Laughlin, Charles, Pre- and Peri-natal Psychology Journal
The most difficult research to carry out effectively in pre- and perinatal psychology is longitudinal research; that is, research designed to uncover causal relations between pre- and perinatal events and the subsequent psychosocial development of the person. It is often very difficult to pin down the causes of adult psychological problems, even when the anecdotal, clinical or intuitive view makes it nearly certain that the roots of the problems are to be found in early life. Most often researchers resort to correlational methods-that is, ways of documenting that changes in the distribution of one variable are related in a patterned way to changes in the distribution of another variable-as such research may be carried out in a short time and is resource-efficient. But we must always remember one simple fact about correlational results: a correlation is not a cause. To treat a correlation as a cause is a logical fallacy. A causal explanation requires that the mechanisms or processes that link the cause and the effect be elucidated. Correlational data do not do this.
However, correlational data can often point the way toward causal links between events, and provide very important information when causal data are unavailable. Moreover, correlational methods may be used to test causal hypotheses. Let me give you an example. Swedish researchers Karin Nyberg, Peter Allebeck, Gunnar Eklund and Bertil Jacobson (1992, 1993) have tested which factor better explains the incidence of amphetamine and opiate addiction: socioeconomic level of the addict, or the administration of drugs to the addicts' mothers during birthing. They show a small correlation between amphetamine addiction and socioeconomic level, but none for opiate addiction. …