Research:How much do Wikipedians in the US value editing Wikipedia?

From Meta, a Wikimedia project coordination wiki
Created
17:15, 22 October 2017 (UTC)
Duration:  2017-12 – 2019-06
This page documents a completed research project.


Note: We are putting the study on hold due to low response rate. We will reach out to the community in the coming months to discuss best ways to get respondents for the survey.

Wikipedia is one of the premier sources of information worldwide, created and maintained by a community of volunteer editors. It generates a tremendous amount of economic value to readers (who can read articles for free) and editors (who contribute towards editing and maintaining Wikipedia without obtaining any monetary compensation in return). However, this value is currently unmeasured since users do not pay to read Wikipedia and editors do not get paid to edit Wikipedia. The standard methods of estimating the economic value that is generated by different goods and services hence fall short.

This research project aims to measure the economic surplus created by Wikipedia to its editors and dig deeper into the mechanisms behind editors' motivations to contribute to Wikipedia. Results of this study will let us quantify the total surplus created by Wikipedia to its editors, heterogeneity in valuation across different groups of editors, and the underlying motivations behind why editors value contributing to the encyclopedia. These results will have two important implications. First, they will inform public policy by providing estimates of the value created by Wikipedia. Second, understanding the mechanisms behind editors' motivations will help the community and WMF further support and improve the experience of editors contributing to Wikipedia.

Background[edit]

This project has two main objectives.

First, we propose to measure the economic value created by Wikipedia to its editors. Previous research (Stern 2004[1]) has shown that scientists pay to be scientists, i.e. when confronted with multiple job offers from different firms, they choose to take an offer with lower wages at a firm if the latter allows them to write and publish their research. Inspired by this, we ask of how much money editors of Wikipedia would forego in order to keep access to editing Wikipedia, i.e. their willingness to accept (WTA) to forfeit the option of editing Wikipedia (for a given period of time). Aggregating survey responses from a representative sample of editors would let us estimate the demand curve for editing Wikipedia as well as the mean, median and range of valuations, differences in valuation of editing Wikipedia based on editor characteristics (gender, age, urban vs. rural location, income, profession: academics vs. non-academics, etc.) and prior involvement (power editors vs. ‘hobbyists’, proxied by labor hours per week).

Second, we seek to further develop our understanding of why editors value contributing to Wikipedia. Standard economic theory struggles to explain why an editor would voluntarily contribute a significant amount of their time towards editing Wikipedia without any monetary incentives. There is a growing stream of research within Economics and other fields studying contributions towards open-source software. Lerner and Tirole (2002)[2] find that career motivations (e.g. future job offers based on past open-source contributions) can explain why programmers would voluntarily contribute towards open-source projects. However, these studies are not directly applicable to Wikipedia contributions, since there is no direct career benefit in editing Wikipedia. Zhang and Zhu (2011)[3] exploit a natural experiment on Chinese Wikipedia and find that contributions of editors are linked with the total number of other editors and readers of Wikipedia. As this number drops, editors' contributions also decrease. They attribute this result to "social effects": the smaller the group size, the fewer the social benefits from editing. However, as the authors point out, a limitation of this study is that it does not allow the authors to disentangle the various motivations that actually give rise to these social effects (e.g., altruism vs. social image related benefits). We aim to fill this research gap and disentangle these motivations. Prior research suggests several motivations, such as pro-social motivation, learning, fun/enjoyment, and reputation building, and we propose to tease out these different sources (mechanisms) in our study.

Methods[edit]

We will attempt to recruit a maximum of 1000 editors to participate in a short survey by posting an invitation on their talk pages. The survey should take less than 10 minutes to complete. All responses will be kept strictly confidential and anonymized before the aggregate data are published. As a reward for their participation, 1 in 5 participants will be given a $25 voucher to purchase any goods of their choice from the Wikipedia store.

In the main part of the survey, we will ask how much money we would have to give a participant so that they stop editing Wikipedia for 1 month. This amount will be determined through the Becker–DeGroot–Marschak method. To make the responses incentive compatible, we will randomly draw 1 in 50 participants (i.e. a maximum of 20 editors if we recruit 1000 participants) and they will be given the option of claiming their disclosed amount if they do not edit Wikipedia for 1 month (note that they can change their mind and resume editing Wikipedia at any point in time). After we verify that they have not made any edits at the end of the month, we will send a voucher for the disclosed amount to these selected participants. The participants will always be free to revoke their decision and continue editing Wikipedia at any point during this period; in this case, they will just not receive the voucher.

To tease out various motivations for contributing to Wikipedia, we will randomly assign the participants to different groups. Within each group, we will manipulate the wording of the survey to focus on one of the motivational mechanisms (pro-social motivations, learning, fun/ enjoyment and reputation building). We will also ask anonymized demographic questions in the survey. We will complement this information with data on participants' edit history.

Survey and Invitation[edit]

Here is a link to the survey questions, and here is a link to the survey invitation that we will post on editors' talk pages. Here is a link to the final survey. We have now made it open for all US based Wikipedians to take part in it. Please fill it out.

Timeline[edit]

Study paused as of January 2019.

Date Goal
December 2017 Obtained IRB approval
November - December 2018 Run a pilot
November 2018 – May 2019 Run the main study
June - July 2019 Give rewards to selected participants
July- August 2019 Analyze data
August - October 2019 Write up a draft of the research paper, update the results section of this meta-wiki
October 2019 Upload the working paper online (ungated) on SSRN and/or authors' webpages

Policy, Ethics, and Human Subjects Research[edit]

We have obtained approval from the IRB (Institutional Review Boards) at MIT.

As noted above, the surveys will be short and will require less than 10–15 minutes to complete. On average, a participant will get a reward worth $5 for filling out the survey (1 in 5 participants get a voucher worth $25).

The survey will require at most 20 editors (1 in 50 out of a maximum of 1,000 participants) to stop editing Wikipedia for 1 month. However, they will be compensated the exact amount they ask for, and they will always have the choice to end their participation and forego the reward.

Results[edit]

We will post our results here once the study is complete. Any feedback on the study is welcome on the talk page.

References[edit]

  1. Stern, Scott (2004-06-01). "Do Scientists Pay to Be Scientists?". Management Science 50 (6): 835–853. ISSN 0025-1909. doi:10.1287/mnsc.1040.0241. 
  2. Lerner, Josh; Tirole, Jean (2002-06-01). "Some Simple Economics of Open Source". The Journal of Industrial Economics 50 (2): 197–234. ISSN 1467-6451. doi:10.1111/1467-6451.00174. 
  3. Zhang, Xiaoquan (Michael); Zhu, Feng. "Group Size and Incentives to Contribute: A Natural Experiment at Chinese Wikipedia". American Economic Review 101 (4): 1601–1615. doi:10.1257/aer.101.4.1601.