Abstract

This study examined the reliability and validity of the Death Implicit Association Test (D-IAT) for suicide risk assessment. It was conducted from September to mid-December of 2020, targeting 179 adults aged over 19 years. D-IAT and suicide-related measures were administered. D-IAT was re-administered after 2 weeks and suicide risk was reassessed after 4 weeks. All procedures were conducted online. The split-half reliability of D-IAT was .73. The test-retest reliability over two weeks showed r value of .30. Convergent validity was partly supported by results showing that the D-IAT was positively correlated with depression (r=.21) and negatively correlated with reasons for living (r=-.15). There was no significant difference between D-IAT groups according to the existence of lifelong suicidal thoughts, attempts, or recent suicidal thoughts. Thus, known-group validity was not supported. Predictive validity was partially supported by results showing that D-IAT significantly predicted the presence or absence of desire for self-harm in suicide risk after one month. Moreover, D-IAT predicted wish for death at a marginally significant level. Current results suggest that psychometric properties of D-IAT are rather limited as a tool for suicide risk assessment. Further research on this is warranted.