One of the most basic tenets of good social or market research design – particularly surveys – is to make answering as easy as possible for your respondents. Each respondent only has a finite amount of willingness to complete your survey, and this is depleted more rapidly if entering their responses is difficult.
Take the following example from a survey I received administered by an international customer insight group. Overall the survey was well-structured and relatively straightforward, if a little long. But one question stood out as being unnecessarily convoluted:
The question asks you to enter your postcode, but rather than offering a standard text box to enter your full postcode it asks you to select each individual element of your postcode from a drop down list. This includes:
- Area (the first letter(s)
- District (the following one or two numbers)
For most respondents that is arduous. There are probably hundreds (at least tens) of options for each element, so that means a very long drop down box. If you don’t know that you can begin typing characters on your keyboard to move to the relevant place of the drop down box, that is a very long and tedious process of scrolling down to find the right letter or number for your postcode. Then you have to do that another three times to complete all four elements. As the question is optional, I would imagine a large proportion of respondents will simply not complete the question.
I can only think of two reasons why the survey question is set up in this way. One (and the most likely) is to reduce administrative overheads and the time needed to check, cleanse and verify the data. The second reason is perhaps that it was felt that this way looked more impressive or sophisticated.
In either case, I think a better option would be to provide a text box. People use these every day: if I provide a text box for you to enter your postcode, you can simply type it in. And most people will enter it in the correct format as they’re used to doing it on a regular basis. Although this will mean someone will need to check and cleanse the data, it’s actually a very straightforward procedure and shouldn’t take too long with a spreadsheet. The point is, the cognitive work is being done by the people administering the survey, not the respondent. The benefit of this is that it should improve the response rate for the question and the survey overall.
If it was done to look more sophisticated, then I think that desire is misplaced, and that effort should instead be spent making the survey look as simple and straightforward as possible to maximise response rate.
So there you have it. Like so many things in life, simple and straightforward is usually the best approach to take. In statistics we call it the principle of parsimony (or Occam’s Razor), but maybe there’s irony there in not calling it the principle of simplicity.