精神障碍者受审能力评定量表的编制

Development of Competency to Stand Trial Rating Scale in Offenders with Mental Disorders

目的研制符合中国法律体系的精神障碍者受审能力评定量表。方法从法学要件着手，提取出15个条目，编制出初步的精神障碍者受审能力评定量表，从决断值、条目与总分相关、校正条目与总分相关、条目删除后的d值、条目共同性、因素负荷量6个方面对量表进行条目分析，并以Logistic回归方程与ROC曲线划定界值两种方式进行诊断效能的判断。结果各条目极端组的决断值18．390～46．763，各条目与总分相关系数0．639～0．952，校正条目与总分相关系数0．582～0．944，条目共同性0．377～0．916，因素负荷量0．614～0．957。共7个条目进入回归方程，总样本的回判正确率为96．0％。ROC拟合曲线分析出宜以33分作为量表划界分，与专家鉴定意见的重叠率为95．8％，敏感性为0．938，特异性为0．966，阳性似然比为27．67，阴性似然比为0．06。结论量表编制合理，各条目符合同质性检验要求，诊断性评价指标较好。

Objective According with Chinese legal system, to develop a competency to stand trial rating scale in offenders with mental disorders. Methods Proceeding from the juristical elements, 15 items were extracted and formulated a preliminary instrument named the competency to stand trial rating scale in offenders with mental disorders. The item analysis included six aspects, which were critical ratio, item-total correlation, corrected item-total correlation, c~ value if item deleted, communalities of items, and factor loading. The Logistic regression equation and cut-off score of ROC curve were used to ex- plore the diagnostic efficiency. Results The data of critical ratio of extreme group were 18.390-46.763; item-total correlation, 0.639-0.952; corrected item-total correlation, 0.582-0.944; communalities of items, 0.377-0.916; and factor loadings, 0.614-0.957. Seven items were included in the regression equation and the accuracy of back substitution test was 96.0%. The score of 33 was ascertained as the cut-off score by ROC fitting curve, the overlapping ratio compared with the expertise was 95.8%. The sensibility and the specificity were 0.938 and 0.966, respectively, while the positive and negative likelihood ratios were 27.67 and 0.06, respectively. Conclusion With all items satisfied the requirement of homogeneity test, the rating scale has a reasonable construct and excellent diagnostic efficiency.