Presentation Title

Investigating Bias Within Demographics Across Occupations

Faculty Mentor

Phil Mui

Start Date

23-11-2019 9:45 AM

End Date

23-11-2019 10:00 AM

Location

Markstein 211

Session

oral 1

Type of Presentation

Oral Talk

Subject Area

behavioral_social_sciences

Abstract

Through analysis of weekly wages and employment percentages, our research paper explores discrimination in the workplace and investigates the amount of gender and ethnic bias in different industries. We acknowledge the existence of the severity of the wage gap targeted towards certain demographic groups but more importantly, we identify the specific industries that show extraordinary discriminatory trends from 2007 to 2017. Through cleansing 2017 and 2007 datasets from the Bureau of Labor & Statistics with statistical formulas, we create comparable objective values for wage and employment percentages in order to analyze changes over a ten year period. A negative deviation from the mean wage and employment percentage for each ethnicity indicates derogatory bias against a demographic and a positive value indicates preferential bias for a certain demographic. We convert the deviation values into an x-y coordinate system, representing wage and employment variables respectively. Each quadrant of the graph displays certain biases in one or both variables for each ethnicity. Using Python machine learning, we utilize the objective values to find outliers which represent the occupations that heavily discriminate against certain demographics with two major gender and ethnic biases: wage bias and demographic employment bias. We examined discrepancies across a ten year period, highlighting occupations that maintained or experienced new biases from 2007 to 2017.

This document is currently not available here.

Share

COinS
 
Nov 23rd, 9:45 AM Nov 23rd, 10:00 AM

Investigating Bias Within Demographics Across Occupations

Markstein 211

Through analysis of weekly wages and employment percentages, our research paper explores discrimination in the workplace and investigates the amount of gender and ethnic bias in different industries. We acknowledge the existence of the severity of the wage gap targeted towards certain demographic groups but more importantly, we identify the specific industries that show extraordinary discriminatory trends from 2007 to 2017. Through cleansing 2017 and 2007 datasets from the Bureau of Labor & Statistics with statistical formulas, we create comparable objective values for wage and employment percentages in order to analyze changes over a ten year period. A negative deviation from the mean wage and employment percentage for each ethnicity indicates derogatory bias against a demographic and a positive value indicates preferential bias for a certain demographic. We convert the deviation values into an x-y coordinate system, representing wage and employment variables respectively. Each quadrant of the graph displays certain biases in one or both variables for each ethnicity. Using Python machine learning, we utilize the objective values to find outliers which represent the occupations that heavily discriminate against certain demographics with two major gender and ethnic biases: wage bias and demographic employment bias. We examined discrepancies across a ten year period, highlighting occupations that maintained or experienced new biases from 2007 to 2017.