Re: Biggest Jumbo?

From:         Pete Mellor <pm@csr.city.ac.uk>
Date:         08 Feb 95 01:21:18 
References:   1
View raw article
  or MIME structure

Kit Linder <left@thecurb.ebay.sun.com> writes 06 Feb 95 02:45:22:- 

> Because "biggest" is a subjective term, we agreed to average the largest of
> the following parameters:
> 
> 1. Wingspan   2. Length   3. Max TOW   4. Number of passengers

Mmm ... To start with, how can you average the "largest" of four numbers, 
since that is, by definition, a single number? 

Also (assuming that you actually intend to sum all four quantities and 
divide by four), what would the resulting "average" figure actually *mean*? 
It would tell you nothing about any useful characteristic of the aircraft 
whatsoever, whereas the individual statistics are quite informative. 
Beware of averaging apples and oranges and thinking it tells you 
something about fruit! 

For comparison, consider the following arguments:- 

A. Because "complexity" is a subjective term where software is concerned, 
we agreed to average the following parameters:- 

  1. No. of display screens  2. No. of files  3. No. of inputs 
  4. No. of outputs 

B. Because "safety" is a subjective term where aircraft are concerned, 
we agreed to average the number of fatal accidents since each model was 
first delivered over the number of examples currently in service. 

I am not offering any free breakfasts for guessing that A) I think Function 
Point Analysis is a load of crap, and B) I don't think the A320 is the 
safest airliner ever built! :-) 

Kung hoi fat choi! 

Peter Mellor, Centre for Software Reliability, 
City University, Northampton Square, London EC1V 0HB 
Tel: +44 (171) 477-8422, Fax.: +44 (171) 477-8585, 
E-mail (JANET): p.mellor@csr.city.ac.uk 
-----------------------------------------------------------------------------