Search All of the Math Forum:
Views expressed in these public forums are not endorsed by
Drexel University or The Math Forum.



stats textbooks
Posted:
Feb 29, 1996 1:11 AM


I think it is important to use a "postTukey" statistics textbook  one that reflects the revolutionary changes in statistics associated with the name of John Tukey. Since these changes took place in the 1960's, I could also say I think it is important to use a textbook that is less than 30 years out of date.
I have been thinking about how a high school mathematics teacher who might be teaching an Advanced Placement Statistics course might identify such texts  especially if that teacher does not have a lot of prior training in post(or even pre!)Tukey statistics.
I came up with a few such criteria and checked some textbooks against them, with the results in the table below. Many people have said that the three (or four) R's of modern statistics are
robustness and/or resistance residuals reexpression (or transformation)
so I simply counted the number of references in a book's index to these topics. I did the same for "outliers", which are closely tied to the first two R's. I decided "boxplots" was NOT a good thing to check because they had been added "cosmetically" to a great many books. By that I mean that they are explained once in a brief section early in the book and never used for anything. The books that DO use them do not put each use in the index. So, since the principal use of boxplots is comparing several groups, I counted the number of sets of parallel boxplots in each book's chapter on Analysis of Variance rather thana the number of references in the index. (This also indicates whether the book encourages the reader to LOOK AT THE DATA!)
Outliers Robust/Resist. Trans. Resids. Par.BP
M&M2 16 21 12 >19 4
BPS 24 6 2 13 6
S1 19 13 92 15 6
S2 13 6 18 7 11
J5 0 0 0 0 0
B&B5 2 1 0 2 0
The books are
M&M=Int. to the Practice of Stats. by Moore & McCabe BPS=Basic Practice of Stats., by aforementioned Moore S=Stats.&Data Analysis by Andrew Siegel J=Elementary Statistics by Robert Johnson B&B=Understandable Statistics by Brase & Brase
The first three books are on the recommended list for AP Stats. I have taught statistics courses using every one of them. The other two are books people enquired about on the AP Stats. email list.
This is just one "quick and dirty" rating system meant for rapid screening. Other things to look at are the amount of real data in a book and the extent to which the reader is taught how to check the assumptions underlying the procedures. 
_   Robert W. Hayden   Department of Mathematics /  Plymouth State College   Plymouth, New Hampshire 03264 USA  *  Rural Route 1, Box 10 /  Ashland, NH 032179702  ) (603) 9689914 (home) L_____/ hayden@oz.plymouth.edu fax (603) 5352943 (work)



