Tuesday, April 17, 2012

Why I type in Dvorak and you should too

The Dvorak keyboard is an alternative to the traditional Qwerty layout. Proponents (like me) claim that it is faster and easier to use.  Dvorak himself claimed in a 1943 National Business Education Quarterly paper "There is a better typewriter keyboard" that experts could type 35% faster in the Dvorak layout than in the Qwerty layout.  (value cited in this paper, I could not locate original)

I started using Dvorak during my freshman year of college because some guy told me it was cool. I converted my computer's keyboard format to Dvorak and re-arranged all the keys of my 1st generation iMac.

I feel old.

I was not much of a 'typer' before attempting Dvorak. I was a step above 'hunt and peck' (I used multiple fingers), but I couldn't type without looking at the keyboard. It wasn't long before I became much faster typing in Dvorak than in Qwerty, and could touch-type for the first time in my life.

I now change all computers I use to Dvorak, but do not change the physical keys on the keyboard. This has resulted in some lovely events such as my work-study boss in college thinking her computer was 'haunted' because I forgot to change the format back before leaving the office. It has also resulted in some embarrassing moments for me when I am forced to return to a Qwerty layout. During a presentation on some new neuro-software, I volunteered to test it out. This was a bad idea, because of course the presenter's computer was set to Qwerty. I not only typed super-slowly, but I couldn't put in a familiar password at one point. I knew the password by touch, and without the letters showing up as feedback, I literally could not type it correctly. 

Despite the occasional problem, I love typing in Dvorak. I find it much easier and more natural than typing in Qwerty. However, since I have been typing in Dvorak since iMacs were cool, my favoritism is probably due to familiarty more than some inherent 'betterness'. I can hardly be objective here.

For some real objective analysis we need some peer-reviewed studies. Luckily the Human Factors and Ergonomics Society cares about this sort of thing.

In a 2009 paper Anderson et al. investigated just how steep the learning curve was for a variety of alternative keyboards. 

Anderson et al., 2009
Figure 1:  chord, contoured split Qwerty, Dvorak, and angle split Qwerty
In this study, participants typed a familiar passage (having practiced it 10 times with the normal Qwerty keyboard) 5 times on an 'alternative' keyboard.  The researchers then plotted the time it took to type the passage. 

Anderson et al., 2009 Figure 3
The split keyboards are Qwerty layout keyboards, just angled differently for ergonomic purposes, so it is not too surprising that they resulted in fast typing times.  The Dvorak and chord keyboards were more difficult for the participants, but both showed strong learning curves. 

This study says nothing about how 'experts' type on any of these keyboards, so I decided to test myself. 

Online, you can test your typing speed by typing in random words or passages for 1 minute.
I tried these tests 3 times each in Dvorak and Qwerty (alternating).  Not surprisingly, I was much better in Dvorak.
open symbols= random words test, filled symbols=passages test

The random words test is much easier than the passages test which includes punctuation, but in both tests I was faster in Dvorak. 

But of course I don't type in Qwerty regularly, so this isn't exactly the right comparison.  To rectify this, I got help from a Qwerty user who was so kind as to try the passages test 3 times for me. My Dvorak passages test were slightly better than the Qwerty-user's passages tests (filled red circles compared to blue squares). One person per group is hardly proof and couldn't even count for preliminary data, so don't quote this figure as proof that Dvorak is faster or anything. It could just as easily be proof that people with brown eyes (me) are better typers that people with blue eyes (Qwerty-user). This was just some good old fashioned dorky fun-with-data. 

If you want to add data points to my table, go ahead and take the typing tests yourself:

Random words

Both sites are annoyingly stuffed with ads, but you can take the test without clicking on any of them.

Then let me know if you are Dvorak or Qwerty user, what test you took, and how many words per minute you typed.

© TheCellularScale

ResearchBlogging.orgAnderson AM, Mirka GA, Joines SM, & Kaber DB (2009). Analysis of alternative keyboards using learning curves. Human factors, 51 (1), 35-45 PMID: 19634307


  1. I'm a Qwerty user. I got 113 wpm on random words, 105 wpm on passages. Not sure this data is really helping us, though!

    At this point, I'd consider switching to Dvorak for the sheer pleasure of learning something new, but I'm too dependent on a host of keyboard shortcuts (for Vim, especially) laid out with the Qwerty keyboard in mind. I also spend 0.0% of my day (I'm a programmer) feeling limited by my typing speed. Show me the programmer who can think good code faster than he can type, and I'll eat my hat.

    1. Not sure if he can....but you might want to ask the Woz as he is a Dvorak user.

  2. I'm a QWERTY: 59 wpm with 2 errors on random words and 54 wpm on passages with no errors.

    I would change just to help my brain grow.....

  3. Great! I will add these data points to my graph at some time. (Data, however unscientifically collected, can be fun to play with) Also, 113 wpm? wow.

    I realize that I never actually got to the 'why you should too' part of my post. (I was side tracked by data-games) The real reason that I think people should try Dvorak is exactly the reasons these comments point out: To learn something new. Typing in Dvorak is like an advanced version of 'brushing your teeth with the opposite hand' which is a commonly purported method to 'exercise' your brain.

  4. To play the devil's advocate...
    Here's the classic rebuttal to the Dvorak argument:

  5. Qwerty, random words: 101 wpm (correct) / 106 total
    Qwerty, passage: 123 wpm (correct) / 124 total

  6. In Germany we also got an unknown alternative to our commonly used Qwertz keyboard layout: NEO.

    Unfortunately, the english description in this Wikipedia article is missing a really huge point with the NEO keyboard layout: it was purely statistically derivated from the most commonly used words in the german language (first version came out 2004). The most used letters are arranged on the main line and the next most used letters are as near as they can be to the main line. With the NEO layout, your fingers have to pass a much shorter way than with the regular Qwertz layout to write the same german text.

    I guess something similar was made with DVORAK but before high end computers were there to calculate the optimal place for eacht letter.
    It seems that COLEMAK is a big contender to be number one for the english keyboard layouts,