4

INTEREST MEASUREMENT

Oleksandr S. Chernyshenko

NANYANG TECHNOLOGICAL UNIVERSITY, SINGAPORE

Stephen Stark

UNIVERSITY OF SOUTH FLORIDA, USA

Christopher D. Nye

MICHIGAN STATE UNIVERSITY, USA

Vocational interests are among the most important classes of individual differences in applied psychology (Chernyshenko, Stark, & Drasgow, 2010) and can be simply defined as “preferences for certain work activities” (Hansen,1984; Van Iddekinge, Putka, & Campbell, 2011). To measure interests, one typically asks examinees to rate, rank, or choose between work activities, school subjects, or occupations. Because numerous studies have linked interests to occupational preferences (Campbell, 1971; Clark, 1961; Holland, 1997; Kuder, 1977; Strong, 1943), interest measures are widely used by vocational and career counselors in guiding individuals’ choices of study majors, occupations, and alternative career paths. Vocational interests are also gaining traction in personnel selection and classification research as valid predictors of job performance, satisfaction, and turnover (Bizot & Goldman, 1993; Nye, Su, Rounds, & Drasgow, 2012; Van Iddekinge, Roth, Putka, & Lanivich, 2011). Not surprisingly, in the past one hundred years, numerous vocational interest measures have been developed. Readers of this chapter are likely to be among the tens of millions of people who have taken at least one interest measure at some point in their lives, including either the Strong Interest Inventory (SII; Harmon, Hansen, Borgen, & Hammer, 1994), the Kuder Occupational Interest Survey (KOIS; Kuder & Zytowski, 1991), or the Self-Directed Search (SDS; Holland, Fritzsche, & Powell, 1994).

In this chapter, we do not intend to revisit the history of interest measurement nor do we intend to provide descriptions of specific measures as such reviews have already appeared elsewhere (e.g., Harrington & Long, 2013). Our goal is to focus primarily on psychometric considerations involved in the measurement of vocational interests. First, we will discuss dimensions measured by vocational interest inventories. We will review broad factor taxonomies (a.k.a., general occupational themes) and discuss why basic interests have been recently receiving more attention. Next we will focus on the kinds of stimuli and response formats that are used to collect vocational interest preference data. Third, we will mention some psychometric models that are used to scale and score various interest assessments and highlight some of the recent research involving ideal point item response theory (IRT) models. We will conclude the chapter with some suggestions for future directions.

Dimensions Assessed by Interest Inventories

Hansen (1984) and Day and Rounds (1997) noted that nearly all vocational interest taxonomies can be organized into three levels of specificity with general interest factors representing the broadest level, basic interests representing the middle level of specificity, and occupational scales representing the narrowest job-specific interests. Below we highlight some of the most influential interest taxonomies and applications.

The Search for a Broad Factor Taxonomy

The quest to understand the factorial structure of vocational interests has largely mirrored efforts in other individual difference domains (e.g., Carroll, 1993; McCrae & Costa, 1987). The earliest interest inventories, such as the Carnegie Interest Inventory (Carnegie Institute of Technology, 1920) and the Strong Vocational Interest Blank (SVIB; Strong, 1927), addressed primarily applied concerns, for example, documenting the occupational preferences of men and women or trying to differentiate the characteristics of various occupations (e.g., sales engineers vs. design engineers). Soon, however, researchers turned their attention to analyzing responses to interest questions. Thorndike (1935), for example, analyzed Carnegie Interest Inventory data and reported that ratings of sixteen activity preferences were stable over a two-year period. He also noted that, unlike the high correlations observed among measures of cognitive abilities, interests exhibited a more complex pattern of relations, with some interests correlating positively and others correlating negatively or not at all. Thurstone (1931) was among the first researchers to apply newly developed factor analysis methods to interest data. He found, for example, that correlations among interests in eighteen occupations measured using the SVIB could be accounted for by just four factors, which he labeled interest in science, language, people, and business. Strong (1943) later factor analyzed four additional correlation matrices, comprising increasingly larger numbers of occupations, and also found that four to five factors, similar to those found by Thurstone, were “sufficient to account mathematically for all or nearly all of the variations in interests among the occupational groups so far studied” (p. 147). Finally, Ferguson, Humphreys, and Strong (1941) factor analyzed a correlation matrix of several interests and values scales and identified three of the four factors described by Thurstone (i.e., interest in -language, -people, and -science). These and other early papers indicated that the interest domain is multidimensional and likely to have a replicable underlying factor structure.

The next wave of taxonometric studies employed a variety of analytic techniques (factor analysis, cluster analysis, multidimensional scaling) and resulted in several researchers proposing comprehensive general factor models of vocational interests. Guilford, Christensen, Bond, and Sutton (1954) conducted a landmark study where they analyzed preferences for one hundred work activities and found six broad interest factors, often called General Occupational Themes: Mechanical, Scientific, Aesthetic Expression, Social Welfare, Business, and Clerical. These factors were very similar to those proposed by Holland (1959, 1997) as part of his hexagonal RIASEC (Realistic, Investigative, Artistic, Social, Enterprising, and Conventional) model of vocational interests. At about the same time, Roe (1956) proposed a circular model of interests including eight factors (Technology, Science, Outdoor, Arts and Entertainment, Service, General Culture, Organization, and Business). Her model was very similar to the RIASEC model, but split the Realistic factor into Mechanical and Outdoor interests and the Social factor into Service and General Culture interests.

Both Holland and Roe essentially based their models on a two-dimensional multidimensional scaling solution and, as was noted by Meir and Ben-Yehuda (1976), were essentially the same models. Prediger (1976, 1982) and Prediger and Vansickle (1992) defined these two broad dimensions as people-things and ideas-data. Building on their work, Tracey and Rounds (1996) proposed a spherical model of vocational interests that incorporated Prediger’s two bipolar dimensions and a third, orthogonal, prestige dimension.

Finally, Gati (1979, 1991), after conducting a series of factor analyses, proposed a hierarchical model of interests. He postulated that the highest level of interests had two general factors, Soft and Hard Sciences, which could be progressively subdivided to form any number of hexagonal or circular general factors; and the next level had smaller subsets that could be further subdivided. Individual difference researchers familiar with taxonometric research in intelligence (e.g., Carroll, 1993) or personality (DeYoung, Quilty, & Peterson, 2007) would probably find Gati’s representation most appealing, but his model has not received nearly as much attention as Holland’s. Perhaps this is due to the influential paper by Tracey and Rounds (1993), who found that Holland’s representation was a more adequate fit for 104 correlation matrices than Gati’s model.

In sum, it has been nearly one hundred years since the first interest inventories were developed, but the field of vocational interest research has yet to converge on a set of factors to adequately describe the vocational interest domain. Although Holland’s RIASEC model is currently the most popular (its 3-digit system, for example, is used to classify and suggest occupations in the Department of Labor O⋆NET system) and widely supported (Tracey & Rounds, 1993), more studies and frameworks are continuously being proposed to address this important research issue. The lack of consensus may also indicate the fluid nature of vocational preferences. Many activities that historically fell into the mechanical interest domain, for example, are quickly disappearing due to robotization, while those relating to information technology have been rapidly evolving and expanding (see Day & Rounds [1997] for more discussion of the changing nature of work). Liao, Jin, Tay, Su, and Rounds (2015), for example, recently conducted exploratory structural equation modeling analyses using responses to a large set of contemporary work activities. They found a well-defined Information Systems broad factor that reflects interest in analyzing data and managing information including mathematics, programming, and computer tasks. Such a factor was not found fifty years ago and, perhaps, best illustrates why the structure of vocational interests may have to be continuously updated.

Emerging Importance of Basic Interests

The second type of dimensions measured and reported by interest inventories is typically referred to as basic interests. According to Hansen (1984) and Day and Rounds (1997), basic interests represent an intermediate level of aggregation of work activity preferences, lying between the specific occupations and general interest factors. Essentially, these can be seen as narrow interest factors, akin to personality facets or specific intelligence factors. As was noted by Day and Rounds (1997), most people actually describe their vocational interests using the language of basic interests such as “I was born to teach,” or “I like to fix things.”

At the basic interest level, work preference items are grouped into homogenous composites based on some shared property, such as occupational context, work setting, objects of interest, and work processes. There are several reasons why basic interests have emerged as an important component of many interest inventories. First, as was noted by Jackson (1977), early interest inventories reported scores for a very large number of occupations, which could be overwhelming for a typical career counseling client. Also, because many occupations were similar, much of the information provided was redundant. Basic interests organize occupations into clusters, so the volume of information presented to an individual and the level of redundancy is therefore reduced. One of the earliest examples of a basic interest structure, which was developed by Campbell Borgen, Eastes, Johansson, and Peterson (1968) for the SVIB (Campbell, 1971; Strong, 1943), was introduced specifically to provide interpretive information for the Strong Occupational Scale scores.

Second, in comparison to the general factor level, basic interests make finer distinctions between areas of vocational interests, thus allowing for a more precise theoretical interest structure (Rounds, 1995). As a consequence, basic interest models are more nimble than broad factor models in their ability to detect and accommodate changes in the labor market (e.g., emerging work activities and occupations).

Finally, from an applied perspective, basic interests offer incremental validity in predicting occupational group membership, because they “more effectively deal with the reality of a complex multivariate space” (Donnay & Borgen, 1996, p. 288). For example, Ralston, Borgen, Rottinghaus, and Donnay (2004) analyzed Strong Interest Inventory scores from 17,074 college students who majored in 24 different areas and found that the addition of basic interest scores to the general factor scores added significantly to prediction for 22 out of 24 study majors.

In addition to the Campbell et al.’s (1968) basic interest scales for the SVIB, a number of other basic interest models have been introduced; many have been based on factor or cluster analyses of occupational preferences. The most frequently cited primary studies are by Guilford et al. (1954) who identified 12 interest factors, Jackson (1977) who identified 28 primary work role factors, Kuder (1977) who identified 16 factors, Droege and Hawk (1977) who identified 11 factors, and Rounds and Dawis (1979) who identified 14 factors for men and 13 factors for women in the SVIB. Rounds and colleagues (Rounds, 1995; Day & Rounds, 1997; Liao, Armstrong, & Rounds, 2008) wrote a series of papers that attempted to integrate these various taxonomies. Ultimately, they settled on 31 basic interest dimensions and they identified 6–14 markers (activities) that could be used to measure each basic interest factor or dimension (see Liao et al. [2008] for more detail). In Table 4.1, we list those 31 basic interests, together with their brief descriptions and examples of typical measurement markers. Liao et al. (2008) grouped the 31 basic interests into 9 clusters (see the first column in the table), but any other general factor model can be used to aggregate these narrow dimensions. For example, the seven basic interest dimensions from the Physical Activity and Technical clusters would fall under the Realistic General Occupational Theme in the RIASEC taxonomy.

Occupational Interests

Occupational interests are located at the most specific end of the interest range and are often labeled using generic job names such as “nurse,” “reporter,” or “zookeeper.” The number of job-focused, occupational interests measured by most inventories can be very large. For example, the 1977 version of the Strong-Campbell Interest Inventory (Strong & Campbell, 1981) listed 116 occupational interests, Kuder (1977) referenced 217, and the 1994 Strong Interest Inventory contained 211 (Harmon et al., 1994). It is important to note that these occupational interests were actually the main reason why interest inventories had been developed. After completing an inventory, each test taker would receive a list of most/least suitable occupations and their brief descriptions, and vocational guidance counselors would use occupational interest information to help their clients in the job search process.

TABLE 4.1 Liao et al. (2007) Basic Interest Taxonomy

image image image image

To develop these job-focused, occupational interest scales, most test developers have relied on empirical keying of work activity items. This is done by determining the most preferred work activities of a criterion group (i.e., individuals working in a particular occupation or studying in a particular major). Subject matter expert opinions may also be used to identify the likely set of most preferred activities. One consequence of empirical keying is that the resulting sets of preferred work activities can be fairly heterogeneous and, thus, responses to the item sets do not exhibit high internal consistency reliability or adequate fit when single-factor models are applied. For this reason, referring to occupational interests as “dimensions” may be misleading. Many researchers instead prefer terms such as “occupational interest scales” or “job-specific scales” and use test-retest correlations as estimates of reliability.

Other Dimensions Measured by Interest Inventories

In addition to the three kinds of interest dimensions discussed above, interest inventories sometimes measure dimensions from other individual difference domains to aid educational or vocational guidance. For example, some inventories measure perceived efficacy in performing certain tasks. The Skills Confidence Inventory is a 60-item supplement to the SII that asks respondents to indicate their confidence in performing a particular task or school subject. The Self-Directed Search–Form R (Holland, 1985) contains two 6-item sets that ask respondents to estimate their proficiency level of a particular ability or skill.

Other inventories may include a handful of personality or values dimensions. The Jackson Vocational Interest Survey (JVIS; Jackson, 2000), for example, makes a distinction between work role scales, which are basic interest scales, and work style scales, which include personality dimensions (e.g., planfulness, interpersonal confidence, dominant leadership) and value dimensions (personal independence, job security, accountability). In addition, the SII includes items assessing various personal styles. In contrast to traditional vocational interest items, which assess more specific aspects of work activities, the Personal Style Scales assess broader preferences for styles of living and working. For example, the 2004 version of the SII includes a Team Orientation scale that differentiates individuals who prefer to work alone from those who enjoy working in teams. In addition, the 2004 SII also includes a Risk-Taking scale to explore preferences for taking chances versus living or working in a safe environment.

Perhaps the most eclectic interest inventory is the Motives, Values, Preference Inventory (MVPI; Hogan & Hogan, 2010). Its authors argue that motives, needs, values, interests, and personality are closely related concepts. The developers cited Super (1973), who saw values and interests as specific avenues to satisfy their needs, Dawis (1980), who suggested that interests, attitudes, needs, and values all represent “an affective orientation toward stimulus objects” (p. 77), and Holland (1997) who construed vocational interests as the expression of personality in work, school subjects, hobbies, recreational activities, and preferences. Consequently, the MVPI groups all these constructs under the term “motives.” For example, the MVPI Power Motive scale reflects “a desire for success, accomplishment, status, competition, and control.”

Summary

In summary, a substantial amount of work has been conducted on the hierarchy and structure of vocational interest dimensions. This research has shown that there are both broad dimensions of work activities (e.g., Prediger, 1982) and narrower dimensions (Liao et al., 2008). Although the exact composition of broad dimensions is still being debated (see, for example, a recent criticism of Prediger’s bipolar dimensions by Tay, Su, and Rounds [2011]), Holland’s (1997) six interest types, which lie somewhere between the broad two-dimensional structure proposed by Prediger and the 31 basic interest dimension structure proposed by Liao et al. (2008), are currently the most dominant and widely used dimensions in vocational interest assessment. The majority of existing interest measures assess Holland’s six types to some degree (i.e., either with scales that assess these types directly or by aggregating scales to get scores for each type). Despite the ubiquity of Holland’s structure in interest assessment, more research is needed to integrate Holland’s interest types with existing research proposing both broader and narrower dimensions. This research may consist of examining hierarchical models consistent with the types of factor structures identified in the literatures on cognitive ability (Carroll, 1993) and personality (DeYoung et al., 2007).

In addition to questions about the number and ordering of interest dimensions, the assessment of interests has also explored both general occupational themes (e.g., RIASEC) and narrower occupational scales. Each of these types of scales has both advantages and disadvantages. For example, although the general occupational themes are closely linked to Holland’s theory and have substantial empirical support for their validity, some have argued that these dimensions may be too broad to predict outcomes like job choice in the modern workplace where jobs and the nature of work are changing rapidly (Liao et al., 2008). On the other end of the spectrum, occupational scales may be too narrow, which raises questions about faking (i.e., it may be easier to identify a high-prestige occupation than the activities associated with it) and/or their generalizability across jobs (i.e., an engineer in one organization may perform different tasks than an engineer in another organization). We discuss these advantages and disadvantages in more detail below but, as a result of these issues, we agree with previous research (Liao et al., 2008) that basic interests may provide a more flexible approach to assessing vocational interests in the future.

General Approaches to Measuring Interests: Stimuli, Response Formats, and Psychometric Models for Scoring

To date, several dozen vocational interest inventories have been developed for use in educational and civilian work settings. Some notable examples include the Strong Interest Inventory (Harmon et al., 1994), the Kuder Occupational Interest Survey (KOIS; Kuder & Zytowski, 1991), the Self-Directed Search (SDS; Holland et al., 1994), the Campbell Interest and Skills Survey (CISS; Campbell, Hyne, & Nilsen, 1992), the O⋆NET Interest Profiler (Lewis & Rivkin, 1999), the Revised Unisex Edition of the American College Testing (ACT) Interest Inventory (UNIACT-R; ACT, 1995), the Oregon Vocational Interest Scales (ORVIS; Pozzebon, Visser, Ashton, Lee, & Goldberg, 2010), and the Jackson Vocational Interest Survey (Jackson, 1977, 2000). Below, we attempt to distill some of the general approaches to scaling vocational interest preferences.

All interest inventories contain large sets of work activity statements, similar to those shown in Column 4 of Table 4.1. Presenting activity statements is, undoubtedly, the most straightforward way to scale interests, as preferences for work activities lie in the heart of the interest construct definition (e.g., Hansen, 1984). Most inventories ask respondents to indicate the degree of their preference for each individual activity; some utilize dichotomous (like/dislike) or trichotomous (like, indifferent/?, dislike) response formats, while many others have settled on a popular five-option Likert response format. Regardless, the response options are typically assigned successive integers and classical test theory methods (a.k.a., total score methods) are used to score respondents. Because activity statements are phrased in a consistent direction, no reverse-scoring is needed prior to computing total scale scores.

A small number of interest inventories have used a forced-choice format (e.g., JVIS, KOIS). In this format, items are shown in blocks of two or three activity statements and respondents are asked to choose an activity they prefer the most/least; scoring is done by summing the observed ranks for statements belonging to the same interest dimension. The main advantage of forced-choice formats is that they appear to be less susceptible to response biases commonly associated with Likert-type formats (e.g., social desirability, halo, acquiescence). The disadvantage is that the resulting scores are often ipsative (see Hicks, 1970; Mead, 2004 for review), which creates difficulties for computing reliabilities, conducting factor analytic studies, and making inter-individual comparisons. Note, however, that many ipsativity issues historically associated with forced-choice measures can now be overcome using more recently developed IRT methods (see Stark, Chernyshenko, & Drasgow, 2005; Brown & Maydeu-Olivares, 2011; Hontangas et al., 2015), but we have yet to see research papers on this topic involving vocational interests.

In addition to presenting activity statements, many interest inventories (e.g., SII) include supplementary items such as checklists of specific occupations, confidence ratings for various abilities, skills, and competencies, or items describing work styles and work environment preferences. Responses to these kinds of items may be added to activity preference responses to compute the final primary, basic, and occupational interest scale scores, or used on their own for vocational counseling and guidance purposes. Although appealing, adding more item types carries some inherent psychometric risks. Respondents are now asked to navigate more complex sets of instructions, counselors must deal with more and possibly conflicting information, and scoring may need to accommodate mixed-item formats (e.g., reconcile combining dichotomous and polytomous items into a single scale score).

In terms of measurement models, the currently used inventories rely exclusively on classical test theory. To that extent, test developers screen and select items based on item-total correlations, estimate internal consistency reliabilities, and report various scale-level statistics for occupational, gender, or age groups. Factor analysis, which involves an underlying item response model, has been used primarily to investigate the factor structure rather than to improve scale construction and scoring (e.g., Nagy, Trautwein, & Lüdtke, 2010).

However, studies involving IRT models are emerging. Some of them have focused on improving existing measures by conducting in-depth analyses to identify poorly discriminating or biased items (e.g., Poitras, Guay, & Ratelle, 2012; Wetzel & Hell, 2014). Other studies have questioned some commonly held assumptions about interest scales. For example, Athanasou (2001) fit a Rasch IRT model to four-option Likert-type interest data and found that the distances between any two of the rating categories varied considerably across items. Based on those results, he cautioned against summing response category values in favor of more accurate, IRT-based scoring. In another paper, Wetzel and Hell (2014) applied multidimensional IRT methods to a German vocational interest inventory designed to assess RIASEC primary dimensions and found about 20% of items were multidimensional (i.e., had high loadings on two or more RIASEC dimensions). Their results have implications for how dimension scores may need to be estimated, because ignoring multidimensionality in data could result in biased trait estimates. Finally, a study by Tay, Drasgow, Rounds, and Williams (2009) questioned whether classical test theory and common factor models even make sense for interest items. Underlying all traditionally used models is the dominance response process assumption, which posits that the probability of endorsing an item increases monotonically as an individual’s standing on the latent trait increases. Tay et al. (2009) suggested that the dominance process is most appropriate for items dealing with maximal performance (e.g., cognitive ability), but, for items requiring introspection (making judgements about likes/dislikes), the ideal point response process assumption would provide a better description of the way individuals respond to interest items. The ideal point response process assumes that individuals are most likely to endorse items near their location on the trait continuum, and they will disagree with items that are too distal in either the positive or negative direction. Tay et al.’s (2009) empirical results supported this claim and indicated that an ideal point model provided better fit to data for three vocational interest measures.

What are the implications of the findings by Tay et al. and other recent psychometric papers? At the very least, they suggest that using more complex and better-fitting item response models will lead to improved scoring accuracy. Assuming equally spaced response options, unidimensional data, and a dominance response process can all result in biased trait estimates. Also, as was discussed in Chernyshenko, Stark, Drasgow, and Roberts (2007), the scale development processes may need to be reconsidered, because the rules for evaluating the quality of items are different when, for example, multidimensional or ideal point IRT models are used. In the personality domain, it has been shown that highly discriminating items in the middle of a trait continuum (a.k.a., neutral items) tend to have low item-total correlations; so they are usually discarded when classical test theory methods are used. However, such items may be particularly informative for respondents at high and low trait levels so there is value in retaining them in interest measures (see Chernyshenko et al., 2007). Finally, the use of more complex models could pave the way for computer adaptive tests (CAT) of vocational interests. CAT could reduce the number of items needed for adequate precision by half, which is particularly important when a large number of dimensions is assessed (e.g., basic interests).

Summary and Conclusions

Psychometricians unfamiliar with the vocational interest domain may likely see the assessment landscape as surprising. On the one hand, interest assessments have nearly a one-hundred-year history with dozens of instruments developed and administered to millions of respondents worldwide. The field has generated a tremendous number of papers on the structure of interests and the effect of vocational interest congruence on important work and life outcomes (e.g., Nye et al., 2012). Finally, there is a sizable infrastructure to support a variety of practitioners that help individuals interpret and use vocational interest scores to make educational and occupational choices.

On the other hand, there is still an apparent lack of consensus concerning the structure of vocational interests, especially at the level of basic interests. Also, despite the availability of large quantities of item response data, relatively little advanced psychometric work has been done to improve assessment processes in this domain. The field continues to rely almost exclusively on classical test theory methods, even though published, albeit limited, research has questioned the tenability of many existing assumptions. Finally, hardly any innovation can be seen in how interest measures are delivered. Most inventories are fixed length, nonadaptive, and use text-based stimuli such as activity descriptors or checklists of occupations.

For these reasons, we see a number of opportunities for advancing assessments in the vocational interest area. First, more item types should be explored. One example is the recent work of Wetzel, Hell, and Passler (2012) who used only verbs instead of more traditional activity descriptors to develop the Verb Interest Inventory. Another example is the JOIN inventory (Farmer, Watson, Alderton, Michael, & Hindelang, 2006), which used pictures to complement text activity descriptors (e.g., images of Navy personnel performing selected work activities). Phan and Rounds (2018) also suggested that capturing affective responses to work activities may lead to new types of congruence indices and better prediction of career outcomes.

A second possible way to advance interest assessment is to use modern psychometric models to develop more informative interest measures and leverage CAT technology to increase the precision of interest assessments or, alternatively, to reduce current testing times. Along these lines, it may be fruitful to consider multidimensional forced choice CAT methods (Stark et al., 2005) that may help to reduce response sets, which can inflate correlations among work activity preferences and distort relationships with criteria in educational and personnel testing contexts. It is no longer beyond our capabilities to combine these technologies to create efficient, video-based CATs that illustrate performance in occupational or educational settings and instruct respondents to “select all” or “choose between” scenarios they find most appealing. It is also worth considering whether game-based assessments can be a more effective and enjoyable method of gathering data, particularly among elementary and high schoolers. With informed consent, it may also be possible to use posts and likes on social media sites or news boards to match website users with information about college majors and careers given that this technology is already widely used for targeted advertising.

As a third opportunity for future interest assessment research, it would be worthwhile to explore what can be learned by using factor-mixture models that account for subpopulation heterogeneity and latent profile methods that may reveal unpredicted patterns of responses. Finally, and perhaps foremost, it is necessary to avidly explore the utility of vocational interest measures for predicting academic performance, work performance, and how interest congruence affects psychological well-being. Until recently, it was widely believed that vocational interests have only weak relationships with motivation and performance in employment settings, but as was demonstrated by Nye et al. (2012), focusing on interest congruence can dramatically increase the magnitude of relationships and may help to clarify and advance theory in this area. Also, as was shown by Ralston, Borgen, Rottinghaus, and Donnay (2004), focusing on narrow factors (i.e., basic interest dimensions) may also provide incremental validity over the currently dominant broad-factor measures. In sum, we see many new and exciting research opportunities in the vocational interest area and are eager to see how future methods of interest assessment can be used to improve outcomes for individuals and organizations alike.

References

  1. American College Testing Program. (1995). Technical manual: Revised unisex edition of the ACT interest inventory (UNIACT). Iowa City, IA: ACT.

  2. Athanasou, J. A. (2001).Analysis of responses to vocational interest items: A study of Australian high school students. Journal of Career Assessment, 9, 61–79.

  3. Bizot, E. B., & Goldman, S. H. (1993). Prediction of satisfactoriness and satisfaction: An 8-year follow-up. Journal of Vocational Behavior, 43, 19–29.

  4. Brown, A., & Maydeu-Olivares, A. (2011). Item response modeling of forced-choice questionnaires. Educational and Psychological Measurement, 71, 460–502.

  5. Campbell, D. P. (1971). Handbook for the Strong Vocational Interest Blank. Stanford, CA: Stanford University Press.

  6. Campbell, D. P., Borgen, F. H., Eastes, S. H., Johansson, C. B., & Peterson, R. A. (1968). A set of basic interest scales for the Strong Vocational Interest Blank for men. Journal of Applied Psychology Monograph, 52, 1–54.

  7. Campbell, D. P., Hyne, S. A., & Nilsen, D. L. (1992). Manual for the Campbell Interest and Skill Survey. Minneapolis, MN: National Computer Systems.

  8. Carnegie Institute of Technology: Bureau of Personnel Research. (1920). Carnegie Interest Inventory. Pittsburgh, PA: Author.

  9. Carroll, J. B. (1993). Human cognitive abilities: A survey of factor-analytic studies. New York: Cambridge University Press.

  10. Chernyshenko, O. S., Stark, S., & Drasgow, F. (2010). Individual differences, their measurement and validity. In S. Zedeck (Ed.), APA Handbook of industrial and organizational psychology (pp. 117–151). Washington, DC: American Psychological Association.

  11. Chernyshenko, O.S., Stark, S., Drasgow, F., & Roberts, B.W. (2007). Constructing personality scales under the assumptions of an ideal point response process: Toward increasing the flexibility of personality measures. Psychological Assessment, 19, 88–106.

  12. Clark, K. E. (1961). Vocational interests of nonprofessional men. Minneapolis: Univesity of Minnesota Press.

  13. Dawis, R. V. (1980). Measuring interests. In D. A. Payne (Ed.), New directions for testing and measurement: Recent developments in affective measurement (pp. 77–92). San Francisco: Jossey-Bass.

  14. Day, S. X., & Rounds, J. (1997). “A little more than kin, and less than kind”: Basic interests in vocational research and career counseling. The Career Development Quarterly, 45, 207–220.

  15. 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.

  16. Donnay, D. A. C., & Borgen, F. H. (1996). Validity, structure, and content of the 1994 Strong Interest Inventory. Journal of Counseling Psychology, 43, 275–291.

  17. Droege, R. C., & Hawk, J. (1977). Development of a U.S. Employment Services interest inventory. Journal of Employment Counseling, 14, 65–71.

  18. Farmer, W. L., Watson, S. E., Alderton, D. L., Michael, P. G., & Hindelang, R. (2006). Improving person-job congruence during the classification process: Item development and initial testing of a pictorial interest instrument (Technical Note No. 06–8). Millington, TN: Navy Personnel Research, Studies, and Technology Division, Bureau of Naval Personnel.

  19. Ferguson, L. W., Humphreys, L. G., & Strong, F. W. (1941). A factorial analysis of interests and values. Journal of Educational Psychology, 32, 197–204.

  20. Gati, I. (1979). A hierarchical model for the structure of interests. Journal of Vocational Behavior, 15, 90–106.

  21. Gati, I. (1991). The structure of vocational interests. Psychological Bulletin, 109, 309–324.

  22. Guilford, J. P., Christensen, P. R., Bond, N. A., & Sutton, M. A. (1954). A factor analysis study of human interests. Psychological Monographs, 68 (4), no. 375, 1–38.

  23. Hansen, J. C. (1984). The measurement of vocational interests: Issues and future directions. In S. D. Brown & R. W. Lent (Eds.), Handbook of counseling psychology (pp. 99–136). New York: Wiley.

  24. Harmon, L. W., Hansen, J. C., Borgen, F. H., & Hammer, A. L. (1994). Strong Interest Inventory: Applications and technical guide. Stanford, CA: Stanford University Press.

  25. Harrington T., & Long, J. (2013). The history of interest inventories and career assessments in career counseling. The Career Development Quarterly, 61, 83–92.

  26. Hicks, L. E. (1970). Some properties of ipsative, normative, and forced-choice normative measures. Psychological Bulletin, 74, 167–184.

  27. Hogan, R., & Hogan, J. (2010). Motives, values, preference inventory manual. Tulsa, OK: Hogan Assessment Systems.

  28. Holland, J. L. (1959). A theory of vocational choice. Journal of Counseling Psychology, 6, 35–45.

  29. Holland, J. L. (1985). The Self-Directed Search professional manual. Odessa, FL: Psychological Assessment Resources.

  30. Holland, J. L. (1997). Making vocational choices (3rd ed.). Odessa, FL: Psychological Assessment Resources.

  31. Holland, J. L., Fritzsche, B. A., & Powell, A. B. (1994). The Self-Directed Search (SDS): Technical manual. Odessa, FL: Psychological Assessment Resources.

  32. Hontangas, P. M, de la Torre, J., Ponsoda, V., Leenen, I., Morillo, D., & Abad, F. J. (2015). Comparing traditional and IRT scoring of forced-choice tests, Applied Psychological Measurement, 39, 598–612.

  33. Jackson, D. N. (1977). Manual for the Jackson Vocational Interest Survey. Port Huron, MI: Research Psychologist Press.

  34. Jackson, D. N. (2000). Jackson vocational interest survey manual (2nd ed.). Port Huron, MI: Research Psychologists Press.

  35. Kuder, F. G. (1977). Activity interests and occupational choice. Chicago: Science Research Associates.

  36. Kuder, F., & Zytowski, D. G. (1991). Kuder Occupational Interest Survey Form DD: General manual (3rd ed.). Monterey, CA: Macmillan/McGraw-Hill.

  37. 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.

  38. Liao, H.-Y., Jin, J., Tay, L., Su, R., & Rounds, J. (2015, May). Finding the big factors in interests: A factor analytic model of basic interests. In R. Su (Chair), New directions in interest assessment: Methodological issues, debates, and solutions. Symposium presented at Association of Psychological Science Annual Convention, New York, NY.

  39. Lewis, P., & Rivkin, D. (1999). Development of the O⋆NET Interest Profiler. Raleigh, NC: National Center for O⋆NET Development.

  40. McCrae, R. R., & Costa, P. T., Jr. (1987). Validation of the five-factor model of personality across instruments and observers. Journal of Personality & Social Psychology, 57, 17–40.

  41. Meade, A. W. (2004). Psychometric problems and issues involved with creating and using ipsative measures for selection. Journal of Occupational and Organizational Psychology, 77, 531–552.

  42. Meir, E. L, & Ben-Yehuda, A. (1976). Inventories based on Roe and Holland yield similar results. Journal of Vocational Behavior, 8, 269–274.

  43. Nagy, G., Trautwein, U., & Lüdtke, O. (2010). The structure of vocational interests in Germany: Different methodologies, different conclusions. Journal of Vocational Behavior 76, 153–169.

  44. 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.

  45. Phan, W. M. J., & Rounds, J. (2018). Examining the duality of Holland’s RIASEC types: Implications for measurement and congruence. Journal of Vocational Behavior, 106, 22–36.

  46. Poitras, S., Guay, F., & Ratelle, C.F. (2012). Using the Self-Directed Search in research: Selecting a representative pool of items to measure vocational interests. Journal of Career Development, 39, 186–207.

  47. Pozzebon, J. A., Visser, B. A., Ashton, M. C., Lee, K., & Goldberg, L. R. (2010). Psychometric characteristics of a public-domain self-report measure of vocational interests: The Oregon Vocational Interest Scales. Journal of Personality Assessment, 92, 168–174.

  48. Prediger, D. J. (1976). A world of work map for career exploration. Vocational Guidance Quarterly, 24, 198–208.

  49. Prediger, D. J. (1982). Dimensions underlying Holland’s hexagon: Missing link between interests and occupations? Journal of Vocational Behavior, 21, 259–287.

  50. Prediger, D. J., & Vansickle, T. R. (1992). Locating occupations on Holland’s hexagon: Beyond RIASEC. Journal of Vocational Behavior; 40, 111–128.

  51. Ralston, C. A., Borgen, F. H., Rottinghaus, P. J., & Donnay, D. A. C. (2004). Specificity in interest measurement: Basic Interest Scales and major field of study. Journal of Vocational Behavior, 62, 203–216.

  52. Roe, A. (1956). The psychology of occupations. New York: Wiley.

  53. Rounds, J. (1995). Vocational interests: Evaluating structural hypotheses. In D. J. Lubinski & R. V. Dawis (Eds.). Assessing individual differences in human behavior: New concepts, methods, and findings (pp. 177–232). Palo Alto, CA: Davies-Black.

  54. Rounds, J. B., Jr., & Dawis, R. V. (1979). Factor analysis of strong vocational interest blank items. Journal of Applied Psychology, 64, 132–143.

  55. Stark, S., Chernyshenko, O. S., & Drasgow, F. (2005). An IRT approach to constructing and scoring pairwise preference items involving stimuli on different dimensions: An application to the problem of faking in personality assessment. Applied Psychological Measurement, 29, 184–201.

  56. Strong, E. K. (1927). Vocational Interest Blank. Palo Alto, CA: Stanford University Press.

  57. Strong, E. K. (1943). Vocational interests of men and women. Palo Alto, CA: Stanford University Press.

  58. Strong, E. K., Jr., & Campbell, D. P. (1981). Strong-Campbell Interest Inventory. Stanford, CA: Stanford University Press.

  59. Super, D. E. (1973). The Work Value Inventory. In D. G. Zytowski (Ed.), Contemporary approaches to interest measurement. Minneapolis, MN: University of Minnesota Press.

  60. Tay, L., Su, R., & Rounds, J. (2011). People—Things and data—Ideas: Bipolar dimensions? Journal of Counseling Psychology, 58, 424–440.

  61. Tay, L., Drasgow, F., Rounds, J. & Williams, B. A. (2009). Fitting measurement models to vocational interest data: Are dominance models ideal? Journal of Applied Psychology, 94, 1287–1304.

  62. Thorndike, E. L. (1935). The psychology of wants, interests and attitudes. Oxford, England: Appleton-Century.

  63. Thurstone, L. L. (1931). A multiple factor study of vocational interests. Personnel Journal, 3, 198–205.

  64. Tracey, T. J., & Rounds, J. (1993). Evaluating Holland’s and Gati’s vocational-interest models: A structural met-analysis. Psychological Bulletin, 113, 229–246.

  65. Tracey, T. J. G., & Rounds, J. (1996). The spherical representation of vocational interests. Journal of Vocational Behavior, 48, 3–41.

  66. Van Iddekinge, C. H., Putka, D. J., & Campbell, J. P. (2011). Reconsidering vocational interests for personnel selection: The validity of an interest-based selection test in relation to job knowledge, job performance, and continuance intentions. Journal of Applied Psychology, 96(1), 13–33.

  67. 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.Washington, DC: American Psychological Association.

  68. Wetzel, E., & Hell, B. (2014). Multidimensional item response theory models in vocational interest measurement: An Illustration Using the AIST-R, Journal of Psychoeducational Assessment, 32, 342–355.

  69. Wetzel, E., Hell, B., & Passler, K. (2012). Comparison of different test construction strategies in the development of a gender fair interest inventory using verbs. Journal of Career Assessment, 20, 86–102.