Catch up on stories from the past week (and beyond) at the Slashdot story archive


Forgot your password?

iPhone Keyboard Leads to Typso 394

jfruhlinger writes "One of the selling points of the iPhone was its revolutionary touch-screen full keyboard. But a study has shown that text messages sent from iPhones contain significantly more typso than messages from phones with other kinds of keyboards — and aren't entered any faster."
This discussion has been archived. No new comments can be posted.

iPhone Keyboard Leads to Typso

Comments Filter:
  • by orta ( 786013 ) on Wednesday November 14, 2007 @11:46AM (#21349657) Homepage Journal
    I was quite slow with my iPhone keyboard till I started to be more trusting of what the spell checker will fix automatically, there's no mention of anything like this in the article.
  • by ByOhTek ( 1181381 ) on Wednesday November 14, 2007 @11:49AM (#21349703) Journal
    Sadly, it happened long before text on cell phones was common.

    It seemed to start growing quickly out of AOL customers starting circa '94-'95, and sadly hasn't slowed down.
  • by letxa2000 ( 215841 ) on Wednesday November 14, 2007 @11:52AM (#21349749)

    Apparently you missed the part of the study that says that messages sent from iPhones have more errors than messages sent from other phones. So while there may be more tolerance for bad spelling in our society, that has nothing to do with the observation that iPhones lead to more typos.

    It doesn't take a rocket scientist to realize that you're going to have more errors with an interface with no tactile response. The Atari 400 was a decent computer back in the early 80's but was generally scoffed at because of its mesh-type keyboard that offered very little tactile response and made touch typing very difficult. The iPhone is the same, but worse, because there is no tactile response.

    I have a hard time believing I ever would get a phone that has no tactile buttons. I have a Treo and while I can dial phone numbers by tapping the screen and can use a virtual keyboard that would require me typing on the touchscreen, I almost always use the tactile keys instead. With the iPhone, that wouldn't be an option.

  • by Red Flayer ( 890720 ) on Wednesday November 14, 2007 @12:05PM (#21349955) Journal
    Good point.

    But it's also important to note two things:

    1. The iPhone hasn't been around that long, it will take time for users to become acclimated
    2. The iPhone may be used by a lot of people that care less about typos in their texts.

    So before one can say this study shows that the UI for the iPhone is flawed, it's important to normalize the results for both 1 and 2.

    Try the study again in 2 years, among people who have been texting on their phone of choice for >2 years who represent similar cross-sections of the texting population at large. Then perhaps we can come to useful conclusions.
  • by kextyn ( 961845 ) on Wednesday November 14, 2007 @12:07PM (#21349977)
    I own a Windows Mobile device with a slide out keyboard as well as an on-screen keyboard. I never have any problems with that because the slide out keyboard offers a tactile response and the on-screen keyboard makes use of a stylus which helps with accuracy. I have used iPhones on several occasions and I always spend about 3x as much time typing in stuff than I would on my phone. You can't use a stylus to improve accuracy, the buttons are too small for large fingers, and the autocorrect feature can be quite annoying.
  • Re:not suprising (Score:2, Interesting)

    by Anonymous Coward on Wednesday November 14, 2007 @12:38PM (#21350381)
    True, the lack of tactile response from the iPhone keyboard makes it almost useless to use if you are moving around while typing (like when you're walking, or in a car... of course you shouldn't be texting while you drive anyway ;-) Still Apple has done a remarkable job in making the keyboard work. I like the way the keys enlarge as you touch them as a visual response - after a while, this does almost make up for not being able to feel the keys. Some of the apps even let you type with the phone turned sideways, making the keys larger & spaced farther apart - I wish all the apps had that option.

    Ironically, most of my misspelling are due to the iPhone's auto-completing spell checker. As you're typing, the spell checker pops up suggestions; to accept the suggestion, you hit the space bar. To decline, you have to tap a little "x" that appears next to the suggested word (which is never conveniently near the rest of the keys on the keyboard. The result is that more often than not - for me, at least - the iPhone will insert the wrong word while I'm typing resulting in horrible misspellings - even though I would have spelled the word correctly without the "help".

    In this respect, I think that Apple got it wrong. The spell checker should be more passive and not interfere while you're typing.
  • by Vishal ( 29839 ) on Wednesday November 14, 2007 @12:53PM (#21350659)
    It is not an issue with tactile response. The keyboard of the iPhone with its predictive correction is so good that I actually miss it on my regular desktop keyboard. The problem is that "texting" has its own dictionary that the iPhone (thankfully) doesn't recognize. So "texters" make more errors. Good I say. If the same study was done with email instead of text, you'd probably see dramatically different results. I type faster on my iPhone than I ever did on any of my Treos (have had 3 over the years).

  • by gknoy ( 899301 ) <<moc.smetsysizasana> <ta> <yonkg>> on Wednesday November 14, 2007 @01:13PM (#21351025)
    So, then a more appropriate thing would be something like,

    "Wo creas as lnog sa u cna raed it?"
    hu kars so lng as u cn reed it?

    Interesting. The former is made of typos I might make (and have, though not at once ;)), the latter is spelled phonetically. Strangely (or not?), I have a really hard time reading phonetical renditions of words, as compared to typos.
  • by 99BottlesOfBeerInMyF ( 813746 ) on Wednesday November 14, 2007 @01:19PM (#21351131)

    Thanks for the link, that answers a number of my questions.

    Likewise, where do you get your figures? How large a majority of the people who own iPhones did not previously own a smartphone, and just who conducted the study?

    That has been remarked upon by a number of analysts, but it can also be easily inferred simply by smart phone sales numbers. Blackberry sales did not decrease. Palm sales did, but not enough to account for more than a small fraction of iPhone buyers. Just looking at the smart phone market shows that the iPhone, as expected, largely reached their target market of people with regular cell phones instead of smart phones.

    What justification do you have for the idea that a large percentage of those people never sent SMS messages?

    Again, just look at the numbers of Americans using SMS regularly. A quick Google search will show you studies with numbers ranging between 25% and 45% of people in the US ever having sent an SMS message, with lower numbers for regular use. Apple's design and marketing strategy for the iPhone was to target users who don't use the advanced features of phones, because it is inconvenient, hard to use, or hard to learn. The idea is to expand the smart phone market by making it accessible to those who currently avoid it. It is the same strategy they used with the iPod, to woo portable CD player users by offering an mp3 player the average person could use easily.

    Think of it this way, half of all iPhones sold to people 35 years of age or older. According to the Pew Internet and American Life Project study, only about 30% of users in that age group have sent even one SMS message.

    The point is, before anyone tries to use this data to support a particular causation, the study should be redone with a larger, random sampling of people, each of whom is given a particular phone, tested with it, uses it for a month or so, and then retested with it.

  • by tommertron ( 640180 ) on Wednesday November 14, 2007 @01:53PM (#21351765) Homepage Journal
    Actually, I believe 'color' was the invention of Daniel Webster and his "American Spelling," along with a lot of the other simplified spellings which were supposed to make spelling and literacy more widespread because they would be easier. Same with theater, laffter, coff, nife, and the other accepted American spellings.

    (Okay, kidding about the last 3)

  • by pokerdad ( 1124121 ) on Wednesday November 14, 2007 @02:07PM (#21352009)

    say it happened when our former colonies broke away and has been getting worse ever since

    Being Canadian I have a fondness for British spellings of words over American (in most cases), but the elitest attitude towards American spelling found in most English speaking countries only shows an ignorance to the history of the English language. During the 18th century and earlier, there was no such thing as a correct spelling, and many words had multiple recognized spellings. Attempts to standardize spellings began only a few decades before the US declared independence, and were not truly complete till well after. Contrary to popular belief American spellings were not dreamed up out of thin air, but were spellings that had been considered correct for centuries. Yes, American spellings were picked precisely because they were not the ones being made standard in Britian, but my point remains that Americans did not invent said spellings and don't deseve all the critism they get for them.

  • by ( 463190 ) * on Wednesday November 14, 2007 @02:24PM (#21352265) Homepage
    It doesn't take a rocket scientist to realize that you're going to have more errors with an interface with no tactile response.

    Really? That's not obvious to me at all. Since I'm not a rocket scientist, I'll argue the opposite:

    On a conventional keyboard, the only information that gets to the CPU for each button press is 1) which button was pressed and 2) when it was pressed and 3) for how long it was pressed.

    The iPhone, on the other hand, despite lacking tactile feedback, receives massively more data. At the software level it receives a detailed image at probably a hundred dpi or so showing where the pressure is being applied, and if you sample that data often enough you can probably figure out things like the angle and rate of movement of the finger in pressing the button.

    So if you mistype an 'o' ionstead of a 'p', a conventional keyboard only knows that you pressed 'o'. But an iphone knows that you typed an 'o' with your finger way over on the right-hand side of the 'o' key. It could also compare more subtle temporal/motion information about the keypress with how you normally type an 'o' compared to a 'p'.

    Sure, a conventional keyboard _could_ try to do smart prediction something like what the iPhone does, but without all that data it is much more limited.

    I am guessing from your comment that you have not actually used an iphone keypad for any meaningful amount of time. Within just a couple days I was already typing faster and with fewer errors than on my old Treo 650. It really works well.

    However, when you do make typos on the iphone they are somewhat more frustrating, because it usually happens when a whole word is replaced by something you didn't intend. Whereas you might let a single letter typo slide, if the whole word is wrong you have to go back and fix it.
  • by mtaco ( 520758 ) on Wednesday November 14, 2007 @02:30PM (#21352375)
    Seems to be a faux paux to inject reality into a SlashDot discussion, but here are some personal experiences. I've owned an iPhone since day 1; tried other smartphones but they never quite cut it for me. Windows Mobile was just awful, which elminated a whole class of phones, and the Treo was just too clunky.

      I could never type successfully on a Treo , buttons were too damn small and there wasn't any autocorrection. iPhone has the same problem, but the auto-correction works pretty good, especially on longer words. It's kind of weird watching yourself type gppndy and having the iPhone turn that into foobar. Is the keyboard perfect? No. In vertical mode, the keyboard is a little bit small for perfect targeting, in horizontal mode its a stretch for "thumb typing".

      So if you just pick up an iPhone in the store, you're not going to think the keyboard is that great, but you're going to think the web browser and email client kicks ass. For me, that was enough.

    After using the iphone for a week you'll like the keyboard a lot better because it seems like magic to type d;sdjfpy and have that turned into "slashdot". Someone else commented they couldn't even type the first letter correctly, and that's part of the iPhone zen you have to get over. English is only really about 3 bits/letter of information. Factor that into the fact that when you're typing, the iphone knows the general area you were trying to hit, and that that's what makes the autocorrection seem like magic. It's not, its probably just running through all the permutations for the letters near what you typed and ranking them against its dictionary. That's why it seems so magical for long words like "permutation" but doesn't do as well for short words.

        So I wouldn't want to type a post like this on the iPhone, but for "Hey, where are you? I'm ready to go." in a store to my wife it's great.

      Typing passwords REALLY, REALLY SUCKS unless you know the secret, which is "dragging" your finger and releasing when the key is right (type via key-up not key-down). Typing a 128-bit hex key for my WiFi network was really painful as a first iPhone experience.

        So there's the good and bad. Is typing as good as Apple's new kick-ass super-thin USB keyboard? No. Is it pretty good when coupled with the auto-correction for a mobile device? Yes. The auto-correction in my opinion makes the keyboard better then the Treo. Is it better then standard phone keyboards? Much, I could never figure out how to get my phone out of its weird "texting mode" (which they didn't give me any documentation for) so I could send "No". So texting is a way better experience for me on my iPhone then my old phone.

    Engineering is about tradeoffs. If I had to carry around my desktop keyboard as my cellphone, I'd leave it at home. Holding it up to my ear to make a call would look bizarre.

          As for non-geek feedback: My wife never, ever sent a text message from her old phone. She's now a texting fiend and reads her email on her phone most of the time.

    Come to think of it, I text more now as well.
  • by relikmu ( 978868 ) on Wednesday November 14, 2007 @02:37PM (#21352501)
    Again the secret is to push down on a letter (it will instantly magnify), if it is the wrong letter move in the direction of the correct one, it is magnified and let go. Next steps learn what part of your finger touches first. Lastly type faster and enjoy the auto corrections.
  • by ceoyoyo ( 59147 ) on Wednesday November 14, 2007 @06:29PM (#21355765)
    Reading the article, iPhone users, who have had their phones for a month, make more mistakes when typing at the same speed than do users with numeric keypads and hardware keyboards, who have been using them for... ten years?

    I'm shocked.
  • by zoney_ie ( 740061 ) on Wednesday November 14, 2007 @08:21PM (#21357121)
    The interesting thing is that we now have "Anglicisation" of English here in Europe; i.e. using all possible British English spelling differences even where the American spelling is allowable as a variant in British English. For example, the propagation of the "ise" endings in British English. This spelling is distinguished from American English, which always uses "ize" endings. Traditionally however, many words can be spelt perfectly correctly in British English with "ize" endings, indeed some have traditionally always used "ize". A British English dictionary of mine from the 1990s has mostly ize spellings, one I bought last year has ise with ize noted as American or variant.

    Wikipedia has to be the most fun place though, with its mix of British and American English.
  • by wfolta ( 603698 ) on Wednesday November 14, 2007 @08:46PM (#21357317)
    First, I haven't seen a good description of the exact tests they did. The task at hand makes a HUGE difference in terms of how well corrective algorithms can do (terrible on phone numbers, URLS, and other arbitrary data, good on real text).

    Second, the sample size was too small.

    Third, so what if you make mistakes? Even more mistakes. Anyone who would type a message that matters and just hit "Send" as soon as they were done is an idiot. You go back, read, and correct an important message. And for my money, a click-to-correct algorithm is better than a cursor-to-correct one. So if you actually measured SENT message errors, perhaps the iPhone would score much better.

    Fourth, your "experienced" users are how experienced? Do they slow down and take advantage of the visual keyboard feedback on arbitrary text? (Plus the fact that a keystroke registers on key release, not press?) And are they experienced at sending SMS, but you asked them to send a two-paragraph email? Or perhaps vice-versa?

    Bah, probably shills for a competing phone technology.

"my terminal is a lethal teaspoon." -- Patricia O Tuama