Financial strategists, medical advisers and venture capitalists that are considered experts in their fields play a crucial role in major organizations, but are more likely than novices to make overconfident predictions after being told they are wrong, according to a Rutgers study.
Analyses of six studies – including data from chief financial officer predictions of stock market returns and calls by Major League Baseball umpires – show that negative feedback about mistakes can result in experts doubling down with over precise predictions and assessments.
“We typically think an expert has the skills and confidence to perform despite adverse conditions,” said Jerry Kim, coauthor of the study and an assistant professor at Rutgers Business School. “But the reality is that confidence is built upon a foundation that can be threatened. Particularly in a time of constant feedback, every expert is subject to extreme scrutiny.”
In lab experiments, the researchers analyzed baseball fans and nonfans after getting negative feedback on how they thought a pitch to a batter would result (i.e., a fly ball, a bunt, etc.). In addition, expert investors and college students watched videos of business startup pitches and assigned probabilities on possible investment ranges.
Subjects primed to think they were experts – namely the baseball fans and investors – made overly precise judgments 16.3 percent of the time, while those who thought of themselves as novices made overly precise predictions 22.4 percent of the time. However, after receiving negative feedback, the expert group’s rate of overly precise predictions increased to 27.1 percent, whereas novices went down to 13.2 percent.
According to the study, negative feedback worsened the performance of the experts, while novices learned from the negative feedback.
The researchers supplemented the experiments by investigating expert responses to negative feedback looking at data from umpire strike balls called during 408 Major League Baseball games from 2014 to 2018 and data from a quarterly survey of 2,000 to 3,000 U.S. chief financial officers (CFOs) on their predictions of Standard and Poor’s 500 index returns.
The researchers found when previous quarterly results were below what CFOs projected for the year in S&P 500 returns, the chances their predictions would be overly precise increased by 26 percent.
“As we saw during the pandemic, the constant criticism toward scientific experts led to a general mistrust in an expert’s judgment,” said Sanghoon Hoonie Kang, a former Rutgers doctoral student who teaches at The Chinese University of Hong Kong Business School. “At the same time, the experts themselves often doubled-down and responded in ways that further alienated the public and undermined their credibility, which we see as the consequence of the threat to their expert identities.”
The researchers suggest an easy fix to help experts in their decision-making: Before the prediction tasks, experts should use a common self-affirmation technique, such as writing a short paragraph about important values and examples of upholding those values. When used, the tendency for overconfidence following negative feedback decreased and experts learned from the negative feedback in subsequent prediction tasks.
Kim and Kang said buffering self-worth in advance of potential negative feedback can help experts produce better predictions.
Business leaders need to think about how they deliver negative feedback to their employees, too. Creating organizational cultures in which errors and mistakes are accepted and seen as opportunities for learning are crucial to maximizing the potential of experts, they said.
Source: RUTGERS UNIVERSITY
Academy of Management Journal