We sociologists love the Matthew effect. An great new article on this topic was just published in Proceedings of the National Academy of Sciences by my advisor Arnout van de Rijt and several co-authors, including me. We demonstrate how initial, arbitrary advantages can lead to higher rates of subsequent success.
If you are interested, here’s the abstract. But why not read the entire paper itself — it’s open access!
Seemingly similar individuals often experience drastically different success trajectories, with some repeatedly failing and others consistently succeeding. One explanation is preexisting variability along unobserved fitness dimensions that is revealed gradually through differential achievement. Alternatively, positive feedback operating on arbitrary initial advantages may increasingly set apart winners from losers, producing runaway inequality. To identify social feedback in human reward systems, we conducted randomized experiments by intervening in live social environments across the domains of funding, status, endorsement, and reputation. In each system we consistently found that early success bestowed upon arbitrarily selected recipients produced significant improvements in subsequent rates of success compared with the control group of nonrecipients. However, success exhibited decreasing marginal returns, with larger initial advantages failing to produce much further differentiation. These findings suggest a lesser degree of vulnerability of reward systems to incidental or fabricated advantages and a more modest role for cumulative advantage in the explanation of social inequality than previously thought.
I was invited to speak at the Provost’s Graduate Student Lecture series at Stony Brook University. Three PhD students are selected to present their research, which highlights some of the most promising scholarship being produced at Stony Brook. Thanks to Tim Moran and Arnout van de Rijt for nominating me.
Here is the abstract of the talk:
Why do people volunteer thousands of hours of free labor to create public goods online such as Wikipedia and open-source software? In this talk, I present evidence from a series of experiments designed to test the theory that in the absence of formal or material incentives, contributors’ efforts can be sustained through community recognition of work performed. According to this theory, there is a social solution to the problem of collective action, where individuals are believed to have a tendency to ‘free ride’ rather than contribute to the public good: groups reward individuals who work hard with status in the community, which in turn motivates additional effort, fostering a virtuous cycle of work and reward. Results from the experiments on Wikipedia’s editing community show that highly productive contributors who are given an informal award, or “Barnstar,” subsequently increase their workload by about 40% compared to a control group. These recipients also accumulate additional awards from third parties at a six-times higher rate than non-rewarded peers, suggesting that pro-social behavior can help sustain contributions to public goods by fostering a self-reinforcing dynamic. I conclude by calculating the surprising magnitude of additional volunteer labor that is generated from such pro-social behaviors.
The theme for the 2013 ASA conference (Linking Micro and Macro) was a good fit for my presentation with Arnout: “No Praise without Effort: Experimental Evidence on the Matthew Effect in Wikipedia.”
The CITASA round table session was well-attended: I don’t remember ever seeing so many people crammed around one table, and in fact there was a second ring of chairs around the presenters! Our paper provided an intriguing contrast of the virtues of “big data” and “small data” – in our case, how small data from an experiment can clarify the causal processes observed in big data studies. Our paper is currently under review at PLoS ONE as a follow-up to our previous study.
Very exciting news: my dissertation improvement grant (well, to be technically correct, Arnout’s dissertation improvement grant) was ranked as “fundable” by the NSF! While this is actually “old” news to me — I received notification in January — I have finally received the official award letter from the NSF. Feels good to know that these dollars won’t be sequestered.
If anyone is interested, here is the research abstract:
Doctoral Dissertation Research: Group Diversity and Knowledge Production
Women’s historical absences from scientific endeavors and other institutions that create knowledge is also seen in the contemporary practice of online peer production. Recent survey research of contributors to Wikipedia, a prominent example of peer produced knowledge, has found that less than 15 percent of its contributors are women. The causes for this high degree of gender inequality have been linked to both self-selection and barriers to participation: we continue to imagine expertise in creating knowledge as a masculine pursuit, and women may buy into this belief as well, undermining their confidence in participating in online knowledge production. However, this gender gap has been shown to be more than simply self-selection: research suggests that some aspects of online participatory culture serve to limit women’s participation due to excessive conflict or contentiousness, devaluation of certain topics or perspectives, and in some instances, overt hostility or other forms of misogyny.
This raises several interesting questions about whether online realms are open to a diverse range of participants and whether they can ever truly represent “disembodied spaces” if participants’ socially-learned and embodied gender, and others’ perceptions thereof, accompany them into virtual spaces. One way that gender may continue to be salient is through presentation – in particular, whether contributors choose to publicly disclose their gender to others in an online community, which can affect how group members perceive one another and alter group interactions. A second way that gender may be salient is through socially-learned gender performances in terms of the roles people adopt and the types of work they perform.
In this research, I analyze the ways in which gender diversity affects the quality of work produced by contributors to Wikipedia. I use the quality of the information in Wikipedia’s encyclopedia articles as an indicator of the performance of the groups who author those articles. Organizational research has long known that diversity – in respect to group composition, the types of tasks being performed, and other group dynamics – affects a group’s performance and shapes its collective output. For example, the quality of work produced by Wikipedia contributors has been shown to vary by the functional diversity (differences in types of work roles performed) and cognitive diversity (differences in the knowledge bases) of group members. However, the potential interaction with gender not been explored in any detail, which is surprising since research on gender differences in online communication and interaction point to the need to incorporate gender more thoroughly into our conceptualization of the problem.
Studying the work of Wikipedia contributors provides an invaluable opportunity to assess how gender diversity, and the different ways of doing gender, affects group performance and shapes the quality of peer produced knowledge. Preliminary analysis of a subset of Wikipedia articles shows that gender diversity leads to favorable outcomes for article quality. The findings from this research will be of interest to gender scholars, organizational researchers, and others who study of online collaboration. More broadly, understanding why and when diversity can be beneficial to groups will be of practical relevance to the people who are building and working within these types of online organizations. Particularly, Wikipedia can take steps to shrink its gender gap by identifying and addressing barriers to inclusive participation from a diverse range of contributors.
This grant will ensure speedy completion of this chapter in my dissertation. I am very excited to be able to pursue this project with institutional support and recognition, because I have long argued that this topic requires attention from sociologists who care about new forms of gender inequality.
This is a great opportunity to expand my research with “big data” capabilities. As a critical-reflexive sociologist, I am wary of the the over-exuberance surrounding this latest trend in science; given the choice, I always prefer better, richer, or more meaningful data rather than simply more data. However, I am very excited about the Institute’s focus on inter-disciplinary collaboration, as well as the obvious methodological sophistication that comes with advanced modeling and statistical computing.
Here is an excellent short talk by Arnout about how the IACS will benefit sociologists trying to better understand the social world:
I attended ASA’s professional conference in Denver where I was on a very thought-provoking panel about Group Procceses.
Organizer was Stephen Benard at Indiana University. Discussant was Robb Willer at UC Berkley. All of the papers had to do with social status in one way or another. I appreciate Stephen’s efforts at finding papers that contribute research on the same topic. Also, Robb’s comments and questions at the end of the session were spot-on and really challenged my thinking, so I am grateful to him for that as well.
My paper with Arnout van de Rijt, entitled “Experimental Study of Informal Rewards in Peer Production,” was based on our PLoS ONE article.
Interestingly, another team of researchers had a paper on the identical subject, which made for a nice discussion where we tried to resolve the different (but complementary) findings. That one was “Social Signaling and Collective Action: A Field Study of Awards on Wikipedia.” The authors were Benjamin Mako Hill at MIT, Aaron Shaw at UC Berkeley, and Yochai Benkler at Harvard.
The room was quite full, and it was a great opportunity to speak to a number of interested scholars about my research. Afterwards, I spoke to Mako a bit more about our papers and I think our work points in the same direction. Overall, a nice experience.