Discussion: Threats to Validity
Discussion: Threats to Validity
There are hallmarks that are necessary in addressing a research problem. In a quantitative approach, this concept includes making the research topic amenable to scientific study. Hence, in framing the problem, researchers must ensure that a scholarly, systematic method of inquiry can be applied to the study (Walden, 2015). I addition, a researcher must maintain scholarly objectivity to mitigate against reliability and validity issues. In this context, reliability denotes a consistency in result from a research instrument, strategy, or approach (Burkholder, Cox, & Crawford, 2016). Overall, a research design must allow for multiple possible conclusions and mitigates against other threats.
Threats to Validity
Validity is a significant aspect of a research design. As it relates to quantitative study, validity denotes the best existing estimation to the truth regarding a proposition (Cook & Campbell, 1979, as cited by (Burkholder, et al., 2016). Therefore, the quality of a research is primarily dependent on the validity of its findings. This approach includes the type of data collected and how that data is used to answer the research question (Burkholder, et al., 2016). In the context of causal mechanism, there are threats that undermines the validity framework—internal and external.
Quantitative research needs to address the internal validity of the target research question(s). Internal validity questions the truthfulness of a given proposition, regarding how a change in one variable causes a change in the outcome (Burkholder, et al., 2016). When there is causal inference, there are also competing explanations—threats to a statement’s validity. In the context of quantitative research, one of the threats to internal validity is selection. This concept refers to the process of selecting participants—self-selection or researcher sampling and assignment procedures (Shadish, Cook, & Campbell, 2002). The problem occurs when “cause and effect” is disputable, because of systematic differences across conditions.
A quantitative approach in research also involve addressing external validity. In this context, the primary issue with external validity is ensuring that research findings hold true across contexts (Burkholder, et al., 2016). However, as with internal validity, there are also threats to external validity. Relative to quantitative research, one of the threats to external validity is the interactions of the observed causal relationship with sample units (Burkholder, et al., 2016). In this case, the results of one particular sample in a research, may not hold across other samples.
Strategies to Mitigate Threats
The quality of a quantitative research necessitates the mitigation of threats—internal and external. A fundamental quality indicator for research addressing causality is the presence of a control condition (Shadish, et al., 2002). This approach is especially significant, for causal inferences. Regarding selection, one approach to mitigating threats to validity is the use of random assignment—coin flips, computer algorithms, etc. (Burkholder, et al., 2016). Mitigating threats to external validity is equally important. Hence, the interactions of the observed causal relationship with sample units can be addressed by strategically incorporating a relevant design and methodology. One approach denotes including a qualitative case study analysis that would elucidate the degree to which these findings might generalize to other contexts (Burkholder, et al., 2016). Hence, mitigating threats to validity impacts research design and is reliant on the strategic incorporation of various methodologies.
Potential Ethical Issues
Potential ethical issues in social science often influence design decisions. Most of these ethical issues involves the methodologies employed (Babbie, 2017). Relative to quantitative research, it is not unusual to select participants—human subjects—for a particular study. The Belmont Report asserts three key principles—Resect for persons, beneficence, and justice—that must be considered (Babbie, 2017). Regardless, there will always be some risks to someone, when conducting research. Nevertheless, some designs are more feasible in mitigating risks than others (Babbie, 2017). Thus, researchers should consider a design that will best safeguard against these risks. One approach, regarding design decisions is to structure the research in a manner that guarantees anonymity.