Correlational Research


Two Kinds of Research

Most research in psychology falls into one of two categories. We will learn about correlational research today and about experimental research next time.

Correlational Research

Correlational research tests for statistical relationships between variables. (I have even seen it referred to as "relational" research.) The researcher begins with the idea that there might be a relationship between two variables. She or he then measures the value of each variable for a large number of cases, and checks to see if they are in fact related. This might involve constructing a contingency table, line graph, or scatterplot and using some other statistical techniques that we will discuss later in the semester.

For example, imagine that a health psychologist is interested in testing the claim that people with more friends tend to be healthier. She surveys 500 people in her community, asking them how many friends they have and getting some measure of their overall health. Then she makes a scatterplot and sees that there is a positive correlation between these variables. Although relationships between two quantitative variables are more likely to be called "correlations," correlational research might involve any of the three types of statistical relationships. For example, imagine that the health psychologist described surveys two groups of people: hospital patients being treated for chronic diseases and healthy community members. Then she compares the two groups in terms of the average number of friends they have and finds that the healthy people have more friends. This is also correlational research. The same would be true if she compared the two groups in terms of the percentage that had a best friend. Whether or not research is correlational does not depend on whether the variables are quantitative, categorical, or one of each. 


A Problem for Correlational Research: Causal Interpretations

Sometimes statistical relationships are interpreted in terms of changes in one variable causing changes in the other. Let us call these causal interpretations. For example, a psychologist might claim that having more friends is not just positively correlated with better health, but that it actually causes people to be healthier. One problem with correlational research, however, is that it provides only weak evidence for causal interpretations. This is because it is possible for Variables X and Y to be related even though X does not cause Y. There are essentially two reasons for this. 

The directionality problem refers to the fact that X and Y will be correlated regardless of whether X causes Y, or Y causes X. Imagine that there is a positive correlation between number of friends and health. This could be because having more friends somehow causes one to be healthier, or it could be because being healthier causes one to have more friends. Maybe healthier people get out of the house more and engage in more social activities, which results in their having more friends.


The third-variable problem refers to the fact that X and Y will also be correlated when there is a third variable that affects both of them. Consider the friends and health example again. Being a happier person might cause one to have more friends and cause one to be healthier. Note that the third-variable problem is not just that there is a third variable that is related to one or both variables; it is that there is a third variable that causes changes in both of the other variables, producing an incidental correlation between them. The thing to remember here is that you have to be very careful about interpreting the results of correlational research as evidence for causal intepretations. Psychologists and psychology majors often remember it this way: "Correlation does not imply causation."


But Do Not Be Too Negative About Correlational Research


Sometimes in a research methods course, we can spend so much time trying to show that correlational studies do not support causal interpretations that we leave the impression that correlational studies are not useful. This is not at all true. They are very useful for at least three reasons. First, correlational research is fine for establishing statistical relationships. A correlational study that shows a positive relationship between number of friends and health has established that this is true, even if it does not establish that more friends causes better health. Sometimes, just knowing that two variables are related is interesting regardless of whether either one has a causal influence on the other. In many situations, after establishing that there is a relationship, researchers will conduct more studies to determine why. Second, even though correlation does not imply causation, causation does imply correlation. For example, if having more friends does cause people to be healthier, then we should observe a positive correlation between these variables. This has two related implications. If we do observe this positive correlation, then at least this is consistent with the causal interpretation and perhaps should count as weak evidence for it. Also, if we do not observe this positive correlation, then this should definitely count as evidence against the causal interpretation. Third, there are situations in which the major alternative to correlational research (experimental research) cannot be conducted for practical or ethical reasons. In such cases, correlational research might be the best approach available. We will discuss examples in class.