Research Methods and Statistics

CHAPTER PREVIEW

If you want to study behavior, it is useful to be able to describe and predict behavior, but it will be more satisfying to know why people act the way they do. It is relatively easy to observe different kinds of behavior and, from there, to make predictions about other behaviors. Most of us have a general sense of how people are going to act in certain circumstances (although we are fooled often enough). The next goal in psychology is to understand the causes of behavior.

In research, we choose experimental designs when we want to discover causation. Descriptive approaches can be quite useful for making predictions about behavior, but they do not inform us about the underlying reasons for those behaviors.

In the simplest experiment, the researcher creates a treatment group that will be compared to an untreated, or control, group. If the two groups start equal but end up different, we presume that the treatment made a difference. In practice, most studies employ more than two groups, but the logic is the same regardless of the number of groups.

Complications arise in any research project because small details of the research situation often have effects on participants’ behaviors that we don't anticipate or even recognize. Further, because an experimental session involves an interaction between people—an experimenter and a participant—social effects can contribute to changes in behavior.

Choosing a Methodology: The Practicalities of Research

In psychology, the word experiment has a specific meaning. It refers to a research design in which the investigator actively manipulates and controls variables. Scientists regard experimental methods of research as the gold standard against which we compare other approaches because experiments let us determine what causes behavior, leading to the ultimate scientific goal—control. In general, researchers often prefer experiments over other methods such as surveys, observational studies, or other descriptive and correlational approaches even though studies that describe and predict behaviors provide important information about thought and behavior.