Abnormal psychology: characteristics and theoretical frameworks

¶ … ABNORMAL PSYCHOLOGY: WHAT DO CLINICAL RESEARCHERS DO?

In the past, our most knowledgeable people have gotten some things very wrong. Aristotle called the brain "an organ of minor importance," and in 1984, genetic researches announced that mammals could not be cloned. Clearly we have things left to learn about humans. This is what makes research so important. The example given in the book of the lobotomy is very important: experts believed it worked when it did not, but before this was recognized, surgeons damaged many thousands of people even more, making their situation worse instead of better. This chapter describes many research approaches.

A case study focuses on one individual. The researcher describes the person's history, family, how the person's problems evolved and developed, and usually, the treatment and how well that treatment worked. This was an approach Sigmund Freud used at least sometimes. He did not know the child well but gathered detailed information from the child's family in the "Case of Little Hans." This approach can document new approaches to old problems. However they are vulnerable to the bias of the reporter, and it is not wise to generalize to others based one case.

Case studies are important because they put a real face to a label. "Little Hans" isn't just another anxiety-driven individual. He becomes a real, breathing individual who did not happen by accident and who is more than the sum of all the vents that molded him. However, case studies will be biased by the observer's views. Freud saw early childhood as the cause of many psychiatric disorders, so that's what he looked for in his case studies. However, case studies can also be used to challenge the status quo. In addition, case studies can look at individuals who seem to present unique difficulties.

The Correlational Method: This approach looks at relationships between two events. The example given, whether stress leads to depression or depression leads to stress, reminds the reader of the old debate, "Which came first -- the chicken or the egg?" Those in such a study are called subjects or participants, within defined restraints: all children, all adults, all female, all male, etc. If they find two events together, this is called a "correlation." This does not explain the cause of the correlation, but researchers can look at the data and determine the probability that the two events are related.

Correlational studies can find a negative as well as a positive connection. The example given is between depression and activity -- depressed individuals tend to be less active than those not depressed, but a causal connection can only be guessed. Maybe depression makes us less active; maybe lower activity leads to depression. Maybe there are other influences not identified in the study. We can't tell from a correlational study. A statistical analysis, the correlation coefficient, tells the researcher how close the relationship between the two variables are. Epidemiological studies look at correlations.

The Experimental Method: Correlations just look at events. In the experimental method, events are manipulated and changed to see what the outcome will be. The factors manipulated are called variables. The variable observed is the dependent variable, and the variable that is manipulated is the independent variable. The researcher guesses that changing the independent variable will change the dependent one. The researchers have to be careful to identify everything that might change the outcome. To avoid this they use a control group, random assignment to those groups, and a "blind" design. Those in the control group do not receive the independent variable to see if the independent really is the cause of any change in the dependent one. They will receive a "placebo," or something not likely to not change the dependent variable. "Blinding" helps eliminate researcher bias. The person evaluating the study does not know which participants received the independent variable and which received the placebo. In fully blinded studies, the researchers themselves don't know who gets the placebo. An experimental design can follow up a correlational study to start idenfitying how the two correlated events are related. Blinded studies are often used to test medications.

Alternative Experimental Designs: Sometimes not every variable can be controlled, and sometimes, natural groups of possible participants already exist. For example we cannot abuse children so we can study abused children afterwards. We have to find a group of children who have already suffered that. It is harder to control such groups from other variables, such as economic or racial status. Other groups might be survivors of some disaster.

Quasi-experiments, or mixed designs, do not control for every variable. Their advantage is that they use real-life events without any changes applied to them. One way to handle the lack of controls is to match each affected participant with an unaffected applicant. So, every abused child might be matched with a child of similar age, sex, intelligence, economic and racial background who had not been abused.

One choice researchers have is use an analogue design, such as working with animals, and subject them to the independent variable. This is done for early drug studies. Skillfully done, animal studies can provide valuable information about human functioning.

Single-Subject Experiment: This form may be used if a person has particularly unusual circumstances that makes him or her a "class of one," or it would be very difficult to find similar subject to add to the study, or the information sought is how an intervention affects that one person. One approach is the "ABAB" design, where A represents one set of circumstances for that person and B. another. If the question is how the person responds to a medication, A might be the medication and B. A placebo.