Table of Contents

• Topics
  • Why we divide by n-1 for the unbiased sample

Topics

Why we divide by n-1 for the unbiased sample

• Since we usually don't know the real \(\mu\), the population mean.
• So, we resort to use the sample mean to calculate the sample variance.
• The problem is:
  1. Since the sample size is smaller than the population size
  2. it is more likely that the values are more close to each other.
  3. Therefore, \(\sum(x - \bar{x})^2\) is usually smaller.
• To compensate this, divide by \(n-1\), which results in the larger value.