| dc.description.abstract | Video observation is gaining popularity as a data collection technique to study eating behavior. Traditional approaches to data analysis of video observations are based on static counts reporting the number of times an action occurs, disregarding the dynamic aspects captured in the video. In this paper, we investigate French consumers’ bread consumption patterns during a restaurant meal. We propose a novel approach, the Temporal Dominance of Behavior (TDB), to visualise the dynamic information contained in video material in terms of frequency, duration, sequence and simultaneity of actions based on standard data analysis principles from temporal methods such as Temporal Dominance of Sensations (TDS) and Temporal Check-All-That-Apply (T-CATA). TDB allows reporting hours of videos across several subjects in a single graphical output, providing an efficient summary overview suitable for results interpretation and communication. Regarding bread consumption patterns during a restaurant meal in French consumers, three groups of consumers are drawn: the No-breaders (43%), the Bread-as-a-tool group (48%) and the Bread lovers (9%). Eating behavior varies in the consumed quantity of bread as well as in consumption dynamics throughout the meal, and it is related to both the type of dish that is consumed and to consumer characteristics. The Temporal Observation Curves approach provides a graphical synthesis of the dynamic information. We discuss the added-value of the TDB method compared to single-point analysis, provide recommendations for future developments and suggest potential applications in the consumer and food domain. | |
