USMLE Forum Archives - USMLE Step 1 - research biases
research biases
sshahraz - 02-23-08 14:46
I don't like to learn the type of biases by their name! They are simple in itself but people have put names on them and they might be asked in the exam. So, I want to explain three of them and get your feedbacks as well. Let's explain them by asking for the answer to this question:
If the subjects systematically change their behavior while they are being observed the bias produced is called:
A. Berkson's
B. Howthorne effect
C. Pygmalion effect
E. Lead-time bias
F. Telescope
sshahraz - 02-23-08 14:46
I don't like to learn the type of biases by their name! They are simple in itself but people have put names on them and they might be asked in the exam. So, I want to explain three of them and get your feedbacks as well. Let's explain them by asking for the answer to this question:
If the subjects systematically change their behavior while they are being observed the bias produced is called:
A. Berkson's
B. Howthorne effect
C. Pygmalion effect
E. Lead-time bias
F. Telescope
Page 1
#2
Re: research biases
sshahraz - 02-25-08 17:10 B is correct.
In summary:
Berkson is a type of selection bias
Pygmalion: a type of researcher bias when researcher biases toward one of the possible results before looking at the results (very very common! and hard to prove)
Lead time: screening chronic conditions (say cancer) early in the course= better prognosis [which is biased]
Telescope: is the reverse of recall bias but subjects tend to more remember severe cases that happened in the far past (they telescope them!)
Page 1