Sampling Distribution of the Sample Variance
You'll learn how the variance of different samples from the same population forms its own predictable distribution.
The Core Idea
When you draw a sample and calculate its variance, that value varies from sample to sample. Just like the sample mean has a sampling distribution, so does the sample variance. If you took many samples from a population and computed the variance of each, those variances would cluster around the true population variance (σ²), forming a distribution with its own shape and properties.
How It Works
Imagine a population with a true variance σ². When you:
- Draw a random sample of size n
- Calculate the sample variance (s²)
- Repeat this process many times
...you'll get a collection of sample variances that form the sampling distribution of the sample variance.
Key characteristic: Unlike the sampling distribution of the mean (which becomes normal thanks to the CLT), the sampling distribution of the variance follows a chi-squared distribution when the population is normal. The sample variance tends to be an unbiased estimator of the population variance when calculated with the n-1 divisor.
Why This Matters
Understanding this distribution helps you:
- Build confidence intervals for population variance
- Test hypotheses about variability
- Understand why we use n-1 instead of n in the sample variance formula
- Assess the reliability of variance estimates from your sample
Key Takeaway: Sample variances from repeated samples form their own distribution around the true population variance, and this distribution's properties (especially its connection to the chi-squared distribution) enable us to make statistical inferences about population variability.