Recently, while I was in the market for a custom keyboard, a friend of mine suggested that I choose a keyboard with the Dvorak layout. I had never heard of Dvorak before, so I decided to do a little research. I discovered a whole mess of articles, many of which indicated that Dvorak was the superior layout, while others claimed that those articles were biased or based on anecdotal evidence. I decided to gather my own data to compare Dvorak, Qwerty, and Colemak (a little-known third layout option).
Before I get to the study, there is an important distinction between the quality of a layout and the cost of switching layouts. The my.opera.com user Vorlath said it best: “Cost of switching is different than the benefit of Dvorak. If it were a straight up determination of which one is best, you’d take two untrained clones and have each one trained in a different keyboard layout. But switching means that they already know Qwerty. That’s a whole different ballgame and while switching costs should be considered, it should not be used to determine if Dvorak is better or not.” I wanted to determine which was actually better, and not take Cost of Switching into account just yet. The easiest way to do that was to listen to the numbers, and I planned to get those numbers my damn self.
To begin with, let’s take a look at some basic things to look for in an efficient keyboard layout:
Things to look for:
- Key placement by row: Keys in the home row are the easiest and quickest keys to press. The top row is the second easiest row to hit while the bottom is the most strenuous to reach. We’ll need to look at which rows are used most often in each layout.
- Key placement by finger strength: Next, we know that the middle and index fingers are stronger than the ring finger and pinkie, so the keys toward the center of the keyboard are easier to hit. That means we’ll also need to look at which fingers are used most often in each layout.
- Key placement by alternating hands: Next, we also know that letter combinations that are pressed with the same finger aren’t as fast as letter combinations pressed with different fingers. Therefore you can achieve better speed and rhythm when common letter combinations let you alternate hands. We’ll take a look at how often same-hand and same-finger letter combinations appear in each layout.
- Finger movement: Lastly, Dvorak advocates often claim that a typist’s fingers move less with the Dvorak layout. To test this claim, we’ll compare total finger movement for these three layouts(more on how this is calculated later).
You may notice that I didn’t mention anything about words per minute. I believe that all of these factors directly contribute to higher typing speeds, less strain on your hands, and overall effectiveness of the layout. These factors are also much easier to measure because they are not dependent on a single person’s typing ability in any given layout.
To do this comparison, I used texts and articles written by many different people, including myself, as I didn’t want the comparison to rely on my writing habits alone. Here’s what I found:
Key Placement by Row:
I see a problem with the amount of usage that Qwerty’s home row is getting, but at least the rest is weighted more toward the top row than the bottom row. Dvorak and Colemak seem to have their row prioritizing right, with relatively small variation between the two.
Key Placement by Finger:
|Fingers||Left Pinkie||Left Ring||Left Middle||Left Index||Right Index||Right Middle||Right Ring||Right Pinkie|
*Percentages may not total 100 due to rounding.
Both Dvorak and Colemak favor the right hand, albeit slightly, while Qwerty favors the left. All three layouts seem to favor the stronger fingers, save for Qwerty’s Right Middle finger, which is surprisingly low.
Here’s a visual representation of the keys that get the most use and the overall layout for each keyboard.
*Note that common hot-keys (Ctrl+C, Ctrl+V, etc.) are largely unchanged with Colemak.
Same finger/same hand letter combinations:
This is a percentage of key-presses that are done with the same finger/hand as the letter before it. Same letters don’t count. i.e. The word “cool” (on a Qwerty keyboard) counts as 1, the two “O”s don’t count, but the “L” does as it is typed with the same finger as the “O.” Same finger presses are totaled and divided by total keystrokes. Same hand letter combinations are calculated in the same way.
Same finger letter combinations:
Same hand letter combinations:
Okay, for same finger combinations, Dvorak is less than half Qwerty, and Colemak is almost half of Dvorak. Pretty fascinating stuff.
For same hand combinations, however, Dvorak takes the lead while Colemak only shows a marginal improvement over Qwerty. Although, it should be mentioned that hand alternation was not a goal in the design of Colemak as it was in Dvorak.
Lastly, we’re going to look at overall finger travel distance. This is calculated assuming standard 1.8cm keys, to include the gaps between keys. This measures the distance the finger moves from home row to the key and back to home row. If the key is pressed with the same finger as the last, it measures the distance between the two keys and then back to home row. The distance of depressing the key is not calculated here as it varies too much on different keyboards.
For all of the text that I measured, Qwerty had the highest distance, so for simplicity’s sake, I’m going to give these as a percentage of that distance.
Overall finger travel distance:
So we seem to have a clear winner here. The difference between Dvorak and Colemak is 8.2% as compared to Qwerty. To put it to another scale, Colemak’s distance is 85.5% of Dvorak’s, which makes it sound like a more impressive improvement.
Overall, however, it’s hard to make a determination between Dvorak and Colemak. In this writer’s opinion, Colemak has a subtle lead on Dvorak. At the same time, it seems clear that Qwerty has lost in every one of these comparisons.
One thing I mentioned at the beginning of this article is that I wanted to compare these layouts on even ground, as opposed to a “Cost of Switching v. Benefits of Switching” analysis. However, it is a very important factor in deciding if you want to make the transition. The biggest problem with this kind of analysis is that it ends up being based on mostly anecdotal evidence, but even that should be considered when deciding.
The cost is simple, and comes in a couple forms.
A new keyboard, if you need it. Some can just be rearranged.
Whether or not your Operating System supports it. Most support Dvorak and can be changed in as little as 30 seconds. However, few Operating Systems support Colemak be default, but that can be fixed with a small download from colemak.com. (Colemak is the least common of the three layouts, and is also relatively new.)
Retraining. This seems to be the bigger issue, so I’ll go into it in more detail.
A 1973 study based on six typists at Western Electric found that after 104 hours of training on Dvorak, typists were 2.6 percent faster than they had been on QWERTY. Similarly, a 1978 study at Oregon State University indicated that after 100 hours of training, typists were up to 97.6 percent of their old Qwerty speed.
I’m not sure exactly how much typing one can get done in 100 hours, but that’s apparently what it takes to match your old speed. But these were also Typists. For the average user things seem to go a little differently.
On average, it seems that it takes about a month to become comfortable with a new layout and type close to their old Qwerty speeds, and after about three months, users are able to type above their old Qwerty speed in the new layout and still use Qwerty with close to the same efficiency that they had before switching. I think that’s worth mentioning seeing that a lot of people’s objection to a new layout is having to use Qwerty when not at their home computer.
As for the benefits, the obvious one is faster typing speed, but I think that the reduction in finger movement is far more important and issue. I’ve heard tales (both in my research and personally) of pain in the hands after typing for extended periods of time. I constantly find myself cracking my knuckles and shaking my hands off after being at the computer more than a couple hours. I don’t type professionally, but this can be a big concern for those that do.
I’ve been told that switching layouts is along the same lines as learning to speak Esperanto. Qwerty doesn’t have much, if anything, going for it. Neither does English, but we already know it. It’s the de facto standard, so we kinda have to live with it. Lucky for us, switching keyboard layouts isn’t as difficult or time consuming as learning a new language.
So, is it worth it?
I would argue that it is. If not for all of these reasons, than for just one. Many testimonies by various forum users show that almost everyone who makes the switch stays with it. I’ve only found one, maybe two, accounts of someone switching to Dvorak or Colemak and then deciding to go back to Qwerty.
Dvorak and Colemak show to be clearly better than Qwerty, and I would suggest learning one of those two. For me, I’m most likely going to go with Colemak, but depending on your specific needs, Dvorak might be better for you. The decision between Dvorak and Colemak is one that only you can decide for yourself. All I can say with any level of certainty is that you should switch to something that was designed with ergonomics and efficiency in mind.