By Dr. Linh Phung
I have read research, reviewed research, and done research. I’ve also guided ESL students to conduct their small research projects. I’ve never taught any research method course though, so I don't claim to be an expert in this. However, the controversy regarding a test item in an important English exam in Vietnam prompted me to revisit some materials I used in the past few years to help students to listen to academic lectures in the Coursera course “The Science of Wellbeing” with some understanding of the research presented in those lectures.
In the English exam under scrutiny here, students were asked to identify one erroneous part among the four underlined options in this sentence: Their (A)pioneering research showed that the learning motivation of the two groups of learners was quite (B)distinctive from each other, and the (C)comparative group whose learning motivation was stronger performed better than the (D)control group.
The item seems to assess students’ understanding of highly technical concepts in experimental research: comparison and control group. So what’s an experiment? What are experimental, control, and comparison groups?
Let me start by summarizing a research study explained in a lecture given in The Science of Wellbeing and then explain these concepts in the social research involving human participants.
A research study conducted by Sonja Lyubomirsky and her colleagues aimed to investigate whether doing random acts of kindness boosted people’s subjective happiness measured by participants’ ratings of their own happiness. In the study, they asked one group of people to perform five random acts of kindness in one day during the following week, one group to perform five random acts of kindness across the whole week, and one group to just write about events of their day during the following week.
As you can see, in this research study, the first two groups are experimental groups because they had an intervention of interest (doing random acts of kindness). The third group is the control group because they didn’t get the intervention. This control group can be considered the comparison group for the two experimental groups. In the test item above, comparing the comparison group and the control group does not make sense because the control group is also known as the comparison group in experimental research.
What is an experiment? It’s understood as a research design that collects data in controlled conditions (usually in a laboratory) to test hypotheses. An experiment has some key features:
1. Independent and dependent variables (acts of kindness and subjective happiness in the example above)
2. Pretesting and post-testing (subjective happiness measured at two times in the example above)
3. Experimental and control groups. Participants (say 200) should be randomly assigned to these groups.
In educational research, participants are often recruited from existing classes and are not randomly assigned, so this is often considered quasi-experimental. In this case, there’s not really a control group because groups occur in the “natural setting.” Researchers may then have a comparison group instead. The features above still apply except random assignment.
Do you know the findings of the above-described research? The happiness of the control group went down. The subjective happiness of the group that did five acts of kindness across the whole week didn’t change much. The biggest increase came from the group that did five acts of kindness on one day. The explanation was that doing five acts of kindness in a single day forced participants to think of them all at once when they rated their happiness, which boosted their rating.