Utility of predicting group membership and the role of spatial visualization in becoming an engineer, physical scientist, or artist.

Abstract

This article has two themes: First, we explicate how the prediction of group membership can augment test validation designs restricted to prediction of individual differences in criterion performance. Second, we illustrate the utility of this methodology by documenting the importance of spatial visualization for becoming an engineer, physical scientist, or artist. This involved various longitudinal analyses on a sample of 400,000 high school students tracked after 11 years following their high school graduation. The predictive validities of Spatial-Math and Verbal-Math ability composites were established by successfully differentiating a variety of educational and occupational groups. One implication of our findings is that physical science and engineering disciplines appear to be losing many talented persons by restricting assessment to conventional mathematical and verbal abilities, such as those of the Scholastic Aptitude Test (SAT) and the Graduate Record Examination (GRE).

Women's underrepresentation in science: Sociocultural and biological considerations.

“…The presence of fewer women at the right tail in mathematical and spatial ability renders fewer available for some math-intensive graduate programs on account of their GRE-Q scores. Having stated this, we note that data linking math and spatial abilities to STEM success are indirect, although strongly suggestive of playing a causal role in women's underrepresentation (Humphreys et al, 1993;Shea et al, 2001, for spatial support). However, if each sex's representation were solely a function of math ability, there would be roughly double the number of women in math-intensive careers as now exists, because assuming a 2:1 male-to-female ratio at the top 1% of math ability, women would be expected to comprise 33% of the professorships in math-intensive fields.…”

Spatial ability: A neglected dimension in talent searches for intellectually precocious youth.

“…It has been shown that, when intellectually talented students are appreciably more verbally than mathematically talented, their verbal gifts tend to propel them toward areas outside the math-science pipeline (Humphreys et al, 1993;Lubinski, Webb, et al, 2001;Shea et al, 2001). The discriminant functions in Phase 3 appear to capture these differential inclinations emanating from differences in specific ability profiles.…”

“…However, although the direction of the comparisons was in the predicted direction for each of the measures, there were no statistically significant differences for the boys between the profile similarities of each The differences in nonintellectual attributes between high-space and low-space intellectually able adolescents observed in Phase 1 were expected to portend individual differences in temporally remote developmental choices and outcomes. Participants' group memberships for five longitudinally assessed criterion variables (favorite and least favorite high school course, leisure activities, undergraduate majors, and anticipated occupations) were categorized to examine the validity of spatial ability using the group membership approach (Humphreys et al, 1993;Rulon, Tiedmen, Tatsuoka, & Langmuie, 1967;Tatsuoka, 1988). C. P. Snow's (1965) "two cultures," further informed by Ackerman's (1996;Ackerman & Heggestad, 1997) science-math and intellectualcultural "trait clusters," provides the rationale for isolating constellations of personal attributes as indicative of humanists (intellectually talented but with low spatial ability) and scientists (intellectually talented and with high spatial ability).…”

“…All three specific abilities manifested incremental validity relative to the other two in the prediction of educationaloccupational outcomes 20 years later . Given that Shea et al's findings mirrored longitudinal findings from older students (Gohm, Humphreys, & Yao, 1998;Humphreys & Lubinski, 1996;Humphreys, Lubinski, & Yao, 1993), the authors suggested that program developers begin to experiment with augmenting talent search selection criteria (beyond quantitative and verbal reasoning measures) with spatial ability: Educators were encouraged to think about experimenting with curriculum modifications for intellectually precocious learners especially able at nonverbal ideation, and theoreticians were encouraged to consider incorporating measures of spatial ability in longitudinal examinations of intellectually precocious youth to forestall misspecified or underdetermined causal models (cf. Lubinski, 2000).…”

“…If identified, would adolescents with high spatial ability be good prospects for STEM? In Phase 1, the interest and value profiles of adolescents with relatively high or low spatial ability were compared, by sex, to the profiles of math-science graduate students 1 In the 9th-grade cohort of Project Talent (Flanagan et al, 1962), for example, we analyzed the quantitative and verbal reasoning ability composites that Humphreys et al (1993) designed to mirror the SAT in conjunction with their spatial visualization composite. Within this sample (N Ͼ 100,000), only 30% of the top 1% on spatial ability scored within the top 1% on either mathematical or verbal reasoning.…”