HPC · Science Deep Dive 18 June 2026 · revised 15 October 2026

First Principles Thinking: What the Science Shows About Checking Your First Answer.

Your brain reaches for a familiar answer before it reasons. Brain scans show which areas do the work when people check that answer, and training has improved some of these skills on tests. Here is what the science actually says, and what to do with it.

Section: Decisions · Mental Models
Reading time: 22 min read
Sources: 31 · reviewed

01The Override Problem

Smart Brains Default to Pattern, Not Analysis

First principles thinking means setting aside the answer that comes to mind first and rebuilding one from the basic facts. In this article, the term names a family of skills. You notice that a quick answer may not fit, hold it back, and work out how the facts relate. No study has tested first principles thinking by that name; the studies below tested parts of it, using reasoning puzzles, a short test of reflection, bias tasks, logic training, news headlines and argument mapping.

Why do smart people still reason badly? Rational thinking and intelligence are related, but research that measures both finds they are separable.^[2] Intelligence does not protect against some biases. One of them is myside bias, the habit of judging evidence in favour of what you already believe. It shows almost no link to IQ.^[13] So the trouble is often not a lack of brainpower. It is which kind of thinking the brain is running.

Psychologists describe two kinds of thinking, which this article calls fast thinking and slow thinking. Fast thinking is automatic and effortless.^[1] Researchers call it Type 1 processing: answers reached through pattern-matching, memory and mental shortcuts.^[3] This article's approach draws on slow thinking, or Type 2 processing: deliberate thought that uses working memory to consider alternatives.^[3]

The history

The stakes are real. One review looked at diagnostic errors at a few large U.S. academic hospitals. Thinking errors played a part in about three in four of the cases, among them bias, mental shortcuts and premature closure (settling on a diagnosis too early).^[18] The review counted contributing factors in errors already found, so it does not show how often doctors make mistakes.

What does the brain do when it overrides a gut answer? This article groups the brain findings into three steps: noticing a conflict, holding back the quick answer, and combining the facts into a new one. A pooled analysis of dozens of brain-imaging studies looked at two kinds of reasoning: generalising from examples (induction) and drawing a conclusion that must follow (deduction). Both drew on the prefrontal cortex, at the front of the brain, and on areas toward the top and back.^[6] This is the frontoparietal network, which this article calls the reasoning network. The pattern varied by task: geometric tasks showed activity at the front alone, and induction drew on extra areas.^[6]

Can the skill be trained? Some parts can, as measured on specific tests. In one small adult study, people trained to hold back a misleading shortcut on a logic task, a skill called inhibitory control, and brain activity then shifted toward the front of the brain.^[7] Reasoning training has also raised test scores in studies the Protocol section describes after its steps.

How do you use this at your next decision? The Protocol section proposes four steps: list your assumptions, map how the facts relate, look for counterexamples, and keep updating your view as evidence comes in. The first three cover the three brain steps; the fourth keeps your answer open to new evidence.

02The Mechanism

The Override Circuit: How the Prefrontal Cortex Reasons from First Principles

A common account of thinking starts with rapid, automatic responses. What does this look like? What happened last time?^[1] Deliberate thought can check or revise those first answers.^[3] Researchers call the pull toward the quick answer cognitive miserliness: a strong tendency to rely on fast thinking, which reaches answers by pattern rather than by analysis.^[3]

First principles thinking takes a different route. The brain must notice the quick answer and judge whether it fits this problem, and if it does not, hold it back in favour of slower work. Part of that work is cognitive decoupling. You keep the real situation fixed in working memory, the mind's short-term workspace, while you test a "what if" version beside it.^[3] The pooled analysis described earlier tied reasoning of this kind to frontal and parietal areas, in a pattern that varied by task.^[6]

That matters because this article treats the override as a sequence, not one event. In that model, each step can fail on its own, which would produce different kinds of error.

An editorial sketch of checking and rebuilding an answer; the studies did not test this anatomical sequence as a whole.

Diagram · HPC

A computer model of the prefrontal cortex, tested against brain scans, gives one account of how people weigh a strategy against others.^[5] In that brain-imaging study of a changing choice task, the model described two interacting tracks. A midline (medial) track, which includes the dorsal anterior cingulate cortex, estimated how reliable the current strategy was. A side (lateral) track, in areas at the very front and the sides of the prefrontal cortex, estimated how reliable the alternatives were.^[5]

This article compares the side track's work to the last of its three steps, building a new answer. That comparison is an editorial analogy; the study did not test it.

A second pooled analysis of imaging studies found activity in regions also linked to cognitive control, the steering of thought toward a goal, including the anterior cingulate and the insula.^[10] It did not separate fast thinking from slow thinking, so it cannot show when a check starts.^[10]

The third step is building the new answer. The area most often linked to that work is the rostrolateral prefrontal cortex, the front tip of the brain, also known as Brodmann Area 10. In a brain-imaging study of reasoning puzzles, it was active during relational integration. That means holding several abstract relations in mind at once and drawing a conclusion from how they fit together, not from how they look.^[8] Puzzles that required combining two relations drove more activity there than one-relation puzzles, even when the two kinds were matched for speed and accuracy.^[8]

A later study added a detail. During abstract reasoning tasks, areas of the reasoning network grouped together differently than they did at rest. Areas that regrouped more also showed larger changes in brain activity during the task.^[9]

Studies of working scientists show a parallel in behaviour. In observations at lab meetings, scientists often used analogies based on structure: how the parts of a problem relate, not how the problem looks.^[17] Experiments found that instructions and the complexity of the scenario affected whether people used structure or surface features.^[17] These were studies of behaviour, not brain scans.

In the reviewed cases, the trouble was rarely a lack of medical knowledge. The review often found that doctors put the facts together badly, for example by settling on a diagnosis too soon.^[18] Related biases include the pull of a recent similar case and anchoring bias, sticking with a first impression.^[18]^[19] Thinking and system factors often overlapped.^[18] A case review cannot show what happened in the doctors' brains.

Strong intuitions can trip up experts too. In a brain-imaging study, physicists with doctorates judged true and false statements about science. Statements that went against intuitive ideas drew more errors, slower answers and more activity in frontal areas than comparable statements that did not.^[30] The authors read this as a possible sign of extra cognitive control.^[30]

That matters because cognitive reflection, the habit of questioning the first answer that comes to mind, can be measured. The Cognitive Reflection Test poses three short puzzles, each with a tempting wrong answer. The test is designed to measure a tendency to question that first response.^[11]

03Evidence

The Five Strongest Studies on First Principles Reasoning

01The claim

The single load-bearing finding

The hero study finds 47 studies.

Studies differ in how much weight they can bear. This article scored the five below on six criteria: Design, Sample, Rigour, Causality, Replication and Citations. The scores and ranks are its editorial judgement, and the note on each study also says how closely it bears on this article's question. Each study tested one part of the family of skills described at the start, such as a reasoning puzzle or a news-judging task. A high score rates the study's design, not the size or reach of its finding.

Key study result

47 studies

02How we measured

Grading the reasoning studies

Studies scored on design, sample, rigour, causality, replication, citations.

For reasoning science, causality is the hard criterion: brain scans show where activity occurs, not what causes what, so training studies carry extra weight.

Rubric weights

Design/30

Sample/20

Rigour/15

Causality/15

Replication/10

Citations/10

03The spread

Heterogeneity across 5 studies

Methodological quality across the ranked studies.

The five studies agree from different directions, and each has its own limit. Brain scans show where reasoning happens, not what causes what. Tests tell rational thinking apart from intelligence.^[2] In one study, Cognitive Reflection Test scores predicted performance on bias tasks better than intelligence, working memory or thinking dispositions did on their own.^[14] The news study shows a link, not a cause. Training studies measured gains on specific reasoning tests, including tasks close to those practised; they do not show better everyday decisions.^[7]^[25]^[27]

Rubric spread

→ 76 /100

Highest to lowest rubric score across the ranked studies.

04Beyond the top five

Wider evidence

What the wider evidence adds.

The Cognitive Reflection Test is one measure relevant to the skills this article groups as first principles thinking.^[11] A separate recent study, not one of the five, found that the test predicted some judgements beyond a matched maths-aptitude test.^[31] Neither study tested first principles thinking by that name. Together they show that a tendency to question a tempting first answer can be measured. They do not show that the whole family of skills can be, or that a high score means better real decisions.

The studies

5 studies, ranked.

Below: the anchor study in full; then every study's rubric score; then the supporting studies in ranked order.

01Anchor

The Neural Correlates of Relational Reasoning: A Meta-analysis of 47 Functional Magnetic Resonance Studies

Wertheim & Ragni 2018 Meta-Analysis · Neuroimaging · Convergent Design

This meta-analysis pooled brain-imaging studies of relational reasoning. Induction and deduction both drew on frontal and parietal areas. The pattern varied by task: geometric tasks showed prefrontal activity alone, and induction (generalising from examples) drew on extra areas.

Rubric breakdown

Design27/30

Sample17/20

Rigour13/15

Causality10/15

Replication10/10

Citations10/10

Total 87/100

The strongest studies, ranked by methodological weight.

Each scored 0 to 100 against a six-criterion rubric, tagged by design and year; the anchor leads. No study in this set reaches the rubric-90 tier.

Anchor (Rank 1) Supporting

Rank Authors & title Journal · Year Finding Score

Pennycook & Rand

Lazy, not biased: Susceptibility to partisan fake news is better explained by lack of reasoning than by motivated reasoning

2019

In two online studies of paid U.S. volunteers, one with pre-registered analyses, higher scores on the Cognitive Reflection Test were linked to better skill at telling real news headlines from fake ones. The link held for headlines that matched people's own politics, which counts against the idea that people believe fake news mainly because they want to.

84/100

Donoso, Collins & Koechlin

Foundations of human reasoning in the prefrontal cortex

2014

In a brain-imaging study of a changing choice task, a model of the prefrontal cortex described two interacting tracks. A midline track estimated how reliable the current strategy was; a side track estimated how reliable the alternatives were. This article's analogy, which the study did not test, links the side track to the rebuilding step of first principles thinking.

82/100

Houdé & Borst

Evidence for an inhibitory-control theory of the reasoning brain

2015

The review describes a small adult study. After training to resist a misleading shortcut on a logic task, brain scans showed activity shifting from areas toward the back of the brain to frontal areas. Separate studies in children linked success on a number task to activity in frontal and parietal areas.

79/100

Stanovich, West & Toplak

The Rationality Quotient: Toward a Test of Rational Thinking

2016

The Comprehensive Assessment of Rational Thinking (CART) has 20 subtests and was tested on a large sample. It found that intelligence and rational thinking are related but separable. The gap is widest for myside bias, which shows almost no link to IQ.

76/100

04Stakes

The cost of reasoning on autopilot

Thinking errors show up in medicine, money, management and judging the news, though each study measured them in its own way.

01 System 01

Clinical Medicine

The case review above found that thinking errors played a part in most of the reviewed diagnostic errors, often alongside system factors.^[18] In that case review, premature closure was the most common thinking error that played a part.^[18] A systematic review linked several biases and risk tolerance to diagnostic inaccuracies in 36.5 to 77% of the case scenarios studied.^[20] Most of the few studies that looked at treatment errors linked them to bias.^[20] Slowing down and self-checking have not consistently improved diagnostic accuracy.^[24]

In practice

You trust the diagnosis you just gave, but you reached it in four seconds and never questioned the first impression

02 System 02

Financial Decisions

On financial decision tasks, the tested biases, such as loss aversion and overconfidence, showed up at similar rates in low-income adults and in adults who grew up disadvantaged but are now better off.^[23] A review describes anchoring, availability and framing as possible sources of error in anaesthesia and intensive care.^[24]

In practice

The investment feels safe because you can easily imagine it succeeding, not because you checked the base rate

System 03

Organisational Strategy

A broad review describes interventions meant to reduce bias in decisions made by organisations.^[22] It sorts these approaches into a framework rather than testing them itself.

In practice

You are solving today's problem with the framework that worked last time, even though the situation is different in kind

04 System 04

Information Evaluation

In two large online studies of paid U.S. volunteers, one with pre-registered analyses, higher reasoning-test scores were linked to better skill at telling real news headlines from fake ones, including headlines that matched their own politics.^[4] In this task, believing false headlines was linked more to not thinking it through than to wanting them to be true. A review of four professions found that biases hurt decisions in healthcare, finance, law and management.^[21]

In practice

The headline feels true because it fits what you already believe, and nothing in your mind raises an alarm

05Protocol

A 4-Step First Principles Reasoning Protocol

The four steps are this article's own synthesis of the studies above: a reasoned approach, not a tested fix. No study has tested them as a set. Each step draws on research into one part of reasoning, and each finding applies to the kind of people that study tested.

The protocol, as a sequence.

Before any solution → After decomposition → After mapping → Ongoing

01 Step 01 · Before any solution

Assumption Decomposition

List every assumption in the problem. Then challenge each one: "Is this actually true, or is it borrowed from past experience?"

Why

One framework distinguishes missing reasoning tools, failure to notice the need to check, and failure to keep checking.^[12] Listing assumptions is this article's proposed application. In a small adult study, training to hold back a misleading shortcut on a logic task went with more activity in frontal areas.^[7]

Common mistake

Confusing a list of constraints with a challenge to them: asking "What are the constraints?" instead of "Which constraints are real, and which are borrowed from a past case?"

02 Step 02 · After decomposition

Structural Mapping

Identify what relates to what, not what resembles what. Draw the causal and structural map of the problem.

Why

In brain imaging, combining several relations at once is linked to the front tip of the brain.^[8] This article treats that combining as the core of first principles thinking. At lab meetings, scientists often drew analogies based on structure, not surface.^[17]

Common mistake

Surface-feature fixation: noting that two things look alike instead of mapping how their parts relate.

03 Step 03 · After mapping

Counterfactual Pressure

Search on purpose for a counterexample. For each claim you assume is true, ask: "What would need to be false for this conclusion to be wrong?"

Why

Searching for counterexamples is this article's own instruction. A review links overcoming misleading shortcuts to inhibitory control, the skill of holding back a quick answer.^[29]

Common mistake

Stopping at the first counterexample instead of collecting several ways the conclusion could fail.

04 Step 04 · Ongoing

Active Updating

Track how confident you are in your predictions over time. When evidence contradicts an expectation, treat it as information, not an anomaly. Also check the base rate: how often similar cases turned out each way.

Why

In a long forecasting project, the most accurate forecasters kept updating their beliefs as news came in.^[16] Actively open-minded thinking, a willingness to weigh evidence against your own view, predicted rational-thinking scores beyond IQ on 17 of the 20 subtests of the Comprehensive Assessment of Rational Thinking (CART).^[15] Checking the base rate, called reference-class forecasting, is this article's own rule.

Common mistake

Defensive thinking: explaining away contrary evidence as "a special case" instead of a sign that your model of the problem needs revising.

Operational logic

Why these four steps? The first three cover the three steps of the brain account, and the fourth adds one after it. Listing assumptions helps you notice a conflict. Mapping and counterexamples help you hold back the shortcut and build a new answer. Updating keeps that answer open to evidence. The steps are simple on purpose, with no special tools. In one training study, university students improved their informal reasoning, measured on a critical thinking test, after 12 weeks of deliberate practice in mapping arguments.^[25] The more practice they logged, the more they improved.^[25]

Other training studies point the same way, each in its own group. In children and adolescents, a systematic review found that every included study showed gains in fluid reasoning on tasks close to the ones trained.^[27] In schools, a large review looked at metacognitive teaching, which shows pupils how to watch and steer their own thinking. It found that such teaching adds about eight months of progress over a year.^[28] A later analysis of a randomised trial looked at healthy adults aged 50 and over.^[26] After six months of online reasoning training, with the training still running, these adults scored better than an active control group on reasoning and some memory tests.^[26]

These gains were measured on the tests each study used, in the group it tested, so they show better scores, not a stronger brain circuit.

06Verdict

The verdict.

"Intelligence gets you into the room. Cognitive reflection decides whether you actually think once you're there." Editorial synthesis

Bottom line

Shortcuts are how the brain saves effort. Spend that effort where a wrong answer costs most.

The argument

When you face a problem, your brain reaches first for a familiar answer. Brain-imaging studies of reasoning tasks link the work of checking and rebuilding that answer to a network across the front, top and back of the brain. Intelligence does not protect against some biases. Training has improved some reasoning skills on tests, in children, students and older adults. This article's proposal, a reasoned approach rather than a tested fix: at decisions that matter, stop and run the four steps.

What changes when you see first principles thinking this way is how you explain your own mistakes. The question stops being "Was I smart enough?" and becomes "Did I stop to check?" Research on rational thinking supports that shift. Intelligence matters, but whether you put it to use depends partly on noticing when a quick answer needs checking.^[2]^[12]

The most practical point is also the most hopeful. You cannot always notice when a quick answer is wrong, but you can decide in advance which decisions deserve the four steps. Teams can do the same by listing assumptions before anyone proposes a solution.

No comparison figure runs here. The prose above does not resolve to one clean effect size to set against another, and this magazine does not manufacture a number to fill the space. The verdict stands on the evidence as written.

01Claim

Parts of the override can be mapped

Brain-imaging studies of reasoning tasks link parts of checking and rebuilding an answer to a network across the front, top and back of the brain. Areas tied to cognitive control are active in these tasks, and the front tip of the brain is active when people combine relations. These are links seen in lab tasks, not a map of real decisions.

02Consequence

Errors show up in many fields

Thinking errors played a part in most reviewed diagnostic errors at a few hospitals. On financial tasks, the tested biases appeared at similar rates in low-income adults and in better-off adults who grew up disadvantaged. Reflection scores were linked to telling real from fake news headlines in online studies. Some findings came from hospital cases; others came from online judgement tasks. They do not establish one common cause of costly real-world errors.

03Lever

Training can help

In one study, university students improved their informal reasoning after 12 weeks of deliberate practice, and recorded practice was linked to the amount of improvement.^[25]

Put it to work

Where this science goes next on HPC

AssessmentMental Models AssessmentScore where you actually stand Protocol90-Day Mental Models ProtocolTurn the mechanism into practice GlossaryFirst Principles ThinkingThe definition, held to the evidence

07Bibliography

The bibliography.

31 sources · ~4h est. corpus read · 31 visible

Meta · 4 Review · 2 Journal · 24 Book · 1

Type

Sort

01 Journal

Kahneman, D2011

Thinking, fast and slow

Farrar, Straus and Giroux

no doi
a b
02 Book

Stanovich, K. E., West, R. F., & Toplak, M. E2016

The rationality quotient: Toward a test of rational thinking

MIT Press

no doi
a b c
03 Journal

Evans, J. S. B. T., & Stanovich, K. E2013

Dual-process theories of higher cognition: Advancing the debate

Perspectives on Psychological Science223–241

doi: 10.1177/1745691612460685
a b c d e
04 Journal

Pennycook, G., & Rand, D. G2019

Lazy, not biased: Susceptibility to partisan fake news is better explained by lack of reasoning than by motivated reasoning

Cognition39–50

doi: 10.1016/j.cognition.2018.06.011
↩
05 Journal

Donoso, M., Collins, A. G. E., & Koechlin, E2014

Foundations of human reasoning in the prefrontal cortex

Science1481–1486

doi: 10.1126/science.1252254
a b
06 Meta

Rubric 87/100

Wertheim, J., & Ragni, M2018

The neural correlates of relational reasoning: A meta-analysis of 47 functional magnetic resonance studies

Journal of Cognitive Neuroscience1734–1748

doi: 10.1162/jocn_a_01311
a b c
07 Journal

Houdé, O., & Borst, G2015

Evidence for an inhibitory-control theory of the reasoning brain

Frontiers in Human Neuroscience

doi: 10.3389/fnhum.2015.00148
a b c
08 Journal

Christoff, K., Prabhakaran, V., Dorfman, J., Zhao, Z., Kroger, J. K., Holyoak, K. J., & Gabrieli, J. D. E2001

Rostrolateral prefrontal cortex involvement in relational integration during reasoning

NeuroImage1136–1149

doi: 10.1006/nimg.2001.0922
a b c
09 Journal

Morin, T. M., Moore, K. N., Isenburg, K., Ma, W., & Stern, C. E2023

Functional reconfiguration of task-active frontoparietal control network facilitates abstract reasoning

Cerebral Cortex5761–5773

doi: 10.1093/cercor/bhac457
↩
10 Meta

Gronchi, G., et al2024

Dual-process theory of thought and inhibitory control: An ALE meta-analysis

Brain Sciences

doi: 10.3390/brainsci14010101
a b
11 Journal

Frederick, S2005

Cognitive reflection and decision making

Journal of Economic Perspectives25–42

doi: 10.1257/089533005775196732
a b
12 Journal

Stanovich, K. E2018

Miserliness in human cognition: The interaction of detection, override and mindware

Thinking & Reasoning423–444

doi: 10.1080/13546783.2018.1459314
a b
13 Journal

Stanovich, K. E., & West, R. F2008

On the relative independence of thinking biases and cognitive ability

Journal of Personality and Social Psychology672–695

doi: 10.1037/0022-3514.94.4.672
↩
14 Journal

Toplak, M. E., West, R. F., & Stanovich, K. E2011

The cognitive reflection test as a predictor of performance on heuristics-and-biases tasks

Memory & Cognition1275–1289

doi: 10.3758/s13421-011-0104-1
↩
15 Journal

Stanovich, K. E., & Toplak, M. E2023

Actively open-minded thinking and its measurement

Journal of Intelligence

doi: 10.3390/jintelligence11020027
↩
16 Journal

Tetlock, P. E., & Gardner, D2015

Superforecasting: The art and science of prediction

Crown

no doi
↩
17 Journal

Dunbar, K., & Blanchette, I.2001

The in vivo/in vitro approach to cognition: The case of analogy

Trends in Cognitive Sciences5(8) · 334-339

doi: 10.1016/S1364-6613(00)01698-3
a b c
18 Journal

Graber, M. L., Franklin, N., & Gordon, R2005

Diagnostic error in internal medicine

Archives of Internal Medicine1493–1499

doi: 10.1001/archinte.165.13.1493
a b c d e f
19 Journal

Croskerry, P., Singhal, G., & Mamede, S2013

Cognitive debiasing 1: Origins of bias and theory of debiasing

BMJ Quality & Safety

doi: 10.1136/bmjqs-2012-001712
↩
20 Meta

Saposnik, G., Redelmeier, D., Ruff, C. C., & Tobler, P. N2016

Cognitive biases associated with medical decisions: A systematic review

BMC Medical Informatics and Decision Making

doi: 10.1186/s12911-016-0377-1
a b
21 Review

Berthet, V2022

The impact of cognitive biases on professionals' decision-making: A review of four occupational areas

Frontiers in Psychology

doi: 10.3389/fpsyg.2021.802439
↩
22 Review

Fasolo, B., Heard, C., & Scopelliti, I2025

Mitigating cognitive bias to improve organizational decisions: An integrative review, framework, and research agenda

Journal of Management

doi: 10.1177/01492063241287188
↩
23 Journal

Ruggeri, K., Ashcroft-Jones, S., Abate Romero Landini, G., Al-Zahli, N., Alexander, N., Andersen, M. H., Bibilouri, K., Busch, K., Cafarelli, V., Chen, J., Doubravová, B., Dugué, T., Durrani, A. A., Dutra, N., Garcia-Garzon, E., Gomes, C., Gracheva, A., Grilc, N., Gürol, D. M., ... Stock, F.2023

The persistence of cognitive biases in financial decisions across economic groups

Scientific Reports13(1) · 10329

doi: 10.1038/s41598-023-36339-2
↩
24 Journal

Webster, C. S., Taylor, S., & Weller, J. M2021

Cognitive biases in diagnosis and decision making during anaesthesia and intensive care

BJA Education420–425

doi: 10.1016/j.bjae.2021.07.004
a b
25 Journal

van Gelder, T., Bissett, M., & Cumming, G.2004

Cultivating expertise in informal reasoning

Canadian Journal of Experimental Psychology / Revue canadienne de psychologie expérimentale58(2) · 142-152

doi: 10.1037/h0085794
a b c d
26 Journal

Roheger, M., Kalbe, E., Corbett, A., Brooker, H., & Ballard, C2020

Predictors of changes after reasoning training in healthy adults

Brain and Behavior

doi: 10.1002/brb3.1861
a b
27 Meta

Guerin, J. M., Wade, S. L., & Mano, Q. R2021

Does reasoning training improve fluid reasoning and academic achievement for children and adolescents? A systematic review

Trends in Neuroscience and Education

doi: 10.1016/j.tine.2021.100153
a b
28 Journal

Education Endowment Foundation2025

Metacognition and self-regulation: Teaching and Learning Toolkit

source
↩
29 Journal

Borst, G., Aïte, A., & Houdé, O.2015

Inhibition of misleading heuristics as a core mechanism for typical cognitive development: Evidence from behavioural and brain-imaging studies

Developmental Medicine & Child Neurology57(s2) · 21-25

doi: 10.1111/dmcn.12688
↩
30 Journal

Allaire-Duquette, G., Brault Foisy, L.-M., Potvin, P., Riopel, M., Larose, M., & Masson, S2021

An fMRI study of scientists with a Ph.D. in physics confronted with naive ideas in science

npj Science of Learning6, 11

doi: 10.1038/s41539-021-00091-x
a b
31 Journal

Meyer, A., Attali, Y., Bar-Hillel, M., Frederick, S., & Kahneman, D2024

Cognitive reflection is a distinct and measurable trait

Proceedings of the National Academy of Sciencese2409191121

doi: 10.1073/pnas.2409191121
↩

The protocol card