Tag: projects

Note: I have not published blog posts about my academic papers over the past few years. To ensure that my blog contains a more comprehensive record of my published papers and to surface these for folks who missed them, I will be periodically (re)publishing blog posts about some “older” published projects.

It seems natural to think of online communities competing for the time and attention of their participants. Over the last few years, I’ve worked with a team of collaborators—led by Nathan TeBlunthuis—to use mathematical and statistical techniques from ecology to understand these dynamics. What we’ve found surprised us: competition between online communities is rare and typically short-lived.

When we started this research, we figured competition would be most likely among communities discussing similar topics. As a first step, we identified clusters of such communities on Reddit. One surprising thing we noticed in our Reddit data was that many of these communities that used similar language also had very high levels of overlap among their users. This was puzzling: why were the same groups of people talking to each other about the same things in different places? And why don’t they appear to be in competition with each other for their users’ time and activity?

We didn’t know how to answer this question using quantitative methods. As a result, we recruited and interviewed 20 active participants in clusters of highly related subreddits with overlapping user bases (for example, one cluster was focused on vintage audio).

We found that the answer to the puzzle lay in the fact that the people we talked to were looking for three distinct things from the communities they worked in:

1. The ability to connect to specific information and narrowly scoped discussions.
2. The ability to socialize with people who are similar to themselves.
3. Attention from the largest possible audience.

Critically, we also found that these three things represented a “trilemma,” and that no single community can meet all three needs. You might find two of the three in a single community, but you could never have all three.

The end result is something I recognize in how I engage with online communities on platforms like Reddit. People tend to engage with a portfolio of communities that vary in size, specialization, topical focus, and rules. Compared with any single community, such overlapping systems can provide a wider range of benefits. No community can do everything.

This work was published as a paper at CSCW: TeBlunthuis, Nathan, Charles Kiene, Isabella Brown, Laura (Alia) Levi, Nicole McGinnis, and Benjamin Mako Hill. 2022. “No Community Can Do Everything: Why People Participate in Similar Online Communities.” Proceedings of the ACM on Human-Computer Interaction 6 (CSCW1): 61:1-61:25. https://doi.org/10.1145/3512908.

This work was supported by the National Science Foundation (awards IIS-1908850, IIS-1910202, and GRFP-2016220885). A full list of acknowledgements is in the paper.

Note: I have not published blog posts about my academic papers over the past few years. To ensure that my blog contains a more comprehensive record of my published papers and to surface these for folks who missed them, I will be periodically (re)publishing blog posts about some “older” published projects. This post is closely based on a previously published post by Kaylea Champion on the Community Data Science Blog.

One popular approach to protecting our privacy online is to use the Tor network. Tor protects users from being identified by their IP address, which can be tied to a physical location. However, if you’d like to contribute to Wikipedia using Tor, you’ll run into a problem. Although most IP addresses can edit without an account, Tor users are blocked from editing.

Other research by my team has shown that Wikipedia’s attempt to block Tor is imperfect and that some people have been able to edit despite the ban. As part of this work, we built a dataset of more than 11,000 contributions to Wikipedia via Tor and used quantitative analysis to show that contributions from Tor were of about the same quality as contributions from other new editors and other contributors without accounts. Of course, given the unusual circumstances Tor-based contributors face, we wondered whether a deeper look at the content of their edits might reveal more about their motives and the kinds of contributions they seek to make. Kaylea Champion (then a student, now faculty at UW Bothell) led a qualitative investigation to explore these questions.

Given the challenges of studying anonymity seekers, we designed a novel “forensic” qualitative approach inspired by techniques common in computer security and criminal investigation. We applied this new technique to a sample of 500 editing sessions and categorized each session based on what the editor seemed to be intending to do.

Most of the contributions we found fell into one of the two following categories:

• Many contributions were quotidian attempts to add to the encyclopedia. Tor-based editors added facts, fixed typos, and updated train schedules. There’s no way to know whether these individuals knew they were just getting lucky in their ability to edit or were patiently reloading to evade the ban.
• Second, we found harassing comments and vandalism. Unwelcome conduct is common in online environments, and it is sometimes more common when the likelihood of being identified decreases. Some of the harassing comments we observed were direct responses to being banned as a Tor user.

Although these were most of what we observed, we also found evidence of several types of contributor intent:

• We observed activism: when a contributor tried to draw attention to journalistic accounts of environmental and human rights abuses committed by a mining company, only to have editors linked to the mining company repeatedly remove their edits. Another example involved an editor trying to diminish the influence of proponents of alternative medicine.
• We also observed quality maintenance activities when editors used Wikipedia’s rules on appropriate sourcing to remove personal websites being cited in conspiracy theories.
• We saw edit wars with Tor editors participating in a back-and-forth removal and replacement of content as part of a dispute, in some cases countering the work of an experienced Wikipedia editor whom even other experienced editors had gauged to be biased.
• Finally, we saw Tor-based editors participating in non-article discussions, such as investigations into administrator misconduct and protests against the Wikipedia platform’s mistrust of Tor editors.

In all, these themes led us to reflect on how the risks individuals face when contributing to online communities sometimes diverge from the risks the communities face in accepting their work. Expressing minoritized perspectives, maintaining community standards even when you may be targeted by the rulebreaker, highlighting injustice, or acting as a whistleblower can be very risky for an individual, and may not be possible without privacy protections. Of course, in platforms seeking to support the public good, such knowledge and accountability may be crucial.

This work was published as a paper at CSCW: Kaylea Champion, Nora McDonald, Stephanie Bankes, Joseph Zhang, Rachel Greenstadt, Andrea Forte, and Benjamin Mako Hill. 2019. A Forensic Qualitative Analysis of Contributions to Wikipedia from Anonymity Seeking Users. Proceedings of the ACM on Human-Computer Interaction. 3, CSCW, Article 53 (November 2019), 26 pages. https://doi.org/10.1145/3359155

This project was conducted by Kaylea Champion, Nora McDonald, Stephanie Bankes, Joseph Zhang, Rachel Greenstadt, Andrea Forte, and Benjamin Mako Hill. This work was supported by the National Science Foundation (awards CNS-1703736 and CNS-1703049) and included the work of two undergraduates supported through an NSF REU supplement.

Note: I have not published blog posts about my academic papers over the past few years. To ensure that my blog contains a more comprehensive record of my published papers and to surface these for folks who missed them, I will be periodically (re)publishing blog posts about some “older” published projects. This particular post is closely based on a previously published post by Nate TeBlunthuis from the Community Data Science Blog.

Many online platforms are adopting AI and machine learning as a tool to maintain order and high-quality information in the face of massive influxes of user-generated content. Of course, AI algorithms can be inaccurate, biased, or unfair. How do signals from AI predictions shape the fairness of online content moderation? How can we measure an algorithmic flagging system’s effects?

In our paper published at CSCW, Nate TeBlunthuis, together with myself and Aaron Halfaker, analyzed the RCFilters system: an add-on to Wikipedia that highlights and filters edits that a machine learning algorithm called ORES identifies as likely to be damaging to Wikipedia. This system has been deployed on large Wikipedia language editions and is similar to other algorithmic flagging systems that are becoming increasingly widespread. Our work measures the causal effect of being flagged in the RCFilters user interface.

Our work takes advantage of the fact that RCFilters, like many algorithmic flagging systems, create discontinuities in the relationship between the probability that a moderator should take action and whether a moderator actually does. This happens because the output of machine learning systems like ORES is typically a continuous score (in RCFilters, an estimated probability that a Wikipedia edit is damaging), while the flags (in RCFilters, the yellow, orange, or red highlights) are either on or off and are triggered when the score crosses some arbitrary threshold. As a result, edits slightly above the threshold are both more visible to moderators and appear more likely to be damaging than edits slightly below. Even though edits on either side of the threshold have virtually the same likelihood of truly being damaging, the flagged edits are substantially more likely to be reverted. This fact lets us use a method called regression discontinuity to make causal estimates of the effect of being flagged in RCFilters.

To understand how this system may affect the fairness of Wikipedia moderation, we estimate the effects of flagging on edits on different groups of editors. Comparing the magnitude of these estimates lets us measure how flagging is associated with several different definitions of fairness. Surprisingly, we found evidence that these flags improved fairness for categories of editors that have been widely perceived as troublesome—particularly unregistered (anonymous) editors. This occurred because flagging has a much stronger effect on edits by the registered than on edits by the unregistered.

We believe that our results are driven by the fact that algorithmic flags are especially helpful for finding damage that can’t be easily detected otherwise. Wikipedia moderators can see the editor’s registration status in the recent changes, watchlists, and edit history. Because unregistered editors are often troublesome, Wikipedia moderators’ attention is often focused on their contributions, with or without algorithmic flags. Algorithmic flags make damage by registered editors (in addition to unregistered editors) much more detectable to moderators and so help moderators focus on damage overall, not just damage by suspicious editors. As a result, the algorithmic flagging system decreases the bias that moderators have against unregistered editors.

This finding is particularly surprising because the ORES algorithm we analyzed was itself demonstrably biased against unregistered editors (i.e., the algorithm tended to greatly overestimate the probability that edits by these editors were damaging). Despite the fact that the algorithms were biased, their introduction could still lead to less biased outcomes overall.

Our work shows that although it is important to design predictive algorithms to avoid such biases, it is equally important to study fairness at the level of the broader sociotechnical system. Since we first published a preprint of our paper, a follow-up piece by Leijie Wang and Haiyi Zhu replicated much of our work and showed that differences between different Wikipedia communities may be another important factor driving the effect of the system. Overall, this work suggests that social signals and social context can interact with algorithmic signals, and together these can influence behavior in important and unexpected ways.

The full citation for the paper is: TeBlunthuis, Nathan, Benjamin Mako Hill, and Aaron Halfaker. 2021. “Effects of Algorithmic Flagging on Fairness: Quasi-Experimental Evidence from Wikipedia.” Proceedings of the ACM on Human-Computer Interaction 5 (CSCW): 56:1-56:27. https://doi.org/10.1145/3449130.

We have also released replication materials for the paper, including all the data and code used to conduct the analysis and compile the paper itself.

Informal online learning communities are one of the most exciting and successful ways to engage young people in technology. As the most successful example of the approach, over 40 million children from around the world have created accounts on the Scratch online community where they learn to code by creating interactive art, games, and stories. However, despite its enormous reach and its focus on inclusiveness, participation in Scratch is not as broad as one would hope. For example, reflecting a trend in the broader computing community, more boys have signed up on the Scratch website than girls.

In a recently published paper, I worked with several colleagues from the Community Data Science Collective to unpack the dynamics of unequal participation by gender in Scratch by looking at whether Scratch users choose to share the projects they create. Our analysis took advantage of the fact that less than a third of projects created in Scratch are ever shared publicly. By never sharing, creators never open themselves to the benefits associated with interaction, feedback, socialization, and learning—all things that research has shown participation in Scratch can support.

Overall, we found that boys on Scratch share their projects at a slightly higher rate than girls. Digging deeper, we found that this overall average hid an important dynamic that emerged over time. The graph below shows the proportion of Scratch projects shared for male and female Scratch users’ 1st created projects, 2nd created projects, 3rd created projects, and so on. It reflects the fact that although girls share less often initially, this trend flips over time. Experienced girls share much more than often than boys!

We unpacked this dynamic using a series of statistical models estimated using data from over 5 million projects by over a million Scratch users. This set of analyses echoed our earlier preliminary finding—while girls were less likely to share initially, more experienced girls shared projects at consistently higher rates than boys. We further found that initial differences in sharing between boys and girls could be explained by controlling for differences in project complexity and in the social connectedness of the project creator.

Another surprising finding is that users who had received more positive peer feedback, at least as measured by receipt of “love its” (similar to “likes” on Facebook), were less likely to share their subsequent projects than users who had received less. This relation was especially strong for boys and for more experienced Scratch users. We speculate that this could be due to a phenomenon known in the music industry as “sophomore album syndrome” or “second album syndrome”—a term used to describe a musician who has had a successful first album but struggles to produce a second because of increased pressure and expectations caused by their previous success

This blog post (published first on the Community Data Science Collective blog) and the paper are collaborative work with Emilia Gan and Sayamindu Dasgupta. You can find more details about our methodology and results in the text of our paper, “Gender, Feedback, and Learners’ Decisions to Share Their Creative Computing Projects” which is freely available and published open access in the Proceedings of the ACM on Human-Computer Interaction 2 (CSCW): 54:1-54:23.

Leaders and scholars of online communities tend of think of community growth as the aggregate effect of inexperienced individuals arriving one-by-one. However, there is increasing evidence that growth in many online communities today involves newcomers arriving in groups with previous experience together in other communities. This difference has deep implications for how we think about the process of integrating newcomers. Instead of focusing only on individual socialization into the group culture, we must also understand how to manage mergers of existing groups with distinct cultures. Unfortunately, online community mergers have, to our knowledge, never been studied systematically.

To better understand mergers, my student Charlie Kiene spent six months in 2017 conducting ethnographic participant observation in two World of Warcraft raid guilds planning and undergoing mergers. The results—visible in the attendance plot below—shows that the top merger led to a thriving and sustainable community while the bottom merger led to failure and the eventual dissolution of the group. Why did one merger succeed while the other failed? What can managers of other communities learn from these examples?

In a new paper that will be published in the Proceedings of of the ACM Conference on Computer-supported Cooperative Work and Social Computing (CSCW) and that Charlie will present in New Jersey next month, I teamed up with Charlie and Aaron Shaw try to answer these questions.

In our research setting, World of Warcraft (WoW), players form organized groups called “guilds” to take on the game’s toughest bosses in virtual dungeons that are called “raids.” Raids can be extremely challenging, and they require a large number of players to be successful. Below is a video demonstrating the kind of communication and coordination needed to be successful as a raid team in WoW.

Because participation in a raid guild requires time, discipline, and emotional investment, raid guilds are constantly losing members and recruiting new ones to resupply their ranks. One common strategy for doing so is arranging formal mergers. Our study involved following two such groups as they completed mergers. To collect data for our study, Charlie joined both groups, attended and recorded all activities, took copious field notes, and spent hours interviewing leaders.

Although our team did not anticipate the divergent outcomes shown in the figure above when we began, we analyzed our data with an eye toward identifying themes that might point to reasons for the success of one merger and the failure of the other. The answers that emerged from our analysis suggest that the key differences that led one merger to be successful and the other to fail revolved around differences in the ways that the two mergers managed organizational culture. This basic insight is supported by a body of research about organizational culture in firms but seem to have not made it onto the radar of most members or scholars of online communities. My coauthors and I think more attention to the role that organizational culture plays in online communities is essential.

We found evidence of cultural incompatibility in both mergers and it seems likely that some degree of cultural clashes is inevitable in any merger. The most important result of our analysis are three observations we drew about specific things that the successful merger did to effectively manage organizational culture. Drawn from our analysis, these themes point to concrete things that other communities facing mergers—either formal or informal—can do.

First, when planning mergers, groups can strategically select other groups with similar organizational culture. The successful merger in our study involved a carefully planned process of advertising for a potential merger on forums, testing out group compatibility by participating in “trial” raid activities with potential guilds, and selecting the guild that most closely matched their own group’s culture. In our settings, this process helped prevent conflict from emerging and ensured that there was enough common ground to resolve it when it did.

Second, leaders can plan intentional opportunities to socialize members of the merged or acquired group. The leaders of the successful merger held community-wide social events in the game to help new members learn their community’s norms. They spelled out these norms in a visible list of rules. They even included the new members in both the brainstorming and voting process of changing the guild’s name to reflect that they were a single, new, cohesive unit. The leaders of the failed merger lacked any explicitly stated community rules, and opportunities for socializing the members of the new group were virtually absent. Newcomers from the merged group would only learn community norms when they broke one of the unstated social codes.

Third and finally, our study suggested that social activities can be used to cultivate solidarity between the two merged groups, leading to increased retention of new members. We found that the successful guild merger organized an additional night of activity that was socially-oriented. In doing so, they provided a setting where solidarity between new and existing members can cultivate and motivate their members to stick around and keep playing with each other—even when it gets frustrating.

Our results suggest that by preparing in advance, ensuring some degree of cultural compatibility, and providing opportunities to socialize newcomers and cultivate solidarity, the potential for conflict resulting from mergers can be mitigated. While mergers between firms often occur to make more money or consolidate resources, the experience of the failed merger in our study shows that mergers between online communities put their entire communities at stake. We hope our work can be used by leaders in online communities to successfully manage potential conflict resulting from merging or acquiring members of other groups in a wide range of settings.

Much more detail is available our paper which will be published open access and which is currently available as a preprint.

Both this blog post and the paper it is based on are collaborative work by Charles Kiene from the University of Washington, Aaron Shaw from Northwestern University, and Benjamin Mako Hill from the University of Washington. We are also thrilled to mention that the paper received a Best Paper Honorable Mention award at CSCW 2018!

Couchsurfing and Airbnb are websites that connect people with an extra guest room or couch with random strangers on the Internet who are looking for a place to stay. Although Couchsurfing predates Airbnb by about five years, the two sites are designed to help people do the same basic thing and they work in extremely similar ways. They differ, however, in one crucial respect. On Couchsurfing, the exchange of money in return for hosting is explicitly banned. In other words, couchsurfing only supports the social exchange of hospitality. On Airbnb, users must use money: the website is a market on which people can buy and sell hospitality.

The figure above compares the number of people with at least some trust or verification on both  Couchsurfing and Airbnb based on when each user signed up. The picture, as I have argued elsewhere, reflects a broader pattern that has occurred on the web over the last 15 years. Increasingly, social-based systems of production and exchange, many like Couchsurfing created during the first decade of the Internet boom, are being supplanted and eclipsed by similar market-based players like Airbnb.

In a paper led by Max Klein that was recently published and will be presented at the ACM Conference on Computer-supported Cooperative Work and Social Computing (CSCW) which will be held in Jersey City in early November 2018, we sought to provide a window into what this change means and what might be at stake. At the core of our research were a set of interviews we conducted with “dual-users” (i.e. users experienced on both Couchsurfing and Airbnb). Analyses of these interviews pointed to three major differences, which we explored quantitatively from public data on the two sites.

First, we found that users felt that hosting on Airbnb appears to require higher quality services than Couchsurfing. For example, we found that people who at some point only hosted on Couchsurfing often said that they did not host on Airbnb because they felt that their homes weren’t of sufficient quality. One participant explained that:

“I always wanted to host on Airbnb but I didn’t actually have a bedroom that I felt would be sufficient for guests who are paying for it.”

An another interviewee said:

“If I were to be paying for it, I’d expect a nice stay. This is why I never Airbnb-hosted before, because recently I couldn’t enable that [kind of hosting].”

We conducted a quantitative analysis of rates of Airbnb and Couchsurfing in different cities in the United States and found that median home prices are positively related to number of per capita Airbnb hosts and a negatively related to the number of Couchsurfing hosts. Our exploratory models predicted that for each $100,000 increase in median house price in a city, there will be about 43.4 more Airbnb hosts per 100,000 citizens, and 3.8 fewer hosts on Couchsurfing.

A second major theme we identified was that, while Couchsurfing emphasizes people, Airbnb places more emphasis on places. One of our participants explained:

“People who go on Airbnb, they are looking for a specific goal, a specific service, expecting the place is going to be clean […] the water isn’t leaking from the sink. I know people who do Couchsurfing even though they could definitely afford to use Airbnb every time they travel, because they want that human experience.”

In a follow-up quantitative analysis we conducted of the profile text from hosts on the two websites with a commonly-used system for text analysis called LIWC, we found that, compared to Couchsurfing, a lower proportion of words in Airbnb profiles were classified as being about people while a larger proportion of words were classified as being about places.

Finally, our research suggested that although hosts are the powerful parties in exchange on Couchsurfing, social power shifts from hosts to guests on Airbnb. Reflecting a much broader theme in our interviews, one of our participants expressed this concisely, saying:

“On Airbnb the host is trying to attract the guest, whereas on Couchsurfing, it works the other way round. It’s the guest that has to make an effort for the host to accept them.”

Previous research on Airbnb has shown that guests tend to give their hosts lower ratings than vice versa. Sociologists have suggested that this asymmetry in ratings will tend to reflect the direction of underlying social power balances.

We both replicated this finding from previous work and found that, as suggested in our interviews, the relationship is reversed on Couchsurfing. As shown in the figure above, we found Airbnb guests will typically give a less positive review to their host than vice-versa while in Couchsurfing guests will typically a more positive review to the host.

As Internet-based hospitality shifts from social systems to the market, we hope that our paper can point to some of what is changing and some of what is lost. For example, our first result suggests that less wealthy participants may be cut out by market-based platforms. Our second theme suggests a shift toward less human-focused modes of interaction brought on by increased “marketization.” We see the third theme as providing somewhat of a silver-lining in that shifting power toward guests was seen by some of our participants as a positive change in terms of safety and trust in that guests. Travelers in unfamiliar places often are often vulnerable and shifting power toward guests can be helpful.

Although our study is only of Couchsurfing and Airbnb, we believe that the shift away from social exchange and toward markets has broad implications across the sharing economy. We end our paper by speculating a little about the generalizability of our results. I have recently spoken at much more length about the underlying dynamics driving the shift we describe in  my recent LibrePlanet keynote address.

More details are available in our paper which we have made available as a preprint on our website. The final version is behind a paywall in the ACM digital library.

In 1965, Bruce Tuckman proposed a “developmental sequence in small groups.” According to his influential theory, most successful groups go through four stages with rhyming names:

1. Forming: Group members get to know each other and define their task.
2. Storming: Through argument and disagreement, power dynamics emerge and are negotiated.
3. Norming: After conflict, groups seek to avoid conflict and focus on cooperation and setting norms for acceptable behavior.
4. Performing: There is both cooperation and productive dissent as the team performs the task at a high level.

Fortunately for organizational science, 1965 was hardly the last stage of development for Tuckman’s theory!

Twelve years later, Tuckman suggested that adjourning or mourning reflected potential fifth stages (Tuckman and Jensen 1977). Since then, other organizational researchers have suggested other stages including transforming and reforming (White 2009), re-norming (Biggs), and outperforming (Rickards and Moger 2002).

What does the future hold for this line of research?

To help answer this question, we wrote a regular expression to identify candidate words and placed the full list is at this page in the Community Data Science Collective wiki.

The good news is that despite the active stream of research producing new stages that end or rhyme with -orming, there are tons of great words left!

For example, stages in a group’s development might include:

• Scorning: In this stage, group members begin mocking each other!
• Misinforming: Group that reach this stage start producing fake news.
• Shoehorning: These groups try to make their products fit into ridiculous constraints.
• Chloroforming: Groups become languid and fatigued?

One benefit of keeping our list in the wiki is that the organizational research community can use it to coordinate! If you are planning to use one of these terms—or if you know of a paper that has—feel free to edit the page in our wiki to “claim” it!

Also posted on the Community Data Science Collective blog. Although credit for this post goes primarily to Jeremy Foote and Benjamin Mako Hill, the other Community Data Science Collective members can’t really be called blameless in the matter either.

Seymour Papert is credited as saying that tools to support learning should have “high ceilings” and “low floors.” The phrase is meant to suggest that tools should allow learners to do complex and intellectually sophisticated things but should also be easy to begin using quickly. Mitchel Resnick extended the metaphor to argue that learning toolkits should also have “wide walls” in that they should appeal to diverse groups of learners and allow for a broad variety of creative outcomes. In a new paper, Sayamindu Dasgupta and I attempted to provide an empirical test of Resnick’s wide walls theory. Using a natural experiment in the Scratch online community, we found causal evidence that “widening walls” can, as Resnick suggested, increase both engagement and learning.

Over the last ten years, the “wide walls” design principle has been widely cited in the design of new systems. For example, Resnick and his collaborators relied heavily on the principle in the design of the Scratch programming language. Scratch allows young learners to produce not only games, but also interactive art, music videos, greetings card, stories, and much more. As part of that team, Sayamindu was guided by “wide walls” principle when he designed and implemented the Scratch cloud variables system in 2011-2012.

While designing the system, Sayamindu hoped to “widen walls” by supporting a broader range of ways to use variables and data structures in Scratch. Scratch cloud variables extend the affordances of the normal Scratch variable by adding persistence and shared-ness. A simple example of something possible with cloud variables, but not without them, is a global high-score leaderboard in a game (example code is below). After the system was launched, we saw many young Scratch users using the system to engage with data structures in new and incredibly creative ways.

Although these examples reflected powerful anecdotal evidence, we were also interested in using quantitative data to reflect the causal effect of the system. Understanding the causal effect of a new design in real world settings is a major challenge. To do so, we took advantage of a “natural experiment” and some clever techniques from econometrics to measure how learners’ behavior changed when they were given access to a wider design space.

Understanding the design of our study requires understanding a little bit about how access to the Scratch cloud variable system is granted. Although the system has been accessible to Scratch users since 2013, new Scratch users do not get access immediately. They are granted access only after a certain amount of time and activity on the website (the specific criteria are not public). Our “experiment” involved a sudden change in policy that altered the criteria for who gets access to the cloud variable feature. Through no act of their own, more than 14,000 users were given access to feature, literally overnight. We looked at these Scratch users immediately before and after the policy change to estimate the effect of access to the broader design space that cloud variables afforded.

We found that use of data-related features was, as predicted, increased by both access to and use of cloud variables. We also found that this increase was not only an effect of projects that use cloud variables themselves. In other words, learners with access to cloud variables—and especially those who had used it—were more likely to use “plain-old” data-structures in their projects as well.

The graph below visualizes the results of one of the statistical models in our paper and suggests that we would expect that 33% of projects by a prototypical “average” Scratch user would use data structures if the user in question had never used used cloud variables but that we would expect that 60% of projects by a similar user would if they had used the system.

It is important to note that the estimated effective above is a “local average effect” among people who used the system because they were granted access by the sudden change in policy (this is a subtle but important point that we explain this in some depth in the paper). Although we urge care and skepticism in interpreting our numbers, we believe our results are encouraging evidence in support of the “wide walls” design principle.

Of course, our work is not without important limitations. Critically, we also found that rate of adoption of cloud variables was very low. Although it is hard to pinpoint the exact reason for this from the data we observed, it has been suggested that widening walls may have a potential negative side-effect of making it harder for learners to imagine what the new creative possibilities might be in the absence of targeted support and scaffolding. Also important to remember is that our study measures “wide walls” in a specific way in a specific context and that it is hard to know how well our findings will generalize to other contexts and communities. We discuss these caveats, as well as our methods, models, and theoretical background in detail in our paper which now available for download as an open-access piece from the ACM digital library.

This blog post, and the open access paper that it describes, is a collaborative project with Sayamindu Dasgupta. Financial support came from the eScience Institute and the Department of Communication at the University of Washington. Quantitative analyses for this project were completed using the Hyak high performance computing cluster at the University of Washington.

Scratch is a block-based programming language created by the Lifelong Kindergarten Group (LLK) at the MIT Media Lab. Scratch gives kids the power to use programming to create their own interactive animations and computer games. Since 2007, the online community that allows Scratch programmers to share, remix, and socialize around their projects has drawn more than 16 million users who have shared nearly 20 million projects and more than 100 million comments. It is one of the most popular ways for kids to learn programming and among the larger online communities for kids in general.

Since 2010, I have published a series of papers using quantitative data collected from the database behind the Scratch online community. As the source of data for many of my first quantitative and data scientific papers, it’s not a major exaggeration to say that I have built my academic career on the dataset.

I was able to do this work because I happened to be doing my masters in a research group that shared a physical space (“The Cube”) with LLK and because I was friends with Andrés Monroy-Hernández, who started in my masters cohort at the Media Lab. A year or so after we met, Andrés conceived of the Scratch online community and created the first version for his masters thesis project. Because I was at MIT and because I knew the right people, I was able to get added to the IRB protocols and jump through the hoops necessary to get access to the database.

Over the years, Andrés and I have heard over and over, in conversation and in reviews of our papers, that we were privileged to have access to such a rich dataset. More than three years ago, Andrés and I began trying to figure out how we might broaden this access. Andrés had the idea of taking advantage of the launch of Scratch 2.0 in 2013 to focus on trying to release the first five years of Scratch 1.x online community data (March 2007 through March 2012) — most of the period that the codebase he had written ran the site.

After more work than I have put into any single research paper or project, Andrés and I have published a data descriptor in Nature’s new journal Scientific Data. This means that the data is now accessible to other researchers. The data includes five years of detailed longitudinal data organized in 32 tables with information drawn from more than 1 million Scratch users, nearly 2 million Scratch projects, more than 10 million comments, more than 30 million visits to Scratch projects, and much more. The dataset includes metadata on user behavior as well the full source code for every project. Alongside the data is the source code for all of the software that ran the website and that users used to create the projects as well as the code used to produce the dataset we’ve released.

Releasing the dataset was a complicated process. First, we had navigate important ethical concerns about the the impact that a release of any data might have on Scratch’s users. Toward that end, we worked closely with the Scratch team and the the ethics board at MIT to design a protocol for the release that balanced these risks with the benefit of a release. The most important features of our approach in this regard is that the dataset we’re releasing is limited to only public data. Although the data is public, we understand that computational access to data is different in important ways to access via a browser or API. As a result, we’re requiring anybody interested in the data to tell us who they are and agree to a detailed usage agreement. The Scratch team will vet these applicants. Although we’re worried that this creates a barrier to access, we think this approach strikes a reasonable balance.

Beyond the the social and ethical issues, creating the dataset was an enormous task. Andrés and I spent Sunday afternoons over much of the last three years going column-by-column through the MySQL database that ran Scratch. We looked through the source code and the version control system to figure out how the data was created. We spent an enormous amount of time trying to figure out which columns and rows were public. Most of our work went into creating detailed codebooks and documentation that we hope makes the process of using this data much easier for others (the data descriptor is just a brief overview of what’s available). Serializing some of the larger tables took days of computer time.

In this process, we had a huge amount of help from many others including an enormous amount of time and support from Mitch Resnick, Natalie Rusk, Sayamindu Dasgupta, and Benjamin Berg at MIT as well as from many other on the Scratch Team. We also had an enormous amount of feedback from a group of a couple dozen researchers who tested the release as well as others who helped us work through through the technical, social, and ethical challenges. The National Science Foundation funded both my work on the project and the creation of Scratch itself.

Because access to data has been limited, there has been less research on Scratch than the importance of the system warrants. We hope our work will change this. We can imagine studies using the dataset by scholars in communication, computer science, education, sociology, network science, and beyond. We’re hoping that by opening up this dataset to others, scholars with different interests, different questions, and in different fields can benefit in the way that Andrés and I have. I suspect that there are other careers waiting to be made with this dataset and I’m excited by the prospect of watching those careers develop.

You can find out more about the dataset, and how to apply for access, by reading the data descriptor on Nature’s website.

As children use digital media to learn and socialize, others are collecting and analyzing data about these activities. In school and at play, these children find that they are the subjects of data science. As believers in the power of data analysis, we believe that this approach falls short of data science’s potential to promote innovation, learning, and power.

Motivated by this fact, we have been working over the last three years as part of a team at the MIT Media Lab and the University of Washington to design and build a system that attempts to support an alternative vision: children as data scientists. The system we have built is described in a new paper—Scratch Community Blocks: Supporting Children as Data Scientists—that will be published in the proceedings of CHI 2017.

Our system is built on top of Scratch, a visual, block-based programming language designed for children and youth. Scratch is also an online community with over 15 million registered members who share their Scratch projects, remix each others’ work, have conversations, provide feedback, bookmark or “love” projects they like, follow other users, and more. Over the last decade, researchers—including us—have used the Scratch online community’s database to study the youth using Scratch. With Scratch Community Blocks, we attempt to put the power to programmatically analyze these data into the hands of the users themselves.

To do so, our new system adds a set of new programming primitives (blocks) to Scratch so that users can access public data from the Scratch website from inside Scratch. Blocks in the new system gives users access to project and user metadata, information about social interaction, and data about what types of code are used in projects. The full palette of blocks to access different categories of data is shown below.

The new blocks allow users to programmatically access, filter, and analyze data about their own participation in the community. For example, with the simple script below, we can find whether we have followers in Scratch who report themselves to be from Spain, and what their usernames are.

In designing the system, we had two primary motivations. First, we wanted to support avenues through which children can engage in curiosity-driven, creative explorations of public Scratch data. Second, we wanted to foster self-reflection with data. As children looked back upon their own participation and coding activity in Scratch through the project they and their peers made, we wanted them to reflect on their own behavior and learning in ways that shaped their future behavior and promoted exploration.

After designing and building the system over 2014 and 2015, we invited a group of active Scratch users to beta test the system in early 2016. Over four months, 700 users created more than 1,600 projects. The diversity and depth of users creativity with the new blocks surprised us. Children created projects that gave the viewer of the project a personalized doughnut-chart visualization of their coding vocabulary on Scratch, rendered the viewer’s number of followers as scoops of ice-cream on a cone, attempted to find whether “love-its” for projects are more common on Scratch than “favorites”, and told users how “talkative” they were by counting the cumulative string-length of project titles and descriptions.

We found that children, rather than making canonical visualizations such as pie-charts or bar-graphs, frequently made information representations that spoke to their own identities and aesthetic sensibilities. A 13-year-old girl had made a virtual doll dress-up game where the player’s ability to buy virtual clothes and accessories for the doll was determined by the level of their activity in the Scratch community. When we asked about her motivation for making such a project, she said:

I was trying to think of something that somebody hadn’t done yet, and I didn’t see that. And also I really like to do art on Scratch and that was a good opportunity to use that and mix the two [art and data] together.

We also found at least some evidence that the system supported self-reflection with data. For example, after seeing a project that showed its viewers a visualization of their past coding vocabulary, a 15-year-old realized that he does not do much programming with the pen-related primitives in Scratch, and wrote in a comment, “epic! looks like we need to use more pen blocks. :D.”

Additionally, we noted that that as children made and interacted with projects made with Scratch Community Blocks, they started to critically think about the implications of data collection and analysis. These conversations are the subject of another paper (also being published in CHI 2017).

In a 1971 article called “Teaching Children to be Mathematicians vs. Teaching About Mathematics”, Seymour Papert argued for the need for children doing mathematics vs. learning about it. He showed how Logo, the programming language he was developing at that time with his colleagues, could offer children a space to use and engage with mathematical ideas in creative and personally motivated ways. This, he argued, enabled children to go beyond knowing about mathematics to “doing” mathematics, as a mathematician would.

Scratch Community Blocks has not yet been launched for all Scratch users and has several important limitations we discuss in the paper. That said, we feel that the projects created by children in our the beta test demonstrate the real potential for children to do data science, and not just know about it, provide data for it, and to have their behavior nudged and shaped by it.

With more than 10 million users, the Scratch online community is the largest online community where kids learn to program. Since it was created, a central goal of the community has been to promote “remixing” — the reworking and recombination of existing creative artifacts. As the video above shows, remixing programming projects in the current web-based version of Scratch is as easy is as clicking on the “see inside” button in a project web-page, and then clicking on the “remix” button in the web-based code editor. Today, close to 30% of projects on Scratch are remixes.

Remixing plays such a central role in Scratch because its designers believed that remixing can play an important role in learning. After all, Scratch was designed first and foremost as a learning community with its roots in the Constructionist framework developed at MIT by Seymour Papert and his colleagues. The design of the Scratch online community was inspired by Papert’s vision of a learning community similar to Brazilian Samba schools (Henry Jenkins writes about his experience of Samba schools in the context of Papert’s vision here), and a comment Marvin Minsky made in 1984:

Adults worry a lot these days. Especially, they worry about how to make other people learn more about computers. They want to make us all “computer-literate.” Literacy means both reading and writing, but most books and courses about computers only tell you about writing programs. Worse, they only tell about commands and instructions and programming-language grammar rules. They hardly ever give examples. But real languages are more than words and grammar rules. There’s also literature – what people use the language for. No one ever learns a language from being told its grammar rules. We always start with stories about things that interest us.

In a new paper — titled “Remixing as a pathway to Computational Thinking” — that was recently published at the ACM Conference on Computer Supported Collaborative Work and Social Computing (CSCW) conference, we used a series of quantitative measures of online behavior to try to uncover evidence that might support the theory that remixing in Scratch is positively associated with learning.

Of course, because Scratch is an informal environment with no set path for users, no lesson plan, and no quizzes, measuring learning is an open problem. In our study, we built on two different approaches to measure learning in Scratch. The first approach considers the number of distinct types of programming blocks available in Scratch that a user has used over her lifetime in Scratch (there are 120 in total) — something that can be thought of as a block repertoire or vocabulary. This measure has been used to model informal learning in Scratch in an earlier study. Using this approach, we hypothesized that users who remix more will have a faster rate of growth for their code vocabulary.

Controlling for a number of factors (e.g. age of user, the general level of activity) we found evidence of a small, but positive relationship between the number of remixes a user has shared and her block vocabulary as measured by the unique blocks she used in her non-remix projects. Intriguingly, we also found a strong association between the number of downloads by a user and her vocabulary growth. One interpretation is that this learning might also be associated with less active forms of appropriation, like the process of reading source code described by Minksy.

The second approach we used considered specific concepts in programming, such as loops, or event-handling. To measure this, we utilized a mapping of Scratch blocks to key programming concepts found in this paper by Karen Brennan and Mitchel Resnick. For example, in the image below are all the Scratch blocks mapped to the concept of “loop”.

We looked at six concepts in total (conditionals, data, events, loops, operators, and parallelism). In each case, we hypothesized that if someone has had never used a given concept before, they would be more likely to use that concept after encountering it while remixing an existing project.

Using this second approach, we found that users who had never used a concept were more likely to do so if they had been exposed to the concept through remixing. Although some concepts were more widely used than others, we found a positive relationship between concept use and exposure through remixing for each of the six concepts. We found that this relationship was true even if we ignored obvious examples of cutting and pasting of blocks of code. In all of these models, we found what we believe is evidence of learning through remixing.

Of course, there are many limitations in this work. What we found are all positive correlations — we do not know if these relationships are causal. Moreover, our measures do not really tell us whether someone has “understood” the usage of a given block or programming concept.However, even with these limitations, we are excited by the results of our work, and we plan to build on what we have. Our next steps include developing and utilizing better measures of learning, as well as looking at other methods of appropriation like viewing the source code of a project.

This blog post and the paper it describes are collaborative work with Sayamindu Dasgupta, Andrés Monroy-Hernández, and William Hale. The paper is released as open access so anyone can read the entire paper here. This blog post was also posted on Sayamindu Dasgupta’s blog and on Medium by the MIT Media Lab.