Statistics · Beginner · 4 h

Describing data honestly

Mean, median, spread, percentiles and outliers on real taxi fares, including a fare of 863,372.

Why employers care: Analysts summarise data every day. Choosing the mean when the median is right is the most common way reports mislead.

Lessons, in order

Mean vs median: central tendency and outliersSummarise 7.6 million New York taxi fares honestly, find the fare of 623,259.86 that breaks the average, and decide which number to put in the report.12 min
Variance and standard deviation explainedMeasure spread properly on 96,000 real Brazilian e-commerce deliveries, avoid the numpy versus pandas ddof trap, and rank states by how consistent delivery is, not how fast.11 min
Correlation vs causation: Pearson and SpearmanCompute Pearson, Spearman and Kendall correlations on real penguin and taxi data, watch one bad row destroy a correlation of 0.95, reproduce Simpson's paradox, and explain why correlation is not causation.11 min