• Skip to secondary menu
  • Skip to main content
  • Skip to primary sidebar
  • My Store
  • Glossary
  • Home
  • About Me
  • Contact Me

Statistics By Jim

Making statistics intuitive

  • Graphs
  • Basics
  • Hypothesis Testing
  • Regression
  • ANOVA
  • Probability
  • Time Series
  • Fun

Control Variables: Definition, Uses & Examples

By Jim Frost 1 Comment

What is a Control Variable?

Control variables, also known as controlled variables, are properties that researchers hold constant for all observations in an experiment. While these variables are not the primary focus of the research, keeping their values consistent helps the study establish the true relationships between the independent and dependent variables. Control variables are different from control groups.

Control variables are important in science.In science, researchers assess the effects that the independent variables have on the dependent variable. However, other variables can also affect the outcome. If the scientists do not control these other variables, they can distort the primary results of interest. In other words, left uncontrolled, those other factors become confounders that can bias the findings. The uncontrolled variables may be responsible for the changes in the outcomes rather than your treatment or experimental variables. Consequently, researchers control the values of these other variables.

Suppose you are performing an experiment involving different types of fertilizers and plant growth. Those are your primary variables of interest. However, you also know that soil moisture, sunlight, and temperature affect plant growth. If you don’t hold these variables constant for all observations, they might explain the plant growth differences you observe. Consequently, moisture, sunlight, and temperature are essential control variables for your study. These variables can be controlled by keeping their values constant for all observations in your experiment. That way, if you do see plant growth differences, you can be more confident that the fertilizers caused them.

When researchers use control variables, they should identify them, record their values, and include the details in their write-up. This process helps other researchers understand and replicate the results.

Related posts: Independent and Dependent Variables and Confounding Variables

Control Variables and Internal Validity

By controlling variables, you increase the internal validity of your research. Internal validity is the degree of confidence that a causal relationship exists between the treatment and the difference in outcomes. In other words, how likely is it that your treatment caused the differences you observe? Are the researcher’s conclusions correct? Or, can changes in the outcome be attributed to other causes?

If the relevant variables are not controlled, you might need to attribute the changes to confounders rather than the treatment. Control variables reduce the impact of confounding variables.

Controlled Variable Examples

Experiment Control Variables
Does a medicine reduce illness?
  • Age
  • Health
Are different weight loss programs effective?
  • Activity levels
  • Starting weights
Do kiln time and temperature affect clay pot quality?
  • Ambient humidity levels
  • Type of clay
  • Clay moisture
Does a supplement improve memory recall?
  • Amount of sleep
  • Timing of medication
  • Familiarity with recall subject

How to Control Variables in Science

Scientists can control variables using several methods. In some cases, variables can be controlled directly. For example, researchers can control the growing conditions for the fertilizer experiment. Or use standardized procedures and processes for all subjects to reduce other sources of variation. These efforts attempt to eliminate all differences between the treatment and control groups other than the treatments themselves.

However, sometimes that’s not possible. Fortunately, there are other approaches.

Related post: Control Groups in Experiments

Random assignment

In some experiments, there can be too many variables to control. Additionally, the researchers might not even know all the potential confounding variables. In these cases, they can randomly assign subjects to the experimental groups. This process controls variables by averaging out all traits across the experimental groups, making them roughly equivalent when the experiment begins. The randomness helps prevent any systematic differences between the experimental groups. Learn more in my post about Random Assignment in Experiments.

Statistical control

Directly controlled variables and random assignment are methods that equalize the experimental groups. However, they aren’t always feasible. In some cases, there are too many variables to control. In other situations, random assignment might not be possible. Try randomly assigning people to smoking and non-smoking groups!

Fortunately, statistical techniques, such as multiple regression analysis, don’t balance the groups but instead use a model that statistically controls the variables. The model accounts for confounding variables.

In multiple regression analysis, including a variable in the model holds it constant while the treatment variable fluctuates. This process allows you to isolate the role of the treatment while accounting for confounders. You can also use ANOVA and ANCOVA.

For more information, read my posts, When to Use Regression and ANOVA Overview.

Share this:

  • Tweet

Related

Filed Under: Basics Tagged With: conceptual, experimental design

Reader Interactions

Comments

  1. Dave says

    January 26, 2023 at 7:00 pm

    Dear Jim,

    I have a question please about when a control variable is also itself part of the dependent variable. I see this referred to in the medical research literature as ‘mathematical coupling’, where – for example – the beats per minute (BPM) is the dependent variable and researchers want to put minutes also as a control variable. This seems to create a problem because ‘minutes’ appears on both sides of the equation, and the medical literature talks about spurious correlation, and the model needing to be redesigned. But do you have a simple text or reference – ideally just plain statistics/OLS rather than linked to medical research – where this could be explained in theory terms ? What goes wrong in the regression when a variable is both a control variable and part of the dependent variable (perhaps as part of a ratio or measurement of change)? I just haven’t found a textbook reference that says definitively ‘you can’t have the same variable in both sides of the regression simultaneously’, so I’m not sure whether this violates OLS and so is something to avoid entirely (with a new model design or different research question) or to live with.

    Any help would be great, thank you for your work,

    Dave

    Reply

Comments and Questions Cancel reply

Primary Sidebar

Meet Jim

I’ll help you intuitively understand statistics by focusing on concepts and using plain English so you can concentrate on understanding your results.

Read More...

Buy My Introduction to Statistics Book!

Cover of my Introduction to Statistics: An Intuitive Guide ebook.

Buy My Hypothesis Testing Book!

Cover image of my Hypothesis Testing: An Intuitive Guide ebook.

Buy My Regression Book!

Cover for my ebook, Regression Analysis: An Intuitive Guide for Using and Interpreting Linear Models.

Subscribe by Email

Enter your email address to receive notifications of new posts by email.

    I won't send you spam. Unsubscribe at any time.

    Follow Me

    • FacebookFacebook
    • RSS FeedRSS Feed
    • TwitterTwitter

    Top Posts

    • How to Interpret P-values and Coefficients in Regression Analysis
    • How To Interpret R-squared in Regression Analysis
    • Mean, Median, and Mode: Measures of Central Tendency
    • Multicollinearity in Regression Analysis: Problems, Detection, and Solutions
    • How to Interpret the F-test of Overall Significance in Regression Analysis
    • Choosing the Correct Type of Regression Analysis
    • How to Find the P value: Process and Calculations
    • Interpreting Correlation Coefficients
    • How to do t-Tests in Excel
    • Z-table

    Recent Posts

    • Fishers Exact Test: Using & Interpreting
    • Percent Change: Formula and Calculation Steps
    • X and Y Axis in Graphs
    • Simpsons Paradox Explained
    • Covariates: Definition & Uses
    • Weighted Average: Formula & Calculation Examples

    Recent Comments

    • Dave on Control Variables: Definition, Uses & Examples
    • Jim Frost on How High Does R-squared Need to Be?
    • Mark Solomons on How High Does R-squared Need to Be?
    • John Grenci on Normal Distribution in Statistics
    • Jim Frost on Normal Distribution in Statistics

    Copyright © 2023 · Jim Frost · Privacy Policy