Using Behavioral Economics to Design Surveys | MDRG

Using Behavioral Economics to Design Better Surveys

Ann Marshall Tilton

Strategic Quantitative Analyst

Behavioral economics has made its way into pop culture lately. From Richard Thaler winning the Nobel Prize in 2017 and saying he would try to spend his award money “as irrationally as possible” to Michael Lewis’s 2016 book The Undoing Project about the work of Daniel Kahneman and Amos Tversky.

The discipline applies psychology to economics to understand behavior, and it has a lot to say about how humans (aka survey takers) make decisions (such as deciding how to respond to a question).

Quality consumer insights begin with precise survey design: the insight is only as good as the data behind it, and the data is only as good as the survey. As we design surveys, we consider the behavioral economics principles of anchoring effects and attribute substitution to help us write questions that encourage clear and truthful responses.

Anchoring Effects

Consider the following questions:
Q1. Please rate your interest in purchasing 100 Mbps High-Speed Internet at $60 per month?

Q2. What is the most you would be willing to pay for 100 Mbps High-Speed Internet?

In the above example, the price presented in Q1 anchors the respondent to $60 for 100 Mbps, causing her to consider this price and speed when answering Q2. While the respondent may be willing to go up to $70 per month, she is unlikely to be willing to go up to $90 because it seems so high in comparison to the anchor.

The order of survey questions makes a difference. We can mitigate the effects of anchoring by switching the order of these two questions.

Attribute Substitution

Attribute substitution occurs when a respondent is asked a complex question. Rather than answer the difficult question at hand, she may instead answer a more available or easily calculated question.

For example, a respondent might substitute
“Did you buy this coffee because you saw an advertisement for it?”
“Did you notice the ad for the coffee?” or “Did you like the ad for the coffee?”

The former question is difficult to answer because several factors generally interact to prompt a consumer to purchase. For example, the respondent may have seen a TV ad for the coffee so when a friend ordered it at a restaurant, she recognized the name. Two weeks later the grocery store near her house was offering a discount for the coffee, so she bought it.

It’s clear that these events had some effect on the respondent’s ultimate purchase decision. However, if she happens to remember the TV but did not like the main actor’s voice, she may answer the question “Did you like the ad for the coffee?” and say “no.”

To mitigate substitution effects, we recommend avoiding complex and difficult-to-answer questions in favor of concrete and straightforward questions. While a respondent may not be able to answer, “Did you buy this coffee because you saw an advertisement for it?” she will probably be able to answer the question “Which brand do you associate with [coffee slogan]”?

Further Exploration
If you’re interested in applying behavioral economics to consumer insights, check out our blog post “System 1 vs. System 2: How We Think”.

If you have a business question that you think can be answered with a quantitative survey, fill out the form below and we can talk!

Let's Talk!