Dr. Mario's 2nd đź§ 

Home

❯

pages

❯

Data Analysis Pitfalls

Data Analysis Pitfalls

1 min read

  • Twyman’s law

  • Paradoja de Berkson

  • Simpsons paradox

  • Polanyi’s Paradox

  • Late reported facts: https://jasher.substack.com/p/how-fewer-foreign-passengers-flying?utm_campaign=posts-open-in-app&triedRedirect=true

  • Comparing time to action between cohorts with different time to make that action (without truncation). Modeling conversion rates and saving millions of dollars using Kaplan-Meier and gamma distributions.

  • YoY comparisons must be on the same “cultural” time.

  • We must compare numbers, but choose an appropriate comparison (99,9% caffeine free drink, 1% of searches…)

  • Line chart for variations, bar chart for absolutes.

  • Cuidado con las escalas de ejes dobles: que siempre empiecen en cero. (p. 218 de đź“– Bullshit. Contra la charlatanerĂ­a)


Graph View

Backlinks

  • Data Analysis

Created with Quartz v4.5.1 © 2025