Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: version B 2.10.1 6/24/83; site dciem.UUCP
Path: utzoo!dciem!mmt
From: mmt@dciem.UUCP (Martin Taylor)
Newsgroups: net.cog-eng
Subject: Re: A good human-factored user interface for postnews
Message-ID: <444@dciem.UUCP>
Date: Tue, 25-Oct-83 10:19:30 EDT
Article-I.D.: dciem.444
Posted: Tue Oct 25 10:19:30 1983
Date-Received: Tue, 25-Oct-83 10:21:49 EDT
References: <952@utah-gr.UUCP>
Organization: D.C.I.E.M, Toronto, Canada
Lines: 10

Mark Weiser's comment about 7 +- 2 being wrong is correct, but his
correction is wrong. There is no real upper limit; it depends on
the circumstances (other cognitive load) and on the material being
remembered. A series of studies at this laboratory (J. Mackworth
and co-workers in the 1960's) showed limits of between 2 and 14
for the same short-term memory task using different materials and
loads. The magic number 7 +- 2 should have been revised long ago
to say 8 +- 6.

Martin Taylor