Authors

Warren Allen

Document Type

Honors Project

First Advisor

Dr. David Huffman

Degree Award Date

Spring 2005

Keywords

Heisman Trophy, voting, data points, goodness of fit

Disciplines

Applied Statistics | Data Science | Numerical Analysis and Computation | Statistics and Probability

Abstract

This study uses cross-sectional data from the 1935 through 2004 National Collegiate Athletic Association seasons to derive a function for Heisman Trophy voting. Using that function, the hypothesis of how Heisman Trophy winning points are influenced by team winning percentage, Associated Press team rank, position, yards in season, touchdowns in season, interceptions in season, being a prior finalist, and class will be tested. This study found that only the variables of team winning percentage, position of fullback, and being prior finalist to be accepted as having marginal statistic significance on winning Heisman Trophy points. These findings however are discredited by the low R-square and adjusted R-square values for the equation as a whole, signifying a poor goodness of fit of the equation to the data points.

Share

COinS