st: Output differences, Stata 11 vs Excel 2010

[Yerik Kaslow <yerik.kaslow@gmail.com>]

I apologize, copy+paste lead to sending before I had finished
displaying the data. Sorry!

Hi Statalisters,

I've noticed this difference several times before, and it's always
confused a troubled me a little.

I have clinical trial data I'm analyzing, ttest output is below. The
means are nearly the same, but the standard deviations are different.
I've also noticed this also with p-values. The differences are always
small, but if it's right at the margin of significance, it can
obviously make a difference.

Does anyone have insight as to why when I run tests of the exact same
data using Stata, then using Excel, I get slightly different results?

Any insight is appreciated. Thus far I've been taking Stata's output
over Excel's. I apologize for the bunching together of the output,
plain text mode is what it is.

Thanks in advance,

Excel Output:

N Int N Con Total N Mean Int Mean Con Diff Int Con SD Int SD Con Pvalue =
CD1 68 58 126 6.49 6.45 0.04 2.70 2.21 0.9344
CD2 68 58 126 7.65 7.41 0.23 2.27 2.07 0.5539
CD3 68 58 126 8.01 7.90 0.12 2.12 2.06 0.7542
CD4 68 58 126 7.29 7.22 0.07 2.34 2.01 0.8597
CD5 68 58 126 7.74 6.45 1.29 2.13 2.49 0.2262
CD6 68 58 126 7.74 7.31 0.42 2.15 2.24 0.2841

Stata Output:

N Int N Con Total N Mean Int Mean Con Diff Int Con SD Int SD Con Pvalue =
CD1 68 58 126 6.49 6.45 0.04 2.37 1.97 0.9252
CD2 68 58 126 7.65 7.41 0.23 2.14 1.66 0.5012
CD3 68 58 126 8.01 7.90 0.12 2.00 1.82 0.7313
CD4 68 58 126 7.29 7.22 0.07 2.11 1.77 0.8425
CD5 68 58 126 7.74 6.45 1.29 1.99 2.29 0.1015
CD6 68 58 126 7.74 7.31 0.42 1.89 1.98 0.2223


-- 
Yerik Kaslow
yerik.kaslow@gmail.com
*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/faqs/resources/statalist-faq/
*   http://www.ats.ucla.edu/stat/stata/