Skip to content

Usability testing of Figma prototype on multi-path circuits

Context

Multi-path circuits (aka. Conflux) is a feature that enhances reliability and performance by splitting traffic across multiple circuits at the same time and prioritizing circuits with lower latency.

Right now:

  • On desktop, Tor Browser only displays single-path circuits.
  • On Android, Tor Browser has no circuit display at all.

Design

Prototype

Timeline

At best, this feature will be implemented in Tor Browser 16.0, not before June-July 2026. We're aiming at handing over finished designs by the end of 2025.

  • September
    • Week of September 22
    • Week of September 29
      • Prepare screening
  • October
    • Week of October 6:
      • Recruit
      • Finish basic prototype (@felicia)
    • Week of October 13:
      • ⚠️ YEC Launch
      • Integrate prototype with contextual elements (together with @felicia)
      • Recruit
    • Week of October 20:
      • Schedule participants
      • Run pilot test
    • Week of October 27:
      • Run sessions
  • November
    • Week of November 3:
      • Analyze
    • Week of November 10:
      • Analyze
    • Week of November 17:
      • ⚠️ Tails Sprint
      • Report
    • Week of November 24:
      • Report
  • December
    • Week of December 1:
    • Week of December 8:
    • Week of December 15:
      • Hand over to the Applications Team

Research goals and questions

  • Test the basic version (no bridge + no onion service) of the new circuit display for single-path and multi-path circuits on both desktop and Android, before we hand it over for development.

    • Discoverability

      • D1: Do users know where to find the circuit display?
        • Even more important for Android users since there is no circuit display right now on Android.
        • Use of the circuit icon instead of the shield icon with UTP.
      • D2: Is it intuitive for users to switch between the view of each path?

    • Understanding

      • U1: Do users understand that websites using multi-path circuits are using the different paths at the same time?
      • U2: Do users understand that the same exit relay is used in multi-path circuits?
      • U3: Do users wonder about the implications of Tor Browser using a single-path vs. a multi-path circuit?
        • U3.1: Do users wonder about the benefits or disadvantages of both?
        • U3.2: Do users wonder about when Tor Browser chooses one or the other?
        • U3.3: Do users wonder about how to change between one and the other?
      • U4: Do users understand the relay terminology? (#148)
        • U4.1: Otherwise, how might we improve it?
      • U5: What questions do people have when opening the "Learn more" link? (not written yet)

  • Test the new circuit display within Firefox's new Unified Trusted Panel (UTC).

    • P1: Do the other elements of the UTP interfere with the circuit display?

Out of scope

  • Test variants of the new circuit display

    • Onion sites
      • Multi-path won't work yet with onion sites.
    • Bridges

  • Accessibility

    • Is the new circuit display accessible for screen reader users?

Script

Prep

  • Reset Figma prototype

Intro

  • Hey, nice to meet you. My name is sajolida and I'm a user researcher at Tor.

    How are you doing today?

  • Thanks for being willing to share your time.

  • Today, we'll ask your help to test a new feature of Tor Browser. The feature is not implemented yet so we will ask you to comment on the design of the feature that I will share with you from my screen.

    What we're testing today is the design. We're not testing you. So every comment and every question you might have is very interesting to us, especially if you think that some things are hard to understand.

  • Before we move on, I wanted to take 2 minutes to talk about our data handling practices and let you ask any questions you might have. It's a summary of what I shared in the research information sheet that I sent you by email.

    • Today, and only with your consent, I'll be using a screen recording application to record my screen and the audio of our conversation.

    • The recording would be stored on a platform called Nextcloud. Nextcloud is an open-source file hosting tool, sort of like Google Drive.

    • What we store in Nextcloud lives on servers run by a trusted partner organization called Riseup. The recording would only be accessible to Riseup administrators, Tor Project employees, and authorized contractors.

    • The recording will be deleted no later than 180 days after the session.

    • From my analysis of the recordings, I will create an anonymized report that will be shared with the rest of the Tor Project and published on our website.

    • No personal information that could identify you as a research participant is ever published in any report from the research. For example, we will never publish your name, where you live, the name of your organization, or anything that alone or in combination with other information could be used to identify you.

  • How does this sound? Any questions so far?

  • With that said, would you be comfortable with me recording this session?

  • As we discussed over email, you'll give you 50 USD right after the end of the session to thank you for you time.

  • Alright, so I'll start the screen recording.

Profile

  • I'd love to learn a little bit more about you in your own words. What do you do?

  • How does digital privacy relate to your work or activism, if at all?

  • What might be the implications of a breach in your digital privacy?

  • What are some things you do today to help mitigate risks to your privacy?

  • Tell me a bit about your experience using Tor.

Prototype

Now I'm going to show you a prototype of Tor Browser and ask how you would use it.

You cannot click on the prototype yourself, but I can do it for you. If you want to click on something just tell me where.

Let's imagine that you're at home. You open Tor Browser and visit a news article, for example, on The Guardian, an English news site. You see the following page.

Task 1: Discovery

  • How would you tell in which country this website thinks that you are? (D1)
    • How would you find information about how Tor Browser connects to this website?

Task 2: Change circuit

  • How would you change country? (D1)

    • Step 2: How would you change the relays that Tor is using.

  • In which country does this website think that you are now? (D1 + U2)

    • What makes you say that?

Task 3: Multi-path

  • How do you understand the different elements on this new screen?

  • Could you give me a list of all the countries through which your connection goes? (D2 + U1 + U2)

  • What was easy or difficult to understand in the previous screen, and why?

  • If you could change anything about this screen, what would it be?

Understanding

Task 4: Differences

  • How would you change country again?

  • What is different between the circuit that is being used now and the previous one? (U3)

    • How do you think these differences impact your connection to this website, if at all? (U3.1)
    • When do you think Tor use single-path or multi-path (using the participant's vocabulary)? (U3.2)
    • How would you change between single-path and multi-path (using the participant's vocabulary)? (U3.3)
      • After learning that it's impossible to choose: What do you think about that?

Task 5: Documentation

  • What kind of information do you expect to find if you click on the "Learn more" link? (U5)

Alternates

Task 6: Exploration

  • Show "alternate-1-open.png"

  • Do you have any idea on how to change this screen to make it easier to understand multi-path?

  • Show "alternate-2-list.png"

  • What differences do you notice between these 2 designs?

  • Which design do you think explains multi-path better?

    • What makes you say that?

  • Show "alternate-2-pill.png"

  • What differences do you notice between these 2 designs?

  • Which design do you think is clearer?

    • What makes you say that?

Terminology

Task 7: Relay terminology

  • How would you describe what happens to your connection to a website when you use Tor?

  • What does the word "relay" mean to you?

  • What are the different relays involved in the connection to a website with Tor?

  • What was easy or difficult to understand in the names of the different relays?

Debriefing

Next

  • Send reminder to the next participant

Participant recruitment

Duration

Limit ourselves to 1 hour: not a lot of different things to test + remote + questions more than tasks.

Number of participants

We won't test both desktop and Android with the same participant (between-subjects study), because the interfaces are too similar.

  • 4 participants for Android
    • Ideally people who use Tor Browser on Android already
  • 4 participants for desktop
    • Ideally people who use Tor Browser on desktop already

Screening criteria

  • Occasional and frequent users
  • Location
  • Language
    • Prototype in English only
  • Incentives: #182

Methodology

Figma prototype

We'll show the prototype from our screen:

  • Improve privacy for the participant. Everything happens on Signal and they don't have to share their IP with Figma.
  • Force people to explain better their thoughts.
  • Fix the reliability issue with Figma.

Signal

  • "Create call link"

Deliverables

Checklist

  • Design research
    • Meet with @felicia to agree on research goals and research questions
    • Agree on research timeline
    • Draft research script
    • Review research script (@alicia)
  • Recruit
    • Decide on session duration
    • Decide on number of participants
    • Discuss screening criteria
    • Discuss incentives
    • Schedule slots
    • Create screening survey using template screener
    • Review screening survey (@alicia)
    • Test screening survey
    • Send link to screening survey
    • Schedule participants using screening spreadsheet
    • Answer to all candidates using email templates
    • Get in touch with participants over Signal before the call
    • Send research information sheet to participant when confirming location
  • Prepare prototype
    • Prepare interactive mockups on Figma
    • Test interactive mockups
  • Prepare tests
    • Adapt research information sheet
    • Prepare recording setup
      • Test the recording setup
      • Check available disk space
      • Test screensharing with Signal
    • Prepare incentives
  • Analyze
    • Create rainbow table
      • Mark new issues
      • Document benefit
      • Strike through solved issues
      • Fill in demographics tab of rainbow table
      • Link rainbow table from here: XXX
    • Create public report and link from here
    • Prepare video clips
    • Document on GitLab
      • Create missing issues
      • Document in existing issues as "### User research findings"
      • Link all issues to this one
      • Link related issues together
      • Verify labels on all issues
    • Share highlights on tails-dev and link from here: XXX
    • Prepare video clips
  • Wrap up
Edited by sajolida