In murder mysteries, circumstantial evidence often leads to a false accusation before the true killer is found. In epidemiology, we call this phenomenon â€œconfounding.â€ Confounding occurs when researchers determine through their analyses that an association exists between a variable and an outcome. In reality, however, the true exposureâ€“outcome relationship has been masked by a third, confounding variable.
For this Discussion, you identify potential confounders of epidemiological associations. You also determine how to assess and minimize the confounders through study design and analysis considerations.
Select one of the following five topics and consider the possible associations between them:
- School lunches and occurrence of childhood obesity;
- Frequency of physical activity and occurrence of cardiovascular disease;
- Infections and mortality in hospitalized infants;
- Geographical location and the occurrence of breast cancer;
- Osteoarthritis and the occurrence of falls
Post what a confounding is, a description of the confounding variables that you might need to consider before designing the study. Then explain why you need to consider these variables. Finally, using an observational study design, explain how you might minimize confounding variables in your study in both the data collection phase and the analysis phase.