Developing a student-targeted Japanese version of the Literacy of Suicide Scale by Item Response Theory

Abstract

There is no scale to assess knowledge of suicide in Japan. To develop a Japanese version of the Literacy of Suicide Scale (LOSS) measuring knowledge of suicide, we translated 27 items of the original version which was developed in Australia and added 6 items unique to Japan to create a Japanese Literacy of Suicide Scale (LOSS-J) consisting of 33 correct and incorrect questions. The items were selected using Item Response Theory. The aim of this study was that reliability and validity were verified. We used a total of 747 responses by college students, graduate students, junior college students, and vocational school students for analysis. Two items with low relevance were excluded from the analysis based on the correct answer rate and the point biserial correlation coefficient. The remaining 31 items confirmed the one-dimensionality of the scale. The discriminating power of the items was sufficiently high, and the difficulty parameter was negative on average by a two-parameter logistic model. The test information function showed that the LOSS-J had high discriminating power when applied to participants with low characteristic values of suicide knowledge. The correlation with gatekeeper self-efficacy was weak, but there was a positive correlation with educational experiences on suicide prevention and comprehension of suicide prevention programs. Thus, LOSS-J was confirmed to have reliability and predictive validity.