Research:Codex/Background

[WikiPapers]

Summary

The success of Wikipedia (and, to some extent, its sister sites) has made it a popular subject of academic research. Hundreds of papers and theses have been published on topics relevant to the interests of the Wikimedia movement.

Due to the volume and diversity of this existing research, catching up on, and keeping up with, this body of knowledge is a difficult endeavor, whether for a researcher new to the field, a volunteer editor caring for their communities, or a Wikimedia Foundation designer creating a new feature for the software.

This page describes an initiative to review the relevant scientific production in an iterative manner, and synthesize it for wider consumption by researchers, practitioners (volunteers, WMF staff, affiliates), and interested third parties.

Previous efforts

Literature reviews

Literature reviews are contributions to a field of research that summarize the current state of knowledge about a particular topic. Several literature reviews have focused on Wikipedia since its creation. The earliest review was published by Ayers in 2006 . Several others were conducted and published in 2011-2012 , notably the massive WikiLit Review , which covered nearly all publications between 2001 and the start of the Research Newsletter.

As early as 2011, researchers have estimated the number of publications to be too high to effectively review them all. Several discussions have argued for a central database or wiki to collect the relevant literature, but all such efforts have failed or become mostly inactive.

More details about how this effort differs from traditional literature reviews may be found in the Methodology section.

Lists and compilations

There have been several attempts to monitor scientific production on topics relevant to Wikimedia over the years.

Some on-wiki resources like the List of academic studies of Wikipedia on the English Wikipedia, or the Wikimedia-pedia from the 2010–2015 strategy process, have become incomplete and out of date.

WikiPapers is another wiki-based compilations that indexes academic publications related to wikis. Its scope is much wider than the one envisioned in this project. It offers useful tools, like keywords and categories for papers on a similar topic. However, its format is optimized for exhaustiveness; it doesn't really lend itself to synthesis, which is our objective here. It being hosted on an external wiki is also a potential issue.

Research newsletter

The Research newsletter is a successful and active collaborative endeavor to monitor academic production related to Wikimedia projects. It is very useful as a mechanism to stay informed of recent research. However, due to its sequential nature, it doesn't lend itself to offering a synthetic view of the state of knowledge about content contributors. It remains a useful complementary resource.

Purpose

The goal of this initiative is to undertake a review and synthesis of the body of academic research as it relates to Wikimedia communities. This synthesis will be organized by topic and available publicly for researchers and practitioners to find answers to common questions by themselves. It will link to primary sources and other, complementary secondary sources for further information.

For example, reviewing the scientific literature about contributor demographics, editor life cycles, roles and motivations, is expected to help paint a more accurate picture of Content Contributors, guide Editing product decisions and development, and possibly help evaluate the success of WMF initiatives. It is also expected to increase the bus factor for in-house WMF experts, and provide documentation that experts can refer themselves and others to.

The analysis should also address the confidence level in the conclusions described in the literature (for example, good but outdated data) and highlight the gaps in the literature. Prioritizing those gaps would then make it easier to seek collaborations and encourage third-party research efforts.

A particular emphasis is put on systems analysis and modeling, and how well those models work (to what extent, under what conditions), to make the resulting knowledge more actionable.