Once a Cheater, Always a Cheater. Or So You think.
Ever met someone you just don’t trust? Maybe it’s something about their face. Maybe you heard something about them from a friend that made you wince in disgust. Research shows that this distrust tends to be a stubborn figment in our imaginations—even when we learn that our reasoning for distrusting someone is unfounded, we have a hard time accepting that the person in question is trust-worthy. A group of cognitive psychologists from Japan wondered why this is the case. Their question: why is it that we’re so good at remembering people who are “cheaters?” Given that we’re social animals cooperatively working to make this thing called society work, is it possible that we’re hard-wired to explicitly identify others who take nefarious advantage of our cooperation? Perhaps evolution is at play, and we need this ability to continue to make society viable (Suzuki, Honma, and Suga, 2013). They wondered just that and decided to study this question with a series of experiments testing the durability of stigma participants held in their study.
We know that we’re stubborn. Once our minds are made up about a person, it’s quite difficult to reverse our opinion of them. This is particularly true if we have a negative opinion of someone else. We seem to possess a bias for remembering people who play outside the rules of the game more than we do for people who honor said rules. The question, from a cognitive psychological point of view, then becomes: how can we detect this bias? Do we study how people come to understand someone is a “cheater?” Or do we study what maintains our negative perceptions of them? These cognitive psychologists set out to determine just this. They devised a study that will allow them to tests just how durable our negative perceptions of others are, even when we learn that the basis of perceptions were unfounded.
These scientists recruited a bunch of college students with a roughly fifty-fifty gendered split and asked them to pretend to be bankers. They were to play a game called the Investment Game where they looked at a series of faces and decided whether or not they’d lend to the owners of these faces based on how trustworthy they seemed. No information on their financials or on their occupation was offered. All they had to go on was their faces, and make intuitive judgments accordingly.
Next they were given useful information that would enable them to make sound investment decisions: the participants were shown which faces belonged to bad borrowers who would embezzle their earnings and which faces belonged to good borrowers who would honorably pay back the loan with interest. Their faces were associated with the label good borrower and bad borrower accordingly to the bankers. Before returning to the investment game, participants were asked to fill an inconsequential questionnaire to scramble their memory some.
The Investment Game was reinitiated and the pretend bankers now repeated the process with having prior knowledge of who was a safe bet and who wasn’t. The catch is that they have to remember who was which. After the game was over the researchers informed the pretend bankers that their intelligence was wrong. It turns out they didn’t actually know which faces belonged to wholesome, honest folk and which belonged to conniving deviants, so they were asked to play game again. Now they would have to replay the game, going off intuition like before. This was done to test how well the label of good borrower and bad borrower stuck even after the pretend bankers learned the labels were applied arbitrarily. The researchers then gathered the results and looked at who the participants identified to be trustworthy and untrustworthy across each session of the game.
It turns out the pretend bankers indeed held a bias for remembering stigmatized faces over the faces of good borrowers. Even after they had been told the criteria for good and bad borrowers was broken after the second trial of the game, the participants tended to continue to perceive the allegedly “bad borrowers” to be bad investment decisions.
To make sure the results weren’t tainted by lack of motivation on the part of the participants to unlearn the stigmas attached to the faces during the face-type association task, they ran the procedure again with a different set of college students. The procedure was essentially same—five steps and all—with an added twist. In this second iteration of the experiment, the students (pretend bankers) were told that the intel on the trustworthiness of the faces they were about to see was solid. They had run the experiment before and these people proved themselves to be either honest or dubious borrowers. Basically, if the faces looked trustworthy, they probably were. If they didn’t, they probably weren’t. This gave the pretend bankers an additional reason to decouple labels after they were again informed that the criterion used for determining trustworthiness was broken.
Analyses of the results were the same as before: the participants still showed a bias for remembering faces that were allegedly untrustworthy.
You would think at this point they would hang up their lab coats (assuming they had lab coats), pat themselves on the back, and call it a day. You would think they would be satisfied at this point. No, these psychologists are men of scrupulous academic standards. They initiated a third iteration of the experiment to see if they would get the same results by instead asking the participants to make investment risks as opposed to deciding who was and who wasn’t trustworthy.
They changed the rules around some in this trial. For one, they incentivized the third set of college students by offering them real money if their investment decisions performed beyond a quota. Given that investment options for lending and receiving was literally in hundreds of yen, they had plenty of reason to take themselves seriously as bankers. Another change was the introduction of neutral faces in the experiment. Now, in addition to having a stock of faces of people who were either definitively good borrowers or definitively bad borrowers, there was a “neutral” batch of faces that were a mix of good and bad borrowers unknown to the pretend bankers.
The results revealed, yet again, that people hold a memory bias for the discredited. Even the neutral faces that the participants initially perceived to untrustworthy were remembered as untrustworthy at the end of the experience after the fact.
So where does this leave us? We’ve learned through this experiment that it is more instructive to pay attention to the persistence memory bias rather than how one comes to learn someone else is untrustworthy. In each case the pretend bankers were explicitly told that the information they were given was bad. Yet they still held on to those notions despite the fact that they are aware that basis for their prejudice was wrong. We can infer from this finding that stigmas cannot simply be disregarded because individuals don’t necessarily remember the context under which they learned someone was discredited (Suzuki, Honma, and Suga, 2013). All you need to hear from a credible source is that someone is a liar, a thief, or a cheat, etc. How you heard it or why doesn’t stick; you just remember the main idea: such and such person is insert stigma.
What we don’t know after exploring this experiment is whether it is the label (untrustworthy, cheat, etc.) or the negative undertones of the label that makes it difficult for one to forget. It is one thing to think of someone as a rule-breaker and quite another to view rule-breaking as immoral. You don’t have to look far within our culture to find stories that laud mavericks that broke the rules in the effort to accomplish a great feat. George Washington, the father of our nation, was certainly not considered a hero by the British Crown in his time. In fact, he would have probably been considered a terrorist, to use contemporary parlance. Dr. Martin Luther King, labeled a communist, argued against systemic racial segregation and was assassinated for it. None of these two figures carry a negative bias, but they are the epitome of rule-breaking. However, what we do know is that this particular distinction in regards to persistence memory warrants continued empirical studying. And at the end of the study, our scrupulous academics reflect on a dark truth about the human experience: “Thus, sadly, repeated lies might become accepted truths, especially if they defame others.” Perhaps next time you encounter someone you don’t trust, it may be worthwhile to ask yourself why.
References
Suzuki, A., Honma, Y., & Suga, S. Indelible Distrust: Memory Bias Toward Cheaters Revealed as High Persistence Against Extinction. Journal of Experimental Psychology: Learning, Memory, and Cognition, 39, 1901-1913.
This is a great post. I really make me think about how rumors affect the image of people. Even when they are proven to be untrue, people do continue to associate an individual with that rumor. I think that the more you hear the rumor, the more likely you are not going to forget the rumor. I think the more memory traces there are that connect the individual with the rumor the more likely, you won’t be able to let it go. The rumor in this experiment was the investment success of the people in the pictures and this experiment really does a great job in showing this issue within the memory system. I will definitely ask myself why i don’t trust that person next time.
I think it would be very interesting to do this same experiment to explore the own age Bias and Rase bias. Are the “bankers”more likely to keep the negative label on someone that is of a different age or race. Would you be more likely to lend money to someone of your own race/ age. This could have major implications into everyday society.
I found the experiment very interesting. Its almost illogical how much the memory bias stuck trial after trial. There was definitely some proactive interference occurring as well, because once the participants learned which faces belonged to bad borrowers and good borrowers, that knowledge interfered with any new information offered. The benefit of this is most probably efficiency, like the persistence error of comission. Labeling is faster and easier than constantly reevaluating and analyzing everyone you know. I also like the previous comment that suggests an experiment focusing on age and race, because that seems to be one of the implications of the experiment that was never fully explored.
This experiment really backs up the saying that first impressions matter. It not only can it affect the beginnings of a relationship, but it will affect later parts of a relationship being that its hard to erase the impression especially when it is bad. Though this experiment seems to address memory, I wonder how attention plays into this unbudging judgement of cheaters.
Since the in the first phase of the experiment requires participants to judge a person’s trustworthiness by intuition after looking at faces, I wonder if there are facial features that are seen as more trustworthy or less trustworthy. If so, when told that certain faces belonged to either trustworthy people or cheaters, participants might have attended more to the features that were consistent with their label. In the case of attenuation, context, which in this case is the label of trustworthy or not, plays an important part as to what passes the filter into the perceptual channel. In the case for trustworthy people, “trustworthy” facial features may have been let into the perceptual channels while the “untrustworthy” facial features are ignored. Once this has been done, looking for these features in these faces has already been practiced making it more “automatic.” It would have been easier to be attended to the same features again even when told that the context was incorrect. This would also explain why neutral faces that were deemed as untrustworthy the first round stuck with that title.