Longitudinal measurement invariance of the Beck Scale for Suicide Ideation

• Published in: Psychiatry research Vol. 225; no. 3; pp. 368 - 373
• Main Authors: de Beurs, Derek P., Fokkema, Marjolein, de Groot, Marieke H., de Keijser, Jos, Kerkhof, Ad J.F.M.
• Format: Journal Article
• Language: English
• Published: Ireland Elsevier Ireland Ltd 28.02.2015
• Subjects: Adult; Factor Analysis, Statistical; Female; Humans; Longitudinal Studies; Male; Measurement invariance; Middle Aged; Psychiatric Status Rating Scales - standards; Psychiatry; Reproducibility of Results; Response bias; Response shift; Screener; Suicidal Ideation; Suicide ideation; BSS; CI; PITSTOP suicide; CFI; Response bias; Screener; ROM; RMSEA; Response shift; MI; Suicide ideation; Measurement invariance; d.f; CFA; measurement invariance; degrees of freedom; root mean square error of approximation; professionals in training to STOP suicide; routine outcome monitoring; confirmatory factor analysis; confidence interval; Beck Scale for Suicide ideation; comparative fit index
• DOI: 10.1016/j.psychres.2014.11.075
• ISSN: 0165-1781; 1872-7123

In mental health care, both clinical and scientific decisions are based on within-subject comparisons of test scores on the same self-report questionnaire at different points in time. To establish the validity of test score comparisons over time, longitudinal measurement invariance should be established. The current study tested whether the 19 item Beck Scale for Suicide Ideation (BSS) is measurement invariant (MI) over time. As the first five items of the scale are often used to screen for the presence of suicidal thoughts, we also tested a model consisting of only the first five items. Psychiatric in- and out-patients (n=475) completed the questionnaire upon admission and after 3 months. By means of confirmatory factor analysis (CFA) we tested whether the parameters of a single factor model were equal over time. All fit indices indicated that both the 19-item questionnaire and the five-item screener were measurement invariant over time. This means that changes in test-scores over time can be attributed to true changes in the construct of interest. These findings legitimate the use of the 19 item scale and the five-item screener in longitudinal assessments. •The present study is the first to examine longitudinal measurement invariance of the Beck Scale for Suicide Ideation.•All fit indices indicated that both the 19-item questionnaire and the five-item screener were measurement invariant over time.•Changes in test-scores on the Beck Scale for Suicide Ideation over time can be attributed to true changes in the construct of interest.•These findings legitimate the use of the 19 item scale and the five-item screener in longitudinal assessments.