Education research in the 1970s, like other related areas, was dominated by quantitative work¹ during an era for which social sciences sought to draw upon the successful scientific approach typically used in the physical sciences (in particular) to investigate teaching and learning.¹’² So if we felt a cohort of students did not understand some con¬cept, we tried to find out whether or not a different teach¬ing approach could fix their misconceptions.³ But how to do this? Well, drawing on a scientific approach, we would divide the class or classes up, teach one cohort the same way we always had, and the other cohort in our new way, and evaluate any differences in conceptual understanding using, e.g. a standardized topic test. Differences would be examined for statistical significance of evidence that our new approach to teaching had worked. And this is the way much research was done at the time. Control of variables, randomized sampling, and so on, were all embedded in such an approach to educational research.