Page MenuHomePhabricator

Understanding newcomer onboarding experiences - mentorship, relatedness, and competence
Open, Needs TriagePublic

Description

Only about 3% of newcomers who have been assigned a mentor via the Newcomer Homepage ask their mentor a question. Among those who have asked a question, about a third ask a question as their first edit, and preliminary research has found no significant effect of having a mentor on early activation and retention of newcomers. However, we don’t know why newcomers largely don’t ask questions – is it that they’re finding the information on their own? Is it that they have other spaces where they feel more comfortable asking questions (including external search or asking ChatGPT?)? Are they aware they have a mentor?

To better understand the experiences of newcomers on Wikipedia, we should survey newly active editors (those who have made over 50 edits and registered their accounts or made their first edit in the past 6 months) about their onboarding experiences. This survey can help us to broadly understand successful newcomers’ knowledge of different components of editing (e.g. mentor access, talk pages, village pumps / tea houses, WikiProjects, watchlists, userpages), which editing experiences they have found rewarding, and those which they have found to be demotivating to their editing. This work can help to inform proposed interventions such as a structured progression system for editors, as well as helping us to understand the broad experiences of newcomers and where they most feel (or don’t feel) a sense of autonomy, competence, and relatedness in their onboarding journey.

After the survey, we can likewise conduct further research with editors who opt in to be contacted about their experiences to better understand mechanisms and sentiments which the survey data cannot address.

Steps

  • Develop a survey specific to newcomers which asks questions related to onboarding, discovery, knowledge of tools, and sentiments about different components in collaboration with stakeholder teams which work with newcomer experience
  • Determine which projects to survey newcomers on (for simplicity, this project could start on English Wikipedia and iterate on the questions asked by adding new projects)
  • Determine the population size and coverage needed for deploying QuickSurvey sampling and link-out to a longer survey
  • Set up QuickSurvey messages
  • Deploy and un-deploy the survey
  • Pull eventlogging data for weighting and metadata
  • Clean survey data
  • Conduct analyses
  • Share findings with hypothesis stakeholders
  • Report findings on metawiki

Background

In Q4, we broadly explored contributor motivations (T391499) and identified three core theoretical underpinnings: competence, autonomy, and relatedness. Of these, relatedness was identified as the most under-explored facet of editor motivation. Based on the 2024 Community Insights data, we know that editors who take on advanced roles like adminship and organizing are more likely to find the social aspects of editing to be a motivator for contributing, but we don’t know whether editors in general are not interested in social interactions or they just haven’t had opportunities to positively engage other editors. Mentorship is the most structured means by which relatedness is currently supported on the Wikipedia projects; while mentorship happens incidentally through spaces like the WP:Teahouse, the Newcomer Homepage provides a mentorship module that directly connects newcomers with mentors.

Details

Due Date
Oct 31 2025, 12:00 AM

Event Timeline

Weekly updates:

  • Scoped population viability for editors who registered in the last 6 months, have made 50+ edits (and made an edit in the last month, to approximate likelihood of being sampled on-wiki) with major thanks to @CMyrick-WMF and @Hghani for helping with the query.
    • Finding: Enwiki has a viable population of approximately 2000 qualified editors in any given month; other Wikipedias may have too small of populations and other research approaches may need to be considered if we want to expand to other languages (for example, the estimate for eswiki is around 300 qualified editors).
  • Scheduled meetings with potential stakeholders and staff doing related research for next week.
  • Began brainstorming potential survey questions (will need input here from relevant teams).

Weekly updates:

  • Met with stakeholders to discuss the survey and where it can support related projects
  • Shared a survey question brainstorming document for teams to add potential questions; stakeholder teams will input to these with a deadline of July 24th.
  • Drafted a research brief (internal) to be updated after survey question brainstorming
  • Note: I will be out of office for the next two weeks, so there will be no updates during this time :)

Weekly updates:

  • A draft of the survey is available for feedback and testing - I hope to have this finalized next week after everyone has had a chance to review. Thanks to all who have taken it and given feedback already!
  • An initial codebook has been made for the survey, with strategic mapping of the questions to KRs; this will later become the survey dataset codebook.
  • Next steps:
    • Begin QuickSurvey configuration and message setup
    • Make a metawiki Research: page for the survey
      • Make a resources page for respondents, showing them where to find the tools, pages, and spaces we asked them about
    • Reach out to movement comms for communications guidance
    • Get a survey privacy statement from Legal

Weekly updates:

  • Research page and questionnaire published, resources page drafted (but may need a bit more work to make it digestible and engaging)
  • Privacy statement request submitted and in progress
  • Finalized KR language
  • Next steps:
    • QuickSurvey configuration and message component setup
    • Reach out to movement comms
    • Final tweaks to LimeSurvey before activation (add QS variable, finalize and check skip patterns, add privacy statement link when it's ready, and end-of-survey page pointing to resources)
Miriam set Due Date to Oct 31 2025, 12:00 AM.Aug 26 2025, 1:12 PM

Weekly updates:

  • The survey was announced on the English Wikipedia Village Pump on the 3rd and has had no comments so far (will check again on Monday)
  • QA on the survey flow (on-wiki QuickSurvey, link to survey, parse QuickSurvey token to LimeSurvey) appears to be working well. I will pull eventlogging data for the tests on Monday to make sure everything looks okay on that side.
  • We're scheduled to deploy (that is, increase QuickSurvey coverage) and start collecting data on Tuesday the 9th!

Weekly updates:

  • The survey launched on Tuesday the 9th and is collecting data!
  • An issue with ad blockers affecting event logs is causing a proportion of responses to not have a corresponding quicksurvey event (which we use for weights); @YLiou_WMF and I anticipated this from the QA process and are brainstorming about how to account for this.

Weekly updates:

  • We're ready to close the survey on Monday :)
  • I've started working on the R code for cleaning and question recodes, and will update it with the full dataset once we're done with response collection.

Weekly updates:

  • The survey was undeployed and closed on Monday the 22nd
  • I'm about 95% through the survey data cleaning and recoding, and will consult with @YLiou_WMF next week for weighting strategy.
  • Stakeholders are encouraged to reach out if there are any survey questions that are high priority in the next 1-2 weeks, as I can prioritize those analyses if needed.

Update:

  • Survey findings were presented to stakeholders on November 13th, and the hypothesis was officially closed on November 21st (hypothesis link - internal).
  • The next aim is to publish the results on metawiki and share on the enwiki village pump so that the community can benefit from this research as well. Once this is done, this task can be closed.