Datset for Internal structure of Beck Hopelessness Scale: An analysis of method effects using the CT-C(M–1) model

The construct validity in relation to the dimensionality or factor structure of the Beck Hopelessness Scale (BHS) has long been debated in psychometrics. Irrelevant variance due to item wording (method effects) can distort the factor structure, and recent studies have examined the method factor’s role in the factor structure of the BHS. However, the models used to control the method effects have severe limitations, and new models are needed. One such model is the correlated trait-correlated method minus one (CT-C(M-1)), which is a powerful approach that gives the trait factor an unambiguous meaning and prevents the anomalous results associated with fully symmetrical bifactor modeling. The present work compares the fit and factor structure of the CT-C(M-1) model to bifactor models proposed in previous literature and evaluates the convergent validity of the CT-C(M-1) model and its discriminatory capacity by taking suicidal ideation as the criterion variable. This study used a large and heterogeneous open mode online sample of Argentinian people (N = 2,164). For this study, the adapted Beck Hopelessness Scale (BHS) version for the Argentinian population was used (Mikulic et al., 2009). This scale comprises 20 items with a dichotomous reply format (i.e., true or false) and is used to evaluate the respondent’s negative expectations for the future. In addition, the version of the Inventory of Suicide Orientation-30 (ISO-30) validated by Fernandez-Liporace and Casullo (2006) in Argentina was used for this study. The instrument measures the respondent’s level of agreement with certain statements using a four-point Likert scale (with responses ranging from 1, “I strongly disagree” to 4, “I strongly agree”). From the inventory, only the questions that were related to the dimension of SI were included (e.g., “In order to stop things from getting worse, I believe suicide is the solution”). Within the ISO-30 scale, the suicidal ideation factor is the one that has shown the highest consistency and evidence of measurement validity in the literature on the internal structure of the scale (Vecco et al., 2021). In the present research, the high-SI and no-SI groups were identified based on the scores obtained from the ISO-30 SI subscale (possible range values = 4–16). A total score of 4 (i.e., the participant marked “strongly disagree” on all items) was considered to indicate an absence of SI, while a score of 16 (i.e., the participant marked “strongly agree” on all items) indicated a high SI. Individuals scoring in the middle range were not included in the analyses. The results indicated that the CT-C(M-1) model with positive words as referenced items achieves the most adequate factor structure. The factorial scores derived from this model demonstrate good predictive and discriminating capabilities