Factor Analysis of the Beck Scale for Suicide Ideation With Female Suicide Attempters

Abstract

A sample of 119 female suicide attempters completed the Beck Scale for Suicide Ideation (BSS). Although confirmatory common factor analyses of BSS items failed to support previously hypothesized one-, two-, or three-factor models, confirmatory principal components analyses substantiated hypothesized one- and two-dimensional models. Heuristics for the number of factors converged on two latent dimensions and exploratory principal components analyses verified the presence of two previously hypothesized suicide ideation factors: motivation and preparation. Scales based on this two-dimensional model demonstrated convergent validity with other suicide indices.

Depressive characteristics of adult psychiatric inpatients with a history of multiple versus one or no suicide attempts

“…Joiner (2005) has hypothesized that this factor would also distinguish multiple attempters from those with only one suicide attempt. In the only study of which we are aware to test this hypothesis, Holden and DeLisle (2005) found that, consistent with Joiner’s (2005) hypothesis, women’s number of previous suicide attempts was significantly related to their level of suicide preparation, but not motivation.…”

“…Specifically, although research has provided support for differences in hopelessness (Esposito et al ., 2003; Forman et al ., 2004; Kaslow et al ., 2006; Rudd et al ., 1996), it is not clear whether these differences would extend to other cognitive factors such as dysfunctional attitudes. Finally, although there is clear support for attempt history differences in levels of suicidal ideation (Forman et al ., 2004; Rudd et al ., 1996), research has suggested that overall levels of suicidal ideation encompass two distinct factors, one reflecting motivation for suicide and the other reflecting preparation (Beck, Brown, and Steer, 1997; Holden and DeLisle, 2005; Joiner, Rudd, and Rajab, 1997; Joiner et al ., 2003). There is evidence that the latter of these two factors, termed “resolved plans and preparation”, is more strongly linked to patients’ history of suicide attempts and eventual death by suicide (Joiner et al ., 2003).…”

Assessing suicide risk among callers to crisis hotlines: A confirmatory factor analysis

“…1,085 suicidal callers to crisis hotlines were divided into three sub-samples, which allowed us to conduct an independent Exploratory Factor Analysis (EFA), EFA in a Confirmatory Factor Analysis (EFA/CFA) framework, and CFA. Similar to previous factor analytic studies (Beck et al, 1997; Holden & DeLisle, 2005; Joiner, Rudd, & Rajab, 1997; Witte et al, 2006), we found consistent evidence for a two-factor solution, with one factor representing a more pernicious form of suicide risk (i.e., Resolved Plans and Preparations) and one factor representing more mild suicidal ideation (i.e., Suicidal Desire and Ideation). Using structural equation modeling techniques, we found preliminary evidence that the Resolved Plans and Preparations factor trended toward being more predictive of suicidal ideation than the Suicidal Desire and Ideation factor.…”

“…Nevertheless, in Sample 2, this item’s primary loading still appeared to be on the RPP factor. Furthermore, in previous factor analytic studies on measures of suicidal ideation (Beck et al, 1997; Holden & DeLisle, 2005; Joiner et al, 1997; Witte et al, 2006), a similar item assessing specificity of planning clearly loaded onto the RPP Factor. Thus, in the subsequent CFA, we began under the assumption that this item would have its primary loading on the RPP Factor, keeping in mind that this might be an area in need of modification if the CFA model had a poor fit.…”

“…Based upon the results of the EFA, EFA/CFA, and prior work (Beck et al, 1997; Holden & DeLisle, 2005; Joiner et al, 1997; Witte et al, 2006), we conducted a CFA (WLSMV estimator; pair-wise deletion) using a third sample ( N = 380; see Table 4). See Table 6 for bivariate correlations between all variables in Study 3.…”