There is a slight confusion between the terms “control” and “comparison” groups. However, both are often used interchangeably. This article aims to provide an overview of these terms.
The terms “control” and “comparison” are usually used in an experimental study to determine if there is a significant effect of an intervention on the experiment group without bias. Without control and comparison groups, most experimental studies are often considered insufficient, which we cannot overclaim the results—the biggest weakness in a pre-experimental study.
There are two perspectives to understand the difference.
A Control Group is a group of individuals who received no treatment or a standard treatment. In contrast, a Comparison Group is a group of individuals who received no treatment or an alternative treatment. You may see the similarity between both. If we do not have any treatment for another group, we can call it a control or comparison group.
But I want you to underline the term “standard” and “alternative” treatment to differentiate the two terms. Let me give you examples.
Let’s say I conduct an experimental study in a public health center in persons with hypertension at home. I want to test whether my intervention (music therapy) is effective in controlling their blood pressure. To avoid bias, I compare the intervention with a control group who receives a standard treatment from the public health center, which is received from a medical doctor or a nurse. Let’s say they only get a consultation from a nurse as the standard treatment. If you have any other standard treatment, you can add.
A similar example, I want to determine the effect of music therapy on controlling blood pressure. But I use a comparison group of my study who receive an alternative treatment (spiritual mindfulness intervention), which is not a standard treatment from a public health center.
It is called a control group if a random assignment of the samples is adequately done. In contrast, if a random assignment is not conducted correctly, we can call it a comparison group. A random assignment means we blindly assign the sample to an experimental group and another group.
So, based on the two perspectives, which one do you want to use? If you use a standard treatment, but a random assignment is not done properly, can it still be called a control group? Or, if you use an alternative treatment but a random assignment is adequately done, can it be called a control group as well? Definitely, you cannot combine both perspectives. You just need to hold one.
For me, if I use a standard treatment and random assignment is done, then I call it a control group with a true experiment. If random assignment is not done, I can still call it a control group, but with a quasi-experiment. However, you need a rationale for why the random assignment cannot be done or undone. And other criteria of the true experiment should be met.
Similar to a comparison group. If I use an alternative treatment and random assignment is properly done, I still call it a comparison group with a true experiment. And, if not, I call it a comparison group with a quasi-experiment.
So, basically, I use the first perspective with a different type of experiment.
What’s your point of view about this?
Cite this as: Gunawan, J. (2021). Control group vs. comparison group. Retrieved from https://jokogunawan.com/blog/2021/07/21/research/833/
This is an open access article distributed under Creative Commons Attribution 4.0 International (CC BY 4.0)