A group method for obtaining dietary recalls of children

A self-administered workbook for dietary assessment in small groups was compared with 24-hour dietary recall interviews. Forty-seven 10- to 18-year-olds, in groups of 5 to 8, recorded consumptions during the previous 24 hours and responded to individual recall interviews. Photographs and names of foods served as probes. Standardized food models were used to estimate portions. Nutritionists calculated gram amounts of foods and beverages consumed and analyzed diets for nutrient intakes. Significantly higher niacin (17.4 mg vs. 15.8 mg, p less than .05) and slightly higher protein, animal protein, carbohydrate, and cholesterol (N.S.) were estimated by the group workbook method. Correlations for individual dietary components ranged from 0.69 to 0.93. Dinner accounted for most differences, with statistically greater mean intakes of nine dietary components shown by the workbook method. Thirty-four dinner foods (8%) were not listed in common by the two methods. Sixty-six foods (15%) listed by the two methods had differing gram amounts. Improvements should include increased parental verification of the evening meal, additional training with food models, and periodic rest sessions. The workbook method may expand the time frame studied per individual and increase the number of eating periods examined. Replicate examinations per individual provide measurement of intra-individual variation. This method has applications in studies for which large samples or repeated measures are required.