Seeing things all at once has less of an impact
Image: iStock/Mehmet Salih Guler
-
The Wisdom of Psychopaths
In this engrossing journey into the lives of psychopaths and their infamously crafty behaviors, the renowned psychologist Kevin Dutton reveals that there is a...
Read More »
Rationality is the crowning achievement of our species. The ability to use evidence is true the cornerstone of science, medicine, and our legal system. We use rational methods, too, in daily life – we assess an applicant’s resume, a child’s IQ, or the mileage of a used car to predict the likelihood of good performance later on. Given that we often use information to make decisions both large and small, how good are we at assessing evidence?
There is a line of psychological research that studies precisely this, by measuring how accurate we are at making probability judgments. One way to study this is to control the nature of information itself and see whether people are accurate judges of its strength. Interestingly, people’s responses tend to be conservative: they are less sure of their conclusion than the evidence justifies. Yet we are not just affected by the strength of the evidence, but by how it is presented. A recent study by Jennifer Whitman and Todd Woodward found that when pieces of evidence are doled out one at a time, instead of being shown all at once, people conclude that the evidence is stronger.
Imagine, for example, that you are in a library (assuming people still do such things), and you’ve become lost. Are you in the Science Fiction or the Fantasy section? Of course, you could wander the shelves until you find a helpful sign, but it’s faster to simply look at the books on the shelf next to you. You see:
Book 1: Piers Anthony’s Blue Adept: The Apprentice Adept
Book 2: J. K. Rowling’s Harry Potter
Book 3: J. R. R. Tolkien’s The Hobbit
You’re not sure how to categorize Book 1, so it’s not good evidence for either Science Fiction or Fantasy. Books 2 and 3, however, have wizards or elves on their covers, and you might firmly classify them as Fantasy. By now, you’ve weighed the evidence and concluded you’re in the Fantasy section.
Here’s where things get interesting. If someone had simply handed all three books to you at the same time, you might feel that it’s somewhat likely you are in the Fantasy section. But if someone handed the books to you one at a time, you might conclude very strongly that you’re in the Fantasy section. Even though the books are the same, you would weigh the strength of the evidence more heavily when you processed them in turn, rather than all at once.
Researchers Whitman and Woodward recently demonstrated this effect in a controlled laboratory setting. In their study, people looked at a display on a computer screen. At the bottom of the screen was a little pond connected to two big lakes. The pond contained three fish – say, two white ones and one black one. Then the two lakes big lakes were filled with different proportions of white, black, and yellow fish. People looked at the lakes and the pond, and used a sliding scale to judge the probability that the fish in the pond came from Lake 1 or Lake 2. Sometimes there was strong evidence, or a high probability, that the fish were from Lake 1 (Lake 1 had mostly white fish, and some black fish, like the pond). Sometimes it was weak evidence, or a low probability, that fish were from Lake 1. Also, sometimes the fish in the lakes were added in sequence: all the white fish appeared, then the black ones, then the yellow ones. Other times, all the fish were added all at once. When the fish were added one at a time, people perceived the evidence to be stronger.
This is an intriguing finding about how our minds work: even in dry, laboratory studies, we are imperfect rationalists who judge evidence not just by its actual strength, but also by how it was fed to us.
See what we're tweeting about
12 Comments
Add CommentSounds to me like the Challenger case noted in the article was less an issue of the order in which the information was presented & more to do with the information design, in this case the order & visual prominence.
Reply | Report Abuse | Link to thisHad the cautions come first & been bold or written in red, they almost certainly WOULD have been noted, perhaps even more so had they been revealed one at a time.
Of course, most people draw their conclusions from news reports like this one rather than, in the field of science and engineering, fully analytical research reports upon which the news reports are based. In many cases the public's conclusions are even based solely on hearsay about news reports...
Reply | Report Abuse | Link to thisAn excellent example of this phenomenon in science is the recent preliminary research report that neutrinos slightly exceed the maximum speed of light in a vacuum. Many people accepted the preliminary results because they seem to support many fantastical ideas fostered upon the public about intergalactic space travel, etc. Even the recent revised test providing confirming results does not 'prove' such extraordinary claims: this may occur if and when the FTL neutrinos results are reproduced by independent laboratory tests.
In the meantime, the public's imagination has already generated many stimulating ideas regarding new science!
I suggest that some enterprising student consider this event as the basis for a thesis!
This agrees with the common courtroom tactic where suppressing a piece of evidence until it is revealed at an opportune moment can be quite damaging in a courtroom. Saying "A witness has just come forward..." might be given more weight than if that witness was presented with others.
Reply | Report Abuse | Link to thisThe same could be said for political campaigns. Revealing a hidden fact late in a campaign is a common ploy to damage an opponent beyond recovery before an election.
How about the stock market? All the good analysis in the world might wilt in the face of an "insider" tip.
It seems this study may just recognize a well known fact about manipulation of knowledge.
Isn't the probability different for a sequence of events than for a set, e.g., 3 heads in a row vs. 3 heads showing when all are tossed at once? If so, mightn't that account for the added-value interpretation of sequentially presented data, and if not, is my faulty memory for this kind of high school probability problem further evidence for the bias the article suggests?
Reply | Report Abuse | Link to thisWell, consider the other case: how our minds are *not* swayed by evidence. "Skeptics" and dogmatic stalwarts of all types poo-pooh so-called "conspiracy theories" -- even when those theories are supported by evidence. In the case of 9/11, the collapse of Bldg. 7 continues to be ignored as a challenge to the official story. There are even some who continue to believe Oswald was the sole assassin in the murder of President Kennedy, even though the Warren Commission's own theory required that one bullet stop mid-air for between one and two seconds for him to accomplish that act.
Reply | Report Abuse | Link to thisHere again you might say evidence is based on presentation: when it is presented in the official "credible" media, it is believed. When it is presented by those who have been painted as belonging to the "lunatic fringe" it is discounted no matter what.
More pointedly, however -- as it relates to "conspiracy theories": people quickly become bewildered when it comes to the complexity of the so-called theories -- even though any reasonably detailed historical account of a given event will be similarly complex. But as the conclusions described in the above article note, given a hundred pieces of evidence, an ordinary person is likely to feel *more* uncertain about the findings. Whereas if he or she is presented with one or two telling facts that person is somehow more likely to feel "sure".
Reply | Report Abuse | Link to this"Also, sometimes the fish in the lakes were added in sequence: all the white fish appeared, then the black ones, then the yellow ones. Other times, all the fish were added all at once. When the fish were added one at a time, people perceived the evidence to be stronger."
Reply | Report Abuse | Link to thisPresumably though, if they added the yellow fish (or a fish which was not in the pond) in first, then the evidence would be weaker? The way this experiment has been described, it sounds much like a primacy effect which is biasing an individual's interpretation towards whichever color of fish is seen first. This makes sense, as the individual is exposed to three sets of probabilities in your lake 1 example: 100% white fish, 50% white/black, 33% white/black/yellow. And the individual must update these probabilities as they arrive, possibly weighting the first probabiliy value more strongly than subsequent ones. But, if they reverse the order, is the effect reversed as well? This is something I would like to see. Another potential confound is that recency effects (e.g. the fish color which arrived in the lake last might bias reports) might modulate the probability estimates. Overall, I'd really like to know whether the order in which fish colors were presented has a differential effect on probability estimates, in addition to an effect of presentation method.
On a side note, I sometimes feel as though the greatest strength of psychological experiments (their cleverness) is also their greatest failing: as clever experiments are designed only to tap one very specific aspect of human cognition, and yet this ignores how we actually think about events in our lives: we do so by taking many perspectives all at once. Hence, analyzing any single aspect of human thought in isolation is bound to produce an artificial environment, which may completely alter how a person's cognition occurs. (I it would be remiss not to point out that this is possibly the most common critique of psychological laboratory studies--indeed, of many studies in biology as well.)
I wonder if confirmation bias could be a factor here, i.e., if the first piece of evidence suggests that theory "X" is true, the observer could acquire an initial, possibly weak, belief in theory X. They might then unconsciously (subconsciously?) interpret additional evidence in the light of this belief. It would be interesting to compare cases where the first piece of evidence supported theory X vs cases where the first piece did not.
Reply | Report Abuse | Link to thisWe have to distinguish two separate phenomena: 1st: the time we study a subject. It is very important to have enough time to study one object. It is not only challenger case but J.F. Kennedy as well. He was a fast reader. If he was taking care of the subject maybe there was a report about the plot. 2nd: it is very valuable to have facts, if necessary, all together. For example map of the world. If they show you only the continents, Could you imagine the oceans and how big they are?
Reply | Report Abuse | Link to thisIt's certainly the case that people can be flim-flammed by form over function, by how things appear versus how they are, and I've encountered cases where people have said, "I know the evidence isn't very strong but the way he put it just sounds better."
Reply | Report Abuse | Link to thisAny good trial lawyer, advertising or public relations person or political consultant is aware of this and advises clients accordingly. I've followed the career of Carl Rove for some years and it's hard to find a better flim-flam man. People of similar talent exist among Democratic activists as well.
It's frightening that people are this easily influenced. Several years ago I compiled a collection of 1,125 quotations on this subject entitled Propaganda, Persuasion and Deception. ISBN 0-9761337-0-9 It's available as a free, downloadable PDF file at several locations on the web. I think you'll find it interesting.
You Heinrich66, are the lunatic fringe.
Reply | Report Abuse | Link to thisThis result seems quite reasonable to me. The brain is constantly trying to make patterns out of what it sees. Upon being presented with an initial set of facts (single or multiple), one is going to create an hypothesis. Regardless of how many facts are in the original group, one now wants to confirm or refute this hypothesis.
Reply | Report Abuse | Link to thisBeing initially presented with many black fish in a lake does not necessarily give any more weight to the statement that "all fish in this lake are black" than being shown one black fish, unless you are told that the group was picked at random over a period of time. It would be easy to get all black fish if they were all picked at once from a school of black fish. We intuitively know that facts gathered over time are a better random sample than any snapshot.
Yes, the behaviours we have evolved to deal with understanding the natural world can be tricked by carefully constructed presentations, but the major point here is that by understanding ourselves a little better, we can reduce our risk of manipulation.