Mission: Figure out the usability issues of Tor Browser and document them.
Location: UC Berkeley campus.
Dates: 30 Jan, 31 Jan, 1 Feb (Friday, Saturday, Sunday)
- 30 Jan: Meet in the evening to prepare for user study.
- 31 Jan: Do user studies all day.
- 1 Feb: Start writing a summary of results.
The idea is to hold this ux sprint before org/meetings/2015WinterDevMeeting, so the output can be discussed there.
To-Do:
Reserve meeting roomsPrepare technical infrastructure for user tests (workstations, screen sharing, recording)Recruit user participantsConfirm attendees (#Who)Send announcement to ML
Who: #Who
Results
- blog post: UX Sprint 2015 wrapup
- subtitled screen videos: https://people.torproject.org/~dcf/uxsprint2015/
- HotPETs 2015 paper and video.
Reading
Eliminating Stop-Points in the Installation and Use of Anonymity Systems: a Usability Evaluation of the Tor Browser Bundle\ Greg Norcie, Kelly Caine and Jean Camp. HotPETS 2012.
Why Johnny Can’t Blow the Whistle: Identifying and Reducing Usability Issues in Anonymity Systems\ Greg Norcie, Jim Blythe, Kelly Caine, L Jean Camp. USEC 2014.
Tor User Experience Report: Recommendations on enhancing the user experience of the Tor Browser Bundle\ Jeff Mau, George Rosamond. ISC Project 2013.
Discovering What Your Users Know\ Arne Renkema-Padmos. Includes sample consent form, questions, and researcher script.
A Needfinding Framework for Internet Freedom | Direct link to PDF\ SecondMuse
Understanding Internet Freedom: Vietnam's Digital Activists | Direct link to PDF\ SecondMuse
Crypto Tales from the Trenches (panel discussion of journalists on crypto tools)\ Nadia Heninger, Julia Angwin, Laura Poitras and Jack Gillum. 31c3.
Notes and ideas
The first thing to do is to identify goals.
- Find places where the current interface is hindering users?
- Test proposed new user flows against current ones?
- Give users an opportunity for direct communication with developers?
- Give developers an opportunity to see how the software is used?
Then, identify which users will participate in testing.
- English speakers?
- Users who have not used Tor Browser before? Good if the goal is to find overall user experience flaws.
- Users who already use Tor Browser? Good if the goal is to test new interface ideas.
- Ideas for recruiting users:
- Advertise at a journalism school.
- Advertise at a university.
- Contact activist groups.
- Ask someone we know where the meeting is (a "stringer") to assemble some of their contacts.
- Ask representatives of Open Tech Fund and Radio Free Asia for resources in the meeting city.
- The translators on the Transifex mailing list are also users and many are interested in helping.
We should have specific goals and things we want to work on before assembling people to a meeting. Doing user testing takes more time than you think.
We have a great resource: monthly help desk reports. We can take the most common questions (or the most common answer templates) and treat them as the most pressing UX issues. We should invite members of the help desk to take part in the meeting.
Techniques:
- "Cognitive walkthrough": Have a user explain verbally what they are thinking as they use the software. Demonstrate the idea for them on an unrelated application. "Now there are two buttons, I don't really know what they mean but the one with the arrow looks like the one I want, now it's asking for a username, it's probably the one I set up earlier..."
- Make a video of the screen and an audio recording.
- Present users with paper or wireframe designs. Good for short tests, only takes a few minutes, doesn't require a live computer running.
It's important for those taking part in the meeting to feel comfortable. It is intimidating to have a developer, an expert, present when you are using an unfamiliar interface. Nobody likes to feel stupid. It may be better to have someone not intimately involved with the software help guide the interaction. Non-developers should outnumber developers. Developers can be at arms' length; e.g., watch screencasts after the fact, in order to avoid influencing users during the tests. Even the "dev" in "Minidev" could be offputting if that's how we recruit users.
Simply Secure and SecondMuse offered to look over templates.
Video on adding transcripts to videos: Automatically Subtitling the C3: How speech processing helps the CCC subtitle project, and vice-versa
Who
Ideally, we'd like to have at least one active person from each of the help-desk, Pearl Crescent and TorBrowser teams, plus whoever is interested in usability issues. If you'd like to join us, please add your name in the following table:
| Person | Attending | Lodging | Arriving | Departing | Team |
|---|---|---|---|---|---|
| mrphs | Y | -- | 29 Jan | 2 Feb | GetTor |
| saint | Y | -- | -- | -- | Satori |
| David F. | Y | Y | local | local | Freelance |
| Linda L. | Y | -- | -- | -- | Berkeley |
| Kathy B. | N | -- | -- | -- | TB (Pearl Crescent) |
| Colin C. | M | -- | -- | -- | Support / Translation |
| Arthur E. | Y | -- | -- | -- | TB |
| Arlo | M | -- | -- | -- | TM |
| SpencerOne | N | -- | -- | -- | Freedom |
| Kate K | M | -- | -- | -- |
Tickets
Tickets created as a result of the sprint (/tags/uxsprint2015 tag): [[TicketQuery(keywords=uxsprint2015,format=table,cold=id|summary|keywords|owner|component)]]
Tickets with the "tbb-helpdesk-frequent" tag: [[TicketQuery(status=!closed&keywords=tbb-helpdesk-frequent,format=table,col=id|summary|keywords|owner|component)]]
