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

Linking to current page

  • Reply to topic
  • Post new topic
Reply with quote The thread on "Self-referential links" petered out due to a lack of information, and perhaps slightly too much opinion? Wink

I don't want to start another huge thread, but I've seen the results of both situations many times in usability testing, and it's clear to me which works better in the vast majority of cases.

The amount of times I've seen non-technical people click on links for the current page and get confused is enormous, probably 1/4 of all the (hundreds of) tests I've done in the last 5 years. It's not just me either (Obligatory Neilson quote - number 10).

What generally happens is that the person doesn't find what they are looking for in the content, and sees a promising link - not realising they are on that page already. They click it, and the page reloads. Sometimes (depending on scrolling) they realise it has reloaded, sometimes they don't and just wait. Either way is confusing.

Based on the evidence, we changed the way we did navigation (and usually logo links to).

What happens when the current page isn't a link (in the navigation)? It doesn't work. It fairly obviously (depending on styling) doesn't work, so the user tries something else.

This is primarily a usability issue, but it can affect people using access technologies more severely as it takes longer to orient yourself. I've seen people using screen-magnifiers click on the current page's link in the navigation up to 3 times as the part of the screen they were viewing jumped because of the page refresh, and they couldn't see the link change. It's not so bad for screen reader users, but still at least as bad a fully sited people.

In terms of sites that do this well, I just found that the examples I remember doing this have changed (BBC & IBM's ease of use dept). I tend to find (with clients) that making the current page a link is due to it being easiest to implement, but we generally persuade them to try!

It would be very interesting to check the logs of a site that has links to the current page, to see for how many page hits the URL matches the referer...
Reply with quote
Alastc wrote:
The amount of times I've seen non-technical people click on links for the current page and get confused is enormous, probably 1/4 of all the (hundreds of) tests I've done in the last 5 years.

Have you got any documentation to back this up, that you can make available? Otherwise it's just more anecdotal stuff.... not that I'm saying you're wrong, just without evidence it's merely opinion...

Alastc wrote:
In terms of sites that do this well, I just found that the examples I remember doing this have changed (BBC & IBM's ease of use dept). ...

Just to back you up on this - here's evidence that the BBC has a link on the home page, that takes you back to the home page - and so does IBM - both the US and UK versions.
Reply with quote
JackP wrote:

Have you got any documentation to back this up


I'm afraid client stuff is predictably confidential, and we don't have links to the current page on our own site - not much help really. Doesn't anyone else do usability testing here? It comes up so often I'd be amazed if it wasn't spotted in any test of a site that does that.

JackP wrote:

Just to back you up on this - here's evidence that the BBC has a link on the home page, that takes you back to the home page - and so does IBM - both the US and UK versions.


Admittedly no the kind of backup I was looking for! Sad
IBM's usability section and (I think, but could be wrong) the BBC used to make the current page a non-link (in the navigation at least):
Wayback machine for IBM

I suspect they've changed to due implementation issues.

It's not the end of the world usability or accessibility wise, it's just that given a choice, I would always make the current page not a link.
Reply with quote Thanks for the Wayback link! I never knew that site existed.
Reply with quote Yea, cool but scary if you have a history on the web!
Reply with quote
asaxton wrote:
JackP wrote:
the BBC has a link on the home page, that takes you back to the home page - and so does IBM - both the US and UK versions.


I wasn't going to mention it because it's not exactly the same and it could have tipped the other thread over the edge, but there is a link back to the homepage from the homepage on http://www.sitesurgeon.co.uk/ too Wink (tip: logo in top left)


I similarly didn't have the heart/energy to point out the many popular sites which retained self-referential links ('s-r links'). A little closer to home, our own beloved accessify.com and accessifyforum.com retain both 'home logos' and s-r links.
A few major sites didn't have clickable home logos, MSN.com was one I think and Yahoo.com was another, but both sites used s-r links in their menus.

In my entirely unscientific skim through some of the web's most popular sites showed that a significant majority used 'home' logos and an ever larger proportion retained s-r menu links.

Imho, these sites are as influential as TBL/W3C (probably moreso) when it comes to establishing the kind of usability expectations which lead to usability conventions (particularly in cases where there is no clear right or wrong approach).


A quick straw poll on the use of 'home' logos and s-r menu links…

bbc.co.uk -- both
wanadoo.co.uk -- both
amazon.com & .co.uk -- both
msn.com & .co.uk -- s-r links only
google.com & .co.uk -- neither
yahoo.com & .co.uk -- s-r links only
cnn.com -- both
w3.org -- neither
aol.com -- both
microsoft.com -- s-r links only on root page, but both on sub-sections
excite.com -- home logo only
wired.com -- both
lycos.com -- both, though home logo only on some sub-sections


…fwiw
Reply with quote Popular doesn't equal perfectly usable, I brought up the (old) IBM example as a site produced by people into usability that did it well.

To clarify, I'm primarily refering to navigation bar/lists etc, not logo-links. Not enough people use those to actually make it an issue.

I put a quick post to CHI-web (3000 usability specialists), asking if anyone else had seen it or had some evidence.

There were several replies from people who'd noted the issue, and one who might be able to provide experimental evidence from a recent usability study:
http://listserv.acm.org/scripts/wa.exe?A2=ind0511a&L=chi-web&F=&S=&P=391

Another useful one was from the web-blooper book guy, who pointed to his explanation:
http://www.uiwizards.com/wBloop8.html

A couple of people didn't think it was an issue, so I guess your mileage may vary... the thread continues:
http://listserv.acm.org/scripts/wa.exe?A1=ind0511a&L=chi-web#2
Reply with quote Jakob Nielsen bases his decisions on research but its obvious to de-activate the link where possible.

Johan De Silva / Portfolio
Reply with quote
Johan007 wrote:
Jakob Nielsen bases his decisions on research but its obvious to de-activate the link where possible.


Obvious because…?

Anyone who disagrees is 'obviously' wrong?
Experience which has witnessed a different behaviour is obviously worthless?
Reply with quote Alastc, Your first link was good - while it's still anecdotal evidence it's starting to be supporting anecdotal evidence. Case studies/case histories if you like.

My personal pref would be to still include the text but disable the link but at the moment all I can say is that it's a personal pref - I can't provide evidence to back it up
Alastc wrote:
Another useful one was from the web-blooper book guy, who pointed to his explanation:
http://www.uiwizards.com/wBloop8.html

However, this person makes some good points but there are also flaws in the argument.
the web-blooper book guy wrote:
At best, this wastes people's time as the page reloads. At worst, if they didn't realize the link just redisplays the page they were on, it can be disorienting. Also, any data entered into forms on the page will be lost.

If you're sending someone to the same page, it wastes their time, I agree.
I also agree it may be disorientating - particularly if you had scrolled down the page.
However, when you click on any link on any page (unless it's a submit button) and it takes you to a different page, any data entered into forms in that page will be lost. It seems to imply that if it was sent somewhere other than the same page, that would be ok. I don't imagine many users (*opinion, not proof*) would half-fill a form, and then click on a link, because they would already expect that to take them to a new page and lose their form contents.

What we really need, though is stats to indicate:
--are people confused by self-referential links on a nav bar?
--are people confused by self-referential links in the main content?

(I'd suspect more of the second than the first. Again my opinion : I'd say the second is a big no no, but the first may be open to debate )

and if so:
--what proportion of people find a disabled item in a nav bar confusing?
--what proportion of people find a nav bar with items missing confusing?
Reply with quote
JackP wrote:
when you click on any link on any page (unless it's a submit button) and it takes you to a different page, any data entered into forms in that page will be lost. It seems to imply that if it was sent somewhere other than the same page, that would be ok.

I don't imagine many users (*opinion, not proof*) would half-fill a form, and then click on a link, because they would already expect that to take them to a new page and lose their form contents.


I'm not sure where you're going with that, if you half fill a form and click on any link, you're likely to loose your data. If you wanted to prevent that, you'd disable all links, but I'm sure that's not where your going?
Edit: Sorry, re-read the above post & quote, and yes, the forms bit is rather weak, but it's a small part of the whole.

Again, I'm primarily referring to navigating through a site with main navigation and content area links.

Quote:

What we really need, though is stats to indicate:
--are people confused by self-referential links on a nav bar?
--are people confused by self-referential links in the main content?


Do we? If we assume for a second that there is some evidence, even anecdotal, that people are sometimes confused by self referential links. (Interestingly, Scott Berkun has chipped in, the guy behind IE 1-5's interface who switched to Firefox).

The logical case is there. There is also some evidence that it makes a difference to people's experience of a site.

The stats are then useful to know how much effort it is worth putting in. Is it worthwhile in a project to spend X amount of time solving the issue.


Quote:

--what proportion of people find a disabled item in a nav bar confusing?
--what proportion of people find a nav bar with items missing confusing?


The solution I'd propose (and have tested) is the (old) IBM one.

The item is highlighted and disabled in the main navigation, and shouldn't be a link in the content area. Like nomensa or ukwindsurfing. What proportion find that confusing? Not many, but that isn't actually the point, it isn't a numbers game. When people have moused-over or tried clicking it, it doesn't work. What happens then? They explore the page more fully or try another link, depending on whether they think they are on the right page. A couple of times I've even heard "oh, that must be the current page", but it's rare for people to think about it that much Wink

The point being that even if it is confusing, it quickly leads to a better outcome.

I hadn't considered removing items from the navigation, I'd assume that would be confusing, but I don't remember seeing it done.
Reply with quote
Alastc wrote:
I'm not sure where you're going with that, if you half fill a form and click on any link, you're likely to loose your data. If you wanted to prevent that, you'd disable all links, but I'm sure that's not where your going?

Correct. That's where your web blooper guy seemed to be going. I was pointing out there's a flaw in his argument as he said it was a reason against self-referential links. It's actually an argument against any link on a form page.

Alastc wrote:

Quote:

What we really need, though is stats to indicate:
--are people confused by self-referential links on a nav bar?
--are people confused by self-referential links in the main content?

Do we? If we assume for a second that there is some evidence, even anecdotal, that people are sometimes confused by self referential links. The logical case is there. There is also some evidence that it makes a difference to people's experience of a site.

Yes we do. The way web users browse evolves. More and more sites uses some form of nav bar. I think the number of users who would be confused by a self-referential link on a nav bar is much less than those who would be confused by a self-referential link half way through the content of the page. My opinion (and I'd like to see stats to see if I'm right) is that the second type of self-referential link will confuse many more than the first, and that the first may not confuse many.

Alastc wrote:

The item is highlighted and disabled in the main navigation, and shouldn't be a link in the content area.

As I say, that's probably my preferred option. I just don't have any stats to back it up.
Reply with quote [quote=JackP]I think the number of users who would be confused by a self-referential link on a nav bar is much less than those who would be confused by a self-referential link half way through the content of the page. My opinion (and I'd like to see stats to see if I'm right) is that the second type of self-referential link will confuse many more than the first, and that the first may not confuse many.[/quote]

Ah, right, we basically agree! Smile

I think you are very probably right about self-referential content links being more confusing (although much less common? I've not noticed them in the site's I've tested). I'd also include 'related links' to that category as well, as in when you have a right-hand column of related or news type links. (Which has come up, and have cause problems.)

I guess for me the stats would help define how much effort would be justified.

I.e. when there is an easy way of adjusting the navigation appropriately, why not just do it?

There are easy ways of 'half' doing it, e.g:
http://www.themaninblue.com/writing/perspective/2004/11/19/
(Good for visual browsers, doesn't work elsewhere).

And depending on your set up for navigation, it usually easy to set up some kind of logic for "if this link is for the current page, use a <strong> tag instead of an <a> tag".

I'm interested to know where you've seen self-referential content links? I'd agree it would be an issue, I'm just struggling to remember where I'd have seen one, and also wonder what would cause people to make them?
Reply with quote
Alastc wrote:
What proportion find that confusing? Not many, but that isn't actually the point, it isn't a numbers game. When people have moused-over or tried clicking it, it doesn't work. What happens then? They explore the page more fully or try another link, depending on whether they think they are on the right page. A couple of times I've even heard "oh, that must be the current page", but it's rare for people to think about it that much Wink


What I don't get is why we're meant to believe that people have the ability to realise 'oh, that must be the current page' because a link is disabled/absent (though some users will doubtless consider it to be 'broken'), but can't make the same, simple conclusion when the same page reloads.
If 'user X' has some form of mental deficiency (as had the original hypothetical user), then how can we really trust them to make the mental link on one count (omitted/disabled s-r links), but we're seemingly not supposed to trust them to make the same mental link on the other count (all active menus)?
If we're supposed to avoid s-r inks because users are being confused when 'nothing changes/happens', then how is that much different from disabled s-r links, which will presumably appear equally ineffective when clicked upon.
At least with active s-r menu links the user is given what they appear to want.

We've also heard how users that encounter a disabled s-r link move onto trying other links, but I'm not convinced that such behaviour is really a good outcome in terms of usability. It might be good for site owners in that it ushers users further into and around the site, but that's presumably not what the user wanted when they clicked on the [disabled] 'link'.


Fwiw, it is possible that they have their own genuine reasons for knowingly clicking on an s-r menu link. What with the use of dynamic content and randomising scripts to present certain elements [random selection of suggested buys or simply random graphic headers] it's also possible that a user might wish to reload the current page.
Fwiw, I have witnessed numerous times people using s-r links to reload Ebay item pages as they watch auctioned items approach their close. Yes, they can, of course, do this using the reload button, but the fact that some people need to be told what that [reload] button is and what it does shows that some aren't aware of its existance or its use, but have been able to work out that clicking on an s-r link will do the job.

<aside>
I wonder if those against s-r menus are also against the presence of the reload/refresh browser button. It can easily be argued that it performs an equally useless task as an s-r button.
</aside>


This may only lead to the conclusion that there really is no single way that is right for every user or every site. Given the range of abilities, no single usability method can really be said to be universally or 'obviously' right. The best we can hope to do is cater as best we can for the most likely behaviour of our [target/known] audiences whilst keeping the site and its content accessible to as large a group as is reasonable.


(I'm not sure that I see a use for s-r links within body text. Whilst I clearly have no objection to them in principle, using s-r links in body text simply seem like the product of bad copywriting - i.e. referring to the current page in the third person, so to speak.
Reply with quote Hmm, I don't think I've been clear.

Quote:
we're meant to believe that people have the ability to realise 'oh, that must be the current page' because a link is disabled/absent ... but can't make the same, simple conclusion when the same page reloads.


I said a very few people realise that, most don't think about it. The problem is most severe when people don't realise it's reloaded.

Quote:
If 'user X' has some form of mental deficiency


(This week) I was mostly referring to visual impairments.

Quote:
If we're supposed to avoid s-r inks because users are being confused when 'nothing changes/happens',


The problem is that it appears that something does happen, and it takes a while to realise that it's the same page. I.e. people click a link and expect a new page, the page reloads. Then work out it's the same page, especially when you can only see 1/8th of it at a time.

Quote:
Fwiw, it is possible that they have their own genuine reasons for knowingly clicking on an s-r menu link. What with the use of dynamic content and randomising scripts to present certain elements [random selection of suggested buys or simply random graphic headers] it's also possible that a user might wish to reload the current page.


Whilst true, this represents far fewer occasions overall, and would probably be best served by some kind of 'reload page' button/link?

Quote:
We've also heard how users that encounter a disabled s-r link move onto trying other links, but I'm not convinced that such behaviour is really a good outcome in terms of usability. It might be good for site owners in that it ushers users further into and around the site, but that's presumably not what the user wanted when they clicked on the [disabled] 'link'.


They either try another link, or re-look at the current page. That decision is usually based on how sure they are that they are on the right page, so if there's nothing else in the links that looks more likely, then the current page is re-evaluated.

Quote:
I wonder if those against s-r menus are also against the presence of the reload/refresh browser button. It can easily be argued that it performs an equally useless task as an s-r button.


Except that it's a (useful) browser function that works consistently across sites. The s-r link doesn't do what most links do - go to another page.

Quote:
no single usability method can really be said to be universally or 'obviously' right.


Always the case when dealing with people Smile
However, I'd cater more for those that aren't technical or have restricted access (e.g. screen magnifiers) than the techies of the world. (Unless it's a techie site of course.)

Display posts from previous:   

Page 1 of 3

Goto page 1, 2, 3  Next

All times are GMT

  • Reply to topic
  • Post new topic