[R] OT: Trends in statistical computing?

Joshua Wiley jwiley.psych at gmail.com
Fri Mar 18 06:34:32 CET 2011


Dear List,

First off, this is completely off topic, but I thought others might
find it interesting.  The good people at IBM are apparently providing
written interpretations of results now.  I was working with a student
the other day with simple nonparametric tests using SPSS, and under
the nonparametric test tab, it defaulted to something like "Let SPSS
decide which test to use".  Even when a test was manually selected,
the output was (of course formatted in one of SPSSs (in)famous
tables):

Null Hypothesis: The distribution of DVtest (the variable) is the same
across categories of group.
Test: Independent-Samples Mann-Whitney U Test
Sig: .246
Decision: Retain the null hypothesis.

In a similar vein, here is a quote from Wikipedia
(http://en.wikipedia.org/wiki/Q_research_software) about Q:

"Q uses expert systems that automatically select appropriate
significance tests based on the structure of the data"

Has anyone seen similar trends elsewhere?  It gave me renewed
appreciation for the thoughtful consideration of the R developers in
even making some tests convenient to access (e.g.,
http://r.789695.n4.nabble.com/Standardized-Pearson-residuals-td3355110.html
).

Cheers,

Josh


-- 
Joshua Wiley
Ph.D. Student, Health Psychology
University of California, Los Angeles
http://www.joshuawiley.com/



More information about the R-help mailing list