Want to dive into Scott Alexander's work and his thousands of blog posts? This fan website lets you sort and do semantic search through the whole codex. Enjoy!

See also Top Posts and All Tags.

Tag: study critique

Minutes:
Pick a custom range (minutes). Leave a field empty for no limit.
–
Blog:
Year:
2026
2025
2024
2023
2022
2021
2020
2019
2018
2017
2016
2015
2014
2013
Tags:
Filter by tag...
Exclude tag...
5306 tags
Links:
Filter by linked site (twitter, substack…)
268 posts found
Compact Mode
Save Reads
Aug 03, 2026
acx
Read on

Does Forecasting Have Room At The Top?

25 min • 3,865 words • 170 comments • 177 likes
Scott analyzes whether AI forecasters will plateau near human levels or keep improving, arguing that even non-miraculous progress could yield 4-12 percentage point improvements in prediction market accuracy. Longer summary
Scott examines whether AI forecasters will plateau near current human performance or continue improving far beyond it. He analyzes a 2010 study by Daniel Reeves showing prediction markets only marginally outperformed simple statistical models, but argues this doesn't necessarily limit future progress. The post reframes those results to show they actually represent meaningful gains in difficult-to-predict domains like sports, then applies this reasoning to geopolitical forecasting. Using three different 'anchors' - comparing AI improvements to past technological gains, human-to-superhuman forecaster improvements, and chess AI progress - Scott estimates that maximally capable AI forecasters could improve prediction market accuracy by 4-12 percentage points over current levels, which he argues would be genuinely valuable even if not miraculous. Shorter summary
Jul 28, 2026
acx
Read on

Psychology Research Is Mostly Fine

9 min • 1,370 words • 377 comments • 438 likes • podcast (12 min)
Scott defends psychology against broad dismissals by showing that the replication crisis primarily affected only social priming within social psychology, while the vast majority of the field consists of solid, replicable research across many subdisciplines. Longer summary
Scott Alexander argues that the replication crisis has unfairly tarnished the entire field of psychology when it primarily affected only social psychology, and more specifically social priming research. He demonstrates this by walking through a standard intro psychology textbook, showing that 15 out of 16 chapters cover solid research areas like psychophysics, neuroscience, cognitive psychology, and clinical psychology, while only one chapter deals with the problematic social psychology subfield. He provides a long list of the diverse types of psychologists doing legitimate work, from those studying visual processing to those treating drug abuse, and emphasizes that bad research exists in all scientific fields, not just psychology. The post concludes by urging critics to be specific and target social priming rather than dismissing psychology as a whole. Shorter summary
Jul 17, 2026
acx
Read on

Your Book Review: Great And Desperate Cures

132 min • 20,439 words • 192 comments • 211 likes • podcast (147 min)
A review of 'Great and Desperate Cures' by Elliot Valenstein, which examines the history of lobotomy and psychosurgery, explaining how tens of thousands of patients underwent dangerous brain operations from the 1930s-1950s despite poor evidence of effectiveness. Longer summary
The post reviews 'Great and Desperate Cures' (1986) by psychologist Elliot Valenstein, exploring how lobotomy became widespread despite its horrific consequences. The review details the procedure itself (inserting an ice-pick-like instrument through the eye socket into the brain), then examines six factors Valenstein identifies for its acceptance: desperate patients and families, physician ambition (especially Egas Moniz and Walter Freeman), uncritical acceptance by researchers and media, territorial disputes between medical specialties, and cold economics. The post includes extensive details about key figures, particularly Freeman's aggressive promotion of transorbital lobotomy and his traveling surgery roadshows. It discusses patient outcomes, the lack of theoretical justification, media hype, and the procedure's eventual replacement by psychoactive drugs. The review concludes by reflecting on modern parallels and the tragic story of Howard Dully, lobotomized at age twelve, arguing that while Freeman likely had good intentions, the safeguards that might have prevented such harm were inadequate then and remain concerning today. Shorter summary
Jul 08, 2026
acx
Read on

Your Book Review: The Book Of Abraham

108 min • 16,711 words • 807 comments • 315 likes • podcast (105 min)
A deep dive into the Book of Abraham, focusing on how Joseph Smith's translations of Egyptian papyri were definitively proven false by Egyptologists, yet Mormonism continues to thrive—raising questions about whether instrumental rationality trumps epistemic rationality. Longer summary
This review examines the Book of Abraham, a Mormon scripture that Joseph Smith claimed to translate from Egyptian papyri purchased in 1835. The post methodically presents evidence showing Smith's translation was fraudulent: Egyptologists identified the papyri as standard funerary texts (the Book of Breathings for a priest named Hôr, dated 150 BC), not Abraham's writings from 2000 BC; Smith's interpretations of the three facsimiles contradict expert consensus (including identifying the god Min's erect phallus as God on his throne); and the rediscovery of the original papyri in 1967 confirmed they don't match Smith's translation. The author also covers Smith's history of treasure-digging with seer stones, the fraudulent Kirtland Anti-Banking Company, and the forged Kinderhook Plates. Despite all this evidence, Mormonism continues to grow and produces highly successful, charitable, tight-knit communities. The post concludes by exploring how a demonstrably false belief system can be instrumentally rational through costly signaling, community cohesion, and psychologically optimized doctrines that give members purpose and identity. Shorter summary
Jun 23, 2026
acx
Read on

Should People Avoid Whole-Body Screening Info?

30 min • 4,501 words • 354 comments • 269 likes • podcast (25 min)
Scott analyzes whether whole-body screening MRIs are worth it by doing a detailed cost-benefit calculation, finding they cost about $108,000 per quality-adjusted life-year saved (right around the threshold of cost-effectiveness), and argues that while rich people immune to anxiety might benefit, most people claiming to be rational about medical decisions probably aren't. Longer summary
Scott performs a detailed cost-benefit analysis of whole-body screening MRIs in response to controversy over medical experts recommending against them. Using rough order-of-magnitude estimates, he calculates that screening 1,000 people saves about 32 quality-adjusted life-years (QALYs) at a cost of $2.7 million plus time and anxiety costs, working out to about $108,000 per QALY saved—right at the threshold of cost-effectiveness. He explores whether rich people who don't care about money should get screened, finding a plausible case but with many caveats about unknown factors that could swing the calculation either way. He then applies this analysis to Midjourney's proposed ultrasound scanner, finding it's unlikely to be clearly better than MRI. The post ends with a warning that people who think they're rational enough to ignore false positives are often the same people making irrational medical decisions based on contrarian appeals. Shorter summary
Jun 16, 2026
acx
Read on

Never Cross a River Four Feet Deep on Average

45 min • 6,939 words • 115 comments • 190 likes • podcast (38 min)
A replication of a 2023 neuroscience study on brain entrainment and learning found that the original effect likely doesn't exist, revealing how the study obscured key issues through statistical averaging and highlighting how AI tools are democratizing scientific scrutiny. Longer summary
Sasha Putilin received an ACX grant to replicate a 2023 study claiming that flickering lights synchronized to individual brain rhythms could boost learning speed threefold. The $32,000 replication with 12 participants (versus 80 in the original) found no such effect. Upon examining the original data more closely, Putilin discovered the apparent effect was driven entirely by a few participants with large negative learning rates (getting worse over time) who happened to all be in one experimental group. The original study had obscured this pattern by presenting averaged data rather than individual results. Putilin argues this exemplifies 'cargo-cult statistics' where researchers mechanically apply statistical rituals without critically examining underlying data. The post concludes by suggesting AI tools are democratizing meta-science, enabling anyone to audit published research that previously required expert-level effort. Shorter summary
Mar 31, 2026
acx
Read on

Against The Concept Of Telescopic Altruism

11 min • 1,562 words • 610 comments • 463 likes • podcast (11 min)
Scott argues against the concept of "telescopic altruism" - the claim that liberals care more about distant strangers than nearby people - showing that people who care about faraway causes also care about nearby ones, and that compassion is generally correlated across all distances rather than inversely related. Longer summary
Scott Alexander critiques the concept of "telescopic altruism," which claims that some people (usually liberals) ignore those close to them to care about distant strangers. He argues this accusation collapses under scrutiny: people who care about 50,000 deaths in Gaza would also care about 50,000 deaths among their neighbors. He debunks a commonly cited study whose heatmap visualization is misinterpreted to suggest liberals care more about rocks than family, when it actually just shows the outer limit of their moral concern. Scott proposes instead a "correlated altruism" hypothesis, citing Dave Barry's principle that someone nice to waiters is genuinely nice. He provides evidence that liberals who support foreign aid also support domestic programs like school lunches and COVID measures, and shows statistics suggesting liberals aren't worse at maintaining family relationships. The post concludes by acknowledging that some people do neglect their communities, but argues this happens because they care too much and are incompetent, not because they don't care at all. Shorter summary
Feb 18, 2026
acx
Read on

Record Low Crime Rates Are Real, Not Just Reporting Bias Or Improved Medical Care

17 min • 2,566 words • 716 comments • 390 likes • podcast (17 min)
Scott demonstrates that US crime rates are at historic lows using multiple data sources, then debunks the theories that this is due to reporting bias or improved medical care, before briefly discussing possible explanations for the decline. Longer summary
Scott presents evidence that US crime rates, particularly murder rates, are at or near historic lows, then systematically debunks two common objections: that the decline is due to underreporting bias, and that it's an artifact of improved medical care saving would-be murder victims. He shows that multiple independent data sources (including victim surveys and consistently-reported crimes like car theft) confirm the decline, and that while medical care has improved, gun injuries have become more severe at roughly the same rate, canceling out the effect. The post concludes by listing various theories for why crime has declined and questioning why public perception doesn't match the data. Shorter summary
Feb 11, 2026
acx
Read on

Political Backflow From Europe

14 min • 2,052 words • 1,123 comments • 403 likes • podcast (12 min)
Scott examines how American political discourse absorbs European narratives that don't fit the US context, particularly around immigration and crime statistics. Longer summary
Scott argues that American political discourse sometimes absorbs European issues that don't apply to the US context. He gives two main examples: the narrative about generational wealth transfer through pensions (which happens in Europe but not America), and conservative talking points about immigrants being criminals and welfare recipients (largely true in parts of Europe, largely false in America). He provides detailed statistics showing that most US immigrant groups, including asylum seekers, have lower crime rates than native-born Americans, contrasting this with higher rates in countries like Germany. Scott suggests both liberals and conservatives avoid acknowledging this difference because it's politically inconvenient, but argues liberals should directly challenge conservatives to focus on American rather than European data. Shorter summary
Dec 03, 2025
acx
Read on

The Good News Is That One Side Has Definitively Won The Missing Heritability Debate

13 min • 1,961 words • 590 comments • 420 likes • podcast (13 min)
Scott examines a new genetic study on missing heritability that both hereditarians and nurturists claim vindicates their position, concluding that despite the study's advances, the fundamental debate over how heritable traits like IQ actually are remains unresolved. Longer summary
This post discusses a new genetic study that attempted to resolve the "missing heritability" debate - the gap between high heritability estimates from twin studies (50-80%) and low estimates from molecular genetic studies (10-20%). The study used whole-genome sequencing to include rare genetic variants and found they could account for about 88% of expected heritability, but the actual heritability estimates themselves were only medium (30-40%). Both hereditarians and nurturists claimed victory: hereditarians because the gap was closed (proving the genes exist), nurturists because the total heritability found was still lower than twin studies suggested. Scott examines both sides' arguments, including measurement error corrections and various confounders, and concludes that despite everyone's claims, the debate remains fundamentally unresolved as different methods continue producing different estimates with no clear explanation why. Shorter summary
Oct 01, 2025
acx
Read on

The Fatima Sun Miracle: Much More Than You Wanted To Know

213 min • 32,949 words • 755 comments • 679 likes • podcast (186 min)
A detailed investigation of the Sun Miracle of Fatima in 1917, where 70,000 people witnessed the sun appear to spin, change colors, and fall to earth, analyzing witness testimonies, skeptical explanations, and similar phenomena at other Marian apparition sites. Longer summary
Scott conducts an extensive investigation into the 1917 Miracle of the Sun at Fatima, where tens of thousands witnessed what appeared to be the sun spinning, changing colors, and falling to earth. He examines approximately 60 primary testimonies, evaluates common skeptical explanations (optical phenomena, weather events, mass hallucination), and documents similar sun miracles at other Marian apparition sites worldwide. Scott then critiques Dalleur's theory about distant witnesses and shadow analysis, explores why more people don't see these phenomena normally, and proposes a tentative materialist explanation involving a rare optical illusion modulated by cloud cover and social priming. The post ends by suggesting research directions for those interested in further investigation. Shorter summary
Sep 12, 2025
acx
Read on

Your Review: The Synaptic Plasticity and Memory Hypothesis

77 min • 11,801 words • 180 comments • 212 likes • podcast (66 min)
A review of the synaptic plasticity and memory hypothesis arguing that it is insufficient to explain memory, and proposing a broader cellular processes and memory hypothesis that includes molecular and intracellular mechanisms. Longer summary
The post reviews the synaptic plasticity and memory (SPM) hypothesis, which claims that learning and memory are stored in changes to synaptic weights between neurons. Through several compelling examples, from cannibalism rituals to heart transplant stories and single-cell learning, the author argues that while the SPM hypothesis has been productive, it is incomplete and partially wrong. The post presents evidence that memory can be stored through non-synaptic mechanisms, and proposes an alternative called the cellular processes and memory (CPM) hypothesis, which suggests memory storage involves multiple molecular and intracellular processes beyond just synaptic weights. Shorter summary
Sep 04, 2025
acx
Read on

Links For September 2025

54 min • 8,283 words • 917 comments • 233 likes • podcast (45 min)
Scott Alexander presents his monthly collection of interesting links and developments across technology, science, culture, and policy, with personal commentary on each item. Longer summary
Scott Alexander shares a collection of 61 interesting links, spanning topics from AI development and scientific studies to cultural observations and policy discussions. The post covers diverse subjects including GPT-5's reception, genetic research, religious demographics, urban development, mental health studies, and various scientific discoveries. Scott provides commentary and analysis on many of these items, often connecting them to broader themes or offering his perspective on controversial issues. Shorter summary
Aug 14, 2025
acx
Read on

In Defense Of The Amyloid Hypothesis

70 min • 10,813 words • 188 comments • 279 likes • podcast (35 min)
A guest post defending the amyloid hypothesis of Alzheimer's disease, explaining why amyloid is likely the root cause despite recent criticism and discussing why early treatments have only been partially effective. Longer summary
David Schneider-Joseph presents a detailed defense of the amyloid hypothesis of Alzheimer's disease, which states that the disease is caused by accumulation of amyloid-β peptide. He explains the ATN model (amyloid → tau → neurodegeneration) and provides extensive evidence from genetics, clinical studies, and animal models. The post addresses various criticisms of the hypothesis, including the recent research fraud controversy and apparent failures of early treatments. The author explains why current treatments have only achieved about 30% efficacy and predicts that future treatments targeting amyloid earlier and more effectively could achieve much better results. Shorter summary
Jul 31, 2025
acx
Read on

Suddenly, Trait-Based Embryo Selection

56 min • 8,637 words • 747 comments • 407 likes • podcast (60 min)
Scott analyzes the sudden emergence of commercial trait-based embryo selection services, discussing both the scientific validity and ethical implications of selecting embryos for health outcomes, intelligence, and physical traits. Longer summary
Scott examines the recent development of commercial embryo selection services, particularly focusing on companies like Genomic Prediction, Orchid Health, Nucleus, and Herasight. He explains how the technology works, comparing different companies' claims and methodologies, and discusses both scientific challenges and ethical concerns. The post explores the potential benefits, like reduced disease risk and increased IQ, while acknowledging issues around cost, racial disparities, and social implications. A significant portion focuses on Herasight's critique of competitor Nucleus's scientific claims. The post concludes by placing this technology in the broader context of human enhancement and future technological developments. Shorter summary
Jul 11, 2025
acx
Read on

Your Review: Of Mice, Mechanisms, and Dementia

51 min • 7,754 words • 170 comments • 184 likes • podcast (48 min)
A detailed review of a 1995 paper introducing the PDAPP mouse model for Alzheimer's disease, examining how its limitations were overlooked and shaped decades of potentially misguided Alzheimer's research. Longer summary
The post reviews a landmark 1995 paper introducing the PDAPP mouse model for Alzheimer's disease, analyzing how its technical achievements and limitations shaped three decades of Alzheimer's research. The author examines the paper's methodology, results, and claims, showing how the model's flaws - including extreme protein overexpression and lack of key disease features - were overlooked in favor of a compelling but incomplete amyloid cascade hypothesis, leading to years of failed drug development and missed opportunities to explore alternative approaches. Shorter summary
Jul 04, 2025
acx
Read on

Your Review: School

55 min • 8,460 words • 553 comments • 435 likes • podcast (46 min)
This review explores how schools are primarily designed to maximize motivation rather than learning, explaining why age-graded classrooms and seemingly inefficient group learning have persisted despite numerous attempts at reform. Longer summary
The post examines why schools have maintained their traditional structure of age-graded classrooms where all students learn the same content, despite its apparent inefficiencies. The author argues that schools are designed primarily to maximize motivation rather than learning, using conformity as a key tool. Through analyzing various attempts at personalized learning and their consistent failures to scale beyond about 5% of students, the post explains how students fall into three categories: no-structure learners, low-structure learners, and high-structure learners. The author concludes that while the current system is far from perfect, it has proven more effective at scale than any alternatives, predicting that despite continued attempts at reform, the basic structure of schooling will remain unchanged. Shorter summary
Jul 03, 2025
acx
Read on

Highlights From The Comments On Missing Heritability

66 min • 10,199 words • 131 comments • 94 likes • podcast (61 min)
Scott Alexander responds to comments and criticism on his earlier post about 'Missing Heritability', discussing issues like gene-environment interactions, sequencing technology limitations, and the use of polygenic scores across ancestry groups. Longer summary
This post compiles and responds to notable comments on Scott's earlier post about 'Missing Heritability'. The post is structured in four sections, starting with responses from experts named in the original post, particularly Sasha Gusev who critiques the treatment of gene-environment interactions and cross-population polygenic scores. The second section features detailed technical comments from knowledgeable readers about topics like genetic interactions and sequencing technology limitations. The third section addresses specific corrections to the original post, while the final section covers various other interesting comments and discussions. Throughout, Scott engages with the criticisms and new perspectives while maintaining his original position on most key points. Shorter summary
Jul 01, 2025
acx
Read on

Links For July 2025

28 min • 4,200 words • 581 comments • 210 likes • podcast (32 min)
Scott shares his monthly collection of 56 interesting links and developments from July 2025, covering AI, politics, science, and culture, with brief commentary on each. Longer summary
This is a collection of 56 interesting links and news items from July 2025, covering topics from AI development to politics to scientific research. The post includes updates on OpenAI's status, developments in AI regulation, new medical treatments, cultural trends, and various scientific findings. Scott maintains his usual style of presenting these with brief commentary and occasional humor, while being careful to note that he hasn't independently verified all links. Some notable items include Trump's "Big Beautiful Bill", new AI safety developments, trends in social media usage, and various medical breakthroughs. Shorter summary
Jun 30, 2025
acx
Read on

Contra Skolnick On Schizophrenia Microbes

11 min • 1,619 words • 180 comments • 256 likes • podcast (11 min)
Scott Alexander refutes Stephen Skolnick's theory that schizophrenia is caused by gut microbes rather than genetics, showing why the evidence better supports schizophrenia being a complex genetic neurodevelopmental disorder. Longer summary
Scott Alexander responds to Stephen Skolnick's theory that schizophrenia is caused by gut microbes rather than genetics. He systematically dismantles Skolnick's arguments by showing that twin concordance rates are exactly what we'd expect from a genetic condition, that microbiological inheritance patterns don't match schizophrenia inheritance patterns, and that the gut bacteria evidence cited was likely caused by antipsychotic medication rather than being causative. He concludes by explaining why schizophrenia is best understood as a complex neurodevelopmental disorder with multiple contributing factors rather than having a single cause. Shorter summary
Jun 27, 2025
acx
Read on

Your Review: Alpha School

121 min • 18,662 words • 686 comments • 714 likes • podcast (109 min)
A parent details their year-long experience with Alpha School in Austin, examining how this innovative program achieves accelerated learning through a combination of technology, personalized instruction, and incentive systems. Longer summary
This extensive review details the author's year-long experience with Alpha School, an innovative educational program in Austin that claims to achieve accelerated learning through a '2-hour learning' platform, analyzing its history, methods, effectiveness, and potential for scaling. The author explores how Alpha combines technology, personalized learning, incentive systems, and afternoon workshops to achieve its results, while also examining the challenges it faces in scaling beyond its current elite private school model. The review is particularly careful to separate marketing claims from reality, showing that while the '2-hour learning' actually takes closer to three hours and isn't truly AI-powered, the program does achieve remarkable results through a combination of proven educational techniques. Shorter summary
Jun 26, 2025
acx
Read on

Missing Heritability: Much More Than You Wanted To Know

68 min • 10,491 words • 669 comments • 456 likes • podcast (62 min)
Scott explores the 'missing heritability' problem in genetics, where twin studies show traits like IQ and educational attainment are highly heritable but newer genomic methods find much lower heritability, analyzing various potential explanations for this discrepancy. Longer summary
The post examines the 'missing heritability' problem in genetics, where twin studies consistently show behavioral traits like IQ and educational attainment are substantially heritable (around 40-60%), while newer genomic methods find much lower heritability (around 15-20%). Scott reviews the history of behavioral genetics research, explains various study methodologies and their potential biases, and analyzes different hypotheses for this discrepancy. He examines whether twin studies might be flawed, whether newer methods might be missing important genetic effects, and whether educational attainment might be an unusually problematic trait to study. While acknowledging remaining mysteries, he tentatively concludes that twin studies are probably largely correct and that newer methods may be missing rare variants and genetic interactions. Shorter summary
May 29, 2025
acx
Read on

Sorry, I Still Think MR Is Wrong About USAID

31 min • 4,651 words • 555 comments • 559 likes • podcast (28 min)
Scott Alexander responds to Tyler Cowen about USAID funding, correcting his own previous claims about overhead costs while maintaining that Cowen's criticism of USAID was misleading and potentially harmful. Longer summary
Scott Alexander responds to Tyler Cowen's criticism of his previous post about USAID funding. He addresses several points: whether Cowen endorsed Rubio's claims about USAID waste, the true nature of overhead costs in USAID-funded organizations, and the broader debate about foreign aid effectiveness. Scott shows that actual administrative overhead in major USAID partners like Catholic Relief Services is much lower than previously thought (around 6-7% rather than 30%), admits his mistake on this point, but maintains his criticism of Cowen's original post as misleading. He argues that USAID's work is predominantly focused on essential humanitarian aid rather than wasteful programs. Shorter summary
May 22, 2025
acx
Read on

The Evidence That A Million Americans Died Of COVID

14 min • 2,056 words • 1,209 comments • 490 likes • podcast (13 min)
Scott presents evidence that COVID-19 did kill approximately 1.2 million Americans, addressing skeptics by analyzing excess death data and addressing common counterarguments. Longer summary
In response to skeptics questioning the official COVID-19 death toll, Scott Alexander presents evidence supporting the 1.2 million deaths figure. He shows excess mortality data from multiple sources indicating 500,000-700,000 extra deaths in both 2020 and 2021, closely matching reported COVID deaths. He addresses various counter-arguments, including the 'died with vs of COVID' distinction, the role of treatments like ventilators, and the common experience of not personally knowing COVID victims. The post demonstrates how the data supports COVID being the primary cause, and explains why personal experiences might not reflect the true scale of the pandemic. Shorter summary
May 15, 2025
acx
Read on

Book Review: Selfish Reasons To Have More Kids

43 min • 6,596 words • 724 comments • 497 likes • podcast (41 min)
Scott reviews Bryan Caplan's book arguing that modern parents can relax their intensive parenting, while wrestling with whether this advice still applies in the age of smartphones and social media. Longer summary
Scott reviews Bryan Caplan's book 'Selfish Reasons To Have More Kids', exploring its main arguments about how parents today spend much more time on childcare than previous generations despite evidence that parenting style doesn't greatly affect outcomes. The post explores historical childcare data, the cultural shift away from letting kids play unsupervised, and modern challenges like screen time. Scott, dealing with his own twins, finds the book's advice about relaxing parenting standards compelling but struggles with modern concerns about phones and technology that weren't relevant when the book was written in 2011. Shorter summary
Per page:
Showing 1 to 25 of 268 results
Get these search results in an EPUB

Your filters match 268 posts.

Posts to include
–
Leave empty to keep the defaults. Range cannot exceed 500 posts.
Download now

Generates an EPUB right now and downloads it to your device.

Send to email

Generates an EPUB in the background and emails you a temporary download link.

Your email is not shared with anyone.

Email address

To send to your Kindle, just use this link.