When planning a study, the size of the sample can influence the results of the study. To get around this, some researchers choose a research design specifically meant for small sample sizes.
In this lesson, we’ll look at some small ‘n’ designs.
Juan is a psychologist who is interested in how to get people to engage in healthier behaviors, like exercising and eating right. He has designed an intervention that addresses people’s eating habits and makes them more likely to eat a salad than a piece of chocolate cake. Juan needs to test his intervention, so he decides to do a study. But there’s a problem; he doesn’t have a large research budget and can only try his study out on a small number of people.In research, the sample size of a study is the number of subjects in the study.
The sample size is often referred to as the ‘n‘ of the study because it is represented with the variable ‘n‘ in statistical analysis. In general, a large n is best. The more subjects that you have, the better your results will be. But Juan only has a small n, so what can he do?Small n studies involve looking at the same subjects over time.
For example, instead of having two groups, a control group and a treatment group, Juan can have just one group and measure them before and after his intervention. This way he needs fewer subjects, so a small n is not a big deal. There are many types of small n designs that researchers can use. Let’s look closer at two common small n designs: the A-B-A design and multiple baseline design.
Okay, so Juan has a small sample for his study, and he wants to test whether his intervention will improve the eating habits of people. He knows that he’s going to use a single-group design; that is, he will measure all of his subjects multiple times instead of dividing them into separate groups.One option that Juan has is to use an A-B-A design, or ABA design, which involves measuring, offering, and withdrawing treatment. Essentially, the name of the design explains what happens: the subjects are without treatment, which is denoted with the letter ‘A,’ then, they get the treatment, which is denoted with the letter ‘B,’ and finally, they are without the treatment again, and they go back to the letter ‘A.
‘ If Juan chooses an A-B-A design, there are three stages that his subjects will go through:
- A – Baseline Measurement: This is basically what they do before the intervention. Juan might, for example, have his subjects keep a food diary for a week before the intervention, so that he knows what their baseline, or starting point, for eating is.
- B – Intervention: After the baseline measurement, an intervention is given, and subjects are measured during the intervention to see how their behaviors change. For example, Juan can have his subjects go through his healthy eating course and try out his techniques for a week. During the week, he has them continue to keep their food diary so that he can see how much their eating changes from the baseline.
- A – Withdrawal: The final stage involves taking the intervention away from the subjects and measuring them again. In Juan’s case, he’ll have the subjects stop using his techniques but continue to keep a food diary so that he can see what happens to their behaviors without the intervention.
The withdrawal phase of an A-B-A design is important because it shows that the results of the intervention weren’t just a result of a difference in time. For example, what if Juan’s clients improved their eating habits the second week because they’d looked at the food diary from the week before? What if something happened, like a television commercial, that influenced the way they ate?If one of those is influencing the subjects’ behaviors, Juan would expect to see the subjects eating as well during the withdrawal phase as the intervention phase. But if it’s his intervention that’s working, he’d expect them to backslide towards their eating habits of the first week.
Multiple Baseline Design
An A-B-A design is a good way for Juan to see whether his intervention works or not.
But there could be ethical concerns about him withdrawing his intervention. In Juan’s case, taking away the eating intervention could lead to his clients eating unhealthy, which could lead to detrimental effects on their health.So, what if Juan has a small n, but can’t do A-B-A design for ethical reasons? Another way to look at his intervention would be to use a multiple baseline design, which involves measuring many different behaviors and then showing that only the behavior of interest changes during the intervention.For example, during the first week before the treatment, Juan might measure his subjects’ eating habits, exercise habits, and number of hours spent watching television.
After that first week of baseline measurement, Juan will then administer his intervention and then measure those three factors again. These might all be related to health, and so if something other than his intervention changes their behaviors, he’d expect all of those behaviors to change. But if his intervention is the cause of change, he’d expect only the eating habits to change.Multiple baseline designs are often used for studies looking at treatments for something that could cause adverse effects if the treatment is withdrawn. Examples include drugs or other treatments for mental disorders or lifestyle interventions aimed at improving quality of life.
The sample size of a study is the number of subjects and is sometimes referred to as a study’s ‘n.
‘ Some studies only have a small n, which can complicate research. However, there are two designs that work well for small n studies. The A-B-A design involves measuring, treating, and then withdrawing treatment, while the multiple baseline design involves measuring several traits to see if only the target trait changes.
Multiple baseline designs are usually used when there are ethical reasons to prevent the researcher from withdrawing treatment.
After watching this lesson, you should be able to discuss A-B-A design and multiple baseline design, which are two examples of small n designs.