Psychometrics of the personal questionnaire: a client-generated outcome measure

We present a range of evidence for the reliability and validity of data generated by the Personal Questionnaire (PQ), a client-generated individualized outcome measure, using five data sets from three countries. Overall pre-therapy mean internal consistency (alpha) across clients was .80; within-client alphas averaged .77; clients typically had one or two items that did not vary with the other items. Analyses of temporal structure indicated high levels of between client variance (58%), moderate pre-therapy test-retest correlations (r =.57), and high session-to-session lag-1 autocorrelations (.82). Scores on the PQ provided clear evidence of convergence with a range of outcome measures (within-client r = .41). Mean pre-post effects were large (d = 1.25). The results support a revised caseness cut-off of 3.25 and a reliable change interval of 1.67. We conclude that PQ data meet criteria for evidence-based, norm-referenced measurement of client psychological distress for supporting psychotherapy practice and research.