This blog is the high-level summary of the talk given by Prof. Elissa Redmiles, Max Planck at the PathCheck Global Health Innovators seminar. Check out http://dice.pathcheck.org/talks.html to know more.
Prof. Redmiles discussed some important questions such as – "How do we get people to adopt contact tracing apps (or similar social welfare applications)? How to do so ethically and respect their preferences? etc." When we think about computational problems in such social contexts, they require constant decision-making.
The talk mainly discusses descriptive ethics approaches and aligns technology with people's values. In particular, Prof. Redmiles shared the entire framework and insights of their study in Louisiana for the covid app, Covid Defense, developed by PathCheck Foundation.
We all know there's a benefit if we release a health data set, but earlier, everyone was concerned about privacy. That might be one of the conditions, but it's not the whole story. They also care about the benefits, the providers, cost, accuracy, etc. But which of these matters the most?
Prof. Redmiles and her team experimented by advertising the app "COVID defense," developed by PathCheck Foundation in the state of Louisiana for one month. And found that 30% of "average American" matters is privacy, people also care immensely about the accuracy and benefits, and there are two: Individual benefit, notify your COVID exposure, societal benefit, can reduce COVID infections.
But "is it better to appeal to the individual or collective good?" So the team ran all of these ads and found that collective good appeal was significantly more effective at encouraging adoption and laid out a couple of hypotheses. Additionally, they also noticed a significant impact in privacy statements depending on which appeal was used. One possible reason is the 'privacy calculus.' It helps in being transparent about the data.
At last, being responsible about data use is about more than just privacy. It's about providing technology that respects people's preferences. And that provides the public health benefits that they're hoping to help.
Q&A Highlights
How did you select the group that you want to survey from? Was it completely random, or was there some algorithm behind selecting the cohort?
This survey was from about 4,000 folks in the US of different ages, races, education, etc. and the demographics matched the US census within about 5%. Also, online surveys have two biases (1) it doesn't include someone who does not have consistent internet access and therefore can't take the survey, and (2) it accounts only for people willing to take online surveys because they know how this can impact the community. So I'd say that was the only selection criteria.
What do you think one should focus most on when it comes to digital health solutions? Is it how we deliver the technical progress of our solution?
I would focus on clearly articulating what people expect from participating in the solution. And hope it will accelerate the ability to solve big health problems in the next 5-10 years. Also, privacy sometimes undervalues the amount of social good people want to do and ensure they're not harming themselves. Last, be honest and clear about your privacy protection, but equally important to clarify the benefits of participation.
Thank both our speaker — Prof. Elissa Redmiles — and the organizers — Rohan Sukumaran, Graham Dodge, Ramesh Raskar, Nina Reščič, Shanice Hudson, and Tavpritesh Sethi!
Also, if you're interested in joining our efforts at PathCheck Foundation, head over to — tiny.cc/pathcheckslack, please drop any feedback or let me know if you would like to attend/speak at our event! You can find more such blogs on the PathCheck website — [Here]
This blog was co-authored by - Mohammad Rizin, Tejal Dahake with feedback from our Research Manager Rohan Sukumaran