Date of Award

11-2024

Document Type

Doctoral Research Project

Degree Name

Doctor of Psychology (PsyD)

Department

Psychology

First Advisor

Radhika Krishnamurthy, Psy.D., ABAP

Second Advisor

Scott A. Gustafson, Ph.D., ABPP

Third Advisor

Theodore G. Petersen, Ph.D.

Fourth Advisor

Lisa A. Steelman, Ph.D.

Abstract

Suicide is a significant concern in society and is one of the leading causes of death in the United States and around the world. Suicidality is particularly prevalent in outpatient psychological treatment centers, with some studies estimating a tenfold increase in prevalence in this setting compared to the population at large. Several measures have been developed to assess suicide risk potential based on symptom frequency and combinations of risk factors, and broadband personality measures have also incorporated suicidality scales into their tests. The current study was focused on assessing the effectiveness of the Suicidal/Death Ideation (SUI) scale on the third edition of the Minnesota Multiphasic Personality Inventory (MMPI-3) in a sample of N = 80 (n = 48 women, n = 32 men) mental health outpatients who had undergone MMPI-3 assessment at the start of receiving psychotherapy services. Preliminary analysis of intercorrelations between SUI and the other substantive MMPI-3 scales revealed a broad range of 23 MMPI-3 scales correlated with at least a medium (> .25) effect size with SUI in the total sample. This relational pattern persisted after the sample was divided by gender. The central goal of this study was to determine the effectiveness with which SUI predicted the frequency of suicidal ideation as self-reported on the Outcome Questionnaire (OQ 45.2). Pearson product-moment correlations between SUI scores and responses to the suicide item of the OQ 45.2 were r = .69, .43, and .14 at the beginning, midpoint, and end of treatment, respectively. Results from regression analyses demonstrated that SUI significantly predicted the frequency of suicidal ideation at the start and midpoint of treatment for the total sample (adjusted R2 = .47 and .17, respectively) and for women but not men, but not at the end of treatment. Additional hierarchical regressions assessing SUI’s contribution above that of closely related MMPI- 3 scales to the explained variance in OQ-reported suicidal ideation frequency showed SUI’s contribution progressively declined over the course of treatment for the total sample and women. Classification accuracy analyses comparing SUI to verbal reports of suicidality at any time during treatment demonstrated a relative balance between sensitivity (.65) and specificity (.69). Accounting for the low base-rate frequency of suicidality, SUI produced a PPP of .62 and NPP of .85, indicating the relatively low frequency of false negative responses in comparison to false positives at the established clinical threshold of T= 58 for the scale. The overall hit rate for SUI was a moderate .68, indicating the correct classification of roughly two-thirds of participants. Overall, SUI scale scores appear to have the greatest signaling effect for suicidality at treatment onset and does not appear to be predictive of self-reported suicide disclosure at later stages. Its classification accuracy, while moderate overall, is more sensitive to the production of false-positive results than false negatives. Future studies with larger sample sizes are recommended to specify the duration over which SUI accurately predicts suicidality throughout psychological treatment.

Download

Share

COinS