Abstract

We attempted to find out, and validated, discriminant functions to differentiate high-risk adolescents of game addiction and low-risk adolescents of game addiction. The participants were 2,197 junior high school and high school students from 32 schools in Gwangju metropolitan city, Suncheon, Yeosu, Mokpo city through the stratified sampling. We divided participants from the 2,197 into two samples (1105 for research, 1092 for cross-validation). We classified high and low risk group as the mean +1 standard deviation (SD) and -1 SD among the 1,105 participants. this yielded 205 adolescents as high and 206 adolescents as low in game addiction. Stepwise discriminant function analysis yielded a function containing 5 variables (friends' positive attitudes to the game, aggression, impulsivity, supervision and control by the parents, interpersonal skills), which was 86.6% accurate in classifying the sample into either high or low in game addiction. To test the validity of the discriminant function, we used 1,092 participants. This yielded 208 adolescents as high and 224 adolescents as low in game addiction. On this validation sample, a function containing 5 variables (friends' positive attitudes to the game, aggression, impulsivity, supervision and control by the parents), classifying 84.3% accurately into either high or low in game addiction. Results suggested that high-risk group of game addiction and low-risk group of game addiction can be differentiated with the measure of friends' positive attitudes to the game, aggression, impulsivity, supervision and control by the parents. Implications for a model of adolescents' game addiction and directions for future research are discussed.