# Is there any difference in element minimal sizes on 10 and 24 inch touch screens?

I am developing an application which is supposed to be used on a 24" touch monitor.

There are lots of guides on physical sizes of a touch screen UI elements. All of them are saying that every button/touch-control should be at least 9-11 mm. I think it is a good advice for tablets. But what about 24" touch monitors?

Are minimal dimensions for 10" tablets same for 24" monitors?

-
You'll want to look into Fitts's Law for this one, I think. The buttons should be larger on larger devices because they will be further apart than they would be on a 10" device. – JonW Jun 21 '12 at 8:32
I have solved the Fitts' formula to `W = D/(2^((T-a)/b) - 1)` and got the result, that button size should be more than one meter. :) Either I am wrong, or the formula does not fit my needs. – Vasiliy Borovyak Jun 21 '12 at 9:20
The minimal dimensions would be independent of the screen size, just as the size of your fingers is not dependent on the size of the screen. – Dan D. Jun 21 '12 at 9:55
@VasiliyBorovyak Ah, well you could move them closer to the related items I suppose! However if the actual calculations are not practical, the principal is still the same - the further away from the likely pointer position different actionable items are the larger they need to be in order to accurately select them. The finger / pointer has to physically travel further so accuracy is diminished as a result; larger buttons counteract this loss of accuracy. – JonW Jun 21 '12 at 9:58
@JonW you are correct, of course. I am looking for some numbers or calculation maybe. I believe I have to calculate all the stuff myself using Fitts' formula. Thank you. – Vasiliy Borovyak Jun 21 '12 at 11:31

Microsoft's touch interaction guidelines for Metro apps (generally used on screens like those) vary from 5–9 mm square, depending on the tolerance you provide to undo accidental presses.

Touch screens aren't all built alike, though; some are extremely precise and some are very poor with touch accuracy. As the panels increase in size to 17 inches and above the parallax effect can even make the touch accuracy highly variable on a single panel. With some technologies, the screen edges are much less responsive than the centre.

Parallax is the major effect that would make this difficult. Many touch-screen devices use projected capacitive touch (PCT) sensors which are effectively behind one (or often two) layers of glass. Even relatively thin glass (~3mm) can have a profound impact on the apparent touch location on different parts of the display (an effect exacerbated by the distance of the user from the display—exacerbated by extreme height disparities in users—and the specific area on screen the user is targeting).

It's a bit difficult to describe effectively but here's a diagram to get the point across:

The easiest way to resolve this issue (especially factoring in the expectation that different users will be at different heights and distances from the screen, and thus the yellow "perceived" hit target is in a different place for every user), is to simply expand the hit target itself (either visibly or invisibly) to accommodate fuzzier targeting.

It's worth pointing out that the effect is greatly lessened as the number and thickness of the layers between the LCD panel itself and the user is reduced. Some technology (like the infrared framed displays on most very cheap touch-screens) are practically immune to this issue, but as you can tell, there's more to answering this question than simply the size of the screen.

For what it's worth, our software written for PCT touch-screen kiosks (targeting 17- and 19-inch displays) has no interactive element smaller than 2 cm square (and even those elements are generally non-essential/convenience controls; most of our buttons are as large as 4–5 cm on their smallest dimension). On very large PCT displays (30+ inches), even those may cause issues due to the extreme parallax effects, so I'd then focus my attention on getting the front glass made as thin as humanly possible.

-

Note Fitts's Law is for pointing in two dimensional space by the way; tapping is a different action. Touch occurs in three dimensional space unlike dragging a mouse. There's lots of other stuff I could say about the differences, but the main thing is you shouldn't be trying to directly calculate it in most cases. Look to Fitts's Law as a rough guide, you should really only be actually crunching numbers with it if you're doing formal research.

Also see these interesting posts about touch and Fitts's law.

Now, the reasons for the physical minimum sizes on touch screen devices are completely unrelated to the size of the screen; they're related to what you can touch with your finger with reasonable accuracy and comfort. The minimum effective size is still pretty much the same, baring other considerations. If anything position on screen (how far to I have to reach to press this button?) is more relevant than the size, as long as you meet a reasonable minimum.

As a real word example, let's look at Windows 7. Certain elements like the Taskbar are very well designed for touch; predictable location, large enough to tap without error (mostly), they work pretty well, even though they're relatively tiny compared to the size of the screen, they're usually at least half an inch wide. Even as a compact element they work well, unlike many other default UI elements; checkboxes are quite painful to work with via touch in Windows 7.

However, since you've got much more space, there's really no reason not to make the buttons physically larger on average; it's still easier to tap a 10" wide button than a 1" wide button. Determine the best control sizes based on how important the button is, how much space you need for other elements etc.

In the Windows 7 example above (and the iOS tab bar) a minimum touch size is used because these are almost always present and always in the UI so they shouldn't get in the way. Unless your app needs to make serious and efficient use of the whole 24" display your touch points should be larger on a 24" machine. How much larger? There's really no magic number. If using it feels awkward, it's probably awkward. Go by how it works.

-
This question has also inspired me to ask Has Fitt's Law been adapted to Touch screens? – Ben Brocka Jun 21 '12 at 14:30
"Note Fitt's Law is for dragging a mouse"...well, no, it's for reaching and pointing in the real world. Adapting it to a mouse came later. – Alex Feinman Jun 22 '12 at 14:14
I must disagree with the answer. "The minimum effective size is still pretty much the same" - nope, they are not for many reasons. The distance my users would work with app is at least 1/2 a meter, and 1 meter at most. Moreover, some users would place our app on the left of them, some - on the right, and some - in front. "there's really no reason not to make the buttons physically larger" - nope, there is a reason called Designer, who have never designed anything for huge touch screens, and another reason is Project Manager, who loves the Designer. But these two are my problems, I suppose. – Vasiliy Borovyak Jun 25 '12 at 9:19
@VasiliyBorovyak minimum effective size is different from "what's comfortable" or "what's nice". What you can effectively touch is not a function of the screen size or distance to the screen (ignoring extremes where you can just barely touch the screen). And the latter part of your comment is just silly and unrelated; there might be random business/style/ect reasons, but people do stupid things all the time, that doesn't really mean there's a good reason to do them. – Ben Brocka Jun 25 '12 at 13:24
Voted the answer down because of this statement "Note Fitts's Law is for pointing in two dimensional space by the way; tapping is a different action. Touch occurs in three dimensional space unlike dragging a mouse. " As Alex also commented, this statement is incorrect. See this article for a description of the tasks in his first set of experiments. – user1757436 May 6 '13 at 14:55