Sam Winn and Aisling O’Dell
Sam and Aisling are juniors in their second year of the Princeton High School Research Program.
Alejandro Garcia Fernandez
Alejandro is a graduate student at Princeton University, working towards a PhD in sociology.
Professor Dalton Conley
Dalton Clark Conley is an American sociologist, and the supervisor of this research project. He is the Henry Putnam University Professor of Sociology at Princeton University where he is also an affiliate of the Office of Population Research and the Center for Health and Wellbeing.
Our proposed stratifying inequality project will attempt to highlight, identify and understand existing biases and discrimination in the context of employee evaluations. We will attempt to identify gaps in ratings across four races (White, Black, Latinx, Asian) and two genders (male, female) through a survey taken by actual employers using MTurk. It is hypothesized that White and Asian men will receive the highest ratings, while Black and Latinx women’s ratings will be lower due to racial/gender bias and discrimination in employers. Conjoint analysis will be performed in order to compare each of the racial and gendered groups in order to study the intersectionality that is present in employer biases. We will later conduct qualitative interviews to gain a better understanding of the driving forces behind this discrimination. One of these hypothesized factors is homophily, or the idea that “similarity breeds connection”. This means employers are more likely to give a higher rating to employees who belong to the same ethno-racial group as them. A second factor behind the gap between employee ratings could be the architecture of the evaluation. For example, in previous research, a 6 point scale was shown to reduce bias over a 10 point scale because employers perceived the rating scale differently, and couldn’t make subtle deductions due to racial and gender animosities. Negative employee evaluations can significantly impact one’s ability to receive promotions and bonuses, preventing employee’s ascension on the socio-economic ladder. This effect can compound with time, leading to generational inequality. If we understand the systemic biases and racism laced in employee evaluations, and the driving motives behind them, we can work with organizations to reduce, and eventually eliminate workplace discriminaton.
Background/Rationale for the Study
While discrimination and bias directly and clearly influence employee evaluations, they also impact the career trajectory of the employee. Therefore, this is a necessary experiment to conduct in order to correct the systemic racism existing within the employment sector of our society.
The effect of perceived race through applications has a profound impact on the likelihood that the individual will be hired, let alone advance on the corporate ladder. Discrimination is not always obvious and can disappear in the system of a workplace as time goes on. If someone is given a slightly different starting position in a company and it goes unnoticed, the racism and sexism that existed during the time of the hiring decision is left undiscovered. This shows the importance of determining the presence of racism and sexism in employer decisions as it will affect the employee and their families lives for generations, whether they are aware of it or not. This also makes it increasingly difficult for researchers to analyze the effects of ethno-racially based discrimination in the work place (Pager and Shepherd 2008). Therefore, studies need to be conducted starting with employers looking at resumes of possible job candidates in order to understand the thought process of these employers and attack the root cause of workplace discrimination.
As multiple variables (ethnoracial identity, prestige of college attended, and gender) are being tested in order to determine the gap present between groups of different sexes and races in the job market, conjoint analysis will be used to analyze the results. The purpose of conjoint analysis is to make it easier to find connections between the data sets that are produced, which would be harder to achieve in other circumstances.
While participating in this survey, employers will be given a series of vignettes with colleges (varied for level of prestige) and names (to signify the race and gender of the applicant) that they are asked to evaluate. They will then rate each candidate quantitatively on a variety of job-related characteristics, provide a detailed qualitative explanation of how their hiring decisions are made, and fill out demographic questions that can later be used to study homophily and other driving factors.
This experiment seeks to answer the following questions:
(1) How are employers’ evaluations of job candidates informed by the prestige of applicants’ college degrees, and how are these evaluations racialized?
(2) How are employers’ evaluations of job applicants informed by their own characteristics?
(3) How are employers’ evaluations informed by the design of evaluation systems?
(4) How do employers evaluate job applicants?
(5) How do employers make hiring decisions?
(Question 4 and 5 are based off of the qualitative response given at the end of the survey asking the employer to detail their thought process while evaluating the candidates. These qualitative responses will then be analyzed for repetition/patterns)