Summary
Attribution theory predicts that accounts of failure depend on who is queried. Presumably, the failing person tends to attribute the failure to situational factors beyond his or her control while observers tend to attribute failure to dispositional weaknesses. The present experiment was designed to measure such differences between soldiers being discharged from basic training (attrites) (n = 57), and soldiers succeeding in basic training (n = 149). Both groups were asked to rate socially acceptable counseling problems and socially unacceptable counseling problems on their frequency of occurrence in basic training. The prediction was that attrites would rate socially acceptable problems higher in frequency than less socially acceptable problems relative to successful trainees. The predicted interaction reached borderline significance. After trainees who were independently judged to have legitimate reasons for leaving the Army were excluded, the predicted interaction was significant. Results are discussed in terms of their implications for post hoc data collection efforts in front-end topic analyses.