FW: [plt-scheme] Statistics (V301.5 Speed Up)

From: Matt Jadud (mcj4 at kent.ac.uk)
Date: Sat Feb 11 04:15:35 EST 2006

Statistics is not a "plug-n-chug" endeavor. One of my favorite resources 
to date on exploratory data analysis comes from NIST:

http://www.itl.nist.gov/div898/handbook/eda/eda.htm

I also liked Trisha Greenhalgh's "How to read a paper: The basics of 
evidence-based medicine", a book (surprise!) on interpreting reports and 
results in the area of evidence-based medicine. The full text seems to 
be available here:

http://bmj.bmjjournals.com/collections/read.shtml

and the two articles "Statistics for the non-statistician" and 
"Statistics for the non-statistician II" are good; the remainder of the 
book requires some interpretation to apply out of context, but it is 
still a good resource while performing EDA on unknown datasets.

http://bmj.bmjjournals.com/cgi/content/full/315/7104/364
http://bmj.bmjjournals.com/cgi/content/full/315/7105/422

Sadly, I can't join further in the First International Scheme Contest 
for One Statistical Beer, as I need to take off for parts uknown for the 
day.

*sigh*

No free beer for me.

M

Williams, M. Douglas wrote:

> I ran 1000 runs and made a histogram of the results.  Unfortunately, there
> is one value out at 16353 while the others were between about 6000 and 8000.
> Which made the histogram difficult to interpret.  I have attached the code
> and the resulting histogram.


Posted on the users mailing list.