Reply To Biostatistics Discussion- Confounders

Reply to Discussion (250-275 words) APA format, scholarly sources

After having the opportunity to do some researcher on Confounders, I learned that they are essentially bonus variables that are not necessarily accounted for in a study. According to Trident American InterContinental University (2020), confounders are “linked with a disease, is related with a risk factor, and changes the relationship between the exposure and outcome”. Consider the relationship between consuming alcohol and lung cancer. While there is limited research that supports the idea that alcohol consumption causes lung cancer, there is sufficient evidence that shows that alcohol consumption and tobacco use go hand in hand. In fact, Cancer Treatment Centers of America (2018) states that roughly 70% of alcoholics are also heavy smokers. While alcohol seemingly has no direct link to lung cancer, they do share an indirect relationship with one another.

Confounders can be tricky and researchers must be cognizant of their existence as they can have hidden impacts on an experiment’s outcome such as confounding bias. For example, if a confounding variable is present and presents bias in a way that detracts from the null hypothesis, then one may experience positive confounding (Glen, n.d.) . However if bias is presented in a way that favors the null hypothesis, one will experience negative confounding (Glen, n.d.).

In order to control for confounding variables, researchers must be sure to pay close attention to study design and consider all elements of the study. It is also important to introduce controls so as to eliminate biases. For example, if one finds that age is a confounding variable in a study, they can consider only sampling from one age group.

Glen, S. (n.d.) Confounding Variable: Simple Definition and Example. Retrieved from https://www.statisticshowto.com/experimental-design/confounding-variable/

Trident American InterContinental University (2020). Module 4 Home. Univariate vs. Bivariate Analyses and Regression. Retrieved from https://tlc.trident.edu/d2l/le/content/149416/viewContent/3641402/View