It’s official – two fingers are quantifiably better than one, at least when it comes to completing online surveys! 

To be clear, I’m talking about online surveys that have their fair share of open text (typing in words) and open numeric (typing in numbers) questions. That is, where you need a keyboard or on-screen keypad to type in your response. We see a lot of these in healthcare professional surveys, and it is from this sector that the data discussed here is drawn. 

Whilst it doesn’t take much expertise to foresee that entering responses using an on-screen keypad will take longer than using a conventional keyboard, you may be surprised by how much longer. At First Line, we passively capture data on how long it takes participants to complete each element of a survey. Over time we’ve collected enough evidence to show not only how long certain types of question take to answer, but also the effect of question variations.

The colourful data table below confirms that single-choice questions are the quickest to complete, and that multi-choice questions are the next fastest. Interestingly, we note that – regardless of device used – multi-choice questions take about twice as long to complete as single-choice questions.

If you’re using a conventional keyboard, the next fastest question type to complete is an open numeric, which on average takes about 16 seconds. But for those using an on-screen keypad, open numeric questions take considerably longer (23 seconds on average on tablets/iPads, rising to 27 seconds on smaller screen smartphones).

A similar thing happens with open text questions: they take an average of 22 seconds to complete using a keyboard and half a minute or more on one-finger screen keypads. The relative difference is about the same: both open numeric and open text questions take on average 7 to 11 seconds longer to complete using an on-screen keypad.

A further indication in this data, backed up by other analysis, is that the bigger the screen the faster you complete. Take single-choice, multi-choice, and grid questions: on a truly mobile compatible survey we’d expect the time difference between moving a handheld mouse, gliding a finger across a laptop touch-pad, and swiping across a touch sensitive screen to be minimal. What we wouldn’t expect, and shouldn’t tolerate, is it taking twice as long to answer single-choice and multi-choice questions on a smartphone compared to on a desktop computer (for grid questions, which are by nature more variable, it takes over a third as long).

In conclusion, there is an obvious time-loss associated with asking open text and open numeric questions on touch-screen devices that we can’t do a great deal about – mobile technology just works that way. What we can do something about is the unacceptable time difference between completing other types of questions on different devices. The answer is obvious enough: author materials that we, as consumers of research, would personally be happy to complete using our mobiles, and then program them in a way that is truly technically compatible.

Related blog posts and links:
Demo: Mobile compatible grid questions: https://flres.uk/mobgrid
Blog: We need better online surveys, not shorter ones (11th September 2018): https://flres.uk/bettersurveys
Blog: There are only two types of online survey (13th August 2018): https://flres.uk/surveyscreens