Finding P-Values for Proportion Tests
What you'll learn: How to interpret the z-test statistic for proportions and calculate the corresponding p-value to measure evidence against the null hypothesis.
Understanding the Test Statistic
After calculating your z-test statistic for a proportion test (comparing your sample proportion to a hypothesized value), you need to translate that number into a p-value. The test statistic tells you how many standard errors your sample proportion is from what you'd expect under the null hypothesis.
For example, if z = 2.3, your observed proportion is 2.3 standard errors above the null hypothesis proportion. But what does that mean in terms of evidence?
From Z to P-Value
The p-value answers: "If the null hypothesis were true, what's the probability of getting a test statistic at least as extreme as what we observed?"
Since proportion test statistics follow the standard normal distribution (when conditions are met), you use the z-distribution to find p-values:
Two-Tailed Tests
If testing H₁: p ≠ p₀ (any difference), you want extreme values in both directions:
- Find the probability beyond |z| in both tails
- P-value = 2 × P(Z > |z|)
One-Tailed Tests
If testing H₁: p > p₀ (greater than):
- P-value = P(Z > z)
If testing H₁: p < p₀ (less than):
- P-value = P(Z < z)
Interpretation
A smaller p-value means your observed proportion is more unusual under the null hypothesis, providing stronger evidence against it. You compare this p-value to your chosen significance level (α) to make your decision.
Key Takeaway: The p-value for proportion tests translates your z-test statistic into a probability using the standard normal distribution, with the calculation depending on whether you're conducting a one-tailed or two-tailed test.