top of page
Wikipedia masthead.jpg

Wikipedia Talk Page Redesign

1-Week Design Challenge

12-19 Nov 2021

CHALLENGE

How might the Wikipedia Talk page be redesigned to make it easier for editors to participate?

TOOLS

Figma

ROLE

UX Designer

UX Researcher

Background

Although Wikipedia has been extremely successful, with more than 55 million articles and approximately 1.7 billion unique visitors every month, it has been noted that there is inconsistent quality and accuracy in some domains.

 

Wikipedia also suffers from systemic bias, including gender bias resulting from a huge gender gap among editors (only ~8.5-16% of editors are female). Consequently, Wikipedia has been criticised for a lack of representation and editorial bias.

 

Talk Page

Each Wikipedia article has with a dedicated talk page where editors can discuss proposed changes and edits to the corresponding article.
 

Unfortunately, the talk page can be quite difficult to navigate, especially when there are many different discussions addressing various types of concerns, and starting a new discussion can be intimidating for users.

Before:

The Challenge

Redesign the Wikipedia 'Talk' page to either make it easier for editors to participate, or to encourage minority groups to contribute their perspectives.

Deliverables: Rationale for selected prompt, design and thought process, lo-fi wireframes for redesigned flow and hi-fi mockups of 2–3 most representative of selected flow

After:

Solution

In this design challenge, I identified platform and user goals and redesigned the flow from a locked Wikipedia article page to its corresponding 'Talk' page to:

 

  • Make it easier for editors to participate without feeling uncertain and overwhelmed, while making sure they adhere to platform policies
     

  • Streamline editors' contributions to make it easier to moderate and prevent duplicate discussions on an article's Talk page
     

  • Foster a more positive community culture by allowing users to better understand rationales for edits and reversions, and making it easier for the community to recognise one another's contributions

Before/After
Design Process Overview
DISCOVER

Background Research

User Research

Competitive Analysis

Partial Site Audit

DEFINE

User Persona

UX Goals

User Flow

Site Mapping

DESIGN

 

Ideation

Sketching

Wireframe

Prototype

DELIVER

 

Usability Testing

Reiteration

Next Steps

Getting Started

 

Upon receiving the brief, I did some background research on why minority groups aren't contributing to Wikipedia.


Reasons women might not contribute
despite the ability to be gender anonymous: Avoiding conflictual nature of Wikipedia community and unsafe internet environment, lack of confidence in knowledge and expertise, uncomfortable editing other people’s work, and lack of discretionary time. 
 

Those in religious countries have deeply rooted religious beliefs not verifiable by reliable sources that Wikipedia requires
 

Low education, high poverty, and broadband access/affordability in parts of Asia, Middle East, former soviet countries, Native American countries and black countries in the South
 

War (Afghanistan, Iraq, Syria, Yemen) or isolated, repressive regimes (Myanmar, North Korea) 

Choosing a prompt

 

Given a wide variety of reasons why they do not contribute (most of which stem from systemic issues) which requires time to really understand and solve for, so I chose to work on the first prompt within the context of this challenge.

However, I don't see these two challenges as mutually exclusive, but rather, prioritising the first prompt as I think that solving that will serve as a good foundation to tackle the next prompt.

prompt.png
Usability Issues
Why might it be hard for editors to participate in the 'Talk' page?
Usability Issues
⚠️ USABILITY ISSUES
 
Group 124.png
The very existence of a Talk page

on a Wikipedia article isn't prominent enough, which might deter participation.

Group 125.png
Unclear that user can request edits on locked article pages

Users who want to edit locked pages might not do so as it is unclear they can submit an edit request through ‘View Source’.

Understanding Wikipedia

 

I also did some platform research on Wikipedia, to get a sense of the existing challenges they face, how they are dealing with them and what kind of constraints I need to work within.

💡 KEY TAKEAWAY #1
Moderation on Wikipedia is a very human process,

where the admins or confirmed accounts have to manually moderate a lot of the entries — because while they're constantly experimenting with machine learning to combat vandalism and to make the platform friendlier to newbies, human participation is very much needed for complex technical and ethical challenges like interpersonal conflicts and sock puppetry.

Additional findings:

Wikipedia is powered by a dedicated community of super-editors who are reported to work on it on a near-daily basis. Editors closely observe watchlists and are alerted to even small changes, quickly shutting down vandalism and other edits. 

Only administrators may physically delete entire articles, protect articles from improper editing and disruption, and block users (a physical block implemented by the site software).

💡 KEY TAKEAWAY #2
Wikipedia is not a Forum/Social Media platform

Wikipedia has long lists of policies and guidelines which states that discussions on talk pages should be kept on topic of how to improve the article, and not used as a forum. User 'Talk' pages should also not be treated as a social media platform.

Additional findings:

Wikipedia generally "do not employ hard-and-fast rules for policies and guidelines, which should always be applied using reason and common sense.”

Article edits should be verifiable and from a neutral point of view.

Platform Research
User Research

During the discovery phase, I uncovered key user insights from user research studies done by Wikimedia Foundation and NYU regarding why volunteers contribute to Wikipedia and why new editors drop off.

💡 KEY INSIGHT #1
Wikipedia editors are motivated by altruism, a sense of community and recognition (a reputable identity and feeling valued).
  • Editors feels motivated to make a difference through educating humanity and financial motivation has no impact on their contribution.
     

  • Contributors often treat their pages as resumes, showing their edit contributions and community awards.

💡 KEY INSIGHT #2
New editors face unwelcome behaviour from current community and struggle to learn the landscape of communication, policies and guidelines.
  • Demotivated by edits being reverted without constructive or kind feedback.
     

  • Experience uncivil behaviour from experienced editors.
     

  • Unsure the correct form of communication on article talk pages or user talk pages. E.g. How to seek help / ask questions / post an award.

User Research
Testing the Current Site

In addition to these user research findings, I conducted usability testing with 3 frequent users of Wikipedia to validate my hypothesis on some of the usability issues I identified, and gain insights on how a potential new editor will go about using the site.

3 Participants
1 Researcher, 1 Arts Student, 1 3D Printer
2 edited Wikipedia once before

Average System Usability Score (SUS): 40.8/100

Tasks

Measured with Single-Ease Question after each task on a 7-point scale (1=easy, 7 difficult)

SEQ.png
💡 KEY UT INSIGHT
Overall, users often felt uncertain in their actions and felt they needed to learn a lot before understanding the site.

For example, they think that editing with the markup editor isn't super hard, but it's a bit troublesome to find the place where they are supposed to input their comments, and they felt worried that adding their input anywhere on the editor will screw up the whole page.

✅ HYPOTHESIS VALIDATED
3/3 confused by the purpose of talk page

as it looks more like an article than a discussions portal. They found it difficult to keep up with conversations on talk page and reply to them due to lack of scannability and content segregation.
 

3/3 found it difficult to join in discussions 

3/3 users found it difficult to locate where conversations start and end, and where they could input their responses through markup editor.
 

3/3 unaware that they could request edits on locked pages,
and assumed that they couldn’t do it. They also expressed confusion by the labelling ‘view source’.

💡 NEW INSIGHTS UNCOVERED
2/3 unaware that that editing function works as a live-editing format, hence expected to edit by inserting section through a menu selection e.g. Insert > Text
Screen Recording 2021-11-24 at 12.43.47 AM.gif
2/3 expressed confusion by labellings such as ‘edit source’ and ‘talk’. 2/3 said ‘discuss’ would be a better name for its purpose.

2/3 Users unaware of visual editor while editing unlocked article page They unanimously preferred the visual editor but 2/3 subconsciously closed the pop-up that prompted them to switch from markup editor to visual editor.

UT #1
Defining the user

With the insights gathered from the research papers and usability test, I created a persona for a newbie editor, choosing to define my target user as someone who is less technologically adept to make sure that the solution I come up with can work for a wide range of users.

Define
user persona.jpg
UX Vision

 

While the goal is to make Wikipedia easier for new editors to contribute, I had to bear in mind that constraints should be set so their participation does not overwhelm the moderation process.

For example, new editors should be aware of article policies before submitting an edit request, and their participation on article 'Talk' page should be kept objective.

UX Vision
UX vision.png
Competitive Analysis

 

While I was conducting platform and user research, I also looked at some competitors to see what types of features could be implemented to improve wikipedia's usability.

I analysed popular discussions platforms like Quora and Reddit, as well as Google workspace, one of the most commonly used online collaboration tool.

wiki (5).jpg
Ideation

 

While exploring incorporating relevant features into my solution, I found that some of them didn't really gel with the Wikipedia platform.

Ideation
wiki (9).jpg
Site Mapping

 

With plans to make better use of the space on the left side of the webpage for a sticky anchor menu, I recategorised site links to condense them into a top navbar:

  • 'Contents', 'Current Events' and 'Random article' under 'Explore'

  • 'Donate and upload file' under 'contribute. 

  • 'Contributions', 'Preferences' and 'Watchlist', under the user profile, so anything related to the user will be under that section.
     

Site map.png
Design
User Flow

 

While the challenge only called for a Talk page redesign, I chose to include locked article pages in the user flow leading up to the talk page as edit requests directly affects the legibility of information on talk page. E.g. subject titles, and duplicate entries.

It was also a flow where a visual editor was not available, and users did not realise they could request edits.

wiki user flow.png

Scenario: Eddie wants to edit introduction on Quantum mechanics page to specify large particle scales

Round 1 Wireframes

 

As users were looking to insert sections through a menu selection, I tried to create a flow that guides them. I also wanted to set constraints so that they do not get confused by what they can or cannot do like through the markup editor.

Round 1 wireframes.png
Streamlining the Talk page design

The overall goal of the talk page design was to allow the user to easily understand what is happening and participate, while the secondary goal was to make guidelines and policies clear to newbies, and to encourage friendliness amongst community.

Round 1 wireframes-1.png
Wireframes
Testing the new flow

I tested the first lo-fi prototype with 1 user and it didn't work out very well. Compared to the first UT, the average SUS score improved just slightly from 40.8 to 57.5.

Screenshot 2021-12-05 at 11.46.55 AM.png
Tasks

Measured with Single-Ease Question after each task on a 7-point scale (1=easy, 7 difficult)

  1. Make a donation

  2. Jump to section on article page

  3. Edit article page

  4. Process of submitting edit request

  5. Understanding the purpose of Talk Page

  6. Navigate back to main article page from Talk Page

  7. Joining conversations on Talk Page

  8. Collapsing threads

  9. Receive updates to thread

  10. Start new topic on Talk Page

  11. Find an existing conversation topic on page

  12. Award fellow editor

  13. Find Talk Page from article page

THE MAIN PROBLEM LIED IN THE PROPOSED FLOW:
User found the whole process of submitting edit request too long.

He expected to be able to edit immediately, rather than go through a few steps. He was also confused by policies, and felt like he didn’t want to submit if someone else’s similar submission has been rejected before. 

ISSUES ON PROPOSED TALK PAGE DESIGN
User found it difficult to browse an existing topic of discussion.

He found the active discussions sticky menu not prominent enough, and would just scroll down or use CMD+F to find keywords.

He was unaware that collapsible thread works like Reddit,

and expected to collapse the thread by using the button above it.

Confused if 'award' is for a comment or user.

"You mean I can award every single comment?"

User easily found the back to article nav from talk page, but found it too small.

ISSUES WITH NEW SITE IA
He did not immediately know where to find the 'Donate' button,

going to 'contact us', before changing his mind to go to 'Contribute'.

Round 2 Wireframes

 

Rethinking the flow of Edit Requests

I revised the ‘submit edit request’ flow to allow users to make edits immediately when they want to, and alerted them to a similar thread of discussion topic later on. They will also view the article policies right before they submit.

Round 2 wireframes.png
Creating a flow for Talk Page post creation to stay on topic

On the talk page itself, I created a new flow for post-creation, after I identified some types of common edit requests. 

Round 2 wireframes-1.png
Gifting community awards

Because the user expressed confusion by the award function, i created a new flow for gifting community awards to make it clearer.

Initially, I intended for the awards to appear on discussion threads so that the overall vibe of the page can appear friendlier and encourage newcomers to get involved. But I realised that the current messages Wikipedians write for one another on their user talk pages are quite thoughtful and appreciative.

 

I assume that allowing users to award on discussion threads will dilute the thoughtfulness and intentioanlity of writing a heartfelt message to another, and i think it will also derail the discussions on article talk pages.

 

Hence, i created a flow that allows users to gift awards without having to manually paste codes, while still retain the mechanics of posting to a user's talk page.

 

Round 2 wireframes-2.png
Hi-fi Wireframes

 

First step towards developing a design system

While I worked on streamlining the site’s usability by following modern UI conventions, I think that moving forward it will be interesting to explore bringing back of the 90s charm that Wikipedia currently has, through graphics, icons and UI patterns, but nothing that will distract from the main content of the page.

hifi wireframes.png
Talk Page Post-creator

hifi wireframes-1.png
Next Steps
Technological constraints

When i was researching on why wikipedia looks ugly, i found that their website needs to work for everyone, including users in undeveloped countries with low bandwidth, so huge imageries and animation will reduce loading time for them. 

 

While I tried to keep imagery limited and icons flat (no drop shadows) to reduce loading time, what are the other constraints of developing the site for countries with low bandwidth? Would Wikipedia have the capabilities to create a stripped down version like Facebook Lite?

Copywriting

Overall tone can be refined to be friendlier while still remaining objective e.g. making policies and guidelines easier to understand to a newbie


“Talk” page
Even though new users found the name confusing, it is also something that the seasoned editor community is used to. If the objective is that Wikipedia wants more editors to participate in the Talk page before making edits, a popup tooltip can be implemented to explain its purpose to a potential user who has been trying to edit wikipedia pages

Incorporating the needs of super-editor community

Since the are the primary user base of Wikipedia, the redesign should also cater to their needs as well. Research should be conducted to gain insights on the most common tools they use on the sites, and what problems they face.

Talk Page post-creator

 

Identify the most common types of discussion to determine what options users can choose between. Should users be allowed to post freely like they could in a markup editor?

Research further on uncivil behaviours towards newcomers

 

How are experienced editors uncivil towards newcomers? More research needs to be done on their actions to better understand how to tackle this problem.

New features to foster deeper sense of community

 

Follow ‘topic spaces’
Create topic spaces like in Quora for users to explore topics they are interested in on Wikipedia and connect with like-minded individuals

Explore community and personal chatrooms
to allow users to ask for help or share article pages with other users without derailing discussions on talk pages.

Celebrating contributors
Explore features like ‘Featured contributors’. e.g. ‘Wikipedian of the day’ to highlight the contributions of different editors everyday.

Moving Forward
Thoughts & Reflections

 

👀 Check on existing user research papers

While synthesising my user interview findings, I found existing user research papers done by Wikimedia Foundation and NYU. It will be more efficient to check if there are any existing reliable and substantial research papers are available online before conducting your own interviews!

✅ Testing a lo-fi prototype for the first time

As an ex-advertising art director, I am used to jumping straight into executing mid or hi-fi prototypes as I don't like the idea of leaving things to the viewer's imagination. However, in this case, I found that testing the flow using a lo-fi prototype helped save me some time to reiterate quickly.

bottom of page