Dear all
I am trying to understand the different techniques that can be used in
comparing and benchmarking the performances of companies, business
units, departments, etc., both internally and against similar external
comparators. My concern has always been that many statistical methods
confuse the gaps between error and "structural" differences and hence
there is a danger of over-estimating what can be achieved by choosing
outliers as benchmarks.
A lot of companies are happy to charge large sums of money to
benchmark performance by comparing your performance to those in their
existing dataset. Whilst this is not an entirely unreasonable
approach, I wondered if there was already material about using
statistical methods to improve the robustness of benchmarks. This
seems to be a topic that is more popular in econometrics than
statistics and I would be keen to hear the experiences of practitioners
of both!
I will happily provide an overall summary of the answers I receive to
either individuals or the list as a whole if there is enough interest.
Thanks in advance for your help,
David Smallbone
News, Sport, Celebrity Gossip and Lifestyle Offers from Tiscali - http://www.tiscali.co.uk/
_________________________________________________________________________
|