【作者】 张斌；

【导师】 蔡太生；

【作者基本信息】 中南大学 ， 心理学， 2012， 博士

【摘要】 目的完美主义是一种追求完美无瑕,为自己设定过高的标准并对自己的行为和表现进行批判性自我评价的稳定人格特质倾向。本研究旨在探讨完美主义的结构及其本质、比较不同完美主义类型的心理特点的差异、验证完美主义双重过程模型理论、考察完美主义与抑郁的关系、初步构建并验证自尊、应对方式在完美主义和抑郁关系间的中介模型。为此,我们进行了五个研究。方法1.研究一,584名在校大学生完成Frost多维完美主义量表中文修订版(Frost Multidimensional Perfectionism Scale, FMPS)、Hewitt多维完美主义量表中文修订版(Hewitt Multidimensional Perfec-tionism Scale, HMPS)、Slaney近乎完美主义量表中文修订版(Almost Perfect Scale-Revised, APS-R)、Rosenberg自尊量表(Rosenberge Self-esteem Scale, SES)、Schwarzer一般自我效能感量表(General Self-Efficacy Scale, GSES)、Spielberger状态焦虑问卷(State Anxiety Inventory, SAI)、Beck抑郁问卷(Beck Depression Inventory, BDI),采用皮尔逊相关分析、探索性因素分析、验证性因素分析考察了完美主义的结构及其本质,探讨不同完美主义结构与积极、消极心理指标的关系。2.研究二,261名在校大学生完成Frost多维完美主义量表中文修订版(Frost Multidimensional Perfectionism Scale, FMPS)、Hewitt多维完美主义量表中文修订版(Hewitt Multidimensional Perfec-tionism Scale, HMPS)、Slaney近乎完美主义量表中文修订版(Almost Perfect Scale-Revised, APS-R)、Watson等人的正性、负性情绪量表(Positive and Negative Affect Scale, PANAS)、Rosenberg自尊量表(Rosenberge Self-esteem Scale, SES)、Beck抑郁问卷(Beck De-pression Inventory, BDI),采用聚类分析和多因变量方差分析(MANCOVA)方法对大学生完美主义进行分类,并比较不同完美主义类型心理特点的差异。3.研究三,292名在校大学生完成Frost多维完美主义量表中文修订版(Frost Multidimensional Perfectionism Scale, FMPS)、Hewitt多维完美主义量表中文修订版(Hewitt Multidimensional Perfec-tionism Scale, HMPS)、Slaney近乎完美主义量表中文修订版(Almost Perfect Scale-Revised, APS-R)、Watson等人的正性、负性情绪量表(Positive and Negative Affect Scale, PANAS)采用皮尔逊相关、偏相关分析、结构模型分析(Structural model analyses)探讨不同完美主义结构与正性、负性情绪之间的关系,验证完美主义的双重过程理论模型及其是否具有跨性别的测量恒等性。4.研究四,206名在校大学生完成Frost多维完美主义量表中文修订版(Frost Multidimensional Perfectionism Scale, FMPS)、Hewitt多维完美主义量表中文修订版(Hewitt Multidimensional Perfec-tionism Scale, HMPS)、Slaney近乎完美主义量表中文修订版(AlmostPerfect Scale-Revised, APS-R), Beck抑郁问卷(Beck Depression Inventory, BDI)、Spielberger状态焦虑问卷(State Anxiety Inventory, SAI)采用多元线性回归分析和分层回归分析(Hierarchical multiple regression)比较不同完美主义量表间的各个维度对抑郁症状的预测效力以及不同完美主义维度对4个月后抑郁症状水平变化的预测。5.研究五,412名在校大学生完成Frost多维完美主义量表中文修订版(Frost Multidimensional Perfectionism Scale, FMPS)、Hewitt多维完美主义量表中文修订版(Hewitt Multidimensional Perfectionism Scale, HMPS)、Slaney近乎完美主义量表中文修订版(Almost Perfect Scale-Revised, APS-R), Beck抑郁问卷(Beck Depression Inventory, BDI)、Rosenberg自尊量表(Rosenberge Self-esteem Scale, SES),解亚宁编制的简单应对方式问卷(Simplified Coping Style Questionnaire, SCSQ)采用皮尔逊相关分析、结构方程模型分析考察自尊、应对方式在完美主义和抑郁关系间的中介机制及其中介模型的跨性别一致性。结果1.研究一,完美主义量表可以提取适应不良完美主义、适应完美主义、秩序三个二阶因素。适应不良完美主义与自尊、自我效能等积极心理指标负相关(r分别为-0.46,-0.16,p<0.01),与焦虑、抑郁等消极心理指标正相关(r分别为0.48,0.55,p<0.001)；而适应完美主义与自尊、自我效能等积极心理指标正相关(r分别为0.12,0.19,p<0.05),与焦虑、抑郁等消极心理指标不相关(r分别为0.07,0.10,p>0.05)。2.研究二,以适应不良完美和适应完美为划分尺度,可以将大学生分为三类：适应完美主义者、适应不良完美主义者和非完美主义者。适应完美主义者的正性心理指标得分显著高于适应不良完美主义者(p<0.001),负性心理指标得分显著低于适应不良完美主义者(p<0.001)；非完美主义者的正性心理指标得分介于适应完美主义者和适应不良完美主义者之间,负性心理指标得分更接近适应完美主义者。3.研究三,适应不良完美主义能显著正向预测负性情绪(β=0.48,p<0.001),适应完美主义能显著正向预测正性情绪(β=0.49,p<0.001),完美主义的双重过程结构方程模型拟合结果较好(χ2=124.90,df=32,GFI=0.92,CFI=0.92,NFI=0.90,TLI=0.90,RMSEA=0.100)。结构方程的群组分析验证了完美主义的双重过程模型具有跨性别的测量一致性。4.研究四,完美主义对抑郁症状预测的横断面研究结果表明,适应不良完美主义中的差异性、社会决定完美主义因子能显著正向预测抑郁症状(β=0.372,p<0.001；β=0.264,p<0.01)；适应完美主义各因子均不能显著预测抑郁症状(p>0.05)。完美主义对抑郁症状预测的纵向研究结果显示,完美主义各因子均不能显著地预测4个月后抑郁水平的变化(p>0.05)。5.研究五,结构方程模型结果显示,消极应对方式和积极应对方式能部分中介适应不良完美主义和抑郁症状的关系,积极应对方式能部分中介适应不良完美主义和自尊的关系(χ2=42.24,df=23,GFI=0.96,CFI=0.97,NFI=0.93,TLI=0.95,RMSEA=0.064,AIC=86.23,BIC=159.45)。结构方程的群组分析验证了中介模型具有跨性别的测量一致性。结论1.本研究验证了完美主义量表二阶三因素的理论构建,支持了完美主义区分为适应不良完美主义和适应完美主义两个方面的合理性。2.不同完美主义类型的正性心理指标和负性心理指标得分有显著差异。3.适应不良完美主义和适应完美主义是完美主义的相对独立的两个成分,结果支持完美主义的双重过程模型理论及其跨性别的测量一致性。4.适应不良完美主义中的差异性、社会决定完美主义因子能显著的正向预测抑郁水平。5.消极应对方式和积极应对方式在适应不良完美主义和抑郁症状关系间起部分中介作用,积极应对方式在适应不良完美主义和自尊关系间起部分中介作用且中介模型具有跨性别的测量一致性。更多还原

【Abstract】 ObjectsPerfectionism is commonly conceived of as a personality style characterized by striving for flawless-ness and setting of excessively high standards for performance accompanied by tendencies for overly critical evaluations of one’s behavior. The present study aims to explore the construct and nature of perfectionism, to compare different groups of perfectionists on psychological adjustment, to identify the dual process model of perfectionism, to examine the relationship between perfection-ism and depression, to test the mediation effects of both self-esteem and coping styles between perfectionism and depression. Therefore, we conducted five studies.MethodsIn study1, a sample of584Chinese university students completed measures of the Frost Multidimensional Perfectionism Scale-Chinese Revised (FMPS), Hewitt Multidimensional Perfectionism Scale-Chinese Revised (HMPS), Slaney Almost Perfect Scale-Chinese Revised (APS-R), Rosenberg self-esteem Scale (SES), General Self-Efficacy Scale (GSES), State Anxiety Inventory (SAI), Beck Depression Inven-tory (BDI). Pearson correlation, exploratory factor analysis and confir- matory factor analysis were used to explore the construct and nature of perfectionism and examine the relationships between different perfec-tionism dimensions and psychological adjustment variables.In study2, a sample of261Chinese university students completed measures of the Frost Multidimensional Perfectionism Scale-Chinese Revised (FMPS), Hewitt Multidimensional Perfectionism Scale-Chinese Revised (HMPS), Slaney Almost Perfect Scale-Chinese Revised (APS-R), Waston Positive and Negative Affect Scale (PANAS), Rosenberg self-esteem Scale (SES), Beck Depression Inventory (BDI). Cluster analysis and multivariate analysis of covariance (MANCOVA) were used to empirically identify groups of perfectionists and compare different perfectionists on mental health variables.In study3, a sample of292Chinese university students completed measures of the Frost Multidimensional Perfectionism Scale-Chinese Revised (FMPS), Hewitt Multidimensional Perfectionism Scale-Chinese Revised (HMPS), Slaney Almost Perfect Scale-Chinese Revised (APS-R), Waston Positive and Negative Affect Scale (PANAS). Pearson correlation, partial correlation and structural equation model analysis were used to explore the relations among different perfectionism dimen-sions, positive affect and negative affect, testify the dual process model of perfectionism and test the model whether the measurement and struc-tural weights did not differ by gender. In study4, a sample of206Chinese university students completed measures of the Frost Multidimensional Perfectionism Scale-Chinese Revised (FMPS), Hewitt Multidimensional Perfectionism Scale-Chinese Revised (HMPS)、Slaney Almost Perfect Scale-Chinese Revised (APS-R), State Anxiety Inventory (SAI), Beck Depression Inven-tory(BDI). Multiple Linear regression analysis, hierarchical multiple re-gression were used to compare the predictive validity of the different perfectionism subscales in the prediction of the depressive symptoms and to predict the change in depression across a four-month longitudinal study.In study5, a sample of412Chinese university students completed measures of the Frost Multidimensional Perfectionism Scale-Chinese Revised (FMPS), Hewitt Multidimensional Perfectionism Scale-Chinese Revised (HMPS)、Slaney Almost Perfect Scale-Chinese Revised (APS-R), Beck Depression Inventory (BDI), Rosenberg Self-esteem Scale (SES), Simplified Coping Style Questionnaire (SCSQ). Pearson correlation, structural model analyses were used to test the mediation effects of both self-esteem and coping styles between perfectionism and depression, and test the mediation model whether the measurement and structural weights did not differ by gender.ResultsIn study1, it was found that perfectionism measures were factored as three second order perfectionism factors-adaptive perfectionism, maladaptive perfectionism, and orderliness. Maladaptive perfectionism was significantly and negatively correlated with self-esteem and self-efficacy (r=-0.46,-0.16; p<0.01), and positively correlated with anxiety and depression (r=0.48,0.55; p<.001), while adaptive perfec-tionism was significantly and positively correlated with self-esteem and self-efficacy (r=0.12,0.19;p<0.05), and not significantly correlated with anxiety and depression (r=0.07,0.10;p>0.05).In study2, it was found that cluster analyses were used to identify adaptive and maladaptive perfectionists and nonperfectionists based on maladaptive and adaptive dimensions. Adaptive perfectionists have sig-nificantly higher scores on positive psychological indicators than mal-adaptive perfectionists (p<0.001), whereas Adaptive perfectionists sig-nificantly lower scores on negative psychological indicators than mal-adaptive perfectionists (p<0.001). Scores of nonperfectionists on posi-tive psychological indicators were between adaptive perfectionists and maladaptive perfectionists. Scores of nonperfectionists on negative psy-chological indicators were close to adaptive perfectionists.In study3, it was found that maladaptive perfectionism predicted negative affect significantly and positively (β=0.48, p<0.001), while adaptive perfectionism predicted positive affect significantly and posi-tively (β=0.49, p<0.001). A dual-process structural model of perfection- ism was fitted to the data well (χ2=124.90, df=32, GFI=0.92, CFI=0.92, NFI=0.90, TLI=0.90, RMSEA=0.100). A multi-group analysis in SEM was performed to confirm that the model was measurement invariance between male and female.In study4, in cross-sectional analyses, the discrepancy and the so-cial-prescribed perfectionism subscales of maladaptive perfectionism could significantly predict baseline depressive symptoms (β=0.372, P<0.001;β=0.264, p<0.01), while none of adaptive perfectionism sub-scales could predict depression (p>0.05). In longitudinal analyses, none of perfectionism subscales significantly predicted the change in depres-sion across a four-month lag (p>0.05).In study5, structural equation modeling indicated that negative coping and positive coping both partially mediated the association be-tween maladaptive perfectionism and depression. The final model also revealed that positive coping partially mediated the relationship between maladaptive perfectionism and self-esteem (χ2=42.24, df=23, GFI=0.96, CFI=0.97, NFI=0.93, TLI=0.95, RMSEA=0.064, AIC=86.23, BIC=159.45). A multi-group analysis in SEM was performed to confirm that the mediation model was measurement invariance between male and female.Conclusions1. The current study verifies perfectionism is best explained as a three-factor multidimensional construct. The results also support that perfectionism should be divided into maladaptive perfectionism and adaptive perfectionism.2. There are significant differences on both positive and negative psychological indicators among different perfectionists.3. Maladaptive perfectionism is distinct from adaptive perfectionism. These are two independent components of perfectionism. The results support for the dual process model of perfectionism and the measure-ment invariance in this model between male and female.4. The discrepancy and the social prescribed perfectionism subscales of maladaptive perfectionism can significantly predict depression.5. Negative coping and positive coping both partially mediate the association between maladaptive perfectionism and depression. Positive coping partially mediate the relationship between maladaptive perfec-tionism and self-esteem. The mediation model is the measurement in-variance between male and female.更多还原