The contribution of self-efficacy in assessing interests using the self-directed search.
Betz (2007) stated that "an integration of personality with Holland interests and confidence would be a useful direction of research and theory" (p. 418). Our understanding, however, is that the SDS is designed to incorporate measurement of perceived abilities and skills (competencies) into the overall summary scores of the respective RIASEC domains. Thus, we raise the question, Has this integration of interests and confidence already been accomplished ipso facto into the existing structure of the SDS? Holland (1997), author of the SDS and RIASEC theorist, believed that interest types are actually personality types and that self-efficacy is embedded into one's overall personality structure. For example, Holland asserted that "it is not necessary to administer a self-efficacy measure to identify a person's beliefs about his or her vocational self-efficacy. The low points in an interest profile indicate the areas where a person lacks the self-confidence to perform well" (p. 209).
Thus, the SDS may not require an independent external measure of self-efficacy to incorporate this construct into the assessment of interests. Two of its original scales, Self-Estimates and Competencies, may serve as embedded measures of self-efficacy within each of the six RIASEC domains. Although the constructs and scales of interest (i.e., interest, self-efficacy, ability self-estimates, and competence) in this study have been examined together with previous research, the methods used in this study involved published instruments commonly used by practitioners. This is a unique aspect of this study.
Betz (2007) reviewed research relevant to the assessment of self-efficacy and interests. One such study (Feehan & Johnston, 1999) examined SDS summary scores (Holland, Fritzsche, & Powell, 1994) in relation to self-efficacy as measured by the Task-Specific Occupational Self-Efficacy Scale (Osipow, Temple, & Rooney, 1993). The results were that the measurement of interests by the SDS predicted career self-efficacy. Hansen and Bubany (2008) used measures of self-efficacy and ability self-estimates other than the published instruments used in our study to examine ability judgments of college students. Among other things, their findings supported the proposition that ability self-estimates involve a normative orientation, but self-efficacy measures do not. Feehan and Johnston (1999) specifically suggested examining the SDS in relation to the Skills Confidence Inventory (SCI; Betz, Borgen, & Harmon, 1996) as a measure of self-efficacy. Nevertheless, our review of the literature indicates that there was no follow-up to this proposed path of inquiry.
A major effort to clarify the relationship between self-efficacy and interests was undertaken in a meta-analysis of 60 empirical studies (Rottinghaus, Larson, & Borgen, 2003). This analysis included the Strong Interest Inventory (SII; Harmon, Hansen, Borgen, & Hammer, 1994) and the Campbell Interest and Skills Survey (CISS; Campbell, Hyne, & Nilsen, 1992) but not the SDS, which is the instrument of choice in the present study. The results of Rottinghaus et al.'s (2003) study demonstrated that self-efficacy and interests were viewed as independent constructs that correlated moderately, sharing about one third of their common variance. Rottinghaus et al. found that the self-efficacy and interest relationships between the RIASEC domains were stronger with the CISS than with the SII, possibly because response scale and scoring procedures in the CISS reflect external normative judgments. This CISS measurement scheme is analogous to the one used in the SDS. The magnitude of the correspondence between self-efficacy and interest scores across the respective RIASEC types ranged between r = .42 for Enterprising and r = .67 for Investigative.
In a study of ability self-estimates and occupational interests, the SDS's ability self-estimates scales and occupational interest on the SII were examined (Brown, Lent, & Gore, 2000). The researchers used only the Self-Estimates section of the SDS to measure ability self-estimates, and they modified the instructions of the Occupations section of the SDS to create a measure of occupational self-efficacy beliefs and perceived career options. Occupational interests were measured with 30 occupations taken from the Occupations section of the SII rather than the six General Occupational Theme (GOT) scores. With these modifications to the instruments, occupational self-efficacy beliefs were found to be conceptually related to as well as empirically distinguishable from occupational interests and ability self-estimates. Brown et al. (2000) concluded that ability self-estimates and self-efficacy might function as complementary dimensions in the assessment of interests. Parenthetically, in contrast to the SDS, the SII makes available a separate instrument, the SCI, to measure beliefs in ability to accomplish tasks in the respective RIASEC interest domains.
Although self-confidence and ability self-estimates in the respective interest domains do not represent objectively measured abilities, individuals' perceptions of how effective they can be in a given interest area are positively correlated with interests and are, therefore, a useful predictor of career choice (Betz, Borgen, & Harmon, 2006; Donnay & Borgen, 1999; Gottfredson, 2002; Hansen & Bubany, 2008; Rottinghaus et al., 2003; Tracey & Hopkins, 2001). In fact, Betz et al. (1996) argued that confidence in one's ability to perform is a necessary antecedent to actual career choice. Prediger (2002) noted that the use of ability self-estimates could be an effective way to facilitate identification of plausible career options. This suggestion has led to a view that the joint use of interest and confidence scores in career assessment and counseling would be useful (Betz & Rottinghaus, 2006). From a cognitive information processing perspective (Peterson, Sampson, & Reardon, 1991; Sampson, Reardon, Peterson, & Lenz, 2004), we believe that the consideration of interests and competencies is extremely important when guiding clients through the formulation and evaluation of viable occupational alternatives in the synthesis phase of the career decision-making cycle.
Since the development of the SCI and the Rottinghaus et al. (2003) meta-analysis, researchers have continued to turn their attention to the utility of the SCI and its relation to career interests (e.g., Bonitz, Armstrong, & Larson, 2010; Larson, Wu, Bailey, Gasser, et al., 2010). Larson, Wu, Bailey, Borgen, and Gasser (2010) found that accurate prediction of college students' choice of major was enhanced when considering student SCI and SII results, rather than just considering interests or self-efficacy alone. In a study of the SCI's concurrent validity, Robinson and Betz (2004) found college students' SCI scores to be highly correlated with college majors in the same RIASEC areas, especially for the Enterprising and Investigative areas. Despite this attention to the SCI and its relation to interest, no researchers have attended to the recommendation of Feehan and Johnston (1999) to study the SCI in relation to the SDS.
On the basis of the suggestions of Feehan and Johnston (1999) to study the relationship between the SDS and the SCI and in an effort to inform practitioners and researchers about the extent to which the SDS provides information about interests, perceived abilities and skills, as well as self-efficacy, the following research question was posed: What are the respective contributions to variation of SDS summary scores by (a) SDS ability self-estimates scores, (b) SDS competencies scores, and (c) SCI skills confidence (i.e., self-efficacy) scores? Given Holland's (1997) assertion, we predicted that the inclusion of an external measure of self-efficacy would capture little incremental variation once the variation attributed to SDS self-estimates and SDS competencies scores have been partitioned. A second question concerns the apparent commonalities among ability self-estimates, competencies estimates, and self-efficacy. That is, to what extent are these measures interrelated such that all three predictors may be considered as facets of a single higher order psychological construct (Hansen & Bubany, 2008)?
The 238 participants were male (25%) and female (75%) students between the ages of 18 and 47 years (M= 20.13) from a large northwestern public university who were enrolled in a psychology course. The top five majors represented among the participants were psychology, undecided, biological sciences, economics, and human development. The sample was particularly diverse in that 37% of the participants were European American, 36% were Asian/Pacific Islander, 11% were Hispanic/Latino, 5% were African American, 5% were biracial, 5% were other, and 1% were American Indian. First-letter SDS RIASEC codes for participants were as follows: Realistic, 5%; Investigative, 12%; Artistic, 7%; Social, 47%; Enterprising, 17%; and Conventional, 12%.
The data were obtained in five sessions in a large lecture hall. The research summary was read to inform participants of a study of career decision making, and each received a folder containing informed consent documentation, a demographic form, and the research instruments in alternating order. Participants were asked to complete the forms and answer all questions on the instruments in the order they appeared within the folder. When the students were finished, they each received a debriefing statement referring them to the university career or counseling center for further counseling if they wished as well as documentation of their research participation.
SDS summary scores. The SDS Form R (Holland, Powell, & Fritzsche, 1994) includes 228 items with five scales (Activities, Competencies, Occupations, and two Self-Estimates). Once scored, the SDS uses a six-part RIASEC typology to classify the structure of an individual's career interests. Internal consistency coefficients of the adult norming subgroup for each of the SDS summary scales are greater than .91. Test-retest reliability (4- to 12-week intervals) of the SDS was conducted using a sample of 73 individuals (28 male, 45 female) ranging in age from 14 to 48 years (median age of 20). Test-retest correlations for each of the summary scales were as follows: Realistic, r = .78; Investigative, r = .87; Artistic, r = .89; Social, r = .89; Enterprising, r = .82; and Conventional, r = .76 (Holland, Fritzsche, & Powell, 1994).
SDS competencies. In the SDS Competencies section, the respondents are directed to respond "YES" to activities in which they can do well or competently and "NO" for those activities in which they have never performed or performed poorly regarding 11 activities within each of the six RIASEC categories. Thus, the range for each scale is from 0 to 11. Examples of items taken from the Realistic scale are "I can make a scale drawing" and "I can use many carpentry tools" Holland, Fritzsche, and Powell (1994) cited two studies indicating that competencies scores have low, but statistically significant, correlations with scores on the Armed Services Vocational Aptitude Battery (ASVAB; U.S. Department of Defense, 1994) and nonacademic accomplishments. Internal consistency estimates (Kuder-Richardson reliability index) ranged from .72 to .84 for college men and women (Holland, Fritzsche, & Powell, 1994). Thus, this assessment may be considered as ipsative in nature in which competencies are compared internally with respect to one another.
SDS self-estimates. In the SDS Self-Estimates section, the respondents rate themselves "as you really think you are when compared with other persons your own age" on the six generic abilities associated with Holland's (1997) RIASEC model. Respondents are also asked to "avoid rating yourself the same in each ability." There are two sets of ability ratings for the respective RIASEC dimensions. Each rating is on a 7-point scale, with 1 being low-rated ability, 4 being average, and 7 being high in ability. Thus, when both ratings are combined, self-estimates scores range from 2 to 14 on the respective RIASEC scales. Examples of the two items from the Social area are "Teaching Ability" and "Understanding of Others." According to the authors (Holland, Fritzsche, & Powell, 1994), self-estimate scores are also correlated with ASVAB scores and have low to moderate predictive validity as indicated in several studies. However, reliability estimates for the SDS Self-Estimates section were not reported. This assessment may be regarded as normative in nature in that individuals are asked to rate themselves with respect to peers.
SCI. The SCI (Betz et al., 1996) is a 60-item (10 items per scale) assessment of self-efficacy for each of the RIASEC dimensions and is available as an option in administering the SII (Harmon et al., 1994). Respondents rate their degree of confidence (i.e., self-efficacy) on a 5-point scale (1 = very little, 2 = little, 3 = moderate, 4 = high, and 5 = very high) in their ability to complete tasks or successfully complete school subjects related to one of the six Holland (1997) types. To obtain a General Confidence Theme score for each domain, one sums the responses to the 10 items for each theme and computes an average, with final scores ranging from 1 to 5. Higher scores indicate more confidence in one's ability in that area. Examples of activities and subjects are "Build a doll house, Industrial Arts (Realistic)" and "Sell a product to a customer, Public Speaking (Enterprising)." This instrument is reported to be useful in career counseling and assessment in the exploration of instances of low skill confidence but high interest, or low interest and high skill confidence in a particular area (Harmon et al., 1996). The technical considerations for this instrument support its usage by reporting concurrent and construct validity, as well as Cronbach's alpha coefficients ranging from .84 to .88 on the respective RIASEC scales and 3-week test-retest stability coefficients ranging from .83 to .87 (Betz et al., 1996).
The relationships among the SDS ability self-estimates, SDS competencies scores, and SCI self-efficacy scores with respect to the SDS summary scores were analyzed. All analyses were conducted using Mplus 5.21 (Muthen & Muthen, 2007) and SPSS 16.0. Analyses conducted included hierarchical multiple regression and structural equation modeling (SEM), specifically confirmatory factor analysis (CFA). The SEM framework allows for significance testing between competing models,
rather than having to rely on only descriptive comparisons, as would be the case with exploratory factor analysis.
Correlations between SCI self-efficacy scores and corresponding SDS summary scores, SDS self-estimates, and SDS competencies scores were all positive and showed strong, significant relationships ranging from r = .46 to r = .78 (see Table I). According to Cohen (1988), all correlations between SDS and SCI scores within the same RIASEC domain fell in the large correlation range. The only exception was the correlation (r = .46) between the SDS competencies and the SCI self-efficacy scores for the Conventional type, which was medium according to Cohen's standards. Thus, the external measures of self-efficacy were highly correlated with the corresponding SDS summary scores.
Six two-step hierarchical regression analyses were used to measure the relative contribution of SDS self-estimates, SDS competencies, and SCI self-efficacy to the variation in SDS summary scores (see Table 2). SDS self-estimates and competencies (internal measures) were entered in Step I of the regression model. The external measure, SCI self-efficacy, was entered in the second step. The results of the regression analyses revealed that SDS self-estimates together with SDS competencies followed by SCI self-efficacy captured significant portions of the variance in SDS summary scores, with the only exception being the prediction of the SDS Enterprising summary score. For the SDS Enterprising summary score, SDS self-estimates and SDS competencies accounted for significant variation in the SDS summary score (adjusted [R.sup.2] = .78), but SCI self-efficacy failed to capture additional significant incremental variation. The adjusted [R.sup.2] for Step 1 of the six models ranged from .81 for the Realistic type to .55 for the Conventional type. The incremental variation accounted for by the SCI in the six respective analyses ranged between 0.0% and 2.5%. With the change in adjusted [R.sup.2]s of 2.50 and lower, the practical effect of the addition of an external measure of self-efficacy seems to add little additional useful information about these six personality traits beyond the information already provided by the internal SDS self-estimates and SDS competencies scores.
In a more rigorous empirical test of whether an external measure of self-efficacy belongs to the same higher order construct as the SDS ability scales, a CFA was conducted on SDS self-estimates, SDS competencies, and SCI self-efficacy as indicators for latent variables representing each of the RIASEC domains using Mplus 5.21 (Muthen & Muthen, 2007). Data were first screened for multivariate outliers using Mahalanobis distance, and none were observed. The model was estimated using the Satorra-Bentler chi-square (Satorra & Bentler, 1994) because of issues with normality. Holland's (1997) concept of consistency in RIASEC theory provides for the strength of relationships between domains (e.g., the Enterprising and Social domains should have a stronger relationship than should the relationship between the Enterprising and Investigative domains).
Using this theoretical framework, we correlated the error terms of the indicators with each of their two adjacent domains. Because the ultimate goal of this analysis was to test if the three factor loadings were significantly different from each other and if the scale for the SCI is different from the two SDS scales, all scores were standardized so that the regression weights could be constrained to be equal. Fit of the model was determined based on the criteria by Schumacker and Lomax (2004), which indicate that for the comparative fit index (CFI) and Tucker-Lewis index (TLI) < .90 is a poor fit, .90 to .95 is an adequate fit, and > .95 is a good fit. Schumacker and Lomax indicated that for the root mean square error of approximation (RMSEA), < .05 is a close fit and < .08 is a reasonable fit. Although all factor loadings were statistically significant at p < .001 and 15 out of 18 were greater than .70 (with the lowest at .51), the full model had a poor fit, [chi square](102) = 428.98,p < .001, CFI = .87, TLI = .81 (Hu & Bentler, 1999), and RMSEA = .12 (MacCallum, Browne, & Sugawara, 1996). There was no theoretical justification for the model modifications suggested by examination of the residual matrix and modifications indices.
On the basis of the work of Smart, Feldman, and Ethington (2000), the Realistic and Conventional domains were removed from the model. Smart et al. used Holland's (1997) RIASEC theoretical framework to examine patterns of stability and change among 4,408 college students attending 360 different postsecondary institutions. With this large and diverse sample, they chose to drop the Realistic and Conventional domains from their analysis based on the conclusion that Realistic and Conventional categories had very few students or majors. They further noted that these two disciplines/majors tend not to be represented frequently among students in 4-year institutions. The resulting IASE partial model in the current study, [chi square](39) = 133.00,p < .001, had a reasonably good fit based on the CFI = .94 and TLI = .90 (Hu & Bentler, 1999) but had a poorer fit based on the RMSEA = .10 (MacCallum et al., 1996). The fit statistics were considered adequate to continue with the test to examine if the factor loadings for SDS self-estimates, SDS competencies, and SCI self-efficacy were significantly different from each other in these remaining four domains. The factor loadings are presented in Table 3.
Nested models were tested in which the three factor loadings were constrained to be equal within each domain. Both the traditional change in chi-square and the more recent change in CFI < .01 tests (Byrne, 2010; Cheung & Rensvold, 2002) were used to determine whether the loadings were different from each other. As can be seen in Table 3, the factor loadings for Investigative, Artistic, and Social were not significantly different using the chi-square difference test, and the factor loadings for all four domains were not different using the change in CFI test. Further analysis of the Enterprising domain using the chi-square test revealed that the loadings for SDS competencies and SCI self-efficacy were not significantly different, but the loadings for SDS self-estimates and SCI self-efficacy were significantly different (p < .01). Therefore, in three of the six domains of the SDS, the external measure of self-efficacy is regarded as belonging to the same order construct as the matched SDS RIASEC scale and, consequently, provides no new information regarding the measurement of interest in these domains.
This study addressed the proposition by Holland (1997) that the SDS summary scores take into account self-efficacy with the inclusion of the components of self-estimates of abilities and competencies within the respective summary scores. Furthermore, a literature review revealed no previous studies that examined SCI self-efficacy in relation to both interests and perceived skills as measured by the SDS. Therefore, through hierarchical regression and CFA in this study, we sought to ascertain the possible contribution of an external measure of self-efficacy to account for unique variation in the SDS summary scores.
As expected, the three predictors of SDS summary scores were highly intercorrelated in all six of the RIASEC domains. Moreover, the results of the CFA partially supported the hypothesized model. Investigative, Artistic, and Social self-efficacy did not improve the fit or explanation of interest in those domains. Self-efficacy did improve the fit for the Enterprising domain. Unfortunately, we were unable to explore Realistic and Conventional domains using CFA because of the lack of fit they produced with the overall model. Hierarchical regression analysis showed that self-efficacy did increase the variability explained in SDS summary scores, but the size of the increases, ranging from 0.0% to 2.5%, added minimal practical explanatory power.
In 1996, the SCI was introduced to complement the interests measured by GOT scores of the SIL The SCI measured the extent to which a client possesses confidence with respect to a given RIASEC domain and was viewed as a measure of self-efficacy (Betz et al., 1996). Shortly after, Holland was asked whether a companion measure of self-efficacy should also be developed to complement the SDS in the same way as the SCI complements the SII. His response, as stated earlier, was that it was unnecessary because the self-estimates of abilities and competencies already tapped self-efficacy (Holland, 1997). The findings of the present study provide strong support for Holland's (1997) position in the Investigative, Artistic, and Social domains. The results of our analyses demonstrate the high degree of correspondence of the SDS self-estimates and competencies with matching SCI scales. The CFA revealed that an external measure of self-efficacy may be considered as a measure of the same construct as the SDS Self-Estimates and Competencies scales. Therefore, we conclude that the SDS Self-Estimates and Competencies scales are indicators of self-efficacy for Investigative, Artistic, and Social types and that the development and incorporation of an external measure of self-efficacy in the assessment of interest using the SDS is not likely a worthwhile endeavor.
The findings of the regression analysis with regard to the Enterprising type are somewhat perplexing. Despite a zero-order correlation of r = .68 between the SDS Enterprising and SCI Enterprising scales (see Table 1), the magnitude of contribution of self-efficacy to the summary score of the Enterprising domain does not seem to be statistically significant (see Table 2), and yet it does differ from the three other domains retained in the CFA (see Table 3). What is it about the nature of the SDS Enterprising domain, the SCI Enterprising scale, or population characteristics that renders them relatively independent entities? The age of our sample and wording of our measures may provide a possible explanation. Perhaps the typical college student has opportunities to participate and gain skills in the Investigative, Artistic, and Social domains and fewer opportunities to gain Enterprising skills, while potentially maintaining confidence in their ability to accomplish Enterprising-related tasks. Therefore, the wording of the SDS (i.e., "blacken ... yes for those activities you can do well or competently") would elicit responses that tap their lack of applied skill, but the SCI would elicit responses more relevant to their confidence in spite of no opportunity to test that confidence. Perhaps there are population characteristics that might influence these measured relationships, such as the preponderance of women, lack of business majors, and high diversity within the population.
Despite the significant correlations among self-estimates, competencies, and self-efficacy of the Realistic and Conventional domains, they were removed from the model because of inadequate fit. It is beyond the scope of the present study to conclude exactly why the Realistic and Conventional domains could not be retained. Possible explanations may include differences in the nature of the questions for the Realistic and Conventional domains on the SDS or SCI, or, as cited earlier, idiosyncratic characteristics of the population sample. Realistic was the least represented domain in this sample, which may help to explain some of the findings with regard to this domain. However, Conventional was as well represented as some of the other retained domains. Perhaps, as we observed, Conventional is often a skill that is developed but not an interest area in college populations. Also, few college majors are classifiable as Conventional, giving little impetus for students to be socialized and rewarded for Conventional skills and self-efficacy. These issues may affect some of the measurement properties of the Conventional domain.
Given the focus of our present study on measures of self-rated abilities, competence, and skills confidence across the Holland (1997) types, and the limited information available on these traits found by Smart et al. (2000), we decided to analyze our data without including college students in the Realistic and Conventional areas in the model. This decision took into account Smart et al.'s extensive work using Holland's RIASEC theoretical framework to examine patterns of stability and change among 4,408 college students attending 360 different postsecondary institutions. Several noteworthy and relevant findings germane to our study were reported by Smart et al. First, students initially select environments that will reinforce and reward their stronger abilities and interests, while avoiding environments rewarding or reinforcing their weaker abilities. Second, academic environments differ in the degree to which they are successful in socializing students to their respective patterns of abilities and interests. Artistic and Investigative environments are most successful, whereas Social and Enterprising environments are less so. Thus, we believe that much can be learned about the SDS and self-efficacy through the use of a partial model that excluded the Realistic and Conventional domains.
The findings of this investigation raise important implications for the use of the SDS in career counseling and career decision making. First, client self-estimates of abilities and competencies can be viewed as valid indicators of efficacy expectations within at least three of the RIASEC domains and should be thoroughly reviewed by career counselors with clients as essential considerations when interpreting the meaning of summary scores of the SDS (Holland, Powell, & Fritzsche, 1994). Second, SDS ability self-estimates and competency ratings may best be considered as domain-specific, but not as occupation-specific. After an array of potential occupations has been derived, career counselors should facilitate the exploration of ability self-estimates and competencies for each occupation.
Finally, with regard to career decision making within the synthesis phase of the CASVE (communication, analysis, synthesis, valuing, and execution) cycle of the cognitive information processing paradigm (Peterson et al., 1991; Sampson et al., 2004), synthesis elaboration entails the formulation of possible occupations for consideration followed by synthesis crystallization, which involves the narrowing of options to a select three to five. The results of this study support the use of liking of activities for the synthesis elaboration process for generating plausible options, and the use of measures of efficacy expectations for the synthesis crystallization in which the number of options is reduced to a manageable few for further consideration in the valuing phase of the CASVE cycle.
The findings of this study should be viewed with some degree of caution regarding external validity. The sample was drawn from a sample of students enrolled in a psychology course who received extra credit for participating in this research. It was also 75% female, 37% European American, and 36% Asian/Pacific Islander with a mean age of 20 years. As might be expected, almost half of the participants were Social types in Holland's (1997) typological theory. As with many samples drawn from college populations, RIASEC types are not distributed equally across the college environment (Smart et al., 2000). It is possible that some statistical coefficients may have been under- or overestimated because this sample had a large female and ethnic minority composition. The Social type is common among women and may have larger values within this study than if the male-to-female ratio was equivalent.
In conclusion, this investigation has demonstrated that the assessment of interests using the SDS contains two measures of self-efficacy, the Self-Estimates and Competencies scales, which contribute to summary scores for the Investigative, Artistic, and Social domains. The present study should be replicated with other college populations with differing gender and ethnic composition to determine the nature of Realistic, Conventional, and Enterprising domains.
Finally, we believe that the assessment of interests in career counseling should include the exploration of efficacy expectations within a given domain in addition to the liking or preferences of activities when formulating viable career options. When the SII is used, external measures such as the SCI or the Expanded SCI (Betz et al., 2003) may well serve the purpose of measuring efficacy expectations (the SCI as a measure of efficacy expectations related to the RIASEC domains and the Expanded SCI as a measure of efficacy expectations regarding the 17 basic interest scale domains). Likewise, when using the SDS, we have confidence from our study that efficacy expectations regarding the Investigative, Artistic, and Social domains are measured internally through the use of the Self-Estimates and Competencies scales as proposed by Holland (1997). Nevertheless, regardless of whether a counselor uses the SDS, SII, or another interest measure, this article underscores the importance of a thorough exploration of efficacy expectations in the assessment of interests in career counseling. We hope that this article clarifies the choices a career counselor makes in selecting measures of interest to facilitate career exploration and career decision making.
Betz, N. E. (2007). Career self-efficacy: Exemplary recent research and emerging directions. Journal of Career Assessment, 15, 403-422. doi: 10.1177/1069072707305759
Betz, N. E., Borgen, E, & Harmon, L. (1996). Skills Confidence Inventory applications and technical guide. Palo Alto, CA: Consulting Psychologists Press.
Betz, N. E., Borgen, F. H., & Harmon, L. H. (2006). Vocational confidence and personality in the prediction of occupational group membership. Journal of Career Assessment, 14, 36-55. doi: 10. I 177/1069072705282434
Betz, N. E., Borgen, E H., Rottinghaus, P., Paulsen, A., Halper, C. R., & Harmon, L. W. (2003). The Expanded Skills Confidence Inventory: Measuring basic dimensions of vocational activity. Journal of Vocational Behavior, 62, 76-100. doi: 10.1016/S00018791 (02)00034-9
Betz, N. E., & Rottinghaus, P. J. (2006). Current research on parallel measures on interests and confidence for basic dimensions of vocational activity. Journal of Career Assessment, 14, 56-76. doi: 10.1177/1069072705281348
Bonitz, V., Armstrong, P., & Larson, L. (2010). RIASEC interest and confidence cutoff scores: Implications for career counseling. Journal of Vocational Behavior, 76, 265-276. doi: 10.1016/j. jvb.2009.12.002
Brown, S. D., Lent, R. W., & Gore, P. A. (2000). Serf-rated abilities and self-efficacy beliefs: Are they empirically distinct? Journal of Career Assessment, 8, 223-235. doi: 10.1177/106907270000800302
Byrne, B. M. (2010). Structural equation modeling in AMOS." Basic concepts, applications, and programming (2nd ed.). New York, NY: Routledge.
Campbell, D. P., Hyne, S. A., & Nilsen, D. (1992). Manual for the Campbell Interest and Skill Survey. Minneapolis, MN: National Computer System.
Cheung, G. W., & Rensvold, R. B. (2002). Evaluating goodness-of-fit indexes for testing measurement invariance. Structural-Equation Modeling: A Multidisciplinary Journal 9, 233-255.
Cohen, J. (1988). Statistical power analysis for the behavioral sciences (2nd ed.). Mahwah, N J: Erlbaum.
Dormay, D. A. C., & Borgen, E H. (1999). The incremental validity of vocational self-efficacy: An examination of interest, self-efficacy, and occupation. Journal of Counseling Psychology, 46, 432-447. doi: 10.1037/0022-0184.108.40.2062
Feehan, P. E, & Johnston, J. A. (1999). The Self-Directed Search and career self-efficacy. Journal of Career Assessment, 7, 145-159. doi: 10.1177/106907279900700204
Gottfredson, G. D. (2002). Interests, aspirations, self-estimates, and the Self-Directed Search. Journal of Career Assessment, 10, 200-208. doi: 10.1177/1069072702010002005
Hansen, J. C., & Bubany, S. T. (2008). Do self-efficacy and ability self-estimate scores reflect distinct facets of ability judgments? Measurement and Evaluation in Counseling and Development, 41, 66-88.
Harmon, L., Borgen, E, Berreth, J., King, J., Schauer, D., & Ward, C. (1996). The Skills Confidence Inventory: A measure of self-efficacy. Journal of Career Assessment, 4, 457-477. doi: 10.1177/106907279600400408
Harmon, L. W., Hansen, J. C., Borgen, E H., & Hammer, A. L. (1994). Strong Interest Inventory: Applications and technical guide. Stanford, CA: Stanford University Press.
Holland, J. L. (1997). Making vocational choices: A theory of vocational personalities and work environments (3rd ed.). Odessa, FL: Psychological Assessment Resources.
Holland, J. L., Fritzsche, B. A., & Powell, A. B. (1994). Self-Directed Search technical user's guide. Odessa, FL: Psychological Assessment Resources.
Holland, J. L., Powell, A. B., & Fritzsche, B. A. (1994). Self-Directed Search professional user's guide. Odessa, FL: Psychological Assessment Resources.
Hu, L.-T., & Bentler, P. M. (1999). Cutoff criteria for fit indices in covariance structure analysis: Conventional criteria versus new alternatives. Structural Equation Modeling, 6, 1-55.
Larson, L., Wu, T., Bailey, D., Borgen, E, & Gasser, C. (2010). Male and female college students' college majors: The contribution of basic vocational confidence and interests. Journal of Career Assessment, 18, 16-33. doi: 10.1177/1069072709340520
Larson, L., Wu, T., Bailey, D., Gasser, C., Bonitz, V., & Borgen, E (2010). The role of personality in the selection of a major: With and without vocational self-efficacy and interests. Journal of Vocational Behavior, 76, 211-222. doi: 10.1016/j.jvb.2009.10.007
MacCallum, R. C., Browne, M. W., & Sugawara, H. M. (1996). Power analysis and determination of sample size for covariance structure modeling. Psychological Methods, 1, 130-149. doi:10.1037/I082-989X. 1.2.130
Muthen, L. K., & Muthen, B. O. (2007). Mplus user's guide (5th ed.). Los Angeles, CA: Author.
Osipow, S. H., Temple, R. D., & Rooney, R. A. (1993). The short form of the Task-Specific Occupational Self-Efficacy Scale. Journal of Career Assessment, 1, 13-20. doi: 10.1177/106907279300100103
Peterson, G., Sampson, J., & Reardon, R. (1991). Career development and services: A cognitive approach. Pacific Grove, CA: Brooks/Cole.
Prediger, D. J. (1999). Integrating interests and abilities for career exploration: General considerations. In M. Savickas & A. Spokane (Eds.), Vocational interests: Meaning, measurement, and counseling use (pp. 295-325). Palo Alto, CA: Davies-Black.
Prediger, D. J. (2002). Abilities, interests, and values: Their assessment and their integration via the world of work map. Journal of Career Assessment, 10, 209-232.
Robinson, C., & Betz, N. (2004). Test-retest reliability and concurrent validity of the Expanded Skills Confidence Inventory. Journal of Career Assessment, 12, 407-422. doi: 10.1177/1069072704266671
Rottinghaus, P. J., Larson, L. M., & Borgen, F. H. (2003). The relation of self-efficacy and interests: A meta-analysis of 60 samples. Journal of Vocational Behavior, 62, 221-236. doi: 10.1016/S00018791(02)00039-8
Sampson, J. P., Jr., Reardon, R. C., Peterson, G. W., & Lenz, J. L. (2004). Career counseling and services: A cognitive information processing approach. Pacific Grove, CA: Wadsworth-Brooks/Cole.
Satorra, A., & Bentler, P. M. (1994). Corrections to test statistics and standard errors in covariance structure analysis. In A. von Eye & C. C. Clogg (Eds.), Latent variables analysis:Applications for developmental research (pp. 399-419). Thousand Oaks, CA: Sage.
Schumacker, R. E., & Lomax, R. G. (2004). A beginner's guide to structural equation modeling (2nd ed.). Mahwah, NJ: Erlbaum.
Smart, J. C., Feldman, K. A., & Ethington, C. A. (2000). Academic disciplines: Holland's theory and the study of college students and faculty. Nashville, TN: Vanderbilt University Press.
Tracey, T. J. G., & Hopkins, N. (2001). Correspondence of interests and abilities with occupational choice. Journal of Counseling Psychology, 48, 178-189. doi:10.1037/0022-0220.127.116.11
U.S. Department of Defense. (1994). Technical manual for the ASVAB Career Exploration Program. Chicago, IL: U.S. Military Enlistment Processing Command.
Emily Bullock-Yowell, Department of Psychology, and Richard S. Mohn, Department of Educational Studies and Research, University of Southern Mississippi; Gary W. Peterson and Robert C. Reardon, Career Center, Florida State University, Tallahassee; Laura K. Wright, Counseling and Psychological Services, Florida Gulf Coast University. Correspondence concerning this article should be addressed to Emily Bullock-Yowell, Department of Psychology, University of Southern Mississippi, 118 College Drive, #5025, Hattiesburg, MS 39406-5025 (e-mail: Emily.Bullock@usm.edu).
TABLE 1 Means, Standard Deviations, and Correlations for the Study's Variables Variable M SD SDS summary code 1. Realistic 16.89 9.89 2. Investigative 23.93 9.72 3. Artistic 21.67 10.02 4. Social 33.46 9.42 5. Enterprising 27.00 9.96 6. Conventional 23.48 10.84 SDS self-estimates 1. Realistic 11.71 4.44 2. Investigative 16.26 5.19 3. Artistic 12.19 5.22 4. Social 19.59 4.24 5. Enterprising 15.97 4.94 6. Conventional 17.03 4.87 SDS competencies 1. Realistic 4.22 3.19 2. Investigative 7.27 2.86 3. Artistic 4.69 2.67 4. Social 8.31 2.61 5. Enterprising 6.63 3.01 6. Conventional 7.17 2.76 SCI self-efficacy 1. Realistic 2.98 0.78 2. Investigative 3.10 0.78 3. Artistic 2.97 0.82 4. Social 3.76 0.67 5. Enterprising 3.15 0.72 6. Conventional 3.22 0.73 Skills Confidence Inventory (SCI) Variable 1 2 3 SDS summary code 1. Realistic .65# ** .22 ** .20 ** 2. Investigative .21 ** .78# ** .11 3. Artistic .32 ** .11 .78# ** 4. Social .04 -.04 .23 ** 5. Enterprising .26 ** .07 .14 * 6. Conventional .15 * .06 -.08 SDS self-estimates 1. Realistic .55# ** .23 ** .16 * 2. Investigative .15 * .71# ** -.03 3. Artistic .30 ** .11 .75# ** 4. Social .14 * .06 .34 ** 5. Enterprising .25 ** .04 .14 * 6. Conventional .14 * .20 -.02 SDS competencies 1. Realistic .61# ** .26 ** .26 ** 2. Investigative .28 ** .67# ** .18 ** 3. Artistic .30 ** .14 * .71# ** 4. Social .14 * .02 .29 ** 5. Enterprising .27 ** .18 ** .26 ** 6. Conventional .23 ** .12 .04 SCI self-efficacy 1. Realistic -- 2. Investigative .40 ** -- 3. Artistic .53 ** .27 ** -- 4. Social .20 ** .21 ** .31 ** 5. Enterprising .41 ** .32 ** .35 ** 6. Conventional .51 ** .43 ** .11 Skills Confidence Inventory (SCI) Variable 4 5 6 SDS summary code 1. Realistic -.02 .19 ** .33 ** 2. Investigative .14 * .13 * .28 ** 3. Artistic .24 ** .21 ** -.05 4. Social .69# ** .13 * -.12 5. Enterprising .16 * .68# ** .39 ** 6. Conventional -.04 * .06 .58# ** SDS self-estimates 1. Realistic -.05 .22 ** .33 ** 2. Investigative .02 .14 * .35 ** 3. Artistic .23 ** .16 * -.O6 4. Social .70# ** .32 ** .01 5. Enterprising .19 ** .61# ** .40 ** 6. Conventional .08 .11 .50# ** SDS competencies 1. Realistic .11 .26 ** .28 ** 2. Investigative .22 ** .21 ** .27 ** 3. Artistic .33 ** .33 ** -.01 4. Social .70# ** .32 ** -.04 5. Enterprising .34 ** .72# ** .24 ** 6. Conventional .13 * .07 .46# ** SCI self-efficacy 1. Realistic 2. Investigative 3. Artistic 4. Social -- 5. Enterprising .36 ** -- 6. Conventional .11 .44 ** -- Note. N = 238. Numbers that represent correlations between SCI and Self-Directed Search (SDS) scores in the same RIASEC (Realistic, Investigative, Artistic, Social, Enterprising, and Conventional) domain are in boldface. Note: Numbers that represent correlations between SCI and Self-Directed Search (SDS) scores in the same RIASEC (Realistic, Investigative, Artistic, Social, Enterprising, and Conventional) domain are in boldface is indicated with #. * p < .01. ** p < .001. TABLE 2 Hierarchical Regression Analyses for the Prediction of Self-Directed Search (SDS) Summary Scores From SDS Self-Estimates, SDS Competencies, and Skills Confidence Inventory (SCI) Self-Efficacy Adjusted Predictor B SE B [beta] [R.sup.2] [R.sup.2] Realistic SIDS self-estimates 1.58 .13 .46 SIDS competencies 1.41 .12 .46 .81 .81 SCI self-efficacy 1.54 .47 .12 .82 .81 Investigative SIDS self-estimates 1.17 .14 .35 SIDS competencies 1.48 .13 .44 .80 .80 SCI self-efficacy 3.00 .53 .24 .82 .82 Artistic SIDS self-estimates 1.24 .16 .36 SIDS competencies 1.61 .16 .43 .80 .80 SCI self-efficacy 2.54 .55 .21 .82 .82 Social SIDS self-estimates 1.53 .28 .32 SIDS competencies 1.53 .21 .42 .67 .67 SCI self-efficacy 2.45 .78 .18 .69 .68 Enterprising SIDS self-estimates 1.96 .15 .51 SIDS competencies 1.55 .15 .47 .78 .78 SCI self-efficacy 0.46 .64 .03 .78 .78 Conventional SIDS self-estimates 1.64 .22 .41 SIDS competencies 1.25 .21 .32 .56 .55 SCI self-efficacy 2.67 .75 .18 .58 .57 Predictor [DELTA][[R.sup.2] F Change Realistic SIDS self-estimates SIDS competencies .81 497.29 ** SCI self-efficacy .01 10.81 Investigative SIDS self-estimates SIDS competencies .80 461.22 ** SCI self-efficacy .03 32.30 ** Artistic SIDS self-estimates SIDS competencies .80 480.34 ** SCI self-efficacy .02 21.10 ** Social SIDS self-estimates SIDS competencies .67 243.61 ** SCI self-efficacy .01 0.01 * Enterprising SIDS self-estimates SIDS competencies .78 425.68 ** SCI self-efficacy .00 0.50 Conventional SIDS self-estimates SIDS competencies .56 148.04 ** SCI self-efficacy .02 12.70 ** Note. N= 238. Step 1 included entering code-specific SIDS self-estimates and competencies scores. Step 2 included entering code-specific SCI self-efficacy scores. Information presented in Table 2 is from Step 2 of each hierarchical regression. * p < .01. ** p < .001. TABLE 3 Factor Loadings for Free and Constrained Models and Changes in Chi-Square and Confidence of Fit Index (CFI) Factor Loading Item 1 2 3 4 Investigative .84 ** .76 ** .86 ** .82 ** Artistic .87 ** .83 ** .86 ** .86 ** Social .87 ** .84 ** .82 ** .84 ** Enterprising .68 ** .80 ** .91 ** .81 ** Constrained Versus Free Item [DELTA][chi square] [DELTA]CFI Investigative 1.87 0.00 Artistic 1.26 0.00 Social 0.32 0.00 Enterprising 7.41 * -0.02 Note. The constrained model set all factor loadings within the latent variable to be equal. The free model allowed all factor loadings to be freely estimated. Chi-square difference tests were conducted using the Satorra-Bentler (1994) correction. Factor 1 = free Self-Directed Search (SDS) self-estimates; Factor 2 = free SDS competencies; Factor 4 = free Skills Confidence Inventory self-efficacy; Factor 4 = constrained, all equal. * p < .01. ** p < .001.
|Printer friendly Cite/link Email Feedback|
|Author:||Bullock-Yowell, Emily; Peterson, Gary W.; Wright, Laura K.; Reardon, Robert C.; Mohn, Richard S.|
|Publication:||Journal of Counseling and Development|
|Date:||Sep 22, 2011|
|Previous Article:||An emergent theory of altruism and self-interest.|
|Next Article:||The authorship determination process in student-faculty collaborative research.|