Correlations are often used when it is not possible to do an experiment. For example, because it is unethical to do a controlled experiment in a sexual abuse study, we would study a given population to see whether there is any correlation between past sexual abuse and later difficulties with intimate relationships. The limitation of a correlational study is that it doesn’t provide any information about cause. We would not learn whether past sexual abuse causes intimacy difficulties, even though we may learn that these factors are related. The intimacy difficulties could occur for several other reasons, including factors such as low self-esteem or a personality disorder.
|