Scott examines studies linking sleeping pills to increased mortality, highlighting a new study that found no link after adjusting for 300 confounders, potentially challenging the validity of less thorough studies. Longer summary

Scott Alexander discusses a controversial topic in medical research: the link between sleeping pills and increased mortality. He reviews several studies that found a strong association between sleeping pill use and higher death rates from various causes, even after controlling for confounders. However, he then introduces a new study by Patorno et al. that found no such link for benzodiazepines. The key difference is that this study adjusted for an unprecedented 300 confounders using a new statistical algorithm. Scott suggests that if this study is correct, it could mean that many other medical studies that only control for a handful of confounders might be inadequate. He draws parallels to other fields where dramatically increasing scale or intensity has led to new insights. Shorter summary

Scott Alexander cautions against drawing strong conclusions from regional scatterplots, demonstrating how apparent correlations can be artifacts of regional clustering rather than true relationships. Longer summary

Scott Alexander warns about the potential misinterpretation of regional scatterplots, using an example of a seemingly strong correlation between rainfall and gender balance in US states. He explains that such correlations can be misleading due to regional clustering, where the relationship appears strong between clusters but may not exist within them. The post discusses how this issue affects interpretation of data in various fields, including gun violence, national happiness, and income correlations. Scott emphasizes the importance of careful analysis and consideration of confounding factors when drawing conclusions from such plots. Shorter summary

Scott Alexander examines the relationship between gun ownership and homicide rates, finding a likely causal link but noting it's not the main factor in the US's high murder rate. Longer summary

Scott Alexander analyzes the relationship between gun ownership and homicide rates in the US, critiquing misleading statistics and conducting a more rigorous analysis. He finds that while there is likely a causal relationship between gun ownership and homicide rates, it's not the primary factor explaining the US's high murder rate compared to other developed countries. He tentatively concludes that gun control measures similar to Australia's might be beneficial, but emphasizes the need for further research. Shorter summary

Scott criticizes a study linking childhood bullying to negative adult outcomes, arguing that its method of controlling for confounders is inadequate and proposing alternative explanations for the correlation. Longer summary

Scott criticizes a study claiming that childhood bullying victimization leads to negative adult outcomes. He argues that the study's attempt to control for confounding factors is inadequate, as bullies are likely better at identifying vulnerable children than the researchers' measures. Scott suggests that unmeasured factors like height could explain the correlation, and that the study's method of adjusting for confounders is unreliable. He proposes that a proper study would involve an anti-bullying intervention with control schools. The post also mentions a contrasting study that found no association after adjusting for confounders, and questions the reliability of parent reports on bullying used in the original study. Shorter summary