Submissions/What the %$*! do we know? How (and why) to run an experiment on Wikipedia.

From Wikimania 2013 • Hong Kong
Jump to navigation Jump to search
Symbol wait.svg

This submission is on the wait list for Wikimania 2013.

Submission no.
5076
Subject no.
T5
Title of the submission
What the %$*! do we know? How (and why) to run an experiment on Wikipedia.
Type of submission
presentation
Author of the submission
Steven Walling, Matthew Flaschen and co. from Editor Engagement Experiments
Country of origin
USA
Affiliation
Wikimedia Foundation
E-mail address
swalling@wikimedia.org
Username
Steven (WMF)
Abstract
What do we really know about how Wikimedia communities work? Every day Wikimedia Foundation engineering teams make countless decisions that could fundamentally alter the fabric of our communities through software. The Editor Engagement Experiments team at the Foundation is based on a bold idea for how to do this: we should be making decisions based on data, not just consensus. This presentation will give you an inside look at how this process of experimentation and data analysis happens at the Foundation.
Detailed proposal

The Editor Engagement Experiments team at the Wikimedia Foundation is one part of the group working to save our projects, and especially Wikipedia, from a long slow death. Without new community members, even large and relatively stable communities like Wikipedia will eventually wither, and they certainly won't grow to encompass the sum of all human knowledge. But unlike previous efforts to address this problem, our team of five-plus staff takes a unique approach.

While the wisdom embedded in our communities is essential, controlled experimentation and rigorous data analysis can provide an objective look at what's really going on, good or bad. Since the team was formed in March 2012, we've run more than a dozen experiments for Wikipedia editors – everything from a new signup page redesign, to emails asking lapsed contributors to return to the wiki.

This presentation will give Wikimedians a progress report on what Editor Engagement Experiments have taught us over the last year about how to test new ideas for improving our communities. We'll talk about the features we've tested, the methods we've used, as well as how to run an experiment that collects data accurately and responsibly.

Track
Technology and Infrastructure
Length of presentation/talk
25 minutes including Q&A
Language of presentation/talk
English
Will you attend Wikimania if your submission is not accepted?
Yes


Interested attendees

If you are interested in attending this session, please sign with your username below. This will help reviewers to decide which sessions are of high interest. Sign with four tildes. (~~~~).

  1. Daniel Mietchen (talk) 09:55, 30 April 2013 (UTC)
  2. Fabrice Florin (WMF) (talk) 15:45, 30 April 2013 (UTC)
  3. DarTar (talk) 18:32, 30 April 2013 (UTC)
  4. Ori.livneh (talk) 18:47, 30 April 2013 (UTC)
  5. Madcoverboy (talk) 23:51, 30 April 2013 (UTC)
  6. Man77 (talk) 20:18, 2 May 2013 (UTC)
  7. Add your username here.