아동이 지각한 부모 간의 갈등, 어머니의 양육행동, 사회적 지지 및 안녕감 간의 구조적 관계 분석 = The effects of perceived parental conflict, maternal behaviors, and social support on children's well-being; A structural equation modeling analysis

다국어 초록 (Multilingual Abstract)

The purpose of this study was to develop a model for children's well-being model and examine the effects of perceived parental conflict, maternal behaviors, and social support on school-age children's subjective and psychological well-being. The present study consisted of two phases: Testing the measurement model and development of structural model.
In total, 1,481 children (M=10.65 years, SD= 1.04) were recruited from nine public elementary schools in Seoul, South Korea. Within the entire sample there were 464 boys and 425 girls from grades 4-6. Children were asked to rate the following four instruments on a 4 point Likert scale; The Children’s Perception of Interparental Conflict Scale (CPIC: Grych, et al., 1992), The Parental Bonding Instrument (PBI: Parker, 1979), The Social Support Appraisals Scale (APP: Dubow & Ullman, 1987), and Well-Being Scale (Keyes, 2005). The data were analyzed by descriptive statistics, Cronbach's alpha, explanatory factor analysis (EFA), and confirmatory factor analysis (CFA) using SPSS 12.0 and AMOS 5.0. The results were the following:
In the first phase, the measurement model was tested. First, the means and standard deviations of the measured variables were as follows: Parental conflict, M= 1.77, SD= .53; maternal behaviors, M= 3.21, SD= .44; social support, M= 3.09, SD= .48; children's well-being, M= 2.56, SD= .50.
Second, EFA was conducted using SPSS 12.0 Program. Parental Conflict had three factors: Frequency, stability, and intensity. Maternal Behavior had two factors: Care and autonomy-respect. Children’s Social Support had two factors: Teacher support and peer support. And Children’s Well-being had two factors: Subjective well-being and psychological well-being. The following Cronbach’s α were obtained for each sub-scales: .85, .83, and .82 for parental conflict; .87 and .67 for maternal behaviors; .92 and .83 for social support; and .77 and .63 for children’s well-being.
Third, CFA was conducted using AMOS 5.0 Program. The measurement model consisted of four latent variables (i.e., parental conflict, maternal behaviors, social support, and children’s well-being), each with two or more indicators. The four latent variables were allowed to correlate with one another. The measurement model was fit to the data and resulted in the following excellent fit indices: χ2 (21, N=889)= 34.24, χ2/df = 1.63; TLI= .976; CFI= .986; RMSEA= .046. Convergent validity was supported for the measures (p<.001 level). Furthermore, the inter-correlations among the latent variables were all significant (p<.01 level). Specifically, parental conflict was significantly related to the maternal behavior (r= -.55), social support (r= -.41), and well-being (r= -.42); maternal behavior was related to social support (r= .67), and well-being (r= .56); and social support was related to well-being (r= .66). This means that the measurement model had appropriate discriminant validity.
And also, the measurement model was tested separately for various groups in terms of sociodemographic characteristics such as age and gender. There were no statistically significant differences observed between grade 4 and 5 (χ2= 13.101, df= 11, p= .28 >.05), between grade 5 and 6 (χ2= 14.027, df= 11, p=.23 >. 05), between grade 4 and 6 (χ2= 11.612, df= 11, p= .394 >.05), and between gender (χ2= 16.526, df= 11, p= .123 >.05). These findings suggest that the factor structure of measurement model is applicable for use with Korean children from 9 to 12 years of age regardless of gender. These findings support the generalizability of the factor structure.
In the second phase, the structural equation model was developed to examine the structural relationships among the perceived parental conflict, maternal behaviors, social support, and children's well-being. The main hypotheses were as follows:
First, children's perceived parental conflict is negatively related to children's well-being. Second, the effect of children's perceived parental conflict is mediated by children's perceived maternal behavior. Third, the effect of children's perceived parental conflict is mediated by the two factors, maternal behavior and social support. Fourth, maternal behavior is positively related to children's well-being. Fifth, the effect of children's perceived maternal behavior is mediated by children's perceived social support. Sixth, children's perceived social support is positively related to children's well-being.
Based on these six hypotheses, this study examined how well empirical data fitted the hypothetical model In order to choose the best-fit model, model modification strategy provided by SEM was utilized. The initial model was selected for satisfying parsimoniousness, fitness, and robustness, compared with the modified model. The final structural equation model was fit to the data and resulted in the following excellent fit indices: χ2 (22, N=889)= 69.40, χ2/df = 3.15; TLI= .970; CFI= .982; RMSEA= .049. And also, the final structural equation model was tested separately for various groups in terms of sociodemographic characteristics such as age and gender to see how robust the model was. The results supported the robustness of the final structural equation model.
The findings show that perceived parental conflict is negatively related to children's well-being, and the relationship is also explained by the mediating factor, perceived social support. However, perceived maternal behavior has no direct effect on children's well-being, but has indirect effects only through the mediating factor, perceived social support. This result provides a counter evidence to the general perception that maternal behavior is the most important variable in children's adjustment. Specifically, higher parental conflict decreases maternal behavior, and poor maternal behavior make children perceive less social support, and less social support decreases children's well-being. The study results suggest that it is necessary to recognize the importance of social support in order to understand children's well-being and to develop prevention intervention program.
This study provides important implications for developing preventive interventions for enhancing children's well-being. The findings identify the importance of social support as a mediator that can be the target of intervention. By targeting the mediating factor, the intervention goal might be drawn to improve the school atmosphere to provide children appropriate social support.

목차 (Table of Contents)

• Ⅰ. 서 론 1
• 1. 연구의 필요성 및 목적 1
• 2. 용어의 정의 6
• Ⅱ. 이론적 배경 8
• 1. 안녕감의 정의 9
• 1) 주관적 안녕감 10
• 2) 심리적 안녕감 13
• 2. 안녕감 연구의 최근동향 16
• 3. 아동의 안녕감 18
• 4. 아동의 안녕감에 영향을 미치는 주요 변인들 24
• 1) 부모 간 갈등 25
• 2) 어머니의 양육행동 30
• 3) 사회적 지지 34
• 5. 연구 가설 39
• Ⅲ. 연구방법 및 결과 41
• 1. 척도의 개발 및 타당화 41
• 1) 연구방법 41
• (1) 연구대상 41
• (2) 연구도구 42
• (3) 연구절차 44
• (4) 자료의 분석 46
• 2) 연구결과 50
• (1) 탐색적 요인분석 결과 50
• (2) 확인적 요인분석 결과 73
• 2. 구조모형의 개발 85
• 1) 연구방법 85
• (1) 연구대상 85
• (2) 연구도구 85
• (3) 자료의 분석 86
• 2) 연구결과 97
• (1) 모형의 평가 97
• (2) 모형의 수정 99
• (3) 가설의 검정 102
• (4) 효과의 분해 105
• 3. 모형의 교차 타당화 109
• 1) 연구방법 109
• (1) 연구대상 및 절차 109
• (2) 자료의 분석 111
• 2) 연구결과 112
• (1) 타당화 모형의 평가 112
• (2) 타당화 모형의 가설검정 114
• (3) 타당화 모형의 효과분해 116
• 4. 모형의 적용성 검토 120
• 1) 연구방법 120
• (1) 연구대상 120
• (2) 자료의 분석 121
• 2) 연구결과 122
• (1) 학년별 모형의 평가 122
• (2) 성별 모형의 평가 128
• Ⅳ. 논의 및 결론 132
• 1. 논 의 132
• 2. 연구의 의의, 제한점 및 제언 147
• 3. 결 론 151
• 참고문헌 153
• 부록목차 177
• ABSTRACT 197