home *** CD-ROM | disk | FTP | other *** search
- Comments: Gated by NETNEWS@AUVM.AMERICAN.EDU
- Path: sparky!uunet!paladin.american.edu!auvm!FRANKSTON.COM!MEREDITH_WARSHAW
- Message-ID: <199208281543.AA04340@world.std.com>
- Newsgroups: bit.listserv.stat-l
- Date: Fri, 28 Aug 1992 11:37:00 -0400
- Sender: "STATISTICAL CONSULTING" <STAT-L@MCGILL1.BITNET>
- From: Meredith_Warshaw@FRANKSTON.COM
- Subject: medical stats texts
- X-To: stat-l@frankston.com
- Lines: 81
-
- Awhile ago, I sent out the following message: "A surgeon (actually, my
- step-brother) has requested that I recommend an easy, preferably short, intro
- to medical stats. Basically, he wants to understands t-tests, chi-squares,
- and lifetables so he can read research literature. Does anyone out there
- have any recommendations?". More replies have come in since last I
- summarized, so here is the latest list. Thanks again to all who took the
- time to reply. Meredith
-
- *****************************************************************
- There is a little book by Duncan, Knapp, and Miller published by Wiley. I
- think it's in a second or third edition now. No frills, just the recipes and
- some discussion of the most common methods.
- Russ Lenth rlenth@stat.uiowa.edu
- *****************************************************************
- I recommend the following:
- Norman, Geoffery; Striener, David. PDQ Statistics, 1986, BC Decker, Inc.
- 165pp.
- Striener, David, Norman, Geoffery; Munroe Blum, Heather. PDQ Epidemiology.
- 1989. BC Decker, Inc. Philadelphia. 119pp.
- Written by members of the teaching faculty in Epediology/Health Sciences at
- McMaster University, these books offer a concise introduction to the
- subject. Blended with examples and humour, they are also entertaining to
- read.
- Jay Bodkin
- Statistics and Graphics Consultant
- McGill University Computing Centre
- Montreal, Quebec
- *****************************************************************
- In particular, for someone like a surgeon who is not likely to have to DO
- statistics (for he/she can
- easily afford to pay a proper professional to do the actual work), I wish
- to recommend
-
- William L. Hays, Statistics, Fourth Edition, Holt, Rinehart, & Winston,
- 1988.
-
- Hays writes in " ... an attempt to give the elements of modern statistics
- in a relatively nonmathematical
- form, but in somewhat more detail than is customary in texts designed for
- psychologists,
- and with considerably more emphasis on the theoretical rather than than
- the applied aspects of the
- subject. It is designed as a text for at least an intermediate level of
- difficulty. ..."
-
- Hayes gives a fine exposition of distributions (normal, chi-square,t, F)
- and their relationships, of
- expectations, etc., form the pointof view of the professional
- statistician, not that of the hapless
- non-statistician struggling to find the right equation into which to pour
- some numbers. I find that any
- of my clients who have mastered Hays can much more readily comprehend the
- theoretical
- niceties which permeate virtually any analysis of real, messy data; it
- makes my job much easier and
- prepares them much better to be able to defend their own papers at the
- meetings. I'm not
- terribly upset that my (say) surgeon friends are not doing the messy
- details of the analyses that
- frequently require the advice of five or six PhD statisticians and thre
- or four statistical
- computing pros to get through. If they but understand the principles of
- WHY we had to work so hard
- to produce one little p-value and one sentence for the results section,
- then I'm better
- satisfied; I think Hays helps them a great deal with that.
-
- Phil Gallagher uphilg@unc
- *****************************************************************
- From: MDA1%NIORDS1.EM.CDC.GOV@VM1.MCGILL.CA
- I believe that the British Medical Association publishes a book (or books) on
- stats written specifically
- for physicians. Check out the British Medical Journal for ads. They
- definitely used to be advertised there. I believe the material is well
- regarded.
- *****************************************************************
- The British Medical Assn. published a book called "Statistics at Square One"
- in the late 1970s. I don't know if it is still in print. It was based on a
- series of articles in the BMJ (15?), at a very basic level, but very good as
- an introduction.
- Bob Parker
-