[Factor analysis of Cornell Medical Index and Medical Data Index (author's transl)].

Structures of Cornell Medical Index and Medical Data Index were examined by factor analysis. Both of the translated questionnaires were self-administered by six hundred male employees. Questions of low response rate less than 5 and 6% were excluded, and 117 and 70 questions were analysed finally by the method of varimax rotation for CMI and MDI respectively. Ten factors were extracted. Cummulated contribution ratio of them amounted to 72.0 and 84.1% for CMI and MDI respectively. The first two factors were common for the both questionaires and had pronounced contribution ratio amounting to 36.7 and 29.3% for CMI and MDI respectively. The first factors was composed of questions related to autonomic inbalance, gastrointesinal symptom and fatigue, so it was named a tendency of many physical complaints. This factor was considered to be characteristic and representative one in such general health questionnaires, which was closely related to response set of the respondent. The second factor of the questionnaires was named mental instability. The two major factors would correspond to the first and the second component of the Subjective Fatigue Complaint Test (1970) by Japan Association of Industrial Health. Not also CMI but MDI were found to have a poor scalability except for the two major factors.