|USFDC Home | USF Electronic Theses and Dissertations||| RSS|
This item is only available as the following downloads:
The Validity Of Personality Trait Interacti ons For The Prediction Of Managerial Job Performance by Amy M. Taylor A thesis submitted in partial fulfillment of the requirements for the degree of Master of Arts Department of Psychology College of Arts and Sciences University of South Florida Major Professor: Walter C. Borman, Ph.D. Carnot Nelson, Ph.D. Marcia Finkelstein, Ph.D. Date of Approval: November 14, 2007 Keywords: conscientiousness, agreeableness, extraversi on, neuroticism, managers, personnel selection Copyright 2008, Amy M. Taylor
Acknowledgements The author would like to thank Personne l Decisions, International for providing the dataset used in this study.
i Table of Contents List of Tables ii List of Figures iii Abstract iv Introduction 1 Previous work with trait interactions 5 Interactions between personality traits and other variables 9 Current study 12 Method 16 Participants 16 Measures 16 Results 18 Exploratory Analyses 20 Discussion 30 Study strengths 31 Practical implications 32 Limitations and directions for future research 33 Conclusion 35 References 36 Appendices 41 Appendix A: GPI Personalit y Dimensions and Facets 42
ii List of Tables Table 1 Correlations of Personali ty Traits with Managerial Performance Ratings 19 Table 2 Hierarchical Regression of Performance Ratings on CxA 21 Table 3 Mean Managerial Performance Ratings by Group 21 Table 4 Hierarchical Regression of Performance Ratings on ExNxC 23 Table 5 Managerial Perfor mance Scores by Industry 24 Table 6 Correlations of Personality Traits with Performance Ratings by Industry 26 Table 7 Hierarchical Regression of Performance Ratings on ExNxC for Trade, Transportation, and Utilities Industry 27 Table 8 Hierarchical Regression of Performance Ratings on CxA for Education and Health Industry 28 Table 9 Hierarchical Regression of Performance Ratings on ExN for Education and Health Industry 29 Table 10 Hierarchical Regression of Performance Ratings on ExNxC for Education and Health Industry 30
iii List of Figures Figure 1 Relationship of Job Performance to Conscientiousness by Agreeableness 22
iv The Validity of Personality Trait Interactio ns for the Prediction of Managerial Job Performance Amy M. Taylor ABSTRACT Personality variables have been shown to be significant predictors of job performance (Barrick & Mount, 1991; Te tt, Jackson, & Rothstein, 1991). Recent advances in methodology for analyzing pe rsonality-job performance relationships indicate that interact ions among traits may yield increm ental validity. Job types in which performance has been shown to relate to trait interactions include cler ical jobs, jobs with high interpersonal components, and jobs in realistic and conventional contexts, (Witt, Burke, Barrick, & Mount, 2002; Burke & Witt, 2002; and Burke & Witt, 2004). This study examined the validity of trait interactions for the prediction of managerial job performance. Hypotheses includ ed a main effect for Consci entiousness, an interaction between Conscientiousness and Agreeableness, an interaction between Extraversion and Neuroticism, and finally, a three-way inter action between Extraversion, Neuroticism, and Conscientiousness. An archival dataset from Personnel Decisions International (n=680 managers) containing GPI personality scores and supervisor-rated performance scores was analyzed to test the hypotheses. Correla tions and moderated hierarchical linear regressions were performed to estimate the rela tionships of the predictors to the criterion, and to learn whether examination of trait inte ractions contributes incremental validity to the single trait scales.
v A main effect for Conscientiousness on managerial job performance was found. No trait interactions explained incremental variance in performance scores. Therefore, Conscientiousness is the recommended persona lity scale to use for selecting managers. This finding is consistent with previous re search on the relation of Conscientiousness to job performance in managers (Barrick & Mount, 1991). Managers from diverse organizations and industries comprised the sa mple, increasing the ge neralizability of the results. Directions for future research include the examination of ot her trait interactions, more specific criteria such as competenci es rather than overa ll managerial job performance, and effects of the hierarchical level of the manage r in the organization.
1 Introduction Research on the prediction of managerial job performan ce has explored numerous predictor variables. One popular predictor is the personality of the manager. Although the usefulness of personality testing has recei ved past criticism (Guion & Gottier, 1965), more recent research supports the validity for predicting job performance. Today, these inventories are commonly used in practice. In fact, about 40% of Fortune 500 companies and 100% of the top 100 UK companies use pers onality inventories as selection tools (Rothstein & Goffin, 2006). The correlation between personality and managerial job performance has been empirically demons trated through meta-analysis (Barrick & Mount, 1991; Tett, Jackson, & Rothstein, 1991 ). This study was intended to further examine the predictive validity of personality for managerial performance by examining the extent to which certain traits may impact the predictive ability of other traits. Metaanalytic evidence supports this moderator hypothesis; substa ntial unexplaine d variability remains in the estimated population validity coef ficients for four of the Big Five traits (Barrick & Mount, 1991). This study attempted to explain a significant portion of this variance through trait interactions. Before presenti ng the evidence supporting trait interactions, I first present support for pers onality as a predicto r of managerial job performance. Past behavior is the best predictor of future behavior Hogan (2005) contends that personality is a summary of behavioral patter ns and thus a good source of information on future behavior. An individuals responses on personality tests can inform employers
2 about how the applicant is lik ely to perform on the job. Speci fically, personality measures based on the Big Five theory reflect the attitudi nal, experiential, emotional, interpersonal, and motivational styles of the individual (Truxillo, Bauer, Campion, & Paronto, 2006). The Big Five model is a commonly used framework for describing personality. Under this typology, personality is reported vi a scores on five dimensions, or traits: Extraversion, Conscientiousness, Agreeab leness, Neuroticism, and Openness to experience. Research drawing on the Big Five framework has led to more positive conclusions regarding the predic tive validity of personality th an were previously noted. Guion and Gottiers (1965) review concluded th at personality was often a poor predictor of job performance and recommended against the use of personality measures for personnel selection for most j obs. Since that time, meta-ana lyses using the five-factor framework (Barrick & Mount, 1991; Tett, J ackson, & Rothstein, 1991 ) found that traits are significantly related to work performance. The usefulness of the Five Factor model li es in the broad trait descriptions that allow related traits to be combined into mo re inclusive dimensions (i.e. the Big Five). The first dimension, Extraversion, represents a tendency toward sociability, high energy, and optimism. Neuroticism is associated with emotional instability, hostility, and anxiety. Agreeableness indicates a tendency to be cooperative, trusting, and helpful to others. Openness to Experience refers to ones creative, insigh tful, and free-thinking attributes. Finally, Conscientious individua ls are responsible, organized, and selfdisciplined. These five dimensions provi de a general framework for classifying personality trait predictors th at have been the subject of research. Many personality tests report scores at the more specific facet level which can then be aggregated to the five
3 factor scores (e.g. NEO-PIR; Costa & Mc Crae, 1992; GPI; Schmit, Kihm, & Robie, 2000). An additional reason to use personality tests is that they do not distinguish between protected classes of applicants as do some other selection tools. Tests of cognitive ability typically favor white appli cants, which can result in adverse impact. The reduced risk of facing an EEO lawsuit increas es the appeal of personality tests over certain other selection tools. The relationship between personality and job performance is supported by empirical evidence. Barrick and Mounts ( 1991) meta-analysis found that Extraversion, Conscientiousness, and Agreeableness were statistically signifi cant predictors of managerial performance ( =.10-.22; validities were averaged across job proficiency, training proficiency, and personnel data criteri a). When only job proficiency scores were examined, Conscientiousness was a signifi cant predictor across all job types ( =.23), of which approximately 41% of the samples we re managers. Bono and Judges (2004) metaanalysis of the Big Five and transformational leadership found correct ed validities of .13.24 for each of the five dimensions. Significan t correlations were found for each trait with the leadership criterion. However, it is possible that moderators may explain additional variance. For example, Extraversion was the strongest pred ictor of transformational leader behaviors ( =.24). It is reasonable that a leader should be gregarious, energetic and assertivebut if that leader is also anxious or di sorganized and does not follow through on commitments, he/she may not be able to insp ire or rally support from followers. Thus, an extraverted manager who is high on Neurotic ism may pass his or her negative moods and emotions onto subordinates, resulting in a poor work environment. Similarly, an
4 extraverted manager low on Conscientiousne ss may not carefully monitor subordinate activities or may not accomplish tasks in a timely manner. This type of manager may be seen as all talk and no action. Therefore, it is not only the result of being extraverted, but the interactive eff ect of being extraverted, emotionally stable and conscientious that may result in successful leadership. Results of Barrick and Mounts (1991) meta-analysis reveal substantial variability not explained by ar tifacts in correlations between four of the Big Five traits (Extraversi on, Conscientiousness, Openness, and Emotional stability) and job performance among the managerial samp les. This suggests the presence of moderators. This study examines a potential moderatorwhether the level of one trait impacts the validity of anot her trait. This hypothesis pos its that consideration of interactions or configural scores rather th an single trait scores will yield more accurate predictions for success on the job. For ex ample, Hogan, Hogan, and Roberts (1996) caution that it is ill-advised to interpret single sc ale scores without additional information, providing the following example: persons w ith high scores on measures of service orientation will be tolerant, patient, and frie ndly, but they may not work very hard, (p. 470). This illustrates the usefulness of ex amining both Conscientiousness and Service orientation aspects of personality to predict who will be the top performers in customer service jobs. Similarly, workers high on Cons cientiousness but low on Service orientation may be seen as inflexible or demanding, a nd unable to build rela tionships and loyalty from customers. Foster and Macan (2006) encouraged the an alysis of interactions in personality testing. Single scale scores cannot address the pot ential enhancement or inhibition effects of one trait on another. Inte raction scores allow for thes e contingent relations with
5 criteria. Trait interactions are based on the premise that the value of a certain trait is dependent upon the presence of another tr ait for successful job performance. Conventional approaches to increasing the usefulness of personality tests as predictors focus on resolving measurement i ssues (e.g. faking) or identifying different personality and performance constructs (Foster & Macan, 2006). Foster and Macan (2006) offer an additional suggestion: the us e of alternative sta tistical analyses. One example of this is personality trait interac tions. Recent work with trait interactions has shown promising results for predicting work behaviors. Previous Work with Trait Interactions Openness to experience would seem to be a poor predictor of job performance based on bivariate correlations between th e trait and job performance scores ( = -.03, Barrick & Mount, 1991). However, Burke and Witts (2002) work showed that this low correlation may be enhanced by considering th e moderating effect of other personality traits. Using a sample of 114 clerical wo rkers, they found that Extraversion and Emotional stability interacted with Openne ss to predict job performance. A separate regression was run for each moderator. Th e Openness x moderator (Extraversion or Emotional stability) cross-produc t term yielded significant incr emental validity over the single scale scores and the control variables age, sex and tenure ( R2= .04 Extraversion, R2=.03 Emotional stability; p< .05 for both). Results showed that those with low Openness/high Extraversion or low Openness/low Emotional stability received lower performance ratings. The highest performance ratings were attained by workers high on both Openness to experience and Extrav ersion, and high on both Openness and Emotional stability. They concluded that lo w Openness was more detrimental to those
6 who were likely to express their close-mi ndedness (extraverts) or who were unable to maintain a positive demeanor (emotionally unstable). Witt, Burke, Barrick, and Mount (2002) f ound support for the interactive effect of two other Big Five traits, Conscientiousn ess and Agreeableness. Of seven employee samples included in this study, five consis ted of jobs characterized by frequent interpersonal interactions. The Conscientious ness x Agreeableness interaction term added significant incremental validity in each of the five ( R2= .01-.02, p< .05). The effect sizes found were within the range of moderator effects typical for nonexperimental studies (Champoux and Peters, 1987; Chaplin, 1991). Cons cientious individual s who were also high in Agreeableness tended to have higher job performance scores than Conscientious individuals who were low in Agreeableness. They hypothesize that this is because besides being diligent, dependable and achievement-oriented, employees in jobs with a substantial social interaction requirement must also be c ooperative and considerate of others in order to be successful. Witt et al. present the example of a manager who is highly conscientious, but also disagreeable. This manager is likely to be seen as micromanaging, unreasonably dema nding, inflexible (p. 165). Burke and Witt (2004) also found an inte raction effect of Agreeableness and Conscientiousness for the prediction of em ployee high maintenance behaviors (HMBs; Grensing-Pophal, 2001). HMBs are behaviors th at are annoying or aggravating such as complaining about work, frequently mentioning the desire to quit, and repeatedly causing interpersonal conflicts at wo rk (Burke & Witt, 2004). The interaction of Agreeableness and Conscientiousness accounted fo r significant incremental validity ( R= .01, p< .05), beyond the control variables employee sex, education, tenure, satisfaction with
7 supervisor, and each of the Big Five traits Results showed that Agreeableness was only predictive of HMBs for those high in Consci entiousness, such that individuals who scored high on both traits had the least frequent HMBs. For employees low in Conscientiousness, Agreeableness was not pr edictive of HMBs. Here, the validity of Agreeableness for predicting these negative in terpersonal behaviors was contingent upon the level of Conscientiousness. The studies reviewed thus fa r show support for interactions of personality traits as predictors of work behaviors; however, they do not examine the validity for prediction of managerial job performance. Foster and H ogan (2006) examined managerial performance in relation to personality pr ofiles. Although trait profiles di ffer from trait interactions, these techniques are related because both use mu ltiple trait scores as predictors. Using the Hogan Personality Inventory and the Hogan Development Survey, three personality profiles were generated and compared for predictive validity. The first profile was a bright side composite that measured manage rs standings on the traits of Adjustment, Ambition, Interpersonal sensitivity, and Prude nce, for which higher scores indicated more management potential. The second prof ile focused on the dark side traits of Excitable, Skeptical, Cautious, Bold, Mischi evous, and Imaginative, for which lower scores indicated higher manage ment potential. The third profile was a combination of the bright and dark side traits. For each profile, participants were groupe d as either high or low management potential based on their score percentile for ea ch trait scale within the respective profile. The goal of this study was to determine wh ich profile yielded the highest predictive validity against the performance ratings, determined by the largest between-groups
8 difference score. Data from six studies (N= 810) were combined to produce meta-analytic estimates of mean difference scores in mana gerial performance rati ngs for those in the high vs. low leadership potential groups. Th e bright and dark si de profiles yielded estimated population difference scores of .33 and .36, respectively, between the high and low management potential groups. The total leadership profile (combination of both bright and dark side traits) yielded an estimated population difference score of .44. Incremental validity of the profiles over sing le scale scores was not reported. Although interactions were not specifical ly tested in this study, result s indicate that the use of two profile scores (the combination of bright and dark side traits) predic ted leadership ratings better than either prof ile did separately. Based on the results from the Burke a nd Witt (2002, 2004) and Witt et al. (2002) studies, Foster and Macan (2006) tested the validity of two pairs of interactions for predicting job performance: Conscientiousness with Agr eeableness and Openness to experience with Extraversion. H ogan Archival data were used to compare personality scores from the HPI and job performance ratings. To test the moderation hypotheses, participants were categorized into high, me dium, or low Agreeableness groups and high, medium, or low Extraversion groups. Corre lations between the moderating variable (Conscientiousness or Openness) were calculat ed for each of the three groups across all samples, and the estimates were meta-analyze d. Foster and Macan classified participants jobs according to Hollands (1996) RIASEC model to account for job context. This model is primarily used to categorize job char acteristics to assess pers on-job fit. Jobs can be classified as Realistic, Investigative, Artistic, Social, Enterprising, or Conventional job
9 types. Only three of the six were represente d in this study: Conven tional, Enterprising, and Realistic. Separate effect sizes were estimated for each job family. Foster and Macan found that Agreeablen ess moderated the relationship between Conscientiousness and job performance in C onventional and Realistic job types. Persons in the high and low Agreeableness groups s howed stronger, positive relationships between Conscientiousness and job performan ce with those least agreeable having the strongest correlation ( = .27). Foster and Macan did not report which group had the highest performance ratings. One possible reason for this form of interaction is that in these less social job contexts, those who are lower in Agreeableness tend to focus on the task and avoid interacting with coworkers; they are then judged solely on their task performance. Of those who are highly agr eeable, persons higher in Conscientiousness may engage in and be more effective at ci tizenship behaviors and thus have higher job performance ratings (Motowidlo, Borman & Schmit, 1997). Not all research has shown support for the interaction of persona lity traits. Warr, Bartram, and Martin (2005) examined inte ractions between Conscientiousness and each of the other four Big Five traits for predic ting sales performance. Each interaction term was tested in a separate m oderated regression analysis, bu t none was significant. Interactions between Personalit y Traits and other Variables Personality may interact with other individual differences and contextual variables to predict job performance. Ge orge and Zhou (2001) found a significant threeway interaction between Openne ss to experience, feedback valence, and unclear means for the prediction of crea tive behaviors at work ( R2 = .04, p< .05). In a sample of 149 employees from a mechanical equipment production company, the most creative
10 behaviors were exhibited by those who were high on Openness, received positive feedback, and had unclear means of how to co mplete their tasks. A similar interaction between Openness, feedback valence, and unclear ends was also supported ( R2 = .03, p< .05). As expected, individuals high on Op enness who received positive feedback had higher creativity ratings when the desired re sult of their task was ambiguous. Openness score alone (or within two-way interaction te rms) was not a significant predictor in any analyses. George and Zhou reas on that situations that pres ent unclear expectations for performance, such as tasks with unclear m eans and unclear ends, provide an environment that is conducive to the creative tendencies of individuals high on Openness to experience. The same study also found interactions be tween Conscientiousness and situational factors. A three-way inter action term between Conscien tiousness, supervisor close monitoring, and coworker suppor t yielded incremental validity for predicting creativity behaviors. Conscientiousness alone (or with in two-way interaction terms) was not a significant predictor in any analyses. Three f acets of coworker support were examined in the three-way interaction terms: unhelpful co workers, inaccurate communication from coworkers, and negative work environment (e.g. coworkers who always find fault with others). All three facets mode rated the validity of Conscientiousness and supervisor close monitoring for predicting creativity ( R2 = .04-.06, p< .05). As hypothesized by George and Zhou, the lowest creativity scores were attained by highly conscientious individuals who were closely monitored by their supe rvisors and had unsuppor tive coworkers. Emotional exhaustion is another proposed moderator of the Conscientiousness-job performance relationship. Witt, Andrews, and Carlson (2004 ) studied customer service
11 representatives (CSR) call volumes as a measur e of job performance. The main effect of Conscientiousness was not significant in th e hierarchical moderated multiple regression analysis; however, the emotional exhaus tion x Conscientiousness cross-product term explained a significant portion of variance in call volume ( R2 = .03, p< .05). The relationship between emotiona l exhaustion and call volume was strongest for CSRs high on Conscientiousness. The highest call vo lume was attained by CSRs with high Conscientiousness and low emotional exha ustion. Individuals lo w on Conscientiousness attained similar call volumes across levels of emotional exhaustion. Witt and Ferris (2003) investigated social skill as another potential moderator of the Conscientiousness job performance rela tionship. Specifically, they examined an interpersonal effectiveness component of j ob performance as the performance criterion. They hypothesized that Conscientiousness woul d have the strongest positive relationship to performance ratings for workers high in social skill. They also posited that performance ratings would be lowest for workers high on Conscientiousness and low in social skill; that is, workers who are highl y Conscientious but lack the ability to appropriately read interpers onal situations would be perc eived as demanding, inflexible, and otherwise difficult to work with. Hierarch ical moderated regression analysis showed that the interaction term added significant incremental validity ( R2 = .03, p< .05). Consistent with their first hypothesis, work ers high on both Conscien tiousness and social skill had the highest performance ratings. Among workers low on social skill, Conscientiousness related negatively to perfor mance ratings in one st udy; the relationship was essentially zero in the second study.
12 Perceptions of organizationa l politics may also moderate the Conscientiousness job performance relationship. Organizational po litics refer to behaviors that promote selfinterest without rega rd to, and often in opposition to, organizational goals (Mintzberg, 1983 as cited by Hochwarter, Witt, & Kacmar 2000). Examples include providing more resources to subordinates w ho blindly follow orders rath er than those who question decisions, and sabotaging the work of cowork ers who do not facilita te ones advancement in the organization (Hochwarter, Witt, & K acmar, 2000). Hochwarter, Witt, and Kacmar found evidence for the interaction of per ceptions of organiza tional politics and Conscientiousness. Although employees with hi gh levels of Conscientiousness tended to receive higher performance ra tings regardless of perceptions of organizational politics, those low on Conscientiousness had signifi cantly lower performance ratings if they perceived that political behavi or was highly prevalent. These studies indicate that pe rsonality traits interact w ith other traits as well as organizational variables to predict work beha viors. However, I am aware of no existing research on the validity of trait interactions against the criteri on of managerial job performance. This study contributes to personality literature by examining these relationships. Current Study As mentioned, this study examined sp ecific hypothesized interactions among personality traits for the pr ediction of managerial job performance. Management is a highly interactive, socially-ori ented job that requires a wide range of characteristics such as charisma, confidence, interpersonal skill, and diligence. The absence of one trait may diminish the efficacy of the other traits. Th e inclusion of trait in teraction scores in
13 hierarchical moderated regression analyses is expected to yield incremental validity over the traditional method of exam ining single scale validities. Conscientiousness is the best personal ity predictor of performance across job types (Barrick & Mount, 1991). High Consci entiousness indicates a tendency towards several effective work behavior s such as being diligent, ac hievement-oriented, and selfdisciplined. Managers who are high on this tra it are likely to be effective at meeting deadlines, planning and setting goals, and stri ctly following organiza tional policies. For these reasons, I hypothesize a main effect of Conscientiousness on performance. Hypothesis 1: Conscientious ness will be positively re lated to managerial job performance. Cultivation of interpersonal relationships is an essential skill for managers. The quality of interactions with subordinates ha s consequences for subordinate performance, organizational commitment, and job satis faction (Gerstner & Day, 1997). A manager should be cooperative with subordinates to inspire a sense of collaboration and mutual respect among employees. This suggests that managers should be high on Agreeableness to be effective. However, agreeableness wit hout a strong need for achievement may lead to a manager who is more focused on getting along with and pleasi ng subordinates than on completing tasks. Conversely, a manager wh o is high on Conscientiousness, but low on Agreeableness, may be seen as pushy a nd demanding. As such, Conscientiousness is predicted to interact with agreeableness to predict mana gerial performance. This hypothesis is consistent with findings from Witt et al. (2002) and Burke and Witt (2004). Hypothesis 2: Conscientiousness and Agr eeableness will interact to predict managerial job performance. The rela tionship between Conscientiousness and
14 performance will be stronger for those high on Agreeableness than those low on Agreeableness. Further, managers high on both traits will have the highest job performance. Individuals high on Extraversion tend to in teract more with others. Individuals who have more contact with ot hers have more opportunity to spread their emotions via contagion effects. A manager who scores hi gh on Extraversion will be more socially interactive and better able to share his or her enthusiasm w ith subordinates. However, if the manager is high on Neuroticism, he or she may instead spread negative feelings among subordinates that, over time, could re sult in low job satisfaction and reduced commitment to the manager and organization. Th is manager would not be able to inspire or motivate subordinates, a key component of effective leadership. For this reason, Extraversion and Neuroticism are predicte d to interact to predict managerial performance. Hypothesis 3: Extraversion and Neuroticism will interact to predict managerial job performance. The relationship between Extraversion and performance will be stronger for those low on Ne uroticism than for those high on Neuroticism. Further, managers who are high on Extraversion and low on Neuroticism will have the highest job performance. A final hypothesis is based on the interac tive effects of Extr aversion, Neuroticism and Conscientiousness. As stated in Hypothesis 3, managers who are high on Extraversion and low on Neuroticism are predicte d to attain the highest job performance. Because these managers are more socially dominant and more likely to foster positive feelings in the workplace, subordinates may be more receptive to the managers work
15 values and practices. For this reason these managers may be even more effective when they are also high in Conscientiousness. Conscientious managers who can impart effective work habits associat ed with their diligence, self -discipline, and thoroughness are more likely to meet performance goals than managers who are low on Conscientiousness. Hypothesis 4: Extraversion, Neuroticism, and Conscientiousness will interact to predict ratings of managerial job performance. Managers who are high on Extraversion, low on Neuroticism, and high on Conscienti ousness will have higher job performance than managers who are high on Extra version and low on Neuroticism and Conscientiousness.
16 Method Participants A dataset from Personnel Decisions Intern ational archives pr ovided personality and job performance scores for 680 managers from numerous organizations across 22 industries. The majority of the sample was ma le (71%). Ethnicities represented were 84% White, 4% African-American, and 4% Hispanic (8% we re other or unreported ethnicities). The managers were between 26-60 years of age (M=41, SD=6.5) with a mean organizational tenure of 9.7 years (SD=7.3). Most managers (79%) held a bachelors degree or higher. Measures Personality. This study used the five dimensions of personality as measured by the Global Personality Inventory (GPI; Schmit, Kihm, & Robie, 2000). This test provided scores on thirty-two lower level facets that we re aggregated to the five trait scores per Schmit, Kihm, and Robies designation (2000; see Appendix for subscales comprising each factor). Subject matter experts crea ted the GPI collaboratively across eleven countries and ten languages using a combined emic and etic approach. This development strategy allows for the comparison of scores among applicants of di fferent nationalities. This is an important contribution in light of the growing trend towards multinational personnel recruitment and selec tion. The GPI was created sp ecifically to aid in the prediction of work outcomes; the work context was either explicitly mentioned or implied in the wording of each item.
17 Job performance. Supervisor ratings of job performance are the criterion in this study. Supervisors rated incumbent managers on five items that assess how well he or she gets the job done, gets work done on time, accomplishes a great deal, produces high quality work, and is an effective manager ove rall. Items were rated on a five point scale and were averaged to one ove rall managerial job performan ce score. These ratings were collected for research purposes only and ha d no administrative cons equences. This is a strength of the data, as the supervisors were more likely to give accurate ratings in this appraisal context.
18 Results Cronbachs alpha for the 5-item job pe rformance measure was estimated at .88, indicating a sufficient level of internal consistency. Descriptive statistics and correlations for study variables are presented in Tabl e 1. Conscientiousness was significantly correlated with managerial job performan ce (r = .14, p< .01), fully supporting Hypothesis 1. To test Hypotheses 2-4, correlations, and subsequently, moderated hierarchical linear regressions were conducted. Th e interaction term, Agreeab leness x Conscientiousness, was significantly correlated with job performance. To determin e whether inclusion of this interaction term explained variance in perfor mance ratings above that of the single trait scores, a hierarchical regre ssion was conducted. First, the jo b performance ratings were regressed on the scale scores for Agreeableness and Conscientiousness, and then the interaction term was entered into the re gression to determine the change in R2 between the two models. Regression resu lts are presented in Table 2. R2 was not statistically significant ( R2 = .003; n.s.). Thus, Hypothe sis 2 was not supported. To identify which managers had the highe st performance ratings, managers were classified as low or high on Conscientiousne ss and Agreeableness based on whether they were below the mean score (low) or equal to or above the mean scor e (high) on each trait. Mean performance ratings for each group ar e displayed in Table 3 and represented graphically in Figure 1. ANOVA with Fisher LSD post-hoc te sts revealed that the high Conscientiousness, low Agreeableness group was significantly hi gher on the criterion than the two groups with low Conscientiousne ss scores (F (3, 676) = 3.49, p< .05). The
19 Table 1 Correlations of Personality Traits w ith Managerial Performance Ratings Variable MSD 12 34567 1. Job performance 4.06.57.88 2. C 23.293.02.14** -3. A 34.513.66-.02.45** -4. E 58.127.04.03.33** .43** -5. N 15.143.34.04-.36** -.50**-.56** -6. CxA 808.73165.81.08*.88** .82**.45**-.50** -7. ExN 866.98163.80.07-.24** -.33**-.03.83**-.34** -8. ExNxC 20075.314124.84 .15**.40** -.05.16**.56**.22**.79** Note: n=680 aInternal consistency (alpha) *p< .05; **p< .01
20 high Conscientiousness, high Agreeableness group scored higher than the two low Conscientiousness groups, but the difference did not reach statistic al significance. The two high Conscientiousness groups (low and high Agreeableness) did not differ significantly on performance ratings, indica ting that Conscientiousness, and not the interaction with Agreeableness, was th e driver of higher job performance. The cross-product term Extraversion x Neuroticism was not significantly correlated with performance ratings, so no further regres sion analysis was conducted. Hypothesis 3 was not supported. The Extraversion x Neuroticism x Cons cientiousness term was significantly correlated with job performance ratings, (r= .15, p< .01). I then tested a hierarchical regression in which the single trait scores were entered in step 1, the two-way interaction terms were entered in step 2, and the threeway interaction term was entered in step 3. See Table 4 for regression results. The in teraction of Extraversion by Neuroticism by Conscientiousness did not explain signifi cant incremental variance in performance ratings. Thus, Hypothesis 4 was also not supported. Extraversion, Neuroticism, and Conscientiousne ss scores were dichotomized at the mean, and participants placed into one of eight gr oups depending on whether they were high or low on each variable. ANOVA showed that the groups did not differ significantly on performance ratings, (F (7, 162) = 1.62, n.s ). Exploratory Analyses To investigate potential differences in validities among industr ies, participants were grouped into one of five industry categ ories: natural resour ces; construction and manufacturing; trade, transpor tation, and utilities; financial, business, and professional
21 Table 2 Hierarchical Regression of Performance Ratings on CxA Variable Step 1 Step 2 Step 1: Conscientiousness .18**-.24 Agreeableness -.10*-.46 Step 2: C x A .67 R2 .027**.030** R2 .027**.003 *p< .05, **p< .01 Table 3 Mean Managerial Performance Ratings by Group Group n meansdstd error Low C, Low A 234 4.00.58 .04 Low C, High A 109 3.97.55 .05 High C, Low A 121 4.17.55 .05 High C, High A 216 4.10.57 .04
22 Figure 1 Relationship of Job Performance to Conscientiousness by Agreeableness. services; or education and health. These cate gories are based on the U.S. Bureau of Labor Statistics North American Industry Cl assification System (NAICS; BLS, 2004). However, the NCAIS has 20 sectors that I sort ed into the five broader categories. The industries comprising each cat egory, with the n, mean, and standard deviation of performance scores for each group, are pres ented in Table 5. The groups performance scores did not vary significantly, (F (4, 663) = 1.18, n. s). Correlations between performance ratings and the personality vari ables for each industry group can be found in Table 6; Group 1 (natural resources) was not included due to the small n. Performance ratings in the trade, transportati on, and utilities industry group were significantly correlated with the Extraver sion by Neuroticism by Conscientiousness interaction term, (r= .16, p< .01). Whether incremental variance in performance was
23 Table 4 Hierarchical Regression of Performance Ratings on ExNxC Variable Step 1 Step 2 Step 3 Step 1: Extraversion .05.41.04 Neuroticism .13* *.05-.63 Conscientiousness .16* *.65.25 Step 2: ExN .09.65 ExC -.72-.11 CxN -.04.62 Step 3: ExNxC -.60 R2 .029 ** .033 ** .033 ** R2 .029 ** .004 .000 **p< .01
24 Table 5 Managerial Performance Scores by Industry Industry nM SD Group 1: Natural resources group n=9 4.48 .59 Agriculture, foresty, & fishing 2 Natural resources 7 Group 2: Construction & manufacturing group n=119 4.02 .60 Food manufacturing/food processing 9 Electrical/electrica l manufacturing 24 Other manufacturing 73 Construction 13 Group 3: Trade, transportation, & utilities group n=393 4.05 .59 Wholesale trade 16 Retail trade 353 Transportation 14 Utilities 10 Group 4: Financial, business, & professional services group n=114 4.09 .64 Banking & finance 35 Insurance & real estate 39 Professional services 11 Services 24 Multi-industry holding company 3 E-companies 1 Software 1 Group 5: Education & Health group n=33 4.15 .60 Healthcare 24 Government 5 Foundations & non-profits 2 Education 2 Note: total n=680, "other or no response n=12.
25 explained by this term was again tested us ing hierarchical regression. See Table 7 for regression results. This intera ction term did not explain additional variance in job performance above the trait scores and two-way interaction terms, ( R2= .001, n.s ). The interaction term Conscientiousness by Ag reeableness was significantly correlated with managerial job performance in the educ ation and health indus try group (r= .41, p< .05). Moderated hierarchical linea r regression results showed th at the interaction term did not add significant incremental validity to the Conscientiousness measure, ( R2= .013, n.s ). Regression results for this industry can be found in Tables 8-10. The interaction terms, Extraversion by Neuroticism and Extraversion by Neuroticism by Conscientiousness, were also significantly correlated with performa nce ratings in this industry (r= .43, p< .05; r= .56, p< .01, respectively). Once ag ain, hierarchical regression results showed that the interaction terms did not explain variance beyond the control variables. Note that in Table 10 the two-wa y interaction terms were not included in the regression. This was a result of the colineari ty tolerance statisti c dropping below .000; SPSS would not produce an estimate for the re gression weight for the ExNxC term. As the three-way interaction term does not expl ain variance beyond the single trait scores, testing the two-way interact ions here is unnecessary. It is also important to note that th e job performance rati ngs were negatively skewed (skew= -.57, min= 1.8, max=5). These ratings are subject to range restriction effects that attenuate the co rrelations presented here. Thus the correlation coefficients may be underestimates of the true construct relationships and could explain the lack of support for the moderator hypotheses. The eff ects of range restriction on the data are discussed further in the next section.
26 Table 6 Correlations of Personality Traits w ith Performance Ratings by Industry Group 2: Construction & manufacturing Group 3: Trade, transportation, & utilities Predictor Job Performance Predictor Job Performance C .10C .15** A -.06A -.01 E .02E .01 N -.02N .04 CxA .03CxA .10 ExN .01ExN .06 ExNxC .08 ExNxC .16** Group 4: Financial, business, & professional services Group 5: Education & health Predictor Job Performance Predictor Job Performance C .02C .43* A -.06A .15 E .05E .17 N .02N .30 CxA -.02CxA .41* ExN .05ExN .43* ExNxC .06 ExNxC .56** *p< .05; **p< .01
27 Table 7 Hierarchical Regression of Perfor mance Ratings on ExNxC for Trade, Transportation, and Utilities Industry Variable Step 1 Step 2 Step 3 Step 1: Extraversion .03.321.28 Neuroticism .15*.142.05 Conscientiousness .21**.531.52 Step 2: ExN .02-1.46 ExC -.51-2.08 CxN -.01-1.73 Step 3: ExNxC 1.43 R2 .038* *.039*.040* R2 .038* *.002 .001 *p< .05, **p< .01 apparent discrepancy due to rounding
28 Table 8 Hierarchical Regression of Performance Ratings on CxA for Education and Health Industry Variable Step 1 Step 2 Step 1: Conscientiousness .41*-.92 Agreeableness .06-.82 Step 2: C x A 1.75 R2 .186*.199* R2 .186*.013 *p< .05
29 Table 9 Hierarchical Regression of Performance Ratings on ExN for Education and Health Industry Variable Step 1 Step 2 Step 1: Extraversion .40*-.24 Neuroticism .49*-1.08 Step 2: ExN 1.40 R2 .217*.234* R2 .217* .016 *p< .05
30 Table 10 Hierarchical Regression of Performance Ratings on ExNxC for Education and Health Industry Variable Step 1 Step 2 Step 1: Extraversion .31 1.13* Neuroticism .49** 2.40* Conscientiousness .41* 1.78* Step 2: ExNxC -2.17 R2 .375* .454** R2 .375* .079 *p< .05, **p< .01
31 Discussion This study examined whether personality tr aits interact to predict managerial work performance. Conscientiousness was si gnificantly related to performance ratings, supporting Hypothesis 1. This result is consis tent with meta-analytic findings that Conscientiousness is the best personality predictor of performance in complex jobs (Barrick & Mount, 1991). Agreeableness was exp ected to moderate th is relationship such that those high in both conscientiousness a nd agreeableness would receive the highest performance ratings. However, this interacti on term did not explain incremental variance beyond the Conscientiousness measure, indicati ng that this trait predicts managerial performance equally well across levels of agreeableness. The second hypothesized interaction was Ex traversion by Neuroticism. Managers higher in Extraversion and lower in Neurotic ism were expected to receive the highest performance ratings. This hypothesis was not supported. The three-way interaction between Extraversion, Neuroticism, and Cons cientiousness also failed to contribute incremental validity. Thus, the tr ait interactions examined in this study did not aid in the prediction of managerial job performan ce. Interestingly, Ex traversion was not a significant predictor in this study. Extraversi on was significantly re lated to managerial job performance in Barrick and Mounts ( 1991) meta-analysis a nd was the strongest predictor of leadership ratings in Bono & Judges (2004) meta-analysis. Perhaps those who did not project sufficient gregarious ness and social dominance would not have advanced to this position, thus, this samp le of incumbents would all be adequately
32 extraverted to manage people well. This possibl e range restriction is described further in the discussion of study limitations. Management is classified as an enterp rising job context in the RIASEC model (Holland, 1996). Foster and Macan found th at Agreeableness and Conscientiousness interacted to predic t job performance in realistic a nd conventional jobs, but not in enterprising jobs. The other ca reer types (investigative, ar tistic, and social) were not examined in their study. This study replicates their findings that in dividual personality scales, rather than trait interactions, pr edict job performance in enterprising job environments. As these types of jobs are more competitive and achievement-oriented, the behavioral tendencies associated with Consci entiousness, such as setting goals and being self-disciplined, may be the main drivers of success. Study Strengths This study advances the grow ing literature on moderators of the personalityjob performance relationship by showing that traits do not appear to moderate the predictive validity of other traits fo r manager jobs. A pervasive problem in job performance research is the quality of the criterion scor es. Specifically, when performance ratings are used to make administrative decisions su ch as promotions and salary increases, supervisors tend to give more lenient rati ngs (Bernardin & Orban, 1990). However, in this study, supervisors were aware that the pe rformance ratings were being collected for research purposes only, and held no administra tive consequences. Therefore, in this study it is unlikely that the ratings were subj ect to any systematic rating error. The personality measure used here was developed and validated by experts using a sound methodology for writing and choosing it ems that reflect the work context and
33 generalize across countries. Thus this measure is particularly useful for predicting work outcomes and selecting personnel within a cross-national applicant pool. Further, managers from diverse industries and organiza tions comprised the sample. This adds to the external validity of the study for the larger managerial population. Practical Implications Although trait interactions ha ve been shown to explain incremental variance in performance ratings in clerical jobs, jobs with high interpersonal components, and jobs in realistic and conventional contexts (Burke & Witt, 2002; Witt, Burke, Barrick, & Mount 2002; Foster & Macan, 2006), they do not appear to contribute to the prediction of overall managerial job performance. Huma n resource professionals looking to hire new managers are encouraged to select applican ts based on Conscientiousness scores, as well as other skills and abilities deemed vital to the job via job analysis Personality tests are particularly useful managerial selection tools as they do not produce adverse impact as do some tests of cognitive ability and they require less time and money than assessment centers. As conscientiousness was only modestly correlated (r=.14) with job performance, these measures are recommende d to be used in conjunction with other selection tools. The most efficient use of pe rsonality tests in the hiring process is as a prescreening hurdle; applicants who pass a pr eset cut score on Conscientiousness should proceed to participate in structured interv iews, situational judgment tests, and/or assessment centers to get a more complete picture of the candidates strengths and likelihood for success on the job.
34 Limitations and Directions for Future Research This study examined the relationships be tween personality in teractions and job performance for managerial incumbents. Ther efore some restriction of range in the predictors and criterion is likely to have attenuated the correlations presented here. Specifically, if the managers were hired usi ng personality measures or a correlated test, those scoring very low would not have been se lected, restricting the range of scores on the GPI scales. Similarly, low performing manage rs are less likely to be retained in the organizations, eliminating some of the lower scores on the criterion as well. The true effects of range restriction on the validity of trait interactions can be examined using a longitudinal, predictivevalidity study design. The criterion measure in th is study consisted of five items directed at overall managerial performance. The impact of le vel of Agreeableness on Conscientiousness, or Neuroticism on Extraversion, may only be releva nt for specific managerial competencies, such as coaching or motivating subordinates. The performance measure used here did not assess specific competencies or skills, thus I wa s not able to look at the interac tive effects of traits for dimensions of managerial job performance. Future re searchers may wish to explore these relationships. Specifically, Agre eableness may moderate the effectiveness of Conscientiousness for interp ersonal functions such as coaching and mentoring. Those highest in Agreeableness are likely to perfor m better as suggested by Witt et al.s (2002) study that found support for the Conscientiousne ss by Agreeableness interaction in jobs with primarily interpersonal requirements. Furthe r, they are less likely to engage in highmaintenance behaviors that may annoy s ubordinates and discourage them from approaching the manager (Burke & Witt, 2004).
35 Further, the trait in teractions tested in this study were chosen from theory and review of previous findings on this topic; however, other in teractions may exist that do explain incremental variance in overall mana gerial performance ratings. For example, Burke and Witt (2002) found evidence that Openness to experience interacted with both Extraversion and Neuroticism to predict in cremental variance in job performance for clerical workers. These interactions were not tested in this study. Research has also shown that some traits not directly measured in the five factor framework may be related to managerial performance. Proactive personali ty appears to be one that is important for predicting leadership charisma (Crant & Bateman, 2000) and may moderate the effectiveness of Conscientiousness. It may al so be that the more specific facets of personality traits interact to predict ov erall or competency performance ratings. A final direction for future research relates to the managers level in the organization. High level managers may serve more influencing and persuading functions, requiring them to be more extraverted. These executive level leaders are responsible for setting the vision and mission of their orga nizations and influencing employees to subscribe to the organizations values and goa ls, whereas low to mid level managers are typically involved in monitoring subordinates performance and dealing with day-to-day work issues. In executive leaders, the tendency toward social dominance and persuasiveness represented by higher Extraversi on scores would likely contribute to the efficacy of the achievement-striving and self -efficacy facets of Conscientiousness. Future research may wish to examine these interact ions within a sample of executive leaders.
36 Conclusion The trait interaction hypothe ses in this study were unsupported. Results from this study and previous research (Barrick & Mount 1991) indicate that Conscientiousness is the most useful personality trait for selecti ng managerial personnel. Personality tests are best used in concert with ot her valid selection tools, such as assessment centers and structured interviews. Although the interactions test ed here did not contribute incremental validity to the prediction of managerial job performance, additional research s hould address whether these findings hold for specific job competenci es rather than overall performance, for other personality trait in teractions, and for all levels with in the hierarchical structure of the organization.
37 References Barrick, M. R., & Mount, M. K. (1991). The big five personality dimensions and job performance: A meta-analysis. Personnel Psychology, 44 1-26. Bernardin, H. J., & Orban, J. A. (1990). Lenien cy effect as a function of rating format, purpose for appraisal, and ra ter individual differences. Journal of Business and Psychology, 5 (2), 197-211. Bono, J. E., & Judge, T. A. (2004). Personal ity and transformationa l and transactional leadership: A meta-analysis. Journal of Applied Psychology, 89 901-910. Bureau of Labor Statistics (last updated June 2004). http://www.bls.gov/bls/naics_aggregation.htm. Burke, L. A., & Witt, L. A. (2002). M oderators of the openness to experienceperformance relationship. Journal of Manage rial Psychology, 17 712-721. Burke, L. A., & Witt, L. A. (2004). Pe rsonality and high-maintenance employee behavior. Journal of Business and Psychology, 18 349-363. Champoux, J. E., & Peters, W. S. (1987). Fo rm, effect size and power in moderated regression analysis. Journal of Occupa tional Psychology, 60 243-255.
38 Chaplin, W. F. (1991). The next generation of moderator resear ch in personality psychology. Journal of personality, 59 143-178. Costa, P. T., Jr. & McCrae, R. R. (1992). Manual for the Revised NEO Personality Inventory (NEO-PIR) and the NEO Fi ve-Factor Inventory (NEO-FFI). Odessa, FL: Psychological Assessme nt Resources, Inc. Crant, J. M., & Bateman, T. S. (2000). Char ismatic leadership viewed from above: The impact of proactive personality. Journal of Organizational Behavior, 21 (1), 63-75. Foster, J. & Hogan, J. (2006). Profile analys es of personality-lead ership performance relations. Paper presented in M. Inge rick & L. M. Hough (symposium chairs) What Makes a Great Leader? Re fining the Personality-Leadership Relationship. 21st Annual Conferen ce of Society for Industrial and Organizational Psychology, May 2006, Dallas, Texas. Foster, J. Macan, T. (2006). The use of inte ractions between pers onality variables to predict performance. Poster presented at the 21st Annual C onference of Society for Industrial and Organizational Psychology, May 2006, Dallas, Texas. George, J. M., & Zhou, J. (2001). When ope nness to experience and conscientiousness are related to creative behavior : An interactional approach. Journal of Applied Psychology, 86 513-524. Gerstner, C. R., & Day, D. V. (1997). Meta-a nalytic review of l eader-member exchange theory: Correlates and construct issues. Journal of Applied Psychology, 82 827-844.
39 Goldberg, L. R. (1993). The structur e of phenotypic personality traits. American Psychologist, 48 26-34. Grensing-Pophal, L. (2001). High-maintenance employees. HR Magazine, 46, 86-91. Guion, R. M., & Gottier, R. F. (1965). Valid ity of personality measures in personnel selection. Personnel Psychology, 18 135-164. Hochwarter, W. A., Witt, L. A., & Kacmar, K. M. (2000). Perceptions of organizational politics as a moderator of the relatio nship between conscientiousness and job performance. Journal of Applied Psychology, 85 472-478. Hogan, R. (2005). In defense of personality measurement: New wine for old whiners. Human Performance, 18 331-341. Hogan, R., Hogan, J., & Roberts, B. W. (1996). Personality measurement and employment decisions: Questions and answers. American Psychologist, 51 469-477. Holland, J. L. (1996). Exploring careers with a typology: What we have learned and some new directions. American Psychologist, 51 (4), 397-406. Mintzberg, H. (1983). Power in and Around Organizations. Englewood Cliffs, NJ: Prentice Hall. Motowidlo, S. J., Borman, W. C., & Schmit, M. J. (1997). A theory of individual differences in task and contextual performance. Human Performance, 10 71-83.
40 Rothstein, M. G., & Goffin, R. D. (2006). The use of personality measures in personnel selection: What does current research support? Human Resource Management Review, 16 155-180. Schmit, M. J., Kihm, J. A., & Robie, C. ( 2002). The global personality inventory (GPI). In B. de Raad, & M. Perugini (Eds.), Big five assessment. (pp. 195-236)Hogrefe & Huber Publishers. Tett, R. P., Jackson, D. N., & Rothstein, M. ( 1991). Personality measur es as predictors of job performance: A meta-analytic review. Personnel Psychology, 44 (4), 703-742. Truxillo, D. M., Bauer, T. N., Campion, M. A ., & Paronto, M. E. (2006). A field study of the role of big five personality in applican t perceptions of selection fairness, self, and the hiring organization. International Journal of Selection and Assessment, 14 269277. Tupes, E. C., & Christal, R. C. (1958). Stabilit y of personality trait rating factors obtained under diverse conditions. USAF Wright Air Development Center Technical Note, (58), 16-16. Warr, P., Bartram, D., & Martin, T. ( 2005). Personality and sales performance: Situational variation and in teractions between traits. International Journal of Selection and Assessment, 13 87-91.
41 Witt, L. A., Andrews, M. C., & Carlson, D. S. (2004). When conscientiousness isn't enough: Emotional exhaustion and performa nce among call center customer service representatives. Journal of Management, 30 149-160. Witt, L. A., Burke, L. A., Barrick, M. A., & Mount, M. K. (2002). The interactive effects of conscientiousness and agreeableness on job performance. Journal of Applied Psychology, 87 164-169. Witt, L. A., & Ferris, G. R. (2003). Social skill as moderator of the conscientiousnessperformance relationship: Converg ent results acro ss four studies. Journal of Applied Psychology, 88 809-821.
43 Appendix A GPI Personality Dimensions and Facets (from Schmit, Kihm, and Robie, 2000, p. 28) Agreeableness Consideration Empathy Interdependence Openness Thought agility Trust Conscientiousness Attention to detail Dutifulness Responsibility Work focus Extraversion Adaptability Competitiveness Desire for achievement Desire for advancement Energy level Influence Initiative
44 Risk-taking Sociability Taking charge Neuroticism Emotional control (reverse-scored) Negative affectivity Optimism (reverse-scored) Self-confidence (reverse-scored) Stress tolerance (reverse-scored) Openness to Experience Independence Innovativeness/creativity Social astuteness Thought focus Vision
xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.loc.govstandardsmarcxmlschemaMARC21slim.xsd
leader nam Ka
controlfield tag 001 001993780
007 cr mnu|||uuuuu
008 090326s2008 flu s 000 0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0002411
Taylor, Amy M.
The validity of personality trait interactions for the prediction of managerial job performance
h [electronic resource] /
by Amy M. Taylor.
[Tampa, Fla] :
b University of South Florida,
Title from PDF of title page.
Document formatted into pages; contains 44 pages.
Thesis (M.A.)--University of South Florida, 2008.
Includes bibliographical references.
Text (Electronic thesis) in PDF format.
ABSTRACT: Personality variables have been shown to be significant predictors of job performance (Barrick & Mount, 1991; Tett, Jackson, & Rothstein, 1991). Recent advances in methodology for analyzing personality-job performance relationships indicate that interactions among traits may yield incremental validity. Job types in which performance has been shown to relate to trait interactions include clerical jobs, jobs with high interpersonal components, and jobs in realistic and conventional contexts, (Witt, Burke, Barrick, & Mount, 2002; Burke & Witt, 2002; and Burke & Witt, 2004). This study examined the validity of trait interactions for the prediction of managerial job performance. Hypotheses included a main effect for Conscientiousness, an interaction between Conscientiousness and Agreeableness, an interaction between Extraversion and Neuroticism, and finally, a three-way interaction between Extraversion, Neuroticism, and Conscientiousness.An archival dataset from Personnel Decisions, International (n=680 managers) containing GPI personality scores and supervisor-rated performance scores was analyzed to test the hypotheses. Correlations and moderated hierarchical linear regressions were performed to estimate the relationships of the predictors to the criterion, and to learn whether examination of trait interactions contributes incremental validity to the single trait scales. A main effect for Conscientiousness on managerial job performance was found. No trait interactions explained incremental variance in performance scores. Therefore, Conscientiousness is the recommended personality scale to use for selecting managers. This finding is consistent with previous research on the relation of Conscientiousness to job performance in managers (Barrick & Mount, 1991). Managers from diverse organizations and industries comprised the sample, increasing the generalizability of the results.Directions for future research include the examination of other trait interactions, more specific criteria such as competencies rather than overall managerial job performance, and effects of the hierarchical level of the manager in the organization.
Mode of access: World Wide Web.
System requirements: World Wide Web browser and PDF reader.
Advisor: Walter C. Borman, Ph.D.
t USF Electronic Theses and Dissertations.