How to Conduct a Psychology Experiment
By Kendra Cherry
Updated January 02, 2018
Conducting your first psychology experiment can be a long, complicated, and intimidating process. It can be especially confusing if you are not quite sure where to begin or which steps to take. Like other sciences, psychology utilizes the scientific method and bases conclusions upon empirical evidence. When conducting an experiment, it is important to follow the five basic steps of the scientific method:
Find a Research Problem or Question
Picking a research problem can be one of the most challenging steps. After all, there are so many different topics you might choose to investigate. Stumped for an idea? Consider some of the following:
Investigate a commonly held belief. Folk psychology is a good source of unanswered questions that can serve as the basis for psychological research. For example, many people believe that staying up all night to cram for a big exam can actually hurt test performance. You could conduct a study in which you compare the test scores of students who stayed up all night studying, versus the scores of students who got a full night's sleep prior to the exam. Review psychology literature. Published studies are a great source of unanswered research questions. In many cases, the authors will even note the need for further research. Find a published study that you find intriguing, and then come up with some questions that require further exploration. Think about everyday problems. There are many practical applications for psychology research. Explore various problems that you or others face each day, and then consider how you could research potential solutions. For example, you might investigate different memorization strategies to determine which methods are most effective.
Operationally Define Your Variables
Variables are anything that might impact the outcome of your study. An operational definition describes exactly what the variables are and how they are measured within the context of your study. For example, if you were doing a study on the impact of sleep deprivation on driving performance, you would need to operationally define what you mean by sleep deprivation and driving performance.
In this example you might define sleep deprivation as getting less than seven hours of sleep at night and define driving performance as how well a participant does on a driving test.
What is the purpose of operationally defining variables? The main purpose is control. By understanding what your are measuring, you can control for it by holding the variable constant between all of the groups or manipulating it as an independent variable.
Develop a Hypothesis
The next step is to develop a testable hypothesis that predicts how the operationally defined variables are related. In our example in the previous step, our hypothesis might be: "Students who are sleep deprived will perform worse than students who are not sleep deprived on a test of driving performance."
In order to determine if the results of the study are significant, it is essential to also have a null hypothesis. The null hypothesis is the prediction that one variable will have no association to the other variable. In other words, the null hypothesis assumes that there will be no difference in the effects of the two treatments in our experimental and control groups.
The null hypothesis is assumed to be valid unless contradicted by the results. The experimenters can either reject the null hypothesis in favor of the alternative hypothesis or not reject the null hypothesis.
It is important to remember that not rejecting the null hypothesis does not mean that you are accepting the null hypothesis. To say that you are accepting the null hypothesis is to suggest that something is true simply because you did not find any evidence against it. This represents a logical fallacy that should be avoided in scientific research.
Conduct Background Research
Once you have developed a testable hypothesis, it is important to spend some time doing some background research. What do researchers already know about your topic? What questions remain unanswered? You can learn about previous research on your topic by exploring books, journal articles, online databases, newspapers, and websites devoted to your subject.
Reasons to conduct background research:
Reading previous research helps you gain a better understanding of what you will encounter during your own experiment. Understanding the background of your topic provides a better basis for your own hypothesis. After conducting a thorough review of the literature, you might choose to alter your own hypothesis.
Background research also allows you to explain why you chose to investigate your particular hypothesis and articulate why the topic merits further exploration. As you research the history of your topic, remember to take careful notes and create a working bibliography of your sources. This information will be valuable when you begin to write up your experiment results.
Select an Experimental Design
After conducting background research and finalizing your hypothesis, your next step is to develop an experimental design. There are three basic types of designs that you might utilize. Each has its own strengths and weaknesses.
Pre-Experimental Designs: This type of experimental design does not include a control group. A single group of participants is studied, and there is no comparison between a treatment group and a control group. Examples of pre-experimental designs include case studies (one group is given a treatment and the results are measured) and pre-test/post-test studies (one group is tested, given a treatment and then retested). Quasi-Experimental Designs: This type of experimental design does include a control group, but the design does not include randomization. True Experimental Designs: A true experimental design include both of the elements that the pre-experimental designs and quasi-experimental designs lack on their own - control groups and random assignment to groups.
Standardize Your Procedures
In order to arrive at legitimate conclusions, it is essential to compare apples to apples. Each participant in each group must receive the same treatment under the same conditions. For example, in our hypothetical study on the effects of sleep deprivation on driving performance, the driving test must be administered to each participant in the same way. The driving course must be the same, the obstacles faced must be the same, and the time given must be the same.
Choose Your Participants
In addition to making sure that the testing conditions are standardized, it is also essential to ensure that your pool of participants is the same. If the individuals in your control group (those who are not sleep deprived) all happen to be amateur race car drivers while your experimental group (those that are sleep deprived) are all people who just recently earned their drivers licenses, your experiment will lack standardization.
When choosing subjects, there are a number of different techniques you can use. A simple random sample involves randomly selecting a number of participants from a group. A stratified random sample requires randomly selecting participants from different subsets of the population. These subsets might include characteristics such as geographic location, age, sex, race, or socioeconomic status.
Conduct Tests and Collect Data
After you have selected participants, the next steps are to conduct your tests and collect the data. Prior to doing any testing, however, there are a few important concerns that need to be addressed. First, you need to be sure that your testing procedures are ethical. Generally, you will need to gain permission to conduct any type of testing with human participants by submitting the details of your experiment to your school's Institutional Review Board, sometimes referred to as the 'Human Subjects Committee.'
After you have gained approval from your academic institution's IRB, you will need to present informed consent forms to each of your participants. This form offers information on the study, the data that will be gathered, and how the results will be used. The form also gives participants the option to withdraw from the study at any point in time.
Once this step has been completed, you can begin administering your testing procedures and collecting the data.
Analyze the Results
After collecting your data, it is time to analyze the results of your experiment. Researchers utilize statistics to determine if the results of the study support the original hypothesis and to determine if the results are statistically significant. Statistical significance means that the results of the study are unlikely to have occurred simply by chance.
The types of statistical methods you use to analyze your data depend largely on the type of data that you collected. If you are using a random sample of a larger population, you will need to utilize inferential statistics. These statistical methods make inferences about how the results relate to the population at large. Because you are making inferences based upon a sample, it has to be assumed that there will be a certain margin of error.
Write Up and Share Your Results
Your final task in conducting a psychology experiment is to communicate your results. By sharing your experiment with the scientific community, you are contributing to the knowledge base on that particular topic. One of the most common ways to share research results is to publish the study in a peer-reviewed professional journal. Other methods include sharing results at conferences, in book chapters, or in academic presentations.
In your case, it is likely that your class instructor will expect a formal write-up of your experiment in the same format required in a professional journal article or lab report: