The Experiment in Psychology

In the last lesson, you heard Shallon give her advice on what makes a guy attractive.  But you also realized that her advice was Pop Psychology - it consisted of unsubstantiated claims, without any research evidence to back them up.  So, how can the Scientific method be used to investigate what makes a guy attractive, after all?  One way is by carrying out an experiment.  What is an experiment, and how is it carried out?  That's the focus of this lesson.
Video Activity
As you are watching the video below, keep these questions in mind:

  • What was the aim of this study?  What were the researchers trying to find out?

  • How did the researchers go about testing their hypothesis?

  • How convincing is this experiment?  Can you think of any flaws in the research?
In the video you just saw, the researchers tested the idea that "social and financial status" increases the attractiveness of men.  In order to test their hypothesis, they carried out an experiment.  Here are some of they key features of an experiment:

  • An experiment is a scientific way of testing a hypothesis.  This involves investigating an idea a psychologist has about the causes of behavior

  • Experiments involve three types of variables:  the independent variable, the dependent variable, and extraneous variables

  • The independent variable is what the psychologist is interested in exploring, to see if it is the hidden cause of behavior.  This is manipulated (changed) to see if it has an effect on a particular behavior.  Often, there will be two (or more) groups of people in the experiment, and each group will receive a different "level" of the independent variable.

  • The dependent variable is the outcome or result that is measured.  If the hypothesis was correct, then a change in the independent variable will result in a change in the dependent variable.

  • Extraneous variables are other variables that may affect the results.  To be sure that it is really the independent variable that is making the difference, extraneous variables should be controlled.  That means everything that could affect the results should stay the same for everybody in the study, except the independent variable.

A simple example may be useful here.  Suppose a psychologist believes that caffeine will increase memory ability.  To test this hypothesis, the psychologist gathers a group of people, called participants, who have volunteered to take part in this study.

  • In this experiment, the independent variable is caffeine.  The participants will be divided into two groups - one group will be given a cup of regular coffee to drink before a memory test, the other group will drink decaf coffee (with no caffeine)

  • The dependent variable is the score on the memory test.  If the psychologist is correct, the group of participants who drank a regular coffee will score higher on the memory test.

  • To make sure its really the caffeine (and not some other factor) that is making the difference, the psychologist should control all extraneous variables.  For example, the two groups of participants should be alike as possible - similar age, level of education, memory ability, and so on.  Obviously, the test itself must be exactly the same for all participants.  Even the room where the test is taken, the time of day, and the test instructions should be exactly the same for all participants.

Think Critically
In the video you watched earlier, a simple experiment was carried out on male attractiveness.  In that video, what was:

  • The independent variable?

  • The dependent variable?

  • Extraneous variables which were controlled?

Scroll down to the bottom of this page for the answers!
Types of Experiments

There are three types of experiments, based on how and where they are carried out.

Laboratory experiments take place in a psychology lab, usually a special room set up in a university or research institute. 

  • Strength: It is easy to control extraneous variables.  The psychologist can make sure that everything in the room stays exactly the same for all participants.  The time of day, furniture, instructions, lighting and so on can all controlled by the researcher.

  • Strength: Since everything about the lab experiment is set in advance, it is easy for another researcher to replicate (copy) the experiment and confirm the results

  • Limitation: People may not behave the way they normally do when in a psychology lab, especially since they know they are taking part in a study and are being observed.  If people behave differently during a lab experiment than they normally would, then the results from the lab might not apply to real life

A field experiment takes place in an ordinary, everyday environment, such as on the street, in a subway station, at a shopping mall, or in a classroom.

  • Strength: Since the experiment takes place in a natural setting, its more likely that the behavior of participants will reflect how they would behave in everyday life

  • Limitation: Controlling all extraneous variables is very difficult in a real life environment.  For instance, if you are carrying out your experiment in a subway station, the environment is constantly changing.  One minute the station might be very crowded, the next minute less so.  Or, there might be different music playing in the background from one minute to the next.

A natural experiment is an experiment in which the change in the independent variable occurs naturally.  For instance, imagine two remote islands in the middle of the Pacific Ocean.  One island gets hooked up to satellite TV, while the other island won't get TV for another few years.  That allows researchers to study the effects of television on behavior.

  • Strength: Natural experiments are real-life situations, and hence reflect how people behave in everyday life

  • Strength: Natural experiments can be used in situations where it would be ethically impossible to manipulate the independent variable.  For instance, it would be practically impossible for researchers to withhold television access to a whole island, just to study TV's effects.

  • Limitation: Natural experiments are usually one-time only events, which are impossible to replicate
Issues in Experimental Research

Because Psychology involves research on people (and not, say, on plants), there are lots of issues that crop up in the fine details of carrying out an experiment. These issues need to be addressed carefully to ensure that the experiment produces valid results. 

  • Participant variability - Experiments often involve two or more groups of participants.  However, if the participants in one group are different in any way from the other group, then this may affect the results.  Remember the goal of the experiment is to investigate the effects of the independent varaible - and only the independent variable!  So its important for each group of participants to be as similar as possible - for instance, in terms of age, gender, intelligence, income, cultural background, and anything else that might possible influence the results.  The simplest way to do this is to randomly assign participants to the different groups.  As long as there are enough participants, then one can assume that each group will have a roughly even mix of people with different characteristics, with no major differences between each group.

  • Demand characteristics - People in an experiment are naturally curious, and will often try to "guess" what the experiment is really about.  They may pay attention to subtle cues from the researcher, including tone of voice and body language, in trying to figure out the true aim of the experiment.  If participants know (or think they know) what the researcher is expecting to see, this could cause them to behave unnaturally.  Participants may try to please the researcher by "going along" with the expected results, or they may even rebel and do the opposite of what's expected of them!  (This is known as the "screw you" effect).  To reduce demand characteristics, the researcher may keep the full details of the research hypothesis secret, or may even mislead participants about what the study is really about.  However, withholding information (or lying) to participants raises ethical issues, which will be explored in a later lesson.

  • Researcher bias - Researchers are people too, and getting certain results from their experiment can be very important for their careers and reputations.  They may (consciously or unconsciously) design the experiment to get the results they are hoping for, drop hints (through tone of voice or body language) to participants to encourage certain behavior, or "see what they want to see", ignoring results that don't support their theory.  Making sure that research is carried out without any bias is a challenge for any study.  One way to overcome researcher bias - especially useful in medical trials of new drugs - is called a "double blind".  This means that neither the participant nor the doctor knows who is getting the real drug, and who is getting a sugar pill (known as a placebo).  The idea here is to eliminate any subtle hint or clue from the doctor that could cause the participant to expect improvement (or not) in their medical condition.

TOK Link
Experiments are carried out both in the Natural Sciences (like Chemistry and Physics) and in the Social Sciences (like Psychology).

  • What are some similarities and differences between an experiment in the Natural Sciences and an experiment in the Social Sciences?

  • Is knowledge obtained from Social Science research more, less or equally as valid and reliable as knowledge in the Natural Sciences?  Explain why.

  • What are some unique challenges faced by researchers in carrying out experiments on humans?

  • I can explain the purpose of the independent variable, dependent variable, and extraneous (controlled) variables in testing an experimental hypothesis

  • I can describe three types of experiments, and discuss at least one strength and limitation of each

  • I can discuss three issues in experimental research, and ways to minimize these issues in order to improve the validity of the results

Try it Out

Now that you have learned so much about experiments in Psychology, lets see if you can apply what you've learned to analyze a real research study.

  • Watch the video below on the "Still Face" experiment

  • What hypothesis is being tested in this experiment?

  • What is the independent and dependent variable in this experiment?  What are some extraneous variables that are controlled?

  • Of the three issues in experiments discussed above, which one might pose the greatest cause for concern?  Explain why.

Scroll to the bottom of the page for suggested answers!
Quiz Yourself!

1.  The independent variable in an experiment is the variable which is:

(a) manipulated

(b) measured

(c) controlled

(d) rejected

2.  The dependent variable in an experiment is the variable which is:

(a) manipulated

(b) measured

(c) controlled

(d) rejected

A psychologist tests the hypothesis that people are more attracted to others who share similar interests and values.  The research is carried out by inviting participants to rate their level of interest in a potential romantic partner based on a fake internet dating profile. 

3.  Which of the following is an extraneous variable that should be controlled?

(a) Political identification of the potential romantic partner

(b) Occupation of the potential romantic partner

(c) Favorite sports team of the potential romantic partner

(d) Physical attractiveness of the potential romantic partner

4.  The independent variable in this study is:

(a) Level of interest in meeting the potential romantic partner

(b) Physical attractiveness of the potential romantic partner

(c) Salary of the potential romantic partner

(d) Degree of similarity with the potential romantic partner

A psychologist tests the hypothesis that praise increases student's academic performance.  The research is carried out by instructing some teachers to praise their students often, while instructing other teachers not to praise their students.

5.  The most important dependent variable in this study would likely be:

(a) Final exam grades

(b) Teaching style

(c) Physical attractiveness of the teacher

(d) Student attendance records

6. What sort of experiment is this?

(a) Lab experiment

(b) Field experiment

(c) Natural experiment

(d) Organic experiment

7. Some teachers are better at teaching than others, and so better teachers will produce better academic results.  How can this effect be minimized?

(a) Carry out a double-blind experiment

(b) Randomly assign students to different teachers

(c) Randomly assign teachers to different experimental groups

(d) Carry out a single-blind experiment

8.  Some of the teachers who are instructed to withhold praise may feel they are being set up to fail their students, and so may subsequently put less effort into their lessons.  This is an example of:

(a) Participant variability

(b) Demand characteristics

(c) Researcher bias

(d) The independent variable


Think critically

Independent variable - Social & economic status (Prestigious, high paying job vs. low status, low paying job)

Dependent variable - Attractiveness rating

Extraneous (controlled) variables - The picture and age of the man

Try it out

Hypothesis - Babies do engage in social interaction

Independent variable - Mother's social engagement with the baby

Dependent variable - Baby's behavior (emotions, posture, physical activity)

Issue of concern - Rating the baby's behavior is subjective, and so the researcher may be biased in assigning ratings.  For instance, the researcher may be more sensitive to signs of distress in the baby when the mother is making a still face.  One way to eliminate this bias would be to use an independent rater who assesses the baby's behavior without knowing whether the mother was engaging with the baby or not

Quiz Yourself

1-A, 2-B, 3-D, 4-D, 5-A, 6-B, 7-C, 8-B