Knowing What Variables to Include in Your Psychological Study
Sometimes students are a little unsure of how many variables to include when they’re planning a research study in psychology. But it’s easy to include too many, which has the potential for making your study unfocused and confusing. A manageable research study has a few carefully chosen variables and tells a clear and coherent story.
Review the literature to find out more about the variables related to the psychological constructs you’re interested in – this background research helps you to ensure that you don’t exclude key variables that may be important to your study.
For example, imagine that you’re interested in emotional regulation in young children (that is, their ability to control their emotions), and you hypothesise that shoe size is related to emotional regulation. If you conduct this study, you may find a strong positive relationship between the two variables and conclude that a larger shoe size is related to increased emotional regulation.
However, if you had read the literature thoroughly you would have realised that age is an important predictor of emotional regulation in young children (which, of course, aligns with shoe size, because as young children get older their shoe size tends to increase). If you had included the variable of age in your study, you may have seen that the children’s age predicted emotional regulation, and not their shoe size.
Carefully consider how you plan to combine the variables in your research study. For example, imagine you want to investigate the variables that predict whether psychology students donate money to a mental health charity. Your literature review indicated that two variables commonly predict donating behaviour: the reputation of the charity, and individuals’ prior donating behaviour.
Your study may therefore look first at the influence of the charity’s reputation on donations, and then separately look at the influence of individuals’ prior donating behaviour on donations – a basic experimental design. However, it may be more useful to look at the effect of both the reputation of the charity and individuals’ prior donating behaviour together – a factorial design.
What you may find is that neither the reputation of the charity nor prior donating behaviour can predict donations when considered independently. However, you may find that these two variables can predict donations when they’re combined; that is, individuals will donate if the charity has a good reputation and they’ve previously donated money to charities. This is an interaction effect, and you can only discover these if you include both variables together in a factorial design.