【作者】 杨海燕；

【导师】 蔡太生；

【作者基本信息】 中南大学 ， 应用心理学， 2010， 博士

【摘要】 目的：探讨工读生和普通生社会人口学特征、父母依恋与青少年行为问题的关系；初步建构并验证社会联接、负性情绪及应对方式在青少年父母依恋与行为问题中的中介模型。方法：运用横断面设计方法,采用青少年父母同伴依恋问卷(IPPAX青少年自评量表(YSQ)、学校依恋问卷(SAQ)、儿童青少年社会支持量表(SSRS-CS).攻击问卷(AQ)自评抑郁量表(SDS)及特质应对方式问卷(TCSQ)对248名工读生和678名普通生父母依恋与行为问题的关系进行比较研究,运用两因素方差分析方法、分层线性回归分析及结构方程模型对工读生和普通生父母依恋与行为问题的关系及其影响机制进行研究探讨。结果：(1)卡方检验和方差分析结果显示,工读生和普通生在家庭结构、父母文化水平、父亲依恋、母亲依恋、攻击、违纪、退缩行为和社交问题方面有显著差异(x2=36.06,24.21,31.16,25.22,35.83,80.98,254.82,18.60,62.62,P<0.001)；男女生在母亲信任、母亲沟通、疏离维度、违纪方面存在显著差异(F=5.81,17.40,8.28, 7.48, P=0.024, P<0.001, P=0.016, P=0.01);在分层线性回归分析模型中,在控制了学生组别的条件下,人口学特征和家庭环境特征对于各类行为问题有显著预测作用,其中对攻击行为、违纪行为、社交问题和退缩行为的解释变异率分别为2%、5%、3%、2%(F=87.56,64.99,3.62,5.76,P<0.001)。在控制了学生组别和人口学特征和家庭环境特征的条件下,父母依恋对于各行为问题有显著预测作用,其中对于攻击行为、违纪行为、退缩行为和社交问题的解释变异率分别为4%、6%、7%、9%(F=80.81,66.27,22.03,11.14,P<0.001)。(2)在社会联接、负性情绪、消极应对方式对青少年父母依恋与行为问题的中介模型,模型与数据吻合都很好：χ2=239.66.χ2/df=4.68,GFI=0.96,NFI=0.95；RESEA=0.07；χ2=250.93、χ2/df=4.92. GFI=0.96.NFI=0.96.RESEA=0.07；χ2=140.19.χ2/df=4.52.GFI=0.96. NFI=0.97.RESEA=0.05;非限定多群组预设模型的拟合度也都很好,χ2(df=158,n=898)=424.90,χ2/df=2.69,GFI=0.93,NFI=0.93. RESEA=0.04；χ2=498.63、χ2/df=3.07.GFI=0.91.NFI=0.94. RESEA=0.05；χ2=333.70,χ2/df=3.33.GFI=0.94.NFI=0.94. RESEA=0.05;其次采用极大似然法对多群组模型间嵌套模型进行比较检验,卡方值差异分别为Δχ2=84.89,P=0.001；Δχ2=51.52,P=0.016；Δχ2=52.86,P=0.003表明各群组学生的社会联接、负性情绪、消极应对方式对于父母依恋和行为问题影响的中介路径不同。结论：(1)与普通生相比,工读生单亲家庭居多、父母文化程度较低,在控制了学生组别和人口学特征的情况下,父母依恋对各类行为问题都有显著预测作用；(2)工读生父母依恋各维度、学校依恋、敌意、愤怒、抑郁、消极应对方式得分显著高于普通生,积极应对方式得分显著低于普通生；(3)社会联接各因素(同伴依恋、学校依恋、社会支持)负性情绪各因素(敌意、愤怒、抑郁)应对方式各维度(积极应对方式、消极应对方式)与父母依恋各维度、行为问题各因子显著相关；(4)社会联接、负性情绪、消极应对方式是父母依恋与行为问题的中介因素、且在不同群组中的路径系数各不相同。更多还原

【Abstract】 Objective:To compare the demography character difference between quasi-delinquents and normal middle school students, analysis the relationship among their demography, parent attachment and behavior problems; to develop and testify three structural equation models that assumed a relationship between to the adolescent parent attachment and behavior problems, as mediated by social bond, by negative emotion and by coping.Methods:A cross-sectional study design was used to compare 428 quasi-delinquents and 678 normal middle school students, Students completed the Inventory of parents and Peer Attachment (IPPA), Youth’s self Report (YSR), School Attachment Questionnaire (SAQ), Social Support Rating Scale for children and Adolescents (SSRS-CA), Aggression Questionnaire (AQ), Self-rating Depression Scale (SDS), Trait Coping Style Questionnaire (TCSQ). Use Chi-square test to explore the demography difference of the quasi-delinquents and normal middle school students; Used Two-factor ANOVA to explore the parent attachment and behavior problems difference among the quasi-delinquents and normal middle school students, boys, girls; Used Hierachiacal Regression Analysis to predict the relationship between the demography, parent attachment and behavior problems; Constructed pathway models, testified them and compare them with Structural Equation Model and Structural Weights ModelResults:(1) The Chi-square test and ANOVA result show there are significant different between quasi-delinquents and normal middle school students on family structure, parents education level, father attachment, mother attachment, aggression, delinquency, withdrawn and social problems (χ2=36.06,24.21,31.16,25.22,35.83,80.98,254.82, 18.60,62.62, P<0.001); there are significant different between boys and girls on mother’trust, mother’communication, mother’alienation and delinquency (F=5.81,17.40,8.28,7.48, P=0.024, P<0.001, P=0.016, P=0.01); the Regressions show that under control the students groups, demography character can significant predicted adolescents’aggression, delinquency withdrawn and social problems:2%,5%,3%,2%(F=87.56, 64.99,3.62,5.76, P<0.001); under control the students groups and demography character parents attachment can significant predicted adolescents’aggression, delinquency, withdrawn, and social problems: 4%,6%,7%,9%(F=80.81,66.27,22.03,11.14, P<0.001)(2) Parents attachment, mediated through adolescent social bond, negative emotion and coping, explains behavior problems, as tested by mediate model. All the fit of the mediation are well:χ2=239.66,χ2/df =4.68, GFI=0.96, NFI=0.95; RESEA=0.07;χ2=250.93,χ2/df=4.92, GFI=0.96, NFI=0.96, RESEA=0.07;χ2=140.19,χ2/df=4.52, GFI=0.96, NFI=0.97, RESEA=0.05; the fit of Unconstrained multi-group model are also well:χ2(df=158, n=898)=424.90,χ2/df=2.69, GFI=0.93, NFI=0.93, RESEA=0.04;χ2=498.63,χ2/df=3.07, GFI=0.91, NFI=0.94, RESEA=0.05;χ2=333.70,χ2/df=3.33, GFI=0.94, NFI=0.94, RESEA=0.05. Assuming model measurement weighs to be correct, used maximum likelihood method and nested model comparison provided a significantly better fit to the dataΔχ2=84.89, P=0.001;Δχ2=51.52, P=0.016;Δχ2=52.86, P=0.003. It shows that there is different pathway of the groups parents attachment, the mediated through adolescent social bond, negative emotion and coping, explains behavior problems.Conclusion:(1) Quasi-delinquents students are more come from single parent family and their parents’educational level are lower. Under control the demography character, parents attachment significant predicated adolescent behavior problems; (2) Quasi-delinquents students school attachment, anger, hostile, depression, negtive coping are significant higer than the normal school students; (3) Quasi-delinquents students’social bond (peer attachment, school attachment, social support), negative emotion (anger, hostile, depression), coping are significant relation to parents attachment and behavior problems. (4) Social bond, negative emotion, coping mediated the relation between parents attachment and behavior problems, and there is different pathway among multi-group.更多还原