【作者】 左晓冬；

【导师】 王欣；

【作者基本信息】 河北师范大学 ， 基础心理学， 2004， 硕士

【摘要】 目的：近年来，大学生的心理健康问题越来越受到人们的重视，成为社会关注的焦点。但是，目前对大学生心理健康进行研究的常用工具多系国外引进，且不是以大学生为适用对象，所以研究的效度问题十分突出，因此编制适合大学生这一特殊群体的心理健康测量工具具有重要的意义。本研究将在以前王欣教授研究的基础上：(1)建构对大学生心理健康状况进行评价的模式；(2)对大学生心理健康综合量表进行修订和完善，初步编制一个适合我国大学生的心理健康量表；(3)检验新编制的心理健康量表的信度和效度。 方法：在石家庄市选择三所具有代表性的大学进行分层整群抽样，共抽取大学生1038名，女性667名，男性371名。采用探索性因素分析法发展大学生心理健康量表的初步理论结构，再选用验证性因素分析法证明理论结构的合理性和正确性。选用Cronbach’sα系数、分半相关和重测相关系数检验量表的信度；效度检验主要包括内容效度、结构效度和效标效度。 结果：(1) 通过项目分析和初步的探索性因素分析，确定心理健康量表由54个项目组成。(2) 进一步探索性因素分析确定量表含有9个因素，共解释了总变异的52.88％。这9个因素分别命名为：躯体化，自尊，抑郁，目标感，人际关系，睡眠，焦虑，疲劳，强迫。(3) 内部一致性分析显示：Cronbach’sα系数、分半相关和重测相关系数分量表的分别为0.61—0.86，0.60—0.86，0.63—0.85；全量表的分别为0.95，0.93，0.83。(4) 量表内部相关分析显示，各项目与量表总分之间的相关系数在0.39—0.64之间，各分量表之间的相关系数为0.24—0.63，各分量表与全量表之间的相关系数为0.60—0.80，以上相关均具有显著性(P＜0.01)。(5) 验证性因素分析证明该量表的理论结构模型与测验变量拟合良好，支持该量表的理论结构。(6) 性别差异比较显示：男大学生在“躯体化”和“焦虑”分量表上的得分显著高于女大学生，在“抑郁”分量表上的得分显著低于女大学生，其他各分量表和量表总分两者差异不显著。(7) 非独生子女与独生子女的差异比较显示：非独生子女在“睡眠”、“焦虑”、“强迫”分量表上得分以及量表总分都显著高于独生子女，其他各分量表上的得分两者差异不显著。(8) 贫困生与非贫困生的差异比较显示：贫困生在“自尊”、“抑郁”、“睡眠”、“强迫”四个分量表和量表总分都显著高于非贫困生。 结论：(1) 本研究在以前研究的基础上初步完成了大学生心理健康量表的建构工作，且验证性因素分析证明该量表结构较为合理。(2) 信效度研究表明，该量表的同质信度、重测信度、内容效度、结构效度和效标效度都达到了心理测量学的要求。更多还原

【Abstract】 Objective:Recently,the college students mental health has been paid more attention.But most scales be used came from abroad, and weren’t applicable for college students.Therefor,it was very important to develop a mental health scale for using in college students. This study based on WangXin’s study,it’s aim is to: (1) explore the intra-structure of college student mental health;(2)revise the Comprehensive College Student Mental Health Scale,and develop a new Mental Health Scale for using in college student;(3)examine the reliability and validity of the scale.Methods:Taking 1038 college students(667 females,371 males) from three representative universities of Shijiazhuang by whole group.Exploratory and confirmatory factor analytic approaches were used to develop and evaluate the intra-structure of the scale.The reliability analysis included Cronbach’s a ,split-half reliability,and test-retest reliability.The validity analysis included content,construct,and criterion-related validity.Results:(1)Using item analysis and exploratory factor analysis,54 items are included in the scale.(2)Exploratory principal component analysis of the items indentified that the scale had nine factors,explaining 52.88% of the total variance.Nine factors was labeled as follows:Somatization, Self-esteem, Depression, Aim, Interpersonal relationship, Sleep, Anxiety, Tired-ness, Obsessive.(3)Internal consistency analysis showed that the Cronbach’s a coefficient,split-half and test-retest correlation coefficients ranged from 0.61 to 0.86,0.60 to 0.85, 0.63-0.85 respectively for each factor,and were 0.95,0.93 ,0.83 respectively for the total scale.(4)Inter-scale correlations studies showed that the correlations between items and total scores ranged from 0.39 to 0.64,the correlations among nine factors ranged form 0.24 to 0.63,the correlations between each factors and total scores ranged from 0.60 to 0.80. All correlations were significant(P<0.01).(5)Confirmatory factor analysis were conducted to assess the measurement of factor structure of the scale.The results supported nine-factor model of the scale.(6)Sex difference studies showed that the scores of Somatization .Anxiety in male college students were significantly higher than those in female college students.Significant higher score of Depression was found in female college students than those in male college students.(7)One-child difference studies showed that the scores of Sleep, Anxiety, Obsessive andwhole scale in only-child were significantly lower than those in multi-child.(8)Family income difference studies showed that the scores of Eelf-esteem,Depression,Sleep, Obsessive and whole scale in low family income students were significantly higher than those in high family income students.Conclusions:(1) The study finished developing the structure of college student mental health scale,and confirmatory factor analysis studies showed the structure of the scale is pretty reasonable.(2)Psychometic properteis analysis showed that the scale achieved good to excellent levels of internal consistency, test-retest reliability, and content, construct, and criterion-related validity.更多还原