Tayo Fabusuyi and the Conversations with Practitioners Working Group
Our Conversations with Practitioners working group had the pleasure of interviewing Tiasia O’Brien. In 2018, Tiasia started Seam Social Labs to transform community voices into a valuable data source for equitable design outcomes.
Community voices can reveal root causes
“There is often a gap in information where the data shows one solution, but the community needs a different solution.”
Tiasia gave an example of a community where data showed a lack of affordable housing and the solution given by policymakers was to build more affordable housing. Engagement with the community, however, found that the core issues were not affordable housing, but a lack of jobs and public transportation to those jobs. Workforce development and lack of transportation were the root causes of the problem.
The need for a mixed methods, multi-channel data gathering approach
Despite the need for both quantitative and qualitative data for policy design, few policymakers utilize a mixed methods, multi-channel data gathering approach. Tiasia started her own company, Seam Social Labs in 2018 to empower communities and make sure that both quantitative and qualitative data are incorporated into the civic and community space. Her company developed a cloud-based platform called co:census that allows institutions to import and analyze qualitative data from surveys, interviews, meetings, and social media.
Hyper local data illuminate social and cultural elements that matter
“There are many different cultural undertones that impact on people’s decisions that we might miss.”
California’s Office of Digital Innovation struggled to understand issues related to vaccination and vaccine hesitancy in rural and farming communities. In collaboration with Tiasia and Seam Social Labs, the Office of Digital Innovation used information from local partners, publications in several languages, and text data scraped from Reddit posts to listen to “hard-to-reach populations” (a term Tiasia dislikes as she feels that these populations are not hard to reach). Results from the analysis found remarkable consistency of responses at the zip code level, which implies that there were social and cultural elements at play that broader education campaigns would not impact.
Making sense of free-form data
co:census analyzes free-form data with several approaches including sentiment analysis (positive, negative, neutral), topic modeling (for example inductive looking for patterns or themes; then going back to the community for verification), and emotion detection. co:census does a behavioral analysis, looking at humans’ behaviors, their experiences, along with the condition of the environment.
The civic gap
“The civic gap is the theory and belief that, based on your socio-economic status, based on the level of wealth in your community, your ability to be civically engaged is going to be impacted, and what civic engagement looks like, at those different levels, is going to be very different.”
Remember the human element of data modeling
“That is the challenge, that is the fight — it’s less about the tools and more about the people and their human experience. Data models are typically made by people in Silicon Valley who may have very different life experiences, and the models are trained on data that teaches an entire tool how to work. The challenge with that is … data modeling and analyzing is so focused on the tools and the best practices on how to use those tools that it leaves out the human side. And we need to go back and ask those people if we did this right or not. That is the big inherent challenge.
Preventing bias to the loudest voice
“I hear cities say that all the time. That is not a data analysis challenge. That’s the data collection challenge. Are we utilizing a model that incorporates enough data? Ideally you would focus on quota-based data collection methodology to reduce the bias of the loudest voice.”
Researchers and policymakers need to make time to listen
“We add the word listening to many things, but we don’t really listen to people. For anyone who’s working on a research or policy role, if they could just spend 10% of their time listening to the public.”
The radical idea of treating the public voice as a data point
“What would that look like? That sounds crazy to a lot of people, but I need(ed) to find the people to whom it didn’t sound crazy. What worked for me was accepting that people thought I was crazy and being okay with it. But being confident that I know I’m onto something. And holding on to the network where I fit in. Finding that community.”
