6

PERSONNEL SELECTION AND VOCATIONAL INTERESTS

Recent Research and Future Directions

Frederick L. Oswald

RICE UNIVERSITY

Leaetta M. Hough

THE DUNNETTE GROUP

Chen Zuo

RICE UNIVERSITY

Many of us (present authors included) reached our current jobs by variously walking, skipping, and staggering through our occupational development. A high school biology teacher might turn you away from an interest in medicine—forever. You might watch a TV show about dangerous jobs, where some of those actually seem interesting and realistic to pursue. On the job, you might meet after work with your colleagues to work on electronics projects that grow your knowledge and stimulate further interests. To some extent, all of these scenarios reflect a range of vocational interests (e.g., Realistic, Investigative, Artistic, Social, Enterprising and Conventional [RIASEC] interests) that come to light as a function of the person selecting and interacting with the situation at hand.

For decades, research and application involving vocational interests have fallen within the purview of career counselors and counseling researchers, who ultimately seek to inform students of their occupational potential and occupational choices. Interests pervade the cycle of education and work, long before someone becomes a job applicant (Figure 6.1). As early as kindergarten, vocational information is deliberately conveyed to children in the classroom. One of the reasons to do so (among others) is to develop an early and broad classroom interest in the STEM fields (i.e., science, technology, engineering, and math), that ideally will strengthen, leading to greater STEM-related confidence and competencies as one enters college (Le, Robbins, & Westrick, 2014), with the long-term hopes of increasing economic competitiveness and prosperity at a national level. We know with great certainty that students’ vocational interests, even more than a decade later, can reliably predict occupational choice (Austin & Hansich, 1990), employment status, and income (Stoll et al., 2017). Although much is known about the predictive value of an individual’s vocational interests, we need a better understanding of the developmental process of interests (in STEM and otherwise) that shape a student’s educational and career pursuits long before personnel selection enters into this process. Theories of vocational interest describe how general factors such as occupational exposure (e.g., in the classroom, at home, with peers, and in the media), and a strong sense of a career as a calling (Kaminsky & Behrend, 2015), translate into individuals’ later decisions and commitments over time that lead to gaining deeper occupational knowledge and skill (Hidi & Renninger, 2006). A little occupational or vocational knowledge early on can go a long way to shape interests—in tandem with information from guidance counselors, parents, and peers—to enhance a variety of behaviors and strategies when students are readying themselves for success in school and the workplace (e.g., course taking, extracurricular activities technical training, choice of college major, job search).

Image

FIGURE 6.1 Interests Permeate the Education and Employment Cycle.

In this context, measures of vocational interests usually report information about how a respondent’s interest profile matches the profiles of a wide range of occupations. Whereas the individual (or job candidate) profile of interests is almost always determined through taking a vocational interest inventory, the interest profiles for occupations can be determined using several approaches.

The first way one might determine interest profiles for occupations is through a job analysis. The Department of Labor’s O⋆NET occupational data base serves as a prime example, where job analysts have provided data on RIASEC characteristics across all available occupations (see https://www.onetonline.org/find/descriptor/browse/Interests/). The database of interest profiles is constantly being updated as new occupations are added to the O⋆NET taxonomy (Rounds, Su, Lewis, & Rivkin, 2013).

A second reasonable way to determine interest profiles for occupations is by calculating the average interest profile of employees within each occupation. The assumption here is that when particular interests are higher for a particular occupation (vs. others), then those interests are job relevant for that occupation (see, for example, the Strong Interest Inventory; Harmon, Hansen, Borgen, & Hammer, 1994). This assumption may not be true, of course, and in low-stakes developmental settings, this may not matter; however, if interests are to be used for personnel selection purposes, then the job relevance of vocational interests and interest profiles becomes more important to the organization for ensuring validity and minimizing adverse impact.

A third approach is more indirect than the previous two, as it involves examining complementary profiles. With this approach, job applicants’ profiles of vocational interests (e.g., RIASEC scores) are matched with occupational profiles of characteristics that are relevant to those interests. For example, organizations often seek knowledge, skills, abilities, and other characteristics (KSAOs) that are interest-related and address their business needs, such as the need for technical knowledge and teamwork. These KSAOs might represent the organizational profile. As another example, an organization’s policies or management will vary in terms of their ability to reward applicants’ interest-related outcomes (e.g., greater autonomy, security, variety). The ability to reward these different outcomes could indeed be represented as an organizational profile. Because these organizational profiles are complementary, they relate to applicants’ vocational interest profiles indirectly (not one-to-one). They are thus more conceptually challenging to develop, and more empirically difficult to match to the applicant profile. Nonetheless, organizational profiles are important: They are motivated by the things that applicants and organizations both have to offer one another—and the things that both of them ultimately want—where none of these things may be the same for either party. This sort of complementarity is featured prominently in the Theory of Work Adjustment (e.g., Dawis, 2005).

Note that many vocational interest and work adjustment theories incorporate all three of these profile-based approaches (e.g., Dawis & Lofquist, 1984; Holland, 1997). Also, as the discerning reader might appreciate by now, there is a remarkable alignment between these vocational approaches and those theories found in the organizational literature regarding employees’ fit to jobs, occupations, and organizational environments (e.g., Edwards, Cable, Williamson, Lambert, Shipp, 2006; Ehrhart & Makransky, 2007; Judge, 1994; Kristof-Brown, Zimmerman, & Johnson, 2005).

Vocational Interests in Personnel Selection Research

Industrial-organizational psychologists have extended the phenomenon of interest-job fit to personnel selection using measures that are more traditionally and directly aligned with those found in the vocational counseling literature than in the industrial-organizational literature. For example, in a large military sample of incumbents (N = 408–418), a web-based RIASEC measure of vocational interests was found to yield cross-validated R2 values of .02 to .09 for predicting technical and interpersonal knowledge, task proficiency, effort, and continuance intentions in separate linear regression models (Van Iddekinge, Putka, & Campbell, 2011). Some of the zero-order correlations with other self-report measures, although exploratory, are also noteworthy: realistic interests correlated -.26 with concern for others; artistic interests correlated .31 with innovation; and enterprising interests correlated .23 with leadership orientation. Although these findings involved an incumbent sample as opposed to a higher-stakes applicant sample, the authors do correct for range restriction effects (as there was direct selection on the Armed Services Vocational Aptitude Battery (ASVAB), which correlates with interests). These large-sample findings at least show some initial promise for the use of vocational interests measures in an employment setting.

That said, for vocational interest measures to be useful operationally, they need to show incremental validity above more typical cognitive and personality tests used in employment settings. When that analysis was applied by the authors (see Van Iddekinge, Putka, & Campbell, 2011, Table 3), the incremental R2 values were only .01 at most over cognitive ability and personality for predicting these outcomes. By contrast, when the outcome is the type of employment that one attains after college, RIASEC vocational interests predict well and predict incrementally above the Big Five personality measures (De Fruyt & Mervielde, 1999), reinforcing the profile-matching approach when determining broader person-occupation fit.

Organizations and personnel selection psychologists alike are keenly concerned about mean score differences between ethnic/racial and gender groups, because in upholding Title VII of the Civil Rights Act of 1964, they seek to guard against the use of selection procedures that result in adverse impact and employment discrimination. In the Van Iddekinge, Putka, and Campbell (2011) study, gender differences favoring males were large for Realistic interests (males, d = .83), whereas moderate differences favoring females were found for Social and Conventional interests (d = -.45 and -.34, respectively). Black/White race differences were moderate-to-large for Realistic interests, favoring Whites (d = .62) and small-to-moderate for Artistic, Social, Enterprising, and Conventional, all favoring Blacks (d values of -.38, -.50, -.39 and -.47, respectively). In a much larger military study (Hough, Barge, & Kamp, 2001; Hough & Barge, 2001), similar mean score differences were found between men and women (male sample size ~7,500, female sample size ~850) and between Whites and Blacks (White sample ~5,500, Black sample ~2,200). Likewise, these patterns of findings are rather similar to those found in larger meta-analyses of gender (Su, Rounds, & Armstrong, 2009) and both gender and race (Jones, 2013). More continuous variables might help explain some of these group differences. For example, gender differences may be at least partially explained by continuous measures of masculinity (instrumentality) and femininity (expressiveness; Ludwikowski, Armstrong, & Lannin, 2018). But regardless, the magnitudes of many of these reported race and gender differences are important in that they may or may not tend to result in adverse impact against a protected class.

Two meta-analyses of vocational interests in the workplace complement the findings above. In the first one, the authors (Van Iddekinge, Roth, Putka, & Lanivich, 2011) compared the validity of interests (RIASEC and otherwise) for predicting several organizational outcomes (i.e., job performance, training performance, turnover intentions, and turnover). Several types of validities for vocational interests were provided: the single-interest scale most relevant to the job (e.g., Realistic interests for a hand-on military job), a regression-weighted composite of interest scales, and congruence indices (i.e., any of several indices that relate the interest profiles of applicants with those from jobs or organizations). Operational validities (i.e., validities corrected for measurement error variance in the criterion) were the highest for regression-weighted interest composites. These validities were shrinkage-corrected to avoid capitalizing on chance and were r ≈ .14 for performance, r ≈ .15 for training, r ≈ -.24 for turnover intentions, r ≈ -.13 for actual turnover. Importantly, validities across organizational outcomes were consistently higher when the interest measure was focused on specific job-relevant interests (magnitudes for validity are about .20).

The second is a pair of meta-analyses by the same authors (Nye, Su, Rounds, & Drasgow, 2012, updated and expanded in 2017) that focused on a comparison of the validity of individual interest scales with interest congruence indices. Here, after correcting for both incidental range restriction and measurement error variance in the criterion, the magnitudes of correlations based on interest profiles were uniformly higher than correlations based on individual scales across all organizational criteria: task performance, organizational citizenship behaviors, persistence, and counterproductive work behaviors (for CWBs, also see Iliescu, Ispas, Sulea, & Ilie, 2015). The authors suggested that profiles had higher validity because each profile reflects relative levels of interests aligned with the job or organization, and thus reflects the motivational characteristics of applicants and employees more so than scale scores, which reflect absolute levels of particular interests. Another possibility is that like regression analysis, profiles simply allow more variables to be involved in prediction; these profiles can be “driven” either by multiple variables (i.e., an interaction effect reflecting unique profile characteristics) or by single variables (i.e., a particularly high or low point within a given profile.).

In short, the revival of vocational interests in the domain of personnel selection is promising and very much welcome. Recent research findings are compelling enough to stimulate a wealth of selection-related research and practice that is likely to benefit both employees and organizations alike.

Future Research Directions

Taken together, the studies and meta-analyses described above offer several suggestions for researching vocational interests in personnel selection. First, research can explore how adding interest measures will affect protected classes (e.g., women and Black subgroups) in terms of reductions in adverse impact, also considering those reductions in light of criterion-related validity. Second, research can explore how interests relate to the modern circumstances of a “gig economy,” which has implications for how interests might be measured and used to predict outcomes relevant to employers and job applicants in such an economy. Third, vocational interests that are specific to jobs can be examined further, given the meta-analytic evidence for higher criterion-related validity of specific (vs. general) interests for predicting performance ratings; historically, there was promise for specific interests reducing turnover as well (Bolanovich, 1948). Fourth, if interest measures are to be used in high-stakes personnel selection settings, then continued research on faking such measures is of concern, as it has been for decades with personality testing (see Garry, 1953, who showed how two samples of college students could elevate their vocational interest scores under instructed faking conditions).

Subgroup Differences and Adverse Impact

Because vocational interest measures are rarely used in the employment setting, group mean differences in interests are rarely examined; for instance, they were not mentioned in a wide-ranging research review of group differences found across constructs often used in personnel selection (Hough, Oswald, & Ployhart, 2011). However, we have long known that vocational interest measures are reliable, and they are remarkably stable over time like traits (even over twenty years; see the longitudinal meta-analysis of interests by Low, Yoon, Roberts, & Rounds, 2005). Vocational interests also show at least some evidence for validity and incremental validity, as noted above; and all of this is to justify the final point, that along with the evidence for reliability, stability, and validity, vocational interests also demonstrate durable patterns of mean differences by gender and race/ethnicity.

Personnel selection researchers are well aware of the “validity-adverse impact tradeoff,” namely, that in many selection settings, cognitive ability measures tend to show the highest validity for predicting job performance, yet they also tend to show the largest mean differences by racial/ethnic subgroups, such as a .6 to 1.0 standard deviation difference between the means of African Americans and Whites in the selection context (see Roth, Bevier, Bobko, Switzer, & Tyler, 2001, and the review by Ones, Dilchert, Viswesvaran, & Salgado, 2017). If a cognitive ability test is included in a selection battery, research has demonstrated mathematically that this mean difference is very difficult to remove. For instance, if a cognitive ability measure has a group mean difference of d = 1.0, and an uncorrelated measure of conscientiousness is added to the selection battery d = .00 (no group mean difference), then the mean difference of the composite is not d = .50 as one might expect through averaging, but rather it is d = .71 (the square root of .50; see Sackett & Ellingson, 1997).

A more effective way to counter the adverse impact for cognitive ability would be to locate measures of constructs that are job-relevant yet show an opposite mean difference that favors the minority group. Such a tentative possibility is found with vocational interest measures: Given the d values that are in the .40s but favor African Americans (vs. Whites) for Artistic, Social, Enterprising, and Conventional interests (see Van Iddekinge, Putka, & Campbell, 2011, above), and given that these interests correlate low—and sometimes even negatively—with cognitive ability (Ackerman & Heggestad, 1997), then introducing these interest scales into a selection composite that included cognitive ability could greatly reduce the d-value of the overall selection composite. But this is a tentative suggestion that would need to be implemented with professional care.

Taking a conservative approach (yielding a higher estimate of d), let us assume that a d value between African Americans and Whites involving cognitive ability was as high as d = 1.0, and Enterprising and Social d values were -.39 and -.47 favoring African Americans as reported above; let us also assume the correlations between these three constructs were low, say zero even. Then using the formula for composites (Sackett & Ellingson, 1997), the composite d value would be d = .08. Note that this point in no way solves the “diversity-adverse impact tradeoff” because the incremental validity for vocational interests over cognitive ability is often quite low, although it depends upon the outcome to be predicted. Additional selection research involving interests is needed that pursues the real-world nature of the diversity-validity tradeoff in organizations (see a series of useful recommendations by Ployhart & Holtz, 2008).

Interests in the “Gig Economy”

Typical vocational interest measures center on one’s interest in the skills, activities, and fit of traditional full-time occupations. But consider the gig economy, where organizations call upon people and teams to take on specific jobs or tasks on a temporary (if not fleeting) basis; where the work may be increasingly remote (e.g., work from home online); and the employment relationship might reflect less organizational loyalty and greater economic risks or burdens on the part of the employee (Friedman, 2014). New interest measures should not only involve different modes of administration (e.g., video-based, adaptive, online) but also be tailored to reflect the diverse, yet more specific, work found in the gig economy. This approach has also been recently advocated in the military, in order for recruits to understand that a wide range of possibilities for work and careers is possible, extending well beyond combat-related military occupational specialties (Ingerick & Rumsey, 2014).

Specific Interest Areas

Future research could determine whether these aforementioned types of tailored or specific interest measures might yield higher than the typical near-zero incremental validity for predicting overall job performance when added to personality and cognitive ability test batteries. We suspect that such tailored, specific interest inventories are likely to produce incremental validity, especially when outcomes such as turnover are considered.

The Big Five factor model of personality has been examined in terms of more refined facets, such as a measure of two facets within each of the Big Five (e.g., Industriousness and Orderliness within Conscientiousness; DeYoung, Quilty, & Peterson, 2007). Likewise, RIASEC interest factors have been examined at a facet level, such as the 23 Basic Interest Scales [BIS] measured in the Strong Interest Inventory, where Adventure, Agriculture, Nature, Military, and Mechanical Activities reside within Realistic interests (Hansen & Campbell, 1985); or the 31 Basic Interest Markers that are publicly available (Liao, Armstrong, & Rounds, 2008). Therefore, going beyond an analysis at the broad factor level, where for instance the Extraversion personality factor is related to the Social interest factor (Mount, Barrick, Scullen, & Rounds, 2005), an analysis at a more refined level of measurement interest level may reveal specific vocational interests that distinguish themselves from specific aspects of personality and thus have greater potential for incremental validity (for inspiration, see Larson & Borgen, 2002).

We noted earlier that Van Iddekinge, Roth et al. (2011) found higher validities for vocational interests predicting performance when interest measures were focused on specific jobs. Likewise, Gutentag and Gati (2016) enumerated 31 specific aspect-based career preferences that include vocational interests, but other aspects as well (e.g., amount of travel, management, indoors-outdoors, income). With four samples of college students (reflecting two cultures) who were administered the Career Preferences Questionnaire (CPQ; Gati & Asher, 2001), the authors found that these 31 aspects varied greatly in their stability; but clustering them suggested job-related interest areas beyond RIASEC, of concern to young and educated employees (e.g., interest in the work “package” that includes income, work hours, and prestige; interest in advancement and going into management; interest in organizational skills and teamwork). These clusters and other, more refined interest clusters could be developed (e.g., interest in combat vs. combat support in the military; Hough, Barge, & Kamp, 2001). Clusters could then, for instance, be aligned with the O⋆NET occupational database to enhance career development and selection tools beyond RIASEC interests. Refined interest clusters also could encourage criterion development and even a program of selection research at a commensurate level of refinement, when appropriate (e.g., how does a refined interest profile help an applicant select from among multiple job offers?).

Mean differences by race and/or gender in specific interest areas (as opposed to general RIASEC scores) may also help understand why such differences occur, with implications for how one might intervene (e.g., via recruitment, selection, and training) to mitigate those differences. In the U.S. Air Force context (Johnson, Trent, & Barron, 2017), a large basic training sample (N = 1,008) responded to interest-related items. Large item-level gender differences were found in participants’ interest in job contexts (women tend to prefer jobs that are indoors, in the office, nonhazardous, and predictable; d ≈ .25-.50) and work environments (women tend to prefer work involving analyzing, maintaining, and creating documents; and interacting, training, and serving people and customers; d ≈ .25-.65). Regarding race and job context: African Americans generally preferred jobs that are indoors, in the office, nonhazardous, and predictable, with d ≈ .50-.65 compared to Whites. In terms of work characteristics, African Americans relative to Whites also generally preferred analyzing, maintaining, and creating documents; and interacting, training, and serving people and customers, with d ≈ .25-.60. Interestingly, the authors found these gender and race effects to be independent; that is, race and gender do not appear to be confounded, despite roughly similar patterns of mean differences.

Faking and the Measurement of Interests

Given the infrequency that vocational interests have been used in high-stakes settings, such as personnel selection, intentional distortion (e.g., lying) with regard to one’s measured interests is an area ripe for research. Earlier military research indicates this need, where Hough, Barge, and Kamp (2001) administered a specially constructed interest inventory in a Military Entrance Processing Station, an applicant-like setting where enlistees completed the interest inventory and were told their responses would be used to make decisions about their careers in the military. In addition to this field setting, this interest inventory was also administered to military participants in three experimental conditions: fake a combat applicant, fake a noncombat applicant, and respond-honestly. Perhaps unsurprisingly, results indicated that when instructed to do so, soldiers could distort their responses in the experimental faking conditions. However, comparisons of data from the applicant-like setting and the honest experimental condition did not suggest any particular pattern to the mean score differences, although much more needs to be learned in future research about the impact of high-stakes settings on applicant responses to interest measures.

Summary and Conclusions

In the long term, it remains to be seen how frequently and effectively vocational interest measures might be implemented in personnel selection settings. As with any new constructs and new measures that are introduced, both organizations and applicants alike may have reservations—or at least they would like to obtain a better understanding of the benefits and drawbacks from empirical, rational, and practical standpoints, in comparison with the current state of affairs. More specifically, organizations and HR experts likely have limited knowledge about the nature of vocational interests, its benefits in personnel selection in terms of potential incremental validity and reductions in adverse impact, and its potential legal liabilities. Given the organizational research just reviewed, academics, practitioners, and organizations themselves likely agree that “more research is needed” to understand how, in what contexts, with what measures, at what level of refinement, and for what criteria vocational interests will be most effective for selecting employees in organizations.

Note: The authors wish to thank Amy Shaw, graduate student at Rice University, for contributing an initial search of existing vocational interest measures.

References

  1. Ackerman, P. L., & Heggestad, E. D. (1997). Intelligence, personality, and interests: Evidence for overlapping traits. Psychological Bulletin, 121, 219–245.

  2. Austin, J. T., & Hansich, K. A. (1990). Occupational attainment as a function of abilities and interests: A longitudinal analysis using Project TALENT data. Journal of Applied Psychology, 75, 77–86.

  3. Bolanovich, D. J. (1948). Reduce factory turnover. Personnel Psychology, 1, 81–92.

  4. Dawis, R. V. (2005). The Minnesota theory of work adjustment. In S. D. Brown & R. W. Lent (Eds.), Career development and counseling: Putting theory and research to work (pp. 3–23). Hoboken, NJ: Wiley.

  5. Dawis, R. V. & Lofquist, L. H. (1984). A psychological theory of work adjustment. Minneapolis: University of Minnesota Press.

  6. De Fruyt, F., & Mervielde, I. (1999). RIASEC types and Big Five traits as predictors of employment status and nature of employment. Personnel Psychology, 52, 701–727.

  7. DeYoung, C. G., Quilty, L. C., & Peterson, J. B. (2007). Between facets and domains: 10 aspects of the Big Five. Journal of Personality and Social Psychology, 93, 880–896.

  8. Edwards, J. R., Cable, D. M., Williamson, I. O., Lambert, L. S., & Shipp, A. J. (2006). The phenomenology of fit: Linking the person and environment to the subjective experience of person-environment fit. Journal of Applied Psychology, 91, 802–827.

  9. Ehrhart, K. H., & Makransky, G. (2007). Testing vocational interests and personality as predictors of person-vocation and person-job fit. Journal of Career Assessment, 15, 206–226.

  10. Friedman, G. (2014). Workers without employers: Shadow corporations and the rise of the gig economy. Review of Keynesian Economics, 2, 171–188.

  11. Garry, R. (1953). Individual differences in ability to fake vocational interest tests. Journal of Applied Psychology, 37, 33–37.

  12. Gati, I., & Asher, I. (2001). The PIC model for career decision making: Prescreening, in-depth exploration, and choice. In F. T. Leong & A. Barak (Eds.), Contemporary models in vocational psychology (pp. 7–54). Mahwah, NJ: Erlbaum.

  13. Gutentag, T., & Gati, I., (2016). The consistency and structure of aspect-based career preferences. Journal of Vocational Behavior, 93, 33–46.

  14. Hansen, J. C., & Campbell, D. P. (1985). Manual for the Strong Interest Inventory (4th ed.). Stanford, CA: Stanford University Press.

  15. Harmon, L. W., Hansen, J. C, Borgen, F. H., & Hammer, A. L. (1994). Strong Interest Inventory applications and technical guide. Palo Alto, CA: Consulting Psychologists Press.

  16. Hidi, S., & Renninger, K. A. (2006). A four-phase model of interest development. Educational Psychologist, 41, 111–127.

  17. Holland, J. L. (1997). Making vocational choices: A theory of vocational personalities and work environments (3rd Ed.). Odessa, FL: Psychological Assessment Resources.

  18. Hough, L. M., & Barge, B. (2001). Development of vocational interests and needs inventories: AVOICE and JOB (Institute Report #260). Minneapolis, MN: Personnel Decisions Research Institutes.

  19. Hough, L. M., Barge, B., & Kamp, J. (2001). Assessment of personality, temperament, vocational interests, and work outcome preferences. In J. P. Campbell & D. J. Knapp (Eds.), Exploring the limits of personnel selection and classification (pp. 111–154). Mahwah, NJ: Erlbaum.

  20. Hough, L. M., Oswald, F. L., & Ployhart, R. E. (2001). Determinants, detection and amelioration of adverse impact in personnel selection procedures: Issues, evidence and lessons learned. International Journal of Selection and Assessment, 9(1–2), 152–194.

  21. Iliescu, D., Ispas, D., Sulea, C., & Ilie, A. (2015). Vocational fit and counterproductive work behaviors: A self-regulation perspective. Journal of Applied Psychology, 100, 21–39.

  22. Ingerick, M., & Rumsey, M. (2014). Taking the measure of work interests: Past, present, and future. Military Psychology, 26, 165–181.

  23. Johnson, J. F., Trent, J. D., & Barron, L. G. (2017). Gender, racial, and ethnic differences in job context and work-activity preferences. Military Psychology, 29, 542–559.

  24. Jones, K. S. (2013). Considering vocational interests in adverse impact: A meta-analysis of race, cognitive ability, and vocational interests (Doctoral dissertation). University of Illinois at Urbana-Champaign.

  25. Judge, T. A. (1994). Person–organization fit and the theory of work adjustment: Implications for satisfaction, tenure, and career success. Journal of Vocational Behavior, 44, 32–54.

  26. Kaminsky, S. E., & Behrend, T. S. (2015). Career choice and calling: Integrating calling and social cognitive career theory. Journal of Career Assessment, 23, 383–398.

  27. Kristof-Brown, A. L., Zimmerman, R. D., & Johnson, E. C. (2005). Consequences of individuals’ fit at work: A meta-analysis of person-job, person-organization, person-group, and person-supervisor fit. Personnel Psychology, 58, 281–342.

  28. Larson, L. M., & Borgen, F. H. (2002). Convergence of vocational interests and personality: Examples in an adolescent gifted sample. Journal of Vocational Behavior, 60, 91–112.

  29. Le, H., Robbins, S. B., & Westrick, P. (2014). Predicting student enrollment and persistence in college STEM fields using an expanded PE fit framework: A large-scale multilevel study. Journal of Applied Psychology, 99, 915–947.

  30. Liao, H. -Y., Armstrong, P. I., & Rounds, J. (2008). Development and initial validation of public domain Basic Interest Markers. Journal of Vocational Behavior, 73, 159–183.

  31. Low, K. S., Yoon, M., Roberts, B. W., & Rounds, J. (2005). The stability of vocational interests from early adolescence to middle adulthood: A quantitative review of longitudinal studies. Psychological Bulletin, 131, 713–737.

  32. Ludwikowski, W. M., Armstrong, P. I., & Lannin, D. G. (2018). Explaining gender differences in interests: The roles of instrumentality and expressiveness. Journal of Career Assessment, 26, 240–257.

  33. Mount, M. K., Barrick, M. R., Scullen, S. M., & Rounds, J. (2005). Higher‐order dimensions of the big five personality traits and the big six vocational interest types. Personnel Psychology, 58, 447–478.

  34. Nye, C. D., Su, R., Rounds, J., & Drasgow, F. (2012). Vocational interests and performance: A quantitative summary of over 60 years of research. Perspectives on Psychological Science, 7, 384–403.

  35. Nye, C. D., Su, R., Rounds, J., & Drasgow, F. (2017). Interest congruence and performance: Revisiting recent meta-analytic findings. Journal of Vocational Behavior, 98, 138–151.

  36. Ones, D. S., Dilchert, S., Viswesvaran, C., & Salgado, J. F. (2017). Cognitive ability: Measurement and validity for employee selection. In J. Farr & N. Tippins (Eds.), Handbook of employee selection (2nd ed., pp. 251–276). New York: Routledge.

  37. Ployhart, R. E., & Holtz, B. C. (2008). The diversity-validity dilemma: Strategies for reducing racioethnic and sex subgroup differences and adverse impact in selection. Personnel Psychology, 61, 153–172.

  38. Roth, P. L., Bevier, C. A., Bobko, P., Switzer III, F. S., & Tyler, P. (2001). Ethnic group differences in cognitive ability in employment and educational settings: A meta‐analysis. Personnel Psychology, 54, 297–330.

  39. Rounds, J., Su, R., Lewis, P., & Rivkin, D. (June, 2013). Occupational Interest Profiles for new and emerging occupations in the O* NET system: Summary. Raleigh, NC: National Center for O⋆NET Development.

  40. Sackett, P. R., & Ellingson, J. E. (1997). The effects of forming multi-group predictor composites on group differences and adverse impact. Personnel Psychology, 50, 707–721.

  41. Stoll, G., Rieger, S., Lüdtke, O., Nagengast, B., Trautwein, U., & Roberts, B. W. (2017). Vocational interests assessed at the end of high school predict life outcomes assessed 10 years later over and above IQ and Big Five personality traits. Journal of Personality and Social Psychology, 113, 167–184.

  42. Su, R., Rounds, J., & Armstrong, I. (2009). Men and things, women and people: A meta-analysis of sex differences in interests. Psychological Bulletin, 135, 859–884.

  43. Van Iddekinge, C. H., Putka, D. J., & Campbell, J. P. (2011). Reconsidering vocational interest for personnel selection: The validity of an internet-based selection test in relation to job knowledge, job performance, and continuance intentions. Journal of Applied Psychology, 96, 13–33.

  44. Van Iddekinge, C. H., Roth, P. L., Putka, D. J., & Lanivich, S. E. (2011). Are you interested? A meta-analysis of relations between vocational interests and employee performance and turnover. Journal of Applied Psychology, 96, 1167–1194.