We need to talk about validity – A commentary on “Six solutions for more reliable infant research” from the viewpoint of an early executive functions researcher

Abstract

In their methodological article, “Six solutions for more reliable infant research”, Byers-Heinlein, Bergmann and Savalei (2021) present compelling arguments for why developmental researchers should report and consider measures of reliability more frequently in their work. They also provide useful guidance on solutions to this “reliability crisis”. In this commentary, I highlight a further methodological aspect that I think is key to successful and robust infancy research, that of construct validity. I also discuss recent reliability data from my own research on early executive function development, analyses which were directly inspired by the target article. Highlights: Considering measurement reliability and effect sizes is important for robust infant research and for optimising infant tasks to measure group-level effects or individual differences. Construct validity – making sure that we measure what we think we are measuring – is also important. A robust effect at the group-level may not always restrict reliability – it depends on the amount of true variation between infants.