testing navigation models

21 Mar 2005 - 12:08pm
9 years ago
18 replies
1650 reads
Angela Azzolino
2005

Looking for methods to test navigation models. With no time allotted
to testing in the schedule I'm looking for recommendations from the
community to test a navigation model designed for a non-transactional
information site with six user groups, one primary. What are some
non-html prototype fast and dirty methods out there?

**sorry for the double post folks

Comments

21 Mar 2005 - 2:38pm
David Texidor
2004

On March 21, 2005 Angela Azzolino wrote:
> What are some non-html prototype fast and dirty methods out there?

Uh, sticky notes, dry boards, good ol' pen-n-paper ??

-----Original Message-----
From:
discuss-interactiondesigners.com-bounces at lists.interactiondesigners.com
[mailto:discuss-interactiondesigners.com-bounces at lists.interactiondesign
ers.com] On Behalf Of Angela Azzolino
Sent: Monday, March 21, 2005 11:08 AM
To: discuss at ixdg.org
Subject: [ID Discuss] testing navigation models

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

Looking for methods to test navigation models. With no time allotted
to testing in the schedule I'm looking for recommendations from the
community to test a navigation model designed for a non-transactional
information site with six user groups, one primary. What are some
non-html prototype fast and dirty methods out there?

**sorry for the double post folks
_______________________________________________
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/

21 Mar 2005 - 2:53pm
Peter Boersma
2003

David wrote:
> > What are some non-html prototype fast and dirty methods out there?
>
>
> Uh, sticky notes, dry boards, good ol' pen-n-paper ??

Those are NOT methods, they're (barely) tools.

Guerrilla methods to test navigation models that I can think of include:
- affinity surveys
("dear prospective user, do you feel comfortable with this label for that?
Why (not)?")
- closed card sorts
("dear prospective user, would you please sort these cards into stacks that
make sense to you, and tell us why in the process?")
- expert review
("dear colleague, would you be so kind as to review this navigation model
for me please? Here are the requirements.")
- gap analysis
("guys, can we think of reasons why this would fail? What if we compare the
options for each section and see if they make sense when applied to other
sections?")

Sure, some of these methods could be executed using sticky notes, etc. but
the tools themselves can be used for other methods as well.

Peter
--
Peter Boersma, User Experience Practitioner
http://www.peterboersma.com

21 Mar 2005 - 6:26pm
Lada Gorlenko
2004

PB> Guerrilla methods to test navigation models that I can think of include:
PB> - affinity surveys

Technically yes, and even a user evaluation method, but not the best
one...

PB> - closed card sorts

It's a formative technique, not an evaluation

PB> - expert review

Yes, but not as good as a first-hand user evaluation

PB> - gap analysis

Yes, but not as good as a first-hand user evaluation

User test is still the best option. Since Angela mentioned six user
groups, I would suggest testing at least a couple of them, one primary
and one or two secondary (ideally, all of them, but I don't live in an
ivory tower).

My team test website navigation with a technique we call PathFinder.
We wrote a test application that assists in running the test and
collects lots of data on the back of it, but you can do a simple
version using a word file or even a paper prototype.

The idea: a user performs a sequence of tasks having nothing but the
navigation tree of a site in front of her. No pages, no layout,
no content, just link labels and the structure. At any particular
moment, you display one level of a specific tree branch to the user.
For example, Step 1 shows the primary navigation: if the site has
tabs, show nothing but tab labels. The user picks one tab and you
'unfold' the next level under the chosen tab. Step 2 then shows the
main nav. categories under that tab, such as News, About, Contact Us,
whatever. The user picks one category and at Step 3 you unfold the
next level, and so on. At any single step, you unfold the next level
of the chosen category and fold in the previous one. When a leafnode
is reached, a page title is displayed. The user traverses the nav.
tree until she is believes she has found the target page (judging by
link labels and page title). Alternatively, she can give up at any
point. Than you calculate the rate of success/failures, compare the
actual path with the optimal, and do any other measurements you fancy
(we use about half a dozen of them in the PathFinder test).

In a way, PathFinder is the reverse card sorting: there are
pre-defined categories and you ask the user to find the target info
based on the categories. Due to the bare simplicity and total
absence of any contextual clues, PathFinder is a pure structural test
that gives you a baseline for the worst case scenario.

Lada

21 Mar 2005 - 7:04pm
Clay Newton
2004

> Looking for methods to test navigation models.

Given that you do not have a huge amount of time allocated to testing,
I would recommend building paper prototypes of the various navigation
models you have, and test them using standard paper prototyping
methods. You will likely that after testing in this fashion, the best
choice will be fairly clear.

If you need some help with paper prototyping, I would recommend:

Paper Prototypes by Carolyn Snyder
ISBN: 1558608702
http://www.powells.com/cgi-bin/biblio?inkey=2-1558608702-2

Best of luck!
-Clay

21 Mar 2005 - 8:49pm
Todd Warfel
2003

You should have taken my workshop at the Summit on Paper Prototyping.

You can get the slides here:
http://toddwarfel.com/archives/2005/03/ia_summit_05/

On Mar 21, 2005, at 8:04 PM, Clay Newton wrote:

> If you need some help with paper prototyping, I would recommend:
>
> Paper Prototypes by Carolyn Snyder
> ISBN: 1558608702
> http://www.powells.com/cgi-bin/biblio?inkey=2-1558608702-2
>
> Best of luck!
> -Clay

Cheers!

Todd R. Warfel
Email/AIM: twarfel at mac.com

21 Mar 2005 - 9:27pm
Clay Newton
2004

Hi Todd,

> You should have taken my workshop at the Summit on Paper Prototyping.
>
> You can get the slides here:
> http://toddwarfel.com/archives/2005/03/ia_summit_05/

Ok, so I noticed shortly into your slides that paper is less useful in
prototyping navigation. Maybe this is referring to navigation as in
links, but I have found it pretty useful in looking at global
navigation paradigms. Can you give a little back-fill in the
challenges to paper prototyping navigation models?

I wish I had been lucky enough to attend the conference this year!

-Clay

21 Mar 2005 - 9:47pm
Richard Czerwonka
2005

Try locating a book called "Software by Design - Creating People Friendly Software" by
Penny Bauersfeld ISBN 1-55828-296-3

It's slanted towards the Apple Mac, but there a whole bunch of ideas in it.

On 21 Mar 2005 at 13:08, Angela Azzolino wrote:

> [Please voluntarily trim replies to include only relevant quoted
> material.]
>
> Looking for methods to test navigation models. With no time allotted
> to testing in the schedule I'm looking for recommendations from the
> community to test a navigation model designed for a non-transactional
> information site with six user groups, one primary. What are some
> non-html prototype fast and dirty methods out there?
>
=================
Richard Czerwonka,
Delphi Programmer
ENT Technologies
Mob: 0412 104 042
=================

21 Mar 2005 - 10:08pm
Donna Maurer
2003

I have found paper prototype testing to be useful for interactive sites,
but almost impossible for informational (and therefore navigation-heavy)
sites. There are just too many pages and alternate paths, and it is almost
impossible to really disconnect content from navigation. Carolyn Snyder's
book (as far as I remember) doesn't even attempt to address informational
sites...

Todd may have said the same in his presentation, but the pdf keeps breaking
my browser!

Donna

At 02:27 PM 3/22/2005, you wrote:

>Ok, so I noticed shortly into your slides that paper is less useful in
>prototyping navigation. Maybe this is referring to navigation as in
>links, but I have found it pretty useful in looking at global
>navigation paradigms. Can you give a little back-fill in the
>challenges to paper prototyping navigation models?
>
>I wish I had been lucky enough to attend the conference this year!
>
>-Clay

-------------------------------------------------
Donna Maurer
Usability Specialist
Step Two Designs Pty Ltd
Knowledge Management / Content Management / Intranets

http://www.steptwo.com.au/
donna at steptwo.com.au
(02) 6162 6307

21 Mar 2005 - 11:41pm
Todd Warfel
2003

Not so. We've found paper to be very useful in testing navigation.
However, one of the things we've found is that people don't necessarily
navigate based on global nav elements, but about 50/50 global (top) nav
and contextual (w/in the page or tile). And we've found this to be true
for applications as well as content heavy sites... And this is the case
both in testing both paper and interactive.

What the slides do say, and Carolyn Snyder states this in her book as
well, is that some interactive elements are tougher to do on paper
(e.g.. scrolling text boxes). However, if you were in my class, you
would have seen how to do even that.

On Mar 21, 2005, at 10:27 PM, Clay Newton wrote:

> Ok, so I noticed shortly into your slides that paper is less useful in
> prototyping navigation. Maybe this is referring to navigation as in
> links, but I have found it pretty useful in looking at global
> navigation paradigms. Can you give a little back-fill in the
> challenges to paper prototyping navigation models?

Cheers!

Todd R. Warfel
Email/AIM: twarfel at mac.com

22 Mar 2005 - 3:01am
Peter Boersma
2003

Lada wrote:

> Technically yes, and even a user evaluation method, but not the best
> one... [...] > Yes, but not as good as a first-hand user evaluation

Since Angela mentioned that time for testing was not officially available, I
went for guerrilla methids that may not be the best overall, but the best
for the situation.

> PB> - closed card sorts
>
> It's a formative technique, not an evaluation

Excuse me? Closed card sorts require research and design work up-front and
can definitely be used as an evaluation method! See e.g.
http://iawiki.net/ClosedCardSort

But my main point is that I was going for guerrilla tactics, not "the best"
option.

Peter
--
Peter Boersma - Senior Information Architect - EzGov
Rijnsburgstraat 11 - 1059AT Amsterdam - The Netherlands
t: +31(0)20 7133881 - f: +31(0)20 7133799 - m: +31(0)6 15072747
mailto:peter.boersma at ezgov.com - http://www.ezgov.com

22 Mar 2005 - 4:19am
Lada Gorlenko
2004

PB> Excuse me? Closed card sorts require research and design work up-front and
PB> can definitely be used as an evaluation method! See e.g.
PB> http://iawiki.net/ClosedCardSort

Yep, and then there is an B&A article by Todd Warfel and Donna Maurer
that says: "Card sorting is a user-centered, formative technique
[snip] Card sorting is not an evaluation technique"
http://www.boxesandarrows.com/archives/card_sorting_a_definitive_guide.php

And if Closed Card Sorting is a sub-method of Card Sorting, it
must inherit attributes of its parent method, correct?

In any case, it's not an attempt to prevent anyone from using closed
cart sorting in evaluation. I am a big fun of the method. I am also in
favour of using whatever one can afford to use and get as much out
of any method as possible. Just be aware that some methods are more
effective than others for a specific goal, and some are less. Yes,
card sorting is excellent for testing categories (discrete elements).
No, card sorting is not as excellent for testing path lengths, doing
path comparisons, and getting lots of other useful IA insights that
require working with *navigation process* rather than merely with
*navigation structure*.

PB> But my main point is that I was going for guerrilla tactics, not "the best"
PB> option.

Than we complement, rather than contradict, each other. I was aiming
for the most effective of the cheep and cheerful options I can
sensibly comment on. Apologies if it sounded as disregarding other
methods, it was meant to be just a quick evaluation of them.

Lada

22 Mar 2005 - 3:50am
bruhns
2005

Hi,

I can recommend the "low cost menu test" as detailed by the inventor here:
<URL: http://www.georg.dk/lowcost.htm >

Basically you make a matrix with your menu items on one axis and some
important content or tasks from these items on the other axis. You
then ask the test participants to tell you which menu item they would
choose to find the content. It is very simple and works well in my
experience.

All the best,

Jakob

22 Mar 2005 - 7:39am
Todd Warfel
2003

It is not a good evaluation method to test navigation. It is a good
evaluation method to validate groupings. That's not what they're after
here.

On Mar 22, 2005, at 4:01 AM, Peter Boersma wrote:

> Excuse me? Closed card sorts require research and design work up-front
> and
> can definitely be used as an evaluation method! See e.g.
> http://iawiki.net/ClosedCardSort

Cheers!

Todd R. Warfel
Partner, Design & Usability Specialist
MessageFirst | making products easier to use
--------------------------------------
Contact Info
V: (607) 339-9640
E: twarfel at messagefirst.com
W: messagefirst.com
AIM: twarfel at mac.com
--------------------------------------
In theory, theory and practice are the same.
In practice, they are not.

22 Mar 2005 - 4:18pm
Donna Maurer
2003

At 09:19 PM 3/22/2005, you wrote:

>And if Closed Card Sorting is a sub-method of Card Sorting, it
>must inherit attributes of its parent method, correct?

I don't think closed card sorting is a sub-method of regular card sorting.
It is a quite different technique that happens to also use index cards and
content ;) More like a sibling than a child...

Donna

-------------------------------------------------
Donna Maurer
Usability Specialist
Step Two Designs Pty Ltd
Knowledge Management / Content Management / Intranets

http://www.steptwo.com.au/
donna at steptwo.com.au
(02) 6162 6307

22 Mar 2005 - 5:04pm
Todd Warfel
2003

Yeah, or even distant cousin.

On Mar 22, 2005, at 5:18 PM, Donna Maurer wrote:

> I don't think closed card sorting is a sub-method of regular card
> sorting. It is a quite different technique that happens to also use
> index cards and content ;) More like a sibling than a child...
>
> Donna

Cheers!

Todd R. Warfel
Partner, Design & Usability Specialist
MessageFirst | making products easier to use
--------------------------------------
Contact Info
V: (607) 339-9640
E: twarfel at messagefirst.com
W: messagefirst.com
AIM: twarfel at mac.com
Blog: http://toddwarfel.com/
--------------------------------------
In theory, theory and practice are the same.
In practice, they are not.

22 Mar 2005 - 10:29pm
sudhindra
2004

Angela...
>>Looking for methods to test navigation models. With no time allotted
>>to testing in the schedule I'm looking for recommendations from the
>>community to test a navigation model designed for a non-transactional
>>information site with six user groups, one primary. What are some
>>non-html prototype fast and dirty methods out there?

Just an idea.. Since time is a constraint here.. Would you it not be a quick
way to create wire frame prototypes using image editing software tools like
photoshop?... You can simulate them for a real time informational site by
linking them from one page to another... (no need to create html here, just
simulate using image maps)....And then user-test them?? They are almost as
fast as paper prototyping and also provide users with an on-screen
experience...

What do ya think??

Sudhindra

23 Mar 2005 - 3:31am
Lokesh
2005

one of the fastest way of creating wire frame prototypes that I use is
VISIO. Its quick, hassle free and you can create wire frames in a jiffy. The
advantage of using visio, unlike other software like photoshop is that since
this wire frame prototype is a live document - with each user testing
feedback - you 'll keep updating it - all yu need to do is make changes to
the stencil drawings. Like photoshop - there is no need to create layers etc
to make your navigation. All click and drop thing here.

-----Original Message-----
From:
discuss-interactiondesigners.com-bounces at lists.interactiondesigners.com
[mailto:discuss-interactiondesigners.com-bounces at lists.interactiondesign
ers.com]On Behalf Of Angela Azzolino
Sent: Monday, March 21, 2005 11:38 PM
To: discuss at ixdg.org
Subject: [ID Discuss] testing navigation models

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

Looking for methods to test navigation models. With no time allotted
to testing in the schedule I'm looking for recommendations from the
community to test a navigation model designed for a non-transactional
information site with six user groups, one primary. What are some
non-html prototype fast and dirty methods out there?

**sorry for the double post folks
_______________________________________________
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 Mar 2005 - 11:02am
Kyle Pero
2005

Jakob, I like this menu test method. It reminds me a lot of Donna's Card-Based
Classification Evaluation -
http://www.boxesandarrows.com/archives/cardbased_classification_evaluation.php.

I've had good results using this method in testing navigation and it can be
executed much quicker using the 3x5 cards although I prefer to do this
evaluation on wireframes.

After 1hr of prep work and maybe 2hrs of execution I have a pretty good idea of
whether or not the naviagtion I'm building is heading in the right direction.
You can't beat that!

Kudos Donna. I think it's an excellent "tool".

Cheers,
Kyle

--
Kyle Pero
Information Architect/Usability Analyst
http://www.usableinterface.com
267-241-4786

Quoting Jakob Bruhns <bruhns at gmail.com>:

> [Please voluntarily trim replies to include only relevant quoted material.]
>
> Hi,
>
> I can recommend the "low cost menu test" as detailed by the inventor here:
> <URL: http://www.georg.dk/lowcost.htm >
>
> Basically you make a matrix with your menu items on one axis and some
> important content or tasks from these items on the other axis. You
> then ask the test participants to tell you which menu item they would
> choose to find the content. It is very simple and works well in my
> experience.
>
> All the best,
>
> Jakob

Syndicate content Get the feed