Epidemiological studies are designed to evaluate a hypothesized relationship between an exposure and an outcome. However, the existence and/or magnitude of these relationships may be erroneously affected by the design and execution of the study itself or by conscious or unconscious errors made by the investigators or the subjects. These systematic errors are called biases. If not avoided or accounted for, biases can completely invalidate the results of an otherwise well-thought-out study.

Are you more of a visual learner? Check out our online video lectures and start your statistics course now for free!

Richard Forshee, Ph.D., Associate Director for Research, Office of Biostatistics and Epidemiology in FDA's Center for Biologics Evaluation and Research explains the importance of quality Infographic Designing in medical research and public health.

Image: “Richard Forshee, Ph.D., Associate Director for Research, Office of Biostatistics and Epidemiology in FDA’s Center for Biologics Evaluation and Research explains the importance of quality Infographic Designing in medical research and public health.” by The U.S. Food and Drug Administration – https://www.flickr.com/photos/fdaphotos/25491597604/. License: Public Domain


Definition of Bias

Definition

A bias is an error in the design, conduct, or analysis of a study that results in a deviation of the test statistic from its true value. This results in an incorrect estimate of an association between an exposure and the target population.

  • Biases cannot be completely eliminated but they can be reduced/minimized through proper study design.
  • Two major subtypes of bias:
    • Selection bias: Bias caused by some preference in the way subjects are chosen which makes the study groups different from the target population.
    • Information bias: Bias that results in a systematic error in the way data is measured or collected.

Video

Definition of Bias by Raywat Deonandan, PhD

Effects of Biases
Masking Hides an existing relationship between the independent variable and a dependent variable
False relationships Create spurious relationships between the independent variable and a dependent variable
Overestimation Exaggerates the size of an existing relationship between the independent variable and a dependent variable
Underestimation Diminishes the size of an existing relationship between the independent variable and a dependent variable

Table 1: Possible effects of biases. 

Selection Bias

Definition

Selection bias is the error introduced when the study population does not represent the target population due to some selection preference. 

  • This results in a nonexistent association between the exposure and outcome.
  • A selection preference can exist: 
    • Between study groups
    • Between the selected and excluded participants for the study
  • Common examples of selection bias are the healthy worker effect, loss-to-follow-up bias, susceptibility bias, survival basis, and Berkson’s bias.
    • Healthy Worker Effect: Bias occurs because subjects recruited from a workplace environment tend to be healthier and more active than the general population.
      • This type of selection bias tends to lead to an underestimation of morbidity or mortality when compared to the general population.
    • Loss-to-follow-up bias: Occurs due to data loss from participants who are lost during the course of the study owing to loss of interest, adverse effect, clinical improvement, etc. 
      • Leads to a reduction of the effective sample size.
      • Results are biased if the study groups become unmatched or if the subjects who are lost to follow-up differ in exposure and/or outcome from those who remain in the study.
    • Susceptibility bias: In the cases where a first disease predisposes to a second disease, the treatment of subjects with the first disease erroneously appears to predispose to the second disease.
    • Survival bias: Occurs when data is collected only regarding subjects who have survived an outcome or have met some other clinical criteria related to it.
    • Berkson’s Bias: Occurs due to subjects being chosen from a specific segment of the population that has more or less an intrinsic risk of having the outcome compared to the general population. 
      • Also called Berkson’s Fallacy or Berkson’s Paradox
      • For example, hospitalized patients have a greater risk for disease than the general population regardless of exposure. Thus, in a study looking at the link between smoking and COPD, choosing hospitalized subjects for the experimental group will result in a greater association than would be for the general public.
      • The higher morbidity of hospital patients relative to the general population can create spurious associations between diseases.

Videos

Selection Bias by Raywat Deonandan, PhD

Healthy Worker Effect by Raywat Deonandan, PhD

Loss to Follow-up Bias by Raywat Deonandan, PhD

Berkson Bias by Raywat Deonandan, PhD

Learning Outcomes I – Statistical Biases by Raywat Deonandan, PhD

Information Bias

Definition

Information bias results from systematic errors in the measurement of some exposure, outcome, or variable. The major types of information bias are misclassification bias, recall bias, interviewer bias, response bias, reporting bias, observer bias, ascertainment bias, and confirmation bias. 

  • Misclassification bias: Results from the misclassification of exposure or health outcome for subjects in a study. People that do not have a disease are classified as having it, and vice versa.
    • Common reasons are inaccurate records, different disease definitions, or different diagnostic criteria.
    • Differential: Caused by a measurement difference that exists between study groups.
    • Non-differential: Caused by equally inaccurate or random measurements across all study groups.
  • Recall bias: Caused by the differential (typically, enhanced) recall of events by case subjects compared to controls.
  • Interviewer bias: Occurs when the interviewer influences the responses of the participant through the method, content, or style of his or her questioning.
  • Response bias: Occurs when subjects report what the researchers want to hear because the subject wants to please the interviewer with his or her response.
  • Reporting bias: Refers to the tendency of researchers to only report statistically significant results motivated by the desire to publish.
  • Observer bias: Occurs due to the differences that exist between observers in the way an outcome is measured or assessed or the difference in the way a subject acts when he or she is being observed.
    • Bias that results from the Hawthorne effect (see Section 4 below).
  • Ascertainment bias: Occurs when certain subjects are more likely to be included in the final results than others.
    • Common sources include a group difference in the screening method or frequency, investigator knowledge of subject group assignments, or allocation preference for subjects into study groups.
  • Confirmation bias: Occurs due to the tendency of investigators to only include data that agrees with their pre-existing beliefs and/or supports their hypothesis.
  • Detection bias: Occurs due to a difference in the way an outcome is measured or detected based on a particular subject characteristic (see Image 1).
  • Experimenter bias: Occurs due to a subject behaving differently based on the investigator’s expectations.
    • Bias that results from the Rosenthal effect (see Section 4 below).

Videos

Information Bias by Raywat Deonandan, PhD

Misclassification Bias by Raywat Deonandan, PhD

Response Bias by Raywat Deonandan, PhD

Detection Bias by Raywat Deonandan, PhD

Learning Outcomes II – Statistical Biases by Raywat Deonandan, PhD

Effect Modification by Raywat Deonandan, PhD

Learning Outcomes III – Statistical Biases by Raywat Deonandan, PhD

Image 1: Example of detection bias. Shown above is a graph of the number of new AIDS cases per year per 100,000 population from 1990–2000. Although the chart seems to suggest that the cases of AIDS in the Caribbean are on the rise (orange line), in reality, the steady increase in new cases was due to the increase in screening number and frequency. This is an example of detection bias, a subtype of information bias. By Lecturio.

Hawthorne and Rosenthal Effects

  • Hawthorne effect: Refers to the tendency of subjects in a study to behave or act differently (i.e. work harder) when they know they are being watched. 
    • Results in and is related to observer bias
    • Especially common in psychiatric studies
    • Bias caused by this effect is difficult to eliminate but can be reduced through hiding from the subjects the knowledge of when/how they are being observed.
  • Rosenthal effect: Refers to the tendency of subjects or investigators to behave differently based on other’s expectations.
    • Also called the Pygmalion effect
    • For example, in a study of the efficacy of a new drug to treat depression, subjects in the treatment group report an improvement in depressive symptoms because the investigators who are assessing them expect subjects who are given the treatment to improve.
    • Results in and is related to experimenter bias
    • Bias that results from this effect can be prevented by blinding.

Videos

Learning Goals – Statistical Biases by Raywat Deonandan, PhD

Hawthorne and Rosenthal Effect by Raywat Deonandan, PhD

Confounding Variables

Definition

A confounding variable is an additional variable other than the independent variable which has an effect on the dependent variable causing an erroneous relationship to be inferred between the independent and dependent variable.

  • Technically, a confounding variable is not a bias.
  • Confounding can be accounted for through proper study design. The following techniques help to reduce the effect of confounding:
    • Randomization: Subjects are randomly assigned to study groups so that confounding variables are spread equally between groups.
    • Controlling: All subjects are chosen so that they have the same confounding variables.
    • Matching: Each subject with a confounding variable in a particular group is matched to a subject in the other groups with the same or similar confounding variable.   
  • Confounding variables are common in observational studies since it is more difficult or impossible to control for such variables.

Confounding variables are different from effect modifiers, a third variable that contributes to the true relationship between exposure and outcome

Videos

Confounding Bias by Raywat Deonandan, PhD

Effect Modification by Raywat Deonandan, PhD

Image 2: A diagram detailing how a confounding variable is related to the exposure and outcome. The confounding variable is related to the exposure and may contribute to or cause the outcome. If not accounted for, it can contribute to or affect the magnitude of the observed (true) relationship. By: Lecturio.

Do you want to learn even more?
Start now with 1,000+ free video lectures
given by award-winning educators!
Yes, let's get started!
No, thanks!

Leave a Reply

Register to leave a comment and get access to everything Lecturio offers!

Free accounts include:

  • 1,000+ free medical videos
  • 2,000+ free recall questions
  • iOS/Android App
  • Much more

Already registered? Login.

Leave a Reply

Your email address will not be published. Required fields are marked *