Comments on: What’s Wrong With this Graphic on the Future of Information?

By: Teddy

Teddy — Sat, 20 Jun 2009 17:06:41 +0000

The chart isn’t that bad. I have nowhere to start in verifying or refuting the data, but if you care about what it is trying to say, you will figure the graph out. If you don’t care (or don’t trust the data) you won’t give it a real look and say its sloppy.

The top of the y axis could read “New / Burgeoning” and the bottom of the y axis could read “Old / Dying”

Then in any given year, there is a cross-section demonstrating the distribution across various media.

For instance, in 1900 about 25% of people got their info from Newspapers/Magazines and %75 from the Local Marketplace with Newspapers being newer (higher on the chart).

By: Rick Schettino

Rick Schettino — Tue, 16 Jun 2009 09:33:49 +0000

It’s an interesting way of illustrating one’s prediction to show the waves of change that are happening in communication. Just another viewpoint albeit a messy one.

By: Jodi Schneider

Jodi Schneider — Sun, 07 Jun 2009 21:03:26 +0000

I like the idea of an “opinion in graph form”. Thanks, Nathan, for that reframing!

By: JÃ©rÃ´me Cukier

JÃ©rÃ´me Cukier — Fri, 05 Jun 2009 10:33:35 +0000

there’s no data, so it cannot be (data) visualization. It’s just a diagram.

still it claims that face-to-face, written and broadcast information dissemination will disappear within 10 years, and be entirely replaced by new paradigms that will appear every year.

As CATS would put it, hahaha.

By: Tim

Tim — Wed, 03 Jun 2009 20:02:24 +0000

@Nathan – It would be nice to get your point of view in this discussion instead of commenting on small pieces that doesn’t satisfy your 5 open questions in this post.

By: Nathan Yau

Nathan Yau — Wed, 03 Jun 2009 19:12:11 +0000

@Tim – Right, that’s what sampling is for. Nielsen ratings are samples, and we use statistics to make an approximation. We know that if we sample properly and enough, the law of large numbers will kick in. As for quantifying trends, well, that’s just time series analysis.

By: Tim

Tim — Wed, 03 Jun 2009 18:44:54 +0000

@Nathan – You should use 30 secs to re-read my comment. I didn’t say it not done or it’s not commen practice but I said that it’s hard and allmost impossible (though some companies try and history is full of examples where they have failed miserably. Just take the opinion measurements up to an election).

If you can quantify trends answer me this question: How many people in the work at this moment thinks that TV is their primary source for news?

You can do an approximation of it but never give an accrurate quantity (simply because you can’t bother every single person in the world).

Taken the approximation into account I still think that the visualication that Thomas did together with the articale is a good guess for the future trends of sources for news.

By: Nathan Yau

Nathan Yau — Wed, 03 Jun 2009 18:02:30 +0000

@Tim – trends are totally possible to quantify – and it’s common practice.

By: Tim

Tim — Wed, 03 Jun 2009 17:31:02 +0000

This discussion doesn’t do the orginal articale justice and to all the people shouting FAIL try to read the the articale and see the graphic (read: not graph) in that context.

I agree with Thomas Baekdal that this might be a possible future for sources of news.

To disguess wheater this is a good or bad graph when it’s actually a graphical visulatiation of trends is irelevant in my eyes. Trends is very hard to quantify (if not impossible) so if you could illustrate the future trends for where people get their news I hereby challenge you!

By: Marlena Compton

Marlena Compton — Wed, 03 Jun 2009 15:25:28 +0000

When this article opened on my reader, showing the graphic, I went slightly cross-eyed for a moment. This reaction made me wonder about the slope of the lines.

I read about sparklines in Tufte’s _Beautiful Evidence_ over the weekend, and he writes about how slopes in a graphic should average 45 deg. over its length(p60). Just a SWAG, but this graphic probably ignores this principle as well.