Challenges for Personal Behavior Change Research on Information Diversity



Researchers have tested a variety of personal informatics systems to encourage diversity in the political leaning, geography, and demographics of information sources, often with a belief in the normative value of exposure to diverse information sources. Methods attempted have included information labeling of media sources, personalized metrics of reading behavior, personalized visualization of social media behavior, recommendation systems, and social introductions. Although some of these systems demonstrate positive results for the metrics they define, substantial questions remain on the interpretation of these results and their implications for future design. We identify challenges in defining normative values of diversity, potential algorithmic exclusion for some groups, and the role of personal tracking as surveillance. Furthermore, we outline challenges for evaluating systems and defining the meaningful social impact for information diversity systems operating at scale.

A Conversation about Legible Cities

Originally published as a post on Unified Field’s blog [Internet Archive].

Legible Cities conversation with Erhardt Graeff from MIT Media Lab

Eli: How can Legible Cities be manifested? Now it is mostly a concept. I was reading Townsend’s book on big data and smart cities, which inspires this question. If city government’s are more interested in using smart city models for promoting economic development and front the street development seems to take the form of Foursquare or Arduino projects, what is there out there for the “people’s” view? Who would own and manage these new types of projects?

Erhardt: This is an interesting and difficult question because the rhetoric of the “front the street” developers like Foursquare is that it is the people’s view which is aggregated, somewhat passively, as people use the app and navigate the city. There is some truth to that idea since passive observation allows us a better sense of what someone’s actual experience of the city is rather than asking them, which brings in all sorts of biases like social desirability. Of course, you can argue that the designs of these systems—their affordances and the behavioral economics imbued in their user experiences—produce new and different biases. The competing paradigm seems to be the open data movement that seeks to release as much public and private data about city resident experiences and give them the chance to analyze it themselves and make their city smarter through some kind of discussion or policy recommendation. It always gets fuzzy from there.

I think the people’s view could look like an app that allows a bit of reflective storytelling to be taken on by the user. We see this working on small scales with organizations that do community mapping projects like Map Kibera. This gives a real, personalized view of the city, where hidden resources relevant to real people are identified and highlighted. The key question is how to scale that. What’s the middle ground between individualized community mapping and Foursquare-like location apps?

Eli: I don’t think aggregated passive data from Foursquare is an indication of the people’s view. Reminds me of the Sanskrit tale of the blind men describing an elephant. Can you tell me more about social desirability? I am not sure exactly what you mean. I am sure you are familiar with the promises of urban dynamism. Like the myth, evidence-based medicine can be a replacement for diagnosis but in many ways is hardly a complete picture.

The open data plans that I’ve seen and heard about in conversations with city officials who manage these open data programs is that they are focused primarily on economic development and in particular mobile apps. For the vast majority of people, the missing piece is not only access, but context and data literacy. Not to say that people should become data scientists, but like STEM and science literacy as a people we need to have a data literate population and workforce.

I cannot believe that people would not be interested in the real picture and status of their city, environment and neighborhood. The question is what does that look like and how can one sustain it? Should it be only mobile apps and websites, or should there be some physical representation that is not just an art project, we has some utilitarian value?

We see a lot of work like Jason Bruges, and other artists and designers doing art installations that are mostly eye candy, not much that is usable. My take on Foursquare is that although it provides a useful service, it’s primarily for the hip urban digerati. This holds true for digital wayfinding systems based solely on smart phones. Although there is evidence in developing countries that the mobile phone is becoming the standard web interface device, we still have situations, at hospitals and other locations and facilities, where you find demographics that either do not have mobile devices or ones that cannot be plugged into a network like Foursquare.

I’d like to see more focus in general on not what would only benefit cities in their planning and management, but what would the people need and use.

Erhardt: I agree with you that the missing value proposition is what everyday people can gain; what do they need and what can they use? Legibility, literacy, comprehension is at the core of this. I think access is a real concern not only in terms of being able to understand the data, but the media it comes through. I agree that it’s interesting and important to consider what a physical instantiation of community data looks like. My colleague Rahul Bhargava is doing great work along those lines, which he calls Data Therapy—check out the mural! I believe that the current obsession with big data and data science is actually hurting the legibility and accessibility of data. It’s making it more elitist. I think a City Data agenda should be organized around empowerment and not the empowerment of coder elites, who will build apps or produce free labor analysis and visualization for themselves.


Erhardt Graeff is a graduate student at the MIT Media Lab and MIT Center for Civic Media, studying information flows across mainstream and social media, and exploring technologies that empower people to be greater agents of change. Erhardt is also a founding trustee of The Awesome Foundation, which gives small grants to awesome projects. He holds an MPhil from the University of Cambridge and two B.S. degrees from Rochester Institute of Technology.