Abstract
Current guidelines for psychological assessment recommend the use of multiple validity measures in an evaluation protocol, particularly in forensic settings. As such, self-report instruments which detect distorted symptom reporting have become essential. We studied a pooled sample of 240 male inmates with the Self-Report Symptom Inventory (SRSI), the Structured Inventory of Malingered Symptomatology (SIMS), and the Symptom Validity Scale–Version 2 (EVS-2). Concurrent validity was analyzed via correlations between all three symptom validity tests (SVTs), revealing strong associations (rho ranging from .72 to .79), and ROC analyses yielded areas under the curve (AUC) values over .9 for every model. Base rates of SVT failure ranged from 7.9 to 13.3% using the most conservative cutoff scores, although true and false positive rates were not established. Education was shown to have a statistically significant effect on the mean results of the three SVTs (rho ranging from − 162 to − 283), associating lower educational levels with more bogus symptom endorsement. The influence of age and conviction status (pre-trial vs. post-trial) on the SIMS results was statistically significant as well (rho estimates of .171 and − 232). With data showing robust construct validity and excellent predictive accuracy, the instruments were shown to be adequate for use in the Portuguese prison context, but further research is required, in particular using forensic inpatients and clinical populations, and including systematic data on the presence of mental disorders.