Motivating Contributions to Public Information Goods: A Field Experiment at Wikipedia

How to cite

Zhang, Fangzhou; Chen, Yan. 2018. "Motivating Contributions to Public Information Goods: A Field Experiment at Wikipedia." AEA RCT Registry. AEARCTR-0002920. https://doi.org/10.1257/rct.2920-5.0

Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.

Fangzhou Zhang* , Yan Chen

* Primary investigator

Registry ID
AEARCTR-0002920
DOI
https://doi.org/10.1257/rct.2920-5.0
First registered
Last updated
Status
Completed
Trial dates
2016-05-06 – 2016-12-22
Countries
United States of America ()
JEL codes
H4
Registry record
http://www.socialscienceregistry.org/trials/2920

Abstract

Wikipedia is among the most important information sources for the general public. Motivating domain experts to contribute to Wikipedia can improve the accuracy and completeness of its content. In a field experiment at Wikipedia, we examine individual motivations to contribute to public information goods. Using a 2-by-3 factorial design, we vary the expectation on the number of recipients along one dimension and the amount of private benefit along the other dimension. In the analysis, we will investigate how our interventions affect the experts' willingness to participate and contribution measured by both quantity and quality.

Intervention

We implement a 2-by-3 between-experts factorial design in which we vary the email content inviting experts' contribution to Wikipedia articles. Along one dimension, we vary the experts' expectation on the number of recipients of the public information goods which they contribute to. In the average view condition, we provide the experts with only the average number of views a typical Wikipedia article received in the past 30 days. This information serves to set the experts' expectation on the readership of a typical Wikipedia article. In the high view condition, we provide the experts with the additional information on the number of views the recommended articles received in the past 30 days.

Along the second dimension, we vary the experts' expectation on the amount of private benefit they receive from their contribution. We include three conditions: a baseline condition, a citation condition and an acknowledgement condition. The baseline condition serves as a control and no private benefit is mentioned in the email. In the citation condition, we mention that the articles recommended to the experts are likely to cite their research. The acknowledgement condition strengthens the private benefit by including acknowledgement as an additional benefit. The experts are told in the email message that their contributions will be addressed on a WikiProject Economics page at Wikipedia. WikiProject Economics is a collection of editors who work together as a team to improve articles related to economics. Being acknowledged for one's contribution in the WikiProject Economics thus serves as a private benefit in additional to the citation benefit. To avoid potential confound due to the experts' sequential contribution, we only post the acknowledgement to the contributions from the experts in our pilot stage and keep the acknowledgement page frozen at the through the main experiment.

Experimental design

We implement a 2-by-3 between-experts factorial design in which we vary the email content inviting experts' contribution to Wikipedia articles. Along one dimension, we vary the experts' expectation on the number of recipients of the public information goods which they contribute to. In the average view condition, we provide the experts with only the average number of views a typical Wikipedia article received in the past 30 days. This information serves to set the experts' expectation on the readership of a typical Wikipedia article. In the high view condition, we provide the experts with the additional information on the number of views the recommended articles received in the past 30 days.

Along the second dimension, we vary the experts' expectation on the amount of private benefit they receive from their contribution. We include three conditions: a baseline condition, a citation condition and an acknowledgement condition. The baseline condition serves as a control and no private benefit is mentioned in the email. In the citation condition, we mention that the articles recommended to the experts are likely to cite their research. The acknowledgement condition strengthens the private benefit by including acknowledgement as an additional benefit. The experts are told in the email message that their contributions will be addressed on a WikiProject Economics page at Wikipedia. WikiProject Economics is a collection of editors who work together as a team to improve articles related to economics. Being acknowledged for one's contribution in the WikiProject Economics thus serves as a private benefit in additional to the citation benefit. To avoid potential confound due to the experts' sequential contribution, we only post the acknowledgement to the contributions from the experts in our pilot stage and keep the acknowledgement page frozen at the through the main experiment.

Design details

Randomization method
randomization done in office by a computer
Randomization unit
expert
Clusters
4000 experts
Observations
4000 experts
Arms
600 experts per experimental condition
Primary outcomes
willingness to participate; number of words in the comment; median of overall quality of the comment

Cited by

No citation data available for this registration yet.

Keywords