[jasonwatkinspdx]

I'm not feeling smug, I'm feeling tired of people using statistical concepts outside the context of statistical evidence.

I don't see anyone making an argument as extreme as saying that tablet native journalism is impossible. I do think people are suggesting that narrowing their market may have been a key factor in exhausting their runway, and I think that's a reasonable question to ask.

But no one is presenting any statistical evidence either way, so a concept like correlation plays no part. We can quote cute maxims all day in order to try to sound smart, but if they aren't applicable to the discussion, what's the point other than as theater: an attempt at rhetorical persuasion by performing familiarity with irrelevant concepts.

As a separate topic, if you'll indulge my own theater:

The phrase "correlation is not causation" is broadly misunderstood as being a stronger criticism than it is. My way of summarizing this is that "correlation is not causation, but causation causes correlation." For a fully rigorous and enjoyable treatment of how causal concepts relate to statistical evidence I recommend Judea Pearl's book Causality, which elaborates the concepts for which he won the Turing prize. If you don't have time for the book, the epilog is up on the web a few different places and is a great summary.

[MaysonL]

I don't see anyone making an argument as extreme as saying that tablet native journalism is impossible.

Read the headline of the Felix Salmon blog post which Gruber was criticizing: "The impossibility of tablet-native journalism"

http://blogs.reuters.com/felix-salmon/2012/12/03/the-impossi...

[jasonwatkinspdx]

Oh man, you're right. I never read the headline. How embarrassing.