Log in   Register a New Account

Accessify Forum - Discuss Website Accessibility

New to the forum?

Only an email address is required.

Register Here

Already registered? Log In

Self-referential links

  • This topic is locked: you cannot edit posts or make replies.
  • Post new topic
Reply with quote What's everyone opinion of self-referential links in relation to checkpoint 13.1 Clearly identify the target of each link? Are self-referential links a failure under this checkpoint, I see no direct comment in the WCAG techniques but always have thought of it as best practice. Recommendation or failure?
Reply with quote can you just give an example of what you mean by "self-referential"?

Patrick H. Lauke / splintered
Reply with quote If we're talking about links which point to the same page they appear on (e.g. an active "Home" link on the home page), I don't *think* they come under a specific WAI checkpoint. A link that reads "Home" is pretty clear re where the link goes. The fact that you're already on the Home page doesn't make the link itself any less clear.

I've always regarded this as very much a usability issue rather than one of accessibility - having self-referential links doesn't actually prevent anyone from accessing anything, whereas it does create the potential for confusion, so best practice would be to avoid creating such links.

Donna Smillie / Senior Web Accessibility Consultant, RNIB
Web Access Centre / WAC Blog
Reply with quote Redux dms just quoted an example of what I mean. Basically a link which reloads the current page is what I was referring to.
Reply with quote A link should not point to the current page as it is a usability problem. Clicking a link only for the same page to be displayed is a confusing response; users expect a hyperlink to replace the current page with a new page of information (or jump to a section within the page). A link which does neither of these things seems to be acting erroneously and undermines a user's confidence in the navigation system of a website. They won't be sure whether they really are where the link said they should have gone to.

I don't think this would come under accessibility but it's certainly a poor piece of usability.

If a page displays content which frequently changes, there are a couple of option:-
  • An Applet could be used to update the page programmatically
  • A "Reload Page" command button could be used

Because an Applet can produce accessibility problems, a command button which refreshes the page would be the better solution, imho. Dynamic navigation systems should be sufficiently intelligent to remove the anchor from links which point to the current page (unless they are fragment links within the page).
Reply with quote For any page I work on which refers to itself than I avoid making the link live, as Donna says, it just confuses people. I also avoid this in navigation lists
Code:
<ul>
<li title="This page">Home</li>
<li><a href="...." title="Next page">Rhubarb</a></li>
</ul>
Don't forget that reloading the same page can cause confusion as the users may lose their place Crying or Very sad

Mike Abbott
Accessible to everyone
Reply with quote
Quote:

I also avoid this in navigation lists
Code:
<ul>
<li title="This page">Home</li>
<li><a href="...." title="Next page">Rhubarb</a></li>
</ul>



Sorry, I'm a little confused. What is that you avoid?

I'm currently inclined to leave the structure of the navigation in place, but remove the link, and some how indicate the active status of the ls element with css. The only usability issue I see with it is that my menu contains a choice with doesn't have any action associated with it, which is a little confusing. But alternatively, having links disappearing and reappearing is also rather confusing?

Craig Loftus
Reply with quote I'm not convinced by the argument stating that 'self referential' links create (the potential for) confusion, so should be removed or disabled.
If a person clicks on the 'home' link unaware that they are already on the 'home' page, then it seems obvious to me that such a person is just as likely to be confused by not being having the means to navigate to the 'home' page whilst being unaware that they're already there.

The basis for both points is that a user doesn't know where they currently are in the site. I don't see how the removal of the self-referring link improves usability any more than keeping it, given that some people will find either scenario equally confusing.

My view is that navigation should stay consistent throughout the site. I feel that the inconsistency that is being recommended could easily confuse those it aims to assist.

Of course, it should be made clear, both through the use of CSS and, more robustly, through the use of section heading text, where the user currently is on the site. If they can be made aware of where they are through the use of headings and the re-styling of (still active) self-referential menu links, then there is no real need to disrupt the consistency of a site's navigation.

Consistency in UI design is considered by many to be one of the cornerstones of good usability.
I can certainly recognise that consistency would be highly appreciated by users with a predisposition towards confusion.
Reply with quote Sorry Craig (BTW welcome on board Very Happy )

I avoid having links to the open page on the open page, which is effectively a loop.

Bill, my thought was directed to those with a mental incapacity who may not understand why it does not re-direct to another page. A bit extreme I think, but nonetheless valid.

I do not understand some of your argument
Bill Posters wrote:
My view is that navigation should stay consistent throughout the site. I feel that the inconsistency that is being recommended could easily confuse those it aims to assist
can you explain the inconsistency please

Mike Abbott
Accessible to everyone
Reply with quote
Mikea wrote:
I do not understand some of your argument

can you explain the inconsistency please

The gist of what's being recommended is to either remove or disable self-referential menu links?
Menus built according to that ethos won't appear or work the same on every page. The menu will change from section to section.
That's the inconsistency to which I'm referring and it's an inconsistency which I*, and many developers working in UI development, consider to have a potentially negative impact on usability, something which is likely to be exacerbated for those with the kind of mental incapacity you aim to assist.

* I do not consider myself to be a UI developer, per se. My only experience is through my work as a graphic designer. I'm happy to defer to and recognise the experience and knowledge of others on the subject - many (if not most) of whom consider consistency in UI design to be of fundamental importance to good usability. It is a conclusion which seems both straight-forward and fairly obvious to my way of thinking.

As a user convention, messing about with the menus simply seems like a counter-intuitive thing to do and only serves to muddy the water.
Rather than the menu doing the same thing on every page, users are required to learn the pattern of menu behaviour. To me, that seems like you're making more work for users, not less; something which is likely to have the greatest impact on those with forms of mental incapacity.


Last edited by Bill Posters on 27 Oct 2005 10:12 am; edited 1 time in total
Reply with quote I always used to deactivate self referential links, but since using styled lists it's not always possible.

A lot of navigation that uses styled lists, such as those featured on listamatic, style the "a" element. If the link is removed the style has to be applied some other way.

By styling the "a" element you can also create a larger clickable area, which you could argue actually improves accessibility and/or usability.
Reply with quote for me it's not so much the styling issue (you could always put a span around the self-referential, rather than an a, and style it accordingly), but the fact that you can't simply have a straight include with your navigation...you need some logic that says "if this link refers to the current page, output a span instead".

this could be done with some nice DOM scripting though as a "sugar coating" extra usability functionality.

Patrick H. Lauke / splintered
Reply with quote
asaxton wrote:
Cerbera wrote:
Clicking a link only for the same page to be displayed is a confusing response; users expect a hyperlink to replace the current page with a new page of information (or jump to a section within the page).
Where did you get this information from? Without wanting to sound confrontational; is this something you can back up with references / research or is it just your opinion?
No, it isn't just my opinion. I would reference my sources more regularly but my bookmarks are a mess. Sad

I've used many sources and observations to arrive at my current view. The most important one was somewhere in Tim Berners-Lee's Design Issues section, where he specifically describes the hyperlink used by HTML as "replacing the current page with a new page" or words to that effect. I can't find the damn page, though! I found a couple of pages which mention it casually but I can't find the one where it's really specific.

Anyway, this is what the vast majority of links one may click on will do so it's the expected behaviour. Although some designers go on a power trip and force new windows on the user (often using Javascript) and other unwanted complications, replacing the current page with a new page is what a link will normally do. Replacing the current page with the current page is not.

One of the main conclusions from usability studies which convinced me of self-referring hyperlinks being damaging to navigation was number ten in Nielson's 10 Most Violated Homepage Guidelines. He uses strong emphasis to urge against self-referrential links. One particular point he makes: "Worst of all, if users do follow these no-op links they'll be confused as to their new location, particularly if the page is scrolled back to the top."

asaxton wrote:
Cerbera wrote:
A link which does neither of these things seems to be acting erroneously and undermines a user's confidence in the navigation system of a website. They won't be sure whether they really are where the link said they should have gone to.
But shurely the same can be said for the method you suggest using instead?

When users expect something to be a link (because they have "just" clicked on it) and then it is suddenly not a link anymore (read: disabled) this causes just as much confusion. I have personally observed this and it still happens even when the links appearence is changed significantly (background colour changed / underline removed etc)

I witnessed this very problem (Disabled "links") only a couple of days ago with my wifes dad who is new to computers, let alone the internet. I had to tell him to look for the hand cursor to make sure its a link and not just text.
Very new users to computers will always have difficulties getting started. It takes a while to learn how your stereo works, how your TV works, hell even the washing machine! Learning about how the cursor changes to indicate a link comes before learning how to use Website navigation systems. It's like learning how the volume works before setting up the equaliser, imho.

Disabling (or removing) useless menu items has been present in user interfaces for many, many decades. Check the Edit menu of your browser when no text is selected; the options for Cut, Copy and Delete are disabled because they would not do anything. If your browser has a Window menu, an item is removed when the window it refers to is closed. This is to avoid redundant commands. In the case of Website navigation, self-referring links are worse than a "do nothing" command since they do something which is confusingly close to being nothing (indicate the link goes to a new page but reload the current, scrolled to the start).

Self-referring links behave like "Return to top" links but in the wrong location with the wrong text. It seems like an error has occurred, like the link is broken. In navigation boxes, highlighting and delinking the current location helps users to visualise where they are amongst the navigation options and prevents them following links which don't go anywhere.


asaxton wrote:
Not sure on the percentages to be honest but Nielsen looks for a certain level of implementation before something is assmed to be a "norm".

The best recent example is clicking on a sites logo to go to the homepage. This is a relatively new "standard behaviour" which i noticed gathering n popularity.
This isn't recent and it's extremely widespread.

asaxton wrote:
I certainly wouldnt say that what you are suggesting is the norm at the moment, but having said that it doesnt mean it isn't a good idea.

May just mean that most people havent picked up on it yet.
I think the problem is the complexity of implementation. Any developer can see that a hyperlink to the current page is an entirely useless feature but writing the code which adapts the navigation to prevent this situation is complicated. Developers often place their own conveniance over their user's conveniance, so they don't bother writing tricky code or fixing broken links and many other things which would be beneficial to users.

One of the main reasons I create my pages in text editors and don't use dynamic navigation is because writing the code to prevent self-linking would be far beyond my ability. I do it manually, which isn't practical for most Websites. It makes sense to avoid pointless links and highlighting the current location within navigation is helpful, even though both of these things are difficult to code.


Last edited by Ben Millard on 27 Oct 2005 03:05 pm; edited 1 time in total
Reply with quote Just some personal reflections on this based on my own experience of using the web:

Over time, as a result of using many websites, I've developed an expectation that an active link will take me somewhere else - somewhere other than where I currently am. For years, main site navigation menus have been constructed in such a way that the link to the page you're currently on is de-activated, so that's something which, without having ever thought about it until now, I pick up on as a confirmation that I *am* where I *think* I am. So now, with sites increasingly using navigation menus where every link is active, regardless of where you are, I find that my confidence in my knowledge of where I am in the site is knocked back momentarily if I find that the link that points to where I think I am is active - to me, that means I must be somewhere else. And if I then click on that link in order to go where I thought I already was, and find that I'm still on that same page - for me it's only a momentary annoyance, but for a less experienced user I suspect (and this is just a personal opinion) that the confusion is likely to be greater.

And to add an anecdotal reference of my own Smile I've seen this element of confusion in my mother. She's technically pretty competent (VERY competent given her age - almost 80), but can become confused when using the PC if things suddenly do something different to what she is expecting. And she too has developed an inner mental map of how websites work that includes the expectation that if a link is active, it means you'll be taken somewhere other than where you currently are. She'll move the mouse cursor over the link, and if she sees that the text identifying the page she is is an active link, she starts to doubt that she's actually where she thought she was, and will click on that link in order to get where she thought she was. And is then doubly puzzled when it lands her right back on the same page, with that link still active.

Entirely subjective and anecdotal I know, but our own experiences can't help but inform our judgements on these things, even if we ensure we don't base those judgements solely on our own experience.

And now I'm starting to witter I think - dosed up for a bad cold at the moment, and my brain feels rather fuzzy and a bit wobbly - kind of like a baby hedgehog...

Donna Smillie / Senior Web Accessibility Consultant, RNIB
Web Access Centre / WAC Blog
Reply with quote
dms wrote:
Over time, as a result of using many websites, I've developed an expectation that an active link will take me somewhere else - somewhere other than where I currently am. For years, main site navigation menus have been constructed in such a way that the link to the page you're currently on is de-activated,…

I can't personally recall any site which does that.
A list of less recents sites would be genuinely appreciated that demonstrate this long history of the type of menu manipulation which you describe.


It's all getting quite anecdotal, which is rarely ever the best way to thrash out the policy behind best practice.
My own experience of the web, as well as those whom I've observed, runs along decidedly different lines and leads me to the conclusions at which I've arrived.
Even Nielsen, cited in defense of the removal or disabling of self-referential link can also be quoted as driving home the importance of consistency.
Quote:
Since the dawn of time (1984), we have known that consistency is one of the strongest contributors to usability.
(source)



Consistency is the key to usable interaction design: when all interface elements look and function the same, users feel more confident using the site because they can transfer their learning from one subsite to the next rather than having to learn everything over again for each new page.
(source)

(emphasis added by original author)

Lies, damned lies and statistics - and Nielsen, perhaps? Wink


I feel that a valid, though not necessarily convincing point has been made in favour of disabling self-referential or other 'redundant' links, by taking software UI as an example. However, I'm still some way off of being convinced that removing said links entirely can ever be a good idea.

I still feel that an entirely consistent and entirely functioning menu combined with considered visual and textual cues is the very best way to present menus whilst clearly imparting the user's location to them. The devising and implementation of such visual and textual cues should be well within the means of any designer worthy of the job title.

Display posts from previous:   

Page 1 of 3

Goto page 1, 2, 3  Next

All times are GMT

  • This topic is locked: you cannot edit posts or make replies.
  • Post new topic