Why touch screens?

Question

There is a question on this site on Touchscreens in car consoles which asks on the advantages of touch screens in car consoles. Most answers rather point out why we should not use touch screens in car consoles, and point out numerous disadvantages, as do the answers to this Aviation.SE question. The top answer to the former question does state that...

...Touchscreens are great for computing devices or a thermostat or...

What are the advantages to using a touch screen for a user interface in a computing device or a thermostat, compared to an input device that provides tactile feedback? When typing on a computer, keying in a phone number on a telephone, or a PI Number on an ATM or POS, isn't tactile feedback just as relevant as it is for other applications? Are touch screens just a way to save space (larger screen) and money (interface fully in software), and is inferior user experience considered as collateral damage? Or is there a positive user experience reason to use them?

Why do we use touch screens at all?

Answer 1

When typing on a computer, keying in a phone number on a telephone, or a PI Number on an ATM or POS, isn't tactile feedback just as relevant as it is for other applications?

Absolutely yes. That's why you can't (reliably) use your touchscreen smartphone if you can't look at it. I used to write SMSs doing something else because muscle memory did all the job of typing. Now I can't with my latest full-featured touchscreen-only smartphone.

It's even more important for impaired people. Touchscreen is (usually) flat and only feedback you may have are vibration, sound and light (however there is hope). It may be enough or not but it depends on context. Visually impaired people or people with impaired motor-coordination won't be able to use them effectively unless they use a Touch Screen for the Visually Impaired or a refreshable Braille display (see also Ollie Ford's suggested abstract) and you provide an adapted UX (for a better and easier experience). Note that in this case a proper designed system may provide a better experience with a touchscreen than with traditional mechanical controls.

Are touch screens just a way to save space (larger screen) and money (interface fully in software)...

Yes. On the same space you can put more stuff. More controls and feedback (also using colors and styles) something you simply can't have with a mechanical knob (unless you add more and more LEDs and small displays). However they have one shape whereas mechanical devices may have different shapes and tactile feedback.

Touchscreens also make changes/updates relatively cheap (you can, for example, take advantage of new software features with the same old and cheap hardware).

...is inferior user experience considered as collateral damage?

Sometimes...yes however benefits may often (but not always) compensate these drawbacks. I wouldn't go back to a physical keyboard for my smartphone (even if I need to use it constantly watching the screen).

It's not always true then you may need to find compromises (or to avoid them all together) but it's true for almost every interaction device (unless you feel comfortable with a panel with 100 knobs, 30 switches, 10 sliders and 20 buttons).

Each input/interaction device has its own advantages and application scenarios. Can you imagine to replace your keyboard with voice control system? I wouldn't do it for my programming tasks...

Or is there a positive user experience reason to use them?

They give you the ability to use new gestures unavailable on mechanical devices (pinches, rotation, multitap).

They also compact more controls in less space hiding (at first) what you don't need in that exact moment. Note that context is not a prerogative of touchscreen (and screens in general) but it's also applied to mechanical devices like knobs (where function changes according to state, selected for example with another switch). However, because screen will (should) greatly change according to state then they will minimize possible Mode-Error-Slips (when a device has different states in which the same controls have different meanings). Norman reports an accident of an Airbus because of this:

The flight control equipment (...) had two modes, one for controlling vertical speed , the other for controlling flight's path angle of descent. In one case, when pilots were attempting to land, the pilots thought that they were controlling the angle of descent whereas they had accidentally selected the mode that controlled speed of descent. The number (-3.3) ... was to steep rate of descent when interpreted as vertical speed (-3,300 feet/minute): -3.3° would be only -800 feet/minute...

Of course mechanical devices may be properly designed to minimize this kind of errors (in this case they changed descent speed to be always displayed with four digits) but a well done touchscreen UX (when applicable, there are also other factors to consider) may greatly reduce them (for example with well-visible labels, full length numbers and even different gestures).

Why do we use touch screens at all?

Because most of times their benefits are more and more important than they drawbacks.

It doesn't mean you can/should use them everywhere. Sometimes because of interaction style itself (turbolence on airplanes, in link you provided) and sometimes because they slow-down normal operation: think about difference between radio receiver designed for military operations and for (amateur) pro-users: in military devices you have few essential controls while pro devices are usually full of controls.

Let's pick these two amateur radio receivers and imagine to use them:

Compared to:

In my opinion (unless you constantly use each single knob multiple times per minute) the second device (with touchscreen) may provide a much better UX. Of course these pictures show two extremes, I think (in this case!) best option is half-way. A good UX job has been done (IMO) for digital oscilloscopes (less influenced by marketing trends than radios and digital cameras):

I do not think touchscreen is best and I do not even think it is bad...it's a good option in many scenarios but not always, both from UX perspective (because of UI usability and accessibility) and from usage patterns perspective (because of environment conditions and typical usage). Cost and marketing trends will also play their roles.

To summarize, you should at least consider this (not exhaustive!) list of factors:

• Environment constraints (temperature range, weather, illumination, dust or chemical agents).
• Usage constraints (with gloves, with wet/humid hands).
• Usage conditions (in airplane during turbulence, in car where you can't constantly watch the display).
• UX considerations (immediately accessible features vs more compact navigable organization).
• UI constraints (it has to be used by visual impaired people or by people with motor-coordination impairment?)
• Need for constant updates (mechanical devices can't - usually - be easy replaced but a SW UI can - usually - be).
• Users expectancy (it doesn't matter what's better, sometimes users want something made in the way they know) and preferences: some gamers buy a case with knobs (yes, more than one) to fine control CPU/case fans' speed (!!!).
• Marketing trends (again it doesn't really matter what's better but what users want because it's trendy).
• Users ability, time and motivation to learn how to use a device. Touchscreen (but in general any non mechanical-only interface) is good for a progressive disclosure while with a full-of-stuff panel you immediately see almost everything you can do. Actually you can see this both as a benefit (easier to use) or a drawback (lack of affordance).
• Cost and availability.
• Safety (sometimes mechanical controls are required because of increased security, users are used to confirm UI actions but to unlock a switch requires an active action).
• Law (sometimes usage of one device instead of another is simply required by law).

Give all these factors (and others!) you will pick better compromise from a completely mechanical UX to a full touchscreen UX through different levels and UX solutions for hybrid models.

Answer 2

Advantages

A touch screen has the advantage of greater context sensitivity. The sizes, shapes, colors, and labels of controls can change during operation to reflect what is needed for a particular step of a process. Non-touch devices often simulate this by placing a control next to the bottom or side of the screen and displaying a label next to that control, such as the buttons on the side of a gas pump's screen or the left, OK, and right action buttons of a flip phone.

Multitouch screens give an additional axis of displacement between touch points without needing much additional space. This is used for pinch gestures to zoom and rotate objects on the display, so as not to clutter the screen with rotation and size handles.

Disadvantages

But a flat sheet of glass is a poor choice when the user's eyes are diverted elsewhere. In the case of a car's entertainment system, the user's eyes are on the road. In the case of text editing, the user's eyes are on the document to catch mistakes at the insertion point. In the case of a video game that isn't point-and-click, the user's eyes are on the main character that he is controlling. This will likely become less of an issue once haptic feedback matures.

And even when the eyes are focused on the display, finger touch has another disadvantage. A mouse, trackball, trackpad, or stylus can acquire finer targets than a finger. The iOS Human Interface Guidelines recommend giving a control's hitbox a width of 44px.* Controls in a mouse- or stylus-driven GUI can comfortably be as small as half that size in both directions, such as on a toolbar, or one-third the size vertically, such as for a button with a text label. Or if someone is working on a drawing, it can become tedious to pinch zoom in to make fine strokes with the finger, pinch zoom back out to see the context, and repeat. A larger touch screen can work around the context problem by showing zoomed-in and zoomed-out views side by side, but a larger touch screen also tends to invoke the gorilla arm effect when the user's arm is raised too long.

Finally, affordable touch screens might not be quite as rugged as physical controls, and use with dirty or gloved hands is often problematic.

_{* In CSS, 1px refers to approximately 1/2700 of the distance from the eye to the display. This translates to 44 pixels on a standard display or 88 pixels on a Retina® display or other high-DPI display.}

Answer 3

One less device

We use touch screens because it removes the device that acts as a proxy for our hands.

Humans act on the physical world haptically—through touch and pressure.

A mouse, or other pointing device, lets the user move a pointer that represents a finger.

Touch removes the need for the proxy device. It also enables multi-touch in a way that would be challenging to do with a pointing device. (Imagine using two mouses to "pinch" a photo, to zoom out.)

Technology has further to go. Currently, a touch screen feels like a touch screen—a piece of glass. Could it feel hot or cold? Could it feel sharp or fuzzy?

Haptic feedback

The thing that will make touch interfaces truly wonderful will be when they can "touch us back" by creating the illusion of ridges or of movement on the screen.

That is, if the application displays a button on the screen, when I push it with my finger, I want to feel it clicking. Similarly, when I nudge an object, I want to feel its edge on screen.

Eventually, touch screens will be able to reproduce a variety of textures. So when you say "Show me the kittens" you can also pet them.

Answer 4

I have a critical view on touchscreen usage, but their huge advantage is the flexibility of dynamic interaction and representations. It is not related strictly with money -- you simply can't create some complex transitions with physical controls.

Another reason is direct manipulation. It is still not perfect because of lack of tactile feedback and some user interface solutions, but it is the real direct manipulation, when you connect different building blocks in an interactive scheme or move one data item from one group to another. There is no mouse, there are no buttons between you and the data display and so on.

These are main interactive advantages, that combined with low price and relatively high level of reliability (one touch screen against tens of buttons, joysticks and knobs) makes this decision most popular in most cases, where tactile feedback and other concerns can be unnecessary. But the reason of car or other usage -- that is another case, sometimes it is because of low understanding of the difference and peculiarities of different interaction ways and pros/cons of certain equipment types. Also it is some kind of fashion influence on decision making -- for instance, currently touch screens can be perceived as more effective and "cool" simply by visual design, not behavior) and so on.

Answer 5

A little about touchscreen POSes in a service industry environment.
They are actually enormously advantageous over tactile buttons for a few reasons.

• Button assignments change frequently
  In the older registers with tactile buttons, the function of the button would have to be reprogrammed when an item changed. Not only would the price, item categories, etc., have to be reprogrammed, the machine would have to be partly disassembled to change the button label. Buttons on modern touchscreen POS machines can be removed and reassigned in seconds.
• Thru-put, A.K.A. Speed.
  This is a huge deal in the service industry. It seems inconsequential, but the extra time required to push a mechanical button adds up when you're considering 300 button presses per hour. Operating a touchscreen POS is much, much faster than operating one with mechanical buttons.
• Buttons die
  Buttons assigned to popular items went out constantly on those older POS systems due to mechanical failure. And the buttons that don't fail become completely illegible.
• Cleanliness
  POSes are gross regardless, but touchscreens are far easier to clean.

Answer 6

Touchscreens are great for when you can take your eyes off the road. Which, while driving, is a vanishingly small part of the time. Most of the commonly used functions in a car remain with knobs (NOT sliders which are horrible in a bouncing vehicle) buttons and stalks because you can reach in the right direction, grasp the proper control by feel, and actuate it correctly, without being able to see it. That would be: driving controls like turn signals, air controls like fan and heat and which heat/ac function, and possibly music volume and a very few buttons.

So the idea for car controls is that you have to treat it like an interface designed for a blind person, because eyes are not reliably available, and even so, lighting varies and it is a strain to adjust focal distance from infinity to reading distance and back a lot.

Have you used an MP3 player connected to a car, one with buttons? Have you compared that with the music play function of a phone while driving? Huge difference in convenience and SAFETY. Buttons, knobs and stalks will not be going away until people no longer need to see the road.

Aside from safety and convenience, well, of course a touch screen is massively cheaper in terms of ease of design, programming, updates, etc. Just as a digital watch is massively cheaper than a grandfather clock. (So why didn't we invent them first?)

Answer 7

What are the advantages to using a touch screen for a user interface in a computing device or a thermostat, compared to an input device that provides tactile feedback?

Since you are asking for advantages I will try to just list the advantages here I see:

1. State can be saved and restored quite simple. A physical slider would require to be motor driven to return to a state automatically when it is loaded from a file.
2. Everything moveable is likely to produce other operative problems (consider dirt in knobs)
3. You can have a lot of different (convenient) controls on a very limited space (not at the same time, but a too small keyboard may still be better than to turn and push a knob to type in text letter by letter).
4. Some kinds of input, like point your destination on the map can be done more convenient on a touch screen than, let’s say, have a knob which switches between move the map left-right, move the map up-down, and zoom on push, and performs the action on turn. Push longer than 2 seconds to confirm. Also simple things, as sliding a list, can be more convenient because you can give a fast touch gesture to scroll fast and a slow gesture to scroll slow, where turn buttons usually only behave linear.

Answer 8

Touch screens are not in and of themselves good or bad at any given application. As someone who designs, builds and tests kiosk and kiosk software for a living it might be surprising to hear that I think physical interfaces are better (currently) at all kinds of tasks, especially in the car.

One of the problems is that touch screens are relatively new tech (in that they have only recently become both cheap and good), and traditional knob-and-button interfaces are much, much older. That means for now most touch screen interfaces still more or less emulate old school interactions (with usually poor substitutes for the old controls represented as pixels on screen instead of physical, tactile objects). That isn't an intrinsic property of touch screens though; it's just an easy way to translate old systems to the new ones through a period of transition.

Touch screens offer some amazing opportunities to improve eyes-free interaction in a car or other situation; it's a large, easy to target surface that can accept input anywhere on its surface. Pair a touch screen with a heads up display or some other kind of feedback mechanism and you could easily allow the user to turn up the volume by sliding their finger vertically anywhere on the touch surface (in one mode) or to advance a heads up menu using the same gesture at other times.

There have been some investigatory designs in this space of pure gestural interaction (this one comes to mind), which indicate some of the ideas.

Setting aside the future possibilities for a second though, the other big reason companies like touch screens right now is that it provides a way to fully brand the experience. While that was technically possible in old-school scenarios, now Tesla can make its dash unit reflect its branding in a way that other manufacturers can't (or haven't). It's also able to (theoretically) be adapted as the company's brand changes through updates.

Answer 9

While this doesn't necessarily apply to thermostat controls or other complicated screens, touch screens are excellent to use because they're intuitive.

I know a lot of people who are not very savvy with computers, but show them how to swipe and pinch and they'll immediately be able to organise their photos and they're not likely to forget the basic actions. With a mouse, there are two things to learn: which commands do what (scrolling enlarges/shrinks, holding a click moves the images) and how the peripheral works (moving the mouse moves the pointer, the middle wheel thing moves up and down with a stroke, there are two buttons but mostly the left one is important) instead of just learning the logical commands with respect to your fingers (grabbing and pulling enlarges, dragging your finger moves it). Look at a toddler, an octogenarian or a computer-illiterate on an iPad, they're much quicker to pick it up.

Answer 10

There is one clear advantage to use touch screens. Its not really a UX question though.

• Touchscreens are cheaper for the manufacturer to design for. Designing the buttons and fixing the interaction is a costly process requiring mechanical engineering and sourcing.

A touch screen is essentially easy to place in. It is also easy to developers of the software. They can easily test without building a mock-up etc. The fact that it is hip and cool makes this an easy decisions.