Path: utzoo!attcan!uunet!ginosko!gem.mps.ohio-state.edu!pacific.mps.ohio-state.edu!tut.cis.ohio-state.edu!mailrus!ames!pacbell!pbhyg!ria
From: ria@PacBell.COM (Richard I Anderson)
Newsgroups: comp.cog-eng
Subject: Re: Menu Interaction Techniques
Message-ID: <1788@pbhyg.PacBell.COM>
Date: 27 Sep 89 18:13:35 GMT
References: <2722@trantor.harris-atd.com>
Reply-To: ria@PacBell.COM (Richard I Anderson)
Organization: Pacific * Bell, San Ramon, CA
Lines: 87

In article <2722@trantor.harris-atd.com> chuck@trantor.harris-atd.com (Chuck Musciano) writes:
>
>     I am currently in the midst of a quandary involving two different
>menu interaction techniques, and I was wondering if there is anything in the
>literature regarding either of the following two hypotheses.
>
>     I have been of the opinion that menus used in an interface should remain
>"static": the position and contents of the menu should not change over the
>course of using the interface.  ...

A relevant study is reported in:
  Mitchell, J. & Shneiderman, B. (April, 1989) Dynamic versus static menus:
  An exploratory comparison. SIGCHI Bulletin, 20 (4), 33-37.

This study did NOT investigate menus that changed in accordance with
context, but it did look at menus that changed versus menus that did not.

Ablex is publishing a book by Kent Norman entitled, "The psychology of
menu selection..."; I would hope that it addresses the issues of concern
to you.

I collected some somewhat relevant data via usability testing I did early this
year.  Menu item position did not change in accordance with context, but
cursor placement within the menu did.  The initial "qualitative" analysis 
suggested that this change prompted user confusion.  However, subsequent
examination of the tapes and preliminary quantitative analyses suggest
that the detected confusion may have actually been due to design weaknesses
encountered earlier and that the changing cursor placement was significantly
beneficial.

Jonathon Grudin has a nice discussion about such menus in:
  Grudin, J. (January, 1989)  The case against user interface consistency
  (Technical Report Number ACA-HI-002-89)  Austin, TX: Microelectronics and
  Computer Technology Corporation.

In article <2722@trantor.harris-atd.com> chuck@trantor.harris-atd.com (Chuck Musciano) writes:
>                           ...  I was under the impression that menu usage
>becomes an instance of "gesturing" or "stroking" and the cognitive loading
>is moved into the "muscle memory" of the user.

In article <618@cs.yale.edu> engelson@cs.yale.edu (Sean Engelson) writes:
>Supposedly there's been some research that's shown that pie-menus
>(menus with selections in wedges around a circle) allow muscular
>memory to take over.  My experience with pie-menus under NeWS bore
>this out ...

Aah - muscle memory again.  Don Norman talked about muscle/motor memory
the last time this topic surfaced here.  I happened to have saved his words:

    >From: norman@sdics.ucsd.EDU (Donald A. Norman)
    Newsgroups: comp.cog-eng
    Subject: Re: A Dvorak keyboard experiment
    Date: 24 Jun 88 17:10:56 GMT
    Organization: UC San Diego Institute for Cognitive Science
    
    ...
    
    Do not get too excited about the advantages of "motor memory."
    Nothing special about "motor."  The same phenomenon works with
    anything that is over-learned, over-practiced.  Such skills become
    automated and, thereby, sub-conscious.  These skills can be done with
    minimal interference to other ongoing tasks, with little or no
    conscious attention, and with great precision and speed.  And once a
    skill reaches this level of automation, it is very difficult to
    change.
    
    PIE MENUS:
    
    I suspect that pie-menus are good things, but not because they are
    motor.  Rather, they are good because items appear in a consistent
    place and it is easy for a single movement to select them.
    
    Regular pop-up or pull-down menus may be consistent, but visual
    attention is needed to find the desired target.  Of course, pie menus
    only work with a limited number of entries (so that the selection
    stroke does not require high angular precision).  I predict that if
    you add too many entries, they will require visual attention just like
    regular menus.
    
    Don Norman


R. I. Anderson
Human Factors Consultant
Pacific Bell
2600 Camino Ramon, Room 2E850                     (415)823-3715
San Ramon, CA 94583                           ria@pbhyg.PacBell.COM