Confounding Variable in Psychology (Examples + Definition)

There are 4 types of variables that are mostly focused on. These are dependent, independent, extraneous, and confounding variables. Confounding variables play a huge role in both dependent and independent variables. 

An external factor known as a confounding variable alters the relation between dependent and independent variables. The outcome of a study design is influenced by this unintentional factor. An additional element that was not considered is known as a confounding variable.

Confounding variables can be tricky as they have the ability to muddy study results. However, there is a way to try and stop that from happening. Here is what you need to know about confounding variables, including helpful examples.

All About Confounding Variables

Confounding variables are one type of extraneous variable. Confounding variables mainly have one answer that needs to be addressed. How does one know that the change that occurs in the dependent variable being observed is caused by the independent variable?

Confounding variables are generally defined as elements showing that the independent variable is not the only one influencing the dependent variable. Simply put, it is a factor that is related to both the independent and dependent variables but was excluded from your analysis.

It is referred to as a confounding variable. In a statistical model, a variable is confounded with other variables if they cannot be estimated separately from the data. If two variables in a statistical model can’t be evaluated independently from the data, they are said to be confounded.

Examples:

For instance, the study could be about memory recollection. Such as how many items a particular number of people can recall and how accurate that recollection is. Only when half of that group first underwent memory training while the other half did not. The participants were recruited and randomly placed in two different groups.

One group is trained in memory recollection studies, and the other is not. In this case, it would be clear that the independent variable affects the dependent variable as the training would make the first group recall more items and more accurately.

However, the confounding variable here could be age. Suppose the two training groups were not recruited according to a specific age, and their ages do not match. In that case, differences in the number of items recalled and the accuracy cannot be attributed to the training offered to the one group.

Age means different cognitive competencies, which leads to the vast differences in the study independent of the memory training.

The correlation need not be causal. In this example, it simply happens in that way. For instance, you might be looking at whether a certain treatment helps people recover from colds more quickly. 50 individuals with colds receive the treatment, and the other 50 more do not receive it, but all are tracked.

However, by coincidence, you administered the treatment to 30 individuals suffering from common colds and 20 individuals with more severe flu, compared to the control group’s 30 individuals with flu and 20 individuals with colds. You conclude that the medication is quite effective. Still, it could also be the case that the group it was administered to was already on the road to recovery.

How Do Confounding Variables Interfere In Studies?

Having any kind of confounding variables interfere in the study makes it quite challenging to determine and clearly isolate whether or not changes in the dependent variable we are studying were because of the independent variable or if there is another factor that somehow casts the same effect on our dependent variables, as the independent variable in our study would have.

How Can You Ensure Compounding Variables Don’t Ruin Studies?

Confounding variables are additional variables that we neglected to account for during the experimentation. Because they tend to raise variation and create bias, confounding variables can render our results meaningless. To avoid this happening, simply include a control variable in your study.

For instance, if you determine whether a lack of physical activity causes weight gain, age would be the confounding variable in the study. This is because it affects weight gain. Therefore, by including a control variable that is simply a fixed age in the study, you can reduce the impact of the confounding variables.

Observational studies are more problematic when confounding variables are present. The other half solution is to examine populations both qualitatively and quantitatively using all available metrics. This may make you aware of potential confounding factors.

Why Study Results Are Adjusted For Confounding Variables

Unsurprisingly, results that are observed from studies of dependent and independent variables are often adjusted for confounding factors such as age, sex, etc. So, what exactly does adjustment mean? For example, if there were to be a study conducted with dependent and independent variables, weight and height, you would be able to see a strong relation between the two.

However, generally speaking, it is widely known that gender typically influences height and weight. Therefore this would be a confounding factor. Therefore, the individuals should be split according to their sex. When this is done, the relation between the height and weight of both groups is still strong, but not like before.

Therefore, this means that the relationship is adjusted for sex. Simply put, the confounding variable, which is sex, is kept fixed. However, as we know, according to the definition of confounding variables, there is more than one factor to adjust for, as the definition states that it is any factor in relation to both the dependent and independent variables.

Therefore, it makes sense that in practice, studies adjust for many factors all in one go. Here is an example that considers other confounding variables to give you an idea of what that would be like. Suppose this study deals with dependent and independent variables of exercise and heart disease.

In that case, we are aware that a person’s age can influence how often they exercise and their risk of heart disease. Therefore, unless age is considered, it would interfere in understanding the relation between the two variables. Because of that interference, age is therefore known as a confounder. There are many confounders. Others are diet, smoking, etc.

What to keep in mind is that these factors may not directly affect the exercise and heart disease variables. Still, they are linked to both the variables we are interested in. Therefore, the goal would be to estimate the relation between the two variables of interest while keeping the confounding variables fixed.

How is that done? Say there is a group of individuals. Data must be collected on all of the factors that affect them. Afterward, a statistical method must be used. The suitable method for this is regression analysis. This method is important if you do not want confounding factors rendering your results meaningless.

This is because we can then estimate the relationship between two variables of interest, all while keeping any confounding variables fixed. This is known as adjustment. As much as adjustment can help with considering confounding factors, it is also inadequate. This is because there may be other confounding variables that we just don’t know about and, therefore, have not measured.

There may also be confounding factors that get measured incorrectly. For example, people can lie when asked about something they feel uncomfortable discussing. They may understate how much they smoke or not fully disclose what they eat. Although adjustment has its faults, it is still essential when dealing with confounding factors.

Are Confounding Variables The Same As Extraneous Variables?

One mustn’t confuse the extraneous variable with the confounding variable. To make it clear, an extraneous variable is usually any variable that relates to any of either the dependent or independent variables that are the primary focus that we simply did not take into account, as we should have when the study was designed.

This may seem tricky as the definitions are almost similar. The word is ‘and’ when it comes to extraneous and compound variables. First of all, all compound variables are extraneous variables, but not all extraneous variables are compounds.

Therefore, extraneous variables are any variables relating to either any of the independent or independent variables. Whereas confounding variables are any variables relating to any of the independent and dependent variables.

With confounding variables, you can also say that maybe something either than the variable we are interested in is what caused an effect on the dependent variable. Basically, could some other variable be an alternative explanation for the study’s findings?

Furthermore, the variable must relate to both the dependent and independent variables. That is, the dependent variable and at least one independent variable must change due to the confounding variable.

Example Of Extraneous And Confounding Variables In Action

The same study will show how the extraneous variables differ from the confounding variables.

Suppose there is a study on whether people with children stay longer in rented apartments or for short periods as opposed to those that don’t have children. One extraneous variable is the fatigue levels. Raising a kid isn’t a joke, and parents are bound to be more exhausted than those that do not have kids.

Another extraneous variable could be the age of the parents with children compared to tenants that do not have children. Typically, older people are the ones that have children compared to teenagers. When looking at the extraneous variables mentioned, they only relate to the independent variable.

When you take a look at what kind of extraneous variables could influence the dependent variable, then you would be looking at a job that requires relocation or the apartment building no longer being safe. As you can see, all the extraneous variables mentioned only relate to either the dependent or the independent variable, not both.

In contrast, confounding variables relate to both the dependent and independent variables. Let’s look at how this is possible using the same study. These are the same variables of interest, having children, how long they stay in an apartment, and the same extraneous variables.

For this part, you simply need to determine if any of the extraneous variables are also confounding variables. For example, fatigue levels are related to whether people have children, but is it also related to the dependent variable? Is it also related to how long someone will live in an apartment complex? The answer is no. tiredness is in no way related to how long someone will live in an apartment.

Therefore this is not a confounding variable as it does not relate to both the dependent and independent variables. Take relocating for a new job, for example. It is related to the dependent variable. However, is it also related to the independent variable? Not exactly. Relocating for a job will influence moving out.

Still, it doesn’t influence whether someone has children as they may have those children already. Therefore, this is just an extraneous variable. If you look at age, it’s clear that it is related to whether or not an individual has children. However, is it related to how long someone lives in an apartment? Absolutely.

Simply put, whether a family decides to stay in a complex may not have to do with the fact that they are parents. It could be because those parents have aged and have no desire to live in an apartment anymore.

It relates to both the dependent and independent variables, making it a confounding variable. Telling the difference between the extraneous and confounding variables may be tricky, but it gets better when you keep the definitions in mind and practice using various examples.

Conclusion

As you can see, extraneous and confounding variables are not the same, although they are similar. Knowing the difference will help you compile a useful list of factors for your studies. Ensuring that you are aware of confounding variables when conducting studies can go a long way in helping you get results that are not influenced to the extent that you are unable to determine which factor made the changes.

Theodore T.

Theodore is a professional psychology educator with over 10 years of experience creating educational content on the internet. PracticalPsychology started as a helpful collection of psychological articles to help other students, which has expanded to a Youtube channel with over 2,000,000 subscribers and an online website with 500+ posts.