Objective: An empirical comparison of treatment efficacy estimates as based on psychometric measures of intra-individual change (Reliable Change methods).
Method: All seven different methods of assessing Reliable Change that have been advocated in the past two decades are compared empirically in a large in-patient sample (n = 107). Estimates of treatment efficacy by each of these seven Reliable Change methods are computed, using pre-/post-score changes on the Brief Psychiatric Rating Scale (BPRS).
Results: It is demonstrated that Reliable Change methods may yield very different estimates of treatment efficacy. The Reliable Change method with the fewest statistical assumptions is one of the least sensitive Reliable Change methods.
Conclusion: Disagreement on the proper definition of Reliable Change is not merely of theoretical importance, but also has major practical implications.