Posts

Posted on

Future Tools for Youth Data Literacies

Workshop Report From Connected Learning Summit 2021

What are data literacies? What should they be? How can we best support youth in developing them via future tools? On July 13th and July 15th 2021, we held a two-day workshop at the Connected Learning Summit to explore these questions. Over the course of two very-full one-hour sessions, 40 participants from a range of backgrounds got to know each other, shared their knowledge and expertise, and engaged in brainstorming to identify important pressing questions around youth data literacies as well as promising ways to design future tools to support youth in developing them. In this blog post, we provide a full report from our workshop, links to the notes and boards we created during the workshop, and a description of  how anyone can get involved in the community around youth data literacies that we have begun to build.

Caption: We opened our sessions by encouraging participants to share and synthesize what youth data literacies meant to them. This affinity diagram is the result. 

How this workshop came to be

As part of the research team interested in research about learning at the Community Data Science Collective, we have long been fascinated with how youth and adults learn how to ask and answer questions with data  While we have engaged with these questions ourselves by looking to Scratch and Cognimates, we are always curious about how we might design tools to promote youth data literacies in the future in other contexts. 

The Connected Learning Summit is a unique gathering of practitioners, researchers, teachers, educators, industry professionals, and others, all interested in formal and informal learning and the impact of new media on current and future communities of learners. When the Connected Learning Summit put up a call for workshops, we thought this was a great opportunity to engage the broader community on the topic of youth data literacies. 

Several months ago, the four of us (Stefania, Regina, Emilia and Mako) started to brainstorm ideas for potential proposals. We started by listing potential aspects and elements of data literacies such as: finding & curating data, visualizing & analyzing it, programming with data, and engaging in critical reflection. We then started to identify tools that can be used to accomplish each goal and tied to identify opportunities and gaps. See some examples of these tools on our workshop website.

Caption: Workshop core team and co-organizers community. Find out more here http://www.dataliteracies.com/

As part of this process, we  identified a number of leaders in the space. This included people who have built tools like Rahul Bhargava and Catherine D’Ignazio who designed Databasic.io,Andee Rubinwho contributed to CODAP, and Victor Lee who focused on tools that link personal informatics and data. Other leaders included scholars who researched how existing tools are being used to support data literacies, including Tammy Clegg who has researched how college athletes develop data literacy skills, Yasmin Kafai who has looked at e-textile projects, and Camillia Matuk who has done research on data literacy curricula. Happily, all of these leaders agreed to join us as co-organizers for the workshop. 

The workshop and what we learned from it

Our workshop took place on July 13th and July 15th as part of the 2021 Connected Learning Summit. Participants came from diverse backgrounds and the group included academic researchers, industry practitioners, K-12 teachers, and librarians. On the first day we focused on exploring existing learning scenarios designed to promote youth data literacies. On the second day we built on big questions raised in the initial session and brainstormed features for future systems. Both workshop sessions were composed of several breakout sessions. We took notes in a shared editor and encouraged participants to add their ideas and comments on sticky notes on collaborative digital white boards and share their definitions and questions around data literacies. 

Caption: organizers and participants sharing past projects and ideas in a breakout session. 

Day 1 Highlights

On Day 1, we explored a variety of existing tools designed to promote youth data literacies. We had a total of 28 participants who attended the session. We began with a group exercise where we shared their own definitions of youth data literacies before dividing into 3 groups: a group focusing on tools for data visualization and storytelling, a group focusing on block-based tools, and a group focusing on data literacy curricula. In each breakout session, our co-organizers first demonstrated one or two existing tools. Each group then discussed how the demo tool might support a single learning scenario based on the following prompt: “Imagine a six-grader who just learned basic concepts about central tendency, how might she use these tools to apply this concept on real world data?” Each group generated many reflective questions and ideas that would prompt and help inform the design of future data literacies tools. Results of our process are captured in the boards linked below. 

Caption: Activities on Miro boards during the workshop.

Data visualization and storytelling

Click here to see the activities on Miro board for this breakout session. 

 

In the sub-section focusing on data visualization and storytelling, Victor Lee first demonstrated Tinkerplots, a desktop-based software that allows students to explore a variety of visualizations with simple click-button interaction using data in .csv format. Andee Rubin then demonstrated CODAP, a web-based tool similar to Tinkerplots that supports drag-and-drop with data, additional visual representation options including maps, and connection between representations. 

Caption: CODAP and Tinkerplots—two tools demonstrated during the workshop.

We discussed how various features of these tools could support youth data literacies in specific learning scenarios. We saw flexibility as one of the most important factors in tool use, both for learners and teachers. Both tools are topic-agnostic and compatible with any data in .csv format. This allows students to explore data of any topics that interest them. Simplicity in interaction is another important advantage. Students can easily see the links between tabular data and visualizations and try out different representations using simple interactions like drag-and-drop, check boxes, and button clicks. Features of these tools can also support students in performing aggregation on data and telling stories about trends and outliers. 

We further discussed potential learning needs beyond what the current features could support. Before creating visualizations, students may need scaffolds during the process of data collection, as well as in the stage of programming with and preprocessing data. Story telling about the process of working with data was another theme that came up a lot from our discussion. Open questions include how features can be designed to support reproducibility, how we can design scaffolds for students to explain what they are doing with data in diary style stories, and how we can help students narrate what they think about a dataset and why they generate particular visualizations.

Block-based tools

Click here to see the activities on Miro board for this breakout session. 

The breakout section about block-based tools started with PhD candidate Stefania Druga demonstrating a program in Scratch and how users could interact with data using the Scratch Cloud Data. We brainstormed about the kind of data students could collect and explore and the kind of visualization, game-based, or other creative interactions youth could create with the help of block-based tools. As a group, we came up with many creative ideas. For example, students can collect and visualize “the newest COVID tweet at the time you touched” a sensor and make “sound effect every time you count a face-touch.” 

Caption: A Scratch project demonstrated during the workshop made with Cloud Data.

We discussed how interaction with data was part of an enterprise that is larger than any particular digital scaffold. After all, data exploration is embedded in social context and might reflect hot topics and recent trends. For instance, many of our ideas about data explorations were around COVID-19 related data and topics. 

Our group also felt that interaction with data should not be limited to a single digital software. Many scenarios we came up with were centered on personal data collection in physical spaces (e.g., counting the number of times a student touches their own face). This points to a future design direction of how we can connect multiple tools that support interaction in both digital and physical spaces and encourage students to explore questions using different tools. 

A final theme from our discussion was around how we can use block-based tools to allow engagement with data among a wider audience. For example, accessible and interesting activities and experience with block-based tools could be designed so that librarians can get involved in meaningful ways to introduce people to data. 

Data literacy curriculum

Click here to see the activities on Miro board for this breakout session. 

In the breakout section emphasizing on curriculum design, we started with an introduction by Catherine D’Ignazio and Rahul Bhargava on DataBasic.io’s Word Counter: a tool that allows users to paste in text to see word counts in various ways. We also walked through some curricula that the team created to guide students through the process of telling stories with data. 

We talked about how this design was powerful in that it allows students to bring their own data and context, and to share knowledge about what they expect to find. Some of the scenarios we imagined included students analyzing their own writings, favorite songs, and favorite texts, and how they might use data to tell personalized stories from there. The specificity of the task supported by the tool enables students to deepen concepts about data by asking specific questions and looking at different datasets to explore the same question. 

Caption: dataBASIC.io helps users explore data.

We also reflected on the fact that tools provided in Databasic.io are easy to use precisely because they are quite narrowly focused on a specific analytic task. This is a major strength of the tools, as they are intended as transitional bridges to help users develop foundational skills for data analysis. Using these tools should help answer questions, but should also encourage users to ask even more.

This led to a new set of issues discussed during the breakout session: How do we chain collections of small tools that might serve as one part of a data literacies pipeline together? This is where we felt curricular design could really come into play. Rather than having tools that try to “be everything,” using well-designed tools that address one aspect of an analysis can provide more flexibility and freedom to explore. Our group felt that curriculum can help learners reach the most important step in their learning, going from data to story to the bigger world—and to understanding why the data might matter. 

Day 2 Highlights

The goal for the Day 2 of our workshop was to speculate and brainstorm future designs of tools that support youth data literacies. After our tool exploration and discussions on Day 1, three interesting brainstorming questions emerged across the breakout sections described above:

  • How can we close the gap between general purpose tools and specific learning goals?
  • How can we support storytelling using data?
  • How can we support insights into the messiness of data and hidden decisions

We focused on discussing these questions on Day 2. A total of 29 participants attended and we once again divided into breakout groups based on the three questions above. For each brainstorming question, we considered the key questions in terms of the following three sub-questions: What are some helpful tools or features that can help answer the question? What are some pitfalls? And what new ideas can we come up with?

Caption: Workshop activities generated an abundance of ideas.

How can we close the gap between general purpose tools and specific learning goals?

Click here to see the activities on Miro board for this breakout session. 

Often tools designed to solve a range of potential problems. That said, learners attempting to engage in data analysis are frequently faced with extremely specific questions about their analysis and datasets. Where does their data come from? How is it structured? How can it be collected? How do we balance the desire to serve many specific learners’ goals with general tools against the desire to handle specific challenges well?

As one approach, we drew lines between different parts of doing data analysis and frequently required features in different tools. Of course, data analysis is rarely a simple linear process. We also concluded that perhaps not everything needs to happen in one place or with one tool, and that this should be acknowledged and considered during the design process.  We also discussed the importance of providing context within more general data analytic tools. We also talked about how learners need to think about the purpose of their analysis before they consider what tool to use and how, ideally, youth would learn to see patterns in data and to understand the significance of the patterns they find. Finally, we agreed that tools that help students understand the limitations of data and the uncertainty inherent in the data are also important.

Challenges and opportunities for telling stories with data

Click here to see the activities on Miro board for this breakout session. 

In this section, we discussed challenges and opportunities around supporting students to tell stories with data. We talked about enabling students to recognize and represent the backstory of data. Open questions included: How do we make sure learners are aware of bias? And how can we help people recognize and document the decision of what to include and exclude?

As for telling stories about students’ own experience of working with data, collaboration was also a topic that came up frequently. We agreed that narrative with data is never an individual process. We discussed that future tools should be designed to support critique, iteration, and collaboration among storytellers, audiences, and maybe also between tellers and audiences.

Finally, we talked about future directions. This included taking a crowdsourced, community-driven approach to tell stories with data. We also noted that we had seen a lot of research effort to support storytelling about data in visualization systems or computational notebooks. We agreed that storytelling should not be limited to digital format and speculated that future designs could extend the storytelling process to unplugged, physical activities. For example, we can design to encourage students to create artefacts and monuments as part of the data storytelling process. We also talked about designing  to engage people from diverse backgrounds and communities to contribute to and explore data together. 

Challenges and opportunities for helping students to understand the messiness of data

Click here to see the activities on Miro board for this breakout session. 

In this section, we talked about the tension between the need to make data clean and easy to use for students and the need to let youth understand the messiness of real world data. We shared our own experiences helping students engage with real or realistic data. A common way is to engage students in collaborative data production and have them compare the outcomes of a similar analysis between each other. For instance, students can document their weekly groceries and find that different people record the same items under different names. They can then come up with a plan to name things consistently and clean their data.

One very interesting point that came up from our discussion was what we really mean by “messy data.” “Messy,” incomplete, or inconsistent data may be unusable for computers while still comprehensible by humans. Therefore to be able to work with messy data does not only mean to have the skills to preprocess, but also involve the recognition of hidden human decisions and assumptions. 

We came up with many ideas regarding future system design. We suggested designing to support crowdsourced data storytelling. For example, students can each contribute a small piece of documentation about the background of a dataset. Features might also be designed to support students to collect and represent the backstory of data in innovative ways. For example, functions that support the generation of rich media, such as videos, drawings, journal entries, can be embedded into data representation systems. We might also innovate on the way we design the interface of data storage so that students can interact with rich background information and metadata while still keeping the data “clean” for computation.

Next steps & community

We intend for this workshop to be only the beginning of our learning and exploration in the space of youth data literacies. We also hope to continue building the community we built. In particular, we have started a mailing list where we can continue our ongoing discussion. Please feel free to add yourself to the mailing list if you would like to be kept informed about our ongoing activities.

Although the workshop has ended, we have included links to many resources on the workshop website, and we invite you to explore the site. We also encourage you to contribute to a crowdsourced list of papers on data literacies by filling out this form.  

This blog was collaboratively written by Regina Cheng, Stefania Druga, Emilia Gan, and Benjamin Mako Hill.

Stefania Druga is a PhD candidate in the Information School at University of Washington. Her research centers on AI literacy for families and designing tools for interest-based creative coding. In her most recent project, she focuses on building a platform that leverages youth creative confidence via coding with AI agents. 

Regina Cheng is a PhD candidate in the Human Centered Design and Engineering department at University of Washington. Her research centers on broadening and facilitating participation in online informal learning communities. In her most recent work, she focuses on designing for novices’ engagement with data in online communities.

Emilia Gan is a graduate student in the Paul G. Allen School of Computer Science and Engineering (UW-Seattle). Her research explores factors that lead to continued participation of novices in computing.

Benjamin Mako Hill is an Assistant Professor at UW. His research involves democratizing data science—and doing it from time to time as well.

Posted on

Do generous attitudes underlie contributions to user-generated content?

User-generated content on the Internet provides the basis for some of the most popular websites, such as Wikipedia, crowdsourced question-and-answer sites like Stack Overflow, video-sharing sites like YouTube, and social media platforms like Reddit. Much (or in some cases all) of the content on these sites is created by unpaid volunteers, who invest substantial time and effort to produce high quality information resources. So are these volunteers and content contributors more generous in general than people who don’t contribute their time, knowledge, or information online?

We (Floor Fiers, Aaron Shaw, and Eszter Hargittai) consider this question in a recent paper published in The Journal of Quantitative Description: Digital Media (JQD:DM). The publication of this particularly is exciting because it pursues a new angle on these questions, and also because it’s part of the inaugural issue of JQD:DM, a new open-access venue for research that seeks to advance descriptive (as opposed to analytic or causal) knowledge about digital media.

The study uses data from a national survey of U.S. adult internet users that includes questions about many kinds of online contribution activities, various demographic and background attributes, as well as a dictator game to measure generosity. In the dictator game, each participant has an opportunity to make an anonymous donation of some unanticipated funds to another participant in the study. Prior experimental research across the social sciences has used dictator games, but no studies we know of had compared dictator game donations with online content contributions.

Sharing content. GotCredit via flickr.

Overall, we find that people who contribute some kind of content online exhibit more generosity in the dictator game. More specifically, we find that people producing any type of user-generated content tend to donate more in the dictator game than those who do not produce any such content. We also disaggregate the analysis by type of content contribution and find that donating in the dictator game only correlates with content contribution for those who write reviews, upload public videos, pose or answer questions, and contribute to encyclopedic knowledge collections.

So, generous attitudes and behaviors may help explain contributions to some types of user-generated content, but not others. This implies that user-generated content is not a homogeneous activity, since variations exist between different types of content contribution.

The (open access!) paper has many more details, so we hope you’ll download, read, and cite it. Please feel free to leave a comment below too.

Paper Citation: Fiers, Floor, Aaron Shaw, and Eszter Hargittai. 2021. “Generous Attitudes and Online Participation”. Journal of Quantitative Description: Digital Media 1 (April). https://doi.org/10.51685/jqd.2021.008.

Posted on

CDSC is hiring a staff person!

Group photo of many of the collective members at a virtual retreat in Spring 2021.

Do you (or someone you know) care about online communities and organizing, scientific research, education, and sharing ideas? We are looking for a person to join us and help grow our research and public impact. The (paid, part-time with benefits) position will focus on responsibilities such as research assistance, research administration, communications and outreach. 

This is a new position and will be the first dedicated staff member with the group. The person who takes the job will shape the role together with us based on their interests and skills.  While we have some ideas about the qualifications that might make somebody a compelling candidate (see below), we are eager to hear from anyone who is willing to get involved, learn on the job, and collaborate with us. You do not need to be an expert or have decades of experience to apply for this job. We aim to value and build on applicants’ experiences.

The position is about half time (25 hours per week) through Northwestern University and could be performed almost entirely remotely (the collective hosts in-person meetings and workshops when public health/safety allows). The salary will start at around $30,000 per year and includes excellent benefits through Northwestern. We’re looking for a minimum 1 year commitment.

Expected responsibilities will likely fall into three areas:

  • Support research execution (example: develop materials to recruit study participants)
  • Research administration (example: manage project tracking, documentation)
  • Community management (example: plan meetings with partner organizations)

Candidates must hold at least a bachelor’s degree. Familiarity with scientific research, project management, higher education, and/or event planning is a plus, as is prior experience in the social or computer sciences, research organizations, online communities, and/or public interest technology and advocacy projects of any kind.

To learn more about the Community Data Science Collective, you should check out our wiki, read previous posts on this blog, and look at some of our recent publications. Please feel free to contact anyone in the group with questions. We are committed to creating a diverse, inclusive, equitable, and accessible work environment within our collective and we look forward to working with someone who shares these values.

Ready to apply? Please do so via this Northwestern University job posting.  We are reviewing applications on a rolling basis and hope to hire someone to begin later this summer.

Posted on

Workshop Announcement: Imagining Future Tools for Youth Data Literacies @ CLS2021

As today’s youth come of age in an increasingly data-driven world, the development of new literacies is increasingly important. Young people need both skills to work with, analyze, and interpret data, as well as an understanding of the complex social issues surrounding the collection and use of data. But how can today’s youth develop the skills they need?

We will exploring this question during an upcoming workshop on Imagining Future Designs of Tools for Youth Data Literacies, one of the offerings at this year’s Connected Learning Summit. As co-organizers for this workshop, we are motivated by our interest in how young people learn to work with and understand data. We are also curious about how other people working in this area define the term ‘data literacy’ and what they feel are the most critical skills for young people to learn. As there are a number of great tools available to help young people learn about and use data, we  also hope to explore which features of these tools made them most effective. We are looking forward to discussions on all of these issues during the workshops.

This workshop promises to be an engaging discussion of existing tools available to help young people work with and understand data (Session 1) and an exploration of what future tools might offer (Session 2). We invite all researchers, educators, and other practitioners to join us for one or both of these sessions. We’re hoping for all attendees to come away with a deeper understanding of data literacies and how to support youth in developing data literacy skills.

Information on registering for the Connected Learning Summit available at: https://connectedlearningsummit.org/

To register interest in attending the Youth Data Literacies Workshop, please complete the pre-registration form at: http://dataliteracies.com/

The workshop is organized by Community Data Science Collective members Regina Cheng, Stefania Druga, Emilia Gan, and Benjamin Mako Hill in collaboration with Rahul Bhargava, Tamara Clegg, Catherine D’Ignazio, Yasmin Kafai, Victor Lee, Camillia Matuk, and Andee Rubin.

Posted on

Community Data Science Collective at ICA 2021

As we do every year, members of the Community Data Science Collective will be presenting work at the International Communication Association (ICA)’s 71st Annual Conference which will take place virtually next week. Due to the asynchronous format of ICA this year, none of the talks will happen at specific times. Although the downside of the virtual conference is that we won’t be able to meet up with you all in person, the good news is that you’ll be able to watch our talks and engage with us on whatever timeline suits you best between May 27 and and 31st.

This year’s offerings from the collective include:

Nathan TeBlunthuis will be presenting work with Benjamin Mako Hill as part of the ICA Computational Methods section on “Time Series and Trends in Communication Research.” The name of their talk is “A Community Ecology Approach for Identifying Competitive and Mutualistic Relationships Between Online Communities.”

Aaron Shaw is presenting a paper on “Participation Inequality in the Gig Economy” on behalf of himself, Floor Fiers and Eszter Hargittai . The talk will be as part of a session organized by the ICA Communication and Technology section on “From Autism to Uber: The Digital Divide and Vulnerable Populations.”

Floor Fiers collaborated with Nathan Walter on a poster titled “Sharing Unfairly: Racial Bias on Airbnb and the Effect of Review Valence.” The poster is part of the interactive poster session of the ICA Ethnicity and Race section.

Nick Hager will be talking about his paper with Aaron Shaw titled “Randomly-Generated Inequality in Online News Communities,” which is part of a high density session on “Social Networks and Influence.”

Finally, Jeremy Foote will be chairing a session on “Cyber Communities: Conflicts and Collaborations” as part of the ICA Communication and Technology division.

We look forward to sharing our research and connecting with you at ICA!

UPDATE: The paper led by Nathan TeBlunthuis won the best paper award from the ICA Computational Methods section! Congratulations, Nate!

Posted on

Newcomers, Help, Feedback, Critical Infrastructure….: Social Computing Scholarship at SANER 2021

This year I was fortunate to present to the 2021 IEEE International Conference on Software Analysis, Evolution and Re-engineering or “SANER 2021.” You can see the write-up of my own presentation on “underproduction” elsewhere on this blog.

SANER is primarily focused on software engineering practices, and several of the projects presented this year were of interest for social computing scholars. Here’s a quick rundown of presentations I particularly enjoyed:

Newcomers: Does marking a bug as a ‘Good First Issue’ help retain newcomers? These results from Hyuga Horiguchi, Itsuki Omori and Masao Ohira suggest the answer is “yes.” However, marking documentation tasks as a ‘Good First Issue’ doesn’t seem to help with the onboarding process. Read more or watch the talk at: Onboarding to Open Source Projects with Good First Issues: A Preliminary Analysis [VIDEO]

Comparison of online help communities: This article by Mahshid Naghashzadeh, Amir Haghshenas, Ashkan Sami and David Lo compares two question/answer environments that we might imagine as competitors—the Matlab community of Stack Overflow versus the Matlab community hosted by Matlab. These sites have similar affordances and topics, however, the two sites seem to draw distinctly different types of questions. This article features an extensive hand-coded dataset by subject matter experts: How Do Users Answer MATLAB Questions on Q&A Sites? A Case Study on Stack Overflow and MathWorks [VIDEO]

Feedback: What goes wrong when software developers give one another feedback on their code? This study by a large team (Moataz Chouchen, Ali Ouni, Raula Gaikovina Kula, Dong Wang, Patanamon Thongtanunam, Mohamed Wiem Mkaouer and Kenichi Matsumoto) offers an ontology of the pitfalls and negative interactions that can occur during the popular code feedback practice known as code review: confused reviewers, divergent reviewers, low review participation, shallow review, and toxic review:
Anti-patterns in Modern Code Review: Symptoms and Prevalence [VIDEO]

Critical Infrastructure: This study by Mahmoud Alfadel, Diego Elias Costa and Emad Shihab was focused on traits of security problems in Python and made some comparisons to npm. This got me thinking about different community-level factors (like bug release/security alert policies) that may influence underproduction. I also found myself wondering about inter-rater reliability for bug triage in communities like Python. The paper showed a very similar survival curve for bugs of varying severities, whereas my work in Debian showed distinct per-severity curves. One explanation for uniform resolution rate across severities could be high variability in how severity ratings are applied. Another factor worth considering may be the role of library abandonment: Empirical analysis of security vulnerabilities in python packages [VIDEO]

Posted on

Mako Hill gets an NSF CAREER Award!

In exciting collective news, the US National Science Foundation announced that Benjamin Mako Hill has received of one of this year’s CAREER awards. The CAREER is the most prestigious grant that the NSF gives to early career scientists in all fields.

You can read lots more about the award in a detailed announcement that the University of Washington Department of Communication put out, on Mako’s personal blog (or in this Twitter thread and this Fediverse thread), or on the NSF website itself. The grant itself—about $550,000 over five years—will support a ton of Community Data Science Collective research and outreach work over the next half-decade. Congratulations, Mako!

Posted on

Detecting At-Risk Software Infrastructure

A span of cracked concrete with exposed rebar.
Crumbling infrastructure. J.C. Burns (jcburns) via flickr, CC BY-NC-ND 2.0

Critical software we all rely on can silently crumble away beneath us. Unfortunately, we often don’t find out software infrastructure is in poor condition until it is too late. Over the last year or so, I have been leading a project I announced earlier to measure software underproduction—a term I use to describe software that is low in quality but high in importance.

Underproduction reflects an important type of risk in widely used free/libre open source software (FLOSS) because participants often choose their own projects and tasks. Because FLOSS contributors work as volunteers and choose what they work on, important projects aren’t always the ones to which FLOSS developers devote the most attention. Even when developers want to work on important projects, relative neglect among important projects is often difficult for FLOSS contributors to see.

Given all this, what can we do to detect problems in FLOSS infrastructure before major failures occur? I recently published and presented a paper laying out our new method for measuring underproduction at the IEEE International Conference on Software Analysis, Evolution and Reengineering (SANER) 2021 that I believe provides one important answer to this question.

A conceptual diagram of underproduction. The x-axis shows relative importance, the y-axis relative quality. The top left area of the graph described by these axes is 'overproduction' -- high quality, low importance. The diagonal is Alignment: quality and importance are approximately the same. The lower right depicts underproduction -- high importance, low quality -- the area of potential risk.
Conceptual diagram showing how our conception of underproduction relates to quality and importance of software.

In the paper—coauthored with Benjamin Mako Hill—we describe a general approach for detecting “underproduced” software infrastructure that consists of five steps: (1) identifying a body of digital infrastructure (like a code repository); (2) identifying a measure of quality (like the time to takes to fix bugs); (3) identifying a measure of importance (like install base); (4) specifying a hypothesized relationship linking quality and importance if quality and importance are in perfect alignment; and (5) quantifying deviation from this theoretical baseline to find relative underproduction.

To show how our method works in practice, we applied the technique to an important collection of FLOSS infrastructure: 21,902 packages in the Debian GNU/Linux distribution. Although there are many ways to measure quality, we used a measure of how quickly Debian maintainers have historically dealt with 461,656 bugs that have been filed over the last three decades. To measure importance, we used data from Debian’s Popularity Contest opt-in survey. After some statistical machinations that are documented in our paper, the result was an estimate of relative underproduction for the 21,902 packages in Debian we looked at.

One of our key findings is that underproduction is very common in Debian. By our estimates, at least 4,327 packages in Debian are underproduced. As you can see in the list of the “most underproduced” packages—again, as estimated using just one more measure—many of the most at risk packages are associated with the desktop and windowing environments where there are many users but also many extremely tricky integration-related bugs.

This table shows the 30 packages with the most severe underproduction problem in Debian, shown as a series of boxplots.
These 30 packages have the highest level of underproduction in Debian according to our analysis.

We hope these results are useful to folks at Debian and the Debian QA team. We also hope that the basic method we’ve laid out is something that others will build off in other contexts and apply to other software repositories.

In addition to the paper itself and the video of the conference presentation on Youtube, we’ve put a repository with all our code and data in an archival repository Harvard Dataverse and we’d love to work with others interested in applying our approach in other software ecosytems.

For more details, check out the full paper which is available as a freely accessible preprint.

This project was supported by the Ford/Sloan Digital Infrastructure Initiative. Wm Salt Hale of the Community Data Science Collective and Debian Developers Paul Wise and Don Armstrong provided valuable assistance in accessing and interpreting Debian bug data. René Just generously provided insight and feedback on the manuscript.

Paper Citation: Kaylea Champion and Benjamin Mako Hill. 2021. “Underproduction: An Approach for Measuring Risk in Open Source Software.” In Proceedings of the IEEE International Conference on Software Analysis, Evolution and Reengineering (SANER 2021). IEEE.

Contact Kaylea Champion (kaylea@uw.edu) with any questions or if you are interested in following up.

Posted on

A round-up of our recent research

Data (Alice Design, cc-by, via the noun project)

We try to keep this blog updated with new research and presentations from members of the group, but we often fall behind. With that in mind, this post is more of a listicle: 22 things you might not have seen from the CDSC in the past year! We’ve included links to (hopefully un-paywalled copies) of just about everything.

Papers and book chapters

Presentations and panels

  • Champion, Kaylea. (2020) How to build a zombie detector: Identifying software quality problems. Seattle Gnu/Linux Users Conference, November, 2020.
  • Hwang, Sohyeon and Aaron Shaw. (2020) Heterogeneous practices in collective governance. Presented at Collective Intelligence 2020 (CI 2020). Boston & Copenhagen (Virtually held).
  • Shaw, Aaron. The importance of thinking big: Convergence, divergence, and independence among wikis and peer production communities. WIkiResearch Showcase. January 20, 2021.
  • TeBlunthuis Nathan E., Benjamin Mako Hill. Aaron Halfaker. “Algorithmic flags and Identity-Based Signals in Online Community Moderation” Session on Social media 2, International Conference on Computational Social Science (IC2S2 2020), Cambridge, MA, July 19, 2020.
  • TeBlunthuis Nathan E.., Aaron Shaw, *Benjamin Mako Hill. “The Population Ecology of Online Collective Action.” Session on Culture and fairness, International Conference on Computational Social Science (IC2S2 2020), Cambridge, MA, July 19, 2020.
  • TeBlunthuis Nathan E., Aaron Shaw, Benjamin Mako Hill. “The Population Ecology of Online Collective Action.” Session on Collective Action, ACM Conference on Collective Intelligence (CI 2020), Boston, MA, June 18, 2020.
Posted on

CDSC is hiring research assistants

The Northwestern University branch of the Community Data Science Collective (CDSC) is hiring research assistants. CDSC is an interdisciplinary research group made of up of faculty and students at multiple institutions, including Northwestern University, Purdue University, and the University of Washington. We’re social and computer scientists studying online communities such as Wikipedia, Reddit, Scratch, and more.

Screenshot from a recent remove meeting of the CDSC
A screenshot from a recent remote meeting of the CDSC…

Recent work by the group includes studies of participation inequalities in online communities and the gig economy, comparisons of different online community rules and norms, and evaluations of design changes deployed across thousands of sites. More examples and information can be found on our list of publications and our research blog (you’re probably reading our blog right now).

This posting is specifically to work on some projects through the Northwestern University part of the CDSC. Northwestern Research Assistants will contribute to data collection, analysis, documentation, and administration on one (or more) of the group’s ongoing projects. Some research projects you might help with include:

  • A study of rules across the five largest language editions of Wikipedia.
  • A systematic literature review on the gig economy.
  • Interviews with contributors to small, niche subreddit communities.
  • A large-scale analysis of the relationships between communities.

Successful applicants will have an interest in online communities, social science or social computing research, and the ability to balance collaborative and independent work. No specialized skills are required and we will adapt work assignments and training to the skills and interests of the person(s) hired. Relevant skills might include: coursework, research, and/or familiarity with digital media, online communities, human computer interaction, social science research methods such as interviewing, applied statistics, and/or data science. Relevant software experience might include: R, Python, Git, Zotero, or LaTeX. Again, no prior experience or specialized skills are required. 

Expected minimum time commitment is 10 hours per week through the remainder of the Winter quarter (late March) with the possibility of working additional hours and/or continuing into the Spring quarter (April-June). All work will be performed remotely.

Interested applicants should submit a resume (or CV) along with a short cover letter explaining your interest in the position and any relevant experience or skills. Applicants should indicate whether you would prefer to pursue this through Federal work-study, for course credit (most likely available only to current students at one of the institutions where CDSC affiliates work), or as a paid position (not Federal work-study). For paid positions, compensation will be $15 per hour. Some funding may be restricted to current undergraduate students (at any institution), which may impact hiring decisions.

Questions and/or applications should be sent to Professor Aaron Shaw. Work-study eligible Northwestern University students should indicate this in their cover letter. Applications will be reviewed by Professor Shaw and current CDSC-NU team members on a rolling basis and finalists will be contacted for an interview.

The CDSC strives to be an inclusive and accessible research community. We particularly welcome applications from members of groups historically underrepresented in computing and/or data sciences. Some of these positions funded through a U.S. National Science Foundation Research Experience for Undergraduates (REU) supplement to awards numbers: IIS-1910202 and IIS-1617468.