Measuring cognitive load

The paper ‘Cognitive load predicts point-of-care ultrasound simulator performance’ by Aldekhyl, Cavalcanti, and Naismith, in this issue of Perspectives on Medical Education [1], is an important paper that adds to work on cognitive load theory and medical education [2–4]. The implications of the findings of this paper extend substantially beyond the confines of medical practice that is the focus of the work. In this commentary, I will discuss issues associated with obtaining measures of cognitive load independently of content task performance during instruction. I will begin with a brief history of attempts to provide independent measures of cognitive load. In the 1980s, cognitive load was used as a theoretical construct to explain experimental results with very little attempt to directly measure load [5]. The theory was used to predict differential learning using particular instructional designs. Randomized controlled trials were run to test the predictions and if the hypothesized results were obtained they were attributed to cognitive load factors. The distinction between extraneous and intrinsic cognitive load had not been specified but the results were due to what was called and continues to be called extraneous cognitive load. Cognitive load was an assumed rather than a measured construct. At that time, the only attempt to provide an independent indicator of load was to use computational models [6] with quantitative differences between models used as cognitive load proxies. The first rating scale measure of cognitive load was introduced in the early 1990s by Fred Paas [7]. The Paas scale continues to be the most popular measure of cognitive load and was used by Aldekhyl et al. to validate alternative measures of load. It is very easy to use and requires no more than a minute or so of a participant’s time. Used primarily to measure extraneous cognitive load it has repeatedly indicated that instructional designs hypothesized to decrease