心理筛查人格量表的修订及初步验证

REVISION AND VERIFICATION OF MINNESOTA MULTIPHASIC PERSONALITY INVENTORY

目的对目前广泛运用的明尼苏达多相人格量表(MMPI)中的L分量表进行修订,以期提高其信效度。方法使用DXC-6型群体心理测评仪对1290名新兵进行MMPI测验,编制拟修订题目后对另外470名新兵进行L量表及拟修订题目、16PF测验,间隔3周后进行重复测验。结果修订了5道题后的L量表比原量表平均分降低约1分(6.3±2.8降至5.2±2.4;6.5±3.0降至5.6±2.9;P<0.01),修订前L量表与修订后L量表得分均与16PF量表24项因子分中的18项具有显著相关性(P<0.05),上述18项相关值两组间差异无显著性(P>0.05),修订后量表两次测验间相关系数(重测信度)为0.85(P<0.01),优于原量表的0.76(P<0.01);修订后量表两次测验的内部一致性信度(KR-20)分别为0.73、0.76(P<0.01),优于原量表的0.63、0.65(P<0.01)。结论本次修订后的L量表信效度好于原量表,更适宜用于新兵心理筛查。

Objective To revise the subscale L of Minnesota Multiphasic Personality Inventory（MMPI）,in order to improve its reliability and validity. Methods A total of 1290 recruits were measured with MMPI by DXC-6 machine of psychometrics. Another 470 recruits were measured with the revised scale and 16 PF,and undergone repeated test after three weeks. Results After revised five questions,the average score of L scale in two groups was about 1 point lower than original scale（6. 3 ± 2. 8 vs 5. 2 ± 2. 4,6. 5 ± 3. 0 vs 5. 6 ± 2. 9,P〈0. 01）;both the score of L scale before and afer revision were correlated with 18 factors in 16PF;there was no difference in the above 18 correlation values between two groups（P〈0. 05）. The coefficient of correlation of revised scale between twice test（Test-retest reliability） was 0. 85（P〉0. 01）,which was superior to the original scale（0. 76,P〈0. 01）. The internal reliability of revised scale between twice test was 0. 73 and 0. 75 respectively（P〈0. 01）,which was superior to the original one（0. 63 and 0. 65,P〈0. 01）. Conclusion The reliability and validity of revised L scale are better than the original one,which is more suitable for mental health screening of recruits.