A thread through the archive
Essays tagged “study critique”. Follow an idea across the archive.
Scott examines reader tests and discussions of AI's GeoGuessr abilities, revealing that AIs perform best with tourist locations and are roughly on par with human professionals.
Scott analyzes Ted Cruz's database of 'woke' NSF grants and finds that only 40% were actually woke, with many regular science grants included simply for having a mandatory diversity statement.
Scott shares various interesting links and news items from January 2025, covering topics from AI development and politics to historical curiosities and economic trends.
Scott responds to comments about Lynn's IQ data, addressing how IQ testing might break down in under-educated populations and how this relates to the apparent disconnect between test scores and real-world capabilities.
Scott Alexander examines Richard Lynn's controversial national IQ estimates and argues they are actually consistent with environmental rather than genetic explanations of IQ differences, while explaining common misconceptions about what low IQ means.
Scott examines the likelihood and potential severity of an H5N1 bird flu pandemic, analyzing prediction markets and historical data to estimate a 5% chance of pandemic in the next year with most likely moderate severity.
Scott analyzes research on whether longer prison sentences reduce crime, examining studies on deterrence, incapacitation, and aftereffects, to conclude that while prison does decrease crime through incapacitation, it's less cost-effective than other methods like policing.
Scott Alexander attempts to replicate a poll claiming high rates of COVID vaccine deaths, finds much lower rates, and concludes such polls are unreliable due to bias.
Scott Alexander examines the effectiveness of education by analyzing knowledge retention, questioning the value of schooling beyond basic skills.
Scott Alexander reviews a book on Internal Family Systems therapy that controversially claims some mental health issues are caused by literal demons.
Scott Alexander summarizes comments on his healthcare effectiveness debate with Robin Hanson, including clarifications, study author responses, and reader perspectives on various healthcare topics.
Scott Alexander responds to Robin Hanson's reply on medical effectiveness, clarifying interpretations and reiterating arguments about the limitations of insurance experiments in evaluating medical care.
Scott Alexander challenges Robin Hanson's claim that medicine doesn't work by analyzing health insurance studies and presenting evidence of medicine's effectiveness.
Scott Alexander argues that schizophrenia should be described as predominantly genetic, addressing counterarguments and emphasizing the importance of this perspective for prevention strategies.
Scott Alexander reviews a study on AI sleeper agents, discussing implications for AI safety and the potential for deceptive AI behavior.
Scott Alexander examines a study challenging ketamine's effectiveness as an antidepressant, discussing potential confounding factors and maintaining his belief in ketamine's utility despite the study's results.
Scott Alexander examines recent studies on the Fraternal Birth Order Effect, finding evidence for a broader sibling birth order effect on homosexuality, though its exact nature and mechanism remain debated.
A critical review of 'Why Nations Fail' that examines the authors' academic work and methodology, while questioning the book's quality and practical implications.
Scott Alexander argues that describable dating preferences are valuable, despite studies suggesting otherwise, by pointing out flaws in research and emphasizing real-world evidence.
Scott Alexander critiques a study claiming an illusion of moral decline, pointing out methodological flaws and suggesting alternative explanations for the perception of declining morality.
Scott Alexander explores the difficulties in contextualizing statistics, providing numerous examples to show how the same data can be presented to seem significant or trivial.
The post challenges the notion that effect sizes below 0.50 for medications are clinically insignificant, using antidepressants as a case study and comparing them to other common drugs.
Scott Alexander examines hypergamy in modern marriages, finding educational hypergamy has reversed while income hypergamy persists, with strong class homogamy for both genders.
Scott Alexander addresses feedback on his post about bisexuality and Long COVID, discussing alternative explanations and defending his methodology.
The post explores AI forecasting capabilities, compares prediction market performances, and provides updates on various ongoing predictions in technology and politics.