/bias-sim

Primary LanguageJavaScript

Bias Simulation

The goal of this experiment is to get an impression on how unconscious bias can affect the structure of a company. This was inspired by this talk.

Let's explain what the simulation is about:

  • We assume that there is an organization, with 7 layers
    • On the level 1 there are the senior members of the organization
    • On level 7 we have the most junior members
  • We assume that at each level, there are 50% man and 50% women
  • Before beginning the simulation, two parameters can be set:
    • Attrition rate: On every step of the simulation, as assume that some employees leave the company. We assume that half of these employees are women and half man. The attrition rate gives the percentage of employees who decide to leave.
    • Bias level: We'll explain this later.

Starting the simulation: The simulation has 20 steps. Here is what happens during each steps:

  • We start from Level 1. We assume that the percentage of employees, given by the attrition rate leave the company.
  • We have two ways of filling in the open positions: We assume that we fill in half of the open positions by hiring new employees and the other half by promoting existing employees from the previous level.
  • We assume that both the hiring and the promotion processes are biased towards men, by a factor given by the bias level. See the next chapters about the Hiring process and Promotion process below.

Hiring process

Let's assume that we have 10 positions to fill. We assume that we have 100 candidates, 50 women and 50 men.

Promotion process

<script type="text/javascript" src="https://d3js.org/d3.v5.min.js"></script> <script type="text/javascript" src="https://naver.github.io/billboard.js/release/latest/dist/billboard.js"></script>
Attrition Rate: Bias Level:

Start Simulation Reset Simulation

Attrition Rate: 15
Bias Level: 1