Research:Ideas/How has the retention of female editors been affected by the decline?

From Meta, a Wikimedia project coordination wiki
Jump to: navigation, search


The English Wikipedia Decline.png
Created
November 9, 2013
Contact
EpochFail
Light-Bulb icon by Till Teenck.svg

This page documents a proposed research project.
Information may be incomplete and may change before the project starts.

El Seano (@evilseanbot) asked on twitter[1] asked the following question:

@halfak Heya, I was curious if the skewed male demographic of wiki editors has increased with the decline of users or not.

The number of monthly active editors is plotted over time.
The English Wikipedia Decline. The number of monthly active editors is plotted over time.

I think it's a good question and one that we don't have a good answer for. In order to answer this question we have to identify the gender of newcomers to Wikipedia who only stayed a short while. This project will involve a review of previous analyses of the "Gender Bias" in Wikipedia[1] and the strategies they used to identify gender and correct for survey bias[2].

Registration-->Edit funnel micro-survey[edit]

The proportion of female readers on Wikipedia does not suffer from the same gender gap that the the population of editors do[citation needed]. So where are we losing them? Are female readers less likely to register an account? Are they less likely to try editing once registered?

The path from reader to editor can be summarized as follows:

 [unregistered reader] --> [new registration] --> [first edit link click] --> [first edit saved]

It would be nice to know where in this funnel female editors are being lost. To do this, we could extend MediaWiki to present a popover micro-survey to a random sample of newcomers as they reach each state of the funnel (unregsistered reader, new registration, etc.) From these sampled micro-surveys, we can generate proportions. By comparing the surviving proportion of editors between states, we can find out where female editors are mostly being lost -- or if they are even clicking "edit" in the first place.

Micro-survey formats[edit]

Multiple choice (3)[edit]

Thanks for <last_action>. We (researchers at the Wikimedia Foundation) would like to ask you a few demographic questions. See our privacy policy and consent form.

What is your gender?

MaleFemaleOtherPrefer not to say

Free-form textbox[edit]

Thanks for <last_action>. We (researchers at the Wikimedia Foundation) would like to ask you a few demographic questions. See our privacy policy and consent form.

What is your gender?

<type here>

SubmitSkip

The multiple choice (3) micro-survey has the advantage of producing clean data. In the free-form text micro-survey we exchange clean data for allowing the user to choose how they would like to refer to themselves. See talk discussion: Micro-survey style

Support needed[edit]

  • Someone with experience parsing userboxes that suggest gender to help gather data
  • People to help with hand-coding to check the accuracy of our algorithms for classifying editor gender against content on Wikipedia
  • Someone with dev experience for MediaWiki

Ready to create a project page?


References[edit]

  1. Shyong (Tony) K. Lam, Anuradha Uduwage, Zhenhua Dong, Shilad Sen, David R. Musicant, Loren Terveen, and John Riedl. 2011. WP:clubhouse?: an exploration of Wikipedia's gender imbalance. In Proceedings of the 7th International Symposium on Wikis and Open Collaboration (WikiSym '11). ACM, New York, NY, USA, 1-10. DOI=10.1145/2038558.2038560
  2. Hill, B. M., & Shaw, A. (2013). The Wikipedia Gender Gap Revisited: Characterizing Survey Response Bias with Propensity Score Estimation. PloS one, 8(6), e65782.