General Discussion
In reply to the discussion: Can anyone provide a reason for more than 3 rounds? [View all]Major Nikon
(36,827 posts)Case controlled studies exclude cases that are non-homogeneous. I'm just making up numbers to explain how it works, but let's say 90% of the cases were from victims with a criminal record, yet only 20% of the general population had a criminal record, then the study would be obviously flawed as you describe. That's why they correct for those types of situations by excluding or oversampling cases. It's pretty typical for a case controlled study to exclude more than half of the cases or even the vast majority of them before any statistical analysis is determined to control for exactly the types of situations you describe.
And no, I don't believe any such ridiculous notion that you are describing. I'm simply pointing to a case controlled study derived using empirical methods that were evaluated and approved by the CDC with the results published in a peer reviewed journal. If you think you can dismiss it that easily, be my guest, but I find your arguments wholly unconvincing.