Bias

The topic of alternative keyboard layouts inevitably comes up from time to time on programmers' forums: is Dvorak really better than Qwerty?  Is it objectively proven?  If nobody has proven it, why would anyone switch?  Also if it's unproven, why aren't there any satisfactory studies?

It is this latter point that fascinates me.  We have studies by August Dvorak, inventor of the Dvorak keyboard layout, which purportedly show that it's faster than Qwerty.  We also have studies by Strong, commissioned by the GSA in 1956, which claim to show the opposite; supporters of the Dvorak keyboard claim that Strong was biased in favor of Qwerty, and most accept that Dvorak may have been biased in favor of his own creation.

How does one design and execute an unbiased study in a world where only the biased even care about the outcome?  If such a study were performed, would the 'losing' side accept the results?  Or would the study get ignored (or contested) for all time?  If an alternative keyboard were, in fact, proven to be better than the existing choice, how many people would actually switch?

Other things to consider besides raw metrics

Right now, Qwerty enjoys an enormous network effect.  It's easiest to be proficient on only one keyboard layout at a given time, and you can effortlessly type on everyone else's computer that has that particular layout set.  (Qwerty on computers enjoys the additional advantage that if you don't touch-type, you can still type in the letters you want, because they match the letters printed on the keycaps.)  The vast majority of "other keyboards" are Qwerty: not only are most people not typing at the computer frequently enough to switch, but the majority of the ones who are were trained in Qwerty, and continue to use it.

Also, the majority of other people can only handle Qwerty keyboards.  If I loan my computer to someone, I have to remember to set it up as Qwerty for them, or they'll be really annoyed with me.

To back up these wild assertions, I can think of only three other people IRL who actually use Dvorak, and one who tried it until it killed his Qwerty skill, which he needed to repair everyone else's computers.  I would consider them all techies, although one of them is more aligned in art and cooking than machines.  This is why I think it's safe to say that a minority-of-a-minority use Dvorak.  (And it's the popular alternative: I'm the only one I know IRL who has even heard of Maltron, QGMLWB, or Colemak.  If you don't count me telling Eric about Colemak.)

Another interesting thing a modern study of Dvorak could consider would be coding: in Dvorak, compared to Qwerty or Colemak, the punctuation is rearranged, and programming text relies a lot more heavily on it than English prose.  I find Dvorak's choices a lot more suitable for typing arrows in PHP and Perl, since they alternate hands, and don't require a same-hand jump from the number row to the bottom row.  Although that means other frequently-used punctuation like brackets and braces move up into the number row, they don't seem to run into the same slowdowns as the arrows did.

Finally, the actual keyboard in use may affect the results.  My wife has difficulty typing on my MS Natural 4000.  (I had difficulty typing on it with the Ergonomist Approved Reverse Slope board installed, which also made my wrists hurt worse than they ever had before.  Quite possibly, I needed a drafting chair to keep it from pushing my wrists into the sky.)  Any modern test of keyboarding prowess is going to have to let people use their own keyboards, or give them time to train exclusively on the standardized keyboard hardware used in the test.

The market argument

Another reason Dvorak vs. Qwerty seems to come up a lot is that it is the primary example of a technology that is considered technically superior, that has failed in the marketplace.  It's the quintessential example of the first-mover advantage leading to the entrenchment of Qwerty, and of the fact that technical superiority doesn't matter.

Unfortunately, neither of these theories are laws of the universe.  There are plenty of examples of the first mover failing.  Just ask Sega.  Or perhaps Apple: the company that brought WIMP environments to the masses, only to be out-executed by the Amiga for a while, and then the PC.  (Apple got the last laugh vs. the Amiga, though: they successfully switched to PowerPC, to OS X, and once again to Intel.  They also are managing/have managed the 64-bit transition quite smoothly, at least from my vantage point outside the system.)  For all their prowess, though, the only Apple product leading the market years after introduction is the iPod—it's too soon to tell for the iPhone, although their marketshare is uncertain in the face of Android.  The iPod was in the unique position of being both desirable and DRM-locked, so again we see other factors at play besides "first" or "best".

A prediction

I doubt that any sound study of Dvorak is going to show such incredible gains in speed as Dvorak's studies did, but on the other hand, I don't think that they're going to show incredible losses, either.  Such a study would also need to take self-reported comfort into account, in case Dvorak typists are self-limiting in speed in order to self-limit pain.  Life is rarely perfect and simple.