Dvorak keyboard layout, anyone used it?

Splinter · Post by **Splinter** » Fri Jul 09, 2004 4:04 pm

I used to do Data Entry for a living, you needed at least 85wpm and some ridiculous number on the numeric keypad as well.

Worst job ever.

shathal · Post by **shathal** » Fri Jul 09, 2004 5:29 pm

At least that way you only get RSI.

I thought working at Mc D's is the worst (or any such chain).

Depends on perspective I guess.

Splinter · Post by **Splinter** » Fri Jul 09, 2004 5:54 pm

Alright, worst job I've ever hard. I never did the fast food thing

fanerman91 · Post by **fanerman91** » Sat Aug 07, 2004 12:12 pm

Just took Huck Finn at 3 minutes after practicing Dvorak on and off for a month.

Gross Speed 57 WPM
Errors 6 Words
Net Speed 55WPM
Accuracy 96%

Not bad. I still make a bunch of mistakes though.

shathal · Post by **shathal** » Sat Aug 07, 2004 1:51 pm

Huck Finn?

I am not assuming that this is the novel, so some test of some sort...?

fanerman91 · Post by **fanerman91** » Sat Aug 07, 2004 2:08 pm

Oops. I meant here.
http://www.typingtest.com/
as outlined on the first page.

shathal · Post by **shathal** » Sat Aug 07, 2004 2:43 pm

Interesting.

It gave me an accuracy of 0%.

Though I thought the whole point of Dvorak was to type faster, rather than on a "normal" keyboard sort of thing?

Obviously, I see that the flaw in my argument is that your WPM will vary to mine, and so on, ad infinitum. So, we'd only be able to compare if we check your Dvorak WPM versus your "normal" WPM, same with me, and so on.

How much time would be "enough" to get into Dvorak? Hard to say, since I've been touch-typing now for ... urm ... *thinks* ... probably something close to 20 years now, and that'd hardly fair to compare Dvorak to, even if I were to start training it today

.

Zyzzyx · Post by **Zyzzyx** » Sat Aug 07, 2004 3:49 pm

shathal wrote:How much time would be "enough" to get into Dvorak? Hard to say, since I've been touch-typing now for ... urm ... *thinks* ... probably something close to 20 years now, and that'd hardly fair to compare Dvorak to, even if I were to start training it today .

I started this thread when I swapped to the Dvorak layout, back in... <checks date>... January. Aside from using 'normal' keyboards around the office, I've been using 100% Dvorak layout, on the same Kinesis Contoured keyboard at work and home. I am 30 years old, and remember learning to touch type while in high school, on our IBM XT (overclocked to 8mhz!!); so its been 15 years of Qwerty typing for me.

I still don't think I've quite reclaimed the speed that I had using Qwerty. Part of that is that I'm still making more mistakes, and using backspace, than I was with 15 years on the Qwerty layout. I'm getting close though. The hard parts right now is typing random-lettered passwords instead of a normal word, that still slows me down to thinking about what key is where. I did a three minute test the other day and came out at 55wpm, so there has definitely been improvement from when I started with Dvorak. The biggest improvement though is the comfort level while typing. With the Dvorak layout I can type much longer without my fingers, hands, or forearms getting tired or sore. Might not quite have the speed back yet, but with the way the keys are distributed on the Dvorak layout, there is much more of a rhythm when typing, as letters are alternated between hands far more often than with Qwerty. And the most common letters are on the home row.

So even with your 20 years of Qwerty typing, I would recommend changing to the Dvorak layout. If you change though, go all the way. Don't use Dvorak for only an hour or two in the evening. I tried that way twice before, never lasted more than a couple weeks. Been six months for me of nearly 100% Dvorak usage, and I am pleased with where I'm at for speed and accuracy. Especially knowing that it will continue to get better for probably another six months. I also strongly recommend the Kinesis keyboard that I use. Its an oddball, but haven't found anything that is as comfy to use, even when I was still on Qwerty.

fanerman91 · Post by **fanerman91** » Fri Sep 17, 2004 5:11 pm

It's been over 2 months... and another update (for no reason at all)

I did 93 WPM in a 1 minute burst and 79 in 3 minutes. At typingtest.com again.

BTW... the first couple weeks were the hardest. I was practicing only a couple hours a day. But I was practicing drills (linked in this thread). The worst stretch was... around the middle of the 2nd week. After the enthusiasm of the first week wore off and I was typing too slow, it was hard to keep going. But by the end of the 2nd week, I was able to type at something close to conversational speed... maybe 20 wpm. After I got over that hump, it was easy to keep going. I never finished the drills btw. After I finished the top 2 rows, I more or less stopped with them.

I don't care too much about speed though. This was for the sake of less finger strain (and paranoia) I make a few more typos than I'd like. I might practice more drills later. But I'm pleased with the results right now.

mathias · Post by **mathias** » Sun Sep 19, 2004 9:38 am

I've just switched. Last night I stuck on some cut up envelope labels with the new layout. I'm not doing too bad, down from around 35 WPM to 12, after about 11 years of querty. Makes me glad I never took keyboarding class and never tried to learn to type the right way(my hands kind of flail around when typing, probably all the more reason to switch).

I figured I should do this cause I have it a lot easier than some people here, I don't have to use any PCs but my own. And I didn't yet have a job that involves typing.

Just took the test again, up to 15WPM now.

ascii · Post by **ascii** » Tue Sep 21, 2004 1:20 pm

Hmmm, not sure if people still read this thread or anything, but I've always been interested in the Dvorak/Qwerty myth.

The basic design goal of Qwerty is to make words with letters very close to each other as rare as possible, so that old typewriters would be less likely to jam. This design rewards typing with all fingers and alternating between hands, which is also desirable from a speed perspective. The Dvorak keyboard lets the fingers stay at the home row a lot, which is good for speed, but has a slightly worse spread between fingers, meaning that you wont be able to use all the fingers of your hands as effictively while typing as a Qwerty keyboard. In the end, none of these effects make a very big difference. Several tests regaring typing speed using different keyboard layouts have been done, and only one such test pointed to large differences between layout. This test was conducted by Professor August Dvorak. You may have heard of him. He once designed his own keyboard layout.

I have read a few articles touching the subject, but I'm to lazy to dig out any scientific data myself. This page that i Googled up does contains a few pointers to articles on the subject if anyone is interested.

sthayashi · Post by **sthayashi** » Tue Sep 21, 2004 1:41 pm

The full version of the article 'debunking' Dvorak is here:
http://reason.com/9606/Fe.QWERTY.shtml

I've found this site to give a bit of a rebuttal for Dvorak;
http://www.mwbrooks.com/dvorak/dissent.html

Zyzzyx · Post by **Zyzzyx** » Tue Sep 21, 2004 8:27 pm

About the best comments I can say in rebuttal to the 'debunking of Dvorak' is that it is by far more comfortable to type, and I haven't lost any of the speed that I had with Qwerty.

The Dvorak layout seems to have more 'flow' across the keyboard than what the Qwerty does. Anytime I have to use a Qwerty layout (usually on a flat keyboard, but sometimes I'll swap my Kinesis) it feels... rough and unrefined. That to me is enough reason to change. I was fairly comfortable with the Qwerty layout on my Kinesis board, with the Dvorak its nice and cozy.

ascii · Post by **ascii** » Tue Sep 21, 2004 11:45 pm

Hey, if Dvorak feels more comfortable to use, you should stick with it. All I'm saying is that evidence points to Qwerty beeing just as fast as Dvorak. That is not really a reson aginst using Dvorak if you prefer it - none of the tests concluded that Qwerty is substantially faster than Dvorak either.

fanerman91 · Post by **fanerman91** » Wed Sep 22, 2004 12:34 am

I don't understand how comfort and speed can't be related. It seems like minimal finger movement and switching between hands are the keys to fast typing, and are the keys to comfortable typing as well. But I have no real hard evidence. Edit: Actually, I might be wrong here after reading more. I suppose there's more to fast typing than what I mentioned. And QWERTY may handle it okay.

Looking at the websites you linked make it seem like both are too one-sided to be objective. The only test (from the links) that seems to have any weight is the Navy test. Which both sides use to argue their points. Can you link other tests?

ascii · Post by **ascii** » Wed Sep 22, 2004 3:45 am

OK, I googled a bit more and found a review article complete with a rather lengthy reference list.

Fanerman, I am completely taken back by how you can think that "The only test (from the links) that seems to have any weight is the Navy test". Did you not read the part about Dvorak himself beeing the test supervisor? Dvorak had a rather large interest in the sucess of the Dvorak keyboard, since he held the patent for it. Did you not read the part about measurements beeing made in different ways, about the typists in the control group not beeing compareable in skill, etc, etc. Out of all the tests conducted on keyboards, one test stands out very clearly as beeing flawed by a lack of objectivety AND by flawed methods, and even some implications of tampering, and it also happens to be the ONLY test whose results point to the superiority of Dvorak. This is the navy article that you seem to think is "The only test ... that seems to have any weight"? Please read again. And again. And again.

If you want more tests, check out the reference list of the article I linked in this post. The Strong article seems to hold it's water rather well when compared to Dvoraks navy test. See also Miller and Thomas, Norman and Rummelhart and a heap of other test.

Also PLEASE read the section on Ergonomic tests in the same article. Just as I stated in my previous post, it clearly states that while the Dvorak has a slightly better layout in some regard (more activity on the home row), the Qwerty layout is excelent at minimizing both same-hand and same-finger typing, which is also very important for maximizing typing speed. For further information, please read Norman and Rummelhart.

Regarding keyboard comfort, like I said before, if you are more comfortable with the Dvorak layout, go for it. I can see how staying more at home row could be more comfortable as your wrists have to move less. But you really shouldn't claim that the Dvorak layout makes you a faster typer as there are dozens of strict scientific tests claiming that there is very little difference and only ONE very, very flawed report and a heap of subjective anecdotal evidence claiming the opposite.

EDIT: Spelling mistake. Maybe I need a Dvorak keyboard after all!

wim · Post by **wim** » Wed Sep 22, 2004 5:48 am

if you doubt the dvorak layout for whatever reason, the best thing you can do is give up a month or so and learn it. until you've personally tried both layouts, your opinion is invalid.

no amount of reading articles and 'studies' can show you the experience. you'll see!

ascii wrote:the Qwerty layout is excelent at minimizing both same-hand and same-finger typing, which is also very important for maximizing typing speed.

WHAT!
which article was this in?
(and where's the actual proof? this doesn't count as evidence just because some article claims it)

sounds like bullcrap to me. i mean how are you going to beat the classic one-hand words like "stewardesses" and "lollipop" .. i challenge you to find an equally awkward word on the dvorak layout

sthayashi · Post by **sthayashi** » Wed Sep 22, 2004 6:52 am

The M. W. Brooks article generally addresses most of the claims made by Liebowitz and Margolis (though he doesn't provide a reference list for his rebuttals)

BTW, I don't think Dvorak really help with spelling that much. I hit the backspace a lot when I type and I think that my problems are just with my head.

But for everyone's benefit: Here's a list of 100 words commonly mispelled in Qwerty and Dvorak. This list was compiled by Dvorak himself, so take whatever grains of salt you need.

ascii · Post by **ascii** » Wed Sep 22, 2004 8:15 am

Sthayashi, Brooks claims that the Strong studies are invalid because Strong didn't like Dvorak, which makes his study unobjective. He does this, while relying on a study performed by Dvorak and anecdotal evidence about Strong. Maybe Strong didn't like Dvorak, but the I would say that Strongs study as a lot more valid than Dvoraks own, the only study I have ever seen that implies the Dvorak keyboard is superior. It becomes even more silly when people like Cassingham, who makes a living selling books about the Dvorak keyboard accuse Strong of beeing impartial whith handwaving arguments. Oh, and a reference list doesn't prove anything in itself. But the reason why Brooks doesn't have one is that he has no valid studies backing him. The reason why Liebowitz and Margolis do have a reference list is that there are numerous studies supporting their claims. In the end I feel that Brooks debunking is a very typical debunking, be glosses over the dozens of different studies that have been done showing Dvorak and Qwerty to be comparable and claims that Strongs study is invalid because (he claims) Strong didn't like Dvorak. Brooks also says that L&M show no proof of the Navy report beeing partial, which they do in quotes like:

'The authors of the Navy study do seem to have their minds made up concerning the superiority of Dvorak. In discussing the background of the Dvorak keyboard and prior to introducing the results of the study, the report claims: ''Indisputably, it is obvious that the Simplified Keyboard is easier to master than the Standard Keyboard.".26 Later they refer to Qwerty as an "ox'' and Dvorak as a ''jeep'' and add: ''no amount of goading the oxen can materially change the end result."'

Also note how he suggests that the Qwerty keyboard had no real initial competition in the section "Claim: The QWERTY Standard was Established by Competition", but in the previous section he has just showed that Shole (Qwertys designer) actually mede at least one other, completely different keyboard layout. Brook even goes as far as to make a smug comment about it's similarity to Dvorak. But wait... In spite of the fact that Shole himself made a rather Dvorak-like keyboard layout Qwerty remained the standard. The list goes on...

Wim, that claim about alternating hands and fingers comes from ergonomics studies.

From Liebowitz and Margolis: "Although the Sholes (Qwerty) keyboard fails at conditions A and B (most typing is done on the top row and the balance between the two hands is 57% and 43%), the policy to put successively typed keys as far apart as possible favors factor C, thus leading to relatively rapid typing. The explanation for Norman and Rummelhart's factor C is that during a keystroke, the idle hand prepares for its next keystroke. Thus Sholes's decision to solve a mechanical problem through careful keyboard arrangement may have inadvertently satisfied a fairly important requirement for efficient typing."

A, B, and C in the above quote are conditions for rapid typing on a keyboard originally described by Norman and Rummelhart. Conditiion C is "The frequency of alternating hand sequences is maximized and the frequency of same-finger typing is minimized."

Oh, and Wim. I think it's perfectly valid to have an opinion on keyboard layouts without trying them. The question I'm discussing is "Will the Dvorak layout usually make people better typists", not whether it FEELS better or makes you cooler or something subjective like that. Whether people learn to type faster and more efficient using the Dvorak layout is an objective question that has an objective answer. Numerous studies have been made and the answer is always no. But I get the feeling that no number of studies could convince you of what you 'know' to be true, or, as someone once said 'We can't let mere facts stand in the way of the TRUTH'.

When I originally heard of the Dvorak keyboard, I wanted to start using it, but before that I read up on it and found a huge pile of articles debunking the claims of Dvorak. Peoiple who have taken the time to learn them should by all means stick by them, no studis have shown tem to be worse than regular Qwertys, but I will always be intrigued by how myths such as the superiority of Dvorak just never die.

sthayashi · Post by **sthayashi** » Wed Sep 22, 2004 9:23 am

I could only find Strong study, and I do have to agree with Brooks on his accusation of Strong's training methodology. 4-hours of training a day is insane and hardly reasonable.

I haven't seen/couldn't find the other studies referenced by L&M. They only referenced them through Yamada (and I couldn't find his paper either). It's interesting that Yamada and L&M view the results of those studies differently.

But allow me to digress a bit. What these studies appear to conclude is that typing in Dvorak is not significantly faster than typing in Qwerty (according to L&M). However, I think more Dvorak proponents are (or SHOULD be) saying that Dvorak is much more comfortable. I don't think there are any studies investigating that claim. It'd be nice if there were. But until there is, we'll have to rely on anecdotal evidence for now.

If you want a reason to switch though, let it be that one.

fanerman91 · Post by **fanerman91** » Wed Sep 22, 2004 10:35 am

ascii

I meant that the navy test was the only one that was cited heavily by both sides (though I suppose the Strong one got a lot of attention as well). But hardly any other tests were mentioned, and those that were were mentioned only briefly. I was hoping you could link more tests. When I did research about Dvorak, I found the 2 links you linked already, but I didn't see any other tests. I was hoping to see some real, unobjective data for either side, particularly data that shows that QWERTY is good for switching hands and fingers (which I still haven't really, though I haven't had a chance to look myself yet). That's all I meant to say and nothing more. It was late when I typed that so maybe I didn't articulate it properly.

Thanks for the link. It makes sense.

I also think (because it's pretty much the only reason I switched) that a good layout not only depends on efficiency, but also comfort. I don't know about any other people that have advocated Dvorak, but when I mention it, I always mention comfort as a big plus as well. I haven't seen any tests that argue that Dvorak is more (or not more) comfortable to type with than QWERTY. All the tests I've seen argue only about speed and not any advantage to comfort. I haven't seen any tests that look at short term and long term comfort (if it's possible to test it), but I'll look later. It's possible I may be wrong here too, but after using both, Dvorak does feel more comfortable.

ascii · Post by **ascii** » Wed Sep 22, 2004 2:41 pm

sthayashi: Fair enough. Might actually be more comfortable to keep your hands near home row a lot. It would probably be hard to test how comfortable something is, peoples answer would greatly depend on what layout they have decided to like.

Fanerman: Often cited, yes. But for very different reasons.

oh, and wim: I tried to come up with a few words that are hard to type with a Dvorak keyboard layout. How about these:

twitch, seppuku, upkeep, yuppie, puppylike, quippu.

ascii · Post by **ascii** » Thu Sep 23, 2004 5:28 am

Some more words that are hard to type with a Dvorak keyboard layout: okupukupu, kukupa, Kippy, Bifid, biddy-biddy.

Ok, I should stop now.

rbsteffes · Post by **rbsteffes** » Thu Sep 23, 2004 5:35 am

This isn't based on test data, it's based on theory. Healthwise, wouldn't a Dvorak keyboard be worse for you in the long run? I know, you're going to jump to comfort level and all that, but here's my line of thought. The Qwerty keyboard was designed to force you to move your hands more, which I believe is why the D-key fans say it's less comfortable. The D-key lets you perform smaller movements in a smaller area. What jumps out at me when I hear that last sentence is RSI. Wouldn't making excessive small movements be worse for your fingers? I thought that's what CAUSES RSI.

ascii · Post by **ascii** » Thu Sep 23, 2004 5:41 am

rbsteffes: I think that there is really to little evidence to jump to that kind of conclusion. There are just WAY to many unknowns here. Trouble is, whereas trying to device a test to see which keyboard allows you to type faster is really hard, trying to device a test to see which keyboard caues less injury, or (heaven forbid) is more COMFORTABLE is nearly impossible with the kind of funding that ergonomics research gets.

sthayashi · Post by **sthayashi** » Thu Sep 23, 2004 7:55 am

rbsteffes wrote:The D-key lets you perform smaller movements in a smaller area. What jumps out at me when I hear that last sentence is RSI. Wouldn't making excessive small movements be worse for your fingers? I thought that's what CAUSES RSI.

By that logic shouldn't not moving your fingers cause RSI? I mean if you make your movements infinitely small, wouldn't you in effect NOT be moving your hands at all?

rbsteffes · Post by **rbsteffes** » Thu Sep 23, 2004 10:45 am

Actually, no. RSI, from the scant bit of research I've read, is caused largely by making the identical movements. (Honestly, a good deal of the 'research' done for either RSI or Carpal Tunnel tends to be on the same level of quality with research paid for by Microsoft showing OSS is bad) Think of it as stress testing a cabinet, if you open and close the same drawer the same amount it'll wear out the fastest in the points that are worked by that movement.

I know it's not based on any actual data of dvorak keyboards, it was just a thought.

sthayashi · Post by **sthayashi** » Thu Sep 23, 2004 11:22 am

rbsteffes wrote:Think of it as stress testing a cabinet, if you open and close the same drawer the same amount it'll wear out the fastest in the points that are worked by that movement.

I know it's not based on any actual data of dvorak keyboards, it was just a thought.

But will it wear out faster if you pull it out 3/4 of the way compared to 3/8s of the way? Probably not. So assuming that Dvorak users utilize smaller movements and range, they will presumably last longer (whether it's nominal or not remains to be seen).

rbsteffes · Post by **rbsteffes** » Thu Sep 23, 2004 12:39 pm

Actually, yes. If you focus the wear and tear to a smaller region, that region will wear out faster. Think of the path your fingers take in order to type, what would be different movements on a regular keyboard become similiar movements on the dvorak keyboard. For example, just hitting the E key goes from being a partial movement of two joints on the left hand to become a single movement on one finger of the left hand.

How much it adds up, I don't know. It was largely hypothetical anyway!

wim · Post by **wim** » Fri Sep 24, 2004 3:28 am

ascii wrote:Wim, that claim about alternating hands and fingers comes from ergonomics studies.
...
A, B, and C in the above quote are conditions for rapid typing on a keyboard originally described by Norman and Rummelhart. Conditiion C is "The frequency of alternating hand sequences is maximized and the frequency of same-finger typing is minimized."

am i missing something? sorry but i simply don't believe this claim. so i wrote a quick program in MATLAB to convince you otherwise. here it is:

Code: Select all

left_q = '`123456qwertasdfgzxcvb~!@#$%^QWERTASDFGZXCVB';
left_d = '`123456,.pyaoeui;qjkx~!@#$%^"<>PYAOEUI:QJKX';
right_q = '7890-=yuiop[]\hjkl;nm,./&*()_+YUIOP{}|HJKL:"NM<>?';
right_d = '7890[]fgcrl/=\dhtns-bmwvz&*(){}FGCRL?+|DHTNS_BMWVZ';

str = '(your message was here)'; same_d = 0; same_q = 0;
for i = 2:length(str)
    if ~((str(i-1)==' ') | (str(i)==' '))
        if sum(str(i-1) == left_q) == sum(str(i) == left_q)
            same_q = same_q + 1;
        end
        if sum(str(i-1) == left_d) == sum(str(i) == left_d)
            same_d = same_d + 1;
        end
    end
end
disp(same_d); disp(same_q);

in place of the "str" variable i pasted your entire message dated Thu Sep 23, 2004 2:15 am, which was 4279 characters long and which i suppose is representative of normal typing for yourself. (i've cropped your message out of the code because it was so long. for what it's worth i first used "find and replace" in notepad to replace all occurences of the character ' with u (a key which is also left on dvorak but right on qwerty), and all carriage returns with m (a key which is also right on both layouts) - this was necessary to avoid interfering with MATLAB syntax)

the result was 946 same-hand presses for the dvorak layout and 1262 same-hand presses for the qwerty layout. that's an extra 33% disadvantage.

for fun i also tried it out on the first 15 pages of douglas adams' the hitchhiker's guide to the galaxy (60,000 characters) resulting in 12701 same-hand presses for dvorak and 19335 for qwerty. that's a 52% disadvantage! i tried it on 17375 character snippet from stephen king's the shining, 3657 dvorak and 5489 qwerty (50% disadvantage).

so what of condition C? The frequency of alternating hand sequences is maximized? doesn't seem that way to me.

Oh, and Wim. I think it's perfectly valid to have an opinion on keyboard layouts without trying them.

you can have your opinion. but i'll tell you that i, at least, am less likely to think your opinion is valid.

for example if you had actually learned the dvorak mapping then you would surely have been less likely to believe the above claim, and probably would have checked it yourself before posting it in the forum as part of your argument. (then again, if you'd tried dvorak i think you wouldn't even be arguing)

I will always be intrigued by how myths such as the superiority of Dvorak just never die.

how can you claim this as a myth without having even bothered to try and disprove it to yourself? you could never be a real scientist.

Dvorak keyboard layout, anyone used it?

UPDATE