Random and Systematic Bias in Population Oral Health Research: an introduction

Jamieson LM

Random and Systematic Bias in Population Oral Health Research: an introduction

Authors: Jamieson LM
doi: 10.1922/CDH_SpecialIssueJamiesonIntro01


Bias in population oral health research is a form of systematic error that can affect scientific investigations and distort inference (i.e under or over confidence in an estimate). A biased study loses validity in relation to the degree of the bias. While some study designs are more prone to bias, its presence is universal. It is difficult to completely eliminate bias; in the process of attempting to do so, new bias may be introduced or a study may be rendered less generalizable. The goals are to therefore minimize bias and for investigators and readers to comprehend its residual effects, limiting misinterpretation and misuse of data. In the four papers that follow, we seek to contribute to the discourse around random and systematic bias in population oral health research through the lens of case controlled studies, longitudinal studies and genomics re search. The papers formed the basis of a symposium entitled ‘Random and Systematic Bias in Population Oral Health Research’ at the 98th General Session of the International Association of Dental Research held March 2020 in Washington DC, United States. Using tangible examples, the collected authors describe the intractable role of bias in population oral health research; how to minimise, identify and articulate; demonstrate the increasingly sophisticated techniques for addressing measurement error and bias in population oral health research, including specific statistical software and codes; and discuss the implications of addressing (or not addressing) bias in population oral health research at an international level, including the role of advocacy and engaging with oral health policymakers to both minimize bias and to increase comprehension of its residual effects that may lead to misinformed policy. Mittinty describes simple methods for conducting sensitivity analysis for unmeasured confounders in his paper entitled ‘Estimating Bias Due to Unmeasured Confounding in Oral Health Epidemiology’, with examples provided through case studies and vignettes. He explains how confounding arises when variables are associated with both exposure and outcome but are not on the causal pathway. Because unmeasured confounders can have a cumulative effect, can make associations seem bigger when the true effect is smaller (or vice-versa) or can make associations appear negative when they are actually positive, understanding and accounting for confounding is essential. Duran and colleagues, in their paper entitled ‘Quantitative Bias



Online (Single user only)
Print & Online (Single user only)
Institution Online (IP address validation)

Back issues may be obtained from the publisher

Consider recommending subscription to your institution's library

You can view Open Access papers without a subscription.