PCs and Macs?

Gerard Kleywegt gerard at xray.bmc.uu.se
Fri Dec 19 14:11:46 EST 1997


In article <67cocs$im4$1 at dismay.ucs.indiana.edu>,
gilbertd at chipmunk.bio.indiana.edu (Don Gilbert) wrote:

>   
>Biology software and data:  20836  61% of total
>   MS Windows browser: 10302  -- 49% of biology 
>   Macintosh browser :  5095  -- 25% of biology 
>   XWindow browser   :  1679  --  8% of biology 
>   remainder         :  3760  -- 18% 
>   
>This is a quick and dirty look at platform usage. 
>

errrr - Occam's razor says that we should conclude
no more than that there are twice as many accesses
originating from windows machines as there are from
macintoshes within the biosciences

any of the following *could* be concluded (just to
prove that statistics can be used to prove anything)
if one cares to ignore Occam:

- there are twice as many windows as mac machines

- there are an equal number of windows and mac
  machines, but windows browsers crash twice as
  often

- there are an equal number of windows and mac
  machines, but windows machines are shared on
  average by twice as many people

- there are ten times as many mac as windows
  machines, but for every solid application on
  the mac, there are twenty crappy clones for
  windows, each of which is so poor that the
  user, after downloading one, has to get on
  and try the next one

- there are an equal number of macs and windows
  machines, but mac users are clever enough to
  spend only half the time on the web compared to
  windows users, and can do science during the
  hours they save

- there are five times as many macs as windows
  machines, but windows users are one-tenth as apt
  at locating information on the web

- and loads of linear combinations of the above

facts/data is one thing, (des)information another !

--gerard (a proud owner of "How to lie with statistics")

----------------------------------------------------------
Gerard Kleywegt
mailto:gerard at xray.bmc.uu.se
http://alpha2.bmc.uu.se/~gerard/




More information about the Bio-soft mailing list