Comments on Poor Usability of Award Winning Web Sites

Here are some comments made on the paper. Please note that the original comments were made on the appearance of the blue paper, which you can see here. You can read about the reasons for the different formats of the blue paper and the present appearance. Names of personal communications are not given to respect privacy.

After running these comments for a while, I had the honor of been removed from the list (UTEST) in which communication took place. The reason given was that I made public what should have been private. But I suspect that given the fact that most of the communication was about the appearance of the paper, not its content, that I stepped on usbility expert toes, especially when I made that point.

Of course I agree with comments about the bold font and some other features. But the content was designed for display in a large auditorium, and I didn't bother to adapt it for a PC monitor.

Carl.Myhill@Smallworld.co.uk wrote:

Date: Wed, 24 May 2000 09:15:20 -0400 (EDT)

RE: Poor Usability of Award Winning Web Sites

Interesting article, although interesting from the perspective of being so dazzled by the formatting that I found it impossible to skim the content. Perhaps I'm just in an irritable mood but I was interested enough in the subject line to click the link, so I had some expectation of getting something from the visit to the article.
I was using netscape so I tried it in IE to see if it was better (less colourful), alas no.
I've been working and researching the field of usability for a number of years now and am of the opinion that usability people should set an example to others. Failure to do so discredits the rest of us. I was once in a meeting at the European Commission to discuss Usability Testing (ELPUB 105 I think) and was pretty horrified when one of the respected usability experts present used a full page of 12pt text on an OHP slide as part of their presentation. I find it quite unforgivable for a professor to stand up and lecture about the importance of usability using a full page of A4 text as a visual aid.
Perhaps I'm being too harsh, what do other people think about this?
Carl
(now in an even more irritable mood)

Reply:

Jacques Steyn wrote:

Date: Wed, 24 May 2000 15:57:44 +0200

Carl
Thanks for your comments.
What do you find irritating?
* use of colors
* "linear" presentation
* language usage
* browser incompatability
* etc

I'll be glad to know. I fully agree with leading by example.
Regards
Jacques

Antonio Vallone avallone@csc.com wrote:

Date: Wed, 24 May 2000 13:44:30 -0400 (EDT))

The shoemaker's children (was Re: Poor Usability of Award Winning

Elizabeth Buie (ebuie@csc.com) wrote:
>>Carl.Myhill@Smallworld.co.uk wrote:
>> ... so dazzled by the formatting that I found it impossible to skim the content.
Me too, although I'm not sure that "dazzled" is quite the word I'd use. Maybe I'll set my browser to "use my colours" temporarily and read it in black on white.
I believe that the comments may have been a little too strong toward Jacques Steyn's web rendition of his paper.
I am interested in usability as a systems engineer and here a few points from a user perspective:
1) I like to have a paper on one page. I can skim it quickly and, if I like it, I print it to read it at leisure (on the Metro);
2) Jacques Steyn includes a note explaining the presentation potential problems with current browser versions (I believe that he could have included a plainer version for "handicapped brouwsers");
3) I didn't notice any problem. I got a black text on white background. All links in blue turning magenta after use. I have default colour settings. And yes, I use my "Old Fateful" Netscape Navigator Gold 3.04. (I avoid using Netscape 4 and I don't like IE.)
I agree that accessibility is one of the major problems in Web site design. Several times I get a page telling me that my browser was decrepit, giving me the link for downloading the acceptable latest version, and, of course, turn JAVA on.

Carl.Myhill@Smallworld.co.uk wrote:

Date: Wed, 24 May 2000 14:43:45 +0100

RE: Poor Usability of Award Winning Web Sites

Blocks of colours mainly. Also:
- links stand out quite clearly but don't look like links;
- I think you've also gone a bit overboard with the fonts. I'd keep the number of different fonts minimal.
- also, I font size quite difficult to skim, even though it was large.
Have you tested this page with anyone?
I'm not a web expert at all, I just know that I found your page hard to read. I'm not a great fan of Jakon Nielsen's web pages either. However, I think we probably can learn from the likes of the BBC web page, which is easy to skim.
Thanks for listening.
Carl

Reply:

Jacques Steyn wrote:

Carl Thanks for your reply.
> Blocks of colours mainly.
* do you mean color combination?
* use of color in blocks?
> Also:
> - links stand out quite clearly but don't look like links;
Not underlined, yes. But that's the future. Most sites using style sheets use either color as cue or different font as cue that there is a link. In the beginning Web users had to learn to click those underlined thingies. Now there is another learning curve, again...
> > - I think you've also gone a bit overboard with the fonts. I'd keep the number of different fonts minimal.
What do you mean "different fonts"? There is only one font-family, in different sizes for different level headings. If you refer to font colors, that is another matter.
> > - also, I font size quite difficult to skim, even though it was large.
Interesting. But this is the paper itself, not a presentation. And it actually is not a typical web site either... No product is sold. Its function is an academic analysis for academic discussion -- which obviously due to the nature of the topic, will include the usability of this site itself.
> Have you tested this page with anyone?
Nope, which is why I posted it here.
> I'm not a web expert at all, I just know that I found your page hard to read. I'm not a great fan of Jakon Nielsen's web pages either. However, I think we probably can learn from the likes of the BBC web page, which is easy to skim.
Yeah, for skimming. But again, it was not designed for skimming. You need to follow the arguments.
But hey, please do continue with comments.
Thanks and regards
Jacques

Carl.Myhill@Smallworld.co.uk wrote:

Date: Wed, 24 May 2000 16:03:44 +0100

RE: Poor Usability of Award Winning Web Sites

Far be it for me to re-design you web page for you (I'm working myself!) but...
> > Blocks of colours mainly.
> > * do you mean color combination?
> * use of color in blocks?
Didn't like the use of color in blocks. But was none too impressed with the color combinations either.
> > Also: >> - links stand out quite clearly but don't look like links;
> > Not underlined, yes. But that's the future. Most sites using style sheets use either color as cue or different font as cue that there is a link. In the beginning Web users had to learn to click those underlined thingies. Now there is another learning curve, again...
> Again, I would refer you to the BBC web page. You're article is surely aimed at being read. It would help me as a reader to know what is a link and what isn't in a way that is well known, subtle and doesn't SHOUT at me!
> > What do you mean "different fonts"? There is only one font-family, in different sizes for different level headings. If you refer to font colors, that is another matter.
Have another look at your contents page. If it is all the same font family, it certainly doesn't look like it. You certainly have several different sizes and formats of the font, for example, "Abstract" is much smaller font than the comment next to it.
> > Yeah, for skimming. But again, it was not designed for skimming. You need to follow the arguments.
Have you read the Human Factors Society Authoring Guidelines? I've written several academic papers following these strict guidelines for presentation. The whole purpose of these is so that you can follow the arguments without having to fight with miscellaneous formats. You have a PhD yourself so presumably are well aware of just how many papers your average research student reads. Do you 'follow the arguments' for every paper you pick up? Or do you skim it first to see if it's worth reading?
> But hey, please do continue with comments.
OK
Carl
PS Someone just sent me this...
Hello Carl, I think, this site is meant to be funny. It deliberately breaks with all usability rules. So have a look again and enjoy the mistakes. ...are you really trying to make a joke with your paper?
I don't think so given how much effort it is to write a paper! Or have I missed the joke.

Reply:

Jacques Steyn wrote:

Date: Thu, 25 May 2000 12:05:23 +0200

Carl Again, thanks for your comments.
No, I do not want a re-design, but I think I have made my point. I have not received any rationalizations about the use of color. Comments are emotive/affective, which suggests a typical design problem with usability consequences: personal taste that is often generalized. My own view is that function comes before aesthetics.
As I wanted more conversation, I deliberatly did not say anything about your previous suggestion to use the BBC site as a good example of usability. Here's why:
1. Technical:
The W3C's WAI recommends *not* to use tables to enforce layout, the reason being that not all user agents can render tables, and speech synthesizers may get confused. Web authors hack tables to enforce layout in collumns. To be honest, the W3C also does that on their main page!! Columns will only be handled in CSS 3. Given that CSS 1.0 was released 17 December 1996 and today not a single browser supports this 100%, the columns of CSS 3 may only be implemented by 2010 (if we're lucky?). I authored the markup of my paper technically correct according to the strict DTD of XHTML. That means the document is accessible on all possible devices, thus (globally) usable, while any site that uses tables is not universally accessible.
2. Appearance:
It could be argued that the BBC's main page is cluttered. The reason is obviously to pack in as much info so that the user does not need to scroll. That's a functional solution to prevent too much scrolling, which indeed has a usability case. But you cannot universalize this approach to a particular type of web document to allow for all possible web documents. Just as function often determines the appearance of paper publications, the purpose etc will determine appearance and structure. A glossy magazine does look different from an annual financial account. My audience was semi-academic.
What browsers are supposed to do (but do not do yet as their makers focus on penetrating the market rather than providing functionality) is to allow a user to choose his/her own style sheet. In fact, it should be possible for authors to suggest a few different style sheets frmo which a user can select. A site such as the BBC's does not allow for that, while mine does. You should be able to choose your own style preferances with your browser. THAT's usability! Put the user in control!
3. Accessibility:
In addition to the above, the BBC site is not accessible (according to WAI recommendations). Given the Disabilities Acts today either already law or in the process of becoming law in many countries, and law suits against internet companies (eg AOL), the meaning of usability is much wider than before. A web document must be usable also to the disabled, and shluld be delivered on all kinds of devices (not just a fancy computer screen).
Anyway, on the style of the first posting of the paper: it was not a joke in the funny sense, but yes, I wanted to make a point. As said, I have not had any rational feedback about why those colors do not work. But as I have made my original point, I have changed the basic style sheet to an alternative one with mellow and pastel-like colors.
We have communicated privately, but I would like to put our correspondence on the site. Do you have any objections?
Thanks a lot again.
Your opinions are appreciated.
Regards
Jacques

Tharon howard tharon@hubcap.clemson.edu wrote:

Date: Wed, 24 May 2000 16:03:44 +0100

RE: Poor Usability of Award Winning Web Sites

Alice Preston writes:
> >Carl wrote: "...usability people should set an example to others. Failure to do so discredits the rest of us."
> >I say:
>Hear, Hear!
> >There is nothing so effective as teaching by example (and nothing so effective in undermining your message as unintentionally including a counter-example).
I would agree that we should all try to set an example as a goal, but we also need to apply some charity here. I've seen certain usability folks fail because they took an absolutist view and refused to compromise. They set their usability guidelines up as inviolate laws and refused to consider the particular and contingent needs/demands of their clients. The engineers and project managers with whom they were working got sick of what they regarded as unreasonable demands and went back to their old functionality-first approach. It's possible to go too far, and it's hard for an extremist to build alliances.
Tharon

Merryl Gross merryl@curl.com wrote:

Date: Fri, 26 May 2000 12:08:06 -0400 (EDT)

RE: Poor Usability of Award Winning Web Sites

I managed to wade through the entire paper, neon color scheme, bad list formatting, and all. While I agree with the essentials of the concepts, I thought the author was entirely too subjective and inflexible in his conclusions. I also thought the paper was very unfocused; it was hard to tell WHAT the author was trying to achieve by it. It seemed like he typed it all down as it occurred to him, then posted it, and never revised a thing.
If you have a site that's similar to the ones reviewed (or ARE the ones reviewed), I would treat this almost as a stream-of-consciousness transcript of a single usability test session.
Now for all you Don Norman fans... "It must have won a design prize..." has never been a favorable statement about usability! Sometimes I think we are seeing the equivalent of how architecture awards are given to buildings where the tenants dream of wringing the architect's neck because of the building's lack of usability.
Merryl Gross
Product Designer, Curl Corporation http://www.curl.com

Jurek Kirakowski jzk@ucc.ie jk wrote:

Date: Tue, 30 May 2000 13:59:21 +0100

award winning sites' usability

I read your paper with interest as I too am interested in the usability of web sites for which a lot of money has been spent: has it been well spent?
In comment to your paper, I found it difficult to agree with you that what you were actually reviewing for was 'usability' because you did not appear (and excuse me if I did not see this definition) to define the term operationally. That is, by what sequence of operations can we measure the amount of usability of a web site? Dillon's work is of interest of course, but unless you can develop a method of measuring on this basis of the definitions, it is as good as any other philosophical outlook on usability: ie, interesting but not proven.
Your comments on the web sites you have reviewed, while always informed, are ad hominem arguments based on your personal judgements to vaguely-defined criteria. We all have lots of these, and it's always fun when a group of experts has a 'jam session' where they compare their subjective reviews to a chosen site. But this is not engineering nor is it science.
For usability testing to evolve into an engineering discipline we must develop objective methods of test which will yield independently-corroboratable measures. Which methods we will develop will depend of course on what we perceive usability =is=... in this case, applied to web site design.
The (old) ISO 9241 definition of usability as 'effectiveness, efficiency and satisfaction' is a good working hypothesis. I and my colleagues at Nomos have worked on satisfaction metrics and naturally our experience leads us to believe that these metrics are more important than most. Some pundits in the usability world would maintain that effectiveness metrics (ie, could you do what you wanted to do) of a binary nature (yes/no) are about as much as one can hope for at this stage in the game. And many guidelines treat of efficiency thereby raising this quality of usability to pre-eminence, but few measure it.
What bedevils the whole area is simply the lack of scientific measurement. We are very much on our own with our carefully validated and relibaility-rated, benchmarked tool for satisfaction measurement. It would be very pleasing if your article sparked a wider interest in developing tools for the accurate measurement of usability of web sites althbough interestingly most comments on your paper prefer to focus on the 'usability of a usability paper.'
Thank you for the trouble of sharing your thoughts with us, and your courtesy of allowing your readership to reply.
Yours
Jurek Kirakowski

Jurek Kirakowski jzk@ucc.ie jk wrote:

Date: Tue, 30 May 2000 17:09:57 +0100

Re: award winning sites' usability

Hi Jacques
Just to carry on a bit (instead of getting on with work as I should be doing)...
>Striving for a numbers-based set of criteria is fine as a metric, for example,
>to determine how many mouse-clicks it takes to reach a destination -- and that
>I have done (eg for Nando's). But 19 mouse clicks does not say much. It makes
>no statement about quality or psychological experience and other such matters
>which cannot be quantified, only qualified.
Aye. But 19 mouse clicks is not meaningful because at least : (1) it is not a valid method for measuring anything other than mouse clicks, (2) it may not be a very reliable method for measuring usability (in the sense of correlation with some other kind of usability criterion), and as you note, (3) 19 mouse clicks does not stand in relation to any known set of distributions of mouse clicks, gathered either empirically or through the application of theory to some parameters.
Although we are agreed that mouse clicks don't amount to much, the conclusion, that there are no meaningful metrics of psychological experience, does not follow. This is what psychometrics tries to do and has been moderately successful in doing over the past 60 odd years.
>...depends on your definition of "science"; and then again, you assume that
>engineering is the paradigm that should be used to determine usability. If you
>wish to use the term "engineer" for software developers, I suggest we rather
>talk of "infoneers". Traditional engineers got their hands dirty with grease,
>coal dust and used massive spanners. Software 'engineers' do not work in this
>kind of environment. There may be a set of useful 'metrics' for traditional
>engineering, but not so for infoneering. You cannot use the same tools -- just
>like you cannot use Euclidean mathematics to create complex figures (you need
>fractals for that). A tool can be used only for what it is designed for.
>Metrics can measure how many bulbs on a factory line are faulty given certain
>contexts, and then perhaps some measures can be taken to reduce faults. You
>cannot measure the use of web sites in the same manner.
I'm afraid here is where we part company. I strongly believe, and I have spent a long time gathering data to prove this, that one can devise metrics which measure aspects of usability, and that these metrics can be devised and standardised so as to be valid, reliable, benchmarked and interpretable.
My dictionary gives the primary meaning of 'engineer' as 'one who designs & constructs military works; soldier of branch of army called E~s, trained to engineering.' I would therefore posit that any other use of the word 'engineer' is increasingly metaphorical hence its applicability to civilian artefacts (accepted), software (almost accepted) and then usability (contentious) :-)#
>It is possible to measure anything because the measuring system is designed for
>a purpose. I can set up a measuring system for bad breath and let users assign
>values to different odors. But does that guarantee objectivity? Remember
>Schrodinger's cat...
No, I wish to forget this poor cat. I prefer an operational definition of 'objectivity' in the sense of the mathematical definition of a 'metric.'
>> What bedevils the whole area is simply the lack of scientific measurement.
>> We are very much on our own with our carefully validated and
>> relibaility-rated, benchmarked tool for satisfaction measurement.
> >Yes, to the lack of "scientific measurement", but simultaneously "no". I argue
>that with systems modeling using newer mathematical tools we can develop more
>interesting tools for measurement (the yes). But then you should get out of the
>old paradigm of thinking with respect to words such as "validate", "measure"
>and "benchmark" (which scores a 'no'). I look forward to seeing the results of applying your mathematical tools to systems modelling. There is too much talk in HCI and not enough getting down to get things done. But if your new tools cannot be validated, if they do not appear to measure anything, and if they do not allow us to set benchmarks, then I'm afraid I must ask you: what are they good for?
Best
jk

Jacques Hugo jachugo@icon.co.za wrote:

Date: Tue, 30 May 2000 14:38:11 -0400 (EDT)

Usability of award winning web sites

Hi Jurek,
I'm so glad you picked up this one! Your reply to my namesake was exactly the kind of reponse I would have liked to make, but couldn't, for reasons you'll discover below!
For your information, here is my comment to Elizabeth Reed, who first discovered the paper on the University of Pretoria's site:
=============
I'm a bit hesitant to comment on this, because this paper was presented at CHI-SA 2000, the South African HCI conference earlier this month, of which I was the organiser and chair! That's why I reply to you directly and not to UTEST :)
The paper was presented by Prof Jacques Steyn who is an expert on HTML, XML, W3C etc. etc., but not necessarily usability. In principle I agree with him that award winning web sites are often not judged on usability principles, but more likely in terms of visual appeal, especially from a marketing point of view. Certainly user-centeredness is often overlooked in the kind of sites he had in mind. BTW, in South Africa we have an annual award called the Loerie (named after an exotic bird) which includes awards for web sites and TV advertisements. It probably won't surprise you that this award is given by representatives of the media industry and of course does not include any usability input!
You can probably guess that this paper was one that evoked a lot of controversy and negative comments at the conference, because people felt that his evaluation criteria (emphasis on W3C compliance) did not take adequate cognizance of usability criteria as we (UTESTers) think of them. Also, his choice of visual presentation (which looked much like his on-line paper) didn't go down too well with the audience!
================
Incidentally, I teach HCI part time at the U of Pretoria, and I run into Prof Steyn often - I'm discreetly going to try and find out how he is taking all of this flack! :)
Regards
Jacques Hugo
Usability Consultant

Caroline Jarrett Caroline.Jarrett@effortmark.co.uk wrote:

Date : Wed, 31 May 2000 20:32:30 +0100

Re: Usability of award winning web sites

I, too, found it difficult to absorb the content of the paper and comments due to the choices made visually in their presentation. As I found the paper physically hard to read, I couldn't devote enough brain-power to considering content.
Among Jurek's comments, he mentioned the ISO 9241:11 definition of usability as 'efficiency, effectivness and user satisfaction'.
I've been thinking about considering these as percentages. You could choose to assign a minimum of 1% to each, but the total has to add up to 100%. Then we try to measure how much has been achieved on each scale. So a site might aim for:
effectiveness (can you do the task) 75%
efficiency (can you do it in an appropriate time) 24%
user satisfaction (will you enjoy doing it) 1%
(meaning: you'll hate the experience but you'll be able to get it over with quickly) and it might achieve:
effectiveness 25%
efficiency 3%
user satisfaction 1%
(meaning: yes, it was horrible to do but also we couldn't do it very well).
I don't know whether this works. For example, what is 1% as an achievement? Should I multiply by the orginal goal to get a 'satisfaction contribution' to the overall score of .01%?
This relates back to the orginal paper because I've been thinking that there is a tendency in the usability community to rate effectiveness as independent of user satisfaction. But some sites might split their aims this way:
effectiveness 1%
efficiency 1%
satisfaction 98%
(meaning: you may achieve nothing, it may take forever, but you'll be enjoying yourself). And there are beautiful, engaging, but almost completely useless sites out there that fit this profile rather well.
For example, I recall with pleasure the old directdebit.co.uk (now, sadly, defunct) that we discussed some time ago on this list. For those who didn't see it at the time, this was a Monty-Python-esque voyage through a British method of paying your bills - a boring topic, handled in a way that some of us found amusing and others horribly unusable.
I suggest that prizes are awarded to websites for all sorts of reasons, and it doesn't _necessarily_ mean that the site is without merit if it isn't particularly effective or efficient.
Of course, some sites fail all three criteria spectacularly, but that's another story.
Caroline Jarrett

Whitney Quesenbery whitneyq@cognetics.com wrote:

Date: Thu, 1 Jun 2000 13:38:29 -0400 (EDT)

Re: Usability of award winning web sites (long)

>I've been thinking about considering these as
>percentages.
>You could choose to assign a minimum of 1% to each,
>but the total has to add up to 100%. Then we try
>to measure how much has been achieved on each scale.
We do this exercise with projects on a regular basis, but with five characteristics (the "5 E's")
Effective
Efficient
Engaging
Error Tolerant
Easy to Learn
We also force it to add up to 100%, to avoid simply wanting "everything" (not the 6th 'E').
It has been very useful not only for our own use, but as part of a guided discussion with clients on the different usability goals which might be appropriate for either a specific program or user group.
In one case, for example, there were different groups of users who had very different profiles in this exercise (and the people who made the selection of software had one different from any of the users), and the design challenge became how to meet the needs of each of them.
We also construct statements which reflect the user's perspective on each characteristic. These are done in first person - "I need to do my work as fast as possible, because I am judged on how many ... I can complete" (for a crude example). They may be made up as best guesss at first, but are modified over time as we learn more about users through c-inquiry, interviews, etc.

Merryl Gross merryl@curl.com wrote:

Date: Thu, 1 Jun 2000 13:59:32 -0400

Re: More: Usability of award winning web sites

One thing about the color that I think you are overlooking is that it's not a "mere" personal preferences issue. There are signs from more than one respondent that the layout and color scheme caused this paper to become painful to attempt to read. Many may have fled from it without reading anything. Examples:
- the extreme, unnecessary contrast between the background of the page and the background of the text paragraphs;
- the lack of between-paragraph whitespace (which you continued in your reprint of my critique, to my chagrin!);
- the fact that all the bullets in your bulleted list became nearly invisible on the dark blue background while the text was on the bright blue;
- the unusual style of your hyperlinks and the fact that they jumped around during a mouseover.
Please, PLEASE read Robin Williams' excellent book The Non-Designer's Design Book to understand why this is NOT a personal preference issue, but a violation of all the things we have learned over hundreds of years of setting type. You could certainly get away with most of your color scheme, IF you followed other principles in this book.
Please forgive me if this realization was the underlying purpose behind your layout. However, you would make your point better if you pointed that out... this reader is currently left with the belief that you don't really care to make your work readable for some reason. Or that you have some sort of color vision abnormality. Either way, it casts what conclusions you come to in the paper into a bad light.
Merryl Gross

Reply:

General comments:
1. Personal preferance
The fact that a statistically higher number of respondents dislike certain features does not make it objective. There are respondents who actually like the blue. "Personal preferance" does not necessarily mean idiosyncratic and individualistic. As is evident from this discussion, personal preferances can be grouped. So the question is then, for which group must we develop?
2. Layout -- eg whitespace, font, etc
You cannot compare paper publishing with web publishing. As pointed out in the paper, the Web has its own unique contraints and a long way from being comparable to paper publishing. Although the latest W3C recommendations (eg for HTML, CSS, XHTML, XML etc) move closer to what is possible on paper, implementation of these recommendations is extremely poor. All your comments concerning appearance are relevant, yes. But technically it is impossible to universally implement such things as whitespace between paragraphs -- it can be indicated in markup, but browsers (generally speaking) do not implement this.
Bullet symbols that disappear: this is due to poor style sheet implementation of browsers.
Hyperlinks that jump around: this is due to a specific IE implementation of W3C recommendations. Note that it does not happen in Netscape.
Why don't you join the pressure group of web developers promoting standards so that browser developers follow recommendations and standards and authors can then comply with your dissatisfactions -- http://www.webstandards.org/ . As soon as browser developers get their act together the Web will become more accessible and usable, and then most of the comments on appearance could be avoided.
Shoot again....
Regards
Jacques

Caroline Jarrett Caroline.Jarrett@effortmark.co.uk wrote:

Date: Fri, 2 Jun 2000 08:53:16 +0100

Re: More: Usability of award winning web sites

The issue here isn't about statistics, which tells you whether an opinion from a sample is likely to be the true opinion of your population.
The issue is about _why_ your correspondents are saying they don't like the way your paper is shown. This could apply to a different feature of a Web or other interface, so I will persist with the argument.
Legibility and readability are a mesh of several factors, but to make the discussion a bit simpler I will choose a single one: the use of a large, tightly packed bold font.
There seem to be three available opinions here:
(1) Like it. (I assume you are in this category or you wouldn't have chosen it)
(2) Neutral. (Don't care, don't have an opinion, neither like nor dislike)
(3) Dislike it.
People mentioning opinion (3) have mostly said that they don't like it and also it has prevented them reading your content comfortably.
Leaving aside any question of statistics, your choices are:
(a) Do nothing. This means you lose the readership with opinion(3).
(b) Decide that you want to retain all three groups in your readership.
This means finding a font that is easier to read for opinion (3) people. It seems likely that opinion (2) people will continue not to care, so you won't have affected them. There is a risk that some people with opinion (1) loved your current choice so much that changing to a lighter choice will alienate them. On the other hand, some people with opinion (1) will approve of any choice you make, and others will (perhaps to their surprise) find that they also like your lighter font.
It's your choice. Choice (b) is likely to increase your overall readership, and you will still be designing for all three groups. Only you can balance the risk that the readers you lose from opinion (1) - which won't be all of them - are more important users to you than the readers you gain from opinion (3) - which probably also won't be all of them, people being opinionated.
> 2. Layout -- eg whitespace, font, etc
> You cannot compare paper publishing with web publishing. As pointed out in the
Well, you can compare them in many ways. OK, the web does have some different challenges and constraints, but the eyes and brains of your readers have the same makeup when they're reading paper or on-screen. And some aspects of legibility and readability are pretty much constant on web and paper. Let's look at the issue of contrast. Contrast between text and ground is important for legibility. The makeup of that contrast differs on screen and paper. For example, the screen emits light whereas paper reflects it. However, the issue continues to be important. Yellow text on a white ground will be relatively harder to read than the same text in black on a white ground.

steve.slatcher@pobox.com Steve Slatcher wrote:

Date: Fri, 2 Jun 2000 04:39:15 -0400 (EDT)

Re: More: Usability of award winning web sites

> The fact that a statistically higher number of respondents dislike certain
> features does not make it objective
But it might point to a feature that *is* bad for an objective reason, e.g. how the eye and human perception works.
> Bullet symbols that disappear: this is due to poor style sheet implementation
> of browsers.
> Hyperlinks that jump around: this is due to a specific IE implementation of
> W3C
> recommendations. Note that it does not happen in Netscape.
Yes. Browsers have lousy implementations of standards. I see that as an argument for not using features that are not properly supported. There are many representations of hyperlinks that work fine. Many of these are also closer to what a user would expect.
> Why don't you join the pressure group of web developers promoting standards
> so
> that browser developers follow recommendations and standards and authors can
> then comply with your dissatisfactions -- http://www.webstandards.org/ . As
> soon as browser developers get their act together the Web will become more
> accessible and usable
Agreed 100%. Webstandards goals are spot on.
>, and then most of the comments on appearance could be
> avoided. ..but not sure about this.
> Shoot again....
I think I have :-)

And here the plug was pulled, so unfortunately the discussion could not continue.

Comments on my article "Poor Usability of Award Winning Web Sites"

Comments on my article
"Poor Usability of Award Winning Web Sites"