Favouring more rigour when investigating human eating behaviour is like supporting motherhood and apple pie: A response to Robinson, Bevelander, Field, and Jones (2018).

In a 1987 paper, addressing questions about factors that influence the initiation, maintenance, and termination of food intake, we wrote, "development of systematic procedures to measure eating behaviour is essential if descriptive and inferential statistics are to be applied to answering such questions, giving them power and replicability" (Hetherington & Rolls, 1987 page 77). Therefore, as longstanding advocates of rigorous procedures in laboratory-based investigations of food intake, we welcome Robinson et al.'s (2018) clear recommendations for laboratory studies. However, this is akin to voting for "motherhood and apple pie", and few would argue against deployment of improved procedures for these studies. What then can we contribute to the debate in order to refine the recommendations made or add to them? Our most important message for researchers is that the central hypothesis or main research question will determine the most appropriate methods for any study. If a laboratory-based study is planned, then there are basic methodological questions that must be answered before proceeding to a final protocol. While such guidelines are needed to ensure basic methodological rigour, these should not be so prescriptive as to inhibit creativity. Here we provide several thoughts on how to advance studies of ingestive behaviour, including the need to apply appropriate controls, encouragement to move beyond convenience samples, and to remember the value of exploratory, observational, and naturalistic studies to complement laboratory-based studies.