fongch
Forum Elite
Topics: 71
Posts: 316
|
This is more like a behavioral science Q but I am still quite confused
The purpose of randomization is to control:
A Confounding
B Selection bias
C Observer's bias
D Ascertainment bias
E Effect modification
Please explain, thank you.
|
| robin082006
Forum Hero
Topics: 471
Posts: 5,125
|
B
___________________
The Key to Succeed is Patience.
|
| fongch
Forum Elite
Topics: 71
Posts: 316
|
I thought so too, Robin. But the answer is A.
|
| robin082006
Forum Hero
Topics: 471
Posts: 5,125
|
it depends on the detail of the study.
with such info I will pick up selection bias., it is mentioned in Kaplan.
___________________
The Key to Succeed is Patience.
|
| fongch
Forum Elite
Topics: 71
Posts: 316
|
It's kind of strange but in another question matching is selected as an efficient way to control confounding.
|
| krsna
Forum Senior
Topics: 1
Posts: 211
|
there are three ways in controlling confounding
they are
1.restriction
2.randomisation
3.matching
this is the indesigning satge and
in the analyses satge there are 2 ways of controlling the confounders
1.stratification
2.multivariate analyses
|
| robin082006
Forum Hero
Topics: 471
Posts: 5,125
|
Thanks
___________________
The Key to Succeed is Patience.
|
|
| |
| | | | | | | |
