Construct Validity and Psychometric Properties of the Cornell Critical Thinking Test (Level Z): A Contrasted Groups Analysis
Total scores and section scores (both corrected and uncorrected for guessing) on the Cornell Critical Thinking Test—Level Z were analyzed for evidence of construct validity. The test performance of three ability groupings of college students and a “no-booklet” (guessers) group (Total N = 527) was examined. Statistically significant differences were found among the corrected total score means for all four groups. Differences among the means of the three ability groups were statistically significant on one corrected and one uncorrected section of the test. In addition, the formula which corrects for guessing substantially improved the estimate of internal consistency reliability for the low-ability group only. Support for the heterogeneity of the thinking skills measured by Form Z was mixed and inconclusive. Implications for the use or modification of the test are discussed.