Are measures of self-efficacy reactive?

This study attempted to determine if the correspondence between self-efficacy and behavioral performance was due to reactivity cues provided by the self-efficacy measure. The extent to which the self-efficacy measure enhanced or detracted from the effects of a fear-reducing intervention (pretest sensitization) was also examined. Participants (age M = 21.54, SD = 3.11) included 52 undergraduate students (8 males, 44 females) who indicated a particular fear on the Fear Survey Schedule (FSS). Participants were randomly assigned to a self-efficacy group or a routine-treatment group. In the self-efficacy group, the measure of self-efficacy was administered at preintervention, between filmed modeling segments, postintervention, and follow-up; in contrast, the routine-treatment group completed the self-efficacy measure at fellow-up only. Outcome assessment of fear included a behavioral approach task and measures of self-reported fear (anticipatory and performance). In general, results indicated that group differences regarding the magnitude of correlations between self-efficacy and fear measures were nonsignificant. In addition, results did not yield evidence of pretest sensitization. The findings are interpreted as supporting Bandura's conceptualization of the reactivity issue, and the need for additional research is noted.