Determining test fairness and differential validity of scores for the Torrance tests of creative thinking for kindergarten students



Journal Title

Journal ISSN

Volume Title


Texas Tech University


The purpose of this study was to determine test fairness and differential validity in creative thinking ability and achievement/language measures for kindergarten students. Data collection from 120 kindergarten students examined standard scores of the Torrance Tests of Creative Thinking Figural Form A (TTCT) with the standard scores of the Woodcock Johnson Tests of Achievement (WJ) and the Woodcock Language Proficiency Battery (WLPB). A culture fair creativity test is needed to identify creative thinking ability in young minority students. This research contributed to the investigation of the TTCT as a reliable and valid instrument beneficial in the identification of creative thinking ability in kindergarten students from diverse ethnic and socioeconomic groups. This study purposed that across ethnicity and socioeconomic status the creativity test would be a good predictor w hen the predictive scores were regressed on the criterion scores of the achievement/language tests.

Utilizing descriptive statistics, a multivariate analysis of variance, and Potthoff's procedure, this study intended to provide evidence that ethnicity and socioeconomic status (SES) differences with respect to creativity and achievement/ language were not significant. A multivariate analysis of variance and Potthoff's procedure were utilized in this study to analyze ethnic, SES, and gender differences. The MANOVA found small significant differences for socioeconomic status on fluency, originality, elaboration and resistance to premature closure. Potthoff's procedure was used to analyze slope and intercept differences. A small significant difference was found for slope bias or test bias w hen the creative variable abstractness of titles was regressed on the language variable.

Overall, there appears to be no relationship between creativity measures and achievement/language measures. Based on the statistical test results, differential validity and test fairness were evident; therefore, these tests worked equally well for both populations in the study. Creative thinking ability seemed to be more apparent in minority females accompanied by low socioeconomic status, but not significant. Results of this study support Torrance's belief that the Torrance Tests of Creative Thinking is fair and unbiased. The TTCT seemed to appropriately identify creative thinking ability without ethnic or SES preference.



Difference (Psychology), Creative thinking -- Testing, Torrance tests of creative thinking, Mexican American children -- Education -- Texas, Creative ability -- Testing, Creative thinking in children