Grouping and Chunking -- Jared Spool

Jared Spool, in Web Site Usability, A Designer's Guide, copyright 1999
Morgan Kaufmann Publishers, Inc, ISBN 1-55860-569-X

HB: I paraphrase some of the copyright material. There is much more worth-while
material in the book.

1. Users don't form mental models of sites, they tend to move on from
where they are, more often than going back.

HB: As there is so little similarity among sites, and the dwelling during
many browsing sessions jumps among sites, the value in learning the structural
model of a site is diminished. A user of a portal site may recall its
structure, but not likely the structure of the many other visited sites.

2. Chunking

2.1 Use of hierarchy may impose too-general chunk names linked to more
details. Those generic names may confuse the user, whose target may
not be obviously contained in any of those chunk names.

2.2. Table of contents structure and substructure can show chunking.

HB: particularly if the table of contents can be collapsed/expanded as
a dynamic user option, though starting with the collapsed form has the
same problem with the too-general chunking naming.

2.3. "You are Here" indication within chunking hierarchy, repeating the
chunk names was not so useful as expected.

3. Link Naming

3.1. "Descriptiveness Aids Prediction" Effective link descriptions allow
the user to anticipate what the link is about. This verbose approach
was found effective on the old Edmunds site, pictured in the book as
having a list of 8 links.

HB: Now

     www.edmunds.com

has been redesigned from a long list into a "contemporary" site, with table
layout but still most links are well-described, as are the alt texts.
But my visit left me with two pop-up windows to separately close.

3.2. Ambiguous link names -- Edmund used cross links so that a user that 
wanders
astray can get back to the desired place without necessarily backing out.

3.3. Distinctly partition information space. A user may choose by process of
elimination if none are obviously the proper choice.

3.4. Avoid embedded links in paragraphs. Surrounding text confuses.

3.5. Avoid wrapped (or wrappable under user browser setting) links --
subject to incomplete selection.

4. Link Targets

4.1. Links with targets within the same page are contrary to user
expectations. Short individual pages avoid the need for these. A special
case: an early link to the "content" skipping over a long list of links.

4.2. Links going "off-site" have different navigation models, and can confuse.

5. Readability and the Web

5.1 Testers were given questions and sites to find answers to them.

5.2 Users skim, to eliminate irrelevance to current goal.

5.3 Counter-intuitive result: less readable sites lead to user success
using them. The Gunning Fog index provides the highest correlation with 
success
in using a site. Larger values indicate harder to read -- and hence higher 
user satisfaction. The Fog index 8-step algorithm is given. Also compared are
the Flesch Reading Ease and Flesch-Kincaid Grade level indexes.

[HB: I expect the test results are biased by how they chose the users to be
testers. All had experience using the web. Each tested for only three hours.
The record does not indicate any other measure of user background.]

6. Afterword:

6.1. I went back and chunked (7 +/- 2) the above. [1]

6.2. There are many other topics discussed in Jared's book. I consider the
work important, and the reporting well organized. Other usability test
designs may well differ in the conclusions.

----

[1] We have discussed the number of related ideas that are comprehensible
in short-term memory. The classic reference is:

The Magical Number Seven, Plus or Minus Two: Some Limits on our
Capacity for Processing Information
George A. Miller (1956)
Harvard University
First published in Psychological Review, 63, 81-97.
reproduced with permission, by Stephen Malinowski:
     http://www.yorku.ca/dept/psych/classics/Miller/

It argues that 7 +/- 2 is the useful range for chunking related concepts.

Regards/Harvey Bingham

Received on Monday, 24 July 2000 01:27:29 UTC