Comments on: Too Good To Be True

By: Douglas Knight

Douglas Knight — Sat, 21 Mar 2015 00:26:20 +0000

Roger, Griggs explicitly talked about degrees.

By: Roger Sweeny

Roger Sweeny — Sat, 21 Mar 2015 00:20:21 +0000

Griggs did not say, technically or otherwise, that you can’t use a college degree. The Court very specifically declined to decide anything about degrees (since the case had not involved degrees, they didn’t have to).

Though I think the logic of Griggs applies to degrees, I don’t think anyone has ever gotten a court to do so. Practice since Griggs has allowed degrees to be used as a “condition of employment or advancement,” and no one expects that to change.

Report comment

By: A week of links | EVOLVING ECONOMICS

A week of links | EVOLVING ECONOMICS — Fri, 20 Mar 2015 08:01:06 +0000

[…] Psychological research sucks. […]

Report comment

By: The Psychology of Lifestyle (II) « Econstudentlog

The Psychology of Lifestyle (II) « Econstudentlog — Thu, 19 Mar 2015 22:41:20 +0000

[…] that frequency of past behaviour and habit are separate constructs [I also pointed this out elsewhere recently, but I think it’s an important insight. Revisiting my coverage of Buskirk et […]

Report comment

By: Browsing Catharsis – 03.18.15 | Increasing Marginal Utility

Browsing Catharsis – 03.18.15 | Increasing Marginal Utility — Wed, 18 Mar 2015 12:14:42 +0000

[…] “Too Good To Be True,” by Scott Alexander. […]

Report comment

By: Paul Torek

Paul Torek — Mon, 16 Mar 2015 15:38:16 +0000

Nice find, dlr. Word needs spreading.

Report comment

By: Cauê

Cauê — Sun, 15 Mar 2015 02:43:45 +0000

If a perfect Bayesian AI kept testing hypotheses and finding them false, maybe something would be wrong. But in Science we already know what’s wrong: humans.

It’s probably true that we’re currently testing way too many stupid hypotheses, but even if we got our act together as much as possible it would still be the case that many of our plausible hypotheses would turn out to be wrong.

Report comment

By: Airgap

Airgap — Sat, 14 Mar 2015 23:12:05 +0000

RTFS

Report comment

By: MadRocketSci

MadRocketSci — Sat, 14 Mar 2015 22:55:34 +0000

And you know this before you test it how? Sans prior experience, what process do you turn the crank on to decide, a-priori, what the result of testing it will be?

Report comment

By: Virginia

Virginia — Sat, 14 Mar 2015 20:10:45 +0000

Don’t you think that there’s a difference between college students and “troubled youth”? Comparing these two types of studies is comparing grapes and guavas. College students are motivated to improve because they want good grades, good degrees, and good jobs in a professional field. “Troubled youth” are… well… troubled; first they need to be convinced that good grades, good degrees, and a good professional jobs are both desirable and attainable. THEN we can start talking about plan-making and so on. I’m not a psychologist; maybe that’s why this sounds sensible to me.

Report comment