A validity study of the reasons for life scale with emerging adult college students

Abstract

This study examined the validity of the Reasons for Life Scale (RFLS) with emerging adult college students. The RFLS measures "reasons for life." It was developed for use with Alaska Native youth as a way to assess potential risk of suicide without directly questioning about suicidal ideation or history of suicide attempts. This study sought to adapt the RFLS for use with emerging adult (age 18-25) college students, and to examine its factor structure and convergent validity with this population. First, a focus group was conducted to assist in rewording two Alaska Native-specific items from the RFLS for non-Natives. Then, with the additional items from the focus group, the revised version of the RFLS (RFLS-R) and other suicide-related measures were administered to a sample of 116 emerging adult college students. Exploratory factor analysis indicated a unidimensional factor structure for the RFLS-R with this sample. The RFLS-R showed a significant and strong correlation with the Reasons for Living Inventory (RLI; r = .70), which, like the RFLS-R, measures reasons for living but makes direct reference to suicide. There also were significant moderate negative correlations with the Suicidal Behavior Questionnaire - Revised (SBQ-R; r = -.36) and the Adult Suicidal Ideation Questionnaire (ASIQ; r = -.29). There was a significant moderate correlation between the RFLSR and a measure of socially desirable responding, the Balanced Inventory of Desirable Responding (BIDR; r = .31), with similar correlations found between the BIDR and other suicide-related measures included in this study. The results suggest that socially desirable responding did not strongly affect participants' responding or explain the associations found among the measures. The high correlation with the RLI suggests that the RFLS-R measures a similar construct, providing evidence of convergent validity; however, the RLI was more highly correlated with measures of suicidality than the RFLS-R -- suggesting that while the RFLS was moderately associated with measures of suicidality, it is a weaker predictor of suicide risk than the RLI. Although the RFLS-R was not as highly correlated with measures of suicidality as the RLI, which directly mentions suicide, the RFLS-R is the only known suicide measure that completely avoids items and instructions that mention suicide, therefore it may be useful in contexts where directly discussing suicide is not acceptable or appropriate.