Scott Alexander argues that while IQ is valuable for research, it's not reliable for predicting individual success and shouldn't be a major personal concern. Longer summary

Scott Alexander addresses the issue of people becoming overly concerned about their personal IQ scores. He explains that while IQ is a valuable research tool for understanding populations, it's not as useful or reliable for predicting individual outcomes. He points out that IQ tests often give inaccurate results for individuals, and even accurate IQ scores don't determine a person's potential for success. Alexander uses analogies like family wealth to illustrate how statistical correlations don't necessarily apply to individuals. He emphasizes that many factors beyond IQ contribute to success in various fields. The post aims to reassure readers who may be worried about their own IQ scores, while still maintaining the importance of IQ research for understanding broader societal trends. 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 offers follow-up insights on his guns and states analysis, addressing cultural factors, rape correlations, replication issues, and the complexity of gun control debates. Longer summary

Scott Alexander provides follow-up thoughts on his previous post about guns and states. He discusses the interaction between cultural factors and gun ownership, explores correlations with rape rates, addresses replication issues, and critiques the gun control debate. He also mentions a meta-analysis by Gary Kleck that shows mixed results in guns-homicide correlation studies, emphasizing the importance of considering multiple studies rather than relying on a single one. Shorter summary

Scott Alexander examines how different statistical presentations of the same data in social science studies can lead to vastly different interpretations, potentially misleading readers. Longer summary

Scott Alexander discusses the importance of understanding and interpreting statistical measures in social science studies, particularly focusing on correlation, percent variance explained, and visual representations of data. He examines two studies: one on IQ and state wealth, and another on wealth inheritance. Alexander highlights how different presentations of the same data can lead to vastly different interpretations, potentially misleading readers. He emphasizes the need for a good grasp of statistical concepts and realistic expectations when evaluating social science research, noting that correlations above 0.4 are rare in this field. Shorter summary

Scott Alexander explores how summary statistics can be misleading when describing relationships between variables, using examples of IQ's correlation with crime and income. Longer summary

Scott Alexander discusses two examples where summary statistics can be misleading. The first example involves the relationship between IQ and crime, where a large difference in average IQ between offenders and non-offenders coexists with a low correlation coefficient. The second example concerns the relationship between IQ and income, where substantial differences in average income across IQ deciles coexist with a relatively low correlation coefficient. In both cases, Scott emphasizes the importance of looking beyond summary statistics and considering the full distribution of data, potentially by examining scatter plots. Shorter summary

Scott Alexander examines Google Correlate, finding it produces both obvious and nonsensical correlations, and explores its potential uses and limitations. Longer summary

Scott Alexander explores Google Correlate, a tool that finds search terms correlated with specific datasets. He finds that while some correlations are obvious (like 'skiing' with cold temperatures), others are nonsensical, especially for diseases. He tests various datasets, including parasite load and family ties, finding odd correlations that often seem to be confounded by factors like race or geography. Scott concludes that Google Correlate isn't very useful for revealing deep connections, but might serve as a sanity check for confounding factors in studies or as a cheap method for creating datasets. He ends with a humorous correlation between searches for an anti-parasitic medication and curiosity about Lil' Wayne's sexual orientation. Shorter summary