Year: 2015
Type: dissertation
Abstract: Among available dietary assessment tools the smartphone stands out in up-to-date measurement but requires further development. This study used data from Renner et al.’s (2015) SmartFood study to test the reliability of the smartphone based visual food record Renner et al. (2015) developed for Germany. In the SmartFood study participants photographed their diet using a smartphone within everyday living conditions over eight days. To determine the coding system’s reliability, two coders independently coded all 269 digital food images from 10 participants (six males, four females; age 30–74 years). Renner et al.’s (2015) coding system, comprising a food atlas (digital booklet with food images), and the nutritional software OptiDiet© were used to analyze the visual food record. As intercoder reliability index Krippendorff’s alpha was calculated to measure coder agreement. Across all participants’ meals, reliability was medium to high with α = .71, SD = 0.17, 95% CI [.59, .84] for food weight, α = .76, SD = 0.11, 95% CI [.69, .84] for food labeling, and α = .67, SD = 0.08, 95% CI [.59, .72] for food nutrients. Reliability varied in dependence of participant and nutrient type but was not significantly affected by the coders. The results support the coding system as a promising tool for analyzing smartphone based visual food records. To further develop the coding system, the food atlas should be extended with more images and the coding guidelines revised. Finally, a validation study is required.
Source: KOPS (University of Konstanz)
Author: Sabine Arnold
Language: English
Topic: Nutritional Studies and Diet
Field: Medicine
Domain: Health Sciences
SDG: Zero hunger
Open Access status: green
OpenAlex ID: W2898113322