Scores of high school students responding to the same biology cognitive preference test using both normative and ipsative procedures were compared. All subtests, whether ipsative or normative, had high alpha Cronbach coefficients. A number of similarities were found in the mean scores obtained by the two procedures, as well as moderate positive correlations in each of the cognitive preference areas. It was concluded that when ipsative scores are preferred in terms of the construct validity of cognitive preference tests, the danger of distorted relationships is not severe. This conclusion has significant implications for the findings of a number of previous studies. (Author)