Fig 1.
Result of confirmatory factor analysis of the eating behavior pattern questionnaire for participants.
Model fit indices: χ2/df: 1.28; RMSEA: 0.034; CFI: 0.954; TLI: 0.936; SRMR: 0.047. ɛ1 to ɛ9 represent error variances of variables. Small numbers at the corner of the rectangular objects are standardized intercepts (the predicted value of the variables when all the predictors are zero).
Table 1.
Cronbach’s alpha and scoring for eating behavior patterns questionnaire.
Table 2.
Selected characteristics of the participants by BMI groups (N = 250).
Table 3.
Dietary intake and eating behavior of participants by BMI groups (N = 250).
Table 4.
Correlation among variables used in SEM modeling.
Fig 2.
Path diagram of the studied model in two BMI categories, a: LBMI group (BMI<21.75) (N = 125) and b: HBMI group (21.75≤ BMI) (N = 125).
Note: CHO: carbohydrates. Standardized coefficients are provided. Bold paths are significant (*: p<0.05; **: p<0.01 and ***: p<0.001). Potential confounding variables are illustrated in italic letters. Rectangles and ovals are used to represent observed vs latent variables and ɛ1 to ɛ7 represent error variances of variables. Model fit indices: χ2/df: 1.38; RMSEA: 0.055; CFI: 0.964; GFI: 0.888; TLI: 0.943; SRMR: 0.066.
Table 5.
Direct, indirect and total effects of evaluated models.
Table 6.
Tests for group invariance (low-to-normal vs. normal-to-high BMI groups) of parameters used in SEM modeling.