Latest / The Psychology Undergrad Podcast / Fundamental Attribution Error: How We Judge and Misjudge Other People
Transcript
- 0:00Welcome to the Psychology undergrad. Yes, welcome.
- 0:03We're really glad you're here. You can just pretend
- 0:05you're basically our third study buddy today
- 0:07because we have a huge exam coming up on person
- 0:10perception. Oh, yeah. We have a ton of material
- 0:12to get through for this deep dive. We really
- 0:14do. I want to start us off with something pretty
- 0:17heavy, actually. It's a provocative case study
- 0:20from 1991. Right, the University of Iowa case.
- 0:23Exactly. So in 1991, there was this tragic event.
- 0:27A 28 -year -old graduate physics student named
- 0:30Gang Liu, who had, you know, just lost in academic
- 0:34competition and was really struggling to find
- 0:36a job, he walked into the physics department.
- 0:38Which is just... terrifying to even think about.
- 0:41It is. He shot his advisor, several colleagues
- 0:44and bystanders before taking his own life. It
- 0:46was a completely horrific, devastating event.
- 0:50Just unimaginably awful. Yeah. But what happened
- 0:53next in the media coverage, it actually revealed
- 0:55something profound about how the human brain
- 0:58is wired, like how we assign blame and understand
- 1:01cause and effect. Well, yeah. And that brings
- 1:03us to the research, right? Michael Morris and
- 1:05Kaiping Peng. looked at the aftermath of this
- 1:08in 1994. Right. They analyzed how this exact
- 1:12same event was covered by two completely different
- 1:15newspapers. So they compared the English language
- 1:18New York Times with the Chinese language World
- 1:21Journal. And the contrast was just like night
- 1:24and day, wasn't it? Oh, absolutely. It was absolute
- 1:26night and day. The New York Times painted a picture
- 1:29of a deeply disturbed individual. They interviewed
- 1:32people who claimed there was always this, quote,
- 1:35sinister edge to his character. So they were
- 1:37really leaning into his personality. Exactly.
- 1:40The entire narrative was focused on his internal
- 1:42traits, his psychological makeup. But then the
- 1:46World Journal went in the completely opposite
- 1:48direction. Yeah, they focused almost entirely
- 1:49on the environment, right? They did. They wrote
- 1:51about his really bad relationship with his supervisor.
- 1:54They pointed to the availability of firearms
- 1:56in the United States. Which is a huge situational
- 1:59factor. Totally. They even published an argument
- 2:01suggesting that the tragedy reflected a lack
- 2:05of religion in Chinese culture. Wow. So the American
- 2:08reporters looked at the shooter and just saw
- 2:10a broken person. But the Chinese reporters looked
- 2:13at the exact same guy. and saw a broken situation.
- 2:17Exactly. And, you know, we assume that when we
- 2:19observe an event, we're acting like objective
- 2:22video cameras, just recording reality. Right.
- 2:25Like we're just seeing facts. But we aren't.
- 2:27We're interpreting reality through these incredibly
- 2:29powerful, highly opinionated cultural and biological
- 2:33lenses. Because the brain is like constantly
- 2:36trying to make sense of other people. It literally
- 2:38has to. Yeah. Our survival depends on it. But
- 2:41the mechanisms we use to do this, the ways we
- 2:44form impressions and assign causes to behavior,
- 2:46they operate so far beneath our conscious awareness
- 2:49that we rarely even notice them working. Which
- 2:51is exactly our mission for this deep dive today.
- 2:54We are trying to figure out how our brains perceive
- 2:57others, how we explain extreme behaviors, and
- 3:00where we get it all wrong. And the moral implications
- 3:02of all of that too. Yes. But to really understand
- 3:06how we explain extreme behavior, we have to start
- 3:09at the very beginning. We have to look at the
- 3:12literal fraction of a second when you first lay
- 3:14eyes on someone. Right, the brain's thin slices.
- 3:18Recent functional MRI studies show us exactly
- 3:20what happens in that fraction of a second. It's
- 3:23crazy how fast it is. It's mind blowing. There's
- 3:26a specific network in the prefrontal cortex that
- 3:29just lights up like a Christmas tree when we
- 3:31make judgments about other people. But not for
- 3:33everything, right? No, that's what's so remarkable.
- 3:36This specific neural pathway doesn't activate
- 3:39when we judge an animal like a dog or an inanimate
- 3:43object. We have this dedicated. biologically
- 3:47expensive neural real estate exclusively for
- 3:50perceiving other humans. Which honestly makes
- 3:52sense from an evolutionary standpoint. I mean,
- 3:54interacting with a tree or a dog is a relatively
- 3:56simple calculation. Right, a tree isn't plotting
- 3:58against you. Exactly. But interacting with another
- 4:00human is this complex, high -stakes, two -way
- 4:03street. They are perceiving you while you perceive
- 4:06them. They could deceive you, collaborate with
- 4:08you, manipulate you. Yeah, so the brain just
- 4:11has to figure out their intentions immediately.
- 4:13So how much data does the prefrontal cortex actually
- 4:16need to make that call? Almost none, honestly.
- 4:19We rely on what researchers call thin slices
- 4:22of behavior. Ambedi and Rosenthal demonstrated
- 4:25this brilliantly back in 1993. Oh, is this the
- 4:27teacher evaluation study? Yes. They took videotapes
- 4:30of college instructors teaching their courses.
- 4:33They muted the audio and pulled just three 10
- 4:35-second clips. So 30 seconds total. of complete
- 4:39silence exactly 30 seconds of completely silent
- 4:42video they showed these clips to people who had
- 4:45never met the instructors and asked them to rate
- 4:47the teachers on traits like optimism, confidence,
- 4:51and warmth. OK. And then they compared those
- 4:53snap judgments to the actual official end -of
- 4:56-semester evaluations, right? The ones filled
- 4:58out by students who had been in that teacher's
- 4:59classroom for months. They did. And the 30 -second
- 5:02snap judgments highly correlated with the months
- 5:04-long evaluations. That is wild. Right. The observers
- 5:08accurately predicted how effective the teacher
- 5:10would be without ever hearing them speak a single
- 5:12word. Because the brain is just a master at unpacking
- 5:16those dense packets of nonverbal data. Like posture,
- 5:20micro -expressions, the fluidity of movement.
- 5:23It's all this rich information. It is. But the
- 5:27timeline gets even tighter than 30 seconds. Willis
- 5:30and Todorov found in 2006 that human beings make
- 5:33a solid judgment about a face in 100 milliseconds.
- 5:36Wait. One -tenth of a second. One -tenth of a
- 5:39second. And Todorov followed that up by showing
- 5:41participants the faces of political candidates
- 5:44for just one single second. Okay, I remember
- 5:47this. Yeah, from that one second glance, participants
- 5:50judged to look more competent. And that snap
- 5:53judgment predicted the actual outcome of the
- 5:55election 68 percent of the time. OK, wait, I
- 5:58have to push back here. I really struggle with
- 6:00this idea. Are you saying voters are literally
- 6:02just picking a face in a tenth of a second and
- 6:04that dictates an entire election? Because what
- 6:07about the millions spent on advertising? What
- 6:09about incumbency or policy platforms or, you
- 6:12know, debate performances? It just feels way
- 6:14too reductive to say it comes down to a one second
- 6:16glance. I know, I know. It sounds totally irrational
- 6:19if you view voting as a purely logical exercise.
- 6:22But the point isn't that voters only use faces
- 6:24and ignore everything else. Okay, then what's
- 6:27the mechanism? The mechanism is that the brain
- 6:30establishes a baseline level of trust instantly.
- 6:33It's like a biological veto. If a candidate's
- 6:36face registers in that first tenth of a second
- 6:39is untrustworthy or incompetent, every single
- 6:42policy speech they give is fighting a massive
- 6:45uphill battle. Oh, I see. So it's not just an
- 6:47aesthetic choice. No, we're running an ancient
- 6:50threat detection algorithm. OK, that brings up
- 6:52a great way to think about this, actually, to
- 6:54help you remember this for the exam. Survival
- 6:57over smiles. Ooh. I like that. Yeah, we aren't
- 7:01looking for the prettiest face. We're looking
- 7:02for the most dangerous one. Hanson and Hanson
- 7:05proved this with their faces in the crowd study
- 7:07in 1988. Oh, the angry face study. Yes. So they
- 7:10showed participants pictures of crowds made up
- 7:12of nine faces. Sometimes it was eight happy faces
- 7:15and one angry face. Other times it was eight
- 7:18angry faces and one happy face. And the results
- 7:21are so telling. Totally. The participants found
- 7:24the single angry face hidden among the happy
- 7:26ones significantly faster than they found the
- 7:28happy one hidden among the angry ones. The threatening
- 7:31face basically just leaps right off the screen
- 7:33at you. Exactly. Because positive information
- 7:35is nice, but negative information can literally
- 7:38kill you. The brain prioritizes the angry face
- 7:42because, well... Failing to spot a threat is
- 7:45an evolutionary dead end. The revival over smiles.
- 7:48And our threat detection software is so robust,
- 7:51it doesn't even require a face to operate. Wait,
- 7:53really? Yeah, think about point light display
- 7:55studies. Researchers attached tiny lights to
- 7:58the joints of an actor, just their wrists, elbows,
- 8:01knees, and ankles. Oh, right. And they put them
- 8:03in a pitch black room. Exactly. The actor moves
- 8:06around in the dark. So all the observer sees
- 8:08is this constellation of dots moving. No face,
- 8:11no skin, no clothing. Just floating lights? Just
- 8:14floating lights. Yet from just the kinematics,
- 8:16the understating physics of how the dots move
- 8:18in relation to one another, observers can instantly
- 8:21recognize what the person is doing. And they
- 8:23can even read their emotional state, right? Like
- 8:25if the dots are angry or sad? Yes. They can tell
- 8:29if the dots are happy, sad, angry. It proves
- 8:31that human motion carries an intrinsic emotional
- 8:34signature. We don't need the whole picture. We
- 8:37just need the skeletal geometry of intent. Wow.
- 8:41Okay, so our brains lock in a judgment in a tenth
- 8:43of a second using everything from facial structure
- 8:46to the geometry of movement. But what happens
- 8:49next? Well, we start getting more data. Right,
- 8:52like once we get past that initial millisecond,
- 8:54we start gathering actual facts about a person.
- 8:56We learn they're smart or funny or maybe, I don't
- 8:59know, a little impatient. How does the brain
- 9:01combine this messy list of personality traits
- 9:04into a single cohesive impression? The brain
- 9:08uses cognitive algebra, but it doesn't do it
- 9:12the way we might naturally assume. Right. Here's
- 9:14a major study hack for you. We are average calculators,
- 9:17not addition machines. Exactly. When we want
- 9:20someone to like us, say, on a first date or in
- 9:22a job interview, our instinct is to list every
- 9:25single positive thing about ourselves. We think
- 9:28we're addition machines. We assume a great trait
- 9:30plus a good trait equals a higher total score.
- 9:34But the reality is, as you said, we're average
- 9:36calculators. If I tell you I'm highly intelligent
- 9:39and incredibly loyal, those are high value traits.
- 9:42Let's say they're both a 10 out of 10. Okay,
- 9:44so your average is a 10. Right. If I stop talking,
- 9:46my average score is a 10. But if I keep talking
- 9:49and add, you know, and I'm also fairly organized
- 9:52and I make an okay cup of coffee, those are maybe
- 9:54five out of 10 traits. You haven't added points
- 9:56to your score? No, I've diluted my average. I
- 9:58just dragged my 10 down to a seven and a half.
- 10:01Wow. So I should really stop talking about my
- 10:03okay coffee on dates. Probably. Mills demonstrated
- 10:06this averaging effect in 2007. The mechanism
- 10:10behind it comes down to cognitive load. The brain
- 10:12just doesn't have the bandwidth, right? Exactly.
- 10:15It can't maintain a running tally of dozens of
- 10:17individual data points about every single person
- 10:19you meet. Instead, it compresses the data into
- 10:23a single gestalt, a holistic feeling about the
- 10:28person. So when you introduce moderate information,
- 10:31it waters down the intensity of that overall
- 10:33feeling. If you want to leave a strong impression,
- 10:36you only present your most extreme positive traits.
- 10:39Yep. But we have to mention, not all traits carry
- 10:42the same mathematical weight in that average.
- 10:44Oh, this is the warm and cold thing. Some traits
- 10:47completely rig the scale. They really do. Solomon
- 10:50Ash found this in 1946, and then Harold Kelly
- 10:53expanded on it in 1950. They gave college students
- 10:56a written bio of a guest professor, right, before
- 10:59he came to teach. Yes. For half the students,
- 11:02the bio listed traits like industrious, critical,
- 11:04practical, determined, and warm. But for the
- 11:07other half, The bio was identical, word for word,
- 11:10except warm was replaced with cold. Just one
- 11:13word change. Just one word. And it fundamentally
- 11:16altered reality for the students. When the professor
- 11:18came in and led a 20 -minute discussion, he behaved
- 11:21exactly the same way for everyone. But the students
- 11:23reacted completely differently. Totally. The
- 11:26students who had read the word warm participated
- 11:28heavily. Afterward, they rated him as sociable,
- 11:31good -natured, and genuinely funny. And the students
- 11:34who read cold. They barely participated at all,
- 11:38and they rated him as aloof and humorless. Because
- 11:41warm and cold aren't just descriptions, they
- 11:43are central traits. They signal intent. Like
- 11:46if someone is industrious and determined, that
- 11:48just tells you they are capable. Right, but capable
- 11:51of what? Exactly. Warm tells you they will use
- 11:54those capabilities to help you. Cold tells you
- 11:57they might use those capabilities against you.
- 11:59The brain grabs onto central traits and uses
- 12:01them as a filter for everything else. That filtering
- 12:04process relies heavily on the primacy effect.
- 12:06The very first information we receive about someone
- 12:10creates an anchor. The halo starts with hello.
- 12:13I love that. Yeah, it's a good mnemonic. If your
- 12:16first impression is a positive halo, say, you
- 12:19think someone is warm and brilliant and later
- 12:21they do something rude, you don't adjust your
- 12:23average. No, a psychological process called assimilation
- 12:26takes over. We actively twist and mold any new
- 12:29information to fit the narrative we've already
- 12:31constructed. You literally tell yourself, oh,
- 12:33they aren't being rude. They're just deeply passionate
- 12:36and uncompromising. You change the definition
- 12:39of the behavior to protect the halo. Which brings
- 12:42us to arguably the most complex part of social
- 12:46perception. Yeah. We've talked about how we judge
- 12:48static faces and how we average personality traits.
- 12:51But humans aren't static. We act. We make choices.
- 12:54Right. And when we see someone do something,
- 12:57especially something unexpected, our brains have
- 12:59to immediately figure out why. Which is causal
- 13:02attribution. Exactly. Are they doing this because
- 13:05of who they are internally, their personality?
- 13:08Or are they doing it because of the external
- 13:10situation they are in? When we're trying to solve
- 13:12that puzzle, our brains basically run an internal
- 13:15algorithm based on the covariation principle.
- 13:17We act like detectives looking for patterns.
- 13:20And you have a good way to remember this one
- 13:21too, don't you? I do! For the exam, think of
- 13:24the CDC of crying. The CDC of crying? Okay, walk
- 13:27us through it. Imagine you're at a wedding and
- 13:29you see someone just bawling their eyes out.
- 13:31To figure out if they're crying because they're
- 13:33just an overly emotional person, or if they're
- 13:35crying because the wedding is incredibly moving,
- 13:38your brain checks three variables. Consistency,
- 13:42distinctiveness, and consensus. The CDC. Exactly.
- 13:45So let's break down the algorithm. Consistency
- 13:48asks, does this person always act this way in
- 13:51this specific situation? Like, do they cry at
- 13:54every single wedding they attend? If yes, consistency
- 13:57is high. Right. Next is distinctiveness. Is this
- 14:00behavior unique to this situation? Do they only
- 14:04cry at weddings or do they cry at grocery stores
- 14:06during car commercials and when they drop their
- 14:08keys? Right. If they only cry at weddings, distinctiveness
- 14:11is high. And finally, consensus. Is everyone
- 14:14else at the wedding crying or is this person
- 14:15the only one in tears? If the whole room is sobbing,
- 14:18consensus is high. So, when the CDC is all high,
- 14:21they always cry at weddings, they only cry at
- 14:23weddings, and everyone else is crying right now,
- 14:25the brain instantly calculates a situational
- 14:28attribution. We conclude the environment caused
- 14:30the behavior. It's just a really sad wedding.
- 14:33But flip the variables. Imagine consistency is
- 14:36high, they always cry at weddings. But distinctiveness
- 14:40is low. They cry at everything. All the time.
- 14:43And consensus is low. Literally nobody else at
- 14:46the wedding is crying. Right. When the brain
- 14:48sees that pattern, it makes a personal attribution.
- 14:52It decides the environment has nothing to do
- 14:53with it. The cause is their internal personality.
- 14:57They're just an unstable, highly emotional person.
- 15:00We run this CDC covariation math constantly,
- 15:04like completely subconsciously, every time we
- 15:06watch someone interact with the world. Which
- 15:07would be amazing if the algorithm actually worked
- 15:10perfectly. But it doesn't. We suffer from massive
- 15:12systemic processing errors. The most famous one
- 15:15is the fundamental attribution error. Oh, this
- 15:17is a big one. It is. When we analyze other people,
- 15:20we overwhelmingly ignore the situation and blame
- 15:23their personality. Like if a guy cuts you off
- 15:25on the highway. Your brain doesn't run a nuanced
- 15:27CDC analysis to see if he has a medical emergency.
- 15:31No, you instantly conclude he's a selfish, reckless
- 15:34person. We underweight the situation for strangers
- 15:37because the situation is invisible to us. We
- 15:39can't see their stress or the bad news they just
- 15:41got. All we see is the person, so the person
- 15:44becomes the cause. Exactly. But when we look
- 15:46at our own behavior, the perspective shifts completely.
- 15:50This is the actor -observer bias. Oh, right.
- 15:53When I cut someone off in traffic, I know I'm
- 15:55not a bad person. I know I'm just late for a
- 15:57crucial meeting and the sun was in my eyes. Yep.
- 16:00You give yourself a full situational pardon while
- 16:03acting as a ruthless judge of everyone else's
- 16:05character. Okay, but I have to ask. Why are we
- 16:08so eager to give ourselves a pass, but so rigidly
- 16:11punitive toward others? It feels like it goes
- 16:12beyond just an innocent math error. Well, there
- 16:15is a massive moral weight to it, and it stems
- 16:18from a deep existential vulnerability. What do
- 16:20you mean? So in 1965, a psychologist named Melvin
- 16:24Lerner identified something called the Just World
- 16:26Hypothesis. Lerner argued that human beings have
- 16:30this desperate fundamental psychological need
- 16:32to believe that the world is a fair place. We
- 16:35need to believe that good things happen to good
- 16:37people, and bad things happen to bad people.
- 16:40That people get exactly what they deserve. Because
- 16:42the alternative is just terrifying. Right. If
- 16:46the universe is chaotic and random, then bad
- 16:48things can happen to good people for absolutely
- 16:51no reason. You could do everything right, follow
- 16:54all the rules, and still lose your job or your
- 16:56health or your life. A random universe means
- 17:00you're never truly safe. And the terror of randomness
- 17:03is just unbearable for the human psyche. It really
- 17:05is. Lerner showed how far we will go to protect
- 17:08our belief in a fair world with this brilliant
- 17:11anagram study. You know the coin flip one? Yes.
- 17:14He had participants watch two people working
- 17:16on puzzles. The observers were clearly informed
- 17:19beforehand that one of the workers was going
- 17:21to be selected entirely by random chance, literally
- 17:24a coin flip, to receive a large financial reward
- 17:27while the other would get nothing. Pure unadulterated
- 17:30luck. The observers knew it was random before
- 17:32it even happened. They watched the coin flip
- 17:34themselves. But after the money was awarded to
- 17:37the random winner, something bizarre happened.
- 17:40The observers cognitively rewrote history. They
- 17:43actually convinced themselves that the person
- 17:45who won the money had somehow earned it, right?
- 17:46Yes. They evaluated the randomly rewarded person's
- 17:50puzzle -solving performance as superior. They
- 17:53literally hallucinated competence to justify
- 17:57a random outcome. They needed the outcome to
- 18:00make logical, moral sense. That is wild. And
- 18:04this actually circles all the way back to our
- 18:05opening case study about the shooting at the
- 18:08University of Iowa. Right, the quote from the
- 18:09World Journal. Yeah, that quote suggesting a
- 18:12lack of religion. Religion often provides a formal
- 18:15framework for the just world hypothesis. It promises
- 18:17that the scales of justice will eventually balance,
- 18:20whether through karma, divine intervention, or
- 18:23an afterlife. But even without religion, the
- 18:26human brain creates its own secular karma. It
- 18:29insists the world is fair. And that insistence
- 18:31has a deeply dark side. The just -world hypothesis
- 18:35is basically the psychological engine behind
- 18:37victim blaming. Absolutely. When people encounter
- 18:40someone experiencing severe suffering, perhaps
- 18:42someone who is homeless or struggling with addiction
- 18:45or a victim of a crime, the observer's brain
- 18:48goes into panic mode. Because if the observer
- 18:50admits that the suffering is the result of a
- 18:52tragic situation or systemic failure or just
- 18:55bad luck, it shatters their illusion of safety.
- 18:59It means the observer themselves could become
- 19:01homeless tomorrow through no fault of their own.
- 19:03So to protect their own sanity, they make a personal
- 19:06attribution. They look at the person suffering
- 19:08and say, you must have done something to deserve
- 19:10this. You must be lazy or irresponsible. By blaming
- 19:13the victim's character, the observer neutralizes
- 19:16the threat. The world remains just and the observer
- 19:19remains safe. But at the cost of entirely stripping
- 19:22away their empathy for the person suffering,
- 19:25we preserve our own psychological comfort by
- 19:27stigmatizing the vulnerable. It's a sobering
- 19:30realization. We judge faces in a tenth of a second
- 19:33to detect threats. We average traits and use
- 19:36central anchors like warmth to build these heavily
- 19:38biased initial impressions. We blindly guess
- 19:41at the causes of behavior heavily favoring personality
- 19:44over situation. And we convince ourselves the
- 19:47universe is perfectly fair, even when it forces
- 19:49us to blame the innocent. We are fundamentally
- 19:52flawed observers. But you know, there is a silver
- 19:56lining for us study buddies today. Not everyone
- 19:59operates entirely on autopilot. No, they don't.
- 20:02People who possess a trait psychologists call
- 20:04a high need for cognition. Meaning people who
- 20:07actually enjoy the effortful process of thinking
- 20:09deeply and analyzing nuance. Right. Those people
- 20:13tend to overcome these biases. They're much more
- 20:15likely to pause, look past the initial face,
- 20:18evaluate the actual context of a situation, and
- 20:21make tolerant situational attributions rather
- 20:24than knee -jerk punitive ones. They recognize
- 20:26that the brain's first instinct is often a shortcut,
- 20:29not the truth. They force the prefrontal cortex
- 20:32to do the hard work of actual analysis. Which
- 20:34is exactly what you are doing by exploring these
- 20:36concepts with us on this deep dive. Exactly.
- 20:39But before we wrap up our session today, there
- 20:40is one final thought we should all be chewing
- 20:42on. Everything we've discussed today, the 0 .1
- 20:45second judgments, the thin slices of behavior,
- 20:47the primacy effect. It all evolved over millions
- 20:51of years in the physical world. Right, face to
- 20:53face. But today, we spend a massive portion of
- 20:56our lives navigating the digital world. Social
- 20:59media is essentially a factory for producing
- 21:02artificial, highly weaponized thin slices. Oh,
- 21:05wow, that is so true. A curated Instagram feed,
- 21:08a perfectly polished bio, a strategically edited
- 21:11video. These aren't natural nonverbal behaviors.
- 21:14They are engineered data packets designed to
- 21:17trigger your cognitive algebra. So if our prefrontal
- 21:20cortex is evolutionarily hardwired to lock in
- 21:23an unshakable impression in a tenth of a second,
- 21:25we know we blindly assimilate new data to protect
- 21:28that initial halo. Are we biologically defenseless
- 21:31against a well -crafted digital avatar? As a
- 21:34terrifying question. The next time you instantly
- 21:36feel a deep sense of trust or a visceral dislike
- 21:38for someone online, stop and ask yourself, are
- 21:40you actually seeing them or is your ancient brain
- 21:43just being hacked by their perfectly calculated
- 21:45halo?