Michael Karsh

I am planning on doing a project on the Winner's Curse. The Winner's Curse applies to genetics in the sense that most genetics effects detected are due to luck. Think of the mean as being 0 and a winner gets a positive result just by chance. In the future that same person may get a less positive or even a negative result. In other words most genetic effects detected are not real but due to random effects. Repeating the study will show the genetic effects to be weaker than was previously thought.
I want to show how using replicated samples one can correct for effect sizes using a winner's curse. I may eventually try to develop a statistic to correct for the winner's curse.
Next week I will read the article by Jonathen Pritchard and Sebastian Zollner called "Overcoming the Winner's Curse: Estimating Penetrance Parameters from Case-Control Data"
I deserve an A for doing what I was supposed to this week.

Unless otherwise stated, the content of this page is licensed under Creative Commons Attribution-ShareAlike 3.0 License