SCL90和EPQ及UPI 3个量表关系研究

目的通过SCL-90、EPQ、UPI 3个量表在大学生心理健康调查中的应用,研究它们的内在有机关联。方法对某高校2005年入学新生进行心理健康普查,运用SCL90、EPQ、UPI 3个量表同时进行测量,实际选择了1 655例有效调查对象,通过计算Cronbachα系数、Guttman分半信度和主成分分析、因子分析、分析3个量表主要因子之间的关系。结果SCL90、EPQ、UPI 3个量表的Cronbachα系数分别是0.970 2、0.725 4、0.840 7;Guttman分半信度分别是0.938 9、0.735 6、0.772 3;随机抽取调查对象的5%重复测量3个表的Pearson相关系数分别是0.840、0.704、0.845,说明这3个量表此次评定结果可信;SCL90总分与SCL90各因子之间的相关系数介于0.747~0.910之间;UPI总分与UPI各因子的相关系数介于0.825~0.867之间;SCL90总分与UPI总分的Pear-son相关系数是0.699;SCL90与EPQ各因子之间的相关关系是,与神经质因子的相关系数较高,是0.663,与精神质因子的相关系数是0.301,与内外向因子和掩饰性因子的关系是反向关系;UPI总分与SCL90各因子的相关系数介于0.499~0.699之间;UPI总分与EPQ神经质因子的相关系数较大,为0.772,与精神质因子的相关系数是0.341;UPI总分与EPQ量表内外向因子和掩饰性因子呈反向关系,相关系数分别是-0.337与-0.328,而SCL90总分与内外向因子和掩饰因子的相关系数分别是-0.314和-0.289,2个量表均说明SCL90总分和UPI总分与内外向和掩饰成反相互关联,即内外向因子和掩饰因子越高,心理问题检出严重情况就越低;鉴别抑郁状态与SCL90各因子的相关系数介于0.434~0.690之间;鉴别精神分裂症与SCL90各因子的相关系数介于0.424~0.635之间;鉴别神经症与SCL90各因子的相关系数介于0.322~0.550之间。本次调查从3个量表的18个条目中抽取3个因子,分别代表不同的变量特征。结论心理调查的结果与所采用的心理测量量表有关,不同的量表会有不同的结果。在分析大学生的心理调查报告时,除报告结果外,还应说明使用的心理测量量表的种类。对不同的调查结果进行比较时,应考虑使用的量表是否相同,并应考虑量表之间的关系。

Study on relationship among three psychological checklists SCL90, EPQ and UPI

Objective To explore the relationship among three psychological checklists, SCL90, EPQ and UPI, through their application of mental health survey in the fields. Methods A total of 1 655 freshmen selected from a university were investigated with SCL90, EPQ and UPI checklists in 2005. Cronbach sand Guttman split-half were calculated to analyze the reliability in order to find the inner relationship among the SCL90, EPQ and UPI checklists with the correlation matrix principal component analysis and factor analysis. Results The Cronhach α（s） of SCL90, EPQ and UPI checklists were 0. 970 2, 0. 725 4, 0. 840 7, respectively. The Guttman split-half（s） were 0. 938 9, 0. 735 6 and 0. 772 3, respectively. Repeatedly measure correlations of 5 percent of the investigated students randomly sampling were 0. 840, 0. 704, 0. 845, respectively, which showed that the three checklists were reliable. （1） The correlation coefficients between SCL90 total scores and other components of SCL90 lay from 0. 747 to 0. 910; （2） The correlation coefficients between UPI total scores and other components of UPI lay from 0. 824 to 0. 867; （3） The Pearson＇ s correlation coefficient between SCL90 total scores and UPI total scores was 0. 699; （4） The correlation coefficients between SCL90 total scores and other components of UPI lay from 0. 546 to 0. 699; （5） The correlation coefficients between SCL90 total scores and other components of EPQ （EPQ-N, EPQ-P, EPQ-E and EPQ-L） were 0. 663, 0. 301, -0. 314 and -0. 289, respectively. （6） The correlation coefficients were 0. 772 between UPI total scores and EPQ-N, 0. 341 between UPI total scores and El：Q-P, -0. 337 between UPI total scores and EPQ-E, -0. 328 between UPI total scores and EPQ-L, respectively. These showed that the higher the EPQ-E scores, the better the university students＇ psychological health conditions. Conclusions When reporting psychological health condition results of psychological investigation, the investigator should report the methods and