Types Of Biases

Question 1

Name three sample biases

Answer 1

Healthy-user bias (sample from fitness center)
Berkson’s bias (sample from hospital)
Exclusion bias (excluding pt’s who have certain characteristics)

Question 2

What is the main result of sampling bias?

Answer 2

Decreasing generalizability.

Question 3

What does randomization of a sample into groups remove?

Answer 3

Selction bias - not creating groups based on certain characteristics or from a certain place.

Question 4

Why is it best to randomize people into equivalent groups?

Answer 4

Because this controls for confounding variables.

Question 5

What are methods (2) to minimize investigator bias?

Answer 5

Allocation concealment (blindling assigning groups)
Investigator "blinding" (the one who is evaluating or providing treatment doesn't know who is in what group)

Question 6

What is one example of subject bias?

Answer 6

Hawthorne effect

Question 7

What is incidence equal to?

Answer 7

Absolute risk (# people with disease/total number of people)

Question 8

Measurement bias

Answer 8

Occurs when the methods of measurement are dissimilar among groups of PTs

Question 9

Bias - NOTE whenever something is done incorrectly in constructing a study. General - what is selection bias?

Answer 9

Occurs when comparisons are made between groups of patients that differ in more ways than the one that you are studying.

Question 10

Dealing with drop-outs - best is intention to treat.

What is this?

Answer 10

Include the measurement even if they dropped out. Use this data point, based on their original assignment

Question 11

What is selection vs. sampling bias?

Answer 11

Sampling bias: how you make sample population. Did you choose from a certain place or a specific group that doesn’t generalize to the population?
Selction bias: how you make the groups. Have you chosen who goes in what group?