What is a “controlled variable?”

Travis Dixon Internal Assessment (IB), Research Methodology

A controlled variable (aka a control variable) is any variable held constant to avoid confounding variables affecting a study.

A controlled variable is a variable that’s kept constant between the conditions of the experiment so that the only difference between the groups is the independent variable. 

What extraneous variables would you need to control in an experiment to see if coffee gives you more energy?

Imagine you’re doing an experiment on yourself to see if drinking coffee in the morning gives you energy. On Monday you wake up, do yoga and then have a cup of coffee. At the end of the day you rank your energy and give it a 9/10. On Tuesday you wake up drink orange juice and then at the end of the day you rank your energy 7/10. This simple experiment seems like it’s saying coffee gives you more energy. But there’s a problem – on both days you didn’t just change the drink, you also did yoga on the coffee day. How do you know it was the coffee that made you feel better and not the yoga? The yoga in this instance is an extraneous variable – it’s a variable other than your  independent variable (your morning drink) that could have affected your results.

You want to make sure you’re just testing the effects of coffee. This means you need to control certain variables, like how much sleep you get, the type of exercise you do during the day, stress levels, etc. Anything you do to keep these extraneous variables the same on both days are controlled variables. 

Learn More:

Controlled Variables in Clinical Drug Trials

IB Psychology - A Student's Guide - The IA - eBook

Available here.

The “gold standard” of experimental research in psychology is the randomized double-blind, placebo-controlled experiment. This type of study controls a number of variables that are potentially confounding variables. If an experiment is “well-controlled” we can confidently say any improvement seen in participants is because of the effects of the IV on the DV, e.g. the effects of medication on symptoms (Read more about drug trials using SSRIs for PTSD here).

  • Randomized: participants are randomly allocated to the experimental (treatment) and control (e.g. placebo or no treatment) conditions. The main controlled variable here is between-subject variables (aka participant variability). By randomly allocating people to either group you’re limiting the chances that group differences (e.g. the average age of one group is 20 years older than the other) will affect the results, so between-subject variability becomes a controlled variable.
  • Double-blind means neither the participant nor the data collector (e.g. interviewer) know if the participant took the drug or a placebo. The controlled variables here are the placebo effects and researcher bias. 
  • Placebo-controlled means one group gets a placebo, which controls for the participant expectancy effect (aka the placebo effect). 

Control variable (aka controlled variable): “…any variable apart from the independent variable that is controlled by the experimenter by being randomized, held constant, statistically controlled, or suppressed in some other way.” (Oxford Dictionary of Psychology).


Here is a famous study by Solomon Asch on conformity. If you were designing this experiment, would you use a single-blind or a double-blind design? What would be the advantages and/or disadvantages of both? What specific confounding variable would be reduced if you used a double-blind study?