Global Data and Insights/Reports/2021 Affiliates data survey report/Wiki-contibutions

Affiliate Data Survey Report 2021

• Summary
• Glossary
• Wiki-contributions
• Affiliate Institutions
• Affiliate Spaces
• Appendix
• Methodology

A community of organized editors

When a community member joins a Wikimedia affiliate of their choice, they express an intention of offering their participation, or direct support to the organization, often through volunteered time and skills to advance shared projects in support of the Wikimedia mission. The perennial question in the movement is whether these additional responsibilities impact an affiliate member's editing activity level. This question requires controlled trials to answer, which we are not currently in position to offer, so the next best question is whether affiliated editors have higher/lower levels of activity compared to non-affiliated editors at the moment.

To answer this question, a total of 3,742 affiliated editors’ usernames were harvested from affiliates membership pages on meta wiki, and 18,118 unaffiliated editors usernames were obtained from the Community Insights survey editor sample list. Wiki contributions were pulled for each username over the period of 12 months (from January to December 2020).

A population frame was created to organize affiliated editors into sampling units using affiliate type, geographic region, edit count bin and wikimedia project as factors. Sampling units with at least 30 editors were selected to create the final sampling frame, from which a random sample would be drawn. A minimum of 30 edits was selected, and 4 edit bins were created to represent activity levels of editors. The final list of affiliated editors was filtered down to 902 (24% of the total harvested affiliated usernames) unique editors. A final list of non-affiliated editors was drawn by randomly sampling usernames based on the final sampling frame of the affiliated editors, resulting in a total of 1,128 usernames.

The following findings were made, when comparing the two samples

• The overall editor to active editor ratio for affiliated editors on WIkipedia projects is 3.7 to 1 with 21% of editors active, 5% very active, while non-affiliated editors have a higher ratio of 6 to 1 and have a lower percentage of editors active (14%), and very active (2%) in comparison.

• Affiliated editors outperformed unaffiliated editors when looking at combined contributions across the 113 selected Wikipedia language projects, with 25% more edits (average of nearly 300 edits more) and 81% more pages created (average of 35 more pages).

• However, Affiliated editors wrote 17% shorter pages on average across selected Wikipedia language projects.

• Unaffiliated editors also out-performed affiliated editors on Wikidata, looking at the average number of edits, pages created and the length of pages in words.

• Affiliated editors also performed better than unaffiliated editors within specific edit bins as shown on the info -graphic to the right.

In time, we hope to compare pre- and post- affiliation editor activity to help us better understand the impact on editing of affiliation, but we can confidently assert with the data we have is that the affiliated editor population had a higher percentage of active (26% to 14%) and very active members (6% to 2%) compared to that of unaffiliated editor population across all Wikimedia project comparisons. See Appendix for specific project comparisons.