You also talk about how no one would arrive at a home page without some context “everyone knows what UPS does”, but surely the same argument applies to your address change example – no one would click on the address change link without the context of wanting to change their address thus they’d arrive at the page with an expectation.

I don’t find the distinction clear enough. In fact I’m on the side of it is equally useful for home page (not that I’ve used this technique yet). Say you click a search result link through to a home page – you need to know quickly that the page you hit matches your expectations based on your search query. Same as any other page, if you clicked an internal link to change address you’d need to know quickly that was the right page. Home page is no different other than it might just possibly be a more common entry point to your site, which surely makes it MORE important.

I was recently amazed at the speed that i saw a viusualy impaired user interact with an interface using a screen reader.

Perhaps a 5 second type test would apply under the same constraints. I would imagine that visualy impaired users interact with an interface at a simliar speed. So, I would imagine that an ever so slighlty longer test of the same type could be appropriate.

Anyone else got thoughts on this?….Has anybody tried this?

For users with visual impairments and/or who are blind, would you suggest a longer period of time? If so, for how long?

Priority definitely on what the user thinks they can “do”.

In addition, often organisations will want to know whether users understand “intended business image” For example, does it convey the right image to the user and does the user know its the Home Page of that organisation.

It gives us a chance to get feedback on both quickly and its worked well.