Crowdsourcing for Accessibility
Source: https://www.theatlantic.com/technology/archive/2018/11/city-apps-help-and-hinder-disability/574963/
People have always crowdsourced information, members of the disabled community in particular. They needed to know which public places they could access which they couldn’t. This Atlantic article talked about how accessibility maps made through crowdsourcing makes the assumption that information is provided instantaneously and that it’s always correct when sourced from one person. However, it’s not always the case.
One proposed solution over the years is to conduct accessibility map-a-thons, where people go around noting their understandings of public spaces and determine in groups whether they are “yes” accessible, or “no” not accessible. This information is then compiled and put into an app. While this initiative gets people who generally may not notice accessibility issues to notice them, they do not solve every problem for people with disabilities. The solution tries to rely on a information cascade model to determine areas of accessibility. Groups of people make the decision in the map-a-thon whether to assign the space as accessible or not. Each person will be affected by the decision of the person before them. Accessibility is also not “yes” or “no” binary. People, or groups of people, cannot catch every accessibility issue when making the decision of whether an area should be deemed accessible or not. Some people only look at wheelchair accessibility, others only for visual or hearing accessibility. There are many forms of disabilities and sensitivities out there, and they need to be specified so that the users of informational accessibility apps know exactly what kind of public spaces they’re getting into.
These map-a-thons should be rethought in a way that makes use of multiple models to determine whether an area is accessible for certain disabilities. The binary choice should be narrowed down so that people know exactly what they are looking for, that information then generating a larger database encompassing a holistic range of disabilities. Map-a-thons should not be using the simple information cascade model.