USiT blog

Clicktale (an experience analytics company) published a blog post back in December 2006 smashing the “people don’t scroll myth” to bits. However, 3 ½ years later clients and colleagues still make this claim fairly regularly. So I figured this study is a good one to keep tucked away ready to pull out for myth busting as required. Users don't scroll? In a month long study analysing over 120,000 web pages Clicktale found:

• 91% of the page-views had a scroll-bar.
• 76% of the page-views with a scroll-bar, were scrolled to some extent.
• 22% of the page-views with a scroll-bar, were scrolled all the way to the bottom.

“While 22% may seem low at first, it is actually quite high as many page-views are repeat views where the visitors have previously scrolled all the way to the page bottom and are already familiar with the page.”

Ok, well how long was the page? Further more, when digging deeper, Clicktale did not find that longer pages performed worse than shorter ones. The graph below shows the percentage of users who scrolled more than 90% of the page relative to the page length. More excellent articles on myth busting can be found on http://uxmyths.com/

Can a well considered defiance of a convention help induce an emotional attachment to an object? I own a car that is a bit different to other cars that I have driven:

• The ignition is on the left side of the steering wheel
• It has a foot pedal for a park brake... not a hand brake.
• There's no lever to open the petrol cap. Instead, if the car is unlocked then the petrol cap is unlocked and I'm supposed to just push it to open it (when I bought the car I actually spent a good half hour looking for a lever)

For each of these I consider the opposite - or alternative - to be the convention and the setup in my car to be against my initial expectations. Here's the thing. After owning the car for a while I have discovered that now when I drive a different car - with the usual conventions - that I feel uncomfortable and out of place. My habits have formed strongly enough that now I always have the keys in the left hand, never use the petrol cap lever and always try to engage the park brake with my left leg... repeatedly, and without success in a car configured in a more standard manner. In other words, my experience of the conventional setup has been ruined by long term exposure to something different. I now have a new convention that I strongly associate with a single object. I think that it's worth exploring the idea that a well considered deviation from an interface standard can actually increase the attachment to an object. Here's another example - Gmail. I actually think Gmail is pretty "sucky" in many ways but for me it was the first webmail that wasn't unbearable - which was why I started using it. The no folders approach to gmail is frequently annoying (against convention) but it's actually impacted the way I use Outlook in that I now have no folders and try to use the search function in outlook to find everything (usually unsuccessfully). I now feel uneasy in non-Gmail because the search isn't any good. Now, obviously, this type of approach could only work for interfaces that a user will be engaged with frequently. Without frequent (and probably prolonged) association with an object the new habits will not be formed. In addition, the type of object and the circumstances under which it is used are probably a factor. As Stephen pointed out to me earlier this could be looked at in terms of people's tendency to "anthropomorphise object and tools"... I suspect that some types of object are more likely to have their quirks projected into personalities than others. Then again, I've heard of people developing attachment to a Roomba and to me that's "just a vacuum." As to whether the new way of doing things needs to be better for attachment to occur. I suspect it probably doesn't (Stockholm Syndrome might agree with me on that) but, of course, a positive attachment is better than a negative one. Does anyone else have a different take on this?

I'm not a fan of either WYSIWYG editors or visual GUI development environments such as visual basic. In both cases there is a danger of narcissism where the developer is overly concerned with the interface they interact with at the expense of the interface they are creating for somebody else. From my experience this usually means that the next person to interact with the code suffers - whether that is another programmer or a user. Perhaps for some people a visual development environment does not get in the way in the manner I describe. For my own part though, I still spend most of my development time switching only between a text editor and the actual UI that I'm creating.