I had an interesting discussion this morning which crystallised some ideas on respondent engagement. Since “respondent engagement” is a key element of my job it’s probably past time I started thinking through what it meant.
Anyway, what I realised in this conversation - and it’s really quite an obvious thing - is that people tend to use the terms "more interesting" and "less boring" interchangeably, when - in the case of market research surveys at least - they’re not.
Making the market research experience less boring for the participant means stripping out the things that make it tedious and frustrating: uneccessary duplication, extra screens, poor design, patronising instructions, absurdly long surveys, etc. (The list is endless!) Removing these will in theory improve data quality.
Making the experience more interesting can be done in any number of ways which have no beneficial effect on data quality. If, for instance, a survey of 18-30 year old men had Nuts calendar girls displayed on one side, a proportion would indeed find it a more interesting experience but the impact on quality would - I’m guessing - be zero at best. This is a ridiculous example but it’s not unrelated to some of the bells-and-whistles approaches I’ve seen to “engagement”.
It’s quite possible that any given survey has a minimum level of required boredom and no amount of fannydangle will get around that.
Luckily, I don’t have to directly worry about this stuff much. By turning data into a social object, social media completely explodes ideas of “data quality” anyway, but that’s another story ;)