“…One factor that Weersing and Weisz (2002) did not explicitly pursue as a potential methodological issue in conducting benchmarking studies was the reactivity and specificity of outcome measures, which have repeatedly been shown to significantly affect the effect size estimates of treatment outcomes (Lambert & Bergin, 1994;Lambert, Hatch, Kingston, & Edwards, 1986;Robinson, Berman, & Neimeyer, 1990;Shadish et al, 1993Shadish et al, , 1997Shadish et al, , 2000Shapiro & Shapiro, 1982;Smith, Glass, & Miller, 1980). Reactivity is generally concerned with the sensitivity of the measure produced by the rater of the outcome-notably, an observer (either the treating clinician or an independent rater) or the client.…”