In other words, too loony for the regular Moonies.

This might be a help. I generally go back through my stats notes, but they’re not much help here, and I didn’t see too much good and helpful when I Googled the concept. There’s a lot of it in my field now, though.

Yes, she referred to that multiple-comparison Wikipedia page. Bonferroni adjustments and the like.

No, it hasn’t changed. Not everywhere.

Maximum-likelihood model testing actually does something very much like that. It’s a perfectly legitimate way of assessing n models simultaneously to see which one fits the data better.

In my experience, people who get a dead-on 0.05 wamble on for some time about how this is actually meaningful. Even more when they get a p = 0.051.

No, they’re going to fail to be rejected. A subtle difference, but important.

...statisticians still believe in determinism...

Strange bedfellows.

What got me was the implication (not a stretch, it’s in the way it’s written, but surely to God it can’t be intentional) that “Black-Americans” only speak an indistinct language. What the hell would that sound like?

I thought we’d stopped talking about that.

With Donald Trump, there is no such thing as a new low.

Not in the least. Everyone can dig it.

Yes, these Fish People do indeed look annoyed.

Give yourself more credit. You gave your kids a stable home with adults to look after them. A great good gift.

Actually, since his equally unattractive but less likeable brother is currently running for the head of the Progressive Conservatives in Ontario, one might almost be tempted to say “yes”.

Sure, twinkletoes.

Sure. Like anything else, social programs don’t exhibit any economies of scale. Just like military expenditures don’t.

There is no place more desirable than Canada.

Some of us have indoor biffies. We have to go by whether or not your spit freezes before it hits the ground. Or if you have to open your nostrils by hand after taking an extra-deep breath and having them freeze shut.