Winner's Curse

Welcome to my project page~

I hope you enjoy.


Project Description

The winner's curse is a phenomenon first observed in economics in which a party at auction overestimates the value of an item. It states that the winner ( in this case the party with the highest estimate ) will tend to pay more than the item's intrinsic value.

The winner's curse can be observed in many fields in addition to economics. In genetics, ascertainment bias and oversampling contribute to an overestimated genotype penetrance in initial studies.

The goal of my project is to detect the winner's curse by simulating association studies with a replication sample. A secondary goal is to correct the effect size through these studies.


Biography

I am a second year mathematics of computation major at UCLA. I'm taking this class because my father is a geneticist and I have always wondered what he does with his time.

I have a bunch of hobbies. I'm on the triathlon club at UCLA and subsequently spend the majority of my days swimming, biking or running. I'm really proud of the last race I participated in, the UCSB Kendra Payne Memorial Triathlon. It was an olympic distance race and since my background is in running, I was petrified of the 1500m swim. But I made it and even placed well. Score one for me.

I also play bass. I'm ashamed to say I haven't been playing it as much as I would have liked, but school is a time consuming endeavor. Right now my favorite bassist is Geddy Lee from Rush. And I'm particularly obsessed with the song Limelight.

I also like to knit, watch movies, go to the beach and take my time to chill.

An interesting fact about myself if that I'm going to spain next year to study abroad. I will be in Granada for a year taking math classes and GEs. I'm excited to learn about another culture and even more about myself there next year.


Quarter Goals

1) Understand the Winner's Curse and why it occurs.

2) Demonstrate the existence of the winner's curse in genetics through replication studies.

3) Correct for overestimated effect size using replication studies.

4) Compile my findings into powerpoint form.

5) Give an interesting presentation which the class enjoys.


Week One

Plan

1) Research the Winner's Curse
2) Start Webpage

Progress

Got the web page up and a little bit of reading done. Need to do more tomorrow though.

Problems

Clear sailing for now.

Grade

A.


Week Two

Plan

1) Research the Winner's Curse
2) Go to office hours and ask questions about winner's curse
3) Plan how to simulate association studies

Progress

Began Reading Zöllner and Pritchard article about estimating penetrance parameters

Problems

The article is dense, and I'm having trouble understanding the methods.

Grade

A.


Week Three

Plan

1) Start Introduction to Powerpoint
2) Continue reading article of Winner's Curse
3) Plan how to simulate association studies

Progress

Went to office hours and sharpened my knowledge of the winner's curse.

Problems

Same as last week, but I haven't gone back to look at the article yet.

Grade

B.


Week Four

Plan

1) Learn to graph in R
2) Finish Powerpoint
3) Practice presentation
4) Learn how to upload .ppt to wikidot

Progress

Got very far through the powerpoint. Simulated association studies and graphed results

Problems

Don't know why the effect size bias isn't proportional to N

Grade

A++


Final Week

Plan

NONE

Progress

Finished my powerpoint and presented it to the class. Just need to upload it to the wikidot

Problems

none

Grade

A


My powerpoint and experiment info are uploaded under the files section

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