External Validity

Do the Results Generalize? In addition to guarding internal validity, researchers are concerned about external validity, the extent to which the results obtained under one set of circumstances would also occur in a different set of circumstances. When an experiment has external validity, its findings can be assumed to generalize to other people and to other situations. Both the participants in the experiment and the setting in which it takes place affect external validity.

To help increase external validity, social psychologists would love to conduct their experiments with huge samples of participants that are representative of the general population. Usually, however, they must rely on convenience samples drawn from populations that are readily available to them, which explains why so much research is conducted on college students. There are very practical reasons for the use of convenience samples. Representative samples are fine for surveys that require short answers to a short list of questions. But what about complex, time-consuming experiments? The costs and logistical problems associated with this would be staggering. Advocates of convenience samples contend that the more basic the principle, the less it matters who participates in the research. For example, people from different cultures, regions, and ages might differ in the form of aggression they typically exhibit when angry, but the situational factors that cause people to be more likely to aggress—in whatever way that aggression is expressed—may be similar for most individuals across time and place. Yet in spite of these argu- ments, having the most diverse, representative samples of research participants as possible is ideal. The growing interest in cross-cultural research in the field is certainly one step in the right direction.