A few years back I was tangentially involved in a very clever study involving the executive ethics that was designed to address this concern.

Some of the questions were routine business questions, others were general questions about the material they read, and, the meat of the questions were ones on the moral consequences of executive decisions. It was expected that executives would answer all the questions correctly. Instead, the study scored based on how long it took them to answer the moral questions as opposed to the general questions. These scores were then compared against respondents from a control group of non-executives.

EDIT: Gosh! I hope I'm representing her study accurately... "Business Experience and Moral Awareness: When Less May be More", Jennifer Jordan (2005)

So what happened? Citation? I'd like to read the study too.

Its a trick - hes seeing what type of responses he gets based on his waiting for a set amount of time. We are being tested!

Jesus! Cliff hanger of the week!

More info needed ASAP.

OP here, totally agree with this. That question is almost "How honest are you with yourself?"

Takes a bit of scrolling down the page of results[1], but people are actually pretty honest on that one, the mode response is "Agree somewhat" (this finding is from our initial, seed sample).

Also agree about the ease of getting around this. Particularly if you're a psychopath, and are therefore good at being manipulative. The version of this test used in parole settings[2] deals with this by having professional interviewers who use the person's history and records to make assessments, in addition to the interview. (Note that I'm not actually defending that process; part of the point of this and NPR's recent discussions about psychopathy is to stimulate thinking about whether stuff like this should actually be used for important decisions like parole)

[1]: https://www.statwing.com/open/datasets/6beda395392cc4b0115e0...

[2]: http://en.wikipedia.org/wiki/Hare_Psychopathy_Checklist

edit: clarified that results from [1] are from initial seed sample of respondents, not the HN sample.

> Aside from the false dichotomies, surely any psychopath with even a shred of intelligence could answer strategically and defeat the test's intended purpose. And they use this in prison hearings?

Psychology is a science, just like pizza is a vegetable.

But...congress said so? Why would you ever doubt anything a group of highly functioning psychopaths say?

It does have tomato...

The thing is, the way you deal with those questions is part of the test. You say that you look out for other people because they matter to you. But just by saying that, you give away a very important piece of information: other people matter to you. They matter enough that sometimes you'll even take risks for them. That's a reasonably strong indicator of non-psychopathy in and of itself.

I would argue that the intent of the question is asking what matters to you, and on that basis I picked "disagree somewhat".

Of course I'm still just not-executing-my-nonexistent-utility-function, but the point is that said utility function cares about people other than myself. In aggregate, more so than myself.

> I enjoy manipulating other people's feelings.

I've had this exact same thought before. I love manipulating people's feelings in certain ways. I like making people have fun. If you're in a bad mood and someone goes out of their way to try and put you in a good mood they have intentionally manipulated you.

I suspect you enjoy it when other people are enjoying themselves, and that you don't get your enjoyment out of the manipulation itself.

At their expense is usually assumed.

I got 22 out of 48, but I think that's because I almost never use "Strongly" levels, at least not on general statements like that.

That apparently puts me in the 9th decile. Weird, because I usually come up as neutral or chaotic good on alignment tests.

I'm not sure chaotic good is incompatible with psychopathy.

Best comment/game reference ever. :D

The test has less nuance than D&D alignments, that's for sure. Feels like the 'right' answers are all Lawful Good.