Memory Loss & the Brain

Research is a procedure of careful scientific investigation. For example, research on memory may involve studying how different factors affect memory in normal individuals, studying brain activation during various learning tasks (as through MRI or PET studies), studying memory abnormalities in patients with damage to particular brain regions, studying the effects of a particular drug on memory, and so on. Research on the brain and memory can lead to a better understanding of normal memory and of memory disorders, which in turn can lead to the development of drugs and therapies to treat and prevent those disorders.

The validity of research should be evaluated according to several criteria, including replicability, proper use of controls, and universality. These criteria are especially important in studies involving human subjects but the same basic principles apply to all forms of research.

Replicability. Ideally, after one set of researchers completes a study and reports their findings, the study should be repeated by other, independent researchers to confirm that the same findings are obtained reliably. If results obtained by two different researcher groups do not match, it is important to establish why. For example, did the two studies use patients who were on different kinds of medication? Did they consider different proportions of male and female patients? Were there differences in the testing procedure that could have influenced the results? These differences must be resolved before the results can be accepted as valid.

Proper use of control conditions. Control conditions are the experimenters' attempt to rule out other factors which could have contributed to the pattern of results. There are several applicable procedures. Most involve giving one group of subjects some behavioral treatment and a second group of subjects behavioral treatment which differs only in one particular detail. For example, experimenters testing a new drug for headache relief can give some subjects the experimental drug and some subjects a placebo or sugar pill; only if the experimental drug causes significantly greater relief than the placebo can it be concluded that the drug is effective for headache treatment.

(Notice that a placebo does not have to be a literal pill. For example, in a research study to determine the effectiveness of a particular kind of teaching method, one group might receive one hour daily of the experimental method, while a second group might receive one hour daily of a different kind of therapy which is not believed to influence learning.)

In the most carefully controlled studies (called double-blind studies), the experimenter is unaware of which subjects are receiving which treatment. This eliminates any chance of the experimenter's own expectations biasing the evaluation of subjects' behavior. Only at the end, when the results have all been collected, is it revealed which subjects received which treatment.

Universality. A study which considers, for example, a small group of similarly-aged healthy white males may report findings which do not apply to a large group of people, or to people of different ages, genders, ethnicity or medical history. A well-designed research study should attempt to include as many participants as possible, with as broad a demographic profile as appropriate, to ensure that any results apply to a wide range of people. (Obviously, in some studies it is appropriate to limit types of participants: in a study of menopause, it is appropriate to consider only female subjects; in a study of Alzheimer's disease, it is appropriate to exclude individuals who have other neurological problems; etc.)