Experimental Design

The experimental design is the complete plan for how an experiment will be carried out.

Deciding on an experimental design is the most important decision a researcher makes when planning an experiment. The experimental design encompasses who the subjects will be, how many groups there will be, what variables will be manipulated, what data will be recorded, and other choices that affect the way the data will be collected and analyzed.

The simplest form of experimental design is the two condition design with one independent variable and one dependent variable. The independent variable is the variable that is manipulated by the experimenter. The dependent variable is the outcome of interest that is measured. For example, if an experimenter wanted to study the effect of font size on reading comprehension, she might give participants an essay to read in either regular sized or very small font and then ask the participants to take a reading comprehension quiz to see how well they understood the material. The manipulated variable, in this case the font size, is the independent variable. The measured variable, in this case the score on the reading comprehension quiz, is the dependent variable.

In this experiment there were two groups. The experimental group is the group that receives the manipulation, in this case, the experimental group would be the group that received the essay in very small font. The control group is the group of participants that is used as a baseline against which to compare the scores of the experimental group to determine if there was an effect of the manipulation. The control group was the group who read the essay printed in normal-sized font.

Choosing an appropriate control group sounds easier than it often is in practice. For example, imagine a researcher who is interested in whether a new medication is effective in treating anxiety. The experimental group is easy. He will give its members the new medication. The control group is a little more difficult. Should he give them a previously existing anxiety medication? This type of control group is often used in clinical studies, when the outcome of interest is whether the new medication is more effective or has fewer side effects than the currently available treatment. Perhaps the researcher should not give the participants in the control group any medication at all. This, however, can cause its own problems. Sometimes people experience relief from symptoms simply because they believe they are receiving treatment and not actually because of the treatment. This is called the placebo effect and can be very strong.

The placebo effect can generally be avoided by giving participants in the control group a sugar pill or other non-effective treatment that looks, feels, and tastes as similar to the real medication as possible. In the case of the researcher who wants to test a new anxiety medication, he should make sure that participants in both groups receive some kind of pill and do not know whether it is the real medication or not. Otherwise the new medication might seem like it works, only because the participants in that group know they are receiving it and expect it to work. In research where the independent variable is therapy, an exercise routine, or surgery, it can be much more difficult to determine how to treat participants in the control group to minimize the chance of a placebo effect affecting the results.

Some experiments have only a control group and one experimental group. Other, more complicated studies may have a control group and many experimental groups. For example, a researcher who was interested in studying the effects of alcohol on driving ability might have a control group use a driving simulator after not drinking any alcohol, and then have experimental groups who used the simulator after 1, 2, or 3 drinks.

In some cases, an experiment will measure the dependent variable only once. In other cases in may be measured many times. For example, if a researcher wanted to study the effect of age on math ability, she may give students a math quiz when they are 6, 8, and 10 years of age. A study that follows one group of participants and takes repeated measures over time is called a longitudinal study.

When a researcher expects there to be large individual differences between participants or when it is too difficult to create a control group, sometimes an experimental design called a repeated-measures design is used. In this design, the dependent variable is measured before and after an intervention to evaluate the effect of the intervention. For example, if a researcher wanted to determine whether cognitive behavioral therapy was effective for treating depression, he might measure the level of depression of participants using a depression inventory before starting therapy and then again after six weeks of therapy.

See also Research methodology .


Control group—
The group of subjects without the manipulation or experimental intervention, used as a baseline to measure the effects of the intervention.
Dependent variable—
The characteristic or variable measured by the experimenter that is expected to be affected by the independent variable.
Experiment group—
The group of subjects with the manipulation or experimental intervention.
Independent variable—
The characteristic or variable manipulated by the experimenter.
Placebo effect—
An effect that occurs not because of any actual medication or treatment, but instead because participants believe they have received the medication or treatment.



Bausell, Barker. The Design and Conduct of Meaningful Experiments Involving Human Participants: 25 Principles. New York: Oxford University Press, 2015.

Christensen, Larry, et al. Research Methods, Design, and Analysis, 12th ed. Boston: Pearson, 2014.

Crano, William, et al. Principles and Methods of Social Research, 3rd ed. New York: Routledge, 2014.

Howitt, Dennis. Introduction to Research Methods in Psychology, 4th ed. New York: Pearson, 2014.

Schneider, Sandra. Experimental Design in the Behavioral and Social Sciences. Los Angeles: SAGE, 2013.