Bias of the MLE Sample Variance

Problem

The maximum likelihood sample variance (dividing by n instead of n - 1) is biased. For a true variance sigma_squared the bias is:

bias = -sigma_squared / n.

Return the bias, and 0.0 for n <= 0.

Hint

The MLE underestimates by a factor of (n - 1) / n.

case 1

in: [10,1]

expected: -0.1

case 2

in: [5,2]

expected: -0.4

case 3

in: [1,0.5]

expected: -0.5

case 4

in: [0,1]

expected: 0

Starter codePython

def sample_variance_bias(n, sigma_squared):
    # Your code here
    pass

4 test cases · run your code against them in the editor