Checking Assumptions for Proportion Intervals
What you'll learn: How to verify that your proportion confidence intervals are trustworthy by checking key conditions.
Why Assumptions Matter
Just like a recipe needs the right ingredients in the right amounts, proportion confidence intervals only work correctly when certain conditions are met. If these assumptions fail, your interval might be too narrow (falsely confident) or completely invalid.
The Three Key Assumptions
1. Independence Each observation must be independent of the others. Think of flipping coins—one flip shouldn't influence the next. This breaks down when you have clusters (surveying multiple family members) or time dependencies (repeated measurements on the same person).
Rule of thumb: When sampling without replacement, your sample size should be less than 10% of the population to maintain approximate independence.
2. Sample Size Requirements For the normal approximation to work (the foundation of standard proportion intervals), you need enough successes and failures.
Rule: Check that both np ≥ 10 and n(1-p) ≥ 10, where n is your sample size and p is your sample proportion.
Example: With n=50 and p=0.15, you have 7.5 expected successes—too few! The approximation breaks down.
3. Random Sampling Your sample should be randomly selected from the population, not a convenience sample or self-selected group.
When Approximations Break Down
Standard intervals fail when:
- Your proportion is near 0 or 1 (extreme proportions)
- Sample sizes are small (under 30)
- Independence is violated
In these cases, use the Wilson Score Interval (which you've learned!) or exact methods instead.
Key Takeaway: Before trusting a proportion confidence interval, always verify independence, check that np ≥ 10 and n(1-p) ≥ 10, and confirm random sampling—otherwise your interval may mislead you.