research methodology


An SNJ Associates Series: Research Methodology

Issue 2: Variables in Quantitative Research Methodology

In the second issue of our series titled: ‘Research Methodology’, we are taking a look at the definition of a variable in quantitative research along with different types of variables common to quantitative research methodology. We define both the independent and dependent variables in research along with discussing why a basic understanding of other extraneous variables in a research study are important.

We began the research methodology series with issue 1 and covered an overview of quantitative versus qualitative methodology. If you are not familiar with the basic differences between quantitative and qualitative methodology approaches, click on Issue 1: Quantitative versus Qualitative Research Methodology. Come back here when you’re done to continue.

What you are truly learning in this series is the language associated with research and various terms in epidemiology and sociological research –Whoa that was a mouthful! Please keep going and we promise it will get easier. Today we happen to be talking about ‘variables’. You’re learning the term in research language but in reality, many of the variables you’ll see in health research are already familiar to you. In our media and other key communication avenues, you have seen many variables common in health and well-being research everyday such as weight, height, number of prescriptions, diagnostic scoring (such as pain scales), number of days of lost work, number of admissions to the hospital, number of visits to the doctor just to name a few. All of these are quantitative or numeric variables.


What is a Variable in Quantitative Research Methodology?

Simply, a variable is a specific characteristic a researcher would like to study in their research project.

Variables can range from straightforward characteristics that we can directly observe and measure such as a person’s age or height. Additionally, variables may be multiple characteristics that fit together to form complex scales represented by a value such as a number from 1 through 10.

An example of a more complex variable from medical research would be a scale that is used for diagnosing a condition or illness. The scale is made up of values that represent many different characteristics. The clinician would use the score of a scale to inform their decision about whether or not a patient had the condition or illness the scale represents.

Importantly, scales that are used to represent many variables in research are derived through a great deal of independent work and are referred to as a tested measure. A tested measure means that whatever is being used is statistically valid and reliable. When reviewing research that uses scales, always check the author’s reference for the scale to make sure that the scale or ‘measure’ the researchers used in their study is a tested and accepted reliable measure in the literature.

When you are presented with a statement that goes something like this… “Researchers from [fill in whatever institution you want] have concluded that their study’s data show… [fill in whatever research finding you want]. When there is a statement that ‘the data show‘, what is really being referred to are the results of looking at all the variables in the study. Variables have values associated with them. All the values for each variable in the study make up the data. Thus, if we have 50 people in a study and we record their age, height, and weight, we would have 3 variables with each variable having one value for each person. When we put all the values together, there are 150 data points (3 variables x 50 participants=150 pieces of data).


Types of Variables

There are different types of variables researchers use within quantitative methods. At the most basic level, there are two main types of variables known as the independent and the dependent variable. Remember that before a researcher chooses a methodology (quantitative, qualitative or a mix of both types), they first must have their research question and in quantitative research often a hypothesis too. You will soon learn to quickly determine what the independent versus the dependent variable are in a study. Always start by noting the research question and/or the hypothesis.

Independent Versus Dependent Variables

The independent variable is the variable that the researchers would like to change or alter during the process of completing the study. In a schematic diagram, you will most often see the independent variable labelled as ‘X’.

The dependent variable is what the researchers are interested in examining as the outcome of their study. The dependent variable is often referred to as the outcome variable and is labelled as ‘Y’. You may use the terms dependent and outcome variable interchangeably. In some instances, both the independent and dependent variables are further classified as the primary independent or main dependent variable. Let’s look at an example.

Imagine a study that is looking at the effect of exercise on blood pressure (with the assumption that over time exercise can reduce hypertension or high blood pressure and cause a person’s blood pressure value to go down). Our researchers would like to know if exercising 1 hour per day for a minimum of 3 days per week impacts blood pressure readings (note you just read the skeleton of the research question).

In the exercise and blood pressure study, exercise is the independent variable (x) and blood pressure is the dependent or outcome variable (y). Given we are focused on quantitative methods, it is important to see that both our independent and dependent variables are measured in the form of numerical data (minutes of exercise, the number of days/week and blood pressure (Hg/mm)).

Of course, there are more variables in the study beyond exercise and blood pressure. Other variables such as age, medications and/or supplements, health status, current fitness level just to name a few. A good question is: ‘If our researchers are interested in examining whether or not multiple exercise sessions per week can reduce blood pressure, then why do the researchers need all the other variables in their study?’

Extraneous Variables

Even though our main focus is on the independent and dependent variables in any study, there are other types of variables, which you will see referred to as extraneous variables included in research studies. We like to think of extraneous variables as possible tricksters because extraneous variables can hide between the independent and dependent variables or can also change the relationship between the independent and dependent variables. Ironically, many extraneous variables are not specifically interesting to the researchers; however, extraneous variables are crucial to include in a study because of how they can confound or muck up a study’s results.

Two specific types of extraneous variables are called mediator and moderator variables. Mediators are the variables that come between or can hide between the independent and dependent variable. The moderators are the variables that impact how strong the relationship is between the independent and dependent variable.

These types of variables can be tricksters for researchers as it is often difficult to observe them directly, yet extraneous variables impact what we see in the data. Let’s think about the fitness level of participants in our exercise and blood pressure study. The fitness level of a participant is directly related to exercise (the extraneous variable, fitness level is related to the independent variable, exercise). It may be important to divide our participants into sedentary, light, moderate and high fitness levels. If we had a person, who was very fit and exercised all the time, the exercise variable our researchers use and define in the study may not be measured in a way that will capture the impact on blood pressure readings because some participants are already so fit (the extraneous variable, fitness level has the ability to impact the results of the dependent variable).

Let’s do one more example, our researchers may want to consider those participants in the study who are taking any type of prescription medication, over the counter medication or supplements that could alter the readings of their blood pressure. If our researchers were to conclude in their study that the exercise changed the blood pressure reading; however, the reality is something else was impacting the dependent variable such as medication or a supplement a serious error in the study’s conclusions could occur.

Ah ha, two ways extraneous variables can be tricksters!

Now imagine for a moment that both fitness level and medications and/or supplements were not included as extraneous variables thus, our researchers did not consider how fitness level is related to our independent variable AND also did not consider what blood pressure medications and supplements participants were taking. The exercise and blood pressure study conclusions would be confounded by something other than the independent variable (exercise). I’m sure you are getting the picture as to how the conclusions of the study could be very inaccurate.

The best approach for researchers is to always include all extraneous variables that are known about before the study begins by ensuring they are in the design of a research study; however, that is not always possible. Sometimes the extraneous variables that may confound a study’s results are not known to the researchers before a study begins. When we get to the topic of data analysis in quantitative methodology, we will examine how we are able to statistically account for these tricksters that we didn’t even know about.



Causation Versus Correlation

An independent variable can cause a dependent variable. An independent variable can be associated with a dependent variable or the independent variable may be completely unrelated to the dependent variable. The first two statements are known as causation and correlation respectively. In the first instance, the independent variable directly causes the dependent variable. In the second case, where an independent variable is correlated with a dependent variable, it means the two variables associated but the independent does not necessarily cause the dependent variable to occur. When an independent variable is correlated with a dependent variable, it is good practice to always clarify both the direction of the association along with how strong the correlation is between the two variables. A correlation or association may be a positive or negative relationship and when we look at statistical data analysis for quantitative methods we will cover this in more detail.

The single most important concept thus far in our research methodology series is to absolutely, without a doubt understand the difference between causation, where the independent variable causes the dependent variable versus correlation, where the independent and dependent variables are associated in some specific way. One of the most common errors in research occurs when statements and publications state some independent thing CAUSES some dependent thing when in fact it is simply not true. Learning how to identify study results that are causal versus correlational and to be able to say to what extent the independent and dependent variable are associated or correlated is a huge skill in becoming a critical consumer of research.


Critical Consumer of Research Information

At first, do not get discouraged in trying to sort out all the different types of variables as it becomes easier over time. For now, just focus on thinking very clearly about what the independent and dependent variables are for any given study. Identifying the independent and dependent variables in a study will come quickly for you as you review a few research publications; however, the extraneous variables and specific mediator and moderator variables are a bit more difficult. The more you read research papers where these specific variables are defined, measured, and used to control what is happening between the independent and dependent variables, the more you will begin to ‘see’ the relationships much more easily.

As you increase your skills regarding quantitative methodology, you will be able to critique the results of studies you hear about in the media. You may be surprised about how much research is presented to all of us each day that has some very serious flaws or confounders in the study that no one ever talks about. Because you are a critical consumer of research information or what we like to refer to at SNJ Associates as an SNJ Research Ranger, you will be more informed and able to detect the difference between valid and good research results versus the invalid or questionable research conclusions.

Your Homework

Your homework is to pop over to the open access journal listing and search for a few studies that use a quantitative methodological approach. Choose a few articles and simply read through and identify both the independent and dependent variables. Think about any extraneous variables the study included or perhaps you will be able to identify some they didn’t include!

Also, consider if there are any variables you think mediate or moderate the study results. By putting these ideas into practice you will soon be impressed with your ability to use the information thus far in the Research Methodology series and know the basic differences between quantitative versus qualitative research methodology along with naming the independent and dependent variables in a quantitative study while specifying any important extraneous study variables. Well done!


 Issue 3 introduces the term ‘operationalization’ of variables and provides a good understanding of why the operational definition of variables in a study is so crucial to valid and reliable quantitative research. We discuss measuring variables and some important aspects of the values attached to variables






Leave a Reply

Your email address will not be published.