Looking Out To Sea

There’s a pretty famous usability study done by Tog et al about the apparent advantages of mouse over keyboard for quick execution of computer commands. It’s a famous study because it seems to have escalated out of all reasonable proportion. In the words of the man himself:

We’ve done a cool $50 million of R&D on the Apple Human Interface. We discovered, among other things, two pertinent facts:

• Test subjects consistently report that keyboarding is faster than mousing.
• The stopwatch consistently proves mousing is faster than keyboarding.

The gist of it is that there seems to be a ‘hidden’ delay when using the keyboard which doesn’t happen with the mouse, because the user takes some time to recall the keyboard function before it gets executed. But this memory access time is invisible in the short term… you only notice that you can’t remember a keyboard action when the the answer doesn’t come to mind for a good second or so.

The conclusions as stated are so bald and so strange that I’ve never met anyone that actually agrees with them as written. If keyboards are so slow, why do we not use the mouse in combination with an on-screen character map instead? “Well, we didn’t mean it like that…”

So you have to wonder, what did you mean it like? I don’t think anyone will reach 70 words per minute by clicking buttons with a mouse (though I’m willing to see studies). Of course, I don’t believe Tog was stupid enough to argue that either. John Gruber recently addressed this issue too, which made me re-examine what he must have been talking about. Gruber makes the valid point that keyboards are much faster when you’re doing repetitive actions — ctrl-C, ctrl-V, repeat — but I’d argue that’s not going far enough.

Muscle memory and so on

The key combination to place the letter ‘a’ on the page is quite simple (the keys even come pre-labelled, though you have to realise that pressing the key labelled ‘A’ won’t actually give you a capital A by default; but I digress). But even the simple act of typing takes a long time to learn — but eventually you learn where all the letters are and then it goes somewhere else in your brain. I haven’t got a great mental map of the keyboard but I can still type reasonably well without it, because my fingers know what to do.

Similarly, I know that ‘going’ on a mobile phone with predictive text is ‘left right left right left’ but I have to think about what those keys are. My fingers find it faster than I can think about. If you’re a frequent user of the text editor called vi then you’ve probably left emails and other documents littered with :wq. Because that’s the combination to save and quit in vi but not anywhere else. I know I want to save and exit so that’s what happens without me needing to think about it.

When I speak French — not very often, and not very well — some phrases and conjugations jump quickly to my lips without me thinking about it. If I then think about it, I’ll start to doubt myself and have to conjugate from first principles. And invariably discover I was right all along. This internalisation of action happens with almost everything you learn — grammars, keyboard shortcuts or violin playing. It just happens and it’s the reason we can become experts at things after much study. Tai chi, karate, ballet — they all have set actions and forms which train the body and the mind to moving in a particular way, so that action happens unconsciously.

The unconscious movement happens to me when I’m not editing in vi. (I type ‘escape : w q’ and then nothing happens.) This happens to me when I move from keyboard-friendly window managers like xmonad to more mouse-oriented ones like GNOME. (I type alt-2 to check my email and nothing happens.) It happens all the time and that’s when I notice that these actions have become second nature. At that point I have to think — hold on, how do I save in this program? or how do I move to the left workspace from here?

And that’s the point where recall is an issue. That’s when it seems noticeable, and it slows me down. Is this what Tog meant? I don’t know but it seems more feasible than the notion that all users take two seconds to remember all keyboard combinations. If I can do things without even knowing I do them then it doesn’t seem likely that I’ve had two seconds of down time to get to that point. That would be similar to suggesting that a black belt takes two seconds to remember to block when being attacked.

Keyboard and mouse evangelism

But all the argument in the world isn’t going to help because Tog’s got 50 million dollars of evidence to support his statements. And all my examples aren’t going to prove anything because, remember, users experience “real amnesia!” when performing keyboarding actions.

So the keyboard users will continue typing out their characters one at a time and the mouse users will continue to click on their on-screen character maps — no, I mean the keyboard users will continue to ctrl-S and the mousers will continue to File → Save.

And the real reason for this loss is that we can’t actually see what was studied or what the results were. We’ve got a few all-encompassing phrases which seem to generalise out of all proportion and no solid data. If only we had more evidence of what the studies showed then people would be more inclined to believe it. Saying we’ve done studies doesn’t cut it, guys.

:wq

Trackback URI | Comments RSS