Subscribe to our mailing list

* indicates required
Close

Wednesday, September 16, 2009

The single biggest usability quagmire in computing

It never ceases to amaze me that we're still, in some ways, hamstrung by the mechanical legacy of the Industrial Revolution in our day-to-day computing. The best example is probably the QWERTY keyboard, which impairs usability for millions of computer users daily.

As you probably know, the QWERTY layout, conceived by James Densmore and patented by Christopher Sholes in 1878, was specifically designed to make it difficult for people to type fast on early typewriters. In other words, it was purposely designed and implemented as a usability antipattern. Fast typing caused jamming of mechanical typewriter keys, which were (in Densmore's time) returned to their original "rest" position by weights, not springs. We continue to live with the QWERTY legacy-layout today even though it is well accepted that other keyboard layouts (for English, at any rate) are much more usable.

The best-known alternative layout for Latin-based alphabets is the Dvorak keyboard, which dates to the 1930s. The U.S. Navy did a study in World War Two that found that typing speed was 74 percent faster for Dvorak than for QWERTY, and accuracy better by 68 percent. Other studies (both by private industry and government) have tended to confirm this general result, although there's a considerable cult movement (given impetus by a 1990 article in the Journal of Law and Economics) claiming Dvorak usability to be nothing more than urban legend. (See further discussion here.)

The studies of Dvorak typing accuracy have produced some interesting results. It's instructive to compare the most-mistyped English words for QWERTY users versus Dvorak users.

The mere fact that your fingers travel dramatically less interkey distance when using Dvorak layout means less wrist, finger, and arm movement; thus Dvorak presents the potential for reduced risk of muscle fatigue and injury. This alone would seem to argue for more widespread adoption.

Interestingly, variants of Dvorak are available for Swedish (Sworak), Greek, and other languages. Also, there's a single-handed-typing version of Dvorak, to help with accessibility.

So, but. Let's assume for sake of argument that Dvorak is demonstrably better in some way (speed, accuracy, accessibility, risk of wrist injury) than QWERTY. Why are we still using QWERTY?

It seems an influential 1956 General Services Administration study by Earle Strong, involving ten experienced government typists, concluded that Dvorak retraining of QWERTY typists was cost-ineffective. This study apparently was instrumental in sinking Dvorak's prospects, not so much because people put stock in its results as because of the government's role as a market-mover. The practical fact of the matter is that the U.S. Government is one of the largest keyboard purchasers in the world, and if a large customer convinces manufacturers to settle on a particular device design, it becomes a de facto standard for the rest of the industry, whether that design is good or not. (Today that sort of reasoning is less compelling than in the 1960s, but it's still a factor in market dynamics.)

It turns out to be fairly easy to configure a Vista or Windows XP machine such that you can toggle between QWERTY and Dvorak with Alt-Shift, the way some people do with English and Russian layouts. Basically, to enable Dvorak, you just go to Control Panel, open the Regional and Language Options app, choose the Keyboards and Languages tab, then click the Change Keyboards button, and in the Text Services dialog, click the Add button. When you finally get to the Add Input Language dialog (see below), you can go to your language and locale, flip open the picker, and see if Dvorak is one of the listed options. In U.S. English, it is. (Click the screen shot to enlarge it.)



If you have tried the Dvorak layout yourself, I'd be interested in hearing about your experiences, so please leave a comment.

In the meantime, I hope to give the Dvorak layout a try myself this weekend, to see how it feels. In all honesty, I doubt I'll stay with it long enough to get back up to my QWERTY typing speed. But then again, if it improves my accuracy, I'll have to consider staying with it a while, becuase frankly my accuracy these dsya sucks.

14 comments:

  1. Reminds me for some reason of the noise they are trying to put into electric cars now because cyclists and vision impaired people have come to rely on the noise of old oil-guzzling combustion engines.

    ReplyDelete
  2. The Sholes/Densmore keyboard was NOT designed to slow down typists, it was designed to make keystrokes come from ALTERNATING SIDES OF THE KEYBOARD.

    It was the alternation that prevented the hammers jamming, not the speed. It also caused an accidental *improvement* in typing speed because as one hand was hitting a key, the other hand was targeting the next letter.

    That's why Dvorak's productivity gain is a myth: the real boost in speed came from alternating hand movements, and further tweaks produced improvements that were too marginal to measure. People who switched to Dvorak and found they typed better were--in reality--improving their typing speed because they were paying attention to their typing, not because the layout was superior.

    More here: http://disenchanted.com/dis/technology/usability.html

    ReplyDelete
  3. I recently talked to Dvorak user and he wasn't persuaded of speed benefits either. I'm still open to idea that Dvorak might be better, but I am fairly certain that it has been proven so far that difference isn't nearly wide enough to warrant your title.

    ReplyDelete
  4. why would a developer, who uses ctrl, alt, shift, symbols, function keys, etc, all the time, want to use a Dvorak keyboard? writing code (and everything around it) is not like writing letter, at all

    ReplyDelete
  5. Anonymous12:58 PM

    I tried dvorak, not for speed but less hand movement (less strain on my fingers). I then ran across "colemak". You only had to relearn 17 keys from qwerty (dvorak is 33) and it put less stress on the outer fingers (pinky). I found this to be true; dvorak had me typing commonly used letters (and letter combinations!) on the right pinky. On Colemak, it was like hardly moving your fingers.

    ReplyDelete
  6. Dvorak does not improve speed that much, but it definitely improves comfort. I switched 7 years ago and don't look back.

    ReplyDelete
  7. I have been using Dvorak for about 15 years, and it is my standard configuration when using my own computer. I'm using a Dvorak keyboard to write this comment.

    @Chris: Dvorak promotes using alternating sides of the keyboard much more than qwerty. All of the vowels are in the home row of the left hand, and all of the most common consonants are in the right hand. Most words in English stagger consonants and vowels, so Dvorak is much faster because of this.

    In Dvorak, 80% of the letters based on frequency are in the home row. Therefore, the fingers move a lot less during the course of constructing words and sentences.

    For me, Dvorak really feels a lot better. The speed gain over qwerty is not astronomical. But it is certainly not a myth. It feels like using a sharp knife to cut better instead of a dull knife.

    @BlackTigerX: If you are a good coder, most of your variable names are based on English words. (I am a fulltime Flex programmer.) So the speed and efficiency does carry over. Even alt-combinations usually serve as the first or most common letter of the word, such as (F)ile, (E)dit, Hi(s)tory, (B)ookmarks (to use firefox as an example), and therefore they are letters that tend to be close to the home row.

    ReplyDelete
  8. I've been a happy Dvorak typist for about 6 years. Without stepping into the flamewar over THE ONE TRUE HISTORY OF KEYBOARD LAYOUTS (made that mistake before...), I'll lay out the takeaways from my experience.

    I was a fast typist on QWERTY, and I'm a fast typist now. Speed gains were marginal.

    My hands and wrists hurt less than they used to. One major Dvorak benefit that gets overlooked (because it's not obvious to the casual observer) is reduced finger motion, due to the key-frequency-based layout that Harry B. Garland mentions above.

    The keyboard itself can make a huge difference, regardless of layout. I use a TypeMatrix 2030, and the low-force keys and straight-column layout help me with my RSI pain a lot.

    ReplyDelete
  9. @Harry: This blog post perpetuates a myth, and beside it, the hypothesis that Dvorak's layout improves typing speed is damaged by both subjectivity and the superior hypothesis that the switchers gained their performance from the fact that they were practicing their typing skills, not because of the layout's own qualities.

    In theory there is another layout that promotes alternating hand movements even more than Dvorak, but it will yield nothing in real life. The limit isn't physical but mental.

    Dvorkan and Qwerty are Chocolate and Vanilla. Neither are better than the other. Qwerty will, however, have the advantage that a Dvorak typist who moves from one computer to another won't be reduced to a hunt-and-peck typist because his muscle-memory is fixed on an arbitrary, subjectively "improved" layout.

    This blog post is tragically ill informed and its premise is disingenuous. My problem isn't with the supposed benefits of Dvorak, but because the OP is badly informed.

    ReplyDelete
  10. if you use Dvorak for comfort I'd be ok with that, but if it is because of speed, I've said it before and I'll say it again "Dvorak keyboard layout is for those who can't type fast in QWERTY"

    ReplyDelete
  11. I use Dvorak.

    I switched because it sounded cool and my hands hurt from typing so much.

    I type more now, and I have no pain.

    I don't see how it can get any simpler than that. Watch a Dvorak typist work. My hands barely ever move off the home row. I can still type okay on QWERTY too. It takes about 30 minutes for me to get my body switched over, but then I'm reasonable. I've been doing this for about 10-12 years now.

    The comedy from watching people try and type on my QWERTY-remapped keyboard is, in my opinion, almost worth switching in itself.

    ReplyDelete
  12. To the suggestion that going back to QWERTY on someone else's keyboard after learning Dvorak is a pain: I'm not sure how the fact that some people - in this case a great majority - do things in an old, broken was is an argument against adopting a better way.

    In any case, I'm sure it's a problem for some people, but I spent enough time touch-typing QWERTY that it takes me just a few minutes to get back into it when the rare need arises.

    ReplyDelete
  13. And why is everyone so obsessed with putting a QWERTY keyboard on a phone? How fast can you type with two fingers (or two thumbs)? There are interesting alternatives, like Fitaly, which are optimized for the one-finger typing.

    ReplyDelete
  14. Anonymous2:11 PM

    I've been using Dvorak for 5 years and am very happy with the switch. My typing speed improved from 40wpm to about 90wpm, but mostly because I didn't touch type correctly before. My girlfriend types equally fast on both (80wpm) but says that dvorak is a lot more comfortable and puts less strain on her wrists. You can really _see_ the difference between a dvorak typist and a QWERTY typist. On dvorak all your fingers are planted gently on the home row and once in a while fingers reach out, while QWERTY keyboardists jump around a lot lot more. This is especially evident when you're trying to type while you're hammered. Dvorak is indispensable because of how hard it keeps your fingers stuck to the "home row". Dvorak also works well with programming. I'm a professional C and ASM programmer and find that the symbol keys are placed just fine. The only problem I can see is the Cut+Paste shortcuts which require two hands (probably annoying for Windows users). Perhaps long-time users of Emacs or other programs also might find it annoying to learn new key combinations.

    One word of warning before switching though. I switched and used a program called gtypist with 20 excercises to help you learn the new layout. It took me about 2 weeks to get to my original speed, and about 2 more weeks to double it. However, I've helped convert at least 10 of my friends (all of whom are glad they switched). What I've noticed though, is that my friends' typing speed and accuracy are proportional to how many gtypist lessons they finished. If they did all 20 exercises, they type extremely fast and accurately. Conversely, if they were too eager and stopped after 12 exercises, speed and accuracy were worse.

    Lastly. If you switch, don't switch back and forth between the layouts. It will take you a LOT longer to learn. It's extremely frustrating in the beginning, but will be well worth it!

    ReplyDelete

Add a comment.