Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!watmath!clyde!burl!ulysses!mhuxl!ihnp4!inuxc!pur-ee!uiucdcs!uicsl!preece
From: preece@uicsl.UUCP
Newsgroups: net.flame
Subject: Re: Re: Kennesaw vs. Morton Grove - (nf)
Message-ID: <5344@uiucdcs.UUCP>
Date: Fri, 3-Feb-84 23:10:42 EST
Article-I.D.: uiucdcs.5344
Posted: Fri Feb  3 23:10:42 1984
Date-Received: Wed, 8-Feb-84 05:38:11 EST
Lines: 18

#R:pyuxvv:-19100:uicsl:4300135:000:656
uicsl!preece    Feb  3 10:08:00 1984

	>  But there was also one armed robbery and one rape, representing
	>a 100 percent increase for the year in those crimes.
	
	Does this mean that in the previous year there was 1/2 a rape and 1/2
	an armed robbery?!?
----------
While we usually take the change relative to the starting amount rather than
the ending amount, this does give us a pathological case.  We have no way of
stating the percentage change from 0 to 1 as a percentage of zero.  I'm not
a statistician, so I don't know what is usually done with such cases. Clearly
in this case the change is of interest, but there's no good way to
express it.

scott preece
ihnp4!uiucdcs!uicsl!preece