Statistical Interpretation & Uncertainty - Statistics addresses reasoning under uncertainty, a more complex question than calculus's rate of change [3] - Challenges like climate change, public health, and AI rely on making decisions with incomplete information [4] - The same data can tell different stories based on its presentation [4] Cognitive Biases in Statistics - Uncertainty is unintuitive, making statistical interpretation prone to errors [3] - Selection bias, where sampling skews results, can affect data interpretation [15] - Common sense is crucial for identifying misinformation and statistical misconceptions [26] - Correlation does not imply causation; coinciding trends don't necessarily indicate a causal relationship [27] COVID-19 Testing & Interpretation - A positive COVID-19 test with 80% sensitivity and 98% specificity may result in less than 50% chance of actually having COVID-19 [9][11] - With 1% of 1,000 people having COVID-19, a test with 80% sensitivity identifies 8 true positives, while 98% specificity results in 20 false positives, leading to a 29% chance of actually having the virus if testing positive [12][13] Data Presentation & Misinterpretation - Misleading graphs can distort data interpretation, such as hospitalization rates based on vaccination status without adjusting for cohort size [28][29] - Hospitalization data adjusted for vaccination cohort size reveals that unvaccinated individuals were overrepresented in hospitalizations [31][33] Combating Statistical Misinformation - Three key questions to combat statistical misinformation: How was the data collected? What is the context? Has the data been pulled or combined? [36][38][40]