README.md

In ariel-shin/tourn: Analyzing Debate Data

Authors

Ariel Shin and Norm Matloff

Synopsis

This project explores the special statistical challenges of elimination tournaments, specifically national high school Lincoln-Douglas debate circuit tournaments from the 2015-2016 season.

Motivation

The first author debated on the national circuit from 2010-2014 for Immaculate Heart High School. She reached late elimination rounds of many tournaments including Lexington, Berkeley, Stanford, and many more. She attended the Tournament of Champions her senior year. This project is based in the R programming language and aims to explore questions such as: Does gender affect the outcome of rounds? Does geography play a role in wins/losses? What consitutes an upset? Is there a so-called "shadow effect"?

Graphs

A sample of the project's results are shown in the following graphs.

Note: The red line represents females and the blue line represents males.

MeanWinsVsSeed.png

The graphs shows that generally, women of a given seed fared poorly relative to men who had the same seed. Theoretically, the seed is a measure of a debater's ability. Even though it is an imperfect measure, seedings should average out such that male and females have similar win/loss records. This graph suggests a systemic bias against females. The debate community has recognized this gender bias and has attributed it to debate being a "boys club." Girls participate less because they have less support and fewer role models than their male counterparts. Moreover, girls are judged more harshly on their appearances, voice, and attitude than boys. On the other hand, a reverse bias may be the factor for the differences in performances. For instance, judges in pre-elimination rounds may give favorable seedings to girls than they would give to boys of the same quality. If this were the case, the girls might not do as well in elimination rounds.

SeedingGender.png

This density graph also suggests that men fare better with seeding

SpeakerpointsGender.png

This graph is inconclusive because of the wide range of speakerpoints. Each tournament sums up the total number of speakerpoings minus the round with the highest speakerpoint and the round with the lowest speakerpoing. Some tournaments have 5 rounds, while others have 6 or 7 rounds. The range of rounds and thus total speakerpoings makes this graph inconclusive.

roundsWonGender.png

Female seems to have losing records and males seem to have winning records. This graphs supports the community's thoughts on the gender disparity in debate.

ariel-shin/tourn documentation built on May 10, 2019, 1:27 p.m.