What About Survey Fatigue?

In talking about potential survey projects with clients, I am often asked about concerns that they have about survey administration. I’ve previously discussed the most-oft voiced concern, which is about response rates, in this blog post here. Today I want to discuss another concern, which is about “survey fatigue.”

Survey fatigue is the supposition that if people are asked too often to take surveys, at some point they will get tired of it and just refuse to take any more surveys. Like many things in life, this turns out to be more of an unfounded belief than a fact established through scientific inquiry.

Let’s do a thought experiment here. You are asked to take a survey, which you do. Then years go by, let’s say ten years, before you are asked to take another survey. Are you going to think “wait a minute, I just did one of these ten years ago?” No, probably not. At the other end of the thought experiment is someone getting one survey request every minute. Surely that is too much. So there must be a point in between that is the tipping point. The question is, what is it? What is the time period that makes it safe to ask someone to take another survey? And by safe, we mean will they open the survey and start filling it out.

The thing is that survey response is multifaceted.

It’s not just how often you get a survey request. It’s who makes that request. Is it someone you know of, or a random company? Or is it a thinly veiled “survey” that is really an attempt to get you to contribute money to a politician? That is one thing we know impacts if people respond: who is asking them to do so.

It’s also how well the survey is crafted. Are the questions obviously unbiased? Nobody wants to take a survey that only exists to lend credence to some particular cause or result.

Is the survey interesting to people responding? If it is boring, then people will stop filling it out. I always try to start my surveys with something that will interest the reader. A common mistake by a survey novice is to start by asking boring demographics. How old am I, where do I live, am I married, blah blah blah. That’s boring. I already know the answers to that. Quit out. But if the survey asks me to reflect on what aspects of my schooling were important to me, then that’s more interesting.

Will the results matter? Very important. If the results are going to lead to positive change, then you are going to have people invested in answering.

People think length matters, but it really doesn't, according to the research and my experience. I can write a 15-20 minute survey that will get 95% response rate if the respondents think there is a respectable organization conducting it and that the results will matter to them.

So, those things matter. Survey fatigue? Not so much. In this recent research paper they found mixed evidence that survey fatigue existed. One of the papers they cite was from colleague of mine formerly at Wesleyan University. In a series of elegant experiments they found that it mattered a little bit, but that with college students they could ask for survey responses every few months without reducing the response rates.

Every few months is perfectly reasonable to me.

So we don’t need to be too worried about survey fatigue. I’m more concerned with a well-crafted survey administered in a thoughtful way, which is what we all deserve.