Development and Psychometric Analysis of Internet Addiction Predict Test for Adolescents

范方,苏林雁,曹枫林,高雪屏,金宇,肖汉仕,王玉凤
DOI: https://doi.org/10.3969/j.issn.1005-3611.2008.01.001
2008-01-01
Abstract:Objective: To develop Internet Addiction Predict Test (IAPT) for Adolescents. Methods: 2620 middle school students(1319 boys,1301 girls,mean of age 15.1±1.7 years) were divided into IA group and CONTROL group according to Internet Addiction Test. Items were collected with significant difference (P<0.01) and higher discrimination (D>0.20) between the two groups from SDQ,EPQ,SCARED,SES and ATMD,and reconciled,screened and encoded,and then a draft of IAPT was formed. IAPT draft and IAT were used to investigate 1363 students(709 boys,604 girls,mean of age 13.69±0.95 years) to verify the IAPT’s reliability and validity. Results: The model of IAPT( 5 factors and 56 items) was clear and stable and its construct validity was receptible (df/n=2.5,NFI=0.84,NNFI=0.88 RMSEA=0.046). Its reliability (Cronbach α=0.72,and re-test reliability=0.61,P<0.001) was satisfactory. The correlation coefficients between predictable object (IAT) and IAPT and its factors were 0.35-0.60. The retest reliability after six months was 0.61,P<0.001. The difference of IAPT and its factors between IA and CONTROL group was significant(t=6.8-13.2,P<0.001;Cohen’s d=0.51-1.15>0.5). When the score of IAPT was 39,the Youden index was the biggest(0.62),the true positive fraction(TPF) was 0.83 and false positive fraction(FPF) was 0.21,and correct frequency was 0.80. Conclusion: The IAPT has good reliability and validity. And it could be used to predict adolescents' Internet addiction.
What problem does this paper attempt to address?