139

Most of us use switches every day in various forms. And although there are sometimes general norms in countries as to which direction is "on", there is variation between countries.

So when developing a product with a switch in it, which direction makes the most sense for "on", and why?

I'm not talking about other ways of showing that they will be on or off, I'm more talking about whether "on" is up, down, left, or right. Any research on this would be appreciated.

enter image description here

JohnGB
  • 68,376
  • 26
  • 180
  • 294
  • 14
    How about a pushbutton that is on when you press it, and when you press it again turns off? No direction needed, works everywhere... Just my $0.02... – Jeroen Landheer Mar 12 '13 at 14:34
  • 5
    @JeroenLandheer: Good point that this is another option. But that is another direction (inward). It's also a toggle, and the toggle characteristic could apply to switches that are pushed in other directions as well. – LarsH Mar 12 '13 at 14:55
  • 3
    The image you embedded is a SPDT (single pole, double throw) switch having three states, both up and down can be on and middle is off. So...? (interesting) – Alvin Wong Mar 12 '13 at 15:05
  • 1
    @AlvinWong It was just an example of what I meant by a switch, so that people don't start thinking of networking switches. – JohnGB Mar 12 '13 at 15:42
  • 1
    @LarsH, A pushbutton doesn't have to have an inward state, optionally can have a light associated with it (even on the button itself), is used in all kind of forms (even on ballpoint pens) and works globally the same... Pushbuttons with lights behind it can also used to indicate state of a device or operation, alert conditions, etc. The thing is though, not every pushbutton is technically a 'switch' (as in switching from one state to another), you can look at your keyboard for an example of that :) – Jeroen Landheer Mar 12 '13 at 16:10
  • 1
    Intuitively I would go for up/right is ON. If I had to explain I would say that for slider, you would increase energy by going up. About the horizontal choice, LTR writing direction is influencing me (progress bars go from left to right as well). – Guillaume86 Mar 12 '13 at 17:52
  • 3
    Intuitively I would go for down/right is ON – mplungjan Mar 12 '13 at 18:03
  • 12
    Personally, and there's nothing to support this but my opinion, I prefer off to be down, on to be up. As a SF writer, I've always thought that if anything falls on a switch, brushes against it or it's hit in some other way, I'll want gravity on my side so that as a fail-safe the switch defaults to off in the direction of gravity. If there's no gravity, then on is outward, off is inward (because if it's stuck pressed, it's often better to be off). Generally, make the 'safe' position for the most stable state, whether on or off. – Noein Mar 12 '13 at 18:19
  • @JeroenLandheer: I generally agree with what you said. But the direction I was talking about wasn't the direction of state ('inward state'); rather it's the direction of pushing. The pushbutton you're talking about does have that direction. – LarsH Mar 12 '13 at 20:33
  • 1
    Wonder if NASA has any guidelines for switch positions (operated in anything between 3-g and 0-g environments). – Deer Hunter Mar 12 '13 at 21:18
  • 1
    Symmetrical looking switches are confusing. Apart from that, I find the action "On" or "engaged", more close to actions "lift up" or "move forward to right" (popular reading direction). – user117 Mar 13 '13 at 06:37
  • 2
    Even if we just look at mplungjan and Guillaume86's comments we can see that intuition isn't enough here :-) – Rory Alsop Mar 13 '13 at 10:36
  • Most replies mention either up or right as most natural to reflect on. If the most important thing is that you know what the position means when you see the switch, I would therefore recommend you to place it diagonally where upright means on. – Dennis Jaheruddin Mar 13 '13 at 17:14
  • 1
    "Cultural question: which position is on, which is off?" http://en.wikipedia.org/wiki/Talk%3ALight_switch#Cultural_question:_which_position_is_on.2C_which_is_off.3F – Kris Mar 14 '13 at 06:07
  • Interesting to note that in switch panels (breaker boxes) if you mount the breaker on the left, ON is to the right (center), and if it is on the left, ON is to the left (center) at least in the US where the panel is mounted vertically - which matches the "power is in the middle, neutrals are on the outside so the ON is towards this power bar - such as the manual old fashion clasp type switch. (throw switch) – Mark Schultheiss Mar 14 '13 at 12:15
  • What “is one of the major causes of mishaps” ? – Nicolas Barbulesco Mar 16 '13 at 23:07
  • Funny you should ask this. A few months ago I built a hardware project in which I re-used the power switch (among other components). It is a rocker switch. Annoyingly, in the original device it was installed so that down was on. When I replaced the front panel with my own, I reversed it so that on is up. It is not only my preference, but consistent with other devices in the rack. It's easy to reverse that kind of switch. Just pop it out of the rectangular aperture, turn, and replace. http://www.kylheku.com/~kaz/smf-1/ – Kaz Mar 17 '13 at 10:49
  • @Kaz Some switches object to reversal: http://1drv.ms/1F7qGFG – Tyler James Young Apr 06 '15 at 19:42
  • @TylerJamesYoung LOL; "just say no"! – Kaz Apr 06 '15 at 19:50

19 Answers19

90

The important thing is not so much which direction is right, but that you make it visually clear which direction is 'on'. This can be done by lighting up an LED, by an icon on the display, changing colors, etc. It just needs to be very clear what state the machine is and that this button will toggle the state. This is how single direction switches (buttons) work; it's obvious what a power button will do because I can tell that the device is currently on. And if I can't tell if it's on or off, your standard button direction really doesn't do me a damn bit of good.

Generally I would recommend Up and Right as 'on' but as I've already said, if the switch isn't labeled, any direction has the potential for frustration. I base the direction purely on common experience, but most physical and digital switches seem to prefer right for on:

enter image description here

Why can't you trust one direction? Flip a light switch somewhere near you. Very often these light switches are set up so that down on one light means "on" while it means "off" on another light in the same house; this is especially common when two switches affect the same light, so you can't depend on the physical direction. But this isn't nearly so much as a problem because there is always a visual cue as to which state the light is in. If the light is on... well, you can see the light. If your switch doesn't have that nice visual affordance you'll need to make your own affordance.

JYelton
  • 103
  • 4
Ben Brocka
  • 40,865
  • 10
  • 112
  • 183
  • 18
    It's worth noting that "down = on" light switches aren't necessarily the result of the electrician being sloppy. With a conventional mechanical 3-way switch the on/off direction changes every time you use switch A to turn the light on and B to turn it off. – Dan Is Fiddling By Firelight Mar 12 '13 at 14:13
  • 4
    @DanNeely I point that out, sort of; another way that can happen is two lightswitches in one room affecting the same light. – Ben Brocka Mar 12 '13 at 14:16
  • 18
    There's a visual cue unless you need to change a burned-out light bulb. Then you can't really tell if it's on or off without a voltage tester, or you just screw in a new light bulb and see what happens :) – Danko Durbić Mar 12 '13 at 15:05
  • 1
    @DankoDurbić I've never had a problem screwing in bulbs with the lights on; though they do make lovely switches which light up when they're on, giving proper visual affordance without the actual light. – Ben Brocka Mar 12 '13 at 15:12
  • 7
    The light switches I have have an LED indicating the state (lit when switched off). The benefit is that I can see where the switches are when in dark. (Wait... on stands for off?) – Alvin Wong Mar 12 '13 at 15:16
  • 2
    And then there's the water heater in my basement. (a) The breakers don't make it clear which direction is off or on; but worse (b) the water heater gives absolutely no indication of whether it is off or on. As a result, I burned out an element because I drained the water heater while it was on, because I'd turned it off with a mislabeled breaker. Usability: It's not just for web sites. – Kyralessa Mar 12 '13 at 17:38
  • @BenBrocka What if it's not a light switch? – JohnGB Mar 12 '13 at 20:38
  • 12
    I find the example in the image unclear. Is the text indicating what state the switch is currently in, or the state that it will be in if I move the slider? Having the text inscribed on the slider itself is bad design. If sliding the switch to the right turns it on, then that indicator should be to the right of the part that is moved, not to the left of it. – JYelton Mar 12 '13 at 23:04
  • @JYelton: unless you're color blind, IMO it's pretty clear that the On state has a more vibrant color than the Off state. – Lie Ryan Mar 13 '13 at 05:03
  • @JYelton: Not to mention that if the switch controls some other UI component, you may be able to tell what's going on in the same way an unlabeled light switch still gives feedback in the form of the light turning on or off. Of course, for situations where the switch doesn't control a UI component (say, controlling WiFi on a smartphone), you've still got a good point... – Kevin Mar 13 '13 at 05:19
  • Even in case of light switches the overall action is still to apply pressure like a push button , its easier to push lower part than the upper part, to flip it. This is different from other switches where you have to get hold of it and then move it. – user117 Mar 13 '13 at 06:42
  • Is it me, or is the lighting "off" on those switches? Where is the light-source, and what is the form of the switch when seen from the side? It seems the labels can't slide under the checkered surface. – Chris Wesseling Mar 13 '13 at 13:05
  • I am used to switches like this: http://www.theinternetswitchstore.co.uk/images/product/600/XDQ20NWS.jpg

    If turning it on meant pushing in the top half, you wouldn't be able to see the writing/visual prompt on the switch, so it makes sense to press the bottom and reveal the top (most people's line of sight will be ABOVE the switch if it's wall-mounted).

     – deed02392 Mar 13 '13 at 17:55
  • I've found that on most "cool-looking" on/off widgets on mobile devices, some Windows programs, and some web pages, I can't tell whether the switch is on or off, or what on means, when such widgets have ambiguous labels, with no additional description. – Mark Stewart Jan 22 '15 at 14:09
  • Yeah, there should be a sensorially accessible engraving or a tactorially feelable light switch shape which would make things clear. Still easier though if there is a standard: it would make it faster and would allow the user not to have to remember every switch configuration separately. – Jack Maddington Dec 11 '15 at 13:25
  • Nevertheless, since most people are right-handed (which is the reason the right, part of hot-cold water taps defaults to cold, almost everywhere, everywhere I've been and would expect it to be the same everywhere (perhaps with the exception of some taps for special testing purposes, not sure what their purpose would be). – Jack Maddington Dec 11 '15 at 13:28
  • @MarkStewart, please post your invaluable examples. – Jack Maddington Dec 11 '15 at 13:30
  • @MarkStewart, yeah, Windows has been a bit clunky in former times, but my opinion is that it will come up with intelligent interoperable solutions as it has in the clear recent path. What this intelligence will look like will be interesting to see. – Jack Maddington Dec 11 '15 at 13:34
  • Thanks @DanNeely, that way it makes sense to prop the lever down, just before it props back up, as it takes less effort. – Jack Maddington Dec 11 '15 at 14:53
  • All those combinations without a standards can make things plain dangerable, and can even be ususable. – Jack Maddington Dec 11 '15 at 14:58
  • @Yelton, maybe the switch should be in the middle when off, with the text on both sides when on, with the light turning on whenever it is moved to either of these sides. – Jack Maddington Dec 11 '15 at 15:08
84

It appears to be dependent on country or region, as Wikpedia states in the article Light Switch:

Up or down

The direction which represents "on" also varies by country. In the USA and Canada and Mexico and the rest of North America, it is usual for the "on" position of a toggle switch to be "up", whereas in many other countries such as the UK, Ireland, Australia, and in New Zealand it is "down." ... In countries prone to earthquakes, such as Japan, most switches are positioned sideways to prevent the switch from inadvertently being turned on or off by falling objects.

So there is really no direction globally true, if we believe in Wikipedia. Up or down or sideways differs by region. From what I learned in school (but I can't find a reference too now) is that vertical motion elecric switches use a metaphor for determin which is which. As an eyelid opens, the eye can see, meaning the switch is on. When the eyelid closes, we can't see, meaning the switch is off.

This metaphor works for electric switches, light switches and (modern) fuses.

enter image description here enter image description here enter image description here

But there is a problem. Electronics switches seems to have the metaphor crushed if you do image searches. You find all sorts of switches where Up is off and Down is on.

enter image description here enter image description here

Or at least there are no consistency in direction

enter image description here enter image description here

To conclude - be careful relying only on direction. An informative text or indicator light would certainly make it easier for your users.

Community
  • 1
Benny Skogberg
  • 54,996
  • 22
  • 140
  • 241
  • 24
    Note that of your examples, only the ones with ON and OFF written on it can't trivially be mounted with either short side up. I think that's a big reason why the on state is represented as a simple vertical bar rather than a more elaborate 1: you can mount it either side up and it will look the same to the casual observer. – user Mar 12 '13 at 15:14
  • 19
    Actually, it's just UK and Ireland that are reversed. In Australia and New Zealand, the light switches are the same as ours, it's the countries that are upside-down. – Kevin Mar 12 '13 at 15:58
  • 6
    @MichaelKjörling true, but I often have trouble remembering which symbol represents "on", especially in the heat of troubleshooting when something in "|" acts "off", and I start thinking things like maybe "O" = "open"? etc. – joshuahedlund Mar 12 '13 at 17:56
  • 4
    Those I/O switches aren't inconsistent, one of them is just up-side-down :P – naught101 Mar 13 '13 at 04:13
  • 9
    I/O is the most miserable way to label a switch that I can imagine. Yes, I understand the binary interpretation now. But there have been so many times when I have been trying to get an engine to start when the logical interpretation has seemed to be: "I" means that something is closed, therefore nothing can pass through it, therefore "I" means off. "O" means that something is open, therefore things can pass through it, therefore "O" means on. This leads to frustration. – jvangeld Mar 13 '13 at 22:12
  • 1
    Please note that if you have 2 switches for the same function, you are not guaranteed to have the button state and power state always be consistent. For example: if you have a light switch at each end of a corridor, someone can turn on the light at switch A, and turn it off at switch B, leading to the light being in the same state as before, but both buttons being in the opposite state from before they were pressed. – Nzall Aug 07 '14 at 10:01
  • Was wondering what "parts let us call 'em", of Japan, have their switches' directionaliy set, whether they'd be moving towards a more specific direction, and if so, at what pace if there isn't one standard for it in place already (what if for instance one part of the switch was uncomfortable to touch or sth, maybe it had a protruding light bulb a blind person could easily break if coming into contant with the switch at an extensive speed. – Jack Maddington Dec 11 '15 at 13:13
  • From a side which had not been standardized. – Jack Maddington Dec 11 '15 at 13:15
  • @KateKerhofs, this is what I originally had, parallelly-aligned versus cross-aligned. I would pick, from a purely psychological perspective... well, I couldn't decide. So, suggest some other way. Perhaps two equally functionin switches on each side with two lights, and then the reverse, with two walls. And then with the 4 lights and 8 switches there would still be plenty to decide. Some "derangements" set free would make a king's palpable toy switch. – Jack Maddington Dec 11 '15 at 14:47
  • @NateKerkhofs, yes but this is not the lovelyness of discreteness, or is it perhaps a live of sequences; maybe both, so, whatsoever, the world goes on and on and on. Don't forget we're trying our best to solve a "problem". – Jack Maddington Dec 11 '15 at 15:02
  • 2
    @jackmaddington I know that. The problem is "Which direction should be the on position of a switch?" And I am just trying to say that there are situations where the position of the switch and the state of the device the switch controls are not tightly coupled. In such a case, the direction of the on position of the switch can change depending on the sequence of actions performed on the system as a whole. My second floor hallway light has a switch on both floors to turn it on or off and the position of the switches is not linked. The bottom switch can be in either position with the light on. – Nzall Dec 12 '15 at 20:52
  • @Kevin from what I observe, at least also in Kazakhstan and Mauritius light switches are down-to-turn-on. – Ruslan Feb 22 '23 at 19:36
47

This is one of the first designs of a vertically-mounted electric switch:

Knife-blade switch

It was presumably designed this way to afford an in-built failsafe: it requires physical effort to close ("turn on") by overcoming gravity, which will otherwise open the circuit.

EDIT: At least in the US, electrical codes (see National Electrical Code paragraph 404.6 - 404.7) still require this type of switch to be mounted so that gravity will not tend to close the circuit. This convention extends to modern general-use switches covered under the scope of electrical codes (i.e within buildings) which must be installed with the "on" position up.

Modern switches almost certainly inherit their behaviour from this prototype.

EDIT: From a UI perspective, keep to the golden rule and don't break expectations.

It's perhaps worth noting that while it may appear that horizontally-mounted switches shown in Ben's response follow reading direction, they don't seem to:

iOS Arabic

(Images acquired from http://sketchup.google.com/3dwarehouse/details?mid=a34d66a99817b368978efdfd2ccd8bdb and http://www.iphone-tips-and-advice.com/languages-on-iphone.html )

Theodore
  • 103
  • 4
msanford
  • 5,995
  • 2
  • 33
  • 52
  • 3
    It seems here, even alignment doesn't follow reading direction on iPhone. Not sure if that is good or not. – André Mar 12 '13 at 15:29
  • 1
    @André Indeed! I was a little surprised, to be honest. You'll also note that the entire UI is backwards: the icons should be on the right, with the disclosure chevron on the left and pointed to the left. But of an i18n fail in my opinion, as all they did was RTL the text field. – msanford Mar 12 '13 at 15:46
  • 2
    @msanford While it does look a little strange seeing the Arabic text left-aligned, this makes sense from an international point of view. Take for example a Japanese manga, even when they are translated to English, you still read them right-to-left. – Adam-E Mar 12 '13 at 21:23
  • 2
    however in movies they seem to prefer that down = on for a more dramatic scene... – ratchet freak Mar 14 '13 at 12:20
  • I don't see why a switch should be off if not otherwise intended. Sure, elecrics are dangerous and nothing should flow somewhere just because the gravity turned the lever. But what's avout stuff that's supposed to be on? The engine lever in the airplane or the machine that keeps a patient breething. Shouldn't some things be activated if not otherwise intendet? Whatever position is more dangerous should be protected against gravity (or in earthquake heavy regions by falling objects), or doesn't it? – BlueWizard Dec 12 '15 at 21:50
  • @BlueWizard for things you want to stay on, use a Molly-Guard. –  Jun 14 '17 at 14:03
  • The iPhone screenshot in Arabic is not properly formatted. iOS added RTL support in 2009 so that must be a very old screenshot, or settings was not displayed correctly. Here's a screenshot from a more recent iOS which is properly formatted: https://i.imgur.com/6n3GG68.jpg but doesn't display any switches. Here's a recent screenshot from Android which does feature switches: https://i.imgur.com/5BTebba.png – Basically the whole UI and everything is mirrored. That's a good rule of thumb. Mirror everything for Arabic, except: digit order, media playback, right-handiness. – Liggliluff Aug 02 '20 at 21:21
  • It is an old screenshot, look at the date of the post. Thanks for the contemporary details. It always annoyed me the the text was left-justified. – msanford Aug 03 '20 at 02:50
15

Sadly I don't have research material but some real world examples which uses the top/bottom direction.

Example 1 Fuses

I have found an interesting post and a manual about switches in cockpits of airplanes.

Boeing manual

F16 switches

Charles Wesley
  • 12,218
  • 9
  • 48
  • 74
sysscore
  • 1,521
  • 9
  • 13
  • 2
    Eurocopter EC-135 has "forwards = on" for overhead switches, and "up = on" for ones on vertical panels - like http://www.osimco.de/images/Switch_Panel/Switch%20Unit%20EC135.JPG – OJW Mar 12 '13 at 18:39
  • For those interested, here is a complete reference for the switches in the F16C Block 50/52. Most of the switches are up or forwards in their on positions. – Mankarse Mar 12 '13 at 22:50
  • I'd hope that whichever way switches in an aircraft cockpit are "on" or "off", they are all the same. I doubt I'd care much about whether the "on" state is switch in "up" or "down" position, but it had better be the same as the switch next to it! – user Mar 13 '13 at 09:37
  • +1 for giving links to the refs written in blood, sweat and tears! – Deer Hunter Mar 13 '13 at 11:23
11

If safety is a concern - make sure gravity deactivates the switch.

enter image description here

This kind of switch is designed to never be activated by accident.

Michael Celey
  • 718
  • 1
  • 7
  • 20
Izzy
  • 211
  • 1
  • 3
  • 5
    Unless for safety's sake it's better to make sure gravity doesn't deactivate the switch, which can be the case with light switches, etc. – taswyn Mar 12 '13 at 17:31
  • Especially when it reads "DO NOT PRESS!" – Lourens Mar 13 '13 at 06:28
  • 1
    Gravity does absolutely nothing for switches like these. Or for most switches, for that matter. The springs in these switches require much more force than gravity exerts on them. – André Mar 13 '13 at 09:26
  • 3
    @André, the gravity discussion doesn't refer to gravity force on the switch alone, but aided by a hand or other object. It's easier to push your hand down on the switch than pushing it up on the switch. Likewise, objects falling on the switch are affected by gravity. – Zano Mar 13 '13 at 09:50
  • @Zano ok, that is a fair point. Thanks for the clarification. I think this clarification belongs in the answer. – André Mar 13 '13 at 10:00
  • 7
    if the button reads "do not press", why is the button there in the first place? – David K. Mar 13 '13 at 13:46
  • 1
    @DKOATED SELF DESTRUCT - Do not press unless you really REALLY mean it – user Mar 13 '13 at 14:55
  • 4
    @MichaelKjörling why in the world would i want to destroy a button? ;) – David K. Mar 13 '13 at 15:29
  • @DKOATED That's a good one :D (Maybe there's no other way to make the device stop humming?) – user Mar 13 '13 at 15:30
9

I live in the UK and so am used to UK design conventions. Many UK households have at least one XOR configuration lighting circuits for the communal areas - two switches leading to the scenario where, depending which is used, both can be in the traditional 'off' position. I'd say that it's therefore not really a big mental challenge to have a mixture of switch configurations, although it can be mildly frustrating at times.

Single-position lighting circuit wall switches — and BS 1363 wall sockets — use a downward motion to make the circuit live. This means that, for power sockets, a visible red indicator on the top-facing side of the socket's switch shows the socket is live at all times. (It's not called the "world's safest mains plug and socket" standard for no reason.) Some sockets and extension cables also have illuminated indicators either built into the switch or on the socket for further guidance.

All UK wall switches have a pretty stiff movement allowing you to 'feel' the position of the switch with a moderate amount of pressure without actually triggering it - whereas some much larger European 'panel' switches can be switched with little more than an unintended arm contact when walking by.

Always-vertically mounted RCDs use the standardised up=live, down=neutral paradigm for a few obvious reasons including failing safe and ease of visibility. The context of their usage is fundamentally different from domestic sockets so this is fine by me.

Example: in daily life, I use IKEA standing lamps which use the European up=live, down=neutral (and are gently annoying for this reason) and audio equipment which uses horizontally mounted switches. Each one is different, probably because of their countries of origin. Some other lamps have a switch which passes through the actual neck of the lamp, so you can put two fingers either side and just push in the appropriate direction. As you can rotate the lamp however you like it this solves the problem ;-) Or you can just fit clappers everywhere!

I don't consider French sockets to always follow the up=live, down=neutral paradigm; I think this depends to an extent on the electrician and of course whether they're to control an XOR circuit.

If I was designing a physical product and had the opportunity to be this obsessive over the minutiae (such as which way a switch tilts when it's engaged), electrical safety standards providing I would see about offering a modular switch - for example, protected by a screw-in plate and not easily modifiable without a screwdriver - which allowed the user to modify the switch's operation to his preference.

As it's tangentially related: for any software design, I always recommend eschewing a skeueomorphic pseudorealistic convention and always prefer when there's a nice plain square button, clearly labelled Off or On in the appropriate i18n locale. That, and perhaps colour the switch as red when off and green/blue when on. Although those colour codes aren't universally used they're arguably instantly recognisable given widespread use.

...And now, ignoring everything I've said above: if I had the ultimate choice, I'd say mount your switch so the top-facing half is flush to the shell when the device is powered on (requiring an upward flick to power on) — and cover with a small, clear, hinged door to prevent accidental disabling. Nothing worse than fumbling blind around the rear of a unit only to accidentally power it off.

Chris Woods
  • 373
  • 2
  • 8
5

There isn't a global preferred direction. And there's no psychological reason for a preferred direction. It will depend on the default in people's home culture.

PhillipW
  • 11,290
  • 2
  • 33
  • 55
  • 3
    So what do you do when you're building a product to ship globally? – JohnGB Mar 12 '13 at 12:00
  • @JohnGB - show the on state with a light? – ChrisF Mar 12 '13 at 12:10
  • 3
    @ChrisF Yes, but then you still have to choose which is the on direction. Indicators aside the question is about that choice. – JohnGB Mar 12 '13 at 12:38
  • @JohnGB - with a status light the direction doesn't matter though. – ChrisF Mar 12 '13 at 12:49
  • 2
    @ChrisF I disagree, as flipping a switch is often a reflex, not a thought out action. – JohnGB Mar 12 '13 at 12:57
  • @JohnGB But, as Benny points out, the 'standard' depends on the region. In Australia, the vast majority of switches here are 'on' when in the 'down'/'towards me' direction - but there's always a couple of imported goods that go in the opposite direction. In the end, it's necessary to become familiar with a specific device, and not expect different devices to follow the same conventions globally (with regard to switch direction) - it's too late for that. Perhaps if you had a specific region you were targeting. – Bob Mar 12 '13 at 13:55
  • If it matters choose a different kind of switch: a radio button style push switch does have a tactile 'logic' to it: pushed in is on / pushed out is off - and you don't even need to look at it to see it's status. – PhillipW Mar 12 '13 at 22:25
  • 2
    @JohnGB, that's expected of course, since south of the equator, up is down. – Zano Mar 14 '13 at 09:15
  • 3
    Another rationale might be physicality. Given our hands are usually hanging down at our sides, and light switches are usually mounted waist height or higher, the natural motion of the arm is upwards and away from the body. For a right-handed person, that's up and right (though left if you're entering a room and the light switch is inside, next to the door, which is prolly why most are vertical -- works coming and going). So entering a room, that's how you want to turn ON the light. Again, making it harder to switch off again. – uliwitness Mar 17 '13 at 12:53
5

Depending on the application safety issues should be considered as well as intuition. I would say when you mount a switch vertically it is usually more sensible to have the “on” state on top. Because vertically mounted switches are more likely to be switched down accidentally by taking your hands down, grazing, things falling down or similar. Usually less harm is done if something is accidentally switched off rather than on, especially if you think about rotating machines, etc.

If you think about the worst case, someone injured lying on the floor trying to reach a switch he is more likely to reach the switch to push it down or someone in a rush, it is easier to switch down because you usually take your arm down after you pushed the switch. Also small children might not be able to switch upwards which is usually the prefered situation.

I would orient the switch the way it is easier to reach for the user in whatever indisposable situation you might catch him—even if it’s only inconvenience.

However in some situation you might want to have the user being aware of his decision; An example from another area: gas cylinders have different handed threads for combustible gases than for oxygen, nitrogen, etc.

Of course some of the examples are a bit extreme and you might even want to use a different switch design in these scenarios; nevertheless it shows that some orientations might be more sensible than others.

Velocipede Berserker
  • 51
  • 1
4

For vertically oriented switches, I (as the user) would first expect Up to be On and Down to be Off. This is entirely based on my usage of North American light switches over the course of my life. There is probably some psychology associated with this, in particular, that things "up" have more potential energy than things that are "down," certainly true for physical objects, and the metaphor used when we talk about moods. If you're feeling "up" then you're full of energy, if you're feeling "down", you're low on energy.

For horizontally oriented switches, I initially thought "right is on," but then I thought about when I drive German or American cars where the headlight switch is operated by the left hand. Off is when the control points up, and you rotate it clockwise (top point moving right) to turn on the headlights. Every single time I do this, it feels backwards to me. This would the expected motion if the control were mounted for my right hand, but not my left.

I also thought about a bathroom I have in my house with a pair of horizontally oriented switches. They switch left to turn on, right to turn off, which makes sense to me, but only because the switch is left of the door (when facing the switch). If the switch were on the right side of the door (as it is in another room in the house), I expect the opposite to be true.

These two mini-case studies have drawn me to the conclusion that away is on and towards is off.

John Franklin
  • 141
  • 3
  • Not sure why they moved the high-beam switch off the floor. That was the best place for it. The button next to it that changed the radio station was marvelous also. Now they have tried putting a dozen buttons on the steering wheel, but I can't keep them straight. –  Jun 14 '17 at 14:23
  • No kidding. I bought a new car about a year ago. The steering wheel has two Android-style back buttons (⮌). One is for the trip computer, the other is for the radio/nav/apps suite. From a UX perspective, I can't stand having two identical buttons in the same control suite. – John Franklin Sep 18 '17 at 00:35
3

Not an answer, but a feeling. I do think that if the switch is horizontal, the on needs to be on the right. Right feels like "forward", or "active" to me, perhaps because the flow of time is usually depicted in that direction.

However, most switches seem to be mounted vertically. For that case, I get the feeling that it depend on use, but that there isn't a very strong preference either way. I do think that there might be research on it, in the context of the design of cockpits or control boards for plants. But based on instinct again: in context where you have to be able to quickly turn off, the on direction is up. Turing off a complete row of switches in a single downswipe with your hands seems easier downwards than upwards.

André
  • 6,574
  • 21
  • 39
  • 3
    I agree, but wonder, would that feeling be backwards in right-to-left cultures? – Mark D Mar 12 '13 at 14:16
  • Might be; there have already been discussions on that (which I can't seem to find right now) – André Mar 12 '13 at 14:25
  • @MarkD Never mind those who natively use bidirectional writing systems. – user Mar 12 '13 at 15:18
  • I suspect it has to do with handedness - if on is to the right, with a right-handed user, that's a pulling action, and off is a pushing action. This might change depending on physical location and access. – penguat Mar 12 '13 at 15:40
  • @MichaelKjörling Why don't we hear about boustrophedon writing more often? (or read about it for that matter) It seems to solve the RTL vs LTR debate decisively. –  Jun 14 '17 at 14:27
3

Although I couldn't find any valid research to back me up on this, I have the strongest feeling that ON should be UP. There is just a strong association between things that are up, going, running, working, living, moving, vital and so on. This is in contrast to things that are down, stationary, or shut. These associations are also expressed in common phrases, like systems being ON-line, or a computer being shut-down.

Perhaps this is rooted in the experience of our daily physical lives; lower things are closer to the ground and have little or nowhere to move. In contrast, higher things tend to move, often towards the ground. If this is not the case, it often takes some energy or process to maintain it in its high position. This is all very abstract, but its through these experiences that we have (hopefully) universal associations.

Therefor, if I had only intuition to guess which way is ON, I would guess it's UP. I hope you understand why.

Dvir Adler
  • 5,008
  • 2
  • 28
  • 44
  • 4
    For every metaphor where 'on' maps to 'up,' there is another where 'on' maps to 'down,' and it is also highly culturally dependent. Never make assumptions that your own cultural metaphors universally apply. As some examples: when water flows, it flows down. Reading is in many cultures a top to bottom motion. The top can be seen as the "ready" state, while the bottom is the activated one. Over hand throwing, waving onward, etc are often seen as forward arm motions, which would coincide with a top to bottom switch motion. – taswyn Mar 12 '13 at 17:31
3

In which direction is the switch most likely to spend its life? Which directions are more costly (or dangerous) to have the switch hold?

If it is most likely to be on, then it should be installed in a manner where the mass of the toggle would likely keep it in the on position. If it is off most of it's life, then the mass of the toggle should keep it in the off position.

Exceptions to this rule exist, for example, circuit breakers tend to have the toggle pointing to the center of the panel to indicate connectivity with the mains. Unless you control the entire layout and design, odds are you are not going to be able to leverage additional information in the switch position in this manner.

If the switch is mean to convey additional information, one generally wires in a lamp with the switch; however, this adds additional complexity (and maintenance). The additional complexity is that now one must consider lamp color, and such a color might not have a standard. To point, in most laymen's eyes, green means on; however, in the electrical power industry, red typically means on and green means off (because a live wire will kill you, and green generally means safe to work on).

So if there are no additional cultural norms to fall back on, wire it such that it spends its normal life with the switch down, which tends to work well as falling objects hit switches occasionally, and are less likely to upset the system. Yes, in the rare instance a rising object might hit the switch; however, gravity is ever present, so you're playing with the odds.

This partially explains why computer power switches then to be "on -> down" and light switches tend to be "off -> down"; because, turning off a computer without system initiated shutdown tends to run the risk of corrupting the system's data.

Once, I ran into a switch that was mounted with normal = up. Somehow it kept getting pushed down, which caused an upset, until someone actually mounted a small box over the switch. Likewise we have a mis-mounted light switch that controls an entire circuit in my childhood home. Eventually we attached a little plastic stop to prevent the switch from being turned off, the ultimate in UI failure.

Edwin Buck
  • 131
  • 2
2

The answer is that whichever way you do it, it needs to be clearly labeled. There is no "which way".

The BUDGET answer is to look at the most used applications on the platform for which you are programming and piggy back on their UX research and use the same or similar functionality, so there is an air of familiarity and commonality across the user's device.

All if this is running under the assumption your killer app is not ACTUALLY about amazing on/off switches. In which case, the debate will rage on. :-)

Mak
  • 41
  • 2
  • Now we have "fidget cubes" with buttons and switched on them that do nothing. The day has arrived! –  Jun 14 '17 at 14:25
2

Goes back to the rule of righty tighty lefty loosey — Switches are built (and drawn in electronic diagrams) as a line that goes from left to right then bends up and then there is a gap and then back on the level (y-axis) before the bend there is another straight line... So whenever the line that bent up is "pushed" down (or turned clockwise (righty tighty)) the circuit is completed. Clockwise = on, counter-clockwise = off.

Toggle-switch — up/right = on
Push-switch — down = on

enter image description here

Albert Renshaw
  • 347
  • 1
  • 7
0

You would also consider if the switch has another meaning than the boolean value on/off.

  • Window blinds up/down
  • Audio left/right channel
  • Tollgate up/down
  • Railway track change left/right

All these have a intuitive position i would say.

Viktor Mellgren
  • 565
  • 1
  • 3
  • 14
0

As there is no global standard, and we still want to give the global answer, you just need to mark it to make it clear. Otherwise you need to check it by trial and error.

But in most cases you can tell if the light or whatever else is on or off, so the answer is: the opposite one switches it to the other state.

I think it is more clear for gauge switches, though, where turning right is for on, and left for off. So maybe using these is an option.

Dominik Oslizlo
  • 13,849
  • 2
  • 30
  • 57
  • 2
    Considering that you do bring up the point of globality, I'd wonder if it is so "clear" that turning right turns on and left turns off in a right-to-left culture. – user Mar 13 '13 at 15:10
  • Not 100% sure, forgot to mention that. +1 for your comment. – Dominik Oslizlo Mar 13 '13 at 16:15
-1

The better way refers in term of user,

as it was used like if the nob/switch is in the up state |\ its off, and while its down |/ its is on,

Now while user's dealing with emergency they found this was wrong, so in Emergency switches/ its reversed as its easy to switch it off. in |/ this direction using hand or any other thing.

So, user perception from past makes it easy with practice.

|\ off

|/ on

Practically this new theory is better.(Safety wise)

|/ off

|\ on

I hope safety makes more sense now a day!

MarmiK
  • 109
  • 4
  • In an emergency, people will not think, they will just do whatever they are used to doing. If they are used to off=down, that is exactly what they will reflexively do. –  Jun 14 '17 at 14:31
-1

It all depends on the usage of the switch. Airplanes and choppers use the switch on top to be ON, this ensures that someone by mistake does not switch it on while sitting down. This is one of the major causes of mishaps.

Usability wise, you need to look at the scenario it is being used.

If it is a 2 way switch you need to consider what orientation you are going to keep the switch in, horizontal or vertical.

If there are just too many switches in the place already, then you should follow the rules that have already been set, dont change something that works.

Gravity is another point to consider. What would you like your switch to do when it is combined with gravitation forces. Switches that trip on leakage should not only be sensitive, but should use gravity as a force to switch off, and may be couple other switches with it to switch many more together.

Siddharth
  • 513
  • 1
  • 3
  • 14
-1

I think most of the people over here already mentioned about the different types of switches and probably the up/down switch is the best for representing ON/OFF. In addition to the use of Up/Down button I would like to mention one very important point which I feel should be pointed out.

Why in MCBs and all other Up/Down high current switches, ON situation is always kept in Up state Down?

enter image description here enter image description here

The reason being that for example if something falls on the switch due to gravity or even by mistake then the switch should not ON the system because it may create some dangerous situation.

Community
  • 1
Ankit
  • 896
  • 1
  • 9
  • 15