Quantifying portion sizes for selected snack foods and beverages in 24-hour dietary recalls

The purpose of this study was to determine individual recall accuracy for selected snack foods and beverages using various measurement aids and to investigate the cognitive strategies associated with aid selection for snacks and beverages in a 24-hour recall. Individuals (n=202) estimated the amounts of snack foods and beverages consumed the previous day while watching a 30-minute videotaped television program. Individuals were provided ad libitum amounts of snacks (potato chips, tortilla chips, or popcorn alone or with salsa or mixed nuts) in various sizes bags and or bowls. Beverages were provided with snacks in either standard 20 oz bottles or 8, 10, or 16 oz glasses. During consumption estimation, a variety of aids including aids used in previous National Health and Nutrition Examination Studies (NHANES) were provided. Average recall errors, by weight, for each of the 10 snack/container groups were within 16 g with percentage recall errors for these groups ranging from -36.8 to +38.7%. Bowls were selected to estimate consumption at least eight times more than any other aid and provided an accurate average recall estimation. Beverage consumption was consistently overestimated. Respondent confidence in estimation accuracy did not correlate with actual error for either snacks or beverages. Results suggest that average recall errors for groups consuming snacks are likely to be within one-half of a standard portion size and that bowls provide a means by which individuals can accurately estimate snack consumption in a 24-hour recall. (C) 2000 Elsevier Science Inc.