【作者】 周莉；

【导师】 姜潮；

【作者基本信息】 大连医科大学 ， 流行病与卫生统计学， 2003， 硕士

【摘要】 研究背景：自杀在中国是一个重要的公共卫生问题。中国自杀死亡的绝对人数居世界第一，全世界大约每年42％的自杀死亡发生在占世界人口的25％的中国人口。以往的研究结果表明，自杀与生理、心理、社会因素密切相关。对于自杀死亡者，如何构建有关死者的完善的心理、社会因素是一个难题，而心理解剖方法恰恰是提供解决此问题的有效方法。 所谓心理解剖(Psychological Autopsy)是一种通过与死者有关的活着的人访谈而重建自杀死亡者信息的方法。该方法已经在西方国家应用了近40多年的历史，目前普遍认为，从心理、社会角度研究自杀问题，心理解剖是首选的研究方法，具有较好的信度与效度。 中国大陆的自杀研究较西方国家时间较为落后，从研究方法上看，在自杀死亡的研究中更多的是采用流行病学调查研究。对于心理解剖的有关方法学的研究，尤其是其信度和效度的研究，尚缺少公开发表的证据。 研究目的：本研究是首次在中国大陆对心理解剖方法进行信度和效度的研究，主要探讨心理解剖方法在中国大陆的可行性，弥补我国在自杀研究中方法学上的不足。同时在本研究中，对于访谈工具的评价、信息人的选择、合成信息的处理进行了探讨，为心理解剖的应用性研究提供了一定的借鉴。 研究方法：根据当地卫生局提供自杀死亡者名单，选择辽宁省大连市金州、庄河自杀死亡者(目标人)66例，每例选取两个信息人进行访谈；同时选取与自杀组目标人在年龄(±5岁)、性别、居住地匹配的正常对照组66例，每例选取两个信息人，访谈对象为对照组本人和信息人，因此本研究共计访谈330人。采用结构式访谈问卷，主要问卷有Beck的自杀意图量表、Duke的社会支持量表、Paykel的生活事件量表、汉密顿抑郁量表(HAMD)、DSM-Ⅲ-R诊断结构式临床晤谈指导手册(SCID)、NEO人格问卷、自杀态度问卷(GSS)、流调中心抑郁水平评定量表(CES-D)进行评定。 将所有信息录入SPSS建立数据库，并采用SPSS11.0对主要量表进行信度与效度的研究。在信度评价中，采用分半信度和克朗巴赫(Cronbach)。系数进行内在一致性信度评价，采用借助于方差分析的组内相关系数(I CC via ANovA)和泛用KaPpa值(GK)，进行访谈者之间的一致检验，即评分者信度的评价。在效度评价中，以正常对照组的目标人作为“金标准”，将每个信息人与目标人以及两个信息人的合成信息与目标人的进行一致性检验、配对差异性检验及相关检验进行效度评价。同时将每个信息人与目标人的信息差值作为因变量，以信息人的一般状况作为自变量，进行线性多元逐步回归分析，以探讨信息人对效度的影响，并根据该项结果，探讨自杀组的效度问题。 研究结果:从分半信度和。系数研究结果看，无论自杀组和对照组以及两组的信息人，Beck的自杀意图量表、Duke的社会支持量表、NEO人格问卷、HAMD、CES一D在自杀研究中具有较高的内在一致性，由此也可看出，由信息人所构建的信息，即间接信息，具有可靠性。从评分者信度结果看，尽管在个别量表(可觉察支持量表)、个别项目上评分者的信度欠佳，但是并未影响总体结果，即在总体上本次心理解剖的研究中，Beek的自杀意图量表、社会交往量表、工具支持量表、死O人格问卷、HAMD、CES一D、paykel的生活事件量表、SCID、GSS受主观因素影响较小，同时也说明本次研究的访谈者在访谈工具的理解及掌握上较好，访谈者之间具有一致性，结果可信。 从效度研究结果看，心、理解剖方法具有较高的效度，即通过对信息人的访谈可以较为真实的构建出与自杀者有关的J心理、社会因素。本研究探讨了合成信急的效度问题，结果显示，对于定量资料:如自杀意图量表、社会支持各分量表、汉密顿抑郁量表、NEO人格问卷的N、E、O、A、C各因子，以两个信息人的信息、均值作为对目标人的信J自、估计:对于定性资料:生活事件量表以第二信自、人的信自、、SCID以两个信息人的阳性信息、作为对目标人的信自、估计。该结果适用于本次自杀研究的结果分析。 研究结论:在中国大陆采用心理解剖方法研究自杀具有可行性，并且具有较高的信度与效度。在本次心、理解剖研究中，对于采用的西方工具量表，由信息人所构建的信息具有可信性及有效性，但是在个别量表的翻译上须注重语言文化。在进行群体性心理解剖研究中，访谈者的培训至关重要，是保证研究质量的重要环节。对于信息人的选用，不仅考虑信息人对自杀事件的了解程度、对目标人的了解程度、而且其年龄、性别、教育水平也是需要考虑的因素。在本次研究中，对于不同信息、人的信J息合成，定量资料采用取均值，定性资料取阳性信自、，或最为可能的信息作为对目标人信息的估计。更多还原

【Abstract】 Background: Suicide has been an alarming public health problem in China. China has the highest number of completed suicides in the world each year. About 42% suicidal deaths in the world have occurred in China, which has one-quarter of the world’s population. Reports on suicide showed that suicide is closely associated with physical, psychological, and social factors. It is difficult to reconstruct the psychosocial factors of a completed suicidal, but the Psychological Autopsy maybe the only valid method to deal with this difficult problem.Psychological Autopsy, known as "a procedure for the reconstruction of suicidal death through interviews with survivors" has been employed in the West for suicide studies in the past four decades. It has been believed that Psychological Autopsy may be the best scientific approach to study the social and personal environments prior to the completed suicide. Studies in West have proved the methodology’s reliability and validity.Suicide studies in China started much later than in the West, and most of the Chinese studies in the past 20 years have been of epidemiological method. Suicide studies using Psychological Autopsy are comparatively new in China, and there is lack of published evidence for the method’s application in China, or its reliability and validity.Purpose: This thesis based on one of the very first Psychological Autopsy projects in China, studies the reliability and validity of the methodology, which fills the gap in the area of suicide studies in China. Using the data, I will assess the interview protocol, discuss the choice of informants, and comment on the integration of various data sources. All those issues have laid a solid foundation for future Psychological Autopsy projects to be held in China.Method: Based on the information provided by the Heath Administration of Jinzhou and Zhanghe County, Dalian, Liaoning province, our consecutive samplingyielded a total number of 66 completed suicides. For each suicide case, two informants were interviewed. Sixty-six age, gender, and location matched living individuals were selected as the community control group. For each control, two informants and the control him/herself were interviewed. The total number of interviews is 330. The protocol is a structured questionnaire which has the following major parts: Beck’s Suicide Intent Scale (SIS), Duke Social Support Index(DSSI), Paykel’s Interview for Recent Life Events(IRLE), HAM-D, The Structured Clinical Interview for DSM-III-R (SCID), Personality Measurement (NEO-FF-I), The General Social Survey Attitudes towards Suicide (GSS), and CES-D.All the information is entered into the SPSS Database. The reliability and validity of major instruments are studied with SPSS11.0. Split-half reliability and Cronbach’s Alpha value are used to examine the internal consistency for reliability test. And ICC via ANOVA and Generalized Kappa are used to estimate the inter-rater reliability. To study the validity of some instruments, we used the data from the control group. The community normal living controls provide "gold standard". We compare the responses from both the control him/herself and their informants. So the consistency test with ICC via ANOVA, Paired-Samples T Test and Bi-variate correlation are conducted to test the validity. The difference between each informant and control him/herself is as the dependent variable and the general situation of informants is as the independent variable. I conduct the linear regression (backward) analysis to investigate the validity affected by informants. With this result, the validity of measuring suicide group is to be discussed.Result: From the result of split-half and Alpha value, in the suicide, normal control, and the informants groups, there is high internal consistency in major instrument including SIS, DSSI, NEO-FF-I, HAM-D, and CES-D. Thus, the results show that the information from informants is reliable. From the results of inter-rater reliability, SIS. Social Interaction scale, Instrument Scale, NEO-FF-I, HAM-D更多还原