Measurement reactivity in smartphone-based dietary ecological momentary assessment

Dietary Ecological Momentary Assessment (EMA) is popular for capturing real-time eating behaviors. Yet research participation effects such as measurement reactivity, i.e. changes to cognitions or behaviors due to being measured in a study, may impair data quality. This preregistered study investigated whether measurement reactivity occurred an eight-day event-based smartphone-based dietary assessment. It furthermore tested which aspects of eating behavior were most susceptible, and the moderating roles of active information processing (AIP), previous experience with diet-tracking apps, and the intention to change eating behavior were explored. Healthy behaviors were expected to decline and unhealthy behaviors to increase during the study period. Eighty-two participants (85.4% women; age M = 22.88, SD = 3.91) recorded all meals and snacks via photos and text descriptions. A total of 2,414 meals were recorded. Results indicated measurement reactivity for fruit and vegetable consumption (i.e., decline over the study period), but not for snacks. Behaviors considered easier to change were not more prone to measurement reactivity. Also, previous app experience and intentions to change behavior did not modulate the effects. However, AIP exerted a moderating effect on change in fruit and vegetable intake. Thus, dietary EMA is not a neutral observational process but may serve as a light intervention that influences the intake of certain food groups. Researchers should account for measurement reactivity in study planning (e.g., by focusing on less reactive behaviors) or analytically (e.g., by discarding initial days of data or controlling for change across time).