How Not to Build a House of Cards

Four recent papers on the measurement of avoidant/restrictive food intake disorder (ARFID) advanced the science in this area. Here, I provide relevant background information to aid readers in appreciating the contributions of these papers and context for considering the difficult but critical task of establishing that a measurement instrument does what we think it does. Each of the papers challenges us to question what we already know, what we are comfortable trusting, and what perspective is most useful to take when evaluating a new measure. For instance, did we learn that caregivers and their children disagree about what ARFID symptoms they experience, or did we learn that we assess ARFID differently between caregivers and their children? The implications of the answers to these questions are important. I end audaciously with a short list of recommendations that includes immensely useful resources, the inescapable necessity of theory to guide measure evaluation, how to describe the evaluation of measures and the evidence obtained, and the need to elevate the prominence of intraclass correlation coefficients and discriminative validity in the evaluation of measures.