[prev in list] [next in list] [prev in thread] [next in thread] 

List:       kde-look
Subject:    Re: Icon Usability
From:       "Tom Hoferek" <tomh () corel ! com>
Date:       2001-03-22 19:25:18
[Download RAW message or body]

This is a noble idea, however there is the possibility that conclusions
formed from the data you collect may be misleading unless you take some
important things into consideration and deal with them in the appropriate
manner. You may be doing so, but since I don't know you and your e-mail
hasn't gone into any specifics please bear with me. If all of what I say
below is just stating the obvious to you, great, I'll be looking forward to
seeing your findings.

You want to determine, "...how well icons are perceived by users." Are you
planning on getting accurate profiles of each user that participates? This
is critical to ensure that your data has any meaning, otherwise you will
lump together users who have seen some of the icons with those that have
never seen them and those that have seen all of them and those that may
have seen similar icons on other platforms, etc., etc., etc.. Without
knowing these facts you simply cannot make any conclusions about how the
icons are perceived.

How are you going to take into account context? Some icons will not be
recognizable out of context but may be very effective when presented in
their normal context. If someone doesn't guess the meaning of the icon when
seeing it out of context that doesn't necessarily mean the icon is
ineffective. The same user may identify the meaning of the icon correctly
when seeing it in its normal context.

While some are literal, others are abstract. Abstract icons often represent
concepts that are extremely difficult to convey in a simple (literal)
graphic, thus the meaning of the abstract image must be learned. That's OK,
because once learned, if the icon is effective, it will be immediately
recognizable and therefore will well serve its purpose. So just because
someone doesn't recognize an icon when they first see it, it doesn't mean
there is something wrong with the icon. It could be that there is no better
alternative. For example, the image of  a floppy disk is recognized by many
people as the command, "Save". This image could be interpreted to mean
other things. To people just beginning to use computers, the image may not
be recognizable at all (they may have never seen a floppy disk), but they
will quickly learn what the image means. Traffic signs are another good
example. The language of traffic signs is easy to learn and the signs
effectively communicate their meaning to their users.

I suggest you design your study very carefully and consult a usability
specialist while doing so to be sure the study delivers useful data to you.
If your study is not properly designed or carried out the data may be
useless. Worse yet, if the data is inaccurate and KDE developers and
artists assume the conclusions of the study are accurate, well that would
be a great disservice.

Good luck,

Tom

Jono Bacon wrote:

> Hello all,
>
> Today while in a usability lecture at Uni, Lee and myself had an idea
> of building into the study some icon researach and how well icons are
> percieved by users. It seems we can learn much from Jakob Nielson in
> terms of iconic research.
>
> The idea is this: we will provide a form on the site displaying all of
> the icons that are included with KDE, and next to each icon will be a
> text box. The user will simply enter words that they associate with the
> icon. This research will give us clues on how the user percieves the
> image, and how cultural differences may be taken into account. At the
> top of the page the user can enter which country they are from.
>
> As far as I am personally thinking, I feel the icons in KDE are
> fantastic, but any extra information on user reaction and input will be
> useful.
>
> Unfortunatly we cant hack at this until I recieve confirmation of the
> kde.org webspace for the study. When we recieve confirmation we can
> build the database to house this research.
>
> I have CC'ed this message to kde-look and kde-artist as people on these
> lists may be interested.
>
> Comments?
>
>         Jono
>
> Jono Bacon - [vmlinuz] -- jono@kde.org
> KDE/Qt Developer

[prev in list] [next in list] [prev in thread] [next in thread]