Our Blog


What we're talking about


Why appearance matters when it comes to survey design

by Pat Prunskus.

Let’s face it: the appearance of most online surveys leaves a lot to be desired. Here we are in the midst of an era when infographics dominate presentations and our collective eyes are glued to some sort of screen, and yet panellists are answering questions that look like liner notes from Air Supply albums. We should be ashamed of how our online surveys look. So who’s to blame?

Culprit #1 – Survey Programming

I once had someone say to me with justifiable exasperation, “I don’t care if gender is a picture of a phallus. Just make the surveys more engaging.” And they’re right: why shouldn’t something like this (see the image below) be standard?

What you may not know is that there’s a better suite of ready-to-use tools available to you. Most programming platforms are at a state where they can automatically generate better looking templates.  Don’t let anyone tell you that losing respondents via a tech drop is a justifiable excuse for not making something look good.  It’s the responsibility of those programming the survey to make it work. If there’s a shift in platform functionality, programmers need to move with it.  By the same virtue, it’s on the shoulders of those programming surveys, to make engaging surveys automatic for standard demographic questions.

Culprit #2 – Timelines

Your client wants to go into field as soon as possible, right? Nine times out of ten, you’ll cut any corner to get into field quicker, won’t you? It’s okay: no one can blame you for trying to keep everyone happy. After a gruelling week of persuading your client to sign off on the final version of the survey you find yourself in those anxious moments of relaying a favourable timeline to your client. The visual appeal of your survey dies a quick but honourable death on that very battlefield. Now here’s a fact most project managers will tell you that very few ever consider: the majority of studies finish field the day after a weekend. Apply that backwards and think of what you could have done with an extra day or two with programming.

Culprit #3 – The Language Gap

Our industry has become fluent in programmer. In the process of defining the logic of our surveys we’ve become better at articulating the skips and terminates within surveys. Good for us. It’s important that it works regardless of how it looks. But aren’t we at a point where we can develop a sub-language? Are there industry specific signals that we can include with the programming notes that tell us about the presentation of a question as well as the incumbent logic?  There should be universal codes we can develop for card sort, slider or ranking exercises (you’re thinking of anagrams already, aren’t you?)

Look, no one comes to work in jogging pants, so don’t tell me appearances are not important. I think it’s fair to say that most surveys should look a lot better than they do. You may be the kind of person that doesn’t care about how data is collected just as long as it’s accurate. But I’ll bet the impression you leave with your client of what a survey looks like is something a lot fancier that what you actually produce. By developing a language first, we can create a set of automatic tools. From there, just a bit more time in programming is needed and we can get there. It’s time to pull up those jogging pants.