[R-sig-teaching] t.test from summary of data

[Greg Snow]

Are you interested in this only for simple one sample t-tests? Or do you need something for the more complicated cases (which lead to discussions about approximations to degrees of freedom, etc.)?

For the simple one sample, the test can be done as a one liner:

> pt( mean/(sd/sqrt(n)), n-1 )

Which is not that much more complicated than a modified t-test function would be:

> t.test.summary( mean, sd, n )

Using the pt function would then force the students to think about which tail they are computing the area for and whether the p-value needs to be multiplied by 2 (rather than just specifying an argument).

For the 2 sample case, I have in the past just simulated a dataset with the same means and standard deviations, then used this in place of the raw data, the mvrnorm function in the MASS package has an option to generate normal data with the exact mean and variance specified.

-- 
Gregory (Greg) L. Snow Ph.D.
Statistical Data Center
Intermountain Healthcare
greg.snow at imail.org
801.408.8111


> -----Original Message-----
> From: r-sig-teaching-bounces at r-project.org [mailto:r-sig-teaching-
> bounces at r-project.org] On Behalf Of Byungchul Cha
> Sent: Monday, November 08, 2010 10:26 PM
> To: r-sig-teaching at r-project.org
> Subject: [R-sig-teaching] t.test from summary of data
> 
> I am using R in my intro stat course, and, I myself is a R novice.
> 
> We use "Intro to the practice of statistics" by Moore, McCabe, Craig as
> textbook. Is there an option for the command "t.test" so that it can
> perform t-test from the summary of sample data (that is, sample mean,
> sample sd, and sample size), instead of the (raw) sample data? I've
> been looking to find a way to do so, but, I couldn't.
> 
> I believe it is possible to write codes to do this for my own use, but,
> I am just curious. Is there a pedagogical reason not to do so? This is
> a bit annoying when all the exercises problems in the book needs such a
> functionality, and I can't figure out how to do it in R, especially
> when many other major stat packages (including TI calculators) have
> that capability. What am I missing here?
> 
> --
> Byungchul Cha
> Muhlenberg College
> 
> _______________________________________________
> R-sig-teaching at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-sig-teaching