Validity threats must be considered when designing quantitative research studies. One threat to a studies internal validity is overall mortality. Overall mortality happens when some of those participating in the research take the pretest but drop out before the post-test (Burkholder et al., 2016). Overall mortality can skew data. There could be a characteristic about in those who dropped out. Change of this characteristic in the groups make up could be the reason for the difference between pre and post tests instead of the control variable (Burkholder et al., 2016). A strategy to mediate this would be using random assignment into control and experiment groups. Random assignment would help ensure that characteristics are evenly distributed between groups. Therefore, if a characteristic would increase the chances of dropping out, there should be an equal chance of drop out across groups. This way any change between pre and post-test data can be more confidently attributed to the independent variable.
Context-dependent mediation is a threat to the external validity of a study. Context-dependent mediation is the idea that a factor in one setting is not available in another setting (Burkholder et al., 2016). If this factor contributes to the change observed it makes the results less generalizable across demographics. To mitigate this threat, a researcher can do a thorough study of previous research in this area. A strong understanding of factors that are culture dependent, as well as a clear understanding of how that particular factor changes across cultures, will help with applying the results of the study across other demographics (Burkholder et al., 2016).
Beyond validity, there are ethical and political considerations to make when designing a quantitative research study. Deception is an ethical issue in many studies. Some studies require a form of deception for accuracy. For instance, a researcher may not want participants to know they are being observed because their behavior might be affected by this knowledge. Therefore deception about their identity may be necessary for accurate results (Babbie, 2017). This can change the design of a research study. A researcher must decide how to design their study to include the least amount of deception possible by creating questions and situations with these issues in mind. They must also plan mitigation strategies to combat the negative effects of the deception (Babbie, 2017). For instance, planing in a debrief session with participants after the experiment is over.
Research design requires that the topic be amenable to scientific study. This means that the design of the research is directly related to the topic being researched. The more amenable, the closer the research activities will be to the research topic (Burkholder et al., 2016).
Babbie, E. (2017) Basics of social research (7th ed.). Boston, MA: Cengage Learning.
Burkholder, G. J., Cox, K. A., & Crawford, L. M. (2016). The scholar-practitioner’s guide to research design. Baltimore, MD: Laureate Publishing.