Relationship of Confidence Level and Perception of Healthiness of Food to Accuracy of Reported Dietary Intake

Abstract

The accuracy with which respondents reported dietary intake when using two different sets of measuring aids was investigated. Participants (n=120) were served roast beef, macaroni and cheese, green beans, french fries, salad, corn bread, cake, ice cream, beverage and condiments. Amounts eaten were calculated. Respondents were assigned to an interview group that used either 2- or 3-dimensional (2D or 3D) aids. All were interviewed the following day using the three pass recall method. The respondents’ confidence level with the amount they reported having eaten and their perception of the healthiness of each food (l=very healthy or confident to 5=not at all healthy or confident) was recorded. Data were analyzed using SAS-PC. In this study, persons more frequently over reported than under reported their intake, especially when using 2D aids. On average, all groups of respondents over reported the amount of cake, salad dressing, and water by more than 20%. 2D groups over reported consumption of green beans, ice cream, macaroni and cheese, and iced tea by more than 20%, but amounts were more accurate with the 3D aids for macaroni and cheese, ice cream, and tea. Amount of beef, salad, and corn bread eaten was reported within 20% accuracy with either set of aids. Respondents’ felt more confident in the amounts reported consumed when using the 3D aids. Although 3D aids tended to produce slightly more accurate estimation than 2D aids, differences were not significant. Mean confidence scores were high (1.24-2.06); however confidence in reported intakes did not correlate with the accuracy of the reporting for any of the foods. Mean healthiness scores ranged from 1.08 for water to 3.85 for margarine. Respondents did not significantly under report their intake of “less healthy” foods or over report the intake of “healthy” foods.