Lecture 26: Information bias

Question 1

What is information bias?

Answer 1

“Observation or information bias results from systematic differences in the way data on exposure or outcome are obtained from the various study groups”

Question 2

How can measurement error occur?

Answer 2

Participants provide inaccurate responses
* E.g. they may forget past exposures
* E.g. they may under or overestimate their exposure

Data is collected incorrectly/inaccurately
* E.g. problem with measuring device
* E.g. person collecting the data doesn’t follow the same procedure for all participant

Measurement error can be random or systematic

Question 3

Can you identify random or systematic error on a sheet (slide 11)?

Answer 3

YES or NOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO

Question 4

What is the difference between random error and systematic error?

Answer 4

Random Error is unpredictable and affects precision, whereas Systematic Error is consistent and affects accuracy.

Random error can be minimized by averaging across a larger sample, while systematic error requires changes in methodology, recalibration, or improved study design to address it.

Both types of error can occur in a study, but identifying and reducing systematic error is crucial for accurate, unbiased results.

Question 5

What effect might measurement error have? In a descriptive study?

Answer 5

In a descriptive study:
* Could over/underestimate prevalence of exposure and/or outcome

Question 6

What effect might measurement error have? In an analytic study?

Answer 6

In an analytic study:
* Can lead to misclassification

• People without the exposure may be classified as having the exposure (and vice versa)
• People without the outcome may be classified as having the outcome (and vice versa)

Question 7

What are the two types of misclassification? What one is focused on in this paper?

Answer 7

Non-differential misclassification
“When measurement error and any resulting misclassification occur equally in all groups being compared, they are described as being non-differential…”

Differential misclassification (THIS IS WHAT WE FOCUS ON)
* ‘Different’ between the study groups e.g. exposed/comparison group, or cases/controls

Question 8

What are some examples of differential misclassification in the different study types? (Cross sectional, Case Control and Cohort study)

Answer 8

Cross sectional:
In a cross-sectional study, people with the outcome might report the exposure differently to those without the outcome

Case Control:
➢In a case-control study, cases might more accurately recall past exposures compared to controls
➢In a case-control study, an interviewer who is aware they are interviewing a case might ask more probing questions about the exposure of interest

Cohort study:
➢In a cohort study, an interviewer aware of the exposure status may ask more probing questions about the outcome among those exposed compared with those in the comparison group

Question 9

What is recall bias?

Answer 9

“Systematic error due to differences in accuracy or completeness of recall to memory of past events or experiences”

Question 10

How can we minimse recall bias?

Answer 10

➢Objective measures
➢Validate self-reported measures with other information
➢Memory aids

Question 11

What is interviewer/observer bias?

Answer 11

if interviewer/observer knew the exposure status and examined the outcome differently for those in the exposed group compared with those in the comparison group

Question 12

How can we Minimise interviewer/observer bias?

Answer 12

Examples:
➢Clearly defined study protocol and measures
➢Blinding
➢Training of interviewers
➢Structured questionnaire and standard prompts

Question 13

What information bias can arise in a RCT and how are these controlled?

Answer 13

• Bias could occur if knowledge of the treatment/exposure category influences the assessment of the outcome
  ➢Ensure there is blinding
• Bias could occur if measurements are undertaken differently for different treatment groups
  ➢ Ensure measurements undertaken in the same way

Question 14

How can we minimise measurement error - information bias?

Answer 14

Collecting information from participants:
- Validated survey instruments
- Validate using objective measure

Measurement instruments:
- Use standardised equipment
- Use calibrated equipment

Question 15

How to minimise information bias (more interviewer bias)?

Answer 15

Collecting information via interviewers/observers:
- Ensure blinding
- Training of interviewers
- Use structured interviews and standardised ‘prompts’
- Use objective measures

*Clearly defined study protocol
*Well-defined exposures, outcomes and other factors collected in

Question 16

What is publication bias?

Answer 16

The result of the tendency of authors to submit, organizations to encourage, reviewers to approve, and editors to publish articles containing ‘positive’ findings…, especially ‘new’ results, in contrast to findings or reports that do not report statistically significant or ‘positive’ results”

Basically a lot of studies are not published if they are not ‘revealing’ as it is expensive/not interesting and thus publication bias arises.

Question 17

Explain why objective measures of exposures and outcomes can reduce information bias compared to subjective measures.

Answer 17

Objective measures of exposures and outcomes are independent measures that do not rely on a person’s memory (recall) and/or judgement.

It important to note that objective measures are also not subject to the reluctance of participants to reveal sensitive or embarrassing information related to exposures and outcomes.

For example, using medical records to assess a participant’s exposure history or outcome may supply more accurate information than self-report, without discomforting or distressing the participant or relying on the participant to remember accurately.

Question 18

Describe the information bias that may result when comparing self-reported
exposures in cases and controls.

Answer 18

Recall bias:

When using self-reported (subjective) measures, cases may be more likely to remember past exposures more accurately compared to controls (cases may also exaggerate or underestimate their exposure). Controls, because they do not have the outcome of interest, do not have the same motivation as cases to remember exposure information.