As researchers using traditional panels, we all experience the heart dropping moment when we realise the results don’t look quite right. Sometimes it’s because genuinely surprisingly results have been uncovered – but other times it’s because the data is just unreliable. This leads to frustration – and sometimes anger – at the respondents who are, frankly, just wasting our time.
Or are they? I think sometimes we are too quick to judge. We ask a lot of respondents who are, for the most part, minimally rewarded for their time and effort. Imagine being a respondent, being paid just a few pence to spend 10 minutes trying to decipher confusing, lengthy, repetitive questions – how much effort would you honestly put in to get it right? Perhaps sometimes it is us who are the time wasters.
When writing a survey, we have to spend more time empathising with the respondent. How would we react being faced with the survey we have written? It is easy to read and enjoyable to answer – or clunky and confusing?
There are several simple ways we can make every survey better, in order to get more engaged responses and better quality data.
If the wording of your question or the answers options you’ve given only just about make sense to you (given that you know what data you need, what you’re getting at and what answers would be preferred), we can guarantee it won’t make any sense to a respondent. Make your questions clearer, so anyone – regardless of background – can understand!
If there is any element of a question that is open to interpretation – change it. Otherwise you’ll get some responses from people who understand the question one way, other responses from people who understand it to mean something else, and totally meaningless results.
Don’t be repetitive
In a conversation you wouldn’t tolerate being asked the same thing over and over with only small variations – why would that be different in a survey? If you sense check all data from the perspective of a face-to-face conversation with the consumer, you’ll soon spot areas where you’re being repetitive, or saying things you’d never dream to say face to face.
Guide the respondent through the survey
Instead of a series of unrelated questions, use text cards to introduce distinct sections so the respondent knows what they are being asked about. They can also be used to signpost when the respondent reaches half way, or is close to the end of the survey. This break in questions helps pace the survey, set expectations and keep the respondent engaged.
Speak to the right audience
There is little point in asking questions of a respondents who is not really in a position to answer your questions. Give thought to who you need to qualify in to your survey, and make use of the demographic filters Attest has in-built.
Tempting though it is – don’t be leading
The respondent will not have given themselves a talking-to before starting, about being objective and considering each question in an unbiased way. They are giving a quick and honest response to what they are seeing and being asked. In a survey, everything that has gone before will have an effect on how they consider (and react to) subsequent questions. So, consider the order you ask questions in, so you’re not giving the game away early on and encouraging respondents to choose any specific answer later on.
Finally, and most importantly, be respectful
The respondent is giving up their time to help you – treat them as such and you will get the results you need.