The Value of a Home

Exploring "Bias" and "Fairness" Through Cook County Property Assessments

Created by Sydney Trieu, in collaboration with Carlos Ortiz and building on the work of Mateo Montoya and Owen Hart

Redlined neighborhood

The "Value of a Home" Curriculum Package explores the promise and limits of data science tools to deliver more accurate and fair assessments of house values. It is based upon the open data and informational interviews with the Cook County Assessor's Office in Illinois. 

This project was realized with the funding support from the Mozilla Responsible Computer Science Challenge grant.

Though the CCAO’s transparency initiative is the first of its kind in the country, it is nonetheless open to critique: The Office itself solicits community input on the quality and fairness of its model. And in this homework, we will address the different types of bias that may arise in the modeling process.
Notebook 1

Notebook 1: Exploring "Bias" Through Cook County's Property Assessments

Jupyter Notebook(link is external)

Grading rubrics and solutions available for instructors upon request (hce@berkeley.edu(link sends e-mail)).

Learning outcomes

By working through this notebook, students will be able to:

  • Identify sources of bias within social and technical decisions. 

  • Understand how bias emerges from the human contexts of data science work, specifically through professions and institutions.

  • Recognize that, because of this human context, bias is structural to data science throughout the data lifecycle rather than an individual, subjective variable that can be eradicated.

  • Analyze the effects of both deliberate and unintentional choices made throughout their work in order to meaningfully address questions of fairness.

Notebook 2: Exploring "Fairness" Through Cook County's Property Assessments

Jupyter Notebook(link is external) 

Grading rubrics and solutions available for instructors upon request (hce@berkeley.edu(link sends e-mail)).

Learning outcomes

By working through this notebook, student will be able to:

  • Understand the relationship between bias and fairness.

  • Analyze the technical and performative functions of transparency initiatives.

  • Recognize the social aspects of transparency in regard to the redistribution of power between different stakeholders.

  • Weigh the effectiveness and limitations of transparency in order to reimagine equitable practices in data science.

Mini-Lecture [coming January 2021]

Homeownership, structural racism, and redlining