Validity and Reliability of Provided Constructs in Assessing Death Threat

Abstract

A self-administered form of the Threat Index (TI) was introduced and compared to the original interview form along dimensions of validity, reliability, internal consistency, and independence from social desirability response set. The assumption that a split between self and death is indicative of threat was empirically tested by means of an expanded format. In addition, the death-relevant situations used to generate constructs in the interview form were tested for generalizability. Thirty-two introductory psychology students completed the study which required two administrations of the instruments, with a four week intervel between the sessions. Results supported the theoretical and psychometric soundness of both forms. Advantages, disadvantages, and appropriate applications of each form were discussed.