Measuring psychache as a suicide risk variable: A Mokken analysis of the Holden's Psychache Scale

Cecilia Blandizzi,Leonardo Carlucci,Michela Balsamo,Anna Contardi,Nicole Bungaro,Denise Erbuto,Maurizio Pompili,Marco Innamorati
DOI: https://doi.org/10.1016/j.jad.2024.09.160
2024-09-23
Abstract:Background: The Psychache Scale (PAS) is a questionnaire measuring trait-level psychological pain with satisfactory internal consistency and a strong correlation with suicidal ideation severity. However, inconsistent results have been reported for the PAS dimensionality. In the present study we used a non-parametric item response theory model, called Mokken Scale Analysis (MSA), to refine an unidimensional version of the PAS. Methods: The sample was composed of 400 Italian adults (312 females and 88 males) nonrandomly recruited from the general population. Results: A final set of 10 items satisfied the unidimensionality, local independence, and monotonicity assumptions, although it did not satisfy the double monotonicity assumption. The internal consistency of the PAS-10 was satisfactory (ordinal alpha = 0.98, ω = 0.97, and AVE = 0.82), and ROC curves analysis indicated good discriminant validity when differentiating participants with higher suicide risk from those with lower suicide risk. Limitations: Structural invariance between nonclinical and clinical samples was not investigated, and the presence of suicide ideation and behaviors was assessed with self-report measures with potential under-reporting of the phenomenon. Conclusion: The PAS-10 resulted to be a potentially valid and unidimensional measure of psychological pain (i.e., psychache) that could be used to screen adults at higher risk for suicide. Future studies are needed to investigate psychometric properties of the PAS-10 in clinical samples and to replicate results in independent samples.
What problem does this paper attempt to address?