Development and psychometric properties of the Suicidality of Adolescent Screening Scale (SASS) using Multidimensional Item Response Theory. (30th September 2016)

• Record Type: Journal Article
• Authors: Sukhawaha, Supattra; Arunpongpaisal, Suwanna; Hurst, Cameron
• Is Part Of: Psychiatry research. Volume 243(2016)
• Page Start: 431
• Page End: 438
• Publication Date: 2016-09-30
• Subjects: Psychiatry -- Periodicals; Psychiatry -- periodicals; Psychiatrie -- Périodiques; 616.89
• DOI: 10.1016/j.psychres.2016.07.014
• Languages: English
• ISSNs: 0165-1781

Abstract

Abstract: Suicide prevention in adolescents by early detection using screening tools to identify high suicidal risk is a priority. Our objective was to build a multidimensional scale namely "Suicidality of Adolescent Screening Scale (SASS)" to identify adolescents at risk of suicide. An initial pool of items was developed by using in-depth interview, focus groups and a literature review. Initially, 77 items were administered to 307 adolescents and analyzed using the exploratory Multidimensional Item Response Theory (MIRT) to remove unnecessary items. A subsequent exploratory factor analysis revealed 35 items that collected into 4 factors: Stressors, Pessimism, Suicidality and Depression. To confirm this structure, a new sample of 450 adolescents were collected and confirmatory MIRT factor analysis was performed. The resulting scale was shown to be both construct valid and able to discriminate well between adolescents that had, and hadn't previous attempted suicide. Highlights: Suicide prevention in adolescents by early detection using screening tools to identify high suicidal risk is a priority. Development of a suicidality scale based on both general and Thai-specific risk factors is necessary for an effective school – based suicidal prevention program in Thailand. The SASS demonstrated content validity comprehensive relevant to the context of suicidality in Thai adolescents. The SASS using MIRT performed shown high level of construct validity good fit model and demonstrated the satisfactory internal consistency and a high sensitivity to identify adolescent at higher risk for suicidal behavior in adolescents.