Defining the mental model - Expectations

24 Feb 2005 - 5:06pm
9 years ago
5 replies
1659 reads
Jess McMullin
2004

Gerard wrote:
>Mental models have already been defined at at a high level by Don
Norman. See the Design of Everyday Things, p. 189-90.

I think it's great that Austin (and others) aren't content with a definition
that's a couple decades old...it's how we're going to create a better
understanding of our work.

More to the point, Norman doesn't actually define what makes up a user's
mental model in POET - he just says it's "what the user develops to explain
the operation of the system", and goes on to say that the designer and user
only communicate through the 'system image' and that the user develops their
model through interaction with the system. But is that all there is? The
user brings nothing to the table, they're in a state of some sort of newbie
tabula rasa, just waiting to be imprinted by the system?

Norman's model, as it's presented in POET, is designer and system centric -
it doesn't acknowledge what the user brings to the table, via expectations,
goals, culture, past experience, etc. Austin's model shows that there are
things going on in the users' minds before system interaction.

I'm not unbiased myself - I think that 'Expectation' is an incredibly useful
concept for discussing mental models.

See this snippet of the start of a presentation from 2001 (IE only, I'm
afraid)
http://www.interactionary.com/files/xgap/

and more recently, from 2003-2004 some thoughts on the fact that mental
models, and expectations, are developed through iteration, as part of the
user's experience cycle:

http://www.nform.ca/files/experience_cycle.pdf

Peter Merholz has his own thoughts about expectations and Explicit Design
http://www.peterme.com/archives/000333.html

Most interesting for the "we've been there, done that" point of view is that
Don Norman started talking about "Expectation Design" at last year's NNGroup
tour. (snippet from Peterme's blog)

Expectation Design: The Next Frontier
Don Norman
Good designers already know how to make products attractive (visceral
design) and how to appeal to self- and brand-image (reflective design). Good
behavioral designers know how to make products usable and
understandable--indeed, that's the focus of most of this conference. It's
time now to turn our attention to pleasure and fun. Here, the challenge for
designers is behavioral design, where expectations drive emotions. This is
where hope and fear, and satisfaction and anger reside. Deliver on positive
expectations and people experience pleasure. Deliver something different
than expected, but equally satisfying, and people have fun. Fail to deliver,
or leave people feeling out of control, and you get a wide range of negative
emotions.

Expectation-driven design marks a new dimension for our discipline and
provides a new framework for design. It shifts the emphasis from pure
function to an emphasis on designs that both function well and offer people
pleasure, enjoyment, a sense of accomplishment, and yes, even fun.

Cheers,

Jess

--------------
Jess McMullin
Principal
nForm User Experience
www.nform.ca

Comments

24 Feb 2005 - 8:46pm
Gerard Torenvliet
2004

Jess, and all:

> I think it's great that Austin (and others) aren't content with a definition
> that's a couple decades old...it's how we're going to create a better
> understanding of our work.

Absolutely - no disagreement here at all. I just want to ensure that
we fully interact with the work that's been done before as a basis for
expanding it. Better to stand on the shoulders of giants than to
re-invent.

I also agree that Norman didn't fully expand the topic out in POET -
like I said, he only dealt with the concept at a high level. Lots of
other people have plumbed the topic of mental models deeply.

Expectation is also a useful topic, and something that we need to
leverage (I think the thread on intuitiveness from a while back spoke
to this).

There is only thing that I want to make clear: I don't think that
expectations are part of a mental model, but rather that they inform
mental models. Because I expect a new system to be like one I've used
previously, my expectation causes me to build a rough mental model of
a new system before even interacting with it. Expectation is a tool we
leverage in building mental models.

This works two ways: We can learn a system that follows our
expectations almost for free. Conversely, points where a system
doesn't follow our expectations often present interaction hiccups (or
seizures, depending on how bad they are). IMO these are the places
where people learn most about the systems they are working with.

Ever your resident ideological conservative, :-)

-Gerard

--
Gerard Torenvliet
g.torenvliet at gmail.com

25 Feb 2005 - 2:31pm
Robert Reimann
2003

Let's look at the word "expectation" itself. There are two relevant
meanings in relation to the discussion so far:

1) Something anticipated or awaited
2) Something considered reasonable, due, necessary, or certain

Using definition 1), if we ask *what* is anticipated, we realize
that we must have a clear mental model of the anticipated
actions/objects/context,
(constructed, presumably from past experiences) otherwise we wouldn't
*expect* it.
How can you anticipate something you aren't relatively clear about?

Using definition 2), we are talking about a value judgment of
reasonability, certainty, etc. within a given context. This judgment
must again be made against a mental model of actions/objects/context
that is fairly well-formed, in order for us to make a judgment in the
first place.

Notice that in both cases, the "something" in the expectation is
a mental model of anticipated reality. To me that implies that
expectations are, in fact, mental models, and that the act of
*expecting*
is the result of having an existing mental model that seems to match
an immediate context, rather than being the trigger for model
construction.

However, I also agree that our expectations do, in a manner of speaking,

"inform mental models" at the point they are compared against reality.

To put this more clearly: our mental models (expectations) are
constantly
being adjusted as expected actions, objects, and outcomes are tested
against
reality. Usually mental models are reinforced with experience, but if a
new experience is at odds with our models, cognitive dissonance/friction

ensues with all the negative effects that implies. MILO.

Robert.

-----Original Message-----
From:
discuss-interactiondesigners.com-bounces at lists.interactiondesigners.com
[mailto:discuss-interactiondesigners.com-bounces at lists.interactiondesign
ers.com] On Behalf Of Gerard Torenvliet
Sent: Thursday, February 24, 2005 9:47 PM
To: Jess McMullin
Cc: discuss at interactiondesigners.com
Subject: Re: [ID Discuss] Defining the mental model - Expectations

[Please voluntarily trim replies to include only relevant quoted
material.]

Jess, and all:

> I think it's great that Austin (and others) aren't content with a
> definition that's a couple decades old...it's how we're going to
> create a better understanding of our work.

Absolutely - no disagreement here at all. I just want to ensure that we
fully interact with the work that's been done before as a basis for
expanding it. Better to stand on the shoulders of giants than to
re-invent.

I also agree that Norman didn't fully expand the topic out in POET -
like I said, he only dealt with the concept at a high level. Lots of
other people have plumbed the topic of mental models deeply.

Expectation is also a useful topic, and something that we need to
leverage (I think the thread on intuitiveness from a while back spoke to
this).

There is only thing that I want to make clear: I don't think that
expectations are part of a mental model, but rather that they inform
mental models. Because I expect a new system to be like one I've used
previously, my expectation causes me to build a rough mental model of a
new system before even interacting with it. Expectation is a tool we
leverage in building mental models.

This works two ways: We can learn a system that follows our expectations
almost for free. Conversely, points where a system doesn't follow our
expectations often present interaction hiccups (or seizures, depending
on how bad they are). IMO these are the places where people learn most
about the systems they are working with.

Ever your resident ideological conservative, :-)

-Gerard

--
Gerard Torenvliet
g.torenvliet at gmail.com _______________________________________________
Welcome to the Interaction Design Group!
To post to this list ....... discuss at ixdg.org
(Un)Subscription Options ... http://discuss.ixdg.org/ Announcements List
......... http://subscribe-announce.ixdg.org/
Questions .................. lists at ixdg.org
Home ....................... http://ixdg.org/

25 Feb 2005 - 2:33pm
Robert Reimann
2003

> MILO.

Ahem: IMHO.

Thanks Outlook spellchecker!

Robert.

-----Original Message-----
From: Reimann, Robert
Sent: Friday, February 25, 2005 3:32 PM
To: discuss at interactiondesigners.com
Subject: RE: [ID Discuss] Defining the mental model - Expectations

Let's look at the word "expectation" itself. There are two relevant
meanings in relation to the discussion so far:

1) Something anticipated or awaited
2) Something considered reasonable, due, necessary, or certain

Using definition 1), if we ask *what* is anticipated, we realize
that we must have a clear mental model of the anticipated
actions/objects/context,
(constructed, presumably from past experiences) otherwise we wouldn't
*expect* it.
How can you anticipate something you aren't relatively clear about?

Using definition 2), we are talking about a value judgment of
reasonability, certainty, etc. within a given context. This judgment
must again be made against a mental model of actions/objects/context
that is fairly well-formed, in order for us to make a judgment in the
first place.

Notice that in both cases, the "something" in the expectation is a
mental model of anticipated reality. To me that implies that
expectations are, in fact, mental models, and that the act of
*expecting* is the result of having an existing mental model that seems
to match an immediate context, rather than being the trigger for model
construction.

However, I also agree that our expectations do, in a manner of speaking,

"inform mental models" at the point they are compared against reality.

To put this more clearly: our mental models (expectations) are
constantly
being adjusted as expected actions, objects, and outcomes are tested
against
reality. Usually mental models are reinforced with experience, but if a
new experience is at odds with our models, cognitive dissonance/friction

ensues with all the negative effects that implies. MILO.

Robert.

27 Feb 2005 - 7:04pm
Eugene Chen
2004

If by "expectation" we mean being able to make predictions of system
behavior, then I think there is a mental model that precedes the ability to
make predictions.

I always feel that I'm working on mental model issues when I'm adding or
clarifying terminology within a program. Providing the user with key
concepts. Often, I find developers will not make certain concepts explicit
in the interface. For example, there is a certain object in the system. But
upon examination it turns out that sometimes the object behaves one way and
sometimes another way. I might recommend that the objects be given
differentiating labels, or icons, or some other way of teaching the user the
distinction. In this sense, mental model is closely related to feedback,
except that it addresses with concept as well as percept.

For instance, I know nothing about insects. To me, a box of insects would
appear like an undifferentiated pile. But if I were to spend a few hours
with an entomologist friend, they would point out certain key relationships.
"Notice how there are 6 legged ones and 8 legged ones", and "8 legged ones
never have wings". That allows me to begin to make certain predictions about
the future.

I've always thought the right way to document mental model would be with
concept maps (circle-and-line-diagrams), where you define key nouns within
the system and their relationships from the user's perspective. e.g. "I want
to organize my PHOTOS into ALBUMS. But WITHIN ALBUMS, I have FAVORITES. And
these I want to GIVE to a FRIEND on a CD". So for photos, concepts like
EMAIL, PRINT, and TIME may be primary in the mental model, even if FORMAT,
COPMRESSION, and RESOLUTION are primary in the implementation model.

- eugene

eugene chen | user experience: design, strategy, and usability

9 Mar 2005 - 6:58am
lopez_r6 at tsm.es
2004

Gerard wrote:
>Mental models have already been defined at at a high level by Don
Norman. See the Design of Everyday Things, p. 189-90.

Maybe I'm mistaken but I think that the first person
to speak about metal models is Emmanuel Kant in the
XVIII century.
------------------------------------------------------------
Rafa López Callejón
Dirección de Comunicación
www.empresa.movistar.es
e mail: lopez_r6 at tsm.es
Tel :+34 680 01 86 79
------------------------------------------------------------

"Jess McMullin" <jess.lists at nform.ca>
Para: <discuss at interactiondesigners.com>
Enviado por: cc:
discuss-interactiondesigners.com-bounces at lists.intera Asunto: RE: [ID Discuss] Defining the mental model -
ctiondesigners.com Expectations

25/02/2005 00:06

Telefónica Móviles España, S.A.

[Please voluntarily trim replies to include only relevant quoted material.]

Gerard wrote:
>Mental models have already been defined at at a high level by Don
Norman. See the Design of Everyday Things, p. 189-90.

I think it's great that Austin (and others) aren't content with a
definition
that's a couple decades old...it's how we're going to create a better
understanding of our work.

More to the point, Norman doesn't actually define what makes up a user's
mental model in POET - he just says it's "what the user develops to explain
the operation of the system", and goes on to say that the designer and user
only communicate through the 'system image' and that the user develops
their
model through interaction with the system. But is that all there is? The
user brings nothing to the table, they're in a state of some sort of newbie
tabula rasa, just waiting to be imprinted by the system?

Norman's model, as it's presented in POET, is designer and system centric -
it doesn't acknowledge what the user brings to the table, via expectations,
goals, culture, past experience, etc. Austin's model shows that there are
things going on in the users' minds before system interaction.

I'm not unbiased myself - I think that 'Expectation' is an incredibly
useful
concept for discussing mental models.

See this snippet of the start of a presentation from 2001 (IE only, I'm
afraid)
http://www.interactionary.com/files/xgap/

and more recently, from 2003-2004 some thoughts on the fact that mental
models, and expectations, are developed through iteration, as part of the
user's experience cycle:

http://www.nform.ca/files/experience_cycle.pdf

Peter Merholz has his own thoughts about expectations and Explicit Design
http://www.peterme.com/archives/000333.html

Most interesting for the "we've been there, done that" point of view is
that
Don Norman started talking about "Expectation Design" at last year's
NNGroup
tour. (snippet from Peterme's blog)

Expectation Design: The Next Frontier
Don Norman
Good designers already know how to make products attractive (visceral
design) and how to appeal to self- and brand-image (reflective design).
Good
behavioral designers know how to make products usable and
understandable--indeed, that's the focus of most of this conference. It's
time now to turn our attention to pleasure and fun. Here, the challenge for
designers is behavioral design, where expectations drive emotions. This is
where hope and fear, and satisfaction and anger reside. Deliver on positive
expectations and people experience pleasure. Deliver something different
than expected, but equally satisfying, and people have fun. Fail to
deliver,
or leave people feeling out of control, and you get a wide range of
negative
emotions.

Expectation-driven design marks a new dimension for our discipline and
provides a new framework for design. It shifts the emphasis from pure
function to an emphasis on designs that both function well and offer people
pleasure, enjoyment, a sense of accomplishment, and yes, even fun.

Cheers,

Jess

--------------
Jess McMullin
Principal
nForm User Experience
www.nform.ca

_______________________________________________
Welcome to the Interaction Design Group!
To post to this list ....... discuss at ixdg.org
(Un)Subscription Options ... http://discuss.ixdg.org/
Announcements List ......... http://subscribe-announce.ixdg.org/
Questions .................. lists at ixdg.org
Home ....................... http://ixdg.org/

Syndicate content Get the feed