Omega Squared

Problem

Return omega squared (SS_between - df_between*MSW) / (SS_between + SS_within + MSW), where MSW = SS_within/df_within. Return 0.0 when df_within <= 0 or the denominator is zero. It can be negative when the effect is smaller than sampling noise.

Hint

Unlike eta squared, omega squared corrects for the positive bias.

Test cases

case 1

in: [54,6,2,6]

expected: 0.8524590163934426

case 2

in: [0,10,1,8]

expected: -0.1111111111111111

case 3

in: [24,0,1,4]

expected: 1

case 4

in: [2,2,1,2]

expected: 0.2

case 5

in: [54,6,2,0]

expected: 0

Starter codePython

def omega_squared(ss_between, ss_within, df_between, df_within):
    # Your code here
    pass

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