Lecture #20 - Bias Pt2

Question 1

What is bias?

Answer 1

Any systematic error in an epi study that results in an incorrect estimate of the association between exposure and risk of disease

Question 2

What is information bias?

Answer 2

Observational or information bias results from systematic difference in the way data on exposure/outcome are obtained from the various study groups

Question 3

How is data collected in a study and can error occur in both these methods?

Answer 3

By participants (e.g self report) or collected or measured by someone else

Question 4

How can measurement error occur? (Can it occur in both?) And what kind of measurements are the two?

Answer 4

Participants provide inaccurate responses (subjective - E.g can’t remember clearly)
-it’s self reported info so it’s SUBJECTIVE - susceptible to error bc involves judgement to be made. But some things have to be measured subjective e.g pain

Data is collected incorrectly/inaccurately
-its objective because don’t require that judgement to be made
E.g problem with measuring device (may not have been collaborated right)
Or person collecting data doesn’t follow the same procedure for all participants (we want it to be collected in the same way for everyone)

Question 5

Measurement error - what kind (systematic or random) and why is it a problem?

Answer 5

• measurement error can be random or systemic (random is lack of precision but systemic is lack of accuracy)
• findings could be overestimated, underestimated or not affected. This could lead to bias.

Question 6

What effect can measurement error have in a descriptive study? What about in an analytic study?

Answer 6

1. Descriptive study: could over/underestimate prevalence
2. Analytic study: can lead to misclassification - people without exposure can be classified as having exposure or people without outcome may be classified as having outcome. This misclassification could be due to a random or systematic measurement error

Question 7

Information bias in cross-sectional studies:

1. What is this study? Give a brief low-down
2. Measures what?
3. What two questions regarding information bias would you want to ask yourself?

Answer 7

-

Question 8

What two types of misclassifications are there?

With the first type, what happens to the estimate with regards to the null value?

What’s the definition of non-differential misclassification?

Give me four examples of differential misclassification (cross sectional study, case-control study, case control and cohort and also kinda RCT)

Answer 8

–

Question 9

Cross-sectional studies and info bias:

1. People with outcome might report what differently to who?
2. What about the survey example?

Answer 9

-

Question 10

Case-Control studies and info bias:

1. Problems with what so what type of bias?
2. Define this bias
3. Read through the example that explains this
4. What are the three steps to consider
5. Three ways to minimise recall bias?

Answer 10

-

Question 11

Cohort studies and info bias:

1. Potential for misclassification of what?
2. What are two things to consider carefully?
3. Two things about differential misclassification - what are they? Black point about each plz (what is this special type of bias called?)
4. 4 ways to minimise this type of bias

Answer 11

-

Question 12

Randomised controll trials and info bias:

1. Bias could occur if knowledge of what could influence the assessment of outcome?
2. Bias could occur if measurements are undertaken…..
3. Five main

Answer 12

-