Correlation Is Not Causation: A Lesson in Evidence and Empathy

Yesterday was my first in-person class after a year of doing predominantly online classes. Partly because of the materials I am developing for a course, and the other part because I was mostly ‘away’ due to my responsibilities as an administrator (which is a whole story for another day).

So to celebrate my own ‘return’ to teaching, I prepared an activity that is very much me–a murder mystery-themed activity to introduce my undergraduate students to the research process and the kind of thinking they need to bring in class. In the scenario I set, no one was really murdered, but people were ‘suffering’. A company has observed a 15% decline in productivity, absenteeism has increased, deadlines are being missed, and several employees have resigned. It was a well-being case; obviously, because of my personal research interest, and also because I wanted to get an idea of the kind of ‘future managers’ my students will be.

Students were asked to play the role of ‘detectives’ and identify the possible root cause of this problem. I prepared 10 pieces of evidence that they need to evaluate. Since this is a research methods class, these pieces of evidence were a combination of quantitative and qualitative data. But more importantly, not all of them are useful, and 3 of them were red herrings–carefully designed to mislead them. To my satisfaction, I was successful in doing so.

The majority of the groups said that the root cause was the weak economy, as evident in the newspaper article reporting that the economy is currently facing market volatility and that companies are advised to implement cost-cutting measures. Because of this, the company had to downsize, leading to workload issues, as corroborated by evidence such as the employee survey data where employees report being stressed at work; interview transcripts where long work hours are discussed; and the workload analysis that shows how workload and overtime hours increased after downsizing. My students were making good arguments until I asked them, “Was there a piece of evidence that actually suggests that the downsizing was a response to the economic pressure?”

There were two (2) reasons why I asked: first, because I wanted the students to go beyond the three (3) pieces of evidence they were focusing on and strengthen their argument using the rest of the clues, but second is that as a teacher, I wanted to see whether I myself missed something and to do a personal bias check.
One student answered: “Miss, the economy was down and then the company downsized as a cost-cutting measure; it is kind of implied and common sense will dictate that the two events were related”. I loved this response because I was able to remind them of something they probably learned in the past but forgot: that correlation does not mean causation. Yes, the two events seem related but there is no concrete evidence that can back that up. Besides, a closer look at the pieces of evidence will show that the downsizing happened way before the economic news came about.

On another note, I could not miss out on a teachable moment that this conclusion my students came to. Let’s recap–weak economy, company needed to tighten its belt and downsize, leading to workload issues as employees need to do more. In short, well-being is compromised because the company had to cut costs. So after our discussion on the process and how it mirrors the research process, I pivoted into a discussion of the managerial issue. I asked my students whether downsizing is the only answer to cost-cutting. And should they be the manager, what other steps could they take that will not be at the expense of the well-being of the employees. I told them that I did not just want them to become capable managers in the future, but I wanted them to remember to be a humanistic manager, one with a people-first mindset.

In summary, I hope my students did not just have fun with this activity but also learned the following:

• Assumptions without evidence can lead to flawed conclusions.
• Not all data is useful—researchers must learn to filter what truly supports their analysis.
• Stronger conclusions come from triangulation by combining quantitative, qualitative, and other evidence.
• Management decisions inevitably shape employee well-being. Always choose kindness.

Note: Should you be interested in doing this activity, I am open to sharing my materials, including the Instructor’s Guide I prepared as a companion. Please email me at jesswhoteaches@gmail.com