Saturday, March 17, 2012

Notes on IQ vs. Arguments

I find communications of this type easier to verify and think about, but I don't know how or why. (HT.)

"The upshot of this is that if an argument doesn’t result in an agreement, at least one of the parties involved is being irrational or dishonest."
By far the most common reason for failure to agree is unwillingness or inability to communicate. The former leads to the latter, so almost all disagreement is proximately caused by lack of articulateness. In the comments on Aretae, Jehu adds that since tribes benefit differently from different conclusions, regardless of veracity, they'll lie to try to shift the likely outcome. Which is intentional inability to communicate.

The relevance of the simplification comes from being a single test I can apply to any argument. Are they articulating? If not, the argument is corrupt. At this point it doesn't matter why they're not, only whether it's worth trying to cause them to articulate. I find the listening unimportant. It'd be nice if I could get the debaters to listen to each other, but mainly I want to hear the articulations, so I can get all the evidence.

More generally, you can see the same kind of objection as Jehu's in Zietsman's comments. This means readers are intuitively concluding that this piece is misleading, as applied to their goals. I suspect they're onto something, but I'm unimpressed with their identification of the main issue. For a start, they didn't articulate what their take-away was, and I can't safely work it out from what they've said.

"One should also be aware that the equation gives an average IQ – the actual IQs vary quite a bit around the expected figure. For example, the authors show that threshold effects exist and that the minimum IQ needed to achieve a rating of 2000 is around 85-90. This is 30-35 IQ points lower than the expected IQ. Also from his peak rating Garry Kasparov’s expected IQ is 167 (and wild claims of 180+ have been made) but his actual IQ was measured at 135 (in a test sponsored by Stern magazine), some 32 IQ points lower."
Fascinating. My immediate thought: estimates of IQ for Einstein and Newton. I thought they were epistemically suspect, and now I have evidence. The same methodology gives a 180 IQ for Kasparov. Hey guys, I found some unknown unknowns... Note that this could underestimate IQ too.

"The difference between a grandmaster and a good beginner would be 5623 to 1."
Added to Dunning-Krueger, the beginner has no idea how utterly dominated they are by the grandmaster. Without unmistakeable evidence like a king getting knocked over, the beginner will think they won all five thousand times. Indeed, I can watch people saying they didn't 'really' lose, even when the king bites it. They make excuses. (So do I.)
I suspect this is largely because the brain incorrectly interprets losing at chess as morally relevant to the individual's character - that only bad people lose. This is the flip side of evaluating high-status people as 'better,' and accepting status hierarchies as morally relevant. (This one I don't do so badly on.)

"It would take the result of a string of 21-22 games (of backgammon) to provide the same test of relative skill as does a single game of chess."
In other words it's a lot easier to say you didn't 'really' lose at backgammon, and thus avoid triggering the faulty moral-status evaluation. Which both explains and excuses why games of chance are more comfortable to play and thus popular, especially among friends.

So here's a puzzle. He states his IQ is ~180 in the about section. You have enough evidence to evaluate my comprehension of the piece yourself.
"Some research shows that friends and spouses have an average IQ difference of 12 points, that for IQ differences less than 20 points a reciprocal intellectual relationship is the rule, for IQ differences between 20-30 points the intellectual relationship tends to be one way, and that IQ differences greater than 30 points tend to create real barriers to communication."
(There is a solution: the smart guy can talk down. By definition they can model the conversation, and the student cannot. I'm putting that aside for the moment.)

My IQ rates in the 130 region. If this was my only datum I'd have to conclude that my reading comprehension skill is worth something like 40 points of IQ.
Which observation is in error?


spandrell said...

Where does that quote on the IQ differences impeding communication originally come from? I´ve read several versions, the lower end starting at 15 points making a conversation impossible.

30 does make sense.

Alrenous said...

It's from the very bottom of the post. It was pretty convenient, since I'm comparing my IQ to someone claiming that he shouldn't be able to talk much to anyone below 150.

Delta 30 is confirmed by politician IQs, which rarely range over 130.

Alrenous said...

Oh, you mean originally originally. Auto-facepalm.

No idea where I first saw it.

Alrenous said...

I'll probably clean these comments up. Perhaps one day I'll be able to write while taking into account a lack of edit button.

So you got me curious, I ended up finding it. It's from the research of one Leta Hollingworth. But first...

So there's this. It makes me think someone extrapolated from the psychologist's retard and gifted thresholds. It gets bonus points for mentioning Mangan and OneSTDV.

Now the relevant quote and source,

"“… generally speaking, a leadership pattern will not form--or it will break up--when a discrepancy of more than about 30 points of IQ comes to exist between leader and led” Children Above 180 IQ Stanford Binet: Origin and Development (1942 p. 287)." (Quote source.)

A decent summary, from which I found the name to search for. Ctrl-f "leader " including space.

Aretae said...


Also...written communication and spoken operate in different spaces. I'm inclined to believe (probably your research already uncovered this) that the written mode is far more forgiving of differences in IQ...and far more work (in general) for the author.

Similarly...I (4+ sigma ... but most conventional IQ measures top out in effectiveness near or under IQ 160, and I ring the top-out bell on all of them) have spent my entire professional life ( 20 years) focused on teaching folks who are much lower IQ than me.

I've taught classes full of ~85 IQ folks (vocational school full of lower class secretaries-in-training learning business english and Microsoft Word ... also dyslexics learning to read), and classes full of 120+ IQ folks (Linux, Android, Web Services programming courses to experienced professionals in Silicon Valley). Clearly, I'm playing outside my comfort-space teaching at least some of those folks. And that's true...but the communication is mostly one-way, so I just have to modulate my presentation to their pattern...and 20 years of practice gives me some skill here.

Almost none of those folks become my friends, though. I would be shocked if any of my serious girlfriends (7 of > 6 months over 15 years) clocked in with an IQ of only 130 (Example of someone that didn't make the serious girlfriends cut, who are difficult to talk to, for lack of smarts: she didn't finish her M.S. in math, but has nonetheless worked as a programmer)...and my favorite conversation partners over the last 10 years are clearly in the >4 sigma range (pitifully incapable of communicating with folks who are only 2 sigma, as experienced through many shared interactions -- not that I'm much better -- I am friendlier though).

Also...3/3 immediate family members besides myself measured at >3 sigma. I (mostly) think they're all noticeably slow. And, of course, I was able to do (and did) College Level computer science classes (assume age 18, IQ 110) at Age 9.

Enough biography...My assessment is that the 2 sigma analysis is pretty clearly correct...if you're talking about extended two-way interaction. Writing, Teaching...completely different models.

Alrenous said...

That makes sense. I usually model IQ as clock speed, and conversation is in real time. Consistently outdistance the other side enough, and you're having a conversation with yourself. Or else having half a conversation and letting your mind wander while you wait. Assuming they even have the necessary background information.