Linear T score norms for the clinical analysis questionnaire

Abstract

The Clinical Analysis Questionnaire (CAQ) was designed to measure both normal and pathological traits in order to facilitate diagnosis and treatment planning simultaneously. One problem that appears to limit the CAQ's usefulness is the use of 10-point normalized standard score conversion tables. First, the use of a 10-point scale appears to be too restrictive for scales designed to differentiate among various clinical syndromes. For many scales, a range of two or three raw score standard deviations is represented by a single standard score. Second, the assumption of an underlying normal distribution, which is implicitly made when using normalized transformations, appears untenable, particularly with respect to the pathological scales. For these reasons, revised norm tables for the CAQ were developed using linear conversions of raw scores to T scores. Results of an analysis comparing the tables indicated that for scales information provided by either approach. However, for the pathological scales the normalized conversion tables produced score distributions with means and standard deviations that were generally lower than those obtained from the linear T score tables.