Chapter 31 - Biases in research studies

Abstract

In all aspects of scientific research, there are multiple sources of bias. They can be present at each stage of your research study, and when present can lead to some deviation from the truth in data collection, data analysis, and interpretation that can subsequently lead to false conclusions.1 Therefore, it is imperative to make a strong effort to minimize biases to prevent loss of credibility to your research study. Unfortunately, some degree of bias in a study is nearly inevitable, so it is important to consider how bias might influence a study's conclusions.11 Learning the different types of biases and how they manifest in study designs is the first step to understanding the potential impact on study conclusions. This chapter will highlight the most common biases present in scientific literature and offer guidance on how to best avoid incorporating bias into the study design. Examples of biases discussed in this chapter include selection bias, reporting bias, performance bias, detection bias, and attrition bias.