Àá½Ã¸¸ ±â´Ù·Á ÁÖ¼¼¿ä. ·ÎµùÁßÀÔ´Ï´Ù.
KMID : 0904520150390010049
Health and Medical Sociology
2015 Volume.39 No. 1 p.49 ~ p.73
The Influence of Trajectories for Drinking Behaviors on Those for Suicidal Ideation among Adolescents : Utilizing the Latent Class Growth Analyses
Jekarl Jung

Kwag Kyung-hwa
Kwon Hye-Bee
Kim Kwang-Gi
Abstract
This study sought to identify latent classes of adolescents who exhibited different developmental patterns of drinking behaviors and suicidal ideation. It also investigated whether latent classes of drinking behaviors influenced those of suicidal ideation after controlling for demographic and socio-psychological characteristics. The data for this study came from the second year middle school students of the Korea Youth Panel Survey (KYPS), collected from the 2nd year to the 4th year. To verify the purpose of this study, we performed latent class growth analyses and multinominal logistics regression analyses. Trajectories of drinking behaviors were categorized into four latent classes, low-level, decreasing, increasing, and high-level groups. As well, trajectories of suicidal ideation were categorized into four latent classes, low-level, decreasing, increasing, and high-level groups. Almost latent classes of drinking behaviors significantly influenced those of suicidal ideation. However, after controlling for socio-psychological characteristics, the effects became nonsignificant except the effects of increasing and high-level groups for drinking behaviors on those of high-level group for suicidal ideation. Therefore, the results identified variability in drinking behaviors and suicidal ideation across sub-groups of adolescents with different trajectory patterns. They also suggest that socio-psychological characteristics as well as drinking behaviors were determinants for adolescents¡¯ suicidal ideation. As a result, findings from the study highlight the importance of more sophisticated public policy to prevent and to intervene adolescent suicide.
KEYWORD
Korea Youth Panel Survey, Drinking Behavior, Suicidal Ideation, Latent Class Growth Analyses, Multinominal Logistic Regression Analyses
FullTexts / Linksout information
Listed journal information
ÇмúÁøÈïÀç´Ü(KCI)