Can a Variant of the Implicit Association Test Detect Nonsuicidal Self-Injury in a Clinical Population? A Registered Report.

Abstract

Background: Nonsuicidal self-injury (NSSI) is a severe and prevalent mental health problem. Measures to detect which individuals are at risk of engaging in NSSI would be valuable for clinical practice. However, we still lack strong predictors of future NSSI behaviour, with the most notable exception being prior NSSI behaviour. Yet, the measurement of prior NSSI behaviour with self-report measures can be difficult because individuals may be motivated to conceal this harmful behaviour. To overcome this problem, an implicit measure was developed that assesses automatic responding to statements about prior NSSI behaviour (i.e., the Past Nonsuicidal Self-Injury Implicit Association Test: P-NSSI-IAT). Previous studies tested the predictive utility of this measure in online studies with samples of at-risk participants and produced promising results. The current study examines the predictive utility of the P-NSSI-IAT for NSSI in clinical samples.

Method: Outpatients (N = 68) completed the P-NSSI-IAT and reported on past-year and past-month NSSI, as well as perceived likelihood of future NSSI.

Results: Patients with recent NSSI had significantly higher IAT scores than those without a history of NSSI. The P-NSSI-IAT predicted past-year NSSI but did not significantly predict past-month NSSI or perceived future likelihood of NSSI. It did not add predictive value beyond self-report risk factors in multivariate models.

Conclusion: These findings suggest that the P-NSSI-IAT may help detect prior NSSI in clinical contexts but offers limited added value for the prediction of NSSI beyond existing self-report measures.