Assessing Negative Response Bias with the Inventory of Problems-29 (IOP-29): a Quantitative Literature Review

This article reviews published, journal articles informing on the conditions of use, strengths, weaknesses, and optimal cut scores of the Inventory of Problems-29 (IOP-29; Viglione & Giromini, 2020). To provide more accurate information on the convergent and incremental validity, hit rates, and optimal cut scores of the IOP-29, in addition to reviewing all published IOP-29 studies, we also retrieved the datasets associated with each of those studies and performed some additional analyses. Taken together, the findings presented in this quantitative literature review indicate that (a) the IOP-29 correlates more strongly with other symptom validity tests (SVTs) than with other performance validity tests (PVTs), (b) the IOP-29 yields incremental validity when used together with other validity checks, (c) its classification accuracy compares favorably to that of other established tools, and (d) its suggested cut scores perform similarly well across various diagnoses and contexts. When considering the 3777 IOP-29 protocols included in the statistical analyses comparing credible (k= 16) versus noncredible (k = 17) presentations, the standard IOP-29 cut score of False Disorder probability Score >=.50 yielded a weighted mean sensitivity of .86 (weighted SD = .07; range: .63-.96) at a weighted mean specificity of .92 (weighted SD = .06; range: .79-1.00). The weighted mean Cohen's d was 3.02 (weighted SD = .98; range: 1.48-5.31), and the weighted mean AUC was .95 (weighted SD = .04; range: .83-1.00). These excellent statistics, however, could be inflated by the fact that almost all of the examined studies used a simulation research paradigm.