中.高生의 中途脫落 因果模型 檢證과 判別 尺度 開發 = Causal model of middle/high school students' dropout and development of discriminant measurement tool

다국어 초록 (Multilingual Abstract)

The present study was investigated to test the causal model between dropout antecedents and dropout intention for middle and high school students and to develop a scale for discriminate students prone to dropout before they quit. The results of this research provide some significant findings.
In the first study to derive dropout related items, 200 items were derived from interview with high school teachers about students' school life and the existing dropout scales. 200 items were administered to 640 middle and high school students(male, 316; female, 324). 640 cases were used in the item analysis and factor analysis. After series of the analyses were conducted, 184 items among dropout items were included to the validation study.
Second, the initial causal model was tested with two-steps. First, the initial model include 23-indicator, which measures five exogenous variables(personal characteristics, family situations, schools, peer relationships, local communities) and two endogenous variables(the percentage of attendance and absence, dropout intention). The model was tested with 584 students samples(male, 281; female, 284), and modified based upon modification indices. In specific, original 23 indicators were reduced to 12 indicators. After applying GL(generalized least square method), which assumed multivariate normality, to the modified model, the overall model fits were significantly improved(χ2(61)=197.338(p<.000), RMR=0.0345, GFI=0.950, AGFI=0.914, NFI=0.950, NNFI=0.947). Second, the modified model was tested whether it can be generalized into other samples. The results of cross validation with using the new 405(male, 200; female, 205) samples showed the modified model was acceptable in terms of its goodness of fit indices(χ2(61)=248.044(p<.000), RMR=0.0708, GFI=0.912, AGFI=0.849, NFI=0.930, NNFI=0.919). Third, regarding the relationship between predictors and criterion, most of predictors are significantly related to at least a criterion. However, positive personal characteristics(internal motivation, circumstantial judgement, teacher trust) had not significantly related to any criterion. Therefore, we can conclude that person related variables do not have any role in predicting dropout criteria.
Forth, to examine whether the dropout scale can distinguish current students from dropout students, t-test was employed. The results showed that dropout students have significantly higher mean than current students in terms of low self esteem, retaliate behavior, conflict with parents, study non-preference, negative attitude against teacher, negative attitude against school, negative peer relationship, inappropriate sexual behavior, toxic resident environment, negative class participation, delinquency, absenteeism. On the other hand, dropout students have significantly lower mean than current students in terms of positivity, teacher trust, non-toxic resident environment, sexual characteristic.
Finally, to examine the discriminability of dropout scale, the discriminant analysis was performed. The results showed that nine out of 12-factor can significantly contribute to group categorization, and the accuracy of categorization was approximately 90%. However, there is a possibility that the high accuracy percentage was attributed to the sample. Because the same data was used in both developing discriminant functional equation and examining accuracy percentage.

목차 (Table of Contents)

• 목차 = ⅰ
• Abstract = ⅵ
• Ⅰ. 서론 = 1
• 1. 연구의 필요성 및 목적 = 1
• 2. 연구 문제 = 5
• 3. 연구 범위 = 6
• Ⅱ. 이론적 배경 = 8
• 1. 중도탈락의 개념 = 9
• 2. 중도탈락 현상의 원인분석 = 12
• 3. 중도탈락의 과정 및 결과 = 27
• 4. 중도탈락 현상의 설명모형 = 41
• 5. 중도탈락 판별척도 = 63
• 6. 본 연구의 가설적 모형 = 69
• Ⅲ. 연구 방법 = 73
• 1. 문항 수집 = 73
• 2. 1차 문항분석을 위한 측정도구와 자료 = 78
• 3. 척도의 타당화와 인과모형 검증을 위한 척도와 자료 = 80
• 4. 교차타당화와 중도탈락 판별을 위한 조사도구와 자료 = 82
• Ⅳ. 연구결과 = 85
• 1. 1차 문항개발 및 문항분석 결과 = 85
• 2. 척도의 타당화 및 중도탈락의 인과모형 검증 = 86
• 3. 교차타당화, 변별타당화 및 중도탈락자의 판별 = 112
• Ⅴ. 결론 및 제언 = 120
• 1. 결론 및 논의 = 120
• 2. 제언 = 124
• 참고문헌 = 125
• 부록 = 137