So You Need A Survey, Do You?

Americans are just so gosh darn optimistic.*

Ask ‘em. Ask ‘em what they intend to buy in the next year. Then compare those results to what they did buy last year. Everyone intends to purchase major appliances, a new car, or a Caribbean vacation next year.

But somehow they just won’t be able to come up with the money this year.

That’s the problem with surveying intentions. Your survey is worthless. People’s intentions are not an accurate predictor of their actions. When people are faced with real choices, their decisions are quite different than when they are questioned about hypothetical choices.

For instance, a recent AARP** poll shows 48% of workers over the age of 40 say they are “somewhat confidant” about their retirement plans. Yet 52% of those same people have not even attempted to figure out how much they’ll need to live on. 22% of them have no savings at all.

We all see ourselves differently than we act. People can’t predict what they would do. You can’t predict it, either. You can only observe what they do.

A good question.

A good question asks for only one answer. “Were you satisfied with our food and service?” won’t provide information you can act on.

If a diner answers “no” was it the food, the service, or both that were not satisfactory?

What about confidentiality?

Questions of a sensitive nature require a degree of trust on the part of the respondent. Either make your questionnaire anonymous with no identifying information, or clearly state your confidentiality policy.

Better yet, do both.

How should the questions be sequenced?

Some researchers have reported that the order of the questions can effect the way people respond, claiming that when general questions are asked before specific questions, the answers are unlikely to be affected by each other. Specific questions about crime, for instance, could change the degree of reaction to the more general “crime prevention” questions which followed.

Other investigators have claimed just the opposite: that putting the specific questions first keeps people more interested in the more general questions.

On one thing everyone agrees, however. Each question should flow comfortably from the previous question. Jumping to non-related questions tend to produce low response rates.

Get rid of those multiple choice check boxes.

Case study: a television/newspaper campaign to sell individual units in a condominium development. Over the first ten days of the campaign foot traffic through the models went up 350%.

Imagine my surprise when the sales manager told me the advertising wasn’t working.

How did he come to this conclusion? He gave every person who stopped in a questionnaire. The first question: How did you hear about Chancellor’s Row?

Nearly every one had checked the first box: “Just driving by and saw your sign.”

Well, of course they saw the sign,” I said. “They’ve driven by here hundreds of times. This time, however, they pulled in and took the tour.”

I had the cards reprinted with a big empty space where all of the check boxes used to be. When people weren’t able to take the easiest answer, and had to think about the question, we got answers that made more sense.

Do you actually need a survey?

Is it possible that some other organization has asked the same questions, and has already found the answer?

Perhaps the Federal Government or an institution of higher learning has already conducted a similar survey and has the results already available. Check with your public library or the local Small Business Development Centers.


* I could just as easily said “Russians are optimistic” or “Austrailians are optimistic.” This is a clearly human trait.

** AARP Bulletin / June 2006


Submit a Comment

Your email address will not be published.