Full text PDF

Imagine we want to understand whether musicians are more empathic than non-musicians. We recruit two groups of individuals, musicians and nonmusicians, and administer empathy tests. To compare the empathy scores we could use a two-sample t-test, or, equivalently, a one-way ANOVA. However, before we do so, textbooks advise us we have to investigate whether the two groups are balanced when it comes to other variables, such as age and sex. Why is that important? Students may feel that this is merely a bureaucratic step, a formality of sorts. In this teaching vignette, we explore how group imbalances may affect the conclusions we draw from the dataset. We provide sample datasets as well as an online interactive interface that allows the students to explore when group imbalances are problematic and when they are not. We explore ways to test group imbalances and ways to deal with them, such as including the confounding variables into the analysis model. Overall, the learning activity aims to help sensitize students to the problem of unaccounted group imbalances.

Pages © TQMP;
Website last modified: 2024-11-30.
Template last modified: 2022-03-04 18h27.
Page consulted on .
Be informed of the upcoming issues with RSS feed: RSS