User research shows issues with <select>

At this time there’s only a video: but I’m assuming/hoping that there’s a transcript coming soon somewhere.

In the video, a front-ender for explains how user research showed multiple problems with the plain old selects… since this set of websites must specifically reach people with poor computer skils, older browsers and special needs, not savvy developers with the latest blink nightly.

And while I get the feeling that a lot of developers expect that savvy blink-nightly users are at least a large part of their userbase, if not the majority, I wonder if more research would set this straight?

Anyway the main issues these users had with select dropdowns:

  • people were unable to close the select
  • people tried to type into them
  • people didn’t realise that focussed selects are not the same as selected selects
  • people trying to pinch and zoom (on phones and tablets) were unable to enlarge the dropdown’s lists, meaning for a portion of the userbase, these options were unreadable (remember a lot of people who really could benefit from using a screen reader or a screen magnifier often don’t know these things exist and rely on things like setting the screen resolution very low and sitting centimeters away from the screen).

The speaker shows examples of possible substitutions, and shows an example where some kind of dropdown still makes the best UX sense. However for example one possibility she mentions is a slider… watch the comments at the end as well, where she answers a question about using sliders. They may not be appropriate.

Also, progressive enhancement is mentioned, and how internally requires that their stuff works without Javascript and down to IE7. Why? Because lots of users (regular, non-developer people) have outdated browsers and this should not preclude them from filing taxes or registering to vote. She mentions how many users only have access to a computer via their work (where they don’t get to choose the software) or only via the public library (ditto).
She also talks about a particular government office branch with security clearance levels, where each morning users download Firefox if they want to use it, because each night, everything installed by users is wiped by the system. Woo!

The without Javascript was two-fold: first of all, Gov.Uk isn’t a game site, it’s mostly a site with textual and graphical information, and lots of forms. All of which work by default without Javascript. Users may be mobile-- if they’re accessing in the train and the big bloat of JS didn’t load, why should that stop them from filling out a form? It shouldn’t, forms work by default without JS, naturally. Second, she mentioned when a new code push broke all the Javascript on the sites-- only for about 15 minutes, but this isn’t an outlandish thing at all, we see it on sites quite often (usually only for brief periods before someone does a roll-back).

Anyway, interesting talk, interesting data. If I find a transcript I will post it.

1 Like

Good find! I use SELECT tags quite a lot, as many devs probably do. That will make me think again and look at Radio buttons a bit more :slight_smile: