Reliability Enhancement: Increasing the accuracy of self-report through mediation-based procedures

Eighteen subjects, self-referred for tension and anxiety, were randomly assigned to either a Reliability Enhancement Package (composed of cognitive consistency, consequence clarification, public commitment, and cueing manipulations)or an Attention-Control condition. Subjects in both groups received anxiety-oriented treatment in the form of progressive relaxation training. A pre- and posttreatment bogus biofeedback assessment was conducted during which time subjects self-recorded "time in alpha" as a supposed measure of their relaxation abilities. Experimental procedures (i.e., REP and AC)were administered prior to the posttreatment assessment session. Self-recorded alpha records were then compared to external ratings in order to assess the accuracy of self-report data. Results indicated significantly greater inaccuracy scores at posttreatment for Attention-Control as compared to Reliability Enhancement subjects (p < . 005). Moreover, errors committed were in the direction of overestimating time in alpha (p < . 02). These results were discussed with regard to mediational manipulations and expanding self-observational research strategies.