|
Many
of you reading this review may well already own a copy of the book.
Many of you will have read it too. It has been in the Amazon.com
bestsellers since it was published two months ago. Many of you may
have been greatly impressed by the sycophantic reviews and unbelievable
claims of some reviewers. If the hype is to believed then this book
does indeed, "part the sea" of mediocre web site design
and "lead us all across to the promised land" of a usable
web.
Jakob
Nielsen, however, makes some rather more modest claims. He claims
that this is the first of two books. This one details with the "what"
of web design, identifying the parameters and detailing the elements
which make up web design. The second book will detail the "how".
Further, he sets the modest goal for this book, that if in some
way it makes the web a little better, makes people think a little
more before they lay out a design then it will have achieved its
goal. The goal of this first book is to change our [the web designers]
behavior. This apparently is to be done without teaching us how
we should behave.
A
Bible should tell you "how to ..."
When
Nielsen himself makes such an honest claim in the preface that this
is not a "how to" book, why can it be that the hype has
been so great? Perhaps, the believers are indeed in a dark place,
bereft of light and desperate that as was foretold the prophet would
come unto them and lead them across the sea into the promised land.
So desperate are they that the prophecy be fulfilled, that they
claim false witness upon the first book of Jakob.
To
rightfully claim that a book is a "bible" amongst its
peers, it must surely teach right from wrong, it must define a mode
of proper behavior and it must be applicable to all life, not merely
some subset of circumstances which occurs all too rarely. For were
it not to cover all of life, then those who ought to follow its
teachings will not, claiming "this doesn't apply to me".
Some may claim that a bible should also involve a blind faith or
belief, it should be written in obscure archaic language, and be
full of half truths, contradictions, and require interpretation
for the lay person by a member of the clergy.
Without
doubt Nielsen's book does fit with some of these attributes. Sadly,
it is full of contradictions and in many cases it does ask us to
follow with blind faith for its teachings are not explained or justified.
For example, page 168 tells us that "the news area should
be restrained and leave a large part of the page available for navigation".
However, earlier in Chapter 2 - page 18, we were told that "Navigation
is a necessary evil and should be minimized."
Naturally,
the disclaimer in the preface that everything will be explained
in the second book, covers all of these shortcomings. The book truly
fails on the "bible" claim when you consider whether it
truly teaches and whether it is truly applicable to all of the web.
The answer is flatly no.
Myth
and Legend
So,
we would like to dispel some of the myth and legend surrounding
this book. Perhaps, sadly, it's just a book about web site and web
page design. It's not even a particularly good one. Let's explain
why...
Author's
background
Jakob
Nielsen's background is as an academic in the field of HCI and more
specifically Usability Engineering and the application of Usability
Engineering on the emerging 1980s technology of hypertext. As such,
Nielsen continues to believe that the web is all about hypertext
documents. This background affects the book in two ways.
Firstly,
it's all about hypertext documents. There are only two types of
text in Nielsen's web: regular text; and hypertext - text which
links to another piece of text usually in another document. He does
expand on this concept of links and produces three classifications
for them: structural links (or navigation links as others might
call them); associative links (within the same page); and also links
(for additional references such as those in the "related articles
box" top RHS of this page). Later, he goes on to consider off-site
links and advertising links. What he fails to acknowledge or even
consider is whether a link is just a link, like any other link.
Is it all really hypertext?
Consider,
if you will the "links" across the banner of this page.
To Nielsen, they are just links, hypertext in a page of more hypertext.
He doesn't consider that they have the same semantic effect as buttons
in an application. He never considers whether visual affordance
of a button or alternative navigation control can be achieved. In
Nielsen's world, there is no semantic difference between a hyperlinked
piece of text like this
one, and a navigation element such as you see all around the
edge of this article. This seemed strange to say the least. Chapter
2 which is dedicated to Web Page Design contains considerable analysis
of links and much of it is interesting and useful but leaves as
many questions open as it answers. So, Chapter 2 represents a good
basis for a debate. It certainly isn't the definitive answers on
page layout and design.
Secondly,
we have to look at Nielsen's background in Usability Engineering.
Usability Engineering is an emerging science (some might claim psuedo-science)
which aims to measure the usability of a given piece of software,
device or information appliance. In order to be a science, it must
follow a scientific method. Such a method involves: making some
premise (a guess or judgment); defining a method (of testing); running
a series of tests (measuring); and then proving or disproving the
premise with the results. Science being science, you can then extrapolate
those results across other things where the premise is the same.
There is a good chance that the same rules will apply. Let
us explain.
So,
we have a premise that the effect of gravity is always the same
regardless of the mass of an object. So we devise a test which says
we will take two metal balls of varying size and weight and drop
them from a tower. We do this and measure the result. Both balls
hit the ground at the same time. So our premise was true. From this
we may extrapolate additional premises such as the rule remains
true for all objects.
Next
we try the experiment with a feather. This time the premise is not
proven. Later we discover that a hitherto unknown phenomenon known
as "air resistance" was affecting our experiment. That
is science.
Science
is the art of making educated guesses and proving them by experimental
testing. Jakob Nielsen is a scientist who makes educated guesses
about interfaces and then proves or disproves them by experiment.
He then extrapolates his results to other closely related interfaces
and tests them again. Or
so we must believe!
We
must truly believe this, for this gospel according to Jakob, is
devoid of scientific method. Compare it to Jared
Spool's book from last year. A book which was at pains to lay
out the scientific method used. It gave us the premises. It gave
us the method. It gave us the test objects (the websites tested),
and it gave us the results. Like true science, others could have
repeated these tests in order to validate them.
In
Nielsen's book, there is no such science, and again the disclaimer
that all will be revealed in the second book.
Full
Color Illustrations
So
this beautiful book, in full color, is merely a taster. Something
to whet our appetite for more. In essence it is a collected set
of Nielsen's "Alert Box" columns, pieced together in a
coherent whole as one single story and beautifully illustrated with
color images from web sites showing both good and bad examples,
of web site design.
It
is common for an author to be heavily criticized for "cashing
in" when they repackage previously "free" web material
and sell it as a book or a lecture tour. uidesign.net offers
no such criticism. Nielsen has over the last 5 years worked very
hard producing useit.com and
has had a profound affect on the web design community. He deserves
reward for these efforts and publishing a book which compiles the
Alert Box columns is a just way to realize such a reward. What is
important for the reader is to see the book as Nielsen advertises
it. It's just a book about web design. It should not be held up
in high esteem for if this is the pinnacle, where do we go next?
It
is important also to realize that the realm of science is based
on a sampling of what already exists. Science measures the known
universe and so it is true with the work of Nielsen. Much of this
book focuses on the web of the last 5 years. A web of publishing.
More specifically, of publishing hypertext documents. It is a book
which focuses on brochure ware, and news sites, online journalism,
and information sites such as uidesign.net. There is only
fleeting notice paid to the eCommerce sites for business-to-consumer,
and there is no coverage of the emerging market for extranet application
sites, application service providers, wireless and wap sites, or
business-to-business portals, hubs, exchanges, or markets. This
is the nature of the whole Usability Engineering approach. It seeks
to measure what is already there and often dare not risk extrapolation
beyond the tested data set. It is important to appreciate this when
fully evaluating whether this truly is a book about "Designing
Web Usability".
Yes,
perhaps it is, so long as your web is a information site. A site
predominated by hypertext. If you are tasked as web designer for
a newly formed asp.com startup, you might need to look elsewhere
for advice on designing application services.
What
does the Content deliver?
Chapter
3 focuses on Content Design. This is essentially a summary of good
online journalism techniques. It then looks at multimedia such as
video and sound. This chapter like most of the rest is littered
with interesting little pieces of advice and design heuristics.
The problem with it all is the storytelling style of the book. It
will make it difficult to go back later and extract these useful
hints and tips. Perhaps, you can make notes and pull these gems
out into a separate document for your own use. With this, the book
has some long term lasting value.
Chapter
4 looks at Site Design or Navigation. Again this is heavily biased
in Hypertext research. It all focuses around "Where am I?",
"Where have I been?", "Where can I go?". This
appears to be rooted in the belief that everyone is a surfer. This
begged the question, "Please show us the scientific method?".
It is never mentioned that perhaps the User's Goal might be important
to Navigation.
On
the whole, there is a lot of interesting material covered. Splash
screens must die! Metaphor is by-and-large a bad idea. Deep linking
is good. Deep navigation rather than wide is better. Sadly, none
of this is justified by experimental result and some of it flies
straight in the face of real scientific results published by other
Usability Experts over the last 18 months.
The
advice on reducing navigation clutter is excellent but sadly not
backed with any examples. As was mentioned before, this book does
not intend to tell you "how". It is not a teaching book.
Nielsen goes on to advocate his "breadcrumb" trail control.
Interestingly, he never debates whether the "breadcrumb"
metaphor is actually appropriate or indeed that he previously stated
metaphor was a bad idea. This highlights another aspect of this
book. It is advocacy of preferences in design, dressed up as science
and, for the most part, unsubstantiated by presentation of any real
scientific fact. For example, Nielsen states, on page 168, that
"most users are search-dominant". The uidesign.net
logs disagree. So, do the logs of other websites we know of.
It
is fine for Nielsen to advocate his preferences but it would be
more honest to come out and admit these are preferences, rather
than scientific truths.
Intranets
are different
Chapter
5 moves to the intranet. Where he looks at how web sites can be
used to provide "groupware" type functionality. He points
out that some of his earlier heuristics for the web in general can
be changed because the audience for the site has changed. What he
fails to point out is that with an intranet you can do traditional
User and Task Analysis and design the site to deliver the User's
Goals. This is a most obvious omission and inexcusable.
Amazingly,
Nielsen gives us some clues to the the approach for the whole book
when he says when discussing intranets, "For public Internet
sites, the most important usability attributes are probably
learnability and subjective satisfaction."
Hang
on! Shouldn't something which claims to be science, be able to give
us better than "probably"?
There
were some genuinely useful pieces of advice in here including a
checklist for heuristic evaluation of an intranet design. The one
glaring omission from the chapter was any discussion of Content
Management tools which are surely essential for any half decent
intranet site where many Content Creators are not html savvy. Again,
we are left with the disclaimer that this book does not intend to
show you "how".
Accessibility
and Internationalization
Jakob
shows some breadth and sensitivity in the concluding chapters of
the book. Chapter 6 is about accessibility. It makes one key point.
Web Designers should remember they are not typical. They are mostly
young, healthy with good mobility and good vision. Sadly, not everyone
is so lucky.
Chapter
7 discusses internationalization and localization and spends most
of the text explaining different approaches for Usability testing
in overseas locations. There is some usefulness to this material
at least for some web designers but again it doesn't tell you how.
Chapter
8 fantasizes about the future trends on the web and for Browsers
and internet applications. Some of this was more believable than
others and debate about these future things is fun. Sadly, the chapter
concludes without offering us any decent design advice to guide
us through the next 3 to 5 years. I guess we'll just have to wait
until after the fact, then we can test it, figure out what we did
wrong and wait for the second coming of the Gospel according to
Jakob.
Conclusion
In
summary, it doesn't look rosy. However, we have to balance Nielsen's
claims for the book in the preface against the hype it received.
Nielsen makes humble claims for this book. It is he claims a populist
book for the masses. One which will open their eyes to the problems
on the web and one which will give them some clues and heuristics
with which to address them.
The
hype has claimed that the book is the answer. That it is the saviour
which will "part the sea" and "lead us all across
to the promised land" of a usable web.
It
clearly isn't that. It is not a good teaching book and it is not
a technical reference. It is an interesting story about designing
certain types of web site and it does teach the issues involved
and opens the debate on the answers to the problems. However, it
leaves us tantalized waiting for the second coming. When will the
second book arrive? and will it be the one to lead us to the promised
land?
So,
we refer you to the scoring guide for book reviews at uidesign.net
3/5
A book which is interesting in part but is likely to have little
more than a neutral effect on the practical development of software
and improved user experience.
Recommendation
3/5.
Unfortunately, it doesn't tell you "how", it isn't science,
it's not about every kind of website or every kind of user, but
it will help, in some small way to make the web a better place to
be.
Order
this book...
|
