Better documentation of statistical analyses

[GoogleCodeExporter]

What steps will reproduce the problem?
1. There are three lines of "statistical analyses". The documentation only 
refers to one, the Hillis method. What are the other analyses?

Could you tell me what method is used for the first row of the hypothesis 
testing information? Alternatively, if the Hillis method is the more accurate 
method that you recommend I use, then I can change the values in my manuscript 
to use the p-values and confidence interval from this method, as the 
interpretation of my results will not change.

Response: 
The first line assumes the test statistics follows a normal distribution. The 
confidence interval is a bit narrower than it should be and p-value is a bit 
smaller than it should be.
The second line assumes the test statistic follows a t-distribution, with the 
degrees of freedom method outlined in Obuchowski2012_Acad-Radiol_v19p1508 . The 
validation of this is not documented yet in the literature.
The third line assumes the test statistic follows a t-distribution, with the 
degrees of freedom method outlined in Hillis2008_Acad-Radiol_v15p647 .

For now, I would refer to the Hillis paper and results.

Original issue reported on code.google.com by Brandon.Gallas on 2 Jun 2014 at 3:58