The Spectrum of Response Bias in Trauma Reports: Overreporting, Underreporting, and Mixed Presentation

Abstract

AbstractTypically, research on response bias in symptom reports covers two extreme ends of the spectrum: overreporting and underreporting. Yet, little is known about symptom presentation that includes both types of response bias simultaneously (i.e., mixed presentation). We experimentally checked how overreporting, underreporting, and mixed reporting reflect in trauma symptom reports. Undergraduate students (N = 151) were randomly allocated to one of four conditions: control group (n = 40), overreporting (n = 37), underreporting (n = 36), or a mixed reporting (n = 38) group. The control group was asked to be honest, whereas the simulation groups received instructions they were expected to be able to relate to. Participants were administered a PTSD checklist, the Supernormality Scale-Revised, which screens for symptom denial, and the Inventory of Problems-29, which gauges overreporting. For all three measures, group differences were significant (F(3, 147) > 13.78, ps < . 001, ɳ2 > .22), with the overreporting group exhibiting the most severe symptom presentation and the underreporting group endorsing the fewest symptoms, while the mixed group almost perfectly imitated the results of the control group on all measures. Thus, instructing individuals to strategically exhibit both overreporting, to a functional degree, and underreporting results in a symptom presentation that is relatively similar to genuine performance. To optimize accuracy statistics, researchers might want to consider including such mixed group condition in future simulation studies.