Campaign Website Accessibility Evaluation

A proof of concept to evaluate the accessibility of political campaign digital materials.

Background

To better understand how political campaigns make their platforms accessible to their constituents, we are building tools to help quickly and easily evaluate their campaign website (and potentially, any additionally linked materials -- i.e. PDFs of their plans).

For example, are political campaigns taking the time to ensure all elements of their digital materials are accessible to low-or-no vision individuals. Are political campaigns ensuring that their materials are available in multiple languages?

In many cases we can combine the data produced by our tools with census data regarding disability populations and language-spoken-at-home statistics. But we can also study correlations between web accessibility and campaign spending.

Recreating Analysis

To recreate our analysis of our preliminary 2021 study:

1. Clone the repo
2. In a terminal navigate to the cloned repo
3. See CONTRIBUTING.md for geckodriver installation instructions
4. In the repo directory run:

pip install .
analyze-access-eval-2021-dataset

This will generate all of the plots used in our paper and place them in a directory called plots/. It will also create a file called stats.json which contains the statistics used in our paper (generated from t-tests, ANOVA, etc.)

aXe Scraper Usage

In general, follow the instructions described in CONTRIBUTING.md to install required packages.

Once installation and setup are complete, to generate a full accessibility evaluation report for a website, run:

make generate-report url={URL}

i.e.

make generate-report url="https://evamaxfield.github.io/"

Maintainer GitHub Action

If you are a maintainer of this library (or of a fork of this library), you can generate accessibility reports for any URL by using the Generate Report GitHub Action.