Tilt increases at higher ability levels: Support for differentiation theories

Tilt refers to an ability pattern and is based on differences in two distinct abilities (e.g., math and verbal), yielding relative strength in one ability (e.g., math) and relative weakness in another (e.g., verbal). The current study is the first to examine linear and quadradic effects of ability level on diverse measures of tilt (e.g., math tilt, verbal tilt, tech tilt). Data were obtained from the National Longitudinal Survey of Youth (N = 1950), a representative sample of US students. Ability level was based on g (general intelligence) factor scores from the Armed Services Vocational Aptitude Battery (ASVAB). Tilt was based on math and verbal scores (math minus verbal) on college tests (SAT, ACT, PSAT), yielding math tilt (math>verbal) and verbal tilt (verbal>math). Tilt was also based on technical and academic (math or verbal) scores (tech minus academic) on the ASVAB, yielding tech tilt (tech>academic) and academic tilt (academic>tech). Linear effects of g on tilt were found for math tilt and verbal tilt but not tech tilt. Quadratic effects were not consistently observed for any specific type of tilt (i.e., math tilt, verbal tilt, tech tilt). The linear effects of g on math and verbal tilt suggest that exposure to academic subjects in school facilitates the acquisition of academic tilt at higher ability levels. The results support differentiation theories, which assume that tilt levels increase at higher ability levels due to ability specialization. The results do not support magnification theories, which assume that increases in tilt accelerate at higher ability levels, producing quadratic effects. Future research should examine the moderators and mediators of g-tilt relations (e.g., vocational interests and personality traits).