Well, like the racially-based scholarships I was offered as an undergrad. I turned them down because I wanted any assistance I received to be based on real need and merit, rather than a factor I had no control over (the ethnicity of my parents). Or racial hiring quotas. Things like that.

On the other hand, it's a…

Again, this is why theory matters so much. Large sample size, multicollinearity (or autocollinearity), functional form miss-match, non-random sample (or treatment), and other problems can lead to that 95% significance, but that's all secondary to what you decide to measure in the first place and how you decide to…

I'm actually looking at publishing a piece I wrote about why the media's reaction to violent games is dishonest and disproportionate. The most scholarly work on the effects of violent video games has been done by Bushman and Anderson (both together and individually), using the psychological framework of the General…

No, I do the same thing. I also almost exclusively use GURPS anymore, both because the rules are modular and since the focus of the game is not on combat.

Ahh, a fellow super! Although your power is probably more entertaining than mine. Being super-forgetful can become wearisome . . .

This.

Snarky much?

Also, I agree that systematic differences between demographic groups are very real and pertinent social issues which ought to be addressed. But many existing programs run the danger of saying "your are of X minority, so here's assistance," rather than "the different factors in your life make it statistically unlikely…

That's exactly it, though. There is a huge distinction between, for example, "Latinos make far less than Caucasians and are less likely to attend and graduate from college" and "Latinos are much less likely to attend and graduate from college, but those who do tend to do just as well as Caucasians with similar…

Hmm. Good to know.

Income as well. At least in the United States, the best predictor for income isn't race, gender, family's socioeconomic status, or IQ. It's education. And once education is factored in with years of consistent full-time work, most of the income disparities you see between different groups disappears.

Well, any sort of statistical analysis you do should be back up by theory first. Not accounting for age in a multivariate regression on life expectancy is an amateur mistake and one which would have been avoided with a little more thought.

Yeah, this is why I tell my wife to get more flats.

Yeah. It's often not very easy to tie people to a crime scene—our legal system would get convictions more often and much faster if it did.

*reasonable, that is.

True. Again, the nature of rape just makes it hard to prosecute. It sucks, but it's true.

And that's exactly why, I think, innocent until proven guilty is so important. DNA tracing doesn't necessarily mean someone was raped.

Well, assuming that each side is an independent little screen, just code it with a rand function and change the parameters. It could be done through the tablet (okay, now I need to roll a d73). Heck, if this die was coded into a tablet game, the whole thing could be done automatically—the game tells you to roll three…

Though you could program it to simulate any one of those, or a theoretical die with any number of sides, for that matter.

And how would you do that?