More Intelligent Design: Testing Measures of Merit

Abstract

This Article articulates the theoretical, legal, and policy implications of new and improved theories of intelligence and recent research finding that conventional mass-marketed standardized tests, or 'factorist tests,' have less predictive power and larger racial differences in scores than newer multidimensional 'systems-based' tests. It raises a new question about the fairness of the role that traditional admissions tests like the SAT, GRE, LSAT, and MCAT currently play in selective higher education admissions - whether basing admissions on scores on such tests unfairly distorts the true admissions-related merit of individual applicants and racial groups. The core of this argument is not that selective universities rely on a flawed definition of merit or that traditional factorist tests are racially, economically, or culturally biased. Instead, this Article considers the ramifications of social science evidence suggesting that the admissions tests most commonly relied upon today are less successful in predicting applicants' future academic performance and have more racially skewed scores because they are designed according to a scientifically flawed theory of intelligence.

Keywords: intelligence, standardized testing, test scores, test deficiency, SAT, LSAT, selective admissions, Title VI, disparate impact, Fourteenth Amendment