军人团体智力量表A式的研制

Development of group intelligence scale for military personnel （Part A）

目的研制军人团体智力量表A式，并进行信效度检验。方法以智力震荡理论及晶体智力和流体智力理论为基础，确定量表由2个分量表4个分测验组成。采取随机整群抽样方法，抽取1000名现役健康官兵施测，采用SPSS17．0及AMOS7．0进行条目分析及信效度检验。结果各分测验的平均难度分为0．49（0．43-0．60），鉴别指数为0．43（0．35-0．50）；总量表及分测验的重测信度为0．773-0．883，Cronbacha系数为0．638-0．880，分半信度为0．503-0．831；探索性因素分析抽取的4个因子（常识与相似概括、领悟与词义解释、图画填充及图形推理）解释总方差比例的61．009％；验证性因素分析显示，本量表所有拟合指标均达到可接受标准。结论军人团体智力量表A式的条目质量及信效度符合心理测量学的要求。

Objective To develop the Group Intelligence Scale for Military Personnel （GIMP-A） and verify its reliability and validity. Methods On the basis of intellectual shock theory （IST）and crystal intelligence and fluid intelligence theory, the GIMP-A was made up of 2 subscales and 4 subtests. Tests were carried out in 1000 servicemen selected according to random cluster sampling, items analyzed and reliability and validity verified using the SPSS17.0 and AMOS7.0. Results The average difficulty score of each subtest was 0.49 （0.43~0.60） ; discrimination index was 0.43 （0.35~0.50） ; the test-retest reliability of total scale and subtests was 0. 773 ~ 0. 883; Cronbach coefficient was 0. 638~0. 880; split-half reliability was 0. 503 ~ 0. 831; four factors （normal knowledge and generalizability, insight and word explain, picture completion and figure inference） selected according to exploratory factor analysis which could explain 61. 009% of total variance; analysis of confirmatory factor indicated that all fitting indexes of the scale a- chieved acceptable standard. Conclusion Item quality, reliability and validity of the GIMP-A are up to requirement of psychometrics.