×

Welcome to the Slashdot Beta site -- learn more here. Use the link in the footer or click here to return to the Classic version of Slashdot.

Thank you!

Before you choose to head back to the Classic look of the site, we'd appreciate it if you share your thoughts on the Beta; your feedback is what drives our ongoing development.

Beta is different and we value you taking the time to try it out. Please take a look at the changes we've made in Beta and  learn more about it. Thanks for reading, and for making the site better!

Data Visualization: Too Easy To Be Too Slick?

Soulskill posted about 8 months ago | from the it's-pretty-so-it-must-be-true dept.

Stats 41

jfruh writes "Data visualization tools are finally putting a longtime dream within reach: offering the ability to make beautiful, slick-looking charts out of datasets almost automatically. But are our psyches ready for the shift? Data scientist Pete Warden quickly put together a visualization of Facebook name geography. Though he didn't consider it to be a scientific sample that could drive major decisions, he quickly found that it drove discussion at the New York Times and on white supremacist websites. 'There is an element of "wow, it's so professionally presented that it must be true,"' said Jim Bell, chief marketing officer for Jaspersoft."

cancel ×
This is a preview of your comment

No Comment Title Entered

Anonymous Coward 1 minute ago

No Comment Entered

41 comments

I don't believe this article at all (0)

Anonymous Coward | about 8 months ago | (#44596533)

Can someone draw me an infographic so I know it's correct?

The map (4, Informative)

Garble Snarky (715674) | about 8 months ago | (#44596539)

Re:The map (1)

Anonymous Coward | about 8 months ago | (#44597615)

If this is what they're talking about, then Betteridge's law of Headlines [wikipedia.org] definitely applies here, because that map is awful.

The answer to he headline is a resounding NO.

Clearly it's not easy to be too slick, because the "slick" graph looks like it was made by a 5th grader that just learned how to use a simple graphics package. If you're impressed by that, go see what the kids at your local art college are doing; your jaw will hit the floor if you think this map is impressive.

Re:The map (0)

Anonymous Coward | about 8 months ago | (#44598593)

That is not data visualisation.
This is data visualisation: http://where-are-the-clouds.blogspot.nl/2011/03/fukushima-timelines-fog-of-data.html (top chart)

Yes (1)

Anonymous Coward | about 8 months ago | (#44596561)

The company I work for does a specialized kind of statistical analysis. Doing it properly involves at least four qualities: a lot of knowledge of the math behind the stats, a good understanding of the context of the data, a well-managed dataset, and good judgement.

Yet the software used in industry lets anybody with a half-decent dataset and some poking around produce charts and graphs that look every bit as professional ans the truly professional work.

The problem is manageable, largely with professional associations & standards of qualification for practitioners, but I have seen instances where people have tried to pass of junk work.

Re:Yes (1)

davester666 (731373) | about 8 months ago | (#44596685)

First, even before software was used, any decent graphic artist could draw up a good looking chart or graph.

Second, the idea that there should be a professional association of infographic-makers is idiotic. There are people try to pass off junk work [or even blatantly illegal] in every profession, whether it is in a regulated or unregulated profession, from CEO's to burger-flippers. Photographers can and have taken pictures that are entirely misleading, but have been published in newspapers.

Third, having seen the process first-hand [to generate some marketing statistics to advertise a private telephone book], the "buyer" [the telephone book publisher] only wanted to buy the reputation of the statistical analysis company. It really became a "how can we mince words so that our directory would seem more accurate than it really is".

Re:Yes (1)

gweihir (88907) | about 8 months ago | (#44597777)

Indeed. Understanding is mandatory, the graphical representation is just a tool that makes one specific step a bit easier, namely getting a first overview of the data. Most people do not understand that at all, hence the omnipresent "lying with statistics". In the end, these computer tools make lying with statistics just a bit cheaper, that is all. Maybe this is a good thing, because if every wacko can generate beautiful charts that are wrong/meaningless/lies, people might finally start to catch on that understanding is never easy and a picture never says it all (or even most of it).

Speak For Yourself (4, Informative)

Anonymous Coward | about 8 months ago | (#44596585)

If your psyche can't properly handle a "slick" looking chart or info-graphic then you have some serious issues that you need to deal with right away. Perhaps you should get of Slashdot for awhile.

Frankly, I was wondering what the point of this "article" was. But, now I realize that it is an advertisement for Jaspersoft.

Re:Speak For Yourself (1)

AthanasiusKircher (1333179) | about 8 months ago | (#44598311)

If your psyche can't properly handle a "slick" looking chart or info-graphic

Poor choice of words in the summary perhaps. But the larger point is that people can't help looking at graphical representations of data without making loads of assumptions... particularly if it looks at all "professional."

But this is no different than the kinds of graphical games used to misrepresent data that have been used for a century or more. You can even fake "professional" statistics with no graph at all. "85.7% of men preferred our brand of shaving cream!" A number like 85.7% sounds very exact -- must be a professional survey of a large number of people. Or, well, it could just be that they asked only 7 guys who worked at the company, and 6 of them (85.7%) said they like the shaving cream.

I sincerely doubt that people are making any more mistaken assumptions -- or any significantly new types -- from new graphs than they were when "How to Lie With Statistics" [wikipedia.org] was published in 1954. The only difference is that more people can make more graphs with ease. The problems of interpretation are not new.

Re:Speak For Yourself (1)

oursland (1898514) | about 8 months ago | (#44603347)

Some mass media manipulate graphs to large audiences to alter their opinion. Here's a bunch of good examples: http://foxnewsgraphs.tumblr.com/ [tumblr.com]

One has to look pretty closely to ensure that each graph has a y-axis that starts at 0, a consistent x-axis, that the height of bars and points match the numbers that are presented and other forms of lying with data.

This is basically the way... (-1)

Anonymous Coward | about 8 months ago | (#44596599)

...that the Climate Change scam was done...

Are We Ready For...? (1)

Jane Q. Public (1010737) | about 8 months ago | (#44596633)

I think this question is one of the major exceptions to Betteridge's Law, in that the answer is usually Yes.

Either the answer is "yes", or it doesn't matter because whatever it is, it's coming anyway.

Examples of past headlines like this:

Genetic Engineering Is Coming. Are We Ready?

Digital Cameras And Photo Software Are Becoming Too Sophisticated, Too Fast. Are We Ready For The Inevitable Fake Photographs?

Etc.

Please link to the supposed charts (0)

Anonymous Coward | about 8 months ago | (#44596647)

If you make a claim like "so professionally presented that it must be true", then I want to see the charts.

tl;dr: PICS or GTFO.

Uhhh....can you say slashvertisement (0)

Anonymous Coward | about 8 months ago | (#44596677)

I actually rtf article (well most of it anyways ;-) and it's just an advertisement for chartio

tub6irl (-1)

Anonymous Coward | about 8 months ago | (#44596735)

The 8es1gnation

Might as well be the first to bring up Tufte (2)

AdamHaun (43173) | about 8 months ago | (#44596937)

The Visual Display of Quantitative Information has many examples of slick-looking graphics going back decades, long before computers were any good at graphics. How to Lie with Statistics is even older than that. Newspapers and news magazines have always been infamously bad at showing data. It's a rare data graphic that doesn't focus on decoration over content, and that's ignoring the ones that are deliberately distorted.

That being said, most software (I'm looking at you, Excel) is way too helpful about creating bad data graphics.

Dilbert, over 4 years ago (0)

Anonymous Coward | about 8 months ago | (#44596963)

Relevant link: http://dilbert.com/strips/comic/2009-03-07/

Pretty, but good? (0)

Anonymous Coward | about 8 months ago | (#44597035)

I feel this push from top management at work. Of course the problem is how good is the underlying data? If you are doing marketing and aggregating data across the country, maybe you can pass of errors as noise, but the data sets I am working with are much smaller. A few thousand, or in some cases hundreds of entries per year. Differences in data entry between sites or even between individual users can massively skew any results and do it pretty fast.

An old theme, now faster. (1)

fuzzyfuzzyfungus (1223518) | about 8 months ago | (#44597151)

Who remembers the classic How to Lie With Statistics [archive.org] ?

Unsurprisingly, just as advances in computer technology have allowed us to make bigger, messier, errors faster than ever before, they are allowing us to exploit the fact that human statistical intuition is pretty much shit better than ever.

Re:An old theme, now faster. (1)

darenw (74015) | about 8 months ago | (#44604825)

An excellent book. Recommended by 100% of those who recommend this book.

Re:An old theme, now faster. (1)

tlhIngan (30335) | about 8 months ago | (#44607909)

Unsurprisingly, just as advances in computer technology have allowed us to make bigger, messier, errors faster than ever before, they are allowing us to exploit the fact that human statistical intuition is pretty much shit better than ever.

Or to make content-less graphics as well. Really, an infographic is just a slick form of PowerPoint presentation, and everyone knows how trivially easy it is to make a complete slide deck that's impressive, but in the end, be completely content-free.

Heck, it's such a problem the US military has the same issue - they can produce very slick looking slide decks, but are either incomprehensible or are just worthless.

It isn't limited to PowerPoint, either - a slick graphic can be just as easily misleading, content-free or incomprehensible. And still on first glance look useful. (The eye is drawn to flash, while content analysis requires actually going through the content and critical thinking, a much slower process than the quick glimpse a graph gets you. It's why humans fall for the same traps).

The Computer Said It, So It Must Be True (1)

tutufan (2857787) | about 8 months ago | (#44597183)

This sounds like a rehash of the old "Computer Said It, So It Must Be True" meme from the not-so-distant past. If that experience can provide any guidance, people will mostly learn to ignore these flashy new graphics after 50 or 100 years or so...

Socalistan (1)

geezer nerd (1041858) | about 8 months ago | (#44597685)

When I first looked at the infographic on the referenced blog, I misread the label for California to be "Socialistan", thinking it to be a comment on the liberalism of the place. Later I realized it said "Socalistan", referring to the locale of the center of influence. I wonder if anyone else did the same.

I am surprised that "Socalistan" showed such localized connections on the map. Since so very many people in California are from outside California originally I would have thought there would be really strong connections to "back home".

Nonsensical graphic (1)

bigdavex (155746) | about 8 months ago | (#44597783)

This pie chart is my favorite example of a nonsensical graphic that looks really professional.

http://kick.it/the-facts-about-smoking-infographic/ [kick.it]

Re:Nonsensical graphic (1)

Kaenneth (82978) | about 8 months ago | (#44598335)

I think the anti-smoking ads are so bad because the tobacco companies are being forced to pay for them. so they take the passive-aggressive route.

Tobacco and Cat Pee both have Ammonia! oh dear!, they also both have DHMO!

Re:Nonsensical graphic (1)

Chris Mattern (191822) | about 8 months ago | (#44599401)

Two pie charts, both nonsensical! Racial distribution of adults who smoke is meaningless because total population of each classification varies widely. Meaningful would've been the *proportion* of each racial classification that smokes, which would have been best plotted as a bar chart, not a pie chart. And in the second one, "Funds available for controlling tobacco use"? Who says those funds are supposed to be used for that purpose? Who says that *only* those funds are supposed to be used for that purpose? "Funds available for controlling tobacco use" are the funds the legislature appropriates for that purpose, no more, no less.

Re:Nonsensical graphic (0)

Anonymous Coward | about 8 months ago | (#44601289)

People only believe what they want to believe. The ability to make a profit from a product that is addictive and kills people, shows that we may be stuck in an early stage of evolution, not quite an intelligent species. There are people that know more than others but only in certain areas, yet we assume thats a form of intelligence. Only by examining the big picture can you see that humans are not an intelligent species.

  I have had to learn excel, access, and a couple 3d animation program in order to explain and show results in several experiments that i am working on and unless you have the genetic instinctual reflexes of an used car salesman you are doomed.

PC Mania (1)

PhamNguyen (2695929) | about 8 months ago | (#44597863)

I like the way they consider the fact that a particular graphic was used by "White supremacists" to argue for a particular viewpoint, to prove that that graphic must be unscientific. I wonder if they would consider such graphics to be unscientific if they were used to argue that America was becoming more diverse, rather than that it was being 'swamped'.

The output is... (0)

Anonymous Coward | about 8 months ago | (#44598247)

Someone told me many, many years ago: "But the output is printed on green-bar paper by a computer -- of COURSE it's correct!"

We call it the autocad effect (0)

Anonymous Coward | about 8 months ago | (#44598531)

In my job there are certain schematics/designs stages we choose not to do in cad as it lends it far too much creditability.

People see a plan drawn in cad and it must be right and correct.

no big deal (0)

Anonymous Coward | about 8 months ago | (#44599099)

I don't get the big deal, it doesnt seem like anything really new.

Is that the same logic as (1)

Opportunist (166417) | about 8 months ago | (#44599367)

wow, they use so big words, they must be right?

If you don't know anything, you have to believe everything. Where's the news, that stupid people are easily duped? I knew that before.

People adapt. (1)

GrumpySteen (1250194) | about 8 months ago | (#44599617)

Look at the photos of the Cottingley Fairies [wikipedia.org] . To our eyes today, it's pretty obvious that those are paper cutouts but many people of the day believed that they could be real.

Listen to the radio play of H.G. Wells' "War of the Worlds" [youtube.com] . At the time, it terrified everyone. Now? It seems ridiculous that anyone would believe it.

Re-watch the Star Wars [youtube.com] . The lasers, explosions and other effects are cheesy and lame by today's standards, but it was awe-inspiring to the people who saw it in 1977.

Initially, people may take a nicer presentation of information with less skepticism than they should, but society adapts. In a few years, any widely used automated data visualization will be viewedthe same way a B-movie using bullet-time is viewed today; with a yawn.

Simulation Visualization (1)

friedmud (512466) | about 8 months ago | (#44600957)

I write massively parallel scientific simulation software for a living (the kind that runs on the biggest machines in the world)... and trying to come up with a way to display GBs or TBs of information from some of our largest simulations can be _tough_.

We use several open source packages ( Mostly http://www.paraview.org/paraview/resources/software.php [paraview.org] and https://wci.llnl.gov/codes/visit/ [llnl.gov] ), but most of our best visualizations are actual done using a commercial package ( http://www.ceisoftware.com/ [ceisoftware.com] )

For some examples check out the YouTube video here: http://www.youtube.com/watch?v=V-2VfET8SNw [youtube.com]

(That's me talking in the video). Those aren't necessarily our best visualizations - just some that happen to be on YouTube...

We find that the reactions to these simulations are mixed. They are certainly eye-pleasing... but sometimes if you go too far in making it look good it can actually turn scientists off. They will start to think that it looks "too good to be true" (I literally had a senior scientist in a room of 200 stand up at the end of one of my presentations and proclaim that "This is too good to be true!"). Because of this we try to do do just enough visualization that you can see all of the features of the simulation and understand what's happening without going overboard.

You have to realize that a lot of scientists still remember the days when they created line plots _by hand_ for publications! I suspect that as young scientists come up through the ranks this feeling that "slick graphics = not real" will go away.

At least, I hope....

Re:Simulation Visualization (1)

sjames (1099) | about 8 months ago | (#44603217)

It's a natural effect of having so much of our economy dedicated to marketing where truth is entirely secondary to creating belief. That earth may be even more heavily salted for the younger scientists who grew up associating anything like that with damned lie.

Visual bias (1)

oranGoo (961287) | about 8 months ago | (#44606301)

It is not a matter of psyche, but realizing the difference between visual and conceptual information processing in humans.

It's not unlike a GPU vs CPU: while visual is much faster it brings a certain bias.

(For bias look for "optical illusions" and focus more on static ones).

Check for New Comments
Slashdot Account

Need an Account?

Forgot your password?

Don't worry, we never post anything without your permission.

Submission Text Formatting Tips

We support a small subset of HTML, namely these tags:

  • b
  • i
  • p
  • br
  • a
  • ol
  • ul
  • li
  • dl
  • dt
  • dd
  • em
  • strong
  • tt
  • blockquote
  • div
  • quote
  • ecode

"ecode" can be used for code snippets, for example:

<ecode>    while(1) { do_something(); } </ecode>
Sign up for Slashdot Newsletters
Create a Slashdot Account

Loading...