정보통신부의 인터넷중독 자가진단검사(K척도)는 청소년이 작성하기에 너무 길기 때문에 현장적용성을 높이기 위한 20문항의 간략형 인터넷중독 자가진단검사(K척도 Short form: KS 척도)의 개발 필요성이 제시되었다. 이에 이 연구에서 KS 척도의 타당성을 조사하기 위하여 이 연구가 수행되었다. 이 검사는 정보문화진흥원의 지원 아래 전국 규모로 실시되었다. 자료 수집을 위하여 K척도 40문항과 문제행동을 묻는 문항 17문항을 포함하여 전체 57문항의 초등학생용·중고등학생용 조사지를 구성하였으며, 참여 학생수는 전국적으로 3,016명이었다. 간략형 척도 구성을 위해 2002년 완성된 K척도 검사 결과를 활용하여, 인터넷 중독 고위험군, 잠재적위험군, 일반사용자군을 통계적으로 잘 판별해 내는 척도와 문항들을 판별분석을 통해 구분하고, 전문가 평가와 자문회의를 통하여 KS 척도를 구성하였다. KS척도로 구분한 고위험군, 잠재적 위험군, 일반사용자군과 K척도로 구분한 세 집단을 비교한 결과 의미있는 수준의 중첩이 확인되어 KS척도가 타당하게 구성되었다는 점을 확인하였다. 또한, KS척도의 신뢰도 분석 결과, 통계적으로 유의미하게 높은 수준의 신뢰도 점수를 확인할 수 있었다. KS척도의 타당도 분석 중 준거 타당도 분석을 위해 청소년위기척도를 함께 비교해 본 후, 추가로 조사한 인터넷 중독과 관련된 문항과 함께 살펴보았다. 그 결과, KS척도의 하위척도들과 청소년위기척도와의 상관을 살펴보았을 때도, 통계적으로 유의한 수준의 상관을 보여, 상관성을 유지하고 있다는 점을 확인할 수 있었다. 그 후, 구인 타당도 분석을 위해, 공분산구조방정식 모형 분석을 실시하였다. 분석된 결과는 5개의 요인구조가 타당하며, 각 하위 척도 간의 관계를 보여주었다.
This study was aimed at developing short version of K scale. K Scale that has been used was too long to fill out for youth and hard to screen faking good or bad. Therefore, in this study, we cut down 40 items of K scale into 20 items. These scales were developed for youth who are prone to internet addiction. KS scale increased sensitivity of screening yet decreased items in half for the sake of testee. KS scale was developed based on K scale. 3,016 K12 students participated in this study. The survey included K scale, KS scale, and Youth Risk Scale, which was developed at Korea Youth Counseling Institute. 20 items of KS scale were selected from K scale(2002) using determinant analysis. High risk group was defined above 70 of standardized score and low to medium risk group is between 65 and 69 in standardized score. As a result of comparing the compositions of each group(High risk, low to medium risk and general groups) from KS scale and K scale, both scales predicted 70~80 percent of each group in same way. Therefore, the validity of KS scale was achieved. As a result of reliability analysis of KS scale, Cronbach's alpha for the elementary students score was .887 and Cronbach's alpha for middle school students score was .909. Compared to Cronbach's alpha in K scale(.944 for elementary students and .955 for middle school students), the reliability of KS scale became a little lower. Yet the reliability of the KS scale is statistically significant. For the analysis of criterion-related validity, we compared the score of KS scale with the scores of internet addiction in Youth Risk Scale. The results of correlation coefficient showed .49 for high school students and. 60 for elementary students. In short, the student who has high score in KS scale also has high score in Youth Risk Scale. Also, the correlation scores of sub-factors between KS scale and Youth Risk Scale are around. 40. therefore, we could confirm correlation between KS scale and Youth Risk Scale. For the analysis of construct validity, we tested Structural Equation Model using AMOS 12.0. The result supported the model of 6 factors in the KS scale(NFI = 0.983; CFI = 0.990; RMSEA = 0.034; for elementary students, NFI = 0.985; CFI = 0.997; RMSEA = 0.019 for middle school and high school students). These scales are for youth. KS scale could be a check list for an internet user to test their proneness to internet addiction by him/herself.
This study was aimed at developing short version of K scale. K Scale that has been used was too long to fill out for youth and hard to screen faking good or bad. Therefore, in this study, we cut down 40 items of K scale into 20 items. These scales were developed for youth who are prone to internet addiction. KS scale increased sensitivity of screening yet decreased items in half for the sake of testee. KS scale was developed based on K scale. 3,016 K12 students participated in this study. The survey included K scale, KS scale, and Youth Risk Scale, which was developed at Korea Youth Counseling Institute. 20 items of KS scale were selected from K scale(2002) using determinant analysis. High risk group was defined above 70 of standardized score and low to medium risk group is between 65 and 69 in standardized score. As a result of comparing the compositions of each group(High risk, low to medium risk and general groups) from KS scale and K scale, both scales predicted 70~80 percent of each group in same way. Therefore, the validity of KS scale was achieved. As a result of reliability analysis of KS scale, Cronbach's alpha for the elementary students score was .887 and Cronbach's alpha for middle school students score was .909. Compared to Cronbach's alpha in K scale(.944 for elementary students and .955 for middle school students), the reliability of KS scale became a little lower. Yet the reliability of the KS scale is statistically significant. For the analysis of criterion-related validity, we compared the score of KS scale with the scores of internet addiction in Youth Risk Scale. The results of correlation coefficient showed .49 for high school students and. 60 for elementary students. In short, the student who has high score in KS scale also has high score in Youth Risk Scale. Also, the correlation scores of sub-factors between KS scale and Youth Risk Scale are around. 40. therefore, we could confirm correlation between KS scale and Youth Risk Scale. For the analysis of construct validity, we tested Structural Equation Model using AMOS 12.0. The result supported the model of 6 factors in the KS scale(NFI = 0.983; CFI = 0.990; RMSEA = 0.034; for elementary students, NFI = 0.985; CFI = 0.997; RMSEA = 0.019 for middle school and high school students). These scales are for youth. KS scale could be a check list for an internet user to test their proneness to internet addiction by him/herself.
