Mook’s Defense of External Validity

External validity may be defined as the extent to which the result of a study can be generalized to population, environment or time. Based on Mook (1983) argument about external validity, it is plausible to concur with his point of view regarding generalization and external validity. According to his article “In Defense of External Invalidity” a researcher should be concerned about the external validity (EV) if the aim of an experiment is predicting behaviors in a real world situation. He adds that the relevance of EV is based on the researcher’s willingness to make a generalization. However, he suggests other experiments where generalization may not be of concern. Among them include first, a study which aims at establishing whether something can happen or it usually does occur. Second, a study in which the researcher create conditions that cannot occur in the real world. Third, a research that aims at establishing whether something that is supposed to happen in the lab really does happen. Moreover, Mook emphasized that if a researcher intends to generalize his study, the sample, setting, and manipulation ought to be representative of the target. Additionally, he forms a solid ground based on illustrations regarding how the three conditions may affect the conclusion of a researcher while generalizing his study.

One of the reasons for agreeing with Mook argument is his emphasis on the randomness of the sample that is used in the study. According to Radhakrishna (2008), the sampling method used to select the sample determines whether the outcome may be generalized to the target. The randomness of selecting the sample ensures that the sample selected is not biased and hence representative of the target population. Radhakrishna argument is in line with Mook’s argument since Mook claims that sampling method in the vital determining representativeness of the sample. Additionally, the high rate of non-response may hinder generalization of the experimental outcome. In the article by Radhakrishna (2008), lower response rate results to skepticism about the representativeness of the sample. The non-response rate may be due to change in the ideal setting or total ignorance. The ignorance is exhibited in Mook illustration where he cites that a policeman may fail to respond to the researcher since there is no command but the same may not happen to while responding to his superior.

Another supporting perspective is offered by Ferguson (2004) who, like Mook, posits that not all research studies are conducting with the intention of generalizing to the population. Once an experiment is conducted and the results obtained, it is imperative that the inference made thereof be clearly stated as to whether the intention was to approve or otherwise disapprove the hypothesis. In this instance, the results or rather the experiment does not require the concept of generalization; therefore, external validity is irrelevant. Nevertheless, an alternative conclusion could be made using the same experiment with the intention of generalization. As such, it would be imperative for the researcher to be clear on the representativeness of the chosen sample, settings as well as manipulations made in the experiment. Additionally, Ferguson (2004) postulates that despite internal validity and other scientific standards compliance in an experiment, generalizability of results should not be obvious since ideal natural conditions may not be replicated in the laboratory setting, thereby affecting the research findings.

In addition to the existing evidence that generalization is contingent on a number of factors and not the sole reason for conducting experiments, Lynch (1999) states that external validity is not much about the methodology utilized in the study. Rather, he emphasizes that external validity is dependent of theoretical and conceptual frameworks and discussion of the results obtained. In light of this, he claims that if these sections of the research work do not map a laboratory setting to the real world situation, then a generalization of the results can be irrelevant without further evaluation.


Ferguson, L. (2004). External Validity, Generalizability, and Knowledge Utilization. Journal of Nursing Scholarship, 36(1), 16–22.

Lynch, J. G. (1999). Theory and External Validity. Journal of the Academic of Marketing Science, 27(3), 367–376.

Mook, D.G. (1983). In Defense of External Invalidity. American Psychologist, 38, 379–387.

Radhakrishna, R. (2008). Strategies for Generalizing Findings in Survey Research. Journal of Extension. 46(2), 62.

One clap, two clap, three clap, forty?

By clapping more or less, you can signal to us which stories really stand out.