Kuro5hin.org: technology and culture, from the trenches
create account | help/FAQ | contact | links | search | IRC | site news
[ Everything | Diaries | Technology | Science | Culture | Politics | Media | News | Internet | Op-Ed | Fiction | Meta | MLP ]
We need your support: buy an ad | premium membership

Edward Tufte on Web Design

By adamba in Internet
Sat Sep 08, 2001 at 09:54:44 AM EST
Tags: Internet (all tags)

I recently attended a one-day course taught by Edward Tufte, entitled "Presenting Data and Information." Tufte is a Professor Emeritus at Yale, best-known for his three books on information design, The Visual Display of Quantative Information, Envisioning Information, and Visual Explanations.

Although his books mostly predate the Internet phenomenon, his ideas certainly have applications in web site design. As was pointed out on a handout, they can also be used for teaching, news graphics, technical illustrations, displaying financial data, decision-making, animation, and a host of other areas. It appeared that many in the audience were primarily interested in web design, and Tufte spent some time discussing this.

Tufte spent most of his talk walking around the room while talking on a wireless mike. He had two projectors set up, but for the most part he only displayed pages or pictures from his books, instructing the audience to follow along in their own copies (which had been provided to every attendee). He occasionally carried around some other props, in particular a few 400-year old books from his personal library. This style not only entertained and engaged the audience, it also emphasized one of his main points, which is that progress is often measured in data density - how many bits per unit of area can be accomodated by a hard drive or a display.

In terms of text display, a page in a phone book can hold 36K of information, while the best display can only show about 5K (those are his numbers, I didn't verify them). If you look at something like a topographical map, the resolution available on paper is a factor of ten, at least, beyond what can be shown on a screen.

Thus, paper is a much better way to present information than a computer display, and a speaker reading off a slide with five bullet items on it is actually an awful way to make a presentation. He feels printed handouts are not something to be displaced by the "paperless office," but instead a hallmark of a good presentation - not just for their information density, but as something permanent that can be taken away from a meeting, symbolizing the integrity of the presenter and his or her belief in what was presented.

Tufte feels that the same mantra about data density should be applied to web sites, and in fact to the entire contents of the computer display that the user sees when navigating a web site. Thus, he dislikes task bars, menu bars, status bars, and other GUI screen overhead, since they constrict how much of the display can be used for content. Once you get to the actual site, he has similar disdain for banner ads, navigation bars, graphical frills, and the like. His most withering criticism was for frames, which he views as having no purpose other than to fill even more screen real estate with useless information (he compared frames to a university filled entirely with assistant deans, which got a good laugh from the audience - so maybe it wasn't all web designers out there).

So he dislikes Powerpoint, Windows, Internet Explorer, Microsoft web sites, and Frontpage. He also dislikes designs based on user testing, feeling that it produces only routine competence, and that great designs - he mentioned the Macintosh and Next - come from great designers. For good measure he threw in some Excel jokes for the Seattle audience. So what does he like?

Tufte feels that the main measure of a web site (or any computer interface) should be the percentage of the screen that is actually devoted to the task at hand. He wants web pages to use words instead of icons, because they can display information more compactly (as he demonstrated, most icons require an explanatory word beneath them anyway, making the icon a complete waste). He does not like navigation bars, but instead wants as many choices as possible on the main page. The one "navigation"-type bar he wants is one allowing the user to select a different language (and please, put each language name in its own language, not in English).

He said that Excite had 240 links on its main page, yet remained clear and easy to use. Other sites whose density he admired include Amazon, Arts & Letter Daily and the very similar (design-wise) SciTech Daily Review, Science @ NASA, and photo.net (he is a friend and admirer of Philip Greenspun, and uses the ArsDigita Community System for his own website). He also highlighted Isys Information Architects' site for its discussion of good and bad interfaces. He did not mention Kuro5hin, but in my opinion its rough-hewn, information-dense design is up to his standards.

Tufte feels that good design is "clear thinking made visible." A designer should ask "What is the thinking task that this display is supposed to help with," and then imagine how the audience will respond to it. In fact, he suggested that designers think of their audience in terms of what they read. What he is aiming for is a move away from the "keep it simple" idea of design, in which the audience is viewed as stupid, towards one in which the audience is viewed as a partner in learning.

In other words, more like academia than the business world. In fact, he stated that the best way to learn how to make good presentations was to ignore the business books and instead read books on how to teach. Given the clarity and persuasiveness of his own presentation, I think he has a point.


Voxel dot net
o Managed Hosting
o VoxCAST Content Delivery
o Raw Infrastructure


Related Links
o Kuro5hin
o Edward Tufte
o Yale
o The Visual Display of Quantative Information
o Envisioning Information
o Visual Explanations
o graphical frills
o Excite
o Amazon
o Arts & Letter Daily
o SciTech Daily Review
o Science @ NASA
o photo.net
o ArsDigita Community System
o Isys Information Architects' site
o Also by adamba

Display: Sort:
Edward Tufte on Web Design | 25 comments (24 topical, 1 editorial, 0 hidden)
I disagree. (3.57 / 7) (#1)
by kwsNI on Sat Sep 08, 2001 at 12:03:03 AM EST

Interesting article and interesting philosophy. Unfortunately, I disagree with almost everything. I think a hallmark of a well done website is it's interface. First, I don't want everything condensed into one page like a phone book. How many times have you tried to look up a phone number in the phone book and after you find it, you lose it two or 3 more times before you have it written down or dialed? Although I can see why data should be compressed in the context of a phone book, there is no reason for it online.

An effective web site should give you quick access to the information you are looking for, but that doesn't mean it should all be on one page with a thousand links that fit in an 800x600 screen.

Also, graphics are important. Who cares if a graphic is redundant or wastes space? IMHO, a good web site effectively uses a reasonable amount of graphics to make itself visually appealing. Again, it goes back to his phone-book example. Yes, you can put thousands of numbers on one page, but the businesses that get the attention are the one that put in huge ads. Graphics get attention, and if used effectively and in moderation, make a web site more usable.

Still, it was well written and an intesting read - +1 Section.

I can picture in my mind a world without war, a world without hate. And I can picture us attacking that world, because they'd never expect it. -Jack Handy

I disagree with your disagreement. :-) (4.12 / 8) (#2)
by Dlugar on Sat Sep 08, 2001 at 12:56:04 AM EST

If you actually take a look at the pages that he recommends, you'll notice that they do have graphics, but for important functions. Yes, making the site visually appealing is an important function (see the NASA site for an example--very clean, very well done, and graphics used to catch the eye and make the layout appealing.)

You'll also notice that although some of the sites have more text than I would prefer, none of them have what I would call "phone-book density". I think you simply misunderstood his argument--what he was saying is that there is no way to fit a phone-book page on a computer monitor, because the monitor has less room and far less resolution. Hence we have to conserve our space and use it wisely--and certainly not try and make everything look like a phone book!

Just my take on it, anyway.


[ Parent ]
I don't know. (3.71 / 7) (#3)
by kwsNI on Sat Sep 08, 2001 at 01:13:00 AM EST

Personally, I don't think any of the sites he chose would be on my list of best designs on the net. And no, none of them are "phone-book" density, but some of them (esp. the SciTech) design are just plain crowded.

What I don't see is why these designs are so great. I can't see why it's so great to have all this info in a condensed format. Unless you're doing something like WAP, there is no reason why you need everything to fit on one page that fits on one screen. That's what's nice about the web. You can put information on different levels, and as long as it isn't excessive, it can be much easier to navigate.

Honestly, I think I can see what he's saying and I even understand the reasoning, but I think that there are more effective ways to do things with modern technology.

I can picture in my mind a world without war, a world without hate. And I can picture us attacking that world, because they'd never expect it. -Jack Handy
[ Parent ]

The phone book doesn't need big ads (none / 0) (#25)
by bcaulf on Mon Sep 24, 2001 at 12:36:13 AM EST

Tut, tut. Of course it is true that "the businesses that get the attention are the one that put in huge ads". But for some reason the phone book designers did not choose to fill the phone book with huge ads, inflating it to bookshelf-size proportions. This is because those designers have at least some loyalty to their users. They have created a functional interface instead of exclusively serving the interests of their advertisers. Most of the phone book is done in the usable design.

In such a situation, a designer just has to determine where his loyalty and allegiance lies: with users or with advertisers. I would contend that it is inhumane to side with the advertisers and oppose the users. However this is normal behavior today, so we do not question it too often.

Regarding the original question of whether Tufte's data density focus is right for the web: since virtually all web pages today consist either of text or navigation, it's hard to reasonably apply Tufte's ideas. Jamming a bunch of navigation onto a page guarantees that the relationship between the elements will be questionable. Tufte advocates for graphical works that have integrity, in the sense that the reader should be able to understand the connections between the elements on the page. There is no relationship to convey in a portal front page. It is more like a newspaper front page, where the many elements are on one page just for the sake of convenience. Maximum data density does not help the reader understand a whole in this case.

However I would claim that Tufte would advocate for a wide and flat tree of navigation, which necessarily means a healthy number of links at each navigation level. As far as I'm concerned that is a motherhood and apple pie type issue that reasonable people should not disagree on. But then I will see an expensive piece of trash like kurzweilai.net with a narrow, deep tree that is impossible to use. (I love the content but I despise the site.)

I do think that Tufte's focus on data density is vital for that tiny group of new media graphic designers who are actually concerned with conveying information. Most people are not pursuing that goal. They can ignore Tufte and get back to creating "whack the monkey" ads and streaming commercials.

[ Parent ]

I think he's way off the mark (4.46 / 13) (#5)
by jesterzog on Sat Sep 08, 2001 at 07:26:28 AM EST

With respect and without having read his books, I don't agree with this philosophy at all as you've described it.

I'm sure it's possible to keep 35k of information on one page of a phone book, but this is completely different. In a phone book, all of that information is likely to be exactly the same type. Get a name, scan down the list and voila! If you had a phone book on the web it could work exactly the same way, and be very successful. The key point is that it's all the same type of information and it's formatted uniformly making it easy to follow.

Compare this with excite, for example, and the context is completely different. Excite is trying to cram at least 11 unrelated types of information onto a single page. In my view this is an awful design, since the eye isn't directed towards anything and unless you know exactly where you're going you have to really sit and stare at it. Even if you do know where you're going, how many people are honestly going to use all 11 of those types of information at once? How on earth is this "clear and easy to use"?

Maybe that's why I never use excite for anything.

Compare it with google which is a straightforward, obvious search engine. The entire page focuses on the text box in the middle, and it's immediately obvious what to do. There's no serious clutter distracting or confusing you, and the bonus is that the search engine actually works really well. Sure you can't do as much, but that's because it's a search engine. It's obvious that it's a search engine and therefore it actually has a function and a reason for being.

I'm also a bit distressed by his shrugging off of user testing. Obviously great designers produce great designs, but the people who produced Macintosh and Next and Xerox PARC for the record all have one hell of a lot of user testing in their design processes. You can't get good user interfaces without understanding how users interact with a system. But if he's promoting excite and amazon, I guess he's never seriously watched people using different interfaces.

jesterzog Fight the light

information density (3.80 / 5) (#8)
by adamba on Sat Sep 08, 2001 at 10:33:04 AM EST

I'm sure it's possible to keep 35k of information on one page of a phone book, but this is completely different. In a phone book, all of that information is likely to be exactly the same type. Get a name, scan down the list and voila! If you had a phone book on the web it could work exactly the same way, and be very successful. The key point is that it's all the same type of information and it's formatted uniformly making it easy to follow.

The point about phone books was not to say that phone books are a breakthrough in how to organize information, but simply that paper is denser than a display because of the higher resolution. Thus, if you had a phone book on the web it would *not* work the same way, because you couldn't get that much information on the page (and in fact, phone books on the web *don't* work the same way).

- adam

[ Parent ]

Tuft has kind of missed the point... (2.50 / 6) (#6)
by Daemosthenes on Sat Sep 08, 2001 at 09:56:59 AM EST

Mr. Tuft might not like graphics, but nowadays they are really becoming an essential part of the web's culture. Web design doesn't just include text and buttons and links anymore. Web design is now considered graphic design; the wide array of tools such as flash, CSS, dhtml etc. hav really served to advance the internet's design above the level of simple "text and links" homepages into the level of art. And, in my opinion, Tuft has completel missed out on the art side of it. Check out some links of quite interestingly designed and run sites, which I think are a good example of the artistic side of the web:

Warning: most of the site are fairly flash or graphics heavy, might take a while to load

he is not anti-graphics! (4.00 / 4) (#7)
by adamba on Sat Sep 08, 2001 at 10:29:14 AM EST

His books are almost all about graphics. He is excited about the possibility of animated graphics on the web telling a story better than a static graphic on a page (albeit at lower resolution). In fact, he actually mentioned xl5design.com as a great site to look at (he said follow the "0.4 linkage" link) -- unfortunately I copied down the URL wrong or I would have included a link to it (so thank you for putting one in).

What he doesn't like are useless graphics that clutter a page and icons that serve no purpose -- most of what passes for "graphics" on the web these days.

- adam

P.S. One site he loves is the Music Animation Machine. He showed a video of it during the class -- excellent work.

[ Parent ]

Tufte is an "information" guy... (5.00 / 2) (#11)
by supine on Sat Sep 08, 2001 at 08:39:56 PM EST

You miss Tufte's main focus. His books (and apparently this talk) are on how best to convey "information" from a source to a recipient.

You refer to web sites which would be considered "art". However, Tufte isn't really talking about these sites because in that context the graphics are the "information".

His main target is sites trying to present some other, non-visual data. He advocates that there are good and bad ways to achieve "information" transfer in this context and that graphics are often (but not always) a poor substitute for text on the web.

And to add a comment of my own, Tufte's analysis in "Visual Explanations" of the breakdown in communication at NASA prior to the Challenger disaster is required reading. He compares the inadequacy of "information transfer" in that situation with the work done by Dr John Snow in successfully convincing authorities to close off a water supply which was the source of a cholera outbreak in London in 1854. An absolutely fascinating comparison of the highs and lows of communication.


"No GUI for you! Use lynx!!!, Come back, One year!" -- /avant
[ Parent ]
Challenger disaster (4.00 / 2) (#13)
by adamba on Sun Sep 09, 2001 at 11:18:16 AM EST

Tufte spent some time discussing this. If you read the book, he seems to be laying most of the fault at the feet of the engineers who prepared such an unconvincing presentation for NASA.

In the class, however, he talked about how people making a serious decision have an obligation to make sure they are getting the information they need -- that NASA should have pressed for more information. The questions he feels the recipients of a presentation should ask are, "Show me causality," "Show me all relevant data," and "What do I really need to know?"

I asked him about this apparent shift afterwards and he said, well, there was certainly enough blame to go around.

- adam

[ Parent ]

Text Nazis under the Bed (none / 0) (#22)
by Mabb on Tue Sep 11, 2001 at 01:49:51 AM EST

Why do a certain set of people immediately assume that those who talk about usability are anti-graphics? Perhaps it is some kind of automatic defence mechanism.... Even in his most autocratic heydey, Jakob Neilsen, although accused most strongly, was never anti web graphics. Like Tufte, he was anti UNNECESSARY graphics. Unnecessary meaning: not contributing significantly to the form or content of the object.

In my experience, this knee-jerk reaction is often found in those with the disorder, comprehendus interruptus*.

* a disorder that manifests by being unable to fully read or listen to the source material before jumping into the fray.

QuiltBlog: WIP, SEX, WOW, MQ, LQS, HST...

[ Parent ]

What's wrong with frames? (3.50 / 4) (#9)
by spacejack on Sat Sep 08, 2001 at 06:43:19 PM EST

Aside from the unnecessary use of (which can be a problem with any element of web design), frames can serve a useful purpose when browsing things like image thumbnails or documentation indexes. It saves bandwidth since you can simply re-load the part of the page that is changing.

You can probably do this in Javascript too, although Javascript has just as many critics (some who turn it off entirely). If you want to save bandwidth for dailup users, you choose what you think is the lesser of the two evils for your situation.

Why Frames Suck (4.00 / 1) (#19)
by kmself on Mon Sep 10, 2001 at 04:52:12 PM EST

Jakob Nielsen still nails it.

Even three years later, the criticisms hold.

In the case of thumbnails (your instance), I'd far prefer a set of pages with thumbnails and back/forward nav buttons, than a frame. Among the features highly utilized in my browser of choice is the "open frame in window" context menu.

Frames suck.

Karsten M. Self
SCO -- backgrounder on Caldera/SCO vs IBM
Support the EFF!!
There is no K5 cabal.
[ Parent ]

hmm (3.00 / 1) (#20)
by spacejack on Mon Sep 10, 2001 at 06:11:34 PM EST

None of those articles addressed any of my points -- frames can be faster and provide a better end-user experience.

I will grant you that best of all would be to have a frames/non-frames option. But frames do not universally suck, they only suck in a few situations; the most common of which is probably copying the URL from your address bar. But sometimes you don't intend people to start in the middle of your website.

FWIW I can do what you're doing in Galeon manually, but I almost never need to.

[ Parent ]
A shame about the course (3.00 / 2) (#10)
by jet_silver on Sat Sep 08, 2001 at 06:43:31 PM EST

Tufte's principles are difficult to apply but I try to do that; I'd been looking forward to his course but if all he does is point at the books and make quips I'll forget about taking it.

His comments about the Web are right on the money. There are far too many sites full of gratuitous graphic data on the Web (Tufte would call these sites 'ducks'.) I remember very well the Audi TT site that -required- a Shockwave presentation to even get started. Well, I wasn't interested in having my user experience controlled; I didn't want to sit through whatever it was, and I ended up with a Subaru WRX. Coincidence? Maybe.

Even the sites with fairly simple chrome tend to do bad things to Netscape - especially Netscape 4 series under NT, which I have to use at work - so I have learned to avoid them. The question is whether you want your message -heard- or whether it's more important that you present your message your way. A site with a lot of busy elements is like a bore talking and not knowing when to shut up. It's interactive BDSM, and I'm not interested.

I particularly love the comment "here are some sites that incorporate the elements I like. They require Shockwave and are slow to load." Do you realize what you are saying?
"What they really fear is machine-gunning politicians becoming a popular sport, like skate-boarding." -Nicolas Freeling
What's he smoking? (3.50 / 2) (#12)
by Dialup on Sun Sep 09, 2001 at 09:08:58 AM EST

Anyone who actually recommends the Excite page has to be on *something*..... about the only thing I've seen out there that's worse is http://web.icq.com/ - which is probably a site that this dude loves.

I'm a graphic artist. I get paid to do it for a living. I'm an amateur web designer- considering the amount of emphasis on things like Flash, streaming media and kilobyte-heavy bullshit doohickies that do nothing but keep me from my data, I would never consider doing it professionally. I'd get fired the second someone asked me to do something in Flash and got the finger in return- and the disgusting fact that the majority of people I talk to *like* it - and most of my friends do it for a living.

Try to do everything and you'll do just that- try. And fail. Portal sites are a joke, sold to newbies and suits who see only dollar signs- maximizing data density is maximizing DISTRACTION, not profit or even design. http://web.icq.com/ looks like a cluttered piece of shit- I don't even know where to start, and have to actually search the text with the browser utility just to find whatever it is that I want. This is good design the way the DMCA is good design.

But then, I'm a minimalist and a Mac user. Most people are neither.

Yeah, you should be concerned with data density from the standpoint of DESIGN, NOT how much you can cram onto the friggin screen. Not everyone runs at 800x600, and absolutely NOBODY I know actually runs their browser full screen. Yet just about everything requires at least that at a minimum- and they proceed to fill that space up with whatever they can think of during their lunch breaks.

I design for 600x, using relatively minimal graphics (with the exception of my "art site"), and keep everything as minimal as possible. I don't use Shockwave, Flash, XML, PHP, or any other TLA or buzzword. I use HTML and Javascript. In other words, I attempt to build sites that don't piss me off. I wish these so-called "web designers" had similar sensibilities....

Learn a bit about context.... fool (4.33 / 3) (#14)
by beynos on Sun Sep 09, 2001 at 11:54:58 PM EST

I wish these so-called "web designers" had similar sensibilities....

I use HTML and TCL... I don't use Javascript, because it is too browser dependent. I use TCL on the server end.

You sound like most graphic designers I know... They're all nice people, but, like you, they think they are so f*cking smart when it comes to web design.

Anyone who actually recommends the Excite page has to be on *something*

Did you hear the writer say that Edward Tufte said portal sites were cool? No, but because you think you know everything, you decided to waste a whole post insulting one point in a really good article. He was referring to the fact that they had managed to cram 240 Links onto one page, and it was still laid out clearly!

I am looking at the page right now. You're right, the graphics are poor, and for me, it is a pretty useless site. But if I did need to use a portal, I would rather be using one that is laid out like Excite, than one of your Graphically intensive Javascript pages.

You need to remember, he is not talking about the appearance of the web page. He is not talking about screen resolutions, he is talking about information. What he was saying is that finding the information he would need to find on Excite is easy, because the links are layed out well.

with the exception of my "art site

funny that you don't supply a link to your art site. Funny that you don't post any links to validate your claims that you have more idea about document design than one of the leading professionals in the area.

Have you read anything by Edward Tufte? Do you understand his key principles of document design? Obviously not, else you wouldn't have made this post. Read a bit more about the guy before you start insulting him.

He's a smart man who is concerned with the conveyance of qualitative information so that as wide an audience as possible will understand. This is NOT graphic design. Graphic design aims to generate a certain response from its target audience... Sounds slightly different to me.

I am sorry if I am sounding rude, but I am sick of the aggression that graphic designer types display when talking about web design. Y'all need to remember that Jpegs and Style Sheets and Flash and so on, all of these are worthless if your site can't be read without them. And most graphically intensive, resolution-specific web sites can't be read on my phone, or palm... nor can I read them from my BASH account... I must have internet explorer 4.0 or later.... Sorry, that just doesn't cop it!

Stop looking so hard for faults, and you actually learn something.

[ Parent ]
That's what the ALT tag is for. (none / 0) (#21)
by Dialup on Tue Sep 11, 2001 at 01:14:04 AM EST

Did I not say I was an amateur?

Yes. I'm a graphic artist. A "mac weenie", as it were. And, like I said, I'm not a professional web designer and have no desire whatsoever to be one. I *do* like to do UI design, however, beyond making the widgets do what I need them to, I really loath coding. That's what friends are for- I do some pages for a guy I know, he cranks them through his text editor to clean them up and make them run smoother. And there was much rejoicing.

If I were a field professional, I'm sure I'd be familiar with this guy's work. But I'm not. I do video and multimedia for a living. The web is a hobby. Not a profession.

I say this guy must be smoking something for the simple fact that when I look at the excite page, my retinae overload and all I see is a blob of color. I can't find shit without pouring over the page for a few minutes. That's BAD design, in my opinion- there's simply way, WAY too much information to assimilate. I look at it and I don't know where to go or what the hell I'm doing there.

Examples of well built sites:

www.google.com - I like Minimal. Minimal is Good.

http://seagate.com/ - If you deal with hardware sites on a regular basis, you'll appreciate just how easy this one is to use.

And you asked for some of MY links? Well....

www.414-crew.com - site design and layout . Very biotech. (maintained by the band)

www.414-crew.com/remix (personal site, presently in the process of a MASSIVE upgrade, which is why I didn't mention it initially)

www.secretaboutbox.com (still needs a lot of the content, I didn't know I was going to end up maintaining it) - this one's the graphics and bandwidth-heavy thingy that needs a bigass monitor.

In all three cases, what you're looking for is real easy to get to. The site designs are cohesive, do not break look and feel at all, and are, to the extent of what they're designed for, minimal as all hell. There is no "wall of links", a la mirabilis or excite. No possible way to get lost. The widgets and thumbnails are pretty much it- most of the graphical load is useable in some fashion.

(and the site I'm working on IS graphics heavy- I hate the quality loss on JPEGs and GIFs and I'm doing the entire site in PNGs. Hence, this is the first time I've extensively used the ALT tag.)

To wrap up:
I'm a hobbyist. Don't expect me to know everything about information design. I've never taken a class on it or been trained or been expected to actually DO it- I fly by ear and do what my design instinct tells me to do.
By the same token, I don't expect my friend Matt, a Computer Engineering major, to be able to quote Knuth. I don't think Matt actually knows who Knuth *is*, actually, but that's another story entirely.

And I still hate flash. And portals.

[ Parent ]
html? (5.00 / 2) (#15)
by theg on Mon Sep 10, 2001 at 07:37:16 AM EST

I use HTML and Javascript. In other words, I attempt to build sites that don't piss me off. I wish these so-called "web designers" had similar sensibilities....

You, sir, are a heritic, and should be punished. HTML, indeed! Now tell me what's wrong with plain text?

I only ever publish in plain ascii text and I always enter it into my computer with a hex keypad.

;) theg.

[ Parent ]
Visual Display of Quantitative Information (4.00 / 1) (#16)
by spring on Mon Sep 10, 2001 at 09:20:22 AM EST

For those who haven't read The Visual Display of Quantitative Information, you should give it a try. Its title is scary, but the book itself is fascinating and an easy read. It doesn't address web design directly, although its principles can be used for that. It talks about the most effective ways to get people to understand large quantities of data, which can sometimes mean charts and graphs, but he branches out into some weirder territory too, like using pictures of facial expressions to represent multiple variables simultaneously. It's a classic.

Hmmm... (4.00 / 2) (#17)
by mkelley on Mon Sep 10, 2001 at 12:37:47 PM EST

"dislikes designs based on user testing"

He mentions the above and then he talks about Excite. If I'm not mistaken, Jakob Nielsen was a consultant for that site and used user testing to help in it's initial design.

He might not like how the sites look, but they are functional. Functional more than a site like k10k. Nothing against them, but users using a screen reader would have problems with that site. I think it's beautiful, but I personally wouldn't use it for a design idea for a company or my personal site.

It's funny but for the sites he listed, most have great interfaces that look like Jakob Nielsen's sites. Most of those site went through user testing.

life is like a freeway, if you don't look you could miss it.

Creating metrics? Attend his lecture! (none / 0) (#18)
by raven on Mon Sep 10, 2001 at 02:58:42 PM EST

I ended up attending one of his sessions while I was on a metrics council. Boy, did we need it.

Different teams with radically different jobs had to submit metrics on the same data, no matter how irrelevant. Graphs and charts were unreadable and measured things no one cared about.

After Tufte, I started sitting in on the metrics meetings and critiqueing the information presentation.

That bar graph with 100 really thin bars? Make it a line graph! I want to see a clean, unbroken line of each item's activity, so I can follow it! Not a clump of thin bars of all the items' values at each data point.

Why are you measuring tickets? You're a project group, you work off projects - measure THAT data, not these unrelated tickets! That metric tells me nothing that will help your group improve, grow, or change. Useless!

There was a SERIOUS improvement in the usefullness and readability of our metrics after a few months, and our managers actually had information they could use to drive the team to improvement.

Tufte helped me understand how you have to not only figure out what information is important, but how to present that information so that it GETS used.

The cholera example and the Challenger example are such perfect examples. It almost made me cry when he took the data that the engineers gave NASA and did nothing more but reformat it - no one, not even the most ignorant corporate idiot would have allowed that launch given one simple graph that took the RELEVANT information and presented it clearly.

Interesting comment about user testing (none / 0) (#23)
by Mabb on Tue Sep 11, 2001 at 02:05:02 AM EST


can you expand a little on Tufte's comment re "designs based on user testing"? I'm sure there's more (and less) to this than first appears and I'm not going to assume he's against user testing but perhaps I will assume that he is advocating that one should not always look to existing designs and what users already know for inspiration?

QuiltBlog: WIP, SEX, WOW, MQ, LQS, HST...

user testing (none / 0) (#24)
by adamba on Mon Sep 17, 2001 at 10:01:33 PM EST

He did not say much more about it. The impression I had was that he valued consistency of design very highly. If you start with a consistent design but then start user testing it, changing this bit and that bit but keeping this other part the same, you lose the consistency.

So the ideal situation would be a designer coming up with a design and then having users like it as is. Of course the design can be informed by previous user tests the designer has done.

There is a section on his site called Ask E.T., where he answers questions from people. So perhaps you could pose this question there.

- adam

[ Parent ]

Edward Tufte on Web Design | 25 comments (24 topical, 1 editorial, 0 hidden)
Display: Sort:


All trademarks and copyrights on this page are owned by their respective companies. The Rest 2000 - Present Kuro5hin.org Inc.
See our legalese page for copyright policies. Please also read our Privacy Policy.
Kuro5hin.org is powered by Free Software, including Apache, Perl, and Linux, The Scoop Engine that runs this site is freely available, under the terms of the GPL.
Need some help? Email help@kuro5hin.org.
My heart's the long stairs.

Powered by Scoop create account | help/FAQ | mission | links | search | IRC | YOU choose the stories!