Toward a Fuller Integration of Graphics in Statistical Analysis

The talk will be 20 Apr 2018 at 1:25pm.

And here are some things to read ahead of time, if you’re interested:

[2008] Diagnostics for multivariate imputations. {\em Applied Statistics} {\bf 57}, 273–291.

Usually I don’t go for such nitpicking but this one bothers me for some reason. It at once sounds like it has some extremely technical meaning and is nonsense slang. The term “Fuller Integration” sounds like the name for a type of integration named after an obscure person/town/process/etc, when you mean “More Complete Integration”. Is “fuller” even a word in the sense you are using it? If something is full, that means it has reached max capacity.

On looking it up it seemed that “Fuller Integration” was the name for something relating to experimental theology (not sure what that is, but I wonder if they use NHST?).

Anon:

What I meant by the title is that graphics are currently somewhat integrated into statistical analysis, and I want this integration to be closer to full.

Maybe I can just change the title to, Integration of Graphics in Statistical Analysis.

Integration of Graphics in Statistical Workflow to Encourage (or even Ennoble) better Analysis ;-)

Do think workflow is the key insight in this.

Keith,

Actually, I was going to call my talk Bayesian Workflow, but that was much of the topic of my NY R talk last year: https://www.youtube.com/watch?v=cuE9eHSbjNI

“Look what you’ve done now! How to use statistical graphs”

I’d call it “Adventures of a Methodological Terrorist,” but that wouldn’t go down well in NYC. I’ll have to save that title for my next visit to Princeton.

If you’re going to go with a song lyric, may as well go with a New Yorker:

“Show me what you got: Incorporating graphs into your statistical workflow.”

Or sticking to rock:

“You look so good to me: How to use statistical graphsâ€ť

+1 to Dan’s title.