CLT for Proportions
What You'll Learn
You'll discover how the Central Limit Theorem extends to sample proportions, making them predictably normal even when individual outcomes are just yes/no.
The Core Idea
Until now, you've seen the CLT apply to sample means. But what about proportions—like the percentage of customers who click an ad, or the fraction of defective products?
Great news: the CLT works for proportions too!
When you calculate a sample proportion p̂ (p-hat, the fraction of successes in your sample), it behaves like a sample mean. As your sample size grows, p̂ becomes approximately normally distributed around the true population proportion p.
The Formula
For a sample proportion p̂:
- Mean: The center of the distribution is the true proportion p
- Standard Error: The spread is √(p(1-p)/n), where n is your sample size
This standard error shrinks as n increases (just like with means), making your estimate more precise.
Why This Matters
Think of surveying voters: if 60% of all voters support a candidate (p = 0.60), and you sample 100 people, your sample proportion p̂ won't be exactly 0.60—but it will cluster around 0.60 in a normal bell curve with standard error √(0.60 × 0.40 / 100) ≈ 0.049.
This allows you to build confidence intervals and make predictions about proportions using the same normal distribution tools you already know.
Key Takeaway: Sample proportions follow an approximately normal distribution with mean p and standard error √(p(1-p)/n), letting you apply all your normal distribution knowledge to yes/no data.