Linear and nonlinear associations between general intelligence and personality in Project TALENT

Area of Science:

• Psychometrics
• Personality Psychology
• Cognitive Science

Background:

• Previous research on personality and intelligence focused mainly on linear associations.
• Nonlinear relationships between general intelligence (g) and personality traits have been underexplored.
• A large-scale dataset from Project TALENT (PT) provides an opportunity to investigate these associations.

Purpose of the Study:

• To test for both linear and nonlinear associations between general intelligence (g) and 10 personality scales.
• To model latent quadratic effects directly, controlling for common factors in personality scales.
• To investigate the direction of effect from general intelligence (g) to personality.

Main Methods:

• Utilized quadratic and generalized additive models to analyze data from approximately 400,000 American high school students (Project TALENT).
• Modeled latent quadratic effects directly, controlling for the common factor in personality scales.
• Formulated 17 directional hypotheses for linear and quadratic associations based on existing literature.

Main Results:

• 53% of hypotheses were supported in male grades and 58% in female grades.
• Quadratic associations explained significant variance beyond linear effects for specific traits (e.g., Sociability, Maturity, Vigor, Leadership in males; Sociability, Maturity, Tidiness in females).
• Linear associations were predominant for other personality traits.

Conclusions:

• Nonlinear models provide incremental detail in understanding personality-intelligence associations.
• Current theories may need refinement to account for these nonlinear relationships.
• Research on intellectually gifted samples could further illuminate the personality-intelligence interface.