New approach to the prediction of medical specialization: Student-based Strong Vocational Interest Blank scales.

Studied longitudinal data from 1583 undergraduates in an attempt to devise means of predicting eventual specialization for medical students. Specialty membership was determined in a follow-up study 11 yr. after the SVIB was administered. SVIB scales for surgeons, obstetricians, pediatricians, and psychiatrists were developed, using the items which differentiated the specialist groups as freshmen. Senior-based scales were also tested, but predicted no better than those using freshman responses. The resulting student-based scales were reliable on retest, but on measures of distributional overlap, they did not perform as well as standard SVIB occupational scales. However, their predictive validity was superior to that of previously developed SVIB scales for medical specialists. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Prediction of medical school attendance from college freshman SVIBs.

This study compared the SVIB profiles of 2 groups of college freshmen, 1 group which eventually applied to medical school and another which did not, but all of whom scored an A on the Physican scale. Significant differences were found between the 2 groups on 8 occupational scales, the largest of which were on the Artist and Architecture scales which were highly negatively related to subsequent application to medical school. The results give further evidence of a lack of a common interest factor in Group I of the SVIB. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Vocational interests of college freshmen and their social origins.

The relationship between interests as measured on the SVIB and socioeconomic status of college students was explored. 9 groups of entering college freshmen were selected on the basis of father's occupation and educational level of both parents. Differences among distributions of the 9 groups on each of 48 SVIB scales were tested for significance using the analysis of variance test. Conclusions are: measured vocational interests of college students were not independent of social origin, college students of lesser cultural background tended to identify with occupations requiring quantitative and technical training, extent of overlap between social groups on SVIB scales was high. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Validity studies of the Strong vocational interest inventories.

The concurrent validity of the Strong-Campbell Interest Inventory (SCII) and the predictive validity of the SVIB were studied in an 18-yr follow-up of a university counseling center group of male college graduates. Results were compared with a previous 12-yr follow-up of the same Ss using C. McArthur's method (1954). The SCII had concurrent validity greater than the SVIB (T399) even after correction was made for a higher percentage of high standard scores on occupational scales. Predictive validity for 18 yrs was not quite as good as for 12 yrs. Although group data were similar for the predictive validity studies, classifications of individuals between studies varied substantially. A 6-yr predictive validity study for the SVIB (T399) had validity comparable to predictive studies using an earlier edition of the SVIB. Job satisfaction and satisfaction with type of work were statistically different, but resulted in no practical difference when used as a correlate with standard occupational scale score. (21 ref) (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Locus of control, consistency of vocational interest patterns, and academic achievement.

Administered the SVIB, the W. James I-E Scale, and the American College Testing Program Examination (ACT) to 356 male university freshmen. Ss were classified as internals or externals on locus of control and as consistents or inconsistents on vocational patterns based on J. Holland's classification schema for scales on the SVIB. ACT scores were used to control for ability. Internals had a significantly higher GPA than externals; consistents had a higher GPA than inconsistents, but no significant interaction effects between vocational interest patterns and locus of control were found. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Temporal stability of personality by questionnaire.

The lack of a summary table or normative reliabilities for self-report inventories prompted the present compendium of test–retest reliabilities of several questionnaires that encompass a range of ages and instruments and differ in construction method and number of items/scale. The tests studied were the SVIB, EPPS, High School Personality Questionnaire, 16 PF, California Psychological Inventory, and MMPI. Samples varied in size, sex, and occupation. Findings indicate that tests of general personality had a similar pattern of decline in reliability over time and that homogeneous scales provided greater stability/item. When measured by occupational interest items, similar constructs had greater stability. In general, group profiles remained the same even over long periods, despite considerable reordering of individuals on any trait over time. (30 ref) (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Stability measures of Strong Vocational Interest Blank profiles.

For a sample of 242 freshmen, test and retest profiles on the SVIB were compared on grade changes, and letter grade and group pattern shifts, using Powers' D-score method, rank correlation, and ratings of the extent of interest changes by counselors (validity criterion). The stability measures were all significantly intercorrelated with the validity criterion (.55 to .68). (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

A dimensional approach to the interests measured by the SVIB.

Attempted to identify the 1st-order dimensions of interest measured by the SVIB and the more inclusive dimensions that account for correlations among the 1st-order factor scores. A further aim was to construct a set of content scales potentially useful in typological analysis and other research. Factor analyses of item correlation for 2 samples of 488 men-in-general indicated the presence of 14 dimensions of interest. An analysis of the correlations among the factor scores revealed 4 higher level factors: people-related, mechanical and symbol manipulation, personal expression and the arts, and outdoor activities. The 14 content scales developed (based on 198 SVIB items) were used to derive tentative interest profiles for 8 occupational groups. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

A personality inventory employing occupational titles.

The hypothesis that "a useful personality inventory might be constructed from occupational or interest test content" was explored further with the Holland Vocational Preference Inventory. 10 scales survived cluster analysis, using 300 college freshmen, and to these were added masculinity-femininity, status, infrequency, and acquiescence scales. Estimated retest reliability after 4 months had a median of .75. Differentiation was obtained between matched control and psychiatric samples. Profiles of university freshmen on the basis of college choice were also differentiated. 15 references. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Probability and utility assumptions underlying use of the Strong Vocational Interest Blank.

Decision implications of E. K. Strong's (1955) assumption of equal base rates for his men-in-general and individual occupational groups are explored, using the Strong Vocational Interest Blank (SVIB) Physician's scale as an example. It is demonstrated that when realistic base rates are assumed, total decision error rate using the SVIB-even when optimal cutting scores are employed-is at best about equal to that resulting from prediction using base rates alone. Use of the SVIB as recommended by Strong implicitly assumes that false negative errors are much more serious than are false positives. When "reasonable" utility or gain matrices are assumed, use of the SVIB can reduce total error rate despite the limitations imposed by low occupational base rates; this effect is greater when statistically optimal cutting scores, rather than those recommended by Strong, are used. (15 ref) (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Ten-year follow-up of the Tyler Vocational Card Sort and the Strong Vocational Interest Blank.

The Tyler Vocational Card Sort (TVCS) and SVIB for men were completed by 67 university students in 1965. The 1975 occupations held by 47 of these Ss were identified. The TVCS was more accurate than the SVIB, by a very small amount, in predicting occupation held 10 yrs after original testing; both methods achieved about 50% accuracy. Of the 47 Ss, 23 took the Strong-Campbell Interest Inventory (SCII) 10 yrs after the original testing. The reliability of the SVIB (1965) scores and the SCII (1975) scores was greater by a very small amount than that of the SVIB (1965) and the TVCS (1975) scores and that of the TVCS (1965) and TVCS (1975) scores. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

A note on the measured interests of Canadian psychologists.

In 1969-70 we conducted a small project which aimed to compare the measured interests of psychologists in five countries (Australia, Canada, Great Britain, New Zealand, and South Africa). The well-known Strong Vocational Interest Blank (SVIB) was mailed to random samples of 75 psychologists in each country. In this article, we wish to summarize the findings for readers of this Journal and to refer them to lengthier reports given in the references section. The results suggest this: The commonality of the measured interests of Western psychologists is so great that we may infer that the multiple occupational interest criteria used in the United States to aid students in making occupational and educational decisions may be appropriate for use in other Western countries. But the nature of the criteria used in this empirical method of measuring interests may be so culture-bound that a truly cautious approach should be used when extending the method and its supporting modes of interpretation to other countries. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Internal-external control of reinforcement and the Strong Vocational Interest Blank: A comparison study.

Studied the relationship between the Strong Vocational Interest Blank (SVIB), Form T399, and Rotter's Internal-External Control Scale (I-E scale). A previous study by D. G. Zytowski (see record 1967-07353-001) suggested a relationship between locus of control and occupational interest. To verify this, he correlated the SVIB with the I-E scale and concluded that such a relation did exist. Because the SVIB has been revised since Zytowski's study, a comparable study using the newer SVIB form was completed. Using a college freshman sample of 736, Pearson correlation coefficients were computed between the I-E scale and the SVIB; several correlations were significant at the .10 level; however, the largest correlation was -.16. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Strong Vocational Interest Blank in-general samples.

Throughout the history of the Strong Vocational Interest Blank (SVIB), the aggregation of "in-general" samples to represent an "average" has been of concern because an adequate in-general sample is necessary for the development of valid occupational scales for the SVIB. 6 in-general samples have been generated over the years to fit different developments of the SVIB. In the present study 20 experimental homogeneous scales were used to measure the similarities and differences among the 6 in-general samples. Generally, all samples were strikingly similar, with the greatest differences appearing between male and female in-general samples. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

The SVIB for women and demographic variables in the prediction of occupational tenure.

8 demographic variables and selected scales of the women's SVIB were incorporated in 2 types of prediction equations for 198 occupational therapists and 255 physical therapists. A double cross-validation design was used to develop and test 4 multiple regression and 4 reciprocal averages equations. 5 of the demographic variables correlated significantly with tenure, but none of the SVIB scales proved to be stable predictors. The method of reciprocal averages prediction yielded equations which proved more stable across samples and differed less shrinkage than those produced by the multiple regression technique. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

A verification scale for the Strong Vocational Interest Blank, Men's Form.

A new scoring procedure was constructed for the SVIB. Its purpose is to show the pattern of rare item-responses that differ from the chance pattern indicated by the shaded areas of the profile and from the scores typical of people in the various occupations. Ss were 192 freshmen at the University of Missouri. The cross-validation sample had 908 Ss. "At the present stage of development it is appropriate to recommend the V scale be included among the scales currently scored on the SVIB." From Psyc Abstracts 36:05:5LB18F. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Factor analysis of Strong Vocational Interest Blank items.

Attempted to identify independent interest dimensions that were equivalent across different S samples, as a first step toward mapping the vocational interest domain. The responses to 347 female form items and 357 male form items from the SVIB were obtained, using the appropriate form for the following 5 single-sex samples: women in general, men in general, female occupational, male occupational, and male rehabilitation client (Ns?=?1,000; 1,000; 2,500; 3,600, and 1,874, respectively). The SVIB items were intercorrelated separately for each sample and the correlation matrices were factored by a principal axes technique, using the highest off-diagonal correlations as communality estimates with rotation to a varimax criterion. Factors between same-sex samples were compared using Tucker's coefficient of congruence. The analyses resulted in 11–23 factors, depending on the sample, of which 9 and 8 were judged to be equivalent across the 2 female and 3 male samples, respectively. (38 ref) (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Effects of vocationally relevant information on the expressed and measured interests of freshman males.

46 undergraduate males from private and public school backgrounds who took the SVIB were classified as congruent or incongruent using a discrepancy method and administered a treatment consisting either of the Self-Directed Search (SDS) and relevant vocational information (treatment) or an irrelevant task (control). All Ss were retested 1 wk later using the SVIB, and examined for level of posttest congruence. Results suggest that private boarding school Ss were less congruent than private day or public school graduates. A secondary analysis revealed that Ss initially classified as low congruent significantly and nonartifactually increased in congruence from pretest to posttest. It was further found that increases in congruence could be attributed to changes in measured interests rather than to changes in expressed choice. Counseling implications are discussed. (26 ref) (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Faking "chance" on the SVIB.

Are scores that fall within the so-called "chance" areas of certain occupational scales of the Strong Vocational Interest Blank for Men (SVIB) "easily obtainable by chance"? To answer this question, Ss were selected whose scores, under standard testing conditions, were either higher than chance, lower than chance, or in the chance area itself. These Ss were then instructed to "fake" directionally (in the direction of the chance area) and to "fake chance." The results indicated that Ss who can fake directionally cannot fake chance, even when the chance range is in the same direction as the one they have faked. It was concluded that rather than ignore scores within the chance area, it may be better to ignore the chance areas themselves. (PsycINFO Database Record (c) 2010 APA, all rights reserved)