A note on expectancy ratios, base rates, and the SVIB.

Validity studies of the Strong Vocational Interest Blank have produced consistent results with male samples. This study was designed to compare validities for females and males on the Strong-Campbell Interest Inventory (SCII). A sample of 232 women and 386 men took the SCII and a satisfaction measure 3V2 years after initial testing. Using the McArthur method, excellent predictive validity (good hit) was evidenced for 42.5% of females and 59.3% of males in the direct hit category. Concurrent validities were 58.0% and 64.0%, respectively. A multivariate analysis of variance revealed differences among good hit, poor hit, and clean miss groups on satisfaction, perceived congruence, and Holland's theoretical signs. Limitations with respect to the comparability of male and female validity data and the selection of a follow-up criterion for women are discussed.

Section: Discussionmentioning

confidence: 92%

Occupational preference and the validity of the Strong-Campbell Interest Inventory for college men and women.

Spokane¹

1979

“…When the differ-ence between the 67 male scales and the 57 female occupational scales is disregarded, 11/79 is 14%. But, as Brown (1961) pointed out, differential base rates have an important influence on results. When the difference between 67 male and 57 female scales is taken into account, the percentage becomes 12%, which is comparable to the percentage for the male scales.…”

Section: Resultsmentioning

confidence: 99%

Concurrent validity of other-sex and same-sex twin Strong-Campbell Interest Inventory occupational scales.

Dolliver¹,

Worthington²

1981

Strong-Campbell Interest Inventory (SCII) results were studied for participants whose "own" (most relevant) SCII occupational scale was a twin scale (with a female and a male scale for the same occupation). Participants included 80 male college graduates, 96 male college students, and 135 female college students. The concurrent validity of their own other-sex SCII twin occupational scale was compared with their own same-sex scale. Many differences were obtained among occupational groups, age groups, and gender groups. Overall, the other-sex own SCII twin scale was about as accurate a predictor as the same-sex scale. When the other-sex and the same-sex own SCII twin occupational scales both had high scores, the rate of accurate prediction was higher than when only one of the own twin scales had a high score.The Strong-Campbell Interest Inventory (SCII; D. Campbell, 1974) contains 37 occupational scales that are "twins;" that is, one scale is normed on females in an occupation and the other is normed on males in the same occupation. Those twin occupational scales have been of particular interest to researchers because they represent an intermediate step toward the potential merger of SCII occupational scales. However, they are troublesome to interpret because there is so little information about their validity.Data on the average scores of SCII twin occupational scales have been reported for four general groups of males and females (D.

“…Thus, depending on the gender of the respondent, and whether his/her scale was located among the men's or women's scales, there may have been many or few scales competing for rank. This is important in light of Brown's (1961) discussion of base rates.…”

Section: Selection Of Predictor Scaksmentioning

confidence: 96%

“…Berdie (1960) and Super and Crites (1962) review much of the same material and draw similar conclusions. Brown (1961), Dolliver (1969), and Kunce, Dolliver and Irwin (1972) have been critical of Strong's method of expressing the predictive validity of the SVIB. Both conclude that because of misunderstandings of Strong's qualifying phrases, which are not usually quoted in interpretations, it is better not to use such a statement with clients.…”

mentioning

confidence: 99%

Predictive validity of the Kuder Occupational Interest Survey: A 12- to 19-year follow-up.

Zytowski¹

1976

More than 1,000 men and women were located 12 to 19 years after taking an early form of the Kuder Occupational Interest Survey either in high school or college. Satisfactory data were available for 882 subjects. Fiftyone percent were employed at the time of the follow-up in an occupation that would have been suggested to them had their inventory been interpreted to them. This compares favorably with other validity findings for occupationally scaled interest inventories and represents about 90% of the possible validity as derived from concurrent validity figures. College major scales from a high school level administration correctly predicted 55%. Subjects in occupations consistent with their early interest profiles did not report greater job satisfaction or success but did show greater continuance in their occupational career. Better prediction is achieved when the occupation is one named on the profile and when the person has gone to or graduated from college or entered a high-level or scientifictechnical occupation.