Concurrent administration of ecSI with validated measures of eating behaviors using on-line and paper-pencil formats.
The on-line survey was completed by 370 participants; 462 completed the paper version.
Participants included 863 adults with 832 usable surveys from respondents (mean age 36.2 ± 13.4 years) without eating disorders, mostly female, white, educated, overweight, physically active, and food secure. Of those indicating intent to complete the on-line survey, 80.3 % did so; 54 % of mailed surveys were returned.
Eating and food behaviors compared among EC tertiles and between dichotomous EC categories; internal consistency of ecSI.
Analysis of variance, independent t tests, chi-square, factor analysis, logistic regression. Significance level was P < .05.
Mean ecSI score was 31.1 ± 7.5. ecSI included 4 subscales with internal reliability and content validity. Construct validity was supported by specific behavioral profiles for ecSI tertiles and ecSI dichotomized categories. Persons unsatisfied with weight were 54 % less likely to be EC; unit increase in the food like index was associated with nearly 3 times greater likelihood of being EC.
The ecSatter Inventory is a valid measure of EC and can be used for descriptive and outcome measurements.