Sales Toll Free No: 1-855-666-7446

Confounding Variables


In mathematics, the students study about two types of basic quantities - the variables and the constant. A constant does have fixed value which can not be changed. On the contrary, a variable refers to "able to vary".
Variable is the quantity which can be changed. It is represented by any letter, such as: x, y, z, a, b, s, t etc. A variable has a value assigned in it which depends upon the circumstances.Lets talk more about variables. When the variables are classified, they can be broadly divided into two - independent and dependent variables. As the name suggests independent variables are those which do not need another variables to find its values ; these are basically independent in nature. While, dependent variables are those which are dependent on some independent variable. With any change in the value of independent variable, the value of dependent variable eventually changes.

There are other classifications of variables. Confounding variable is a type of variables that has its own importance. It is a variable that is correlated with the independent variable. Confounding variables are one of the most frequently-used variable, especially in statistics and researches. We are going to focus on confounding variables in this article. Let us go ahead and learn about confounding variables in detail.


Back to Top
A confounding variable is a statistical variable that is correlated with both of the independent and dependent variables. This correlation may be direct or inverse. A confounding variable is defined as an extraneous variable (a variable which is usually not a direct focus of study) which is statistically correlated with the given independent variable. We can also say that if the value of independent variable varies, the confounding variable also changes along with that.

The purpose of a research or an experiment is to provide a situation in which the difference between the conditions is actually the difference in the independent variable. If there exists some other variable which is changing with the change in independent variable, then this variable will be confounding variable. The specific definitions of confounding variable may vary in words. But, this variable should necessarily fit the following four criteria.
Let us suppose that there is a variable of interest denoted by "X". The confounding variable is represented by "C" and the outcome of interest be denoted by "O", then :
1) C is directly or inversely associated with O.

2) C is also associated with O just independent of X.

3) C is inversely or directly associated with X.

4) C does not in the direct pathway from X to O; i.e C is actually not a direct consequence of X.

Extraneous And Confounding Variables

Back to Top
Extraneous variables: is a variable that is said to be a undesirable variable which poses an impact on relationship between the variables that are examined. In other words, we can say that these are the variables that influence the result of an experiment and though these are not the variables of interest.

Extraneous variables add up an error to the research or experiment, therefore they are termed as undesirable variables. In a research, the main goal of researcher is to control or decrease the effect of the extraneous variables as much as possible. These are the variables that are not a direct part of the manipulation.

We can say that these are the factors that we have not controlled. Extraneous variables affect the result. But usually they do not create any bias in the result. More specifically, a variable that the researcher is not studying intentionally and is able to threaten the validity of the result is known as an extraneous variable.
Confounding and extraneous variables are closely related to each other. When an extraneous variable changes along with the studied variables systematically, it is known as a confounding variable.The researcher's goal is to try to control the extraneous variables such that they do not become confounding variables. Extraneous variables do not create bias but confounding variables produce bias in the experiment.


Back to Top
We come across with various practical examples of confounding variables in statistics. Some of them are listed below:
1) Confounding variables are used in risk assessment factors; such as gender, age and educational status. These usually effect on the health status and therefore are supposed to be controlled.

2) Another example is about the study of smoking or tobacco use on human health. These habits are correlated with the risks of heart, lung and various other diseases.

3) Confounding variables are also seen in the review of occupational risk assessments; for example in safety in coal mining.

4) According to a concrete example, a study about the relation between order of  birth ( such as - 1st child, 2nd child and so on) and the possibility of Down Syndrome. In which the age of the parents may be a confounding variable; since the higher the maternal age the more chances of Down Syndrome in the child. So, Down syndrome is directly related to the maternal age.

5) In an example of the study consumption of ice cream for a given period. The confounding variables may be climate condition such as temperature, children's vacations etc; which directed correlate to ice cream consumption.