Independent Samples t-Test - Comparing Two Probability %

From Q
Jump to navigation Jump to search

Where a^1 and a^1 are two probability % and n1, n2 are their effective sample sizes and se1 and se2 are their standard errors, the test statistic is:

t=a^1−a^2se12+se22,

where:

p=2Pr⁡(tv≥|t|),
v=(se12+se22)2se14/(n1−b)+se24/(n2−b),
b=Bessel's correction.