抄録
食行動異常傾向測定尺度の信頼性・妥当性を検討し,カットオフポイントを設定した.女子学生429名(平均年齢20.29±3.67歳)・摂食障害患者27名(平均年齢24.26±6.13歳)を解析対象とした.因子分析の結果,第1因子:食物摂取コントロール不能感,第2因子:不適応的食物排出行動,第3因子:食物摂取コントロールが抽出され,信頼性係数は十分であった(α=0.78〜0.89).EAT-26およびEDI下位尺度で対象者を分類し,各因子得点について一要因分散分析を行った結果,EAT-26高群は中群・低群よりも全因子得点が高く,中群は低群よりも高かった.EDI過食尺度高群は中群・低群よりも第3因子得点が高く,中群は低群よりも第2因子・第3因子得点が高かった.EDIやせ願望高群は低群よりも第1因子・第2因子得点が高かった.以上から尺度の基準関連妥当性が確認された.カットオフポイントは第1因子:16点,第2因子:2点,第3因子:7点であった.
Objectives: Recently, eating disorders and abnormal eating behaviors have been increasing among female college students. The purpose of this study was to develop the Abnormal Eating Behavior Scale (AEBS) and to examine the reliability and the validity of this scale. Subjects: The subjects of this study were 593 students and 31 patients with eating disorder. In these subjects, 429 (the mean age was 20.29±3.67) students and 27 patients with eating disorder (the mean age was 24.26±6.13) answered to these scales completely. Method: Firstly, we examined the structure of the AEBS by the factor analysis, and we also examined the reliability of this scale. Secondly, we tested the validity of this scale. The subjects were classified into 3 groups such as the high, the moderate and the low group according to each score of the EAT-26 and the EDI. The one-way analysis of variance (ANOVA) of 3 groups was performed on the factors of this scale. Thirdly, the cut-off point of this scale for discriminating eating disorder patients from normal subjects was examined by the receiver operating characteristic analysis. Results: The results of the factor analysis of 29 items for measuring an abnormal eating behavior were three factors (19 items) such as Factor 1: uncontrollability of excessive food intake, Factor 2: purging behavior and Factor 3: control of food intake. Cronbach's alpha coefficients demonstrated highly sufficient. The results of the examination for the EAT-26 indicated that all factor scores were higher in the high group than the moderate and low groups, and that all factors showed higher scores in the moderate group than the low group. In addition the results of the examination for the Binge-eating (EDI) indicated that Factor 3 score were higher in the high group than the moderate and low groups, and that Factor 2 and Factor 3 showed higher scores in the moderate group than the low group, while the examination for the Desire to be slim (EDI) indicated that Factor 2 and Factor 3 scores showed higher in the high group than the moderate and low groups. Moreover, the results of the receiver operating characteristic analysis were indicated the following cut-off points: Factor1: point of 16 (Sensitivity: 77.78%; 1-specificity: 83.46%), Factor2: point of 2 (Sensitivity 81.48%; 1-specificity: 92.91%), Factor 3: point of 7 (Sensitivity: 74.07%; 1-specificity: 70.08%). Conclusions: These results showed that the reliability and construct validity for this scale were acceptable. And, we can also say that this scale has criterion-referenced validity. Therefore, this scale may possibly be used to screen female college student for eating disorder.
