The research design refers to the overall strategy that a researcher chooses to investigate the research problem and test their hypothesis. The research design a psychologist chooses may have an effect on the validity of the study.  The two most commonly used designs that we see in IB Psychology are the repeated measures and independent samples design.  However, it is important for you to know that other research designs exist since you will find an example of other ways of doing research throughout the course.

Repeated measures design

In a repeated measures design, you have one sample of participants that receives each condition of an experiment.  If we were testing the effect of music on learning, the same participants would memorize a list of words with music - and then again without music.

It is not always that there are two or more conditions that are tested distinctively.  Sometimes researchers test the conditions concurrently - that is, at the same time.  For example, a psychologist may give a list of words to a group of participants to memorize.  Some of the words are about childhood and the other words are about adulthood. The psychologist wants to see if one type of word is recalled at a higher frequency than another.

The strength of this design is that an individual is only compared to him or herself.  In other words, participant variables are controlled. In the example of having to memorize a list of words while listening to music, an individual's level of English will not make a difference because she is only being compared to herself, not to someone who speaks English more fluently. Another advantage of this design is that fewer participants are needed.

There are also limitations to this design. When participants are asked to take part in more than one condition, they may demonstrate what is called order effects.  For example, boredom, fatigue, or practice effect.  Practice effect is when they get better at something just because they keep doing it.  If they are given a memory test four or five times, they may just get better because they are developing strategies through practice.

Researchers can control for order effects by counter-balancing. This means that one group of participants will start with condition A and then takes part in condition B. The second group of participants starts with condition B and then takes part in condition A.  This makes sure that it is not the order of the conditions that affects the results of the study.

Another limitation is that participants may demonstrate demand characteristics -  that is, participants form an interpretation of the experiment's purpose and subconsciously change their behaviour to fit that interpretation.  They may try to do what they think that the researcher wants them to do - a demand characteristic called the expectancy effect.  Or they may try to disprove the hypothesis with a demand characteristic called the screw you effect.

In repeated measures designs it is often not possible to use the same materials for both conditions. For example, you cannot use the same list of words to memorize under two conditions.  By using two different lists of words, you now have introduced a confounding variable - for example, the difficulty level of the words may be different or some of the words may be more familiar to an individual in one condition than the other. Confounding variables - also called extraneous or third variables - lower the internal validity of an experiment, making it less clear whether it was actually the independent variable that influenced the dependent variable.

Independent samples design

In an independent samples design, members of the sample are randomly allocated to one condition of the experiment. If we were testing the effect of music on learning, the participants could be randomly assigned to the classical music, rock music, pop music, or no music condition.

The strength of this design is that order effects are controlled for since each participant only experiences one condition. In addition, demand characteristics are less likely as the participants will most likely not guess the hypothesis.  Finally, the same materials can be used for all conditions - for example, the same list of words could be used for each of the groups as they listen to different types of music.  Therefore, we could draw a conclusion about the type of music used since all participants were given the same list of words.

As with any design, there are limitations. Unlike the repeated measures design, participant variability may influence the results of the study. - for example,  one group may have more non-native English speakers or better memorizers. Another difficulty of using an independent samples design is that more participants are required.

Matched pairs design

A matched pairs design is an independent samples design in which participants are not randomly allocated to conditions. Instead, they are usually pre-tested with regard to the variable. So, a memory test may be given and then the weakest memorizers are randomly allocated to one of the conditions, then the middle performing memorizers are allocated, and then the top-performing memorizers.  In this way, the researchers guarantee that each condition has the full range of ability.  Random allocation may end up with all of the poor memorizers in a single group. In this way, we lessen the chance that participant variability will affect the results.  It could also be that they are "matched" based on a trait - for example, years speaking English, whether they do regular exercise or if they are a smoker.

 Teacher only box

Does the time your school starts affect student health?

• The IV would be the start time of school (e.g. 7:30 vs. 8:30).  The DV could be the rate of absenteeism, the number of days reported ill in a semester (based on a daily report rather than simply absences), or a survey asking students to comment on health indicators.
• An independent samples design would use two schools.  One school would start at 7:30 and one school would pilot a late-start program, starting at 8:30.  The goal would be to find two schools that have similar attendance behaviours so that the effect can be compared.
• A repeated measures design would most likely provide more valid results. This would mean looking at the health/absenteeism of a school one year when they are still starting at 7:30 and then carrying out the same measures a year later after starting at 8:30. The problem with this design is that the variables in the community are not the same. So, for example, this year there may be a bad flu in the community that was not the case the previous year.  This is what makes research on late starts rather complicated.
• In a matched pair design regarding health, the researcher may want to compare smokers, grade levels, or athletes vs non-athletes.

Does exercise help students to remember more information?

• The IV would be exercise.  This could either be a single experience (running on a treadmill) or asking students to engage in a supervised exercise program for a period of weeks. The DV would be the recall of information.  This could be operationalized as a list of words, a piece of prose, a news story, etc.
• An independent samples design would mean that one group would be asked to exercise before being asked to do the recall task.  The other group would not.
• A repeated measures design works on a "pre-test"/"post-test" design.  The participants would be tested on their ability to recall a story.  Then they would carry out an exercise task.  They would then do a second recall task.  The problem with this design is, unlike the independent samples design, the same text cannot be used. Whereas in the independent samples design the difference in the participants may be a confounding variable, in the case of a repeated measures design, it may be the materials that are problematic.
• In a matched pairs design participants might be matched for level of fitness (eg. regular level of exercise) or their memory skills which would be pre-tested.

Does meditation help students to manage stress?

• The IV would be whether the students engage in meditation or not.  In order for this to be effective, the same routine would have to be used by students and it should be, to some extent, supervised.  Perhaps starting off every TOK class with 5 minutes of meditation.  The DV would be the level of stress.  This could be measured by self-report or by physiological testing - eg. measuring cortisol (a saliva test).
• An independent samples design would mean that one group would be asked to take part in a meditation program and the other would not.  Each group would be asked to measure their stress (self-report or cortisol levels) at the start of the experiment.  Those that are in the control group would be told that they will start a meditation program after the other group finishes their six-week program - or they may be given another task.  For example, the other group could be told to spend ten minutes on the Internet searching for interesting topics every night.
• A repeated measures design works on a "pre-test"/"post-test" design.  This would follow the same plan as the independent samples group but without the control condition.
• In the matched pairs, the researcher may allocate people based on their baseline stress level so that not all the "highly stressed" individuals end up in one group. Other answers, of course, are also possible.

Checking for understanding

Qualitative research methods