Will majoring in (Applied) Mathematics in college increase IQ and make me smarter overall (generalize)?

[Sheldon Cooper]

I know that learning certain things, such as a foreign language and music, can actually increase overall intelligence (it generalizes): http://www.nytimes.com/2012/03/18/opinion/sunday/the-benefits-of-bilingualism.html?_r=0

Anyways, do you think learning (more) mathematics will give the same, generalized increase in mental capacity?

[Payman Saghafi]

Hello, Sheldon.

It is very possible for a person's IQ scores to increase through schooling.  This is a fact, and it is one of the least controversial topics in psychometrics.  Anyone who tries to tell you otherwise is dead wrong.   I'm rarely adamant about topics related to the human mind, but I am very confident about this particular issue.  You can go to the University of Minnesota and speak to some of the best psychometricians in the world, and they will accept this.  I went to the University of Minnesota and took a class under famous psychometrician, Thomas Bouchard.  He is a strong advocate of IQ tests and a staunch proponent of heredeterianism.  He reminds me of the infamous Arthur Jensen.  Even he would have to accept this.

Now for the BAD NEWS.  It can take a long time to increase your IQ scores substantially through typical education.  You have to be very active mentally.  As a hard-working intellectual, your efforts will make you an outlier.  Very few people have the interest or energy to study mathematics for years.  

Some very intellectually active people have shown legitimate increases in IQ scores of nearly 2 standard deviations between the ages of 20 and 50, for example.  (I am referring to score improvements that factor out test error and training effects.)

Notice that your IQ scores probably won't increase by a large amount in a single year.

Now, superficial scores on IQ tests might rise a lot in a short period of time.  It is pretty easy to increase your score on the WAIS IQ test if someone teaches you strategies.  This is not a legitimate increase in intelligence.  It is also possible to increase your SAT scores in a short period of time through advanced training.  Again, part of this is a training effect.

I scored over the 99th percentile on the GRE (Graduate Record Examination), which might technically put my IQ over 160 (with a standard deviation of 15).  In my case, this is an overestimate.  Part of the reason I got a score that high is that I am interested in the testing process.

Also, remember that someone who starts with a higher IQ at a younger age will have an advantage because he or she will perform better earlier in life and get a head start.  It is a huge advantage to have a very high IQ earlier in life.

Pay

[Payman Saghafi]

I should also mention that if you are in your 20's you can gain a few legitimate IQ points a year by staying in school.  So don't think that you have to wait 30 years to show improvements.  Just don't expect to gain 15 genuine IQ points in 1 week of studying mathematics.  :)

[1gn1t0r]

Could you please cite some sources supporting your argument?

[Lumos]

I don't think it's a very good idea to major in something because it has a small chance of increasing your IQ. But if I had to recommend a major to increase your overall brain power it would be philosophy or physics. Analytic philosophy (specifically) majors score highest in the verbal reasoning and analytic writing portions of the GRE exam. Physics majors score the highest on the quantitative reasoning portion and the highest out of all the hard sciences majors on verbal reasoning and analytic writing. Of course it could be the case that the students with the best verbal abilities choose  philosophy and the students with the best quantitative abilities choose physics. But that said, the skills you learn from say, doing philosophy seem to transfer the most widely out of all the majors (yet may be the least practical :D) But some transferable skills you learn may include critical thinking skills, good writing abilities, analytic skills, a solid grasp of logic and language.

http://www.physicscentral.com/buzz/blog/index.cfm?postid=6469561661568777605

[Mercel]

I've been wondering as well if demanding degrees (and later demanding
jobs) increases intelligence. I haven't found any answer to this
question yet however.

The only thing I know coming close to answering this question is that
the time spent doing e.g. mathematics and grey matter proved to be
proportional: http://www.ajnr.org/content/28/10/1859.full

[jttoto2]

It seems reasonable to assume, unless the person has incredibly poor time-management skills or is a complete dullard, that majoring in a math-heavy major will improve your scores in the quant portion of the GRE, with a similar effect with English-heavy majors on the verbal portion.  These effects are likely near-transfer though. 

[Bruno Ponce-Jones]

Such strong assertions do require citations. Could you please provide some evidence, please, Payman? Thank you.

[Payman Saghafi]

Here is some of the evidence.  This is not comprehensive:

1)  Lorge's study of 131 students    Mean effect: about 2 IQ points per additional year of school.
2)  Harnquist Study of 5,000 young people.  About 2-3 points per year.
3)  Jenck's study.  About 1 point per year.
4)  Wolfe study.  About 1 point per year.
5)  Lund and Thrane.  About 3 points per year.
6.  Husen and Tuijnman.  About 4 points per year.

7)  Neal and Johnson study
8) Degroot and Greene study.
9) Cahan and Cohen study.

10)  http://www.voanews.com/content/study-more-education-increases-iq-score-136593433/169492.html
You can go online and learn more about these studies

[Mercel]

Few dispute that schooling increases crystallized intelligence (which
usually is a substantial proportion of IQ tests). Fluid intelligence
on the other hand:

"In contrast, test scores on fluid intelligence tests (spatial and
logic tests) do not increase with additional days of schooling, but do
increase modestly with age."

http://ideas.repec.org/p/iza/izadps/dp6913.html

[Payman Saghafi]

I should also state that this was probably a more controversial topic back in the days of the "Bell Curve."  Hernstein and Murray did try to argue that IQ cannot be increased much by education.

I have strong opinions about this topic because claiming that IQ cannot be changed at all by education is an extreme stance.   I have always had a hard time taking extreme stances very seriously. After close examination, I usually find them to be off the mark.

Pay

[Payman Saghafi]

Also, keep in mind that the studies I cited show mean changes.  They do not tell us much about extreme outliers (like many of us on this web site) who are devoted to understanding and developing the mind.

[Payman Saghafi]

Exactly.  Moreover, crystallized intelligence is very important too.

[Zaraki]

I have strong opinions about this topic because claiming that IQ cannot be changed at all by education is an extreme stance.   I have always had a hard time taking extreme stances very seriously. After close examination, I usually find them to be off the mark.

Pay

I wouldn't consider it an extreme stance considering that nearly all the studies that have tried to increase intelligence through training have failed. Jaeggi doesn't count for much either, considering how riddled with holes her studies are.

I guess that what you have in mind is increasing intelligence by gathering knowledge rather than training WM though. I can think of numerous reasons as to why studies of the effects of schooling would fail, but I am not familiar with the studies you mentioned, so I will let it be for now.

[Payman Saghafi]

Hello, Zaraki.  I agree with some of your points.  Very short term training programs are certainly controversial (ie. dual-n-back).  Also, programs that trained people who did not really want to be trained or who had really bad life circumstances failed (ie. Headstart). 

Long-term studies (involving many years) with healthy people who CONTINUE to actively and willfully pursue education to better THEMSELVES have not failed.  This is why I presented you with so many studies.  You are free to go and review them.

You are right about the fact that much of the improvements seem to be in crystallized intelligence.  One caveat: Let's not make the mistake of assuming that crystallized intelligence is unimportant.  It is very important.  Sometimes it feels like people devalue the importance of this key aspect of the intellect.

Pay

[Sheldon Cooper]

@ Lumos, of course I'm not going to major in Applied Math just because it's going to increase my IQ; I also have a genuine interest in the subject. Right now I'm deciding between Applied Math and Computer Science - two of the hardest majors at Berkeley, but also two of the most interesting majors.

@ Mercel, interesting paper. So more mathematical aptitude = more grey matter?

[Zaraki]

On Saturday, October 20, 2012 8:12:27 PM UTC+2, Payman Saghafi wrote:

Hello, Zaraki.  I agree with some of your points.  Very short term training programs are certainly controversial (ie. dual-n-back).  Also, programs that trained people who did not really want to be trained or who had really bad life circumstances failed (ie. Headstart). 

Yes, lack of motivation in the experimental group is a confounding factor which renders those studies without an active control group even more unusable(since the effects varies depending on the researchers ability to motivate). They were hopefully equally inspiring to both groups for the studies that did contain active groups, but it's not a certainty.

Anyways, my reason for posting was mainly your opinion that saying IQ can't be changed is an extreme stance, which I consider consider to be an extreme stance(your stance that is) - ironically enough.

[Zaraki]

You are right about the fact that much of the improvements seem to be in crystallized intelligence.  One caveat: Let's not make the mistake of assuming that crystallized intelligence is unimportant.  It is very important.  Sometimes it feels like people devalue the importance of this key aspect of the intellect.

Oh right, I forgot about this. One reason for which I continually disagree with you is that I think knowledge might harm rather than do good if it is unrelated to the matter that you are trying to solve. A computer will start up programs more slowly if there are alot of other programs in the system, which would make the forgetfulness of the human mind useful. There are those who remember everything, but I don't think it's a gift. In some cases it starts after accidents in which trauma to the head occurs, which would suggest that a mechanism has been taken out of play, rather than into it. Why would such a mechanism exist if it wasn't important?

[Payman Saghafi]

Zaraki, I need clarification.  I want to make sure I am not defending against a straw man argument.  

A year of schooling can result in a small increase in crystallized intelligence for certain people (2-3 points). Schooling can also result in a modest increase in crystallized intelligence over a very long period of time for certain people.

In a nutshell, this is really what I stated.  Moreover, I also stated that denying this based on existing evidence is an extreme opinion.

Are you implying that my opinion above is extreme?  It is a very "middle of the road."

I am NOT arguing that overall fluid Intelligence can be improved quickly from dual-n-back.

I could understand why you might call THIS an extreme opinion. 

Also, your opinion that increased knowledge can overload the human mind and thereby decrease expert performance is concerning to me.  
I can brainstorm dozens of situations in which this is completely wrong.  Sure, I can think of some situations where it might be true.  For instance, if you are under tons of stress and studying for Jeopardy while also trying to do a great job at school, your mind can get overloaded with info and start making errors.   This type of overload is well known.

You are right.  We totally disagree on this point.

[exigentsky]

That's probably why most healthy brains have a forgetting mechanism. I
don't think you have to worry about learning more information slowing
you down. The brain doesn't work anything like a computer and you're
more liky to get the opposite result.

[Kevin Ann]

Hey Sheldon,

Don't smoke too much weed at Berkeley.  ;)

I got a phd in theoretical physics - which is essentially applied math.

I don't know enough to say that applied math may increase IQ, but it will certainly help you think better about things, even if that involves 'only' crystallized intelligence.  

Don't want to go too far off topic, but what's the point of having high horsepower if you can't direct it well?  Similarly, what's the point of having an extraordinarily high IQ (either 'natively' or through training) if you can't think about things well nor have a good base of knowledge in which to frame ideas?  For example, Bobby Fischer had an estimated 180 IQ and of course he was world champion in chess - an IQ intensive field.  However, outside of chess, he was really an idiot and said stupid things.

Good luck and enjoy your undergrad years, I wish I were back in college - very happy time in my life.  I probably would have majored in computer science - which is a fascinating field.

On Friday, October 19, 2012 9:59:57 PM UTC-4, Sheldon Cooper wrote:

[Mercel]

Before Kasparov was tested by a German team of psychologists, people
believed his IQ was around that of Fischer's estimation. However, they
conducted tests and concluded that his IQ was around 130 and that he
had an exceptional memory. I don't see why the same could be the case
with Fischer. Additionally, you insinuate that high intelligence comes
with disabilities in some respects, which I think is odd.

[Kevin Ann]

Nope, didn't want to insinuate that. Basically, I wanted to say that high intelligence alone is not necessarily a good thing in itself if you can't direct it well. It's like a powerful engine without a good driver, the 'driver' here being education and crystallized intelligence.

I have nothing against high intelligence. I think it's a great thing and I want to increase mine as much as possible, if possible.

[Kevin Ann]

Anyways about Kasparov, that's very interesting. Thanks for that tidbit.

Physicist Richard Feynman tested at 125 when he was in high school.  Years later he jokingly boasted (paraphrased): 

"winning a Nobel prize is impressive, but winning a Nobel with a 125 is even more impressive."

I suspect that Feynman and Kasparov IQ-achievement discrepancy are of a different nature. I bet Feynman's test was heavily loaded on verbal and was not really a good test of IQ. We're talking about one of the most creative physicists of the 20th century.

In general, seems to me that achievements in many fields is more of a 'threshold' effect in the sense that higher achievements are not correlated with higher IQ, 'provided' you are above some threshold. That threshold probably isn't very high either, probably like 130, and not 160.

[☉]

If a test can only measure so high, one is bound to see
extraordinarily intelligent persons pegged lower than expectations.
Feynman, who was self-reportedly pegged at around 125sd15, would have
definitely scored higher given a test which could legitimately measure
higher; I find it amusing he is often used as an example that I.Q.
tests are bunk. In a sense, they are bunk, but only when they're used
in ways for which they weren't designed. Most tests are intentionally
restricted to a little over two standard deviations above the average
(mean), so anyone who is more statistically anomalous will necessarily
appear "ordinary" by such measures.

argumzio

[☉]

Oh, I just saw your post after mine. I mentioned the misuse of the
Feynman example. It isn't unlikely that Kasparov falls in the same
line as Feynman, who certainly had an I.Q. higher than 125.

argumzio

[Mercel]

The test must have been heavily emphasized on language, obviously.
However, arguing for selectively choosing the substance of an IQ test
devaluates it. If everyone could pick and choose the substance of an
IQ test according to their own strengths and abilities, then the
higher percentile of the scale would end up having much higher IQs,
because lopsidedness is more common amongst the members of the right
side of the curve. For instance, when I was about being drafted into
the army as a teenager I scored so high on the verbal testing sections
that the officers wanted me to participate in a prestigious high
intensity Russian and Arabic language course instead of regular
military training. My mathematics score, on the other hand, was
absolutely horrendous, because I had not the slightest interest in
mathematics through middle school and high school. Coupled with a
tendency to choke during severe stress, I was certainly a mathematical
Illiterate.

[Payman Saghafi]

Mercel, I think you have a good point.  One thing to consider is that Feynman scored extremely low in "non-math" areas at certain points in his life. 

Of course, there could be many reasons for this.  See an interesting little debate about Feynman's intellect on the site below.

http://www.fraterslibertas.com/2010/06/power-line-albert-einstein-and-richard.html

[Lumos]

I don't think you can discount the fact that Feynman was a pretty big troll who took pleasure in messing with peoples' expectations. I'm especially surprised when fans who've read all his biographies take the 125IQ thing at face value. This was a man who did quantum physics in a strip club. 

[Payman Saghafi]

We  have to keep in mind that even if he did legitimately score 125 on an IQ test this wouldn't mean he was incapable of his great achievements.  125 is still almost two standard deviations above the norm.  I have seen people with learning disabilities score 85 on the verbal portion of the WAIS and over 130 on the performance section.  These people are outliers.  Feynman could have been a person who would have scored above the ceiling on nonverbal tests and only slightly above average on verbal tests.  This would make a score of 125 possible if he was a true outlier.  Imagine he scored 160 + on nonverbal tests and 110 on verbal tests.  This is certainly possible.

I am not arguing that this WAS the case.  I am only arguing that it isn't out of the realm of possibility.

We have to remember that IQ tests don't measure everything and that Feynman may have had weaknesses that didn't show up when he focused on physics.  Still, he was certainly a "brilliant" guy by my definition.

[Zaraki]

A year of schooling can result in a small increase in crystallized intelligence for certain people (2-3 points). Schooling can also result in a modest increase in crystallized intelligence over a very long period of time for certain people.

I guess I misunderstood the sentence I replied to then, because I thought you were saying that everyone is agreeing that IQ can be increased. I automatically think of fluid intelligence when speaking of IQ since it is the part that interests me the most. Sorry for the misunderstanding.

Exigentsky:

I know that the mind isn't like a computer, but it can have the same problem as one. Going through more(unnecessary) material can mean that it takes you longer to reach the conclusion.

[rgpddt]

On computers it depends entirely on the data structure used. If you know the memory location of a piece of data, then the amount of data stored does not affect the time it takes to retrieve that piece of data.

I don't think it's plausible to assume that if you have a lot of memories unrelated (childhood memories etc.) to a certain topic (physics for example), that they would then somehow slow down the search for information about that specific topic (physics).

[jotaro]

it is also a mistake to assume to human brain search for information inside.

human memory isnt computer memory ,we dont store the information

instead we generate it when the connections gets triggered.

check it out ,look at the cigar in the TV and you will remember how it smells.

look at the cake you eat everyday,you will remember its taste.

same with everything else just a more complicated version.

dont compare human beings to computers. 

[Zaraki]

Well, don't take the comparison too far. The retrieval and storage are different, I agree. But what I mean is that if unrelated data occurs in your mind then cognitive processes are spent on them. This should allow for greater creativity since it comes from vaguely related memories, at least from what I have seen, but they aren't important for solving the task. I.e. aliens are often very humanoid in movies featuring them. Often there is just one unrelated feature, like a fish head or squid arms, but these features are seldom unique; they come from vaguely related memories.

Concentration vs creativity in other words.

I am also wondering why there would be a mechanism for forgetting things if it doesn't slow cognitive processes down. The brain has the capacity to remember everything we have been through after all.

[☉]

Nope, you'll just know more applied mathematics.

argumzio

On Oct 19, 8:59 pm, Sheldon Cooper <angus.tm.h...@gmail.com> wrote:
> I know that learning certain things, such as a foreign language and music,

> can actually increase overall intelligence (it generalizes):http://www.nytimes.com/2012/03/18/opinion/sunday/the-benefits-of-bili...

[☉]

On that note, you'd probably have a better chance of increasing
functional I.Q. through computation, i.e., computer science, since
computers can do things humans can't do all that well. Cyborgs are,
supposedly, more intelligent for just that reason: linking an
autonomous pattern-recognition device (human) with a universal
computer would indeed have to be the best arrangement for human
intelligence being legitimately increased, for its powers of autonomy
would be extended, which would be another measure of intelligence
(getting things done) enhancement.

argumzio

[Payman Saghafi]

Funny.

[☉]

Is that a "logical error" or an intentionally unexplained quip?

argumzio

[☉]

Since the central question is fairly important, I suppose we should
consider actual research. Here are some selections.

"Arthur Jensen, IQ, and Intellectual Desire": http://www.mmisi.org/pr/23_01/lyons.pdf
"Can intelligence change?" http://cranepsych.edublogs.org/files/2010/08/Intelligence_EE.pdf
"How much can we boost IQ and Scholastic Achievement?"
https://docs.google.com/open?id=1MqaB5sNfr-8KQlinCrsH1ZIjoItDKOy7DV8fEsmR9s2W_K_0SPRcktTgQoR4
Blog article: http://abc102.wordpress.com/2011/02/17/spearmans-hypothesis-and-the-jensen-effect/

In the second article, the author suggest that the Flynn effect hints
at the possibility of I.Q. malleability. This is an error, because
I.Q. increases in accord with the Flynn effect are largely empty (not
a Jensen effect). Some history on the subject can be found here:
http://masi.cscs.lsa.umich.edu/~crshalizi/reviews/flynn-beyond/ . In
short, just because one sees an increase in I.Q. scores, one may not
therewith conclude an increase in intelligence. When one looks more
closely at the data, one does not see an increase in the standardized
scores (IQ's), what one really sees are mere increases in raw scores.
Absolute (raw) performance on a test, when normalized w.r.t. a given
population does not show any increase in I.Q. (which is the normalized
score). That is what the Flynn effect tells us, NOT that I.Q. scores
have increased.

If education were all that was needed to increase I.Q., wouldn't it be
reasonable to expect less of a difference across groups? Furthermore,
the claim that I.Q. will increase through education only has relevance
if a particular group (not the entire population) is equivalently
educated; however, the very fact that not everyone has the same set of
interests means that one will find such differences (e.g., in
knowledge) that will show up as differences in various I.Q. measures.

So, a more even-handed response to the question "Will X in college
increase I.Q.?" Not really, and only to the extent that the entire
population doesn't take an interest in the subject. You as an
individual can take a special interest in an arcane subject and quite
possibly learn special knowledge that will set you apart from the rest
of the population - and if such information happens to be relevant in
an I.Q. test, then you just might do better at that test on account of
that kind of knowledge.
For instance, let's say you have a sample of 10000 people and have
them take the GRE (fresh college graduates). You're likely to get a
normalized distribution of scores across the group. Then let's say you
select 100 of them to take 10 practice exams based on the format and
quality of the actual GRE. Have them (all 10000) take a second version
of the GRE. What do you expect to see? Will the selected 100 score
"higher" on the second run at the GRE? Yes, in terms of raw
performance. But will there be a demonstrated increase in I.Q. as a
result of this practice? Yes, because their I.Q. w.r.t. to the newly
normalized group will have been quite a deal greater than the 9900 who
did not take those 10 practice exams. Did they get "smarter"? Yes, but
only "smarter" in the sense that they performed at a higher level than
the other 9900. That's simply how I.Q. works.

If someone can point to a different scenario and legitimately claim
that an increase in I.Q. may result, then I'll take notice.

argumzio

[Payman Saghafi]

Zaraki has written some quips.  A few recent comments by Zaraki included errors in logic. 

I assume you know I wasn't referring to what you wrote.

Pay

[☉]

Ah, thanks for clarifying. :)

argumzio

[☉]

To clarify why I needed clarification: lately I have less time to read
others' responses than to type my own. Nevertheless, I seem to have
responded on a subject which was addressed prior to my response (about
cyborgs).

I found *that* funny, as well. :)

argumzio

[Payman Saghafi]

Argumzio, I want to get some clarification in regards to your recent post that included several papers on race, IQ and the malleability of intelligence.  I want to make sure that we are discussing the same thing.

Let's assume that person X takes 10 random IQ tests at the age of 20 and scores an average of 120 with a standard deviation of 15.  The test is normed against people his or her own age.  After 30 years of being an extremely mentally active person, this now 50 year old individual takes 10 different IQ tests that are normed against other 50 year olds.  His or her IQ now averages out at about 145.  Raw scores are irrelevant here.  After excluding variables like test error, disease, and other obvious confounding factors, it appears the person's higher crystallized intelligence has resulted in the ability to do better on a wide variety of culturally valued tasks that he or she has not specifically trained for.

Are you denying that the scenario above can happen as a result of increased crystallized intelligence? If so, I'm shocked.  Also, do you deny that the increases might be the result of the individual being more effective at solving a wide variety of problems, which are clearly relevant to society?

If you do deny that the person is "smarter" in a general and meaningful way, why do you deny it? 

Pay

[☉]

On Oct 22, 4:42 pm, Payman Saghafi <paymansagh...@gmail.com> wrote:

> Are you denying that the scenario above can happen as a result of increased
> crystallized intelligence?

No, increases to Gc is common knowledge. (If you start out
knowing X on a test of X,Y,Z, you'll only do so well on it; but then
if you learn Y and Z for another test of X,Y,Z, you'll do better on it
in comparison, while others may still only know X over the time
interval between testing sessions.) In fact, I'd say that's probably
the only way to increase one's ability (by knowing things, and even
more things). My GRE example plays into this fact.

> Also, do you deny that the
> increases might be the result of the individual being more effective at
> solving a wide variety of problems, which are clearly relevant to society?

If one knows something important to solving a problem, then one will
be "effective" at solving such problems. Hence, if one may know such
things and learn more such things, then one can become "more
effective" at solving more things over time. So no, I do not deny
this. Otherwise, I'd be denying that anyone can learn, which would be
ridiculous.

> If you do deny that the person is "smarter" in a general and meaningful
> way, why do you deny it?

I.Q. is not the same thing as being "smart", but they may be rather
strongly correlated. (Anyone have a Smartness Test?) However, I would
deny that anyone can straightforwardly increase their ability to learn
things. Even memorization techniques only go so far (from what I've
been told). After all, one cannot - or at least not that well -
memorize how to reason through complex problems about which one does
not have prior knowledge. One may learn various algorithms,
techniques, methods, and so on, but that is a lower order than being
posed with problems about which one has no prior experience, i.e.,
genuinely novel problems. Ever heard of "educated incapacity" a.k.a.
"intellectual stupidity"? It is akin to expert blindness and such
problems show the real material of which one is made. I think any
intervention which claims to increase the hard core of ability should
take a look at how someone can rise above the limitations of their
awareness to solve truly novel problems, rather than teaching them how
to handle overly familiar and straightforward tasks. After all, even
nowadays, computers cannot be said to possess intelligence (yet) - but
they clearly can solve whatever it is we've managed to turn into an
algorithm (computer program).

argumzio

[Payman Saghafi]

As you probably know, Jensen and test maker "ETS" argue that making significant improvements in scores on the GRE is not really possible for most people.  ETS has argued that self-study for the GRE results in almost no improvement in score and they have presented research studies making this claim.  Hernstein and Murray of the Bell Curve would make the same argument.  Ironically, this type of argument mirrors the argument that "functional intelligence" cannot be increased.

So, according to these parties, and assuming there is no cheating, it is almost impossible to improve performance on standardized tests such as the SAT or GRE.

I strongly disagree with them on this.

Pay

[Payman Saghafi]

But Argumzio, the same people who make the argument that IQ cannot be increased meaningfully are often the same people who claim that scores on the SAT and GRE cannot be increased.  You don't find that odd?

Pay

[Payman Saghafi]

Ok, your last post was very elucidating. I feel we are really getting somewhere now.

I understand what you are implying with regards to "novel" problems.  What I am arguing is that almost all thinking skills are to some degree linked with crystallized skills.  As a result, a person with a tremendous amount of important crystallized knowledge, skills, and insight will perform very well on a wide variety of novel tasks that are demanded by society.  This is because even novel tasks will often have some resemblance to intellectual skills that are valued by society at large.  For instance, years of improving reading comprehension will also help a person put together better answers to "similarities" questions on the WAIS.  It will also help a person have more complex thoughts on something as random as "the ethics of polygamy" or "whether Einstein did more for society than Newton." The more advanced your crystallized abilities, the more you can EXTEND them to novel tasks.

Moreover, a 50 year old who is very strong in advanced crystallized skills might "GENERALLY" perform as well intellectually as a 50 year who is strong in fluid intelligence, but limited in crystallized knowledge.  For all practical purposes, I view these two people as equally smart because one cannot become the other in any short period of time. 

Now, you might argue that the person with higher fluid ability had "more potential originally" but that potential cannot be immediately transferred to results.  It took 30 years for the person with less potential to develop his or her intellect.

This is my primary point that I am really trying to get across to people. 

Pay

[Zaraki]

You are awfully quick with proclaiming that I have made logical errors, but your examples are woefully inadequate to prove them as such. I didn't mean that you would have trouble recalling a simple fact, but that an abundance and facts and memories irrelevant to the task you are trying to solve slows you down. And again, I am not talking about recalling a memory, but of solving complex problems.

[Payman Saghafi]

Fair enough.  I guess I should have written that the brevity of what you wrote did not provide clarification of what you meant.   I did not mean to offend.  Peace.

[Payman Saghafi]

You know what, Zaraki, I reread the paragraph I wrote and it comes off as more patronizing than I originally thought.  I'll delete it.

[jttoto2]

I believe I heard that as well but I can't find the link.  It is funny that the makers of the LSAT say prep time is extremely important, and I'd say that is a much harder test IMO.  Perhaps the test-makers and Jensen are making false assumptions based on current data.  If the majority of those taking the GRE have already done the optimal amount of prep-time before their first test, then it would seem that preparation has no importance because they can only go by test and retest results. Before and after course tests from Kaplan or some equivalent would be a more accurate indicator if preparation has any effect.

[Kevin Ann]

Payman and Zaraki, I like your posts and I love this group.

One of the reasons I used my real identity is that I find that anonymity tends to bring out certain behaviors in people, especially when it's so easy to misinterpret what another said and to write something that can be misinterpreted. 

I find that I never really argue with anyone with my real identity, so I don't end up wasting time and effort on pages and pages of flaming (in other message boards).  :)

[jttoto2]

I would also like to add, that I believe that the ETS makes such bold claims because it give more validity to their test as a measure of individual potential.  I wouldn't put too much stock into such claims without harder evidence.

[Payman Saghafi]

Jttoto:

Here is one of the papers by ETS claiming that it is even extensive preparation DOES NOT improve GRE scores.

http://www.ets.org/Media/Research/pdf/RR-85-12-Swinton.pdf

[jttoto2]

"The  results  suggest  that,  for  the  verbal  and  quantitative  sections  of 

the  test,  additional  test  familiarization  activity  beyond  that  normally 

undertaken  by  typical  GRE test  takers  will  probably  not  result  in  higher 

test  scores.  We emphasize  that  this  conclusion  applies  only  to  typical 

examinees  who  engage  in  some  preexamination  preparation  and  only  to  effects 

on  test  scores."

So we still can't safely conclude that preparation does nothing, only that additional preparation beyond what people normally do does nothing.  I suspect that average person taking the GRE had math-heavy (engineers, natural sciences, hell even business from a decent school) majors which would make you well prepared for the quant section as well.  I suspect there is  an optimal prep time, with the quality of undergrad coursework also having an effect.  

This effect does not hold true on the analytical portion though (akin to logic games).  More prep time improves scores even more.  No wonder they took it out.

[Payman Saghafi]

Yes, and keep in mind that this is only one of the papers they have published on this subject.  Others take a more firm stance.

Nonetheless, the tone of what ETS wrote will make a typical test taker think, "as long as I have practiced some GRE problems, and brushed up on my math, additional preparation will probably not improve my score, even if I spend a significant amount of time practicing."

This is clearly the message that ETS wants to send out with their slightly ambiguous statement. 

Pay

[jttoto2]

And what is your opinion?  Do you personally think one can prove one's test score, provided that they studied effectively? (sometimes simply putting time into something is not enough, and yes the question is not rhetorical.)

[jttoto2]

I agree with you that this is the tone they want to put out.  I disagree with their conclusions.  Problems with their data:

-  The sample on average was likely moderately prepared. Hell, they could have been well-prepared even. We can't say for sure since they don't go into the specifics of prep-time before the intervention.  Or perhaps I overlooked it (it was a long paper)

-  Reliance on self-report measures of prep-time.  Not only can this be unreliable, but how do we distinguish from those who used their time wisely to those who needed more time because they were goofing off while they were studying?  In college I was one of those guilty of using my time spent studying woefully inefficiently, but I would still say I studied 10 hours for an exam, even if I really was using only 3 of those 10 hours.  I'm not saying others do this, but that is an example of methodical problems when relying on self-reports.

[jttoto2]

methodological problems when relying on self-reports*

[Payman Saghafi]

I strongly believe that for those students who are motivated to study the test design, to improve their vocabulary, to calm their anxieties, and to practice old tests over the course of a couple of months, significant improvements can be made.

Now, obviously natural talents will limit how much people can improve.

I'm just one example, of course.  But, as I've mentioned before, I basically got a perfect score on the GRE, which would put my IQ at an estimated 160 for this particular test.  I could take the new version of the GRE tomorrow and do just as well.  Now, I know for a fact that my IQ is NOT that high!  I'm definitely dumber than this. :)

I discuss the GRE quite a bit on my site and offer insight into the test.  You might assume that I am biased.  :)  I strongly believe that one of the biggest issues for people with the GRE is what I call the "working memory trap."  If people panic, scores drop quite a bit:

http://www.youtube.com/watch?v=1SUCPjpZUac&feature=youtu.be

Also, if you want proof that a typical person can improve his or her scores on verbal quite a bit, then look at the vocabulary list below.

http://quizlet.com/47571/barrons-gre-wordlist-4759-words-flash-cards/

Have the foreign student do a little preparation for the GRE and take the test.  Then, give this list to the foreign student and have him or her memorize the list for a couple of months.  Have him or her practice a bunch of GRE tests.  Then, compare his or her first "moderate preparation score" to his or her second "extensive preparation score." Note the significant improvement.

The same types of strategies can be used to improve quant scores quite a bit.

ETS is wrong.

[Payman Saghafi]

Jttoto,  I actually did research into the vocabulary level you would need to have to be at the minimum threshold to get a particular score on the GRE.  See these videos on the side bar of my web site if you are curious to see the level of vocabulary generally needed.  I actually looked at populations of people taking professional IQ tests and did some cross correlations:

http://www.gkacademy.org/

[☉]

Payman Saghafi <paymansagh...@gmail.com> wrote:
> ETS has argued that self-study for the GRE results in almost no improvement in score and they have presented
> research studies making this claim.

If I'm not mistaken, I believe they refer to no improvement in the
normalized score (where it counts) as opposed to the raw score (which
means nothing outside of a statistically meaningful context).

> Hernstein and Murray of the Bell Curve would make the same argument.

I think one would do well not to confuse pre-normalized scores with
normalized ones. This often happens when people don't know what
they're talking about.

> Ironically, this type of argument mirrors the argument that "functional intelligence" cannot be increased.

If "intelligence" is understood in purely operational terms, one would
have to concede that this isn't ironic, but merely a matter of course.
Going back to any number of examples, one's functional intelligence as
measured with respect to a given population (otherwise one will not
have any sensible definition of intelligence at the outset) really
cannot change, except insofar as one engages in a sort of intervention
which a portion of that population is not subjected.

I will not and do not argue that group differences cannot be closed so
that no gaps remain, but anyone who seriously claims to put forward a
program to do this has a lot of work ahead of them. However, I detect
some sort of confusion in your messages, particularly on the matter of
not grasping the difference between pre-normalized (unstandardized)
scores and normalized (standardized) scores. Forgive me if I
misunderstand you, but it is a difference of paramount importance. No
one who works in the field denies the existence of confounding
variables which make operational measurements less accurate, but that
isn't the same as saying such confounds cannot be controlled.

> But Argumzio, the same people who make the argument that IQ cannot be
> increased meaningfully are often the same people who claim that scores on
> the SAT and GRE cannot be increased. You don't find that odd?

No, because they're referring to standardized scores, not raw scores.
Raw scores are meaningless statistically. We're only so far in finding
methods to measure "intelligence" as found in the neurons, axon
fibres, interconnections, synapses, biochemical interactions, etc.
Studies involving tests of the kind we currently use to evaluate
ability again and again show no gains in normalized scores. (Even the
studies involving DNB do not actually use properly standardized tests
as they were originally developed.) Going back to my example with the
GRE: I could have said that we should give all 10000 persons ten
practice exams instead of to only 100/10000 of them. In that scenario,
do you honestly believe no gaps between individuals would remain such
that everyone would be at the same level of performance? Of course
not. One would find a similarly normalized distribution of scores
prior to the ten practice exams. The only significant difference would
probably be an upward increase in the mean and less variance in the
scores (greater precision w.r.t. the mean). This is what the data
show. No one denies that raw scores (absolute performance) can be
increased; but that is enormously different from saying that the
standardized scores can be altered significantly.

> What
> I am arguing is that almost all thinking skills are to some degree linked
> with crystallized skills.

Yes, but only insofar as such thinking is *skill based*. I argue it
isn't entirely skill based. And this should be obvious.

> As a result, a person with a tremendous amount
> of important crystallized knowledge, skills, and insight will perform very
> well on a wide variety of novel tasks that are demanded by society.

Novel only to the extent that such problems may be reduced to familiar
tasks. Did you even look into the issue of "intellectual stupidity",
or "expert incapacity"? It is almost as if you're ignoring the
failures of those who cannot see beyond their own acculturated
knowledge.

> This
> is because even novel tasks will often have some resemblance to
> intellectual skills that are valued by society at large.

And what if they don't? What if the problem, even if similar to
something familiar (but that would be to say it *isn't* novel), cannot
be solved except by some method no one actually knows, so that the
method must be invented? Is there a method for inventing methods?
(Partly facetious, but I hope you grasp the point.)

> For instance,
> years of improving reading comprehension will also help a person put
> together better answers to "similarities" questions on the WAIS.

That is a very tame example. You're entitled to whatever examples you
like, but I must speak to a plain fact: the WAIS is a human
construction. Do you honestly believe all problems humans face are
humanly constructed problems? It may be admitted that human problems
are uniquely human, but the problems we find out there in reality are
a great deal less kind to the pretensions of human intellect. Think of
an example, if you prefer.

> It will
> also help a person have more complex thoughts on something as random as
> "the ethics of polygamy" or "whether Einstein did more for society than
> Newton."

That may very well be. But you sound as if you're only wishing to
develop a more literate society. Unfortunately, the "first world"
hasn't much need of increased literacy - although some would lament
its unfortunate degradation, so to speak - and even those parts of the
world which *do* need it aren't exactly wanting in unique methods to
aid them. No, such parts of the world are deprived of the stimulation
they have a right to obtain, which is why they are less developed than
they reasonably could be; however, this is not to imply that increased
stimulus will necessarily result in a systematic and universally
distributed increase in a reliably educated populace. Many factors are
at play, so I must hold back for fear of being required to write a
quarto volume on the topic.
That said, it is clear to me you have some sort of project you wish to
defend here. I have no qualms with you trying to help the guy on the
street get into a good university and make a comfortable lot in life.
Heck, I know a guy who practiced over a 100 SAT exams and finally got
a near-perfect score (on the old SAT). Such dedication is a measure of
overcoming difficulties (or a sign of obsession), but whatever the
case, if everyone had the time, energy, and commitment for 100 SAT
practice exams, then the world would be a radically different place to
begin with.

> The more advanced your crystallized abilities, the more you can
> EXTEND them to novel tasks.

Not so. You do not recognize true novelty as it arises - I do not mean
anything less than true novelty. You're probably one of those people
who is good at handling pseudo-novelty by reasoning patterns an order
above most people, but in any case, no one taught you how to reason in
that fashion. (I'm assuming you don't always find overly familiar
problems.)

Nevertheless, even those who possess the requisite knowledge to solve
a problem do not always manage to solve that problem. Finding ways of
looking at a problem differently than it might initially be conceived,
creating new methods, and so on goes a long ways to solving
intractable problems. Knowledge isn't enough for all cases that exist,
and even if it were, that is no guarantee that someone will apply that
knowledge well enough to solve a problem of some difficulty. Even very
intelligent persons hit a brick wall over which their minds cannot
leap - until someone spontaneously comes along and walks around it.

> Now, you might argue that the person with higher fluid ability had "more
> potential originally" but that potential cannot be immediately transferred
> to results. It took 30 years for the person with less potential to develop
> his or her intellect.

Yes, and it might only take that higher-Gf person 5 years to do the
same. That's partly the difference between the two. More to the point,
the guy with less knowledge just might be able to solve problems the
knowledge-laden guy cannot. Light helps one see, but only when it
isn't too bright to blind one's eyes.

argumzio

[Colin Dickerman]

Haha, I remember Fischer called Kasparov an idiot savant.

On Oct 20, 5:47 pm, Mercel <st.gjeng...@gmail.com> wrote:
> Before Kasparov was tested by a German team of psychologists, people
> believed his IQ was around that of Fischer's estimation. However, they
> conducted tests and concluded that his IQ was around 130 and that he
> had an exceptional memory. I don't see why the same could be the case
> with Fischer. Additionally, you insinuate that high intelligence comes
> with disabilities in some respects, which I think is odd.

[Colin Dickerman]

I think it will make you more intelligent because you'll have a larger
knowledge base to draw on when solving problems. The person with a
masters degree will tend, I think, to make better decisions than the
person that went to work out of high school.

The knowledge you've spent years soaking up will exude into the rest
of your life. I draw on the things I've learned all the time.

[Payman Saghafi]

Interestingly, we actually agree on a lot, but we emphasize different facts.

I will not and do not argue that group differences cannot be closed so
that no gaps remain, but anyone who seriously claims to put forward a
program to do this has a lot of work ahead of them. However, I detect
some sort of confusion in your messages, particularly on the matter of
not grasping the difference between pre-normalized (unstandardized)
scores and normalized (standardized) scores. Forgive me if I
misunderstand you, but it is a difference of paramount importance. No
one who works in the field denies the existence of confounding
variables which make operational measurements less accurate, but that
isn't the same as saying such confounds cannot be controlled.

I forgive you.  :)  Seriously, we are on the same page, argumzio. Let me tell you that I am not confusing raw scores with standardized scores, at all.   I enjoy the language of statistics and am very comfortable with it.  My problem with ETS is more about what they try to imply than what their research actually shows.

If "intelligence" is understood in purely operational terms, one would
have to concede that this isn't ironic, but merely a matter of course.
Going back to any number of examples, one's functional intelligence as
measured with respect to a given population (otherwise one will not
have any sensible definition of intelligence at the outset) really
cannot change, except insofar as one engages in a sort of intervention
which a portion of that population is not subjected.

Agreed. 

And what if they don't? What if the problem, even if similar to
something familiar (but that would be to say it *isn't* novel), cannot
be solved except by some method no one actually knows, so that the
method must be invented? Is there a method for inventing methods?
(Partly facetious, but I hope you grasp the point.)

This is most certainly true in the sense that there are limits to how far
crystallized intelligence can take someone.  This is especially true
in situations where very high levels of fluid ability are needed for abstractions
that are clearly beyond the capacities of an individual's gray matter.

Yes, and it might only take that higher-Gf person 5 years to do the
same. That's partly the difference between the two. More to the point,
the guy with less knowledge just might be able to solve problems the
knowledge-laden guy cannot. Light helps one see, but only when it
isn't too bright to blind one's eyes.

Now, this is something that I definitely know and accept, despite the fact that it
thoroughly displeases me. I simply argue that someone with moderate intellectual potential can, with hard work,
function at a higher level of overall cognitive ability than he or she originally did, and that this
higher level of performance, for all practical purposes, equates to what most any educated and reasonable non-psychometrician
might think of as manifestations of  "increased general intelligence." 

pay

[Zaraki]

Hmm, odd. I agree with you on your latest post, Payman. I usually get the image that you think crystallized intelligence can transfer to the realm of fluid intelligence, or work in its place. Perhaps its more of a minsunderstanding then. If you mean that crystallized intelligence can provide near-transfer to better your reasoning ability then we are on the same page. The quote about standing on the shoulders of giants is fitting here.


Kevin Ann: I would be equally anonymous while using my real name since I doubt anyone knows me here. It was not my intention to flame; I find "haters" and "flamers" childish.

[Payman Saghafi]

Hi, Zaraki.  Yes, I have never implied that crystallized intelligence supplants fluid intelligence.  I would be naive to argue that fluid abilities are not important.  If you are not capable of conducting the visual thought experiments of Einstein today, then the odds of you suddenly being able to do so tomorrow are extremely low.  :)

I repeat what I wrote, above.  This is probably the best representation of what my conclusion is.

I simply argue that someone with moderate intellectual potential can, with hard work,
function at a higher level of overall cognitive ability than he or she originally did, and that this
higher level of performance, for all practical purposes, equates to what most any educated and reasonable non-psychometrician
might think of as manifestations of  "increased general intelligence." 

Now, I understand the limitations of enhancing overall functional intelligence through the increase of crystallized intelligence alone!  I'm not that crazy.  I just don't want us to overextend the limitations.

Herein lies my passion, and it is a passion which sprouted only after careful examination of facts.

Pay