J. Philip Miller Professor of Biostatistics 
Choosing and fitting statistical models to data occupies a
large proportion of the time of many medical statisticians. It is an activity
that requires the use of many specific and welldefined techniques such as
using a certain algorithm to fit a model or calculate a particular
diagnostic. However, the sequence of techniques used to select a model and
decide its suitability will owe more to experience and judgment than
technical expertise. Statistical modeling provides one of the best
illustrations of Healy's remark that the practice of statistics is not a
science but "that blend of knowledge and practical knowhow" that
he describes as a technology; in this respect statistics has much more in
common with medicine than is often supposed. 


How to Practice Statistics
John Bailar 


Please visit THE place for SAS users www.sasCommunity.org 
My office is in the the BarnesJewish HospitalSouth complex. My office is on the third floor of the Barnard building (just below the GCRC). To get there, just take either the Barnard Elevators (by the entrance to the cafeteria) or the Wohl Hospital Elevators. For a map of the main floor of BJHSouth clinic here.
How a sore ankle led to a broken neck.
Member of Executive Board of SAS Global Users Group
Last updated 3/7/2008