availability bias

When I read that Judy Collins would be performing at the Freight and Salvage Coffehouse in Berkeley, I didn't hesitate to order tickets. After I'd shelled out more than $100 for two tickets, I began to question my judgment. Judy Collins was born in 1939. This is 2012. My memories of her are of recordings I own, most of them dating from the 1960s and '70s. She had a beautiful voice and wrote some fantastic songs. My impulsive act is an example of the availability bias at work. The first thing that popped into my mind on reading about her upcoming appearance at The Freight was of her singing songs like "Both Sides Now," "My Father," and "Someday Soon." (Those memories are very pleasant; there was affect bias involved in this decision, as well.) My decision to drive 150 miles (round trip) to see and hear Judy Collins was not based on any knowledge I had of what her voice or her performances are like now. I should have investigated a bit before investing time and money that I might later regret. After the fact, I checked out TicketMaster fan reviews. The first one goes like this:

Not only did we enjoy the show entirely. The non-stop performance that she did was just unbelievable. She did not even drink water. Judy is 71, and her voice has remained the same for all these years! She did a vast array of songs ranging from Christmas Carols to children's lullabies and, of course, the standard favorites. I would go and see her again. Audience participation was great! 

Favorite moment: When she replaced the pianist, who was extraordinary, and played six songs back to back without stopping. She is just an amazing pianist herself!

Wonderful! Now confirmation bias can kick in! The glowing review was a couple of years old, however. The next review on the list was dated just a few days ago and it read: "Beautiful, talented and ever elegant. I saw Judy in 1969 and she is just as good now. It was a terrific night all around." There were six more highly positive reviews, all dated within the last couple of weeks. (There were over 150 reviews and very few were totally negative.) Now I think I've made a good decision, but not for a very rational reason. I based my decision to go to a Judy Collins concert on the first thing that popped into my mind and the pleasantness of the memories aroused. Other things should have been considered before making the decision. I've heard other singers performing past their prime and it wasn't pretty. She's over 70 and her voice is unlikely to be what it was forty years ago. I feel better now that I've found some recent positive reviews. However, for all I know these positive reviews are written by shills or people who are easily pleased. I don't think that's likely, but it's possible. If I were a thorough investigator I'd seek out reviews by professional reviewers, but one of them might be really negative and say things about her recent performances that would make me second-guess my decision. Since I've already made the decision, why look for a wound and some salt to rub into it? (When I told a friend that I had tickets to see Judy Collins, he told me that he had heard her on a PBS program recently and her voice was cracking. She was no longer a solid soprano and her voice was cigarette-rough. But he's old and his hearing is going and ....)

The availability bias is a cognitive bias involving making quick judgments based on the speed with which memories are aroused and become available to the conscious mind. The main factors influencing the speed with which memories present themselves are recent frequency of similar experiences or messages, or the salient, dramatic, or personal nature of experiences. In our culture, the mass media play an important role in affecting what comes to mind quickly when we think of the frequency, importance, or causes of things. Rational judgments should be made on the basis of a consideration of all the relevant evidence, but many judgments we consider rational are made based on the ease with which they come to us. For example, a person might decide not to take a cruise to Alaska that she was about to book when she heard about the cruise ship Costa Concordia striking a reef near the Tuscan island of Giglio, killing more than 20 passengers. The safety of a cruise to Alaska has not diminished because of what happened off the coast of Italy, but the news reports and videos immediately bring to mind the horror of dying on a capsized cruise ship. The decision not to take the planned cruise has been biased by the news of the Costa Concordia. Likewise, many people refuse to fly on a commercial airliner because someone they love died in an airplane crash, yet these same people will drive thousands of miles every year rather than fly, even though they are more likely to be killed in an automobile crash than in an airliner crash.

When asked for your opinion on teenage drug use, premarital sex, morals of politicians, good stocks to invest in, the incidence of violent crime, or any other subject that mass media outlets are likely to cover, the odds are that your answer will be based on what comes immediately to mind and that will be heavily influenced by what you've read, seen, or heard recently in the mass media. Or, your answer will be heavily influenced by personal experience. What is unlikely is that your opinion will be based on objective or scientific knowledge of the subject. This tendency to make judgments by the ease with which ideas come to mind is called the availability heuristic.

Consider, for example, how people think of violent crime in the United States. During a five-year period in the 1990s, the homicide rate nationwide dropped by 20 percent, but coverage of murders on the national evening news programs at ABC, CBS, and NBC went up by 721 percent.  (Vincent Schiraldi, the Justice Policy Institute, 1998, cited in Carroll, Becoming a Critical Thinker, 2005, p. 62.) What do you think had more influence on people's perception of the incidence of homicide in our country? The crime rate decreased all through the 1990s, and for the last decade crime rates have remained steady. Yet, between 52% and 89% of Americans every year since 1990 have thought that crime is on the rise.* It's not just the excessive coverage of crime by the news media that influences our judgments about crime; entertainment programs also play a part in contributing to our misinformed minds. Consider this press release from Purdue University:

People who watch forensic and crime dramas on TV are more likely than non-viewers to have a distorted perception of America's criminal justice system, according to new research from Purdue University.

"These kinds of shows, such as 'CSI: Crime Scene Investigation,' 'Law & Order,' 'Cold Case,' and 'The Closer' are some of the most popular programs on television today, so it's important that we understand how they might influence people," says Glenn Sparks, a professor of communication who studies mass media effects. "We know they have inspired people to pursue careers in forensic science and law enforcement, but what are some of their other effects? We found that people who watch these shows regularly are more likely to overestimate the frequency of serious crimes, misperceive important facts about crime, and misjudge the number of workers in the judicial system."

Heavy TV-crime viewers estimated two and a half times more real-world deaths due to murder than non-viewers. Lawyers and police officers each make up less than 1 percent of the work force, but those surveyed estimated the percentage of the work force in those professions at more than 16 percent and 18 percent, respectively.

My local newspaper, the Sacramento Bee, contracts with the Mervin Field polling agency to do an annual poll on what issues Californians are concerned about. Two things stand out about this poll. 1. Those surveyed must select from a fixed list of issues, all of which are the kinds of things newspapers report on, e.g., crime and law enforcement, the state’s economy, public schools, controlling the spread of AIDS, unemployment, creating more jobs in new industries, illegal drug use, health care, taxes, inflation, illegal immigration, and toxic wastes. The issues are selected, according to Field, because they are public policy issues. 2. The results vary from year to year based on what stories have been in the news shortly before the poll is taken. After the  Chernobyl nuclear accident, guess what topped the list of Californians' concerns? If the economy is the topic of daily mass media coverage, guess what tops the list? If politicians have been coming down heavy on the issue of immigration, guess what happens in the poll? Anyway, when I read a headline that states Californians' top worry is crime, I interpret it to mean that there has been a lot of media attention given to crime stories lately and that Californians' top worry may actually be about keeping their home from being taken by the bank, finding a job, or how they're going to pay for their childrens' education.

The concept of availability as a bias grew out of the work of Daniel Kahneman and Amos Tversky.

One of our projects was the study of what we called the availability heuristic. We thought of that heuristic when we asked ourselves what people actually do when they wish to estimate the frequency of a category, such as “people who divorce after the age of 60” or “dangerous plants.” The answer was straightforward: instances of the class will be retrieved from memory, and if retrieval is easy and fluent, the category will be judged to be large. We defined the availability heuristic as the process of judging frequency by “the ease with which instances come to mind.” The statement seemed clear when we formulated it, but the concept of availability has been refined since then. --Kahneman, Daniel (2011-10-25). Thinking, Fast and Slow (p. 129). Macmillan. Kindle Edition. 

One of the refinements came about by distinguishing natural and automatic thinking from reflective thinking and seeing that availability plays a role in both. In their early work they focused on natural, automatic thinking--what Kahneman calls System I thinking. (System II thinking is deliberative and reflective, rather than automatic.)

The availability heuristic, like other heuristics of judgment, substitutes one question for another: you wish to estimate the size of a category or the frequency of an event, but you report an impression of the ease with which instances come to mind. Substitution of questions inevitably produces systematic errors.

Eventually, they and other researchers expanded the concept of availability to include more than just estimating size of a class or frequency of an event. In How Doctors Think, Jerome Groopman gives several examples of bad judgments in diagnosis made by physicians because they were made on the basis of the ease with which the diagnosis came to mind due to recent experience rather than on a careful consideration of all the patient's symptoms. One of Dr. Groopman's examples was of a doctor who had treated "scores of patients" over a period of several weeks with "a nasty virus" causing viral pneumonia. Then a patient presented herself with similar symptoms except that her chest x-ray "did not show the characteristic white streaks of viral pneumonia." The doctor diagnosed her as being in the early stages of the illness. He was wrong. Another doctor diagnosed her correctly as suffering from aspirin toxicity. The diagnosis of viral pneumonia was available because of the recent experience of many cases of the illness. Had his recent experience not included so many cases of viral pneumonia it is likely the doctor would have made the right diagnosis. After he realized his mistake, he said "it was an absolutely classic case--the rapid breathing, the shift in her blood electrolytes--and I missed it. I got cavalier."

One famous study on the availability heuristic was done by Norbert Schwarz and a group of German psychologists in the 1990s. They tested people to see what effect requiring them to list a specific number of instances of an event would have on their impression of the frequency of that event. You might think that the greater the number of instances you could list, the higher you'd estimate the frequency, but the results didn't work out that way. Why? Because it's easier to come up with few instances of something than a larger number of instances and the harder it is to complete the task the less frequent we think the instances are. This finding opens the door to manipulators. If you want to have a group of people, say, rate your speech favorably, then begin the evaluation form with a request that they list ten ways the speech could have been improved. The difficulty of the task might lead them to think that your speech couldn't have been improved very much and so they might evaluate you more favorably than if you just asked them to rank your speech on a scale of 1 to 5. Schwarz et al. also found that if they provided an explanation for the fluency of retrieval, even a spurious explanation, they could mitigate the difference in effect between the easier and more difficult retrieval tasks. This discovery led Kahneman to suggest that rather than call this the availability heuristic, it might be better to call it the “unexplained unavailability heuristic."

One professor found that teachers might increase favorable course ratings by first asking students to list ten ways the course could be improved (a relatively difficult task). He showed that the ratings were improved by replacing the word 'two' with the word 'ten' in the standard course evaluation form. It didn't matter that those asked to come up with ten ways to improve the course averaged about two suggestions; just asking them to come up with ten made the task harder and apparently motivated them to rate the course more favorably than those asked to come up with just two ways. ("The availability heuristic in the classroom: How soliciting more criticism can boost your course ratings" by Craig R. Fox, UCLA Anderson School and Department of Psychology.)

Another study on the availability heuristic was done by Paul Slovic, Sarah Lichtenstein, et al. They found that people's judgments about the frequency of death from various causes was biased by media coverage which itself is biased toward novelty (death by botulism or some rare virus is going to get more coverage than death by lung cancer) and drama (big accidents or dramatic events like airplane crashes will get much more coverage than, say, the annual death toll of those killed in railroad accidents in India which, by the way, was more than 15,000 last year, which, by the way, is about the same number of people killed by the tsunami in Japan a year ago). The researchers found, for example, that 80% of the 660 adults surveyed thought that there are more deaths from accidents than from strokes. Strokes actually account for about twice as many deaths per year than death from accidents. The majority of respondents thought that death from tornadoes is greater than deaths from asthma, even though the latter accounts for about twenty times as many deaths as the former. Schwarz and his group, however, found that personal involvement in an issue will make a person less likely to go to fluency or whatever pops into one's mind automatically when making judgments and more likely to consider the number of instances they can retrieve from memory.

They recruited two groups of students for a study of risks to cardiac health. Half the students had a family history of cardiac disease and were expected to take the task more seriously than the others, who had no such history. All were asked to recall either three or eight behaviors in their routine that could affect their cardiac health (some were asked for risky behaviors, others for protective behaviors). Students with no family history of heart disease were casual about the task and followed the availability heuristic. Students who found it difficult to find eight instances of risky behavior felt themselves relatively safe, and those who struggled to retrieve examples of safe behaviors felt themselves at risk. The students with a family history of heart disease showed the opposite pattern—they felt safer when they retrieved many instances of safe behavior and felt greater danger when they retrieved many instances of risky behavior. They were also more likely to feel that their future behavior would be affected by the experience of evaluating their risk. Kahneman: Thinking, Fast and Slow (p. 135).

Knowledge and personal experience, however, are often of little help in overcoming the strong emotional reaction people have to vivid scary stories in the media. On an intellectual level, one may know that the chances of the cruise ship to Alaska sinking are slim, but the fear that has been aroused by the media accounts of the sinking of a cruise ship in another part of the world may be impossible to overcome when making the decision to go ahead as planned or cancel the trip.

The way to mitigate the availability bias is to be aware of it and to take the necessary steps to get good data before making a judgment. Unfortunately, scientific studies have shown that certain kinds of personality traits make one more susceptible to the availability bias. Schwarz et al. found that people who have great faith in intuition and people who are powerful (or made to feel powerful) tend to be affected more strongly by ease of retrieval than by the content they retrieve, if indeed they even bother with much content. I leave it to the reader to find the research showing that when the first thing evoked in memory is a stereotype or is accompanied by very pleasant feelings, it becomes more difficult to overcome the availability bias.