Bias and confounding

Question 1

Why do things go wrong in epidemiolgy?

Answer 1

Poor study design
Random error
Selection bias
Information bias
Interpretive bias
Confounding

Question 2

How can the effect of random error be reduced?

Answer 2

Increase sample size

Question 3

Bias occurs due to the way…?

Answer 3

Populations are sampled
Data is collected
Data is analysed
Results are interpreted

Question 4

What does increasing the sample size not reduce?

Answer 4

Bias

Question 5

What is selection bias?

Answer 5

When selected individuals are systematically different to those not selected for a study so the results do not accurately represent the population of interest

Question 6

Give some examples of selection bias

Answer 6

• Volunteer bias
• Non-response bias
• Incidence-prevalence bias
• Missing information in analysis
• Loss to follow up

Question 7

What is information bias?

Answer 7

When there is a misclassification of animals as diseased and non-diseased or inaccurate measurements of study factor

Question 8

What is information bias and give 2 examples of it?

Answer 8

Occurs when different emphasis is given to different evidence when results are evaluated

• Conformation bias
• Rescue bias

Question 9

What is conformation bias?

Answer 9

selectively reporting information that agrees with your prior convictions

Question 10

What is rescue bias?

Answer 10

discounting (uncomfortable) data by finding faults with study but applying less stringent criteria to studies which agree with prior convictions

Question 11

What is cofounding?

Answer 11

Occurs when the apparent association between a study factor and an outcome is distorted by the effect of a 3rd variable which is associated by the outcome and the study factor

Question 12

What are three requirements for a variable to be confounding?

Answer 12

• must be associated with the outcome
• must be associated with the exposure factor
• must not be an immediate step in a casual pathway