Web Users Judge Sites Instantly

Ant writes "This Nature.com news article reports that potential readers can make snap decisions in just 50 milliseconds: 'Like the look of our website? Whatever the answer, the chances are you made your mind up within the first twentieth of a second. A study by researchers in Canada has shown that the snap decisions Internet users make about the quality of a web page have a lasting impact on their opinions...'"

Comments

That's Crap

[Cobralisk]

This article is obviously rubbish

Re:That's Crap

[William Robinson]

This article is obviously rubbish

Yeah, it took additional 50 ms to close the tab.

Re:That's Crap

[malsdavis]

Indeed, it reads just like all those "employers decide whether they will employ you in 50ms" studies.

Re:That's Crap

[Crayon Kid]

Now let's talk about how we recognize ad banners in 50ms and shut them out of our vision.

Duh

[Lord_Dweomer]

Well, let me be the first to say "Duh, of course we do."

Having all this information at our fingertips is awe-inspiring, yet completely useless if we can't sort through it properly. That's why companies like Google and datamining companies make so much money.

As society and people evolve to adapt to the new technology, we build our "defenses" against bad information. We have so much to go through that unless we are able to filter out bad information that quickly, we'll never get anywhere. Not to mention the fact that in this day and age of spyware/adware, plagiarism, virii and big brother everybody needs to learn what information to avoid.

Re:Duh

[Blakey Rat]

So Hitler didn't fund the Illuminadi after being abducted by big-headed aliens in a UFO in the Bermuda Triangle while covering-up his involvement in the American Civil War? Impossible!

Re:Duh

[pomo monster]

That's exactly why I resent grammar/spelling nazis--if everyone followed their advice, it'd be harder at a glance to filter out the idiots.

(That said, some of the smartest people I know are dyslexic. And rulebreakers in general.)

Re:Duh

[boingo82]

Consequently writing teh though...

Um, you spelled "Consistently" wrong.

/. Design

[Anonymous Coward]

This is why I kept away from /. for so many years

"Like the look of our website?"

[strobexii]

It hasn't stopped us from visiting Slashdot. Over and over and over again...

Funny...

[versiondub]

I thought playboy.com was drivel when I was a young lad...but over the course of about 5 years, that all changed.

Re:Funny...

[gangien]

as a funny offtopic info. apparently playboy mirrors files for eclipse, apache, freebsd, and some other stuff! coolness. I fuond this out in some other article clicking around. look for yourself

Ironically...

[saladasalad]

it took me just 50 milliseconds to disagree with that article!

Navigations and ads

[AuMatar]

I think these are the two big determiners- if the first thing I see are 20 banner ads, I'm looking elsewhere. If I can't easily see how to get to the data I want, I'm looking elsewhere. These are easy to tell very quickly (ads on 1 glance, navigation by looking for a left column or top navigation bar). Most sites that have people leave that quickly fail one of these 2 tests, I think.

Re:Navigations and ads

[bit01]

Another big determiner for me, on major sites anyway, is time-to-load. I'll frequently abort a page before it's even finished if I'm not reading something else.

A long time-to-load probably means a badly configured server, or graphics heavy and often content free site. If a graphics rich site like BBC news can get it right, why can't anybody else?

Incidentally, 50ms can't be right - very few web sites take less than that to load.

---

Open source software is everything that closed source software is. Plus the source is available and copyable.

Re:Navigations and ads

[Anonymous Coward]

Incidentally, 50ms can't be right - very few web sites take less than that to load.

Yes, and if you read the article, it's clear that the study does not show what it claims to:

Even though the images flashed up for just 50 milliseconds, roughly the duration of a single frame of standard television footage, their verdicts tallied well with judgements made after a longer period of scrutiny.

But there is a major flaw. When the image is gone, the participants don't automatically stop making judgements about it.

50 ms (a.k.a. three refreshes at 60 Hz) is long enough for a person to see something and remember basically what it looks like. In fact, your mind will continue to perceive the image well after the display has gone away. This phenomenon is part of what used to be called 'persistence of vision'.

So when the experimenters ask the subject a few seconds later what their impression was, and the subject takes a second or two to indicate a preference, this is not necessarily a 50 ms snap judgement. There are whole seconds during which the image was probably being thought about.

Now, it may be possible that a snap judgement really can be made in 50 ms. But this study does nothing to prove that.

Oh Dear

[HotmanParisHiltonKam]

Quoth TFA "Even though the images flashed up for just 50 milliseconds, roughly the duration of a single frame of standard television footage, their verdicts tallied well with judgements made after a longer period of scrutiny."

The human reaction time is about .25 seconds. This study erronseously assumes that the judgement is made during the time the image is displayed - of course, the image retention time on the eye end the lasting photographic imprint on the memory means that the judgement can happen well after the image is gone.

Re:Oh Dear

[SnowZero]

You are definitely on the right track, though its not clear if its the article summary that's botched or the study itself. Trained cognitive psychologists at least should know better, so I'd lean toward an innacurate summary. The title should probably read "users can judge websites after seeing them for only 50 msec". The "photographic" effect you are referring to is called after-image, and can last long after the initial stimulus is gone. They could make a stronger claim than I stated above if they put up a visual distraction image after the 50 msec (visual memory studies do this a lot). But the summary doesn't mention it so we can't know for sure.

Regarding human reaction time, it varies depending on the task, but rarely is less than 100 msec (usually when you expect something to happen, such as runners starting a race). That means some tasks can be completed faster than 250 msec thouch, so that's not a good lower bound to quote if you are trying to debunk something. 50 msec certainly is too fast for anything I'd call "judgement" though, as people usually cannot even press a button that fast in response to an event.

At any rate, the slashdot summary is far from an accurate description of the phenomenon, but since when is that news...

P.S. I am not a psychologist, but I do have a B.S. double in cognitive science.

goatse.cx

[heatdeath]

I guess now would be the most appropriate time for people to start posting goatse.cx links. I can tell you what, I think my reaction time was quicker than 50ms the first time I accidentally clicked on that link at work. *shudder*

Firefox contributes to the effect

[saskboy]

"Firefox prevented this site from opening a popup window."

Whenever I see that on a website, right there I think to myself, "This is an annoying, and/or low quality website with suspect information on it."

Re:Firefox contributes to the effect

[thetzar]

"Firefox prevented this site from opening a popup window."

Whenever I see that on a website, right there I think to myself, "This is an annoying, and/or low quality website with suspect information on it."

Funny, I see that every time I go to cnn.com...

In other news:

[humungusfungus]

"Man judges book by cover"

Looks like someone's been reading...

[Errandboy of Doom]

Blink by Malcolm Gladwell.

Re:Wrong word choice

[foandd]

Excellent... now look up "pedant."

Known for Years at Slashdot

[mattwarden]

Commenters on /. have for years been able to judge articles without reading anything more than a marginally accurate 3-sentence summary riddled with typos. Why would scientists think the same would not apply to impressions of websites?

Obviously another waste of government research funds that could be better applied to [insert controversial proposed government project aimed at protecting against terrorism].

By the way, I didn't have a chance to read the article.

They shut that one down.

[xiphoris]

Goatse.cx has been shut down. You just get an error page about policy violations now. (Don't believe me? Check for yourself!)

Imagine that. A Slashdot post linking to Goatse and *not* being a troll! =)

They have, however, relocated to goatse.ca.

Re:They shut that one down.

[Morlark]

Actually, the goatse.ca mirror is incomplete. It lacks a great deal of content that goatse.cx (and goat.cx) had. The most complete goatse mirror you are likely to find is goatse.ragingfist.net

Re:They shut that one down.

[stevesliva]

(Don't believe me? Check for yourself!)

I'll just take your word for it, thanks.

Web Site Peeves

[queenb**ch]

50 milliseconds huh?

Here's my list of things that almost guarantee that I'll leave your site behind, never to look back.

1 - Music - Your taste in music is not mine. Your music sucks!
2 - Pages that don't load - It's usually the page that looks like it has exactly what you were searching for too!
3 - Pages that don't contain the information "as advertised" - you know the ones...you click on a link and it goes to some search page that tries to reset your home page.
4 - Pages that are more banner ad than web page - Get over it. No one wants to see that much advertising.
5 - Anything that blinks - Thank god the W3C deprecated the blink tag
6 - Anything that demands I install a plug-in for "the user experience" - espeically those stupid cursors
7 - Anything that spawns pop ads
8 - Anything that doesn't present easy to read and use navigation (www.thetrueagency.com/true.html is a prime example of this)
9 - Anything that doesn't have a sufficient amount of contrast between the text and the background.
10 - Anything that uses more than 5 different fonts on the same page - Its a web site, not a comic book.
11 - Sites that redirect to another redirect - We get the idea that you move - a lot.
12 - Anything that uses more than 6 colors on the same page - It looks like a circus barfed on your page.

2 cents,

Queen B

Re:Web Site Peeves

[Justin205]

Hint for Firefox users... about:config -> browser.blink_allowed

Set it to false. It unfortunately defaults to true, at least on 1.0.7.

Re:A perfect example being...

[spectrumCoder]

Wow, if you boycott any place you see a pop up ad for I'm surprised you end up going anywhere.

yeah right...

[Striver]

So you are trying to tell us you just read playboy.com for the mirror files...