The “Over The Phone” Test
Here at Humanized, we use the “Over the Phone” test as a good rule of thumb for interface design.
If your friends, parents, or acquaintances have ever got it into their heads that you might be good at technology, you have no doubt found yourself trying to explain some aspect of computing over the phone. I give you my condolences. Trying to troubleshoot a GUI over the phone is like giving vocal driving directions to someone who doesn’t drive.
For instance, it’s difficult to tell Grandma how to spellcheck in a web mail application that doesn’t have spellcheck:
- “Ok, Grandma, select everything by clicking anywhere in the text and typing Control-A.”
- “Now copy the text by typing Control-C.”
- “Open Word by going to the Start Menu, clicking ‘All Programs’, clicking ‘Microsoft Office’, and finally by clicking ‘Word’.”
- “Paste your text into Word by typing Control-V… No, Grandma, I don’t know why it’s ‘V’–maybe because ‘P’ is already used for printing?”
- “Click the little icon on the top of the bar that has a checkmark and some letters. You don’t see it? Okay, describe what you see. You see something that looks like a coffee cup? I have no idea what that is. Actually, forget it. Just select ‘Spell Check’ from the ‘Edit’ menu.”
- “Click ‘Start Checking’. Grandma, I know you just told it you want to spell check; I don’t know why you have to tell it again.”
- (Time passes as Grandma spell checks her document.)
- “Click the ‘Done’ button. Or maybe it’s called ‘Finish’. Uh, just click either the ‘Done’ or the ‘Finish’ button.”
- “Select everything by typing Control-A.”
- “Copy the text by typing Control-C.”
- “Switch back to the email you were writing. What’s that? You can’t see the email you were writing? Well, um… Move some windows around and try finding it.”
- “Is the text of the email still selected? No? Okay, click anywhere in the text.”
- “Select everything by typing Control-A.”
- “Paste in your spell checked text by typing Control-P. Wait, no, Control-V…”
- “You’re done! wasn’t that easy?”
At every step something can go wrong, your mental model can get out of sync with the state of the computer, or you might remember a button name incorrectly.
Thus our test: We ask ourselves, “Would I be willing to teach my Grandma how to use this over the phone?”. If the answer is “Definitely”, we know we’re doing well; if the answer is “Maybe”, we know we can do better; and if the answer is “No”, then it’s time to rethink the whole thing.
RT @azaaza The “Over The Phone” Test | Follow @azaaza on Twitter | All blog posts
cold wolf
Ha! That’s brilliant! A bit harsh though–as in, setting the standard a bit high. Which is also brilliant, because developers needs to stop being lazy and start building programs for humans, not machines. And that’s my beef with open-source software: they say “we’re doing this on our free time, be grateful” while I end up with more problems than the ones it solved. And they expect me to be grateful? But if I provide constructive criticism, they get so defensive. Honestly…
The biggest problem of all is that we have so much tech to help us, but too few of us know how to turn it into something that genuinely helps.
Richard Karpinski
Two answers for why V for paste:
1. XCV are adjacent on the qwerty keyboard
2. X it out, Copy it, V (plunk) it down
You ask if I remember personal info. I do, sometimes. Some of it I have entered in this site three or four times in the last few hours so maybe you could guess the answer.
My grandma died forty years ago, so I would be quite eager to teach her anything over the phone. That would be wonderful.
Could I see what I am about to post? Oh, no matter, I can undo it If i don’t like it, right? And you’ll remember it if I forget to click on the button, right? Humane software is so nice.
Pgan
Mnemonics for key-combinations are nice, but perhaps a bit ethnocentric, as they may not work in other languages. For example, Ctrl+A is mnemonic in English (“select All”), but not in Bulgarian. Using commands instead of key combinations has the same problem. Either users of some languages have no mnemonic, or they use different key combinations and commands.
On the other hand, it makes sense to me to use three universal, ergonomically convenient, consecutive keys for three related operations.
Admittedly, these are hard to explain to Grandma over the phone, which may point to a weakness in the Over the Phone Test.
Pgan
By the way, a better title may have been “… trying to give cooking directions to someone who has never cooked”. I think anyone who navigates by walking would understand instructions for driving like “turn right, go for 1.3 km, take highway I-12 going east”. But in my experience, it is surprisingly hard for people who have never cooked to grasp a recipe. They apparently do not have the primitive concepts of the cooking language, aquired by watching and doing cooking.
Atul
It seems to me that ultimately, any humane user interface has to be “ethnocentric” in that it needs to be carefully adapted to the locale it will be used in: if an interface can’t make itself visible to its user without using some sort of unambiguous language, then how is its user supposed to know or remember how to use it? Adapting an interface to suit a particular locale seems not only inevitable, but preferable.
And I would also argue that using “three universal, ergonomically convenient, consecutive keys for three related operations” as the sole basis for a user interface wouldn’t do well on the Over the Phone test for quite a good reason: if it weren’t for the heavily language-dependent mnemonics involved in remembering cut, copy, and paste, I’d have lots more trouble remembering which key did what, for the exact same reason that I have trouble remembering which key does what on my cell phone’s voicemail interface. Just imagine if it was Ctrl+1 for cut, Ctrl+2 for copy, and Ctrl+3 for paste: regardless of the ergonomics involved, anyone is going to have a hard time mapping the numbers 1, 2, and 3 to clipboard actions.
ashley
Where did the “Over the phone” test come from? Did someone invent it? Is it one of those things they teach in HCI school or something?
Aza
The “Over the Phone” test was originated at Humanized by yours truly. Hopefully it will become something that they teach at HCI school in the future. :)
Pgan
By “ethnocentric” I meant that ^A would have a mnemonic in some languages but not others, which is unfair. This is not a problem if, say, Bulgarian uses instead ^V (for “vsichko”).
Agreed, Ctrl+1 etc. would be hard to use until you develop the reflexes (The Humane Interface distinguishes ease of learning from ease of use once learned).
Then every command in a shell and in Archy (RaskinCenter.org) should be translated to the user’s language? There is a lot of work for translation.
And situations like kiosks set in a different language — which are bound to occur — would pose great problems. User’s habits would work strongly against them.
Perhaps this is the better option, but I want to point out the trade-off.
Stephen
I actually think keyboard shortcuts are for those users who do in fact have time to memorize things like Ctrl-X,C,V or Ctrl-1,2,3 – trading training for an increase in speed. You could have told your grandmother to right-click, then “Select All”, and the same for copy and paste.
That of course merely shifts the debate to be about the current Windows mouse configuration (left-click for selecting and positioning, right-click for a list of contextual operations).
But your original point still remains in tact (confusing menus, windows; there’s a whole lot of Bad in that conversation).
As an aside, this weblog seems like it could be a prime target for holy wars. I know people who have spent weeks debating where to position the OK and Cancel buttons on an OK/Cancel dialogue. Hope you’re ready ;-)
Vitorio
I always thought the reason Cut, Copy and Paste were X, C and V were from original proofreading conventions… you know, from writing with pens and pencils on paper.
You crossed something out to “cut” it (from the next rewrite), a cross is an X. Copy, which has no proofreading equivalent, is C, because Copy starts with C. And paste (or insert) is V, because when you write text for insertion on paper you write a caret (which looks like a little “v”) above or below the “insertion point.”
It’s just coincidence that they’re next to each other on the keyboard.
Mike
Your grandmother uses webmail? It wasn’t that long ago that I was coaching my mother just on how to turn a computer on.
Though you don’t say it, I get the feeling that the suggestion here is that with a simpler solution (i.e. your solution) even Grandma (or Mom) will love using the computer! I’m not convinced. Okay, so you come up with a simpler way to spellcheck, but just the idea of spellcheck might boggle some people’s minds. For instance, pushing a button isn’t hard, but my mom definitely struggled with it, at least for a while. There’s an issue of intimidation involved.
I’m also not blind to the fact that intimidation and ease (or rather unease) of use are closely conjoined, but really all you can promise to do is to make computers easier for those who already use computers. I think this is an issue of humility that it would behoove you to note.
Atul
Personally, I think this may be an issue of the concept of essential complexity versus accidental complexity. It’s possible, to a certain extent, that the very concept of spell-checking a document with the technologies we currently have available to us will inevitably involve some complexity and unpleasantness. In that case, though, I’d say that our goal would be to create an interface that comes as close as possible to passing the Over the Phone Test. So even if it’s a bit idealistic, I think it’s still quite useful as a sort of “litmus test” for user interfaces.
I understand your comment on humility; my apologies for any negative sentiments our words may have fostered. Being a new company that’s just started a weblog for the first time, I’m certain we have some things to learn about effective writing. Thanks for your feedback.
Mike
Don’t get me wrong–I think it’s a good test. But it can be run on anyone who’s new to the system you’re explaining, computer-saavy or not. The basic idea of the test is “If we don’t abstract away to notions like ‘cut’ and ‘paste’ which the listener already understands, how difficult a process is this”… right? It assesses steepness of the learning curve and the monotony (in the good sense) of it in an easily conceptualized fashion–exactly what a good “litmus test” does.
The contrast between essential vs. accidental complexity is a big part of what I was trying to get at, though I didn’t have those terms–I definitely agree with you there. Some things are essentially complex, and some people are going to be afraid of that complexity. I think that’s just the way it is. I wasn’t trying to get on your case or trash your ideas; I was just getting this quixotic vibe about the changes your software was going to bring.
But hey, by all means, prove me wrong! One less hour coaching my dad how to connect to his wireless network (or whatever) is one more hour I can spend… well… figuring out how to connect to my own network!
urbaer
Control V? Why control v? Personally feel that Shift-Insert makes more sense. Okay, it’s a two handed thing, but still.
Agreed though, over the phone is horrible, especially when your mother is still running Win98…
Clark
As I was reading from a site that included anecdotes about the beginnings of the Macintosh, I mused as to why the GUI paradigm hasn’t seemed to change much in 20+ years…
… then, VOILA! I discovered humanized.com.
I like the idea that a group with serious computing experience cares enough to help bypass this tired Windows world.
After 20 years, we still have a crap GUI, which is basically akin to a few tired bells-and-whistles riding on top of ol’ MS-DOS.
So, why should we give you a hard time, here on this blog?
I think a few well-placed keys on a new-style keyboard would be better than reverting to command-line-type stuff.
The overall idea is wonderful, though, to have a streamlined, universal method to access programs, files, etc..
James
Universal natural syntax commands — I like it! Now, how much longer will I have to wait for voice command?
mrbig
Imagine all this when you’re outside an english speaking country. I’m from Hungary, and usually my first question is: Grandma, are you using english or hungarian windows? And the word is english or hungarian?
And I might remember some menu items only either in english or in hungrian, like: “I know it was named Control Panel in english, but how is it translated?”