cognitive bias in psychology — 1 investigation

cognitive bias in psychology is treated, in the textbook, as a malfunction. in the productivity bro tweet, it is a feature. i looked into both. the textbook is longer. the tweet has more likes. one of these things is winning.

it is friday, 11:47am, and i am at the desk on the second floor with one airpod in and the other one quietly aging on the corner of the keyboard. carla is upstairs in the annual planning meeting, on the third floor, where the slides are honest and the snacks are not. i have approximately 47 minutes before the elevator brings her back down with that look she gets after two coffees and a graph she did not ask for.

so i am going to use these 47 minutes to do what i call a single-investigation. one term. four sources. zero papers. the focus, just so we are clear, is cognitive bias as the psychology textbook tries to explain it, contrasted with how the rest of us use the phrase, which is mostly as a polite word for “i was right and other people were tired”.

cognitive bias in psychology, brief

the short version of cognitive bias in psychology, the one i can hold in one airpod’s worth of attention, is this: the brain prefers a quick answer over a correct answer. then it dresses up the quick answer in a suit and calls it a conclusion. then it goes to the meeting and presents the conclusion as if there had been a process. there had not been a process. there had been a feeling and a stapler.

i am not making this up entirely. the manual they reference on the shows i watch lists dozens of these biases. anchoring. availability. survivorship. the one where you only remember the bus when it was late. there is a name for the one where you only remember the bus when it was late, and i am fairly sure it is not “the bus thing”, although that is what i call it at the bar.

the textbook treats cognitive bias as a leak in the machinery. the productivity bro on a podcast i would not subscribe to treats it as a hack: “use anchoring to win negotiations” — which is just admitting you plan to confuse people for money. the academic frames it as a defect. the bro frames it as a feature. both are correct. only one of them sells a journal.

carla seems to know the textbook version

carla, when she came past my desk last thursday with a coffee and a clipboard, said the phrase “cognitive bias in psychology” out loud, like a person who had read the actual chapter and not just the part the algorithm fed her. she said it the way a doctor says a two-syllable word: with no theatre.

i nodded as if i had also read the chapter. i had not read the chapter. i had skimmed a list. there is a difference, and carla, to her credit, did not make me prove it. she does this thing where she lets me off the hook in real time and adds it to a ledger i will never see. one day the ledger will be presented, possibly during a annual planning meeting on the third floor, and i will be unable to argue with the totals.

the textbook version, as carla seems to understand it, treats biases as a sequence of named, observed, repeatable failures of judgment under specific conditions. that is a lot of qualifiers. the qualifiers are the difference between psychology and a tweet. the qualifiers are also why nobody reads the textbook on the bus.

tom has the volvo version, in his head

tom — owns a house, has a wife, two kids, a volvo, and a pension he understands — does not use the phrase cognitive bias. tom uses a different vocabulary. tom says “i had a feeling” and then is correct. i say “i had a feeling” and then need to call dave to fix the feeling. tom and i are not playing the same game.

tom’s version of cognitive bias in psychology is, as far as i can reverse-engineer it from a christmas card, “other people make decisions for the wrong reasons; i do not”. this is itself a textbook bias. there is a name for it. i am not going to look it up. i am going to keep it as evidence for later, when carla brings the clipboard.

the difference is that tom’s model is short, portable, and gets him into a 401k-equivalent. the textbook is long, accurate, and gets you tenure if you live long enough. one of these things is more useful at a barbecue. the other is more useful in a courtroom. you have to pick a venue. tom has picked. i have not.

the chatgpt version, the modern one

i asked chatgpt to define cognitive bias in psychology, on the principle that if a machine that has read most of the internet cannot summarize a 1970s field, no one can. it gave me a clean three-paragraph answer with a list of examples and a polite warning that biases are normal and i should not be too hard on myself. i was not being hard on myself. i was being hard on the field. there is a bias for that, too.

the chatgpt version is the textbook version with the qualifiers shaved off and the moral lesson glued on. it is, in a strange way, the productivity bro version dressed up as the academic version. the bro is loud. the chapter is long. the bot is the middle floor. carla works on the third. tom owns the building. i rent. for a longer take on whether asking a robot makes you smarter, see my notes on becoming a measurably smarter person, which were also written from this desk on a different wednesday.

the airpod that still works listens

here is a thing nobody mentions in the literature i’m fairly sure exists. the way you discover your own cognitive bias in psychology is not by reading. it is by talking to yourself out loud while one airpod plays a podcast and the other one does nothing. the silent airpod is the witness. it has heard everything. it has not been bought.

i listened, this morning, to myself defend a budget number to no one in particular before the meeting i was not invited to. i had, in my head, four arguments. all four arguments led to the same conclusion. the conclusion had been there before the arguments. the arguments arrived to escort the conclusion to its seat. that is the textbook definition of how a confirmation bias forms in real time, and i was running it in the kitchenette while the kettle boiled.

which is the productivity tweet’s whole pitch, by the way: “use confirmation bias to your advantage”. meaning, find the conclusion first, then go shopping for arguments at the place where they sell them cheap. mike, who has not filed since 2019, runs his entire financial life this way. mike thinks tipping should be a flat 12%. mike has done the math. mike has not done the math. mike has done a feeling and called it the math. it is consistent. it is also, on a long enough timeline, why mike pays more for things. the man at the bar i once watched, played in a film i have on a list, said something similar — see 12 Angry Men on IMDB — although his bias was about a kid and a knife, not a kettle and a tip.

verdict, in psychology the bias is the chapter

so here is what i have, from the second-floor desk, with one airpod and 23% phone battery and a annual planning meeting running upstairs without me. cognitive bias in psychology is not a flaw the smart people identified to embarrass the rest of us. it is the operating system. the textbook describes it in a thousand pages. the tweet describes it in nine words. chatgpt describes it in three paragraphs with a kindness rider. tom does not describe it at all and is doing better than all of us.

the bias is not the bug. the chapter is the bug. or more accurately, the idea that the chapter ends is the bug. the chapter does not end. the chapter is the chapter and you are also the chapter and that is the joke psychology has been politely trying to tell us since the 1970s — that the people writing the textbook are the same people the textbook is about. you cannot get out of it by reading more.

this connects, in case you were wondering whether cognitive bias is just stupid by another name, to the larger investigation we are running here, which is whether stupid is a fixed quantity in a person or a renewable resource you draw from depending on the day. on a friday, with one airpod, i am going to go with renewable. it would be a bias to say otherwise. i have at least four arguments for this. they all arrived after the conclusion.