Take the 2-minute tour ×
Cross Validated is a question and answer site for people interested in statistics, machine learning, data analysis, data mining, and data visualization. It's 100% free, no registration required.

It seems like data visualization experts generally disapprove of gauge charts (see here: What do you call a chart that looks like a half pie chart with a needle indicating a percentage?). The primary reason is that a gauge chart has a low data-to-ink ratio.

Ever since I was exposed to these concepts (a few Tufte books), I generally agreed with them, but today it made me wonder: if gauges are so inefficient at communicating information, then why do cars/boats/planes have lots of gauges on their dashboards? And does the answer to that question have some kind of relevance for creating software dashboards for large enterprises?

Edited to include some additional information I found:

I found a term, "glass cockpit", that refers to an airplance cockpit that has its mechanical gauges replaced with LCD screens. This gives credibility to the "convention" argument put forth by Wayne.

http://en.wikipedia.org/wiki/Glass_cockpit

Here's an iPad app that gives a dashboard-like readout of your car's telemetry, with no gauges to be seen.

http://itunes.apple.com/us/app/dashcommand-obd-ii-gauge-dashboards/id321293183?mt=8

I also found a gross example of digital gauges for cars (viewer discretion is advised).

http://www.chetcodigital.com/index-Automotive.htm

share|improve this question
4  
If pie charts are bad, why do we have pies? ... Oh. ;-) –  cardinal Jun 14 '12 at 20:42
5  
Tufte's data-to-ink ratio does not appear to be relevant in this problem, but Cleveland's studies of the speed and accuracy with which people interpret statistical graphics have a strong bearing on it. These studies suggested that people do not compare angles to each other as quickly or accurately as they compare lengths or parallel positions relative to a common baseline. One big problem with angles is that the comparison may depend on how the angles are oriented; that problem doesn't appear to be an issue with a dynamic gauge. So perhaps circular gauges are close to optimal at what they do. –  whuber Jun 14 '12 at 20:50

3 Answers 3

up vote 12 down vote accepted

A (real) dashboard gauge needs to be: 1) physical, and 2) read quickly under circumstances that disturb concentration. In that sense, you want a low data-to-area ratio. Not to mention that when physical gauges were invented, digital (numeric) displays didn't exist so there was no real choice.

A software dashboard is not physical, and is not generally looked at in a pitching, moving vehicle with other vehicles whirring around it. So the effect of imitating a physical device doesn't buy you much.

EDIT: I'd also add that a physical dashboard only has a couple of key attributes to get across to you at (literally) a glance. A corporate dashboard needs to make a lot more detail visible, though of course things should be drawn/coded/organized in a way to also give a quick status.

That's part of the Tufte philosophy of dense detail in presentations that allow a broad view but also allow you to drill down. You car's dashboard doesn't let you drill down, basically because there's no need to.

share|improve this answer
    
"when physical gauges were invented...there was no real choice" But I own a car built in 1999 that has more gauges than it does digital displays. Are you saying we still use gauges just out of convention? I do agree with your point on reading quickly, but would that not sometimes be a desirable quality in a software dashboard? (I'm playing devil's advocate here.) –  mehaase Jun 14 '12 at 20:29
2  
(+1) Re: your last paragraph: Except of course for software systems meant to be used in-vehicle during operation, as found on many police vehicles and other industrial vehicles like in construction and the military. Interestingly enough, those hardware/software solutions often do look like dashboards. Touch screens, big buttons and judicious use of (high contrast) color play a very important role in usability and safety. These interfaces tend to look like anything but your typical GUI. –  cardinal Jun 14 '12 at 20:39
1  
@mehaase: "Reading quickly" is quite different when you literally have a couple of seconds to glance down from the road versus when you're sitting at a desk in front of a monitor. I guess some financial trader will claim that they only have a couple of seconds to make a buy/sell decision or something, but realistically most corporate "dashboards" aren't used in such conditions. –  Wayne Jun 14 '12 at 21:44
    
My previous comment relates to the paragraph starting "A software dashboard is not physical...". I should have been mindful of future edits and not made references to paragraph placement. :) –  cardinal Jun 15 '12 at 13:51

In supplement to Wayne's fine answer, Robert Kosara has a recent post on his Eager Eye's blog about the very topic, Data Display vs. Data Visualization. In addition to as Wayne mentioned the goals of real-time visualization vs. more static displays might call for differences, he also mentions that gauges aren't very good for displaying multiple values. This is summed up nicely in his comment,

What you want to know is, how fast am I going right now? How much gas do I have left? What your speed was five minutes ago, or how much gas you had in your tank three hours ago, matters little.

So here is any obvious contrast between the goals of data visualization versus car-gauges, we pretty much always want to see multiple data values! And circular car-gauges are certainly a poor tool to do that. Sometimes we don't want to see multiple values though (a few circumstances are given in this question on the GIS site, What is the point of standard symbology?). And so we might expect other rules to which we apply the data visualization techniques in such circumstances. The GIS post I mention uses very flashy symbols/icons for point patterns that attempt to incapsulate the nature of the event (and sometimes visualization techniques like blinking dots to focus attention).

What I find interesting is that the work of Cleveland on comparing angles is still pertinent to car gauges though, and hence we still might expect a linear scale for a car gauge to work better than the circular display. So I suspect there might be more historical context as to why circular gauges were chosen (they are compact?), and it certainly may be this historical inertia as to why they are popular.


This much be a popular topic in the thralls lately, as the Visual.ly blog just came out with a post on the topic as well, Speedometer Design: Why It Works. In there they give credence to some of the things gung mentions in his post that I am somewhat critical about in the comments, in particular how we develop a gestalt for identifying locations around the circular display.

visually relative areas of speed circular display

I think I'm partially coming around to this notion. A circular display provides more visual distinction between general areas than does a linear one. For a general example, it is easier to quickly tell the difference between a needle pointing to 3 o'clock and a needle pointing to 12 o'oclock than it is to tell the difference between 15 and 12 on a linear scale.

I'm still not totally convinced though, and I say rubbish to the notion that acceleration is easier to distinguish on a circular scale (or even if it is information we need the dashboard to inform us about anyway) that the visual.ly blog post mentions. Just my opinion though, I'm not sure any of us have been citing directly pertinent experimental results on human perception. Cleveland's is a start, but not likely to give an entirely satisfactory answer to these particular circumstances.

That being said the multiple data values are still the main crux of the argument, circular displays aren't good for multiple data values.

share|improve this answer
1  
+1 This seems to be on the right track. I think Cleveland's principles can be pushed further. Recall that he asked subjects to compare quantities being displayed on charts. Comparison with angular displays was neither as quick nor accurate as comparison of position along a marked axis. In effect, a dial gauge gives a very clear position along a curved axis. As such it shares some of the good properties of graphics that use position to represent quantities. It also suffers a bit, too: we may have a harder time distinguishing speeds near the top of a gauge compared to speeds at the sides. –  whuber Jul 16 '12 at 15:21
2  
Historically, the reason for a dial gauge is clear: Maxwell's equations suggest the most basic way to translate an electric current into a physical motion is by placing a small coil of wire inside a static magnetic field (e.g., wrap the wire around a magnet). Not only simple, but cheap, reliable, stable, nearly linear over a long range, and easily calibrated. –  whuber Jul 16 '12 at 15:26

There are great answers here. I also like @whuber's comment, especially "[o]ne big problem with angles is that the comparison may depend on how the angles are oriented". Let me throw out one quick note here: it's worth remembering that all car speedometers are oriented in the same way. (What I mean is they all run clockwise, and the physical location of the endpoints are in approximately the same position at the bottom.) In line with @Wayne's point about having to quickly glance at the gauges and then back to the busy road and still have extracted the relevant information, note that to encode magnitude via relative distance (a la Cleveland's dotplots, which I do like a lot), you have to encode the position of the dot and also the positions of both endpoints. With a gauge, you need only notice the angle of the needle, which you can still 'see' in your mind even a few seconds later while looking at the road again. Realize that you get very used to looking at your car's speedometer. Thus, interpreting this angle can become effortless. Moreover, because all car gauges are oriented the same way, it is easy to adapt to an otherwise unfamiliar car, although because the top speed listed can vary (as @cardinal notes) some period of adaptation can be required. On the other hand, although the endpoints would always be in the same place, it would be more difficult to become automatic at reading a horizontal position because your head will always be in a different position and thus the endpoints will be in a different position relative to your head. It is possible to overcome this by making the gauge larger so that the relative position of your head has less influence. In fact, 'linear' gauges were somewhat common in the 70's & early 80's (they were actually horizontal windows over a round gauge), and they generally took up half the dashboard. This will not be a problem for a gauge, though, unless you tilt your head to the side and try to read the speedometer, in which case, it would be harder to read!

share|improve this answer
2  
The psychological conditioning to a particular gauge over time is a solid, valid point. My empirical observations regarding consistency of speedometer gauge orientation are not consistent with yours, though. Nearly all of the vehicles I have owned over my lifetime have had a somewhat different orientation! In particular, this mostly resorts from having different maximum values on the gauge. I've even owned vehicles with (a) a linear gauge(!) and (b) a gauge that changed from miles/hr to km/hour depending on the user's selection. –  cardinal Jul 16 '12 at 14:14
    
(The latter can cause havoc since my experience is we seem to associate a relative gauge position to an absolute speed independent of units.) –  cardinal Jul 16 '12 at 14:15
    
@cardinal, good points, as always. I've updated my answer to address them. I've also had 'linear' gauges, they were fairly common back in the 70's & 80's. Nb, however, that they were actually a horizontal window over a circular gauge & that the angle of the needle did change as it moved from left to right. That extra (redundant) info made it easier to read, but they also had to be larger as I note. I've never seen your last situation, but I guess options like that will probably become more common. However, IMO, your comment about (b) supports my position. –  gung Jul 16 '12 at 15:03
    
I don't agree with this. You seem to be saying the axis is superfluous for a circular (i.e. polar coordinate) because of learning, but it isn't for rectilinear coordinate systems in 1-d or 2-d. The learning is pertinent (and related to what I said about historical inertia) but the part about needing to know (or not) the end-points is not correct. You need to have a documented scale for either to depict position accurately along the scale. –  Andy W Jul 16 '12 at 23:18
    
Also to note dials are frequently used for other instruments on the dashboard, but don't go in the same direction! See the picture of the dial for engine temp in the dashboard in the Robert Kosara post I link to. –  Andy W Jul 16 '12 at 23:19

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.