The Japanese version of Home & Östberg's Morningness-Eveningness Questionnaire (MEQ) was examined of the construct validity, administered to 143 women aged 18 to 23, averaged 19.0 years old, who were college students of nursing and medical technology. 1) The subjects, 143 women, were divided into five types of Morningness-Eveningness by the criteria of Home & Ostberg (1976). One woman was definitely morning type, 14 moderately morning type, 101 neither type, 24 moderately evening type, and 3 definitely evening type women. 2) By the G-P analysis of 35 women of the highest scores (G-group) and 35 women of the lowest scores (P-group), it was found that the MEQ's two items of number 12 and 16 failed to identify G-group from P-group. 3) Spearman's correlation coefficients (rs) obtained between total scores and scores of each item revealed that the three items in the MEQ, number 10, 12 and 16 were not significant. 4) Seven factors were extracted by principal factor analysis, having eigen value of more than 1.00. Only the first factor was accountable for morningnesseveningness with a major contribution rate of 24.5%. The MEQ's item number of 10, 12 and 16 had the lowest factor loading in the first factor. 5) Cronbach's alpha coefficient was 0.78 calculated with the MEQ now in use consisted of 19 items, and increased to 0.81 with the MEQ consisted of 16 items eliminated of the three items, number 10, 12 and 16. 6) These results suggest that the MEQ eliminated of the three items, number 10, 12 and 16, is more valid in the concept to identify morningness-eveningness, at least in our subjects, women college students.
View full abstract