Enhancing External Validity in Psychology Research

By Martin Dempster, Donncha Hanna

One way of maximising external validity in psychology is to ensure, as much as possible, that your sample represents the study population on all the characteristics that are important in your research. You can use a number of different sampling methods to make sure your sample provides a good representation of the study population (its representativeness).

Obtaining a representative sample is related to your sample size. For example, all else being equal, the larger the sample, the more likely it is to be representative of the population. So, in that sense, large samples are good. But how big does your sample need to be? You can consider two things to help you answer this question:

  • How many people do I need in my sample to provide a representative picture of the population?

  • How many people do I need in my sample to allow me to conduct the statistical analysis that I want to, with sufficient power?

To answer the second question, you need to conduct a sample size calculation for the type of analysis that you’re planning. The first question, however, is tougher to address. You need to think about the number of people that you require to convince yourself and others that you have a representative group from the population.

This decision is partly guided by how widespread your topic of interest is in the population; for example, if you want to study depression and you think that levels of depression vary considerably in the population, you need a large number of people to represent this range of outcomes. However, if you believe that the range of depression scores in the population is quite narrow, you can work with a smaller number of people to represent these outcomes.

When you’re thinking about the external validity for your study, consider your sample size in terms of both of these questions – and then work with the largest sample size you anticipate from answering these questions.