In Section 2-5, you learned about Greg, a man who suffered extensive damage to his brain because of a large tumor. Greg’s problems were described by the neurologist, Oliver Sacks (1995; neurology is a medical field in which the structures, functions, and disorders of the nervous system are studied). The description that Sacks provided represents a case study — an intensive examination of an individual suffering from physical and/or mental abnormalities. Case studies in clinical areas, such as medicine and clinical psychology, include information that may help to explain a patient’s abnormalities: medical histories, medical tests, psychological tests, childhood experiences, and so on. A major strength of case studies is the large amount of potentially useful information collected. In neurological case studies such as Greg’s, detailed observations of the location and extent of brain damage are included.
In Section 1-6. you learned about another case study–that of a woman named Christina, who lost the ability to sense her body. Here is the introduction to a documentary about another person who developed a similar disorder:
Clinical researchers collect information that they hope will help them to find factors (causes) responsible for the development of their patients’ abnormalities. Their ultimate goal is to distinguish the effects of one or more of these factors from the effects of extraneous variables. An extraneous variable is a factor, other than the one being investigated, that also causes changes in the events being studied. For example, neurologists reading Greg’s case study immediately would suspect that brain damage caused the changes in his cognitions, emotions, and behaviors. However, one possible extraneous variable is the religious training that he was undergoing. In fact, the members of his religious sect believed that “spiritual enlightenment” was the primary cause of the changes in Greg’s personality:
[Greg] complained that his vision was growing dim, but this was interpreted, by his swami and others, in a spiritual way: he was “an illuminate,” they told him, it was the “inner light” growing. Greg had worried at first about his eyesight, but was reassured by the swami’s spiritual explanation. His sight grew still dimmer, but he offered no further complaints. And indeed, he seemed to be becoming more spiritual by the day — an amazing new serenity had taken hold of him. He no longer showed his previous impatience or appetites [desires], and he was sometimes found in a sort of a daze, with a strange … smile on his face. It is beatitude, said his swami — he is becoming a saint. (Sacks, 1995, pp. 43-44)
Neurologists, however, would immediately have noticed that Greg’s bizarre thinking and behavior, along with his physical problems, were identical to those experienced by neurological patients who have suffered damage to a particular area of the brain. They would have quickly dismissed, therefore, the possibility that Greg’s religious training was an important extraneous variable.
Controlling for Extraneous Variables
If researchers wish to study the effects of one factor, they must control for the possibility that extraneous variables are affecting their results. The term control for refers to setting up a research situation in such a way that the effects of extraneous variables are excluded. In other words, by controlling for extraneous variables, researchers are trying to reduce the possibility that extraneous variables explain the results of a study.
For example, the children of parents with schizophrenia are more likely to develop the disorder themselves when they get older. One possible explanation of this fact is that schizophrenia is caused, in part, by genes — genes that the children of schizophrenic parents inherit (Sullivan, 2005). But there is an important extraneous variable that might explain these results: the effects of being raised in a family with a schizophrenic parent. How might researchers control for that extraneous variable? One commonly used method is to study children who were adopted at a very young age, so that they weren’t raised by a schizophrenic parent. If these children are more likely to develop schizophrenia even though they weren’t raised by their biological parent, it seems likely that genes they inherited from this parent are a cause of their schizophrenia.
More evidence for this claim is illustrated in Figure 1, which shows that the closer one’s biological relationship to a schizophrenic person is — and hence, the more genes one shares with that person — the more likely one is to develop schizophrenia. (Note: as you learned in Section 1-6, the concept of multifactorial causation states that no cause acts alone: its effects depend on interactions with many other causes.)
When performing case-study research, the effects of extraneous variables may be controlled to some extent by comparing and contrasting case studies of people with similar problems in mental or physical functioning. For example, in case-study research on the effects of brain damage, researchers compare and contrast the case reports of people with damage to the same areas of the brain, and look for similarities that might indicate important extraneous variables. If Greg’s case had been included in such a study, the researchers probably would have found that he was the only one receiving this kind of religious training, which would have allowed them to rule out any influence of Greg’s “enlightenment” on the development of his symptoms. They also would have found that the most important similarity among the various cases was extensive damage to particular areas of the brain. They then would have concluded that this brain damage was the main cause of the problems observed in their patients.
When an extraneous variable affects research results in a systematic way (i.e., in an orderly or organized manner), we say that there exists a confound — a term that refers to the inability to distinguish the effects of an extraneous variable from the causal factor being studied. Their effects can’t be distinguished because the two factors are somehow “intertwined” or “bundled together.” The schizophrenia example illustrates what is meant by a confound: if the children of schizophrenic parents are raised by those parents, the effects of the genes they’ve inherited can’t be separated from the effects of how their schizophrenic parents raised them. Thus, the word confound is simply another way of saying that the effects of an extraneous variable have not been controlled for.
The major goal of controlled research is to eliminate confounds so that researchers can investigate the causal effects of the factor they are studying. Case-study research, however, is limited in its ability to do this for at least two reasons.
Reason 1: Researchers may be unaware of some of the important extraneous variables. It’s often impossible to know what all possible confounds might be because our knowledge of the world isn’t complete. For example, if we didn’t know that patents’ belief that a treatment will work may cause them to feel better, medical researchers wouldn’t know that they needed to include a placebo-control group.
Reason 2: Researchers may have biases that affect their observations. These biases are themselves extraneous variables that would be confounded with the effects of a causal variable. For example, if a medical researcher believes that a particular treatment is going to help her patients, then, if she knew which patients received the treatment, she may see them improving more than patients in the placebo-control group. This extraneous variable is called experimenter bias, and it’s difficult to control for in case-study research because the researcher is the one collecting all the information about each patient.
In the next section, you’ll learn about another method of performing research — the correlational study — that has some important advantages over case-study research.
Study Questions for Section 3-2
- How would you define, in your own words, a case study?
- When are case studies most likely to be performed in psychology?
- What is a strength of case studies?
- How would you define “extraneous variable” in your own words?
- If I claimed that the “number of hours you study” determines how well you will do on a test, what would be some important extraneous variables to control for?
- How can case-study research be used to rule out the effects of extraneous variables?
- What is a major weakness of case studies?
- How would you define a “confound” in your own words?
- If I claimed that the “number of hours you study” determines how well you will do on a test, what would be some important confounds?
- What is the major goal of controlled research?
- How would you define “experimenter bias” in your own words?
Gottesman, I. I. (1991). Schizophrenia genesis: The origin of madness. New York: Freeman.
Sacks, O. (1995). An anthropologist on Mars: Seven paradoxical tales. New York: Knopf.
Sullivan, P.F. (2005). The genetics of schizophrenia. PLoS Medicine, 2, 0614-0618. doi: 10.1371/journal.pmed.0020212