True and False Confessions: Comparing Self-Reported Likelihood of Confession with Actual Confession Behavior

Confession rates in the United States criminal justice system are high, and at least some of those who confess are innocent. Research on false confession frequency among justice-involved individuals either relies on historical self-reported confessions or considers the few cases in which DNA exoneration occurred, which are significant limitations in determining the frequency of or risk factors for false confession. Another line of research, undertaken with community samples, uses laboratory manipulations to examine the conditions under which individuals offer false confessions. Because laboratory false confession research has largely used two paradigms, it is difficult to identify the degree to which paradigm-specific factors (e.g., accusation of something the participant could have accidentally done; uncertainty regarding what the confederate will say happened) influence false confession rates. A valid self-report tool would allow for more accurate estimation of false confession frequency in justice-involved populations with whom laboratory studies are not practical or ethical. The Perceptions of Confession behavior during the Holding and Interrogation Process (P-CHIP) is a reliable, examiner-administered measure of self-reported likelihood of false confession if innocent and prediction of true confession behavior if guilty that has been administered to justice-involved populations, but the validity of the measure is unknown. This study introduced a novel laboratory false confession paradigm to examine the ecological validity of the P-CHIP. Participants were 91 undergraduates who believed they would be completing a study of logical reasoning and police perceptions and who were ultimately accused of cheating on a research task. False confession rates were lower and true confession rates higher than in previous research; almost all guilty participants signed a confession admitting to cheating, and 17.9% of innocent participants signed the confession. Likelihood of an innocent participant confessing was significantly associated with P-CHIP score. Suggestibility, decision-making ability, and age significantly predicted false confession behavior. Results suggest that the P-CHIP measures some aspect of false confession vulnerability or likelihood if the individual is accused of wrongdoing they did not commit. Additionally, this study raises questions regarding the potential role of paradigm choices in rates of confession. Future directions for false confession research are discussed.Ph.D., Psychology -- Drexel University, 201