Comparison of ‘Power For Comparing Two Means’
to other public software programs and statistical textbook.
Date last tested: 7/11/05.
|
CI %, |
Mean difference (Group1 – Group2), n1, s1, n2, s2, |
|||
Data input |
95%, 5, 100, 15.34, 100, 18.23 |
95%, -0.24, 50, 0.64, 50, 0.76 |
95%, 5, 100, 18.23, 100, 15.34 |
95%, 5, 30, 18.23, 30, 15.34 |
95%, -4, 30, 18.23, 30, 15.34 |
OpenEpi |
55.52% |
40.06% |
55.52% |
20.89% |
14.89% |
Rosner |
55.5% |
40.1% |
NA |
NA |
NA |
PEPI |
55.1% |
? |
55.1% |
20.35% |
? |
Of
Statistics |
55.1% |
39.4% |
55.1% |
20.3% |
14.5% |
‘CI’ Confidence
interval; ‘n1’ sample size of
Group 1; ‘s1’ standard
deviation of Group 1 mean; ‘n2’ sample size of Group 2; ‘s2’ standard deviation of Group 2
mean;
? Wrong result due to negative mean difference
Reference:
·
Bernard Rosner. Fundamentals of Biostatistics (5th edition).
· PEPI: Abramson JH and Gahlinger
PM (2001). Computer program for epidemiologists: PEPI version 4.0. Sagebrush
Press,
·
UCLA Department Of Statistics. (http://calculators.stat.ucla.edu/powercalc/normal/n-2-unequal/n-2-uneq-var-power.php)