The Between-Not-Within fallacy coined and exemplified: why studying a within-person uniform measurement bias is driven by between-person differences in intensive longitudinal data

Psychological research is currently increasingly conducting studies that use intensive longitudinal data, such as ecological momentary assessment, ambulatory assessment, daily diary studies and experience sampling methods. What these study designs have in common is that we wish to go beyond the long-standing research tradition of studying phenomena between-person thus emphasizing interindividual differences. Furthermore, these lines of research strive and promise that by going beyond cross-sectional data and the study of interindividual differences, we can gain deeper insights into the underlying within-person processes. I agree that research on within-person processes is insightful, and I agree with the potential of studies using intensive longitudinal data. At the same time, I argue in this General Article that there is a real danger that we overgeneralize effects driven by the between-person level to the within-person level. I coin this overgeneralization the Between-Not-Within fallacy. I demonstrate this fallacy using two examples from the field of measurement. More specifically, I take two examples of repeated measurements where there is a uniform measurement bias present within-person. In both examples, I show a) how they are inconsistent with our common modeling of latent variables and b) how they oversimplify uniform measurement bias in the face of true change. Both lines of reasoning lead to the same conclusion: even though we may believe we are delving deep into within-person processes, we are still driven by the between-person level in our research.