Darden School Researcher Explores the Science Behind Snap Decisions

Newswise — The holiday gift-giving season is here.

In a world that is short on time but long on merchandise, how can holiday shoppers make quick decisions about what to buy for the people in their lives? Once they make the decision, how do they know if they got it right? According to Darden Professor Yael Grushka-Cockayne, humans apply “similarity features” associated with certain events or categories to make decisions. Formally, this phenomenon is called the “similarity heuristic.” Grushka-Cockayne explains that it tends to guide people down the right path in a study published in the Journal of Behavioral Decision Making.

“The field of heuristics examines how people make quick decisions based on prior knowledge or experience in place of carrying out exhaustive research. The similarity heuristic can be used when a classification decision must be made,” Grushka-Cockayne says.

So, let’s say a shopper is buying a gift for someone about to enter business school. They might assume that MBA students enjoy reading business magazines and decide to buy a magazine subscription for their gift recipient.

“Much of the time, we humans make good snap decisions,” Grushka-Cockayne adds.

This may be good news for last minute-holiday shoppers, but in business, decisions are more complex. For instance, hiring managers, like shoppers, must choose from a wide range of possibilities.

One camp in the field of heuristics believes that snap judgments lead to biased decision-making, while another camp asserts its merits. To test these beliefs, Grushka-Cockayne and co-author, Professor Daniel Read of the Warwick Business School in the U.K., randomly assigned 160 volunteers to various experimental conditions. They used artificial stimuli to represent different populations.

“We provided a mathematical model of the heuristic and tested it on 120 sets. Each set contained two populations of different colored rectangles that were comparable to many natural populations,” she says. “For example, most cities can be characterized by their population’s ethnic mix.”

The two large populations were made up of a diverse community represented by 100 blue, green and yellow rectangles. There were more blue rectangles than any other color. The researchers paired the two populations to generate smaller samples containing a diverse mix of 25 blue, green and yellow rectangles. The study volunteers had to choose which populations the samples came from based on their mix of colors, much the way a birdwatcher might categorize a species of bird based on its appearance.

The researchers placed the volunteers into one of four groups: two “similarity” groups and two “choice groups.” The two similarity groups assessed whether a sample was more similar to population one or population two. This exercise showed whether making ‘‘useful’’ similarity judgments depends on the volunteers knowing why they were making them. One group knew the context for their assessments and the other did not.

The two “choice” groups included volunteers who chose which of the two populations a sample was drawn from. The study showed whether people use the similarity heuristic and whether prior knowledge about the populations affected their choices and interacted with similarity judgments. One choice group had no knowledge of prior probabilities while the other choice group did.

Grushka-Cockayne and Read uncovered that people successfully made similarity judgments 86 percent of the time, compared to leaving a decision to chance. They suggest that in most instances in life, people do well to make a snap decision.

So what does this mean for hiring managers?

They can trust their instincts. Snap decisions made about candidates using the similarity heuristic are generally accurate.

“In addition to hiring, this area is pertinent to business activities such as project selection,” says Grushka-Cockayne, who also conducts research in the area of project management decision-making and new product development in the Behavioral Research at Darden (BRAD) Laboratory.

Recently, she received a grant from the Batten Institute to study the general causes of time and cost overruns in R&D, new product development projects and to develop a method, in the form of a project classification scheme, for producing more accurate project forecasts. Read will serve as the co-author on this project as well.

Grushka-Cockayne teaches in the area of Quantitative Analysis at Darden. Her courses include “First Year Decision Analysis” and “The Business of Film - the Cannes Film Festival” elective, among others.

For questions or information, contact [email protected].