Statistics - Topics

Misleading charts: how data visualization can lie to you Countries with the longest life expectancy and why Cancer survival rates: what the statistics really mean What is home advantage and does it really exist? What is a triple-double and why does it matter? What are the most visited cities in the world and why? What is xG and how do analytics change football? What is the true unemployment rate and how is it measured? Is religion declining or simply changing? What is the safest form of transport? When a regression is the wrong tool for the question What sample size does to the reliability of a survey result Outlier detection: real signal or a data entry mistake Handling missing values: dropping rows vs imputing them Comparing prices over time means adjusting for inflation — most charts do not Meta-analysis is comparison done rigorously and it still has limits Reading a confidence interval without a statistics degree Why regression to the mean fools almost every before-and-after study Why Anscombe's quartet looks identical until you plot it Bias-variance tradeoff explained without the dartboard metaphor Reading a poll: margin of error vs the headline lead Fixing the unit of analysis before the very first measurement is taken People behave too inconsistently across situations for behavior alone to work as a classification. How much does an interest label actually predict about a person? Epidemic compartment models and the compartments people forget Interests drift fast enough that interest-based categories go stale. Which human characteristics can actually be measured and which only look measurable? Type-2 fuzzy sets and handling uncertainty about uncertainty Survivorship bias: the advice you never hear from the ones who failed Why do two people with identical measured characteristics often behave nothing alike? Why do measurable characteristics crowd out the ones that matter most to the people described? Why model predictions lose skill outside the range they were fitted on