The Psychometric Properties of Four Systems for Scoring Divergent Thinking Tests

The traditional system for scoring divergent thinking tests has been criticized for its lack of predictive and discriminant validity. The present investigation was conducted to evaluate alternative scoring systems. Two divergent thinking tests (Uses and Line Meanings) were administered to 120 seventh- and eighth-grade children; the psychometric properties (i.e., reliability, and predictive and discriminant validity) of four scoring systems were evaluated and compared. Correlational analysis indicated (a) that the Uses test had notably higher validity coefficients than Line Meanings; (b) that the summation score (the sum of fluency, originality, and flexibility), the uncommon score (the number of ideas given by less than 5% of the sample), and the weighted-fluency score had the highest validity coefficients; (c) that ratio scores (e.g., flexibility divided by fluency) were generally unreliable and invalid; and (d) that all divergent thinking test scores were unrelated to IQ, but were related to achievement test scores. These findings have practical and theoretical implications for the testing of divergent thinking and creativity.