Back to projectPublic page

Sample Bias: Why Polls Get It Wrong

Understand how sampling errors led to wrong election predictions and business decisions

Part of Data-Driven Decisions

4 blocks0 nested pages
Last updated Oct 29, 2025. Clone to remix or explore the blocks below.
aa8880db...
deba4807...
content

The 1936 Presidential Poll Disaster

Educational content slides

The 1936 Presidential Poll Disaster

Literary Digest magazine's massive survey:

  • Sent 10 million survey forms
  • Received 2.4 million responses (largest poll in history)
  • Confidently predicted: Alf Landon wins by landslide

Actual election result: Franklin D. Roosevelt won by actual landslide (60.8% of vote)

What went wrong? They surveyed people from car registrations and phone directories. In 1936, only wealthy people owned cars and phones. Wealthy people voted Republican. The majority (poor/middle class) voted Democrat but weren't in the sample.

The lesson: Sample size doesn't matter if your sample is biased.**

e34bcb93...
quiz

Quiz: 3 Questions

Test your understanding with this quiz.

0 / 3

YouTube posts a Community poll asking 'Should we remove the dislike button?' 85% vote YES. What's the sampling problem?

e5840f5a...
feedback

Detect and Fix Sample Bias

Complete this exercise and get AI-powered feedback.

Detect and Fix Sample Bias

Sample bias is everywhere once you know to look for it—reviews, surveys, social media polls.

Think about: Who responds vs who doesn't? What groups are systematically excluded? How would you capture the silent majority?

This reveals: How to design better surveys and question any data based on biased samples.