Typing Errors: The standard typewriter keyboard is Exhibit A... Dvorak not so great either

http://www.reason.com/news/show/29944.html

by nickb • 18 years ago

26 32 18 years ago

ellyagg2• 18 years ago

Based on the distance your fingers have to move to type average words on qwerty and dvorak, it's really impossible that Dvorak wouldn't be loads better.It's sort of like claiming that someone wouldn't be faster at running 50 meters than 100 meters. Of course, where this analogy falls down is that you have to learn a new skill, which means there's an upfront cost and thus makes the whole issue interesting economically.

The article claims that the cost of learning the new skill could never be amortized and uses this evidence:

> In the first phase of the experiment, 10 government typists were retrained on the Dvorak keyboard. It took well over 25 days of four-hour-a-day training for these typists to catch up to their old QWERTY speeds. (Compare this to the Navy study's results.) When the typists had finally caught up to their old speeds, the second phase of the experiment began. The newly trained Dvorak typists continued training and a group of 10 QWERTY typists (matched in skill to the Dvorak typists) began a parallel program to improve their skills. In this second phase the Dvorak typists progressed less quickly with further Dvorak training than did QWERTY typists training on QWERTY keyboards. Thus Strong concluded that Dvorak training would never be able to amortize its costs. He recommended instead that the government provide further training in the QWERTY keyboard for QWERTY typists.

I really can't express how stupid this is. The logical implication the article would have you draw from this study is that qwerty actually may be MORE efficient. To believe this, you'd have to believe that the top end hypothetical speed of qwerty is higher than dvorak. This is literally physically impossible, as I've already pointed out. Moreover, the fastest typist ever used Dvorak. While that's not conclusive, it's suggestive. Folks trying to optimize a skill to the highest level are very good at cutting out all inefficiencies.

To understand the results of the study cited there, we need only look to exercise science. First of all, what most lay people don't realize is that a great deal of exercise results for activities like weight training is explained by the exercise's affect on your nervous system, not your muscles. Properly managing your nervous system response is at the heart of continued gains, and not managing it properly leads to overtraining and plateaus. All top flight Olympic athletic coaches spend much of their time thinking about this.

As a matter of fact, it is a well known and well documented phenomenon in the literature that if you train at close to maximal capacity on the same training regimen for longer than 3 to 4 weeks, you will stop making gains. You'll see big increases the first 3 weeks, and after that, nothing. This is a nervous system effect.

Armed with that knowledge, it becomes clear why the study's results look the way they do: The dvorak trainee were already in a state of CNS fatigue, whereas the qwerty trainees were starting out fresh.

Dvorak is clearly better than qwerty. Colemak is even more better. However, I don't think this situation is much of an indictment of the free market. I don't think the bottleneck in most people's productivity now (or ever) was top end typing speed and it would take quite a long time to recover the cost of the switch. Considering that this is THE MOST concrete example ever cited where you might be able to say that government meddling could induce a positive effect, I'm not inclined to agree that it's in general a good idea to have our economic lives centrally manipulated. There are far more examples of net economic losses from government interference.

pragmatic2• 18 years ago

The point seemed fairly clear. Retraining on a Dvorak held no real advantage. The fact is that almost all people learn on QWERTY so the necessary retraining isn't worth it.

Not to mention the millions/billions of keyboards in existence that would need to be scrapped.

tome• 18 years ago

Touch typists don't look at the keys anyway, so you don't need to scrap the keyboard. Windows keymap switcher, or the equivalent in your OS of choice allows you to quickly go back to QWERTY on the (very) rare occasions where what you need to do corresponds with the actual symbol on the keys.

Rabidmonkey1• 18 years ago

Hmm... Colemak looks interesting. I've always had a hard time learning to use Dvorak, especially since everywhere else you work it's going to be Qwerty by default. Colemak is also public domain, according to Wikipedia. I might have to give it a try.

http://en.wikipedia.org/wiki/Colemak#Colemak

cninja2• 18 years ago

I actually installed Colemak on my computer yesterday. I played with it for a couple hours before uninstalling it. Most of what I type is programing code, not prose. It dawned on my that most my typing delays come from all of the parens and brackets and other programming symbols. Since all these fancy keyboard layouts have been computationally optimized for prose language, they don't address the primary cause issue. I decided that if I'm going to take the time to learn a new keyboard layout, then it will need to be one that has been optimized for programming.

elviejo4• 18 years ago

Qwerty is by default, but on any computer you can simply activate Dvorak from the control panel in your computer.

The problem with Colemak is that you need to download and install software to use it. It isn't an standard yet.

But yeah there are much better options than Qwerty.

PS FYI: to activate dvorak on windows: http://www.microsoft.com/enable/training/windowsxp/keyboardl...

to activate dvorak on Mac http://www.chimoosoft.com/articles/dvorak.php

twopoint7182• 18 years ago

Agreed, the distance your fingers travel is shorter and so it would stand to reason that for any given person they could cover that distance faster.

A large part of this article seems to have a problem with retraining. That the retraining of someone to Dvorak will be inefficient compared to them just getting better at Qwerty; this may be the case. The analogy that comes to mind is that of someone who learns another language in adulthood. In almost all cases that speaker will have an accent in the new language for the rest of their life. But, when new languages are learned as a child this isn't the case. Perhaps if you just lean Dvorak as your "native" keyboard these inefficiencies don't show up and the cost/benefit changes.

One could do what is done with software upgrades. New students learn Dvorak in typing class and everyone else can stay with Qwerty. This would give a generation-long deprecation period where nobody has to re-learn anything if they don't want. Since the complaint is with having to "re-learn" to type, this would avoid that.

mechanical_fish• 18 years ago

Based on the distance your fingers have to move to type average words on qwerty and dvorak, it's really impossible that Dvorak wouldn't be loads better.

The first lesson that every experimental scientist is taught is: Never, ever write a sentence like this. It will only make the other scientists laugh more loudly when your theory explodes on impact with the actual data.

Oh, will you be teased!

I tried using the word "clearly" once in a lab report in junior year. Man, did I hear about that.

ellyagg2• 18 years ago

As Einstein and many others have shown, we can reason correctly about the world without physical experimentation, you smug twit. I'm not Einstein, but then keyboard layouts aren't general relativity.

pragmatic2• 18 years ago

This is not just about keyboard layouts. It's about the concept the poor products (VHS, Windows) win over superior products (Beta, Mac) because of a winner take all economy.

As you can guess the author shows this to be bunk science.

davidw• 18 years ago

What they are afraid of is the idea of the free market not always and in every case producing the best possible result, so they need to aggressively attempt to demonstrate that 1) either market failure just doesn't exist, as they attempt to show here, or 2) that in no case, ever, will government intervention improve the situation, which is the topic of other libertarian writing.

In other words, this is just another episode of "Libertarians vs Socialists":

http://news.ycombinator.com/item?id=419503

For a serious treatment of the economics of digital goods and networked economies, I think Varian and Shapiro are a better read:

http://www.squeezedbooks.com/book/show/7/information-rules-a...

(Edit: to be fair, it turns out the authors have a more scholarly treatment of the subject matter, which is a better read:

http://www.utdallas.edu/~liebowit/harvj/harvard.html

Still though, I think it's pretty clear that they're coming from a point of view, and "backfilling" to try and get there)

twopoint7182• 18 years ago

It would be interesting to construct an experiment to see what kinds optimal solution a free market can and can't find (though I have no idea how this would be done).

In search algorithms you have to worry about merely finding a local maxima rather than a global maximum. In algorithms such as simulated annealing you perturb your solution randomly at the outset, trying to "bump" your solution off of one of those local maxima.

What if the free market is prone to getting "stuck" in some locally optimal, globally sub-optimal solution? Depending on initial conditions (price, number of firms in the marketplace, demand, etc.) we could end up with great solutions, but in others with terrible results.

Hexstream3• 18 years ago

What bullshit. There's a common pattern in Dvorak VS Qwerty: Dvorak advocates sometimes rely on history and sometimes on purely technical grounds to prove their claims, while Qwerty advocates always only rely on history because there's simply absolutely no way to prove Qwerty approaches Dvorak when it comes to logic.

Here are some facts about Dvorak that are easy to verify and the utility of which is obvious:

- Dvorak has the most used keys on the home row. Compare: AOEUIDHTNS VS ASDFGHJKL;

- Dvorak optimizes for alternating hands, which obviously helps speed. Ever tried typing "street" or "states" in Qwerty, two common words with common letters that you have to type all with the left hand? On the Dvorak side the worst word is "joke" I think.

- Dvorak optimizes for typing a bit more with the right hand instead of the left hand, which makes sense because the right hand is usually the strongest.

- Dvorak optimizes for typing from the outside to the inside (try tapping your fingers in rapid succession on the desk from outside to inside, then inside to outside, and you'll see immediately which one is easier and faster).

- Dvorak optimizes for typing more with stronger fingers (index and friends) than weaker ones (pinkies and friends).

- Dvorak has all the vowels on the left hand on the home row (except Y which is conveniently on top of I), one beside another, which is great for learning. It also makes sense because you use consonants more than vowels, (see right hand argument) and since the most used consonants are on the right home row, it optimizes alternating hands.

No need to mention Qwerty makes no such optimisations and the result is correspondly atrocious.

See? No need to bullshit with biased reports about history.

lliiffee3• 18 years ago

You weaken an otherwise useful post by using words like "bullshit". The problem with "bullshit" isn't that you are cursing. It is a way to dismiss an article with out actually engaging in any of the arguments made by the article. At the end you do hint that you believe the article gives a biased report of history, but you give no specifics, and it comes many paragraphs after "what bullshit".

cchooper2• 18 years ago

I see nothing wrong with beginning the post "What bullshit". It's a concise statement of what he believes. And what's wrong with stating your belief up-front and then providing the argument afterwards?

Hexstream3• 18 years ago

I won't address the article's historical claims for the same reason I won't address the Bible's historical claims:

- Addressing those historical claims would be an implicit acknowledgement of their relevance, but I believe they aren't.

- It would take a massive effort of research on my part to verify the historical accuracy of the claims, and this effort would be pretty aggravating for me since as I stated the claims are not relevant whether true or false. Also, it's disproportionally easier to make a new (BS) historical claim than to definitely disprove it so it's a losing battle.

And let's see... First August Dvorak claimed his layout is better, then there was the Fable of the keys, which said it wasn't, then there was the Fable of the Fable which contradicted that... And I'm sure there's a Fable of the Fable of the Fable lurking somewhere. So if you believe this historical BS is important, help yourself.

elviejo4• 18 years ago

If we were designing a keyboard from scratch: What shall we do? How about we find the most popular letters and put those under the strongest fingers. then we favor hand to hand switch, since that is faster. etc. etc.

Seems to me that the logical conclusion of such a study, would be a Designed Keyboard (DK). Said GK has to be better than any alternative that came from a) Random arrangement b) being able to type 'typewriter' at sales demonstrations or c) designed to to avoid jams of moving parts. We would call this Not Designed Keyboard (NDK)

So we have DK and NDK. Almost by definition DK has to be better than NDK.

Now we all know that in order to do a good job we need to use the right tool for the job. If the job is typing which it is for 100,000 of people then the only logical conclusion is that we need to use the better keyboard in this case Designed Keyboard. (DK)

Next what if you had an organization with 1,000 secretaries that their whole work was to type documents day in and day out for the next 20 years and you could improve their typing in just 1% using a better keyboard. Would it be economically worth doing? Of course.

In conclusion switching to a DK is the smart choice, is the cheaper choice.

And the only thing absent from this article in "reason" magazine is precisely that; reasoning.

s_tec• 18 years ago

Our wonderful universe acts in surprising ways. Human intuition is often wrong, and the only way to know something for sure is to test it. The Dvorak keyboard's logical organization seems intuitively faster, but we don't know how the human brain is wired, so we don't really know the limiting factors on typing speed. Anything is possible, and the only way to know the truth is by experiment. That's the scientific method. According to this article, the valid experiments support QWERTY over Dvorak, or at least show little difference. I don't know if the article true without examining the experiments myself, but I wouldn't call the article "unreasonable" for valuing science over fallible human reason.

tokenadult• 18 years ago

Human intuition is often wrong, and the only way to know something for sure is to test it.

Hear. Hear.

http://norvig.com/experiment-design.html

Hexstream3• 18 years ago

Qwerty couldn't be much less scientific. You can spell "TYPEWRITER" with the letters from its top row, for fuck's sake.

Dvorak (the man), on the other hand, dedicated a good part of his life studying Typewriter Behavior (he wrote a book with the same name but it's out of print) by looking at actual people while they type, then formulated theories about what eases typing speed and accuracy, then conducted experiments to validate his models. Wait... Haven't I just described the scientific method?

elviejo4• 18 years ago

Agree on all accounts except on:

"I wouldn't call the article 'unreasonable' for valuing science over fallible human reason. "

Because I don't think this article values science. If it did, then it would simply state: "Here we have 2 keyboards which one is better?"

Defining better as either: a) Fastest to type in for a user with IDK a 2 year experience in it. b) Easiest to learn, from scratch. c) Easiest to retrain for a person that already dominates the other keyboard. d) Healthiest over a period of 10 years (RSI anyone).

Instead of doing an experiment or honestly revising the studies mentioned. This article spends most of its time, attacking the Dvorak history or "hoax" as the author calls it in the end.

So this study is neither truly scientific nor truly logical. Is simply the author trying to support his preconceived notions: "Dvorak is a hoax". And using "ad hominem" attacks or self selecting studies to support his conclusions.

as I mentioned before for real science on keyboards check: http://mkweb.bcgsc.ca/carpalx/

And thank very much for your comment, it is nice to see succinct well thought argument on the internet.

elviejo4• 18 years ago

For "real science" on keyboards go to:

http://mkweb.bcgsc.ca/carpalx/

From the site: "The carpalx project attempts to find the best keyboard layout to minimize typing effort for a given set of input documents. These documents may be English text, programming code, or whatever you find yourself typing for hours"

Take a look at the studies from those results.

silentbicycle• 18 years ago

Thanks for posting this! I was in the midst of writing something very similar.

tlb• 18 years ago

The Dvorak keyboard might have some slight advantage for English text, but it's noticeably worse for programming where you need ;=+-,./{}[]() and the numbers quite often.

There might be a advantage for a programmer's keyboard, basically QWERTY with common punctuation on a fifth row above the numbers. But my few brief experiments with unorthodox layouts caused no end of hassle when trying to use other people's machines, so I gave up.

Typing Errors: The standard typewriter keyboard is Exhibit A... Dvorak not so great either

Related Stories