A few weeks ago I listened to a very interesting Freakonomics podcast hosted by University of Chicago economist Steven Levitt. In the podcast, Why Our Judgment is Flawed — and What to Do About It, Levitt interviewed Daniel Kahneman about his recent book, Noise: A Flow in Human Judgement, co-authored with Olivier Sibony and Cass Sunstein. Kahneman is Professor of Psychology and Public Affairs Emeritus at Princeton University. In 2002, he was awarded the Nobel Prize in Economics “for having integrated insights from psychological research into economic science, especially concerning human judgment and decision-making under uncertainty.”
Thinking, Fast and Slow, Kahneman’s 2011 bestseller, was about the major discoveries by psychologists and cognitive scientists that have led to our current understanding of judgement and decision-making over the past several decades. Up to the 1970s, the prevailing view among social scientists was that people are generally rational and in control of the way they think and make decisions. It was thought that people only departed from rational behaviors because powerful emotions like fear, hatred or love distorted their judgement.
These assumptions were challenged by the pioneering research of Kahneman and his long time collaborator Amos Tversky, who died in 1996. In a series of experiments, they demonstrated that human behavior often deviated from the predictions of the previous rational models, and that these deviations were due to the machinery of cognition, that is, to the biases and mental shortcuts or heuristics that we use for making everyday decisions, rather than to our emotional state.
Kahneman has been studying biases and how they lead to errors in judgement for most of his career. But a few years ago, he encountered another type of error, noise, that he hadn’t thought about earlier. Noise is “the sort of ignored stepsister of bias… the complement of bias. And in terms of the standard way of measuring accuracy and measuring error, bias and noise are completely independent of each other. … bias is really the average error. And noise is the variability of error, the standard deviation of errors.
In a recent NY Times guest essay, Kahneman, Sibony and Sunstein illustrated the difference between bias and noise with a few concrete examples. We use the word bias “when there is discrimination, for instance against women or in favor of Ivy League graduates. But the meaning of the word is broader: A bias is any predictable error that inclines your judgment in a particular direction. For instance, we speak of bias when forecasts of sales are consistently optimistic or investment decisions overly cautious. Society has devoted a lot of attention to the problem of bias - and rightly so.”
Noise, another type of error that leads to mistaken judgements and unfortunate decisions, has attracted far less attention. “To see the difference between bias and noise, consider your bathroom scale. If on average the readings it gives are too high (or too low), the scale is biased. If it shows different readings when you step on it several times in quick succession, the scale is noisy.”
According to the essay, noise is a large source of malfunction in society, including in important institutions like the criminal justice system. “In a 1981 study, for example, 208 federal judges were asked to determine the appropriate sentences for the same 16 cases. The cases were described by the characteristics of the offense (robbery or fraud, violent or not) and of the defendant (young or old, repeat or first-time offender, accomplice or principal).”
“You might have expected judges to agree closely about such vignettes, which were stripped of distracting details and contained only relevant information. But the judges did not agree. The average difference between the sentences that two randomly chosen judges gave for the same crime was more than 3.5 years. Considering that the mean sentence was seven years, that was a disconcerting amount of noise.” That means that the same person might sometimes be sentenced to almost 9 years in person and other times the sentence might be a bit over 5 years.
An essential aspect of justice is the principle of fairness. We want people who have committed the same crime to get a similar sentence. But instead, “It is hard to escape the conclusion that sentencing is in part a lottery, because the punishment can vary by many years depending on which judge is assigned to the case and on the judge’s state of mind on that day. The judicial system is unacceptably noisy.”
A second example comes from a 2015 study of underwriters in a large insurance company that Kahneman was involved with. To evaluate if the underwriters all operated in a similar way, the study ran a series of experiments or noise audits. Forty-eight underwriters were presented realistic summaries of insurance cases to which they assigned a premium based on the risks involved, just as they did in their jobs. Prior to the experiment, the executives of the company were asked how much of a difference, that is noise, they expected between the premium values that two competent underwriters assigned to the same cases. The executives said that they expected about a 10 percent difference. "But the typical difference we found between two underwriters was an astonishing 55 percent of their average premium - more than five times as large as the executives had expected.”
Another experiment in the same study involved payouts to someone who had fallen and gotten hurt. Claim adjusters were given the same exact cases and asked to assign the payment that the insurance company should offer. The actual fair value of the payout was $20,000. Executives expected the difference between two claim adjusters to be between $1,000 and $2,000. But, in fact, the actual difference between two adjusters was closer to $10,000.
Large variances between claim adjusters can be very costly. If the proposed settlement is too high, the company is paying out too much. If it’s too low, there will likely be litigation. The insurance company realized that noise was costing them over $1 billion a year and put in place a program to get rid of a large chunk of the noise.
“Many other studies demonstrate noise in professional judgments. Radiologists disagree on their readings of images and cardiologists on their surgery decisions. Forecasts of economic outcomes are notoriously noisy. Sometimes fingerprint experts disagree about whether there is a ‘match.’ Wherever there is judgment, there is noise - and more of it than you think.”
Where does noise come from? The NY Times essay cites three potential sources of noise:
Irrelevant circumstances can often affect judgements. In criminal sentencing, for example, it could be a judge’s mood, fatigue, the time of day, the weather, or a personal matter that has nothing to do with the case.
Different people have different general tendencies. There are harsh judges and lenient ones; generous claim adjusters and stingy ones; aggressive-treatment physicians and watchful-observance physicians.
Different patterns of assessment. “We celebrate the uniqueness of individuals, but we tend to forget that, when we expect consistency, uniqueness becomes a liability.” Judges may differ on the type of cases that merit being harsh or lenient about; underwriters may differ in their views of what is risky; physicians may have different views on which conditions call for aggressive treatment and which for watchful observance.
How can noise be reduced? Structure, discipline, and a good processes will generally lead to better judgements and decisions. Problems should be broken down into smaller pieces that can be evaluated independently and then put back together for the overall decision. Enough information should be gathered for a well understood and thought-through decision. And averaging independent judgements from multiple people will generally lead to better decisions.
“No noise-reduction techniques will be deployed, however, if we do not first recognize the existence of noise. Noise is too often neglected. But it is a serious issue that results in frequent error and rampant injustice. Organizations and institutions, public and private, will make better decisions if they take noise seriously.”
Thank you for sharing! It is indeed a problem we don't hear often in comparison to bias. As we make decisions, we often ignore the noise in our decisions. In addition, we often ignore how cognitive dissonance affects our decision process.
Best,
Benny Xian
Project AI+Compassion
www.ai-compassion.com
Posted by: Benny Xian | September 20, 2021 at 05:43 PM