 2-factor model was significantly lower than that for the 1-factormodel. This test indicated that the difference in x^ for the two modelswas statistically significant (Ax^ = 99.43, Ad/ = 1; p < .05), with the2-factor model being a better fit to the data. These results suggest that,despite their high correlation, perceived P-J and P-O fit are best treatedas separate variables.

Perceived Fit and Interview Outcomes

Hypothesis 5 predicted that both perceived P-J fit and perceived P-Ofit would explain unique variance in recruiters' hiring recommendations.


TABLE4Prediction of Hiring Recommendations by Perceived P-J and P-O Fit

Hiring recommendations

Control variables:GPAInterview siteCompany

Fit perceptions:P-J FitP-O Fit

.08-.10-.38

.63

.35

Step Variables 0 R^ AH^ df

.12 .12* 3, 78

.80 .68* 5, 78

Notes: 0 = standardized regression coefficient•p < .05

After controlling for the applicant's GPA, interview site (on campus oron site), and the company (A or C), hiring recommendations were re-gressed on the two perceived fit variables (see Tkble 4). Of the controlvariables, only company explained a significant amount of the variancein the criterion. Applicants to Company A were more likely to be recom-mended for hire than applicants to Company C, which may be explainedby the fact that Company C was a small company needing to fill fewerpositions. In addition to the company control variable, recruiters' fitperceptions as a group predicted hiring recommendations, AR'^ = .68,F(5, 78) = 58.49, p < .01. In support of Hypothesis 5, the regressionweights for both perceived P-J fit (P = .63; p < .05) and perceived P-Ofit (P = .39; p < .05) were significant, indicating that each explainedunique variance in hiring recommendations. Perceived P-J fit explainedthe greatest amount of variance; after controlling for perceived P-O fit,R^ increased 21% with the addition of perceived P-J fit. Alternatively,R^ increased only 5% when perceived P-O fit was added after control-ling for perceived P-J fit.

Discussion

Two studies were conducted to assess whether recruiters form dis-tinguishable perceptions of applicants' P-J and P-O fit. Tbgether thesestudies make an important contribution to the recruiting literature byempirically demonstrating that perceived P-J and P-O fit differ in termsof their antecedents and factor composition, and both offer unique pre-diction of recruiters' hiring recommendations.

Results of Study 1 show that recruiters judge P-J and P-O fit usingdifferent types of applicant characteristics. Although some recruitersused applicant KSAs, values and personality traits to judge both P-Jand P-O fit, the relative reliance on each type of characteristic differed


depending on which type of fit was being assessed. As predicted, KSAswere mentioned more firequently as indicators of P-J fit than P-O fit,and personality traits and values were mentioned more frequently asindicators of the latter.

Although it is not statistically meaningful to compare the frequencyof use for one type of trait versus another for judging P-J or P-O fit, it isinteresting to note the high number of personality assessments recruitersmade. This may be because personality traits may be inferred fromnonverbal behaviors and appearance, as well as applicant responses.In addition, Rothstein & Jackson (1984) proposed that recruiters hold"implicit personality theories" that lead them to judge personality byassessing one trait, and then assuming the existence of other traits basedon their perceived covariation with the initial trait. For these reasons,recruiters may make extensive personality assessments after even briefencounters with applicants, and use these assessments to judge their P-Jand P-O fit.

Exploratory analyses revealed that within these broad categories ofapplicant characteristics, recruiters seldom agreed on the specific indi-cators of a good fit of either type. Their perceptions were dominatedby idiosyncratic ideas of what made an applicant a good P-J or P-O fit.These results support Adkins et al.'s (1994) conclusion that recruitersoften vary in the criteria they use to judge an applicant's fit, and areparticularly noteworthy because all recruiters viewed the same set ofapplicants in the current study. One explanation for the high relianceon idiosyncratic criteria is the limited training that participants in thisstudy had received on interviewing. Other studies have shown that re-cruiter training may reduce idiosyncratic biases (Maurer & Fay, 1988;Stevens, 1998). Another possibility is that, as suggested by Judge andFerris (1992), recruiters may use themselves as an individualized bench-mark to judge P-O fit. Van Vianen (2000) suggests that this might notnecessarily be harmful, as value similarity between newcomers and fo-cal others, such as a supervisor or peer (what she termed person-personfit) has a stronger impact on organizational commitment than does sim-ilarity with the organization's values. Finally, it should be noted that theoperationalization of "idiosyncratic" was fairly liberal, with a high levelof agreement (50% of recruiters in firm) required to categorize some-thing as organization-specific. Thus, there may have been a bias towardfinding a relatively high proportion of idiosyncratic criteria.

Whereas Study 1 investigated antecedents. Study 2 assessed the fac-tor structure and predictive validity of perceived P-J and P-O fit. In thisstudy the two constructs were highly correlated (r = .72), which contrastswith the weak correlation between actual P-J and P-O fit reported byO'Reilly et al. (1991). The high correlation was not surprising, however.


because of the overlap in predictors and the short time frame of firstinterviews. TWo CFAs were conducted to compare a 2-factor model, inwhich perceived P-O and P-J fit were distinct, covarying factors, witha 1-factor model of perceived overall fit. The 2-factor model fit thedata well, significantly better than did the 1-factor model. Thus, thereis evidence that recruiters' fit perceptions form two discriminable fac-tors. Results of Study 2 also indicate that both perceived P-O and P-Jfit explained unique variance in recruiters' hiring recommendations.

These results provide evidence that perceived P-J and P-O fit aredistinct constructs. However, there is no question that the two variablesare highly related in the minds of recruiters. Their bivariate correlationis high and recruiters make some use of KSAs, values and personalitytraits when assessing both types of fit. Researchers of distributive andprocedural justice have faced a similar challenge. These constructs havebeen reported to correlate as highly as 0.77, but are still argued to beconceptually distinct (Tyler, 1984). Folger (1987, p. 151) concluded thatempirical evidence supports that "(i) distributive and procedural justiceare interrelated, such that perceptions of one can infiuence perceptionsof the other; and (ii) distributive and procedural justice can neverthelessbe evaluated on independent grounds, such that the two types of fairnessneed not coincide." The results reported in the current paper support asimilar conclusion about recruiters' perceptions of P-J and P-O fit.

When comparing perceived P-J and P-O fit, P-J fit was found to havethe stronger relationship with recruiter recommendations, which is con-sistent with prior research conducted on first interviews (Bretz et al.,1993; Rynes & Gerhart, 1990). Because many first interviews are usedas screening mechanisms, the initial concern should be to eliminate ap-plicants who do not meet the job requirements. Even though the partici-pants in this study were not fulltime recruiters, their objective was to findapplicants who could fill position vacancies defined by particular skillrequirements. In all cases the applicants would complete an additionalinterview with another recruiter before being made a job offer. There-fore, perceived P-O fit may be used later to distinguish among applicantsalready judged to be qualified for the job. Future research which tracksapplicants through all stages of an organization's hiring process shouldbe conducted to assess whether the impact of perceived P-J and P-Ofit varies changes over time. Nevertheless, because both perceived P-Jand P-O fit explained unique variance in hiring recommendations in thisstudy, there is evidence that P-O fit is considered by recruiters even atthe earliest stages of the hiring process.


Limitations and Strengths

These studies have a number of limitations. First, the categoriza-tion scheme used to distinguish between KSAs, values, and personalitytraits in Study 1 is only one way of classifying the data. Although asresearchers we try to make clean distinctions between constructs, theyoften have overlapping domains that make perfect categorizations diffi-cult (Herringer, 1998). Tb deal with this challenge, care was taken to usewidely accepted definitions of each of the three categories. Coders weretrained to use these definitions when making categorizations, which re-sulted in a high degree of consensus. In addition, if a characteristic wasrecategorized the effect would be consistent across perceived P-J andP-O fit, and would not change the results of the hypothesis tests. Thus,there is sufficient reason to believe the categorization schemes used aremeaningful and adequately allow the hypotheses to be tested.

A second concern is that the hypotheses in Study 2 were tested usingcross-sectional, self-report data. This opens the possibility that commonmethod variance may have infiuenced the results by artificially increas-ing relationships between the variables. However, because the focus ofStudy 2 is on the relative explanation provided by perceived P-J versusP-O fit, the absolute size of the parameters is less critical than their rel-ative contribution to predicting hiring recommendations. Because all ofthe relationships are likely inflated, the comparison process is not biased.Other problems such as priming of the participants or self-generated va-lidity may have also resulted from the data collection strategy. However,Tourangeau, and Rasinski (1988) have noted that these effects are lesslikely to occur when respondents are familiar with the issues being as-sessed, have expertise in making similar evaluations, and have involve-ment in the outcome. Therefore, by asking experienced recruiters toevaluate applicants in the context of authentic interviews, the effect ofthese biases should be reduced.

Finally, the limited generalizability of these results should be noted.The sample sizes for both studies were low, and represent recruitersfrom only four companies in a single industry. On average the recruitersdid not have extensive training on interviewing techniques, and onlyserved as recruiters a few days a year. Therefore, whether these find-ings generalize to other industries and to recruiters with more exten-sive training should be explored. In particular research is encouraged tofollow-up on the findings regarding the influence of idiosyncratic pref-erences on fit perceptions. Because the operationalization of idiosyn-cratic, organization-specific, and universal was based on whether a ma-jority of recruiters did or did not endorse a characteristic, the findingsare highly dependent on sample size. A characteristic could have been


classified differently based on responses by one or two additional re-cruiters. Therefore, while the exploratory analyses provide preliminarysupport for the prevalence of idiosyncratic preferences, future researchis needed using larger recruiter samples from multiple organizations.

A number of strengths offset these limitations. Two distinct method-ologies were used to assess the discriminant validity of recruiters' P-J andP-O fit perceptions. Study 1 specifically addressed the question of dif-ferential antecedents; whereas. Study 2 tackled whether perceived P-Jand P-O fit form discriminable factors and provide unique explanatorypower when it comes to predicting hiring recommendations. In addi-tion. Study 1 used a unique methodology, relatively new to industrial-organizational psychology, which followed Bretz et al.'s (1993) recom-mendations for how to study recruiters' perceptions of fit. Moreover,unlike the "paper people" commonly used in this type of study, video-tapes of actual mock interviews and resumes were used. This allowedrecruiters to view the applicants' qualifications, nonverbal behavior, ap-pearance, and interpersonal skills in a realistic setting. The combinationof mock interviews with real recruiters provided a balance of experimen-tal control and realism. This realism was expanded upon in Study 2,which examined recruiters' perceptions in an authentic decision makingenvironment. In addition, by assessing recruiters' perceptions after theyinteracted with real applicants using their natural interview strategy, thelikelihood that recruiters reported their true perceptions of the appli-cants is increased.

Implications and Recommendations for Future Research

Even during short, screening interviews, recruiters appear to dis-tinguish between the two types of fit. This implies that recruiters areconsidering a wide range of applicant characteristics that extend wellbeyond KSAs. However, it is clear that recruiters would benefit fromgreater guidance in terms of what makes an applicant a good fit. Al-though many companies have trained their recruiters to assess specificjob-related qualifications, they may benefit from expanding this train-ing to include what makes an applicant a good cultural fit. This mighthave the additional benefit of reducing the reliance on idiosyncratic pref-erences for judging fit. Research suggests recruiters vary in the extentto which they make valid decisions in interviews (Dougherty, Ebert, &Callender, 1986; Graves & Karren, 1992). Therefore, it may be usefulfor organizations to first determine which of their recruiters make themost valid judgments, and then to ascertain to what extent they rely onjudgments of P-J and P-O fit and the specific characteristics they use to


form these impressions. Other recruiters could then be trained to usetheir criteria and in order to improve selection decision making.

These findings corroborate other reports that recruiters make exten-sive personality inferences during interviews (e.g., Caldwell & Burger,1998; Rothstein & Jackson, 1984). Yet, there is little existing empiri-cal evidence regarding the accuracy of these assessments. Barrick, Pat-ton, and Haugland (2000) report that recruiters' evaluations of appli-cant personality are more similar to applicants' self-reported personal-ity than are evaluations by virtual strangers. However, even for the mosteasily observable personality traits, such as extraversion and agreeable-ness, recruiters' assessments are only moderately correlated with appli-cants' self-reports. Therefore, increased training on personality assess-ment or the use of interview formats specifically designed to measurepersonality (e.g.. Trull et al., 1998) is encouraged if recruiters are go-ing to continue using personality in their judgments of fit. Alternatively,companies should be persuaded to take greater advantage of existingpersonality tests that have proven to be reliable and valid measures ofjob-relevant personality traits (see Schmitt & Chan, 1998, for a review).

The results also have implications for future studies of perceivedP-J and P-O fit in recruiting research. Because recruiters use a varietyof applicant characteristics to judge P-J and P-O fit, existing measuresmay not be sufficient for assessing recruiters' perceptions. For example,to study recruiters' perceptions of P-O fit existing values measures suchas the Organizational Culture Profile (O'Reilly et al., 1991) or the CES(Ravlin & Meglino, 1987) should be supplemented with additional mea-sures to assess fit on personality and KSAs (see Bretz & Judge, 1994).Similarly, perceived P-J fit measures should be expanded to include per-sonality traits as well as KSAs. For applicants, the implication is thatresearching a company and doing self-assessment on a wide variety ofcharacteristics prior to interviewing is important. To convey a good fit,applicants should be prepared to articulate why their particular KSAs,personality traits, and values make them a good match for both the joband the organization.

These studies also help illustrate the relative contribution of per-ceived P-J and P-O fit to selection outcomes. More research is neededon what communication strategies and actions applicants can use to con-vey impressions of both P-J and P-O fit. In addition, because recruitersvary widely in their ability to accurately interpret "nonobservable" char-acteristics such as values (Cable & Judge, 1997) research is needed onwhat can be done by recruiters and applicants to make more reliable as-sessments of these characteristics. By building on the findings presentedin these studies, future research may improve on, as well as better un-derstand, recruiters' perceptions of fit.

668 P E R S O N N E L PSYCHOLOGY

R E F E R E N C E S

Adkins CL, Russel CJ, Werbel JD. (1994). Judgments of fit in the selection process: Therole of work-value congruence, PERSONNEL PSYCHOLOGY, 47,605-623.

Anderson NR. (1990). Repertory grid technique in employee selection. Personnel Review,193,9-15.

Anderson N, Ostroff C. (1997). Selection as socialization. In Anderson N, Herriot P(Eds.), International handbook of selection and assessment (pp. 413-440). Chich-ester, U.K.: Wiley.

Assouline M, Meir El. (1987). Meta-analysis of the relationship between congruence andwell-being measures. Journal of Vocational Behavior, 31,319-332.

Barrick MR, Mount MK. (1991). The Big Five personality dimensions and job perfor-mance: A meta-analysis. PERSONNEL PSYCHOLOGY, 44,1-26.

Barrick MR, Patton GK, Haugland SN. (2000). Accuracy of interviewer judgments of jobapplicant personality traits. East Lansing, MI: Michigan State University.

Bentler PM. (1985). Theory and implementation ofEQS: A structural equations program.Los Angeles, CA: BMDP Statistical Software.

Borman WC, Motowidlo SJ. (1993). Expanding the criterion domain to include elementsof contextual performance. In Schmitt N, Borman WC (Eds.), Personnel selectionin organizations (pp. 71-98). San Francisco: Jossey-Bass.

Bowen DE, Ledford GE Jr, Nathan BR. (1991). Hiring for the organization not the job.Academy of Management Executive, 54,35-51.

Bretz RD Jr, Judge TA. (1994). Person-organization fit and the theory of work adjustment:Implications for satisfaction, tenure, and career success. Journal of Vocational Be-havior, 44, 32-54.

Bretz RD Jr, Rynes SL, Gerhart B. (1993). Recruiter perceptions of applicant fit: Im-plications for individual career preparation and job search behavior. Journal ofVocational Behavior, 43,310-327.

Brook JA (1986). Research applications of the repertory grid technique. InternationalReview of Applied Psychology, 35, 489-500.

Byrne D. (1971). The attraction paradigm. New York: Academic Press.Cable DM, Judge TA. (1997). Interviewers' perceptions of person-organization fit and

organizational selection decisions. Journal of Applied Psychology, 82,546-561.Caldwell DF, Burger JM. (1998). Personality characteristics of job applicants and success

in screening interviews, PERSONNEL PSYCHOLOGY, 51,119-136.Caldwell DF, O'Reilly CA III. (1990). Measuring person-job fit with a profile-comparison

process. Journal of Applied Psychology, 75,648-657.Cascio WF. (1991). Applied psychology in personnel management. Englewood Cliffs, NJ:

Prentice-Hall.Chatman JA. (1989). Improving interactional organizational research: A model of person-

organization fit. Academy of Management Review, 14, 333-349.Coffey A, Atkinson P (1996). Making sense of qualitative data: Complementary research

strategies. Thousand Oaks, CA: Sage.Cohen J. (1960). A coefficient of agreement for nominal scales. Educational and Psycho-

logical Measurement, 20,37-46.Cooper WH. (1981). Ubiquitous halo. Psychological Bulletin, 90, 218-244.Costa PTJr, McCrae RR. (1988). From catalog to classification: Murray's needs and the

five factor model. Journal of Personality and Social Psychology, 49,1266-1282.Dawis RV (1991). Vocational interests, values, and preferences. In Dunnette MD, Hough

LM (Eds.), Handbook of industrial and organizational psychology (Vol. 2, pp. 833-871). Palo Alto, CA: Consulting Psychologists Press, Inc.


Deal TE, Kennedy AA. (1982). Corporate cultures: The rites and rituals of corporate life.Reading, MA: Addison-Wesley.

Dipboye RL. (1992). Selection interviews: Process perspectives. Cincinnati, OH: South-western.

Dougherty TW, Ebert RJ, Callender JC. (1986). Policy capturing in the employmentinterview. Journal of Applied Psychology, 71,9-15.

Edwards JR. (1991). Person-job fit: A conceptual integration, literature review andmethodological critique. International Review of Industrial/Organizational Psychol-ogy, 6,283-357. London: Wiley.

Folger R. (1987). Distributive and procedural justice in the workplace. Social JusticeResearch, 1,143-159.

Gilmore DC, Ferris GR. (1989). The effects of applicant impression management tacticson interviewer judgments. Journal of Management, 15,557-564.

Goldberg LR. (1992). The development of markers for the Big Five factor structure.Psychological Assessment, 4(1), 26-42.

Graves LM, Karren RJ. (1992). Interviewer decision processes and effectiveness: Anexperimental policy-capturing investigation, PERSONNEL PSYCHOLOGY, 45,313-340.

Guion RM. (1987). Changing view for personnel research, PERSONNEL PSYCHOLOGY, 40,199-213.

Harvey RJ. (1991). Job analysis. In Dunnette MD, Hough LM (Eds.), Handbook of indus-trial and organizational psychology (Vol. 2, pp. 71-163). Palo Alto, CA: ConsultingPsychologists Press, Inc.

Herringer LG. (1998). Relating values and personality traits. Psychological Reports, 83,95^954.

Hogan R. (1991). Personality and personality measurement. In Dunnette MD, Hough LM(Eds.), Handbook of industrial and organizational psychology (Vol. 2, pp. 873-919).Palo Alto, CA: Consulting Psychologists Press, Inc.

Holland JL. (1985). Making vocational choices, 2nd ed. Englewood Cliffs, NJ: Prentice-Hall.

Ivancevich JM, Matteson MT (1984). A Type A-B person-work environment interactionmodel for examining occupational stress and consequences. Human Relations, 37,491-513.

Jackson DN, Peacock AC, Smith JP (1980). Impressions of personality in the employmentinterview. Journal of Personality and Social Psychology, 39,294-307.

Judge TA, Ferris GR. (1992). The elusive criterion of fit in human resource staffingdecisions. Human Resource Planning, 154,47-67.

Kelly GA (1955). The psychology of personal constructs (VoL 1 and 2). New York: Norton.Krippendorf K. (1980). Content analysis: An introduction to its methodology. Beverly Hills,

CA: Sage.Kristof AL. (1996). Person-organization fit: An integrative review of its conceptualiza-

tions, measurement, and implications, PERSONNEL PSYCHOLOGY, 49,1-49.Maurer SD, Fay C. (1988). Effect of situational interviews, conventional structured in-

terviews, and training on interview rating agreement: An experimental analysis.PERSONNEL PSYCHOLOGY, 41, 329-344.

Meglino BM, Ravlin EC. (1998). Individual values in organizations: Concepts, controver-sies, and research. Journal of Management, 24,351-389.

MilesMB,Huberman AM. (1994). Qualitative data analysis: An expanded sourcebook (2nded.). Thousand Oaks, CA: Sage.

Motowidlo SJ, Borman WC, Schmit MJ. (1997). A theory of individual differences in taskand contextual performance. Human Performance, 10,71-83.

Motowidlo SJ, Van Scotter JR. (1994). Evidence that task performance should be distin-guished from contextual performance. Journal of Applied Psychology, 79,475-480.


Muchinsky PM, Monahan CJ. (1987). What is person-environment congruence? Supple-mentary versus complementary models of fit. Journal of Vocational Behavior, 31,268-277.

O'Reilly CA III. (1977). Personality-job fit: Implications for individual attitudes andperformance. Organizational Behavior and Human Performance, 78,36-46.

O'Reilly CA III, Chatman JA, Caldwell DF. (1991). People and organizational culture:A profile comparison approach to assessing person-organization fit. Academy ofManagement Journal, 34,487-516.

Paunonen SV, Jackson DN, Oberman SM. (1987). Personnel selection decisions: Effectsof applicant personality and the letter of reference. Organizational Behavior andHuman Decision Processes, 40,96-114.

Phillips JM. (1998). Effects of realistic job previews on multiple organizational outcomes:A meta-analysis. Academy of Management Journal, 41,673-^90.

Posner BZ. (1992). Person-organization values congruence: No support for individualdifferences as a moderating influence. Human Relations, 45,351-361.

Ravlin EC, Meglino BM. (1987). Effects of work values on perception and decision mak-ing: A study of alternative work value measures. Journal of Applied Psychology, 72666-673.

Reger RK. (1990). Managerial thought structures and competitive positioning. In HuffAS (Ed.), Mapping strategic thought (pp. 71-88). Chichester, U.K.: Wiley.

Roberts P (1999). The agenda—total teamwork. Fast Company, 23, 148-153, 156, 158,160,162.

Rokeach M. (1973). The nature of human values. New York: Free Press.Rothstein M, Jackson DN. (1984). Implicit personality theory and the employment in-

terview. In Herriot P, Cook M (Eds.), Issues in person perception (pp. 167-201).London: Methuen.

Rynes SL, Gerhart B. (1990). Interviewer assessments of applicant fit: An exploratoryinvestigation, PERSONNEL PSYCHOLOGY, 43,13-35.

Schein E. (1990). Organizational culture and leadership. San Francisco: Jossey-Bass.Schmitt N, Chan D. (1998). Personnel selection: A theoretical approach. Thousand Oaks,

CA: Sage.Schneider B. (1987). The people make the place, PERSONNEL PSYCHOLOGY, 40,437-453.Smith JM. (1986). A repertory grid analysis of supervisory jobs. International Review of

AppUed Psychology, 35, 501-512.Snow CC, Snell SA. (1993). Staffing as strategy. In Schmitt N, Borman WC (Eds.),

Personnel selection in organizations (pp. 448-478). San Francisco: Jossey-Bass.Spokane AR. (1985). A review of research on person-environment congruence in Hol-

land's theory of careers. Journal of Vocational Behavior, 26,306-343.Stevens CK, (1998). Antecedents of interview interactions, interviewers' ratings, and ap-

plicants' reactions, PERSONNEL PSYCHOLOGY, 57,55-86.Stevens CK, Kristof AL (1995). Making the right impression: A field study of applicant

impression management during job interviews. Journal of AppUed Psychology, 80,587-606.

Tburangeau R, Rasinski KA (1988). Cognitive processes underlying context effects inattitude measurement. Psychological Bulletin, 103,299-314.

TVull TJ, Widiger TA, Useda JD, Holcomb J, Etoan B, Axelrod SR, Stern BL, GershunyBS. (1998). A structured interview for the assessment of the five factor model ofpersonality. Psychological Assessment, 70,229-240.

. (1984). The role of perceived injustice in defendants' evaluations of their court-room experience. Law and Society Review, 18,51-74.


V̂ n Vianen AEM. (2000). Person-organization fit: The match between newcomers' andrecruiters' preferences for organizational culture, PERSONNEL PSYCHOLOGY, 55,113-149.

Wanous JP. (1978). Realistic job previews: Can a procedure to reduce turnover also infiu-ence the relationship between abilities and performance? PERSONNEL PSYCHOLOGY,31,249-258.

Wanous JP. (1980). Organizational entry: Recruitment, selection, and socialization of new-comers. Reading, MA: Addison-Wesley.

Werbel JD, Gilliland SW. (1999). Person-environment fit in the selection process. In FerrisGR (Ed.), Research in personnel and human resource management (Vol 17, pp. 209-243). Stamford, CT: JAI Press.

APPENDIX

Example Repertory Grid Interview Data Sheet(Protocol A: Applicant Triad Comparison Series 4)

Comparison Best fit

1. Applicants 5

4, 5,2 2

2. Applicants3,1,7

3. Applicants4,3,6

4. Applicants1,2,3

5. Applicants5,7,2

6. Applicants4,5,6

7. Applicants2,7,4

8. Applicants6,1,5

9. Applicants3,7,6

Characteristic

P-J: Workexperience*

P-O: Likesvariety*

P-JP-OP-JP-OP-JP-OP-JP-OP-JP-OP-JP-OP-JP-OP-JP-O

Description

P-J: Already heldan internshipin our field*

P-O: Wants to constantlybe working onnew projects*


Opposite

P-J: No workexperiencein our field*

P-O: Someone whois comfortable withroutines and rules*


*Examples of the information provided by recruiters.

Kristof and Brown

Documents

Transcript of Kristof and Brown