- From: David Woolley <david@djwhome.demon.co.uk>
- Date: Wed, 1 Oct 2003 22:48:08 +0100 (BST)
- To: w3c-wai-ig@w3.org
> - they have difficulty reading text on screen Screens tend to fall between the reading and distance prescriptions for spectacles, so are not covered by standard bi-focals, or separate reading glasses. One has to be fairly highly motivated in the first place to buy glasses to a special prescription for screen reading. It is also fiddly to get the text into the right zone on vari-focals, and only a small part of the screen will be in focus. However, for use in private, I suspect a cheaper alternative might be clip on magnifiers. 2 dioptre ones would convert a distance prescription to 50cms, which not to unreasonable a distance from the screen. I can't see many people using them at public terminals, though. > - have difficulty with window management (pop-ups) I think the key problem with these is they are unexpected behaviour and likely to make people think they have done something wrong and panic. > - have difficulty with mouseover navigation Do you mean pull down menus. I have problems with these. You need quite a lot of dexterity, and with some of the standard Windows ones, if you slip off an item you can have to start over (I know that you can click to lock the windows ones down, but I don't, and I haven't noticed that in scripted ones on web pages). They are also very likely to interact badly with accessibility options. E.g. they often assume that colours are enabled, but I've turned them off on occasions because of sites that show white on white text (scripted on, if you enable scripts) and I often like to see visited links distinguished. I recently had a site for which I had to re-enable author font sizes before I could use one because the menu turned off when I went outside the bounding box it would have had if it had used the designer's font specification. Of course turning off scripting, will produce a fallback without these problems on an accessible site; unfortunately that means knowing how to control the option and turning them on and off as you visit bad sites. > > Note: This new design added text to explain icons, made icons links, > simplified choices, reduced unnecessary text, created a my profile page, Once you've got to the stage of a my profile page you should consider that the proper place to do some of these things is in the user agent, and if someone understands how to configure a profile, they should be able to configure the user agent. Of course, in the real world, people don't have this level of skill, may not "own" the browser, and may want to visit sites where settings which improve properly written sites will break the site completely (see pull down menu cases above). > - use of action words appeared to compel older users to click with more > confidence (action terms such as "go to 401K" or "view 401K" vs the simple > "401K" helped users understand that these were links) I feel uncomfortable about this. It goes against the whole hypertext paradigm and it would not, in my view, be necessary if designers didn't disguise links in the first place. It is of course the reason that authors like calling every link "click here". > - the more clear the resulting action for the link, the more likely older > users will click it (and the faster they will click it) I suspect these links are really controls. Maybe they should presented as such, i.e. as buttons. > and layout STABLE...don't introduce needless changes. And don't change from the browser defaults. > Some Suggestions from her findings: > - use a hover color for links (give user feedback) I think you will find that all browsers that implement :hover have an option to do this for all unstyled links. On IE it is under colours, so is not even in the accessibility ghetto. > - user a larger target for clicking (no small images) I would prefer that you didn't use images at all. Images are a primary reason why people are unable to identify links. > - older users don't understand web terms: window, login, link, URL, menu > bar, toolbar, minimize, go back, home etc. Most of these are not web terms, they are windowing GUI terms, which points out that you are not just dealing with a lack of web literacy but a general lack of computer literacy. Kiosk applications will be problematic for this group as they assume familiarity with HCI devices (I actually found that touch screens where very difficult to use on graphical web pages compared with a mouse, given that I'm familiar with a mouse). For using their own browser, I feel that an effort should be made to teach the goback and link concept. Non-kiosk systems should not care about the users home page (even though some sites might like to script the home page to be theirs!) I don't see that this sort of site should ever need to refer to a URL, although it is an important concept for all but the most basic web use. I don't think there is any better non-computer term for login, and if a login is needed for security purposes (rather than simply to join up click trails) it is important that the user understands the concept and that of logging out. > - even though it does not increase performance provide a way to increase > the fonts, like using a visible button to change the text size or use > scalable fonts. (they PREFER larger text) Again, this is really a browser issue, and my advice would be to turn off author fonts on any browser that they control, as it will improve access to most sites, these days. Unfortunately this does break some sites badly. Scalable fonts have been used on Windows since Window 3.1. I think you mean relative font sizes. These should always be used and there is a WCAG guideline to that effect. > - older users have difficulty clicking on text links because the target is > too small. That can, in part, be a fixed by letting the user choose the text size. > - use images to increase the target size of links I strongly disagree. I think the use of images as links is the prime reason that people cannot work out what is and what is not a link. I'd suggest using a large line height (remember to use ems, not px or pt) would be a better solution. Where the link is not really a link, I would suggest buttons instead (this will confuse search engines if used for real links). Note that, long term, all elements will become potential links. Except in lists (which are not strictly within the hypertext paradigm, anyway), real links in real hypertext should not he that dense as to be confused. Note that at least some of the pointing problem is the result of not knowing how to use a mouse, rather than a lack of dexterity, e.g. not resting the heel of the hand on the mat when using one, or stabbing at the buttons rather than simply flexing a finger. > In a third test Ms. Chadwick-Dias found that older users were less trusting > than younger users. However, it was argued that this could have been I think that is probably true. Young people tend to believe that busineses are for their benefit, not the business owner's. Older people know that they must take charge of their own destiny.
Received on Wednesday, 1 October 2003 17:48:40 UTC