DISCUSSION BOARD DUE THURSDAY 250-300 WORDS
This week you will review different types of bias, present an example of a study, and discuss whether bias was a factor in the study outcome. You will also discuss how the study design could have been altered to minimize or eliminate the risk of invalidating the results.
To prepare for this Discussion:
Review the types of bias listed below.
Non-differential recall bias
Differential recall bias
Loss to follow-up
Refusal to participate
Select one type of bias from the list above, and consider the ways that bias could impact a study.
To complete this Discussion, post a real or hypothetical example of one study. You can use one of the studies you designed during Weeks 3 or 4 or search for a different study.
How could bias be introduced in the study?
Would the bias be considered a selection bias or an information bias? Why?
Name two or three variables that might be possible confounders in your study.
Describe at least one method of controlling those confounding variables.
Epidemiology for Public Health Practice
Chapter 2, “Practical Applications of Epidemiology” (Causality in Epidemiological Research, pp. 84–97)Concepts about causality and risk in epidemiological research are covered in this section of the chapter.
Chapter 9, “Measures of Effect” (pp. 420–430)The authors explore statistical measures in determining absolute and relative effects in epidemiologic investigations and evaluating epidemiologic associations.
Chapter 10, “Data Interpretation Issues”Interpreting the findings of an epidemiological study can sometimes be challenging. There are, however, a few guidelines that can point you in the right direction. This chapter covers those guidelines.
Hills Criteria of Causation
This article explores the Hills Criteria of Causation, which outlines the minimal conditions needed to establish a causal relationship between two items. It is a method of determining a causal link between a specific factor(s) (e.g., cigarette smoking) and a disease (such as emphysema or lung cancer).