IQ Distribution

Discussion regarding the Outsider webcomic, science, technology and science fiction.

Moderator: Outsider Moderators

Sweforce
Posts: 546
Joined: Fri Mar 13, 2015 12:00 am

Re: Miscellaneous Loroi question-and-answer thread

Post by Sweforce »

boldilocks wrote:If the loroi are an engineered species I'd expect the IQ distribution to be very clustered around the middle of the bell curve, far more so than in humanity, unless their design allowed for mutations. Then again, telepathy might be such a powerful tool to maintain societal control that extreme IQ outliers would be automatically cleaned out of the gene pool regardless of mutations.
Then again, a telepathic society could counteract the human need for stable middle-of-the-road levels of IQ and therefore allow for a wider distribution without it negatively affecting societal health, thus less clustering around the middle.
A more narrow clustering around the middle could explain the difference in scientific development between loroi and humans. The loroi are further ahead but took a long time getting there while humanity advances rapidly. Loroi are smart but have fewer geniuses.

Gorbash
Posts: 30
Joined: Mon Sep 21, 2015 11:02 pm

Re: Miscellaneous Loroi question-and-answer thread

Post by Gorbash »

orion1836 wrote: To an extent. From what I've read, IQ tests are not biased (as uncomfortable as the results may be when applied to demographics). For one example, pattern matching is pattern matching, no matter your nationality, race, age, etc.
That depends entirely on the test being used, and its generally *not* safe to assume that a test that seems to apply well to one area will apply well to another.

Unfortunately for those of us who like TTRPGs out there, intelligence is generally very hard to define in a way that's both measurable and accurate. The WISC and WAIS tests attempt to make up for this by focusing on spacial reasoning, pattern matching, etc, but even with that focus, some amount of cultural bias will have an impact regardless.

The problem with most IQ tests that remove cultural bias is that they mostly focus on removing the language requirement. Well, surprise surprise, not only do they have to suffer through translation issues, that also fails to cover a plethora of other concepts included within cultural bias that are harder to track.

For example, here's a weird one -- color differentiation. Older languages don't always differentiation between colors as much as modern English does (though its worth noting that some differentiate *more*). Its entirely possible for questions on an IQ test to rely on the participant identifying two colors as separate from each other ("Organize all the red objects into shape X, and the orange objects into shape Y") that simply don't translate well to the person's native culture and language.

Neither the WISC nor the WAIS are as bad about cultural specificity as, say, the SAT or ACT (which are both more measures of 'would this person perform well in a college environment' than actual IQ), but cultural comparison of IQ tests still need to be taken with a grain of salt.

User avatar
Werra
Posts: 840
Joined: Wed Jun 06, 2018 8:27 pm

Re: Miscellaneous Loroi question-and-answer thread

Post by Werra »

That still doesn't explain how various genetic backgrounds within one culture consistently perform different from one another. A performance gap that also remains constant(ish) across different cultures.

Gorbash
Posts: 30
Joined: Mon Sep 21, 2015 11:02 pm

Re: Miscellaneous Loroi question-and-answer thread

Post by Gorbash »

Werra wrote:That still doesn't explain how various genetic backgrounds within one culture consistently perform different from one another. A performance gap that also remains constant(ish) across different cultures.
Sure, but like I said, grain of salt. Just because you see a pattern doesn't mean there isn't an outside factor influencing the result. There can be genetic traits common to specific human phenotypes that could be influencing the result... but there also could be influence from parenting strategies that are cross-culturally common with that phenotype, limited population sizes skewing the results in one direction or another, influence from the cultural position of the type of people most commonly taking the test in that area (for example: Irish immigrants to the US in the 1910s, when IQ tests were first being developed, tended to be of middling education -- because they weren't so poor that they couldn't leave Ireland, but they were poor enough they wanted to), etc etc.

Don't assume IQ tests are perfect, because they're not. They're a measure of a complex system we don't fully understand, by the users actually using that system, of other users using that system. The only proper way to truly determine IQ would be an analysis of brain structure that we're nowhere near ready to conduct.

boldilocks
Posts: 669
Joined: Fri Jun 02, 2017 3:27 pm

Re: Miscellaneous Loroi question-and-answer thread

Post by boldilocks »

That they didn't differentiate between certain colors does not imply that they couldn't look at a blue and a red square and notice that they were different colors. (I believe scandinavians used the word blue even for black, but if you gave them a black ball and a blue ball they'd be able to tell the difference.)
At most you'd be able to find out that oh, maybe pygmies are all color blind and that would invalidate certain IQ tests, but that's largely irrelevant because they also score lower than the chinese on IQ tests that don't rely on color matching.

I'm not sure what kind of 'culture' would suffer some horrendous inability to evaluate this, for example:
Image

Unless that culture was universally unable to distinguish shapes, which would suggest that it's a problem that's not inherently cultural, because all cultures have outliers. (Perhaps all pygmys are far-sighted and so the image appears a blur, for example.)
but there also could be influence from parenting strategies that are cross-culturally common with that phenotype
That is not what the research indicates. The research indicates that phenotypic expression will assert itself regardless of parenting strategy. (Unless the parenting strategy involves such obscene violence that the care-recipient doesn't survive or is entirely mangled by the process.)

Gorbash
Posts: 30
Joined: Mon Sep 21, 2015 11:02 pm

Re: Miscellaneous Loroi question-and-answer thread

Post by Gorbash »

boldilocks wrote:That they didn't differentiate between certain colors does not imply that they couldn't look at a blue and a red square and notice that they were different colors. (I believe scandinavians used the word blue even for black, but if you gave them a black ball and a blue ball they'd be able to tell the difference.)
True, but it does leave them with a disadvantage that could affect scoring. And yes, they should be able to differentiate between the two different balls, but that doesn't mean the fact that they're equating black and blue as the same color won't have an effect on testing (ranging from translations issues around "form a shape using all blue objects" to, potentially, issues with perceiving dark blue objects on a black background).
boldilocks wrote:I'm not sure what kind of 'culture' would suffer some horrendous inability to evaluate this, for example:
Image

Unless that culture was universally unable to distinguish shapes, which would suggest that it's a problem that's not inherently cultural, because all cultures have outliers. (Perhaps all pygmys are far-sighted and so the image appears a blur, for example.)
Or their culture reads from top to bottom rather than left to right. I could easily see getting different answers on that question based on which block is assumed to be the start of the pattern.
boldilocks wrote:That is not what the research indicates. The research indicates that phenotypic expression will assert itself regardless of parenting strategy. (Unless the parenting strategy involves such obscene violence that the care-recipient doesn't survive or is entirely mangled by the process.)
But not necessarily cognitive and behavioral expression.

"Intelligence: Knowns and Unknowns" (1995), which was one of the strongest bits of research arguing your point that I found, does point out the following:
"We should note, however, that low-income and non-white families are poorly represented in existing adoption studies as well as in most twin samples. Thus it is not yet clear whether these studies apply to the population as a whole. It remains possible that, across the full range of income and ethnicity, between-family differences have more lasting consequences for psychometric intelligence."
I can see your argument for genetics playing into phenotypic expression of cognitive ability to some extent (several more recent studies or compilations, like "A Twin Study of the Genetics of High Cognitive Ability Selected from 11,000 Twin Pairs in Six Studies from Four Countries" (2009), somewhat support your argument, but also maintain that environment is a strong factor), but I don't think there has been enough research yet to support the assumption that specific genetic backgrounds correlate to specific amounts of variation in IQ. And most of these studies still remain focused on comparisons between groups within the western world (Europe and the US); if you can find a twin study that covers both a Western and a non-Western nation, and controls for economic differences between the twins involved, please post it up.

User avatar
Werra
Posts: 840
Joined: Wed Jun 06, 2018 8:27 pm

Re: IQ Distribution

Post by Werra »

Gorbash wrote:I don't think there has been enough research yet to support the assumption that specific genetic backgrounds correlate to specific amounts of variation in IQ.
1. Evolution is real.
2. Humans are an evolved species.
3. Intelligence is a result of the physical form of human organs.
Therefore, genetic variations exist which cause differences in intelligence.

Is there even one other trait a human can posses which is not a priori assumed to be influenced by the biological ancestry?
Gorbash wrote:And most of these studies still remain focused on comparisons between groups within the western world (Europe and the US); if you can find a twin study that covers both a Western and a non-Western nation, and controls for economic differences between the twins involved, please post it up.
Why would it matter that twin studies are limited to within Western nations? First of all, they can still show how much of the variation in IQ is genetic and how much stems from the environment, even using only natives. Second, there are lots of people with diverging genetics amongst Western populations.

Gorbash
Posts: 30
Joined: Mon Sep 21, 2015 11:02 pm

Re: IQ Distribution

Post by Gorbash »

Werra wrote: 1. Evolution is real.
2. Humans are an evolved species.
3. Intelligence is a result of the physical form of human organs.
Therefore, genetic variations exist which cause differences in intelligence.
Sure, but that doesn't mean we should start equating "person is in genetic group X" with "person is more likely to be smarter." Not only do we not have a full understanding of the genetics or physical mechanics involved with intelligence, we can prove (via twin studies) that culture and upbringing has an effect on resulting intelligence (somewhere between 20-35% in modern research), and therefore can't prove that any one subgroup of people is always going to be smarter than another *because* of their genetics.
Werra wrote:Why would it matter that twin studies are limited to within Western nations? First of all, they can still show how much of the variation in IQ is genetic and how much stems from the environment, even using only natives. Second, there are lots of people with diverging genetics amongst Western populations.
Because environment and upbringing within Western cultures can end up grouping people of similar genetic subgroups into similar environments and cultural situations, at that can confound the comparison. If we want to prove that one subgroup is more or less intelligent than the others, we have to prove that any variation we see isn't being caused by culture. If that culture tends to be consistent across the population (and many minorities in Western nations of subcultures that impact them regardless of their economic status), it could be the 'cause' of the variation we're observing.

By running twin studies that are not restricted to Western nations, we can see how the results look outside of Western cultures, and compare afterwards to our previous studies to see if the variation is consistent. If its not, its very likely what we're seeing is a direct result of subculture the population being measured is in.

User avatar
Werra
Posts: 840
Joined: Wed Jun 06, 2018 8:27 pm

Re: IQ Distribution

Post by Werra »

Gorbash wrote: Sure, but that doesn't mean we should start equating "person is in genetic group X" with "person is more likely to be smarter." Not only do we not have a full understanding of the genetics or physical mechanics involved with intelligence, we can prove (via twin studies) that culture and upbringing has an effect on resulting intelligence (somewhere between 20-35% in modern research), and therefore can't prove that any one subgroup of people is always going to be smarter than another *because* of their genetics.
80 - 65% influence of genetics on IQ outcome is huge. At best that means that the majority of the difference is due to genetics and at worst, that almost all the difference is genetic. Why do we need a full understanding of genetics and physical mechanics to know what the IQ distribution of a given genetic group is? We don't need the same for height, for example.
Truth is, we don't even need to know what genes are to understand that heredity is deciding in IQ. Simply charting ancestry and performance would be enough to make that judgement. "Person is more likely to be smarter" is exactly what we should think. Because it's true, even at just 65% influence of genetics. Even at 10%, honestly. The likelyhood would just be smaller.
Gorbash wrote:Because environment and upbringing within Western cultures can end up grouping people of similar genetic subgroups into similar environments and cultural situations, at that can confound the comparison.
So what? A study only done on Finns or Scots would still show IQ to be hereditary.
Gorbash wrote:If we want to prove that one subgroup is more or less intelligent than the others, we have to prove that any variation we see isn't being caused by culture.

Any variation? That's way too strict. Especially when we know that genes are at least 65% responsible. Again, can you imagine making the same case for height distribution?
Gorbash wrote:If that culture tends to be consistent across the population (and many minorities in Western nations of subcultures that impact them regardless of their economic status), it could be the 'cause' of the variation we're observing.
The distributions we're seeing hold up in their native countries, as well as with individuals not raised in such subcultures.

Honestly, your standards seem really dubious. Turn the argument around in your head.
"If we want to prove that one subgroup is not more or less intelligent than the others, we have to prove that any variation we see isn't being caused by genetics."

Krulle
Posts: 1413
Joined: Wed May 20, 2015 9:14 am

Re: Miscellaneous Loroi question-and-answer thread

Post by Krulle »

Gorbash wrote:
boldilocks wrote:I'm not sure what kind of 'culture' would suffer some horrendous inability to evaluate this, for example:
Image

Unless that culture was universally unable to distinguish shapes, which would suggest that it's a problem that's not inherently cultural, because all cultures have outliers. (Perhaps all pygmys are far-sighted and so the image appears a blur, for example.)
Or their culture reads from top to bottom rather than left to right. I could easily see getting different answers on that question based on which block is assumed to be the start of the pattern.
Well, I saw two possible correct answers, but only one of them was in the list of possible answers.
There is simply no clue for how two different symbols i the first twoimages translate into the last image in the 3x3 matrix.
I tried it rowby row, and column by column, I did not get a "one single correct answer possibility",... Shape and filling remained ambiguous, and if row by row was treated, only the positions were "lower left and upper right".
Vote for Outsider on TWC: Image
charred steppes, borders of territories: page 59,
jump-map of local stars: page 121, larger map in Loroi: page 118,
System view Leido Crossroads: page 123, after the battle page 195

boldilocks
Posts: 669
Joined: Fri Jun 02, 2017 3:27 pm

Re: Miscellaneous Loroi question-and-answer thread

Post by boldilocks »

Krulle wrote:
Gorbash wrote:
boldilocks wrote:I'm not sure what kind of 'culture' would suffer some horrendous inability to evaluate this, for example:
Image

Unless that culture was universally unable to distinguish shapes, which would suggest that it's a problem that's not inherently cultural, because all cultures have outliers. (Perhaps all pygmys are far-sighted and so the image appears a blur, for example.)
Or their culture reads from top to bottom rather than left to right. I could easily see getting different answers on that question based on which block is assumed to be the start of the pattern.
Well, I saw two possible correct answers, but only one of them was in the list of possible answers.
There is simply no clue for how two different symbols i the first twoimages translate into the last image in the 3x3 matrix.
I tried it rowby row, and column by column, I did not get a "one single correct answer possibility",... Shape and filling remained ambiguous, and if row by row was treated, only the positions were "lower left and upper right".
Going top-bottom or left-right will in fact both lead you to answer #4. (Different rules, though.)
How these IQ test usually work from what I've observed, is that previous problems in the test will be simpler and set some easier rules that you then apply to increasingly complex problems that may add new rules of their own. This being Question B, it will most likely look ambiguous because the test assumes you've already done Question A.

Post Reply