Hi - does anyone make any experiences with designing a query for mobile screens recently ?

Are there any question types to seriously avoid ? Is there a way to still keep the arrays, or some of the arrays ?

What about the LS templates, like Janus ? To what degree do they improve things ?

The core themes in the 3.x version generally work well on smaller mobile devices.

The array questions revert to a block display on smaller screens so are very accessible.

I would avoid the drag-drop ranking question if using lengthy answers. There may be some difficulty dragging the large items on smaller screens.

To add my experience:
Avoid sliders. Also difficult to handle.

But one easy way is:
There is a survey in the manual that contains all available question types.
Here you may try the look and feel on small devices.
www.limesurvey.org/manual/Expression_Man...Types_and_Attributes

Joffm

I would completely avoid matrix/array questions. Technically it is true that they can be shown both on large screen as well as small screens, but the presentation differs and therefore the way people look at the question is different and so there may be different answers from people seeing different presentations of the same question.

I have used sliders on mobile screens, this worked.

I would avoid ranking questions, if you really need them, test thoroughly.

And: always test on all screens that people will use.

Tammo

And: always test on all screens that people will use.

Which today is basically impossible. It was already difficult before (different screen sizes, thus different resolutions), but it was somehow controllable. Now with mobile phones in so many different resolutions, it is almost impossible to make thorough tests.

gs.statcounter.com/screen-resolution-stats

Statcounter offers 14 different resolutions in the graph, yet "other" ist by far the highest share.

Creating applications that work well on all screens probably has never been more difficult than today.[/b]

Right. The way to do minimal testing is to test on laptop screen, iPad (this has several resolutions, try the smallest) and small android screens.

But be prepared to see strange things happening on the large portfolio of possible screen sizes/resolutions.

Developing surveys is a craft.

holch wrote:

And: always test on all screens that people will use.

Which today is basically impossible.

Tammo's recommendation is still correct. SaaS user of other survey packages are used to have a systematic check of different browser done by the SaaS provider. There seems to be the hope that the "community"cover that.
The bugtracker show "issues" with certain browsers/devices, which is not caused by the survey design.
Mostly caused by other bugfixes.

The awareness of these issues is quite low, cause not many respondents will report these issues. They just will close the survey. To have browser meta infos in the results can help to find issues in combination with non-finished responses.

Totally agree that Tammo's recommendation makes sense. I just wanted to add, that it has become really difficult.