Former Contributors Survey Results

History

After some discussion at the community health task force, members of the Wikimedia community felt it would be valuable to gain more data about why volunteers leave the project. This led to a proposed survey, which was developed into an actionable questionnaire. Wikimedia Foundation Staff and volunteers reviewed the data from 1200+ responses, and the analysis and conclusions can be found here.

Note that this survey was aimed at less experienced editors. A further survey aimed at more experienced contributors is in the works.

Key findings

  • Around half of editors leave solely for personal reasons
  • Other half left for "community" or "complexity" reasons, in roughly equal parts
    • "Complexity critics" found Wikipedia difficult or confusing
      • Experienced basic editing and specialized tasks as challenging
      • Felt other editors expected them to know too much too soon
      • Wanted interface improvements to make activities easier to use/learn
    • "Community critics" felt that other editors harmed their experience
      • Bad interactions with editors who were stubborn, biased, reckless, etc.
      • Saw contributions reverted or removed over time
      • More frustration if no one explained why
    • Users with 10+ edits were more likely to be "community critics"

Detailed analysis

Preliminary results

Here are the preliminary results from the Former Contributors Survey. This presentation contains a summary of the 1,200+ responses we received to the survey. The material was reviewed with the Wikimedia Foundation Staff during a brown-bag session in February. People who would like to see the details should contact howief. We'll also be working to get this information into a regular wiki page (vs. a pdf file) (Howief 00:30, 19 February 2010 (UTC))

 

Further analysis

Here is another level of analysis for the Former Contributors Survey. The preliminary results (above) give a thorough summary of all the responses, whereas this analysis tries to make comparisons in order to draw further conclusions. A few main techniques were used:

  • Respondents were divided into segments based on why they left. (Reasons of "complexity", "community", or personal reasons only)
  • The "multiple choice" responses were compared between segments
  • The "open comment" responses were compared between segments, by examining the most frequently used words by each segment

The anonymity of the responses has been protected, and requests for the detailed results should go through HowieF.

 

Recommendations to the Wikimedia Foundation

  1. Don’t 
break 
what
 works. 
Don’t 
get 
lured 
into 
trade‐offs.

  2. Improve the interface
    • Economize time for all editors by 
improving
 the 
interface 
for 
tedious 
tasks.

    • WYSIWYG editing is an important start
    • Improve 
interface 
for 
photos,
 referencing,
 community 
processes,
 etc.

    • Add 
tool tips 
and 
contextual 
help 
for 
jargon
 (e.g.:
 “NPOV”)

  3. Steer 
editors 
towards 
good 
experiences

    •  Lead
 editors to 
articles/sections 
where 
their 
work 
will 
matter

    • E.g.:
 match 
articles 
based 
on 
interest, 
experience 
level, 
etc.

  4. Demarcate 
new 
users 
(less than 10 edits) and "mid" users (less than 100 
edits)
    • Reduce accidents where people "bite newbies" in good faith
  5.  Improve communication and monitoring of changes
    • Allow 
users 
to
 monitor 
sections
 of 
articles 
for 
revisions

    • Design a 
quick
 comment 
tool, so that editors can 
message 
peers 
without 
leaving 
the 
edit
 screen

  6. Improve 
dispute 
resolution
 processes 
(and 
help 
newer 
users 
access 
them)

  7. Wikimedia Foundation cannot 
set 
behavioral 
policy... 
but they 
can 
set 
goals 
to
 influence
 the 
community