MIT Researchers Show Dash Font Choice Affects Distraction

bdking writes "A typeface family commonly found on the devices installed in many modern cars is more likely to cause drivers to spend more time looking away from the road than an alternative typeface tested in two studies, according to new research from MIT's AgeLab." It seems that the closed letter forms of Grotesque type faces require slightly more time to read than open letter forms of Humanist type faces, just enough that it could be problematic at highway speeds.

Comments

Just don't text/SMS!

[fantomas]

Well it wouldn't matter if you weren't texting while doing 70 on the highway! :-)

ok, I am sure the article is about the fonts on the dashboard or something like that but really, the number of drivers I see texting while they are rolling a ton of metal along at high speeds is ridiculous.

Re:Just don't text/SMS!

[Dupple]

Well it wouldn't matter if you weren't texting while doing 70 on the highway! :-)

I'm doing 70 and commenting on /. you insensitive clod!

So what we're saying is...

[Anonymous Coward]

Serif fonts are easier to read than sans-serif fonts?

Who would have thought it!

Bloody graphic designers. They'll join the lawyers, bankers, patent trolls, advertising shills, dodgy stock traders and so on up against the wall when the revolution comes!!!

Hmmmmm - its going to be an effin big wall, or we're going to have to operate in shifts to clear the backlog.......

Re:So what we're saying is...

[digitig]

I thought that was a matter of screen resolution. At the resolutions commonly available when most studies were done, serifs would have been hard to render accurately and consistently. Heck, even on the screen and size I am using to type this, if I switch to Times New Roman the anti-aliasing struggles with the serifs on 's' and 'n' with the result that they look blurry.

Re:So what we're saying is...

[Anonymous Brave Guy]

Serif fonts are easier to read, especially large blocks of text. The serifs "lead your eyes" from one letter to the next, and help your eye group the words.

Actually, that's an old theory that has been solidly debunked on both counts at this point.

For a start, based on experimental research, we know that people don't actually move their eyes continuously across the text as we read. Instead, our eyes make short jumps called saccades, fixating on one point on the line and then another a few characters further along. That immediately makes any argument based on serifs "guiding" anything suspect.

Re:So what we're saying is...

[BetterSense]

I tried to read you post, but all I got was *fap fap*saccades*fap fap*

Our eyes may use saccades at the hardware level, but we compose images with our brains' DSP (ASP?). The fact that the eyes jump around is interesting but means approximately nothing.

I think it's safe to say that nobody sees the world in jerky motion from eyeballs moving jerkily. For that matter, the high-resolution fovea in the human eye only subtends a few degrees of arc, but you just never notice, because the brain has heavy-duty processing power that synthesizes a high-resolution picture of the world through image-stitching. You can only focus at once distance at a time, but we don't really notice that either. We have stereo vision which means that we see double-images of things, but we don't really notice any double-images. We can see our noses 24 hours a day, but don't notice that either. If you want to try a fun experiment, go into an absolutely dark room, stare straight ahead, and fire a camera flash. Do not move your eyes. You will see a bright, very realistic image of the entire room for many seconds if you can avoid moving (saccading?) your eyes. You see the perfectly bright room, even in the absolute dark, because if you don't move your eyes, it's probably still accurate. As soon as you move your eyes, though, the image disappears, as your brain flushes it like cache data that can no longer be trusted. Sort of a visual cortex version of copy-on-write.

Typeface design is visual art. Visual artists have known for hundreds of years that certain shapes are pleasing, and how certain lines can draw your attention to certain features of an image, and how certain colors can influence mood. I'm sure it's all completely bunk though, after all, I read on Slasdot about saccades, so now I can dismiss another huge swath of scary subjective human experience and fence it safely out of the lonely introverted enclave that is my nerd existence.

Re:So what we're saying is...

[Anonymous Brave Guy]

Perhaps you didn't notice the link I posted to Alex Poole's site, where one of the first paragraphs says "In 2003 as part of my master’s degree I reviewed over 50 empirical studies in typography and found a definitive answer" [emphasis added]? It then goes on to describe that work in a lot more detail, complete with numerous citations. I've read some of Alex's work, and I've read quite a few of the other pieces of works he cites. You obviously haven't, but hey, if you prefer to trust in "another huge swath of scary subjective human experience" rather than empirical data collected across a broad set of scientific experiments, knock yourself out.

However, I suggest that you would be more convincing to others if, instead of attempting exactly the kind of unsourced pseudo-science you seem to be accusing me of and then throwing in a strawman or two at the end, you actually bothered to read the work I pointed to before. We don't have to guess at how these effects work or appeal to old wives' tales from the 1800s. We have detailed, properly conducted experiments using techniques like eye tracking and even updating text on a screen as fast as someone is actually reading it to determine what we really do see and how much our brain is filling in for us. Here's another page that describes some more experiments on related topics, which provide further examples of how the people researching this field reach the kinds of conclusions they do.

Just use Comic Sans

[CodeheadUK]

Everyone loves Comic Sans.

Then all drivers will be happy, smiley and give way to old ladies.

Re:Just use Comic Sans

[L4t3r4lu5]

Comic Sans is for schools and community newsletters. Pro desktop publishers and graphics designers use Papyrus

I feel lied to

[Anonymous Coward]

Did anybody else think this post was going to be about hyphens?

I'm blind

[TheInternetGuy]

I'm blind you insensitive clods, the typeface in all cars should be braille and nothing else.

Not "Grotresque", but "Square Grotesque"

[pieleric]

The summary links to Grotesque, but what they use in the article is "Square Grotesque", a modified version which is _really_ square and IMHO hard to read (and which apprently quite appreciated by car manufacturers). Concluding every Grotesque font is hard to read is definitely not what the research demonstrated.

The best is to have a look at the paper, which has good examples. A similar font can be found on wikipedia there: http://en.wikipedia.org/wiki/Eurostile (but I find this one is still slightly easier to read).

Nothing to do with (sans)serif

[A Friendly Troll]

Read the PDF, people, damn it, before jumping to conclusions.

The fonts used in the experiment were Eurostile as the grotesque and Frutiger as the humanist. Both of those are sans serif.

This is about shapes, form and spacing.

Re:Old news

[Kenoli]

This isn't exactly a new finding. Typographers have known this for over a century, if not multiple centuries. Why do you think newspapers are printed in seriffed typefaces?

This research deals with the shapes, proportions, and spacing of characters in square grotesque and humanist typefaces. It doesn't have anything to do with serifs.

Re:How does this affect web design ?

[Rainer]

So what font should you choose on your web site ?

Your user's preferred font in their preferred size and with their preferred colors.

Funny story...

[JimmyVolatile]

This has been "tested" around 2002 in Norway. A car registration plate font redesign was conducted to make all plates issued from that moment look more modern and stylish and a font similar to Eurostile were implemented. All in the name of creating a mono-space font which would make all plates equal width. ("IL 111111" would be just as wide as "MW 123456")
Result: Numbers 3, 6, 8 and 9 went from being easily distinguishable at 80m+ to be undreadable by speed and toll cameras. You could pass speed cameras with little risk of getting fined and drive on any toll road for free. Sombody else would end up with the bill due to the misreading of the license plates.

Scroll down to see examples here:
http://www.typografi.org/bilskilt/bilskilt.html

In 2004 they decided to go for Myriad with variable white-spacing instead. This has not yet been implemented :)

Re:Nice find but....

[clickety6]

Too true, What bugs me these days is how many TV documentaries feature interviews with people who are driving cars. Stop talking to them and let them concentrate on driving. If you want to interview them, hire a bloody studio and sit them on a couch to do it - not while they're trying to guide 2 tonnes of metal through a busy intersection in town at 30 miles per hour!

Re:I think the lesson here is

[Dogtanian]

Eurostile is a pretty terrible font

Nothing wrong with Eurostile for what it is, just that in this case (where *any* tradeoff between legibility and style- however minor- might have an effect), it's probably not the best choice.

In fact, I'd say the fact that it's still "functional-looking" enough is how you could imagine car manufacturers using it in a dashboard whereas (e.g.) a black letter, or cursive/joined-up "handwriting" font would be much worse, but obviously so (and hence not likely to be chosen and hence not an issue here).

It seems that the closed letter forms of Grotesque type faces require slightly more time to read than open letter forms of Humanist type faces

This is true but incomplete; the study used Eurostile (apparently a Square Grotesque font), which is clearly less legible and a stronger example of those claimed issues than Helvetica. Helvetica is still a "Grotesque" type font).

I'm not saying that Helvetica is the perfect choice, or as good as a Humanist font, just that I suspect it doesn't suffer from this problem to anything like the same extent as Eurostile.