Answers to the Titan Test (a phoney IQ test)
gregw...@yahoo.com
world and have formed a large number of pretentiously-named clubs to
revel in their supposed intellectual superiority. This is despite very
little (or rather absolutely nothing) in the way of actual intellectual
accomplishments among them to substantiate their incredible claims of
intelligence.
Their main club is the "Mega Society" which they say is reserved for
those with one-in-a-million IQ's, ie above 176. Of course for good
reason conventional IQ tests do not have a scale that reaches such a
height, so they have created what they claim is a test that can measure
the intelligence of those with very high IQ's.
When I first heard about this four years ago, I decided to put a stop
to this by publishing the answers to their current test, the Titan Test
(notice again their extreme vanity in choosing this name). This had
been done before, to their Mega Test, which they had to then abandon as
a criterion of membership.
After spending an afternoon working on the Titan Test, I came up with
answers to a good part of it, certainly enough to "compromise" it, to
use their phrase for publishing the answers. However I lost interest
soon afterward and never got around to posting these answers until now.
I did not attempt to have my answers scored by the author of the test
for his $30 fee, some questions have more than intelligent answer, and
I have no way of knowing the intended answer so I put what I thought
was the best.
The questions are here
http://www.eskimo.com/~miyaguch/titan.html
And their other tests and other related crap is here
http://www.eskimo.com/~miyaguch/
Here are my answers along with a few parenthetical comments and
explanations
1. STRIP : MึBIUS :: BOTTLE : KLEIN
3. LACKING MONEY : PENURIOUS :: DOTING ON ONE'S WIFE : UXORIOUS
(definitions)
4. MICE : MEN :: CABBAGES : KINGS (from literature)
6. ALL IS ONE : MONISM :: ALL IS SELF : SOLIPSISM (a doctrine in
philosophy, nothing exists except the self, or alternately the
existence of everything else depends on the existence of the self)
9. HOLLOW VICTORY : PYRRHIC :: HOLLOW VILLAGE : Potemkin
14. LEG : AMBULATE :: ARM : Rotate
15. MOSQUITO : MALARIA :: CANNIBALISM : KURU (a disease limited to the
Fore tribe of New Guinea, who eat human brain in their religious
rituals)
16. HEAR : SEE :: TEMPORAL : OCCIPITAL (Lobes of the brain that deal
with hearing and sight.
17. ASTRONOMY AND PHYSICS : ASTROPHYSICS :: HISTORY AND STATISTICS :
CLIOGRAPHY (Cliography is the use of statistics or economics in
historical studies.)
18. JEKYLL : HYDE :: ELOI : MORLOCKS (from the books by Robert Louis
Stevenson and H.G. Wells)
20. SET OF SETS NOT MEMBERS OF THEMSELVES : RUSSELL :: DARKNESS OF THE
NIGHT SKY IN AN INFINITE UNIVERSE : KEPLER (concepts the men wrote
about)
22. LANGUAGE GAMES : LUDWIG :: PIANO CONCERTI FOR THE LEFT HAND : PAUL
(So obscure it is silly. Ludwig and Paul Wittgenstein were brothers.
One was a pianist, the other was the famous philosopher who used the
phrase "language games" in his book Philosophical Investigations.
Knowing this meaningless bit of trivia can hardly be an indication of
great intelligence.)
23. IDOLS : TWILIGHT :: MORALS : GENEALOGY (Titles of Nietzsche's
books)
24. SWEET*NESS* : SUFFIX :: BOAT*SWAIN* : STEM ("ness" is a suffix,
"swain" is the stem of the compound word "boatswain."
29. The answer is 20.
31. The answer is 22.
38. I am pretty close to sure the answer is 1/27
47. 3 (The list is of the digits of pi/4)
gregw...@yahoo.com
(then 66238993967)
Danny Kodicek
<gregw...@yahoo.com> wrote in message
news:1112382703.2...@l41g2000cwc.googlegroups.com...
Their main club is the "Mega Society" which they say is reserved for
those with one-in-a-million IQ's, ie above 176. Of course for good
reason conventional IQ tests do not have a scale that reaches such a
height, so they have created what they claim is a test that can measure
the intelligence of those with very high IQ's.
What I find interesting about this is how little any of these questions have
to do with IQ - they're almost all tests of vocabulary and knowledge.
Danny
gcrh...@yahoo.com
You didn't mention that Potemkin was one of Catherine the Great's
many lovers who tried to impress her by building a really impressive
looking village (really an entire countryside) that turned out to
be a grand illusion.
> 14. LEG : AMBULATE :: ARM : Rotate
>
> 15. MOSQUITO : MALARIA :: CANNIBALISM : KURU (a disease limited to
the
> Fore tribe of New Guinea, who eat human brain in their religious
> rituals)
>
> 16. HEAR : SEE :: TEMPORAL : OCCIPITAL (Lobes of the brain that deal
> with hearing and sight.
>
> 17. ASTRONOMY AND PHYSICS : ASTROPHYSICS :: HISTORY AND STATISTICS :
> CLIOGRAPHY (Cliography is the use of statistics or economics in
> historical studies.)
>
> 18. JEKYLL : HYDE :: ELOI : MORLOCKS (from the books by Robert Louis
> Stevenson and H.G. Wells)
>
> 20. SET OF SETS NOT MEMBERS OF THEMSELVES : RUSSELL :: DARKNESS OF
THE
> NIGHT SKY IN AN INFINITE UNIVERSE : KEPLER (concepts the men wrote
> about)
>
> 22. LANGUAGE GAMES : LUDWIG :: PIANO CONCERTI FOR THE LEFT HAND :
PAUL
> (So obscure it is silly. Ludwig and Paul Wittgenstein were brothers.
> One was a pianist, the other was the famous philosopher who used the
> phrase "language games" in his book Philosophical Investigations.
> Knowing this meaningless bit of trivia can hardly be an indication of
> great intelligence.)
Indeed, I never heard of Paul Wittgenstein.
The first thought that jumped into my head was Maurice for
Maurice Ravel, the composer of the most well known piano
concerto for the left hand. I just looked it up and this
piece was commissioned by Paul Wittgenstein! who lost his
right arm in WWI.
> 23. IDOLS : TWILIGHT :: MORALS : GENEALOGY (Titles of Nietzsche's
> books)
>
> 24. SWEET*NESS* : SUFFIX :: BOAT*SWAIN* : STEM ("ness" is a suffix,
> "swain" is the stem of the compound word "boatswain."
>
> 29. The answer is 20.
>
> 31. The answer is 22.
>
> 38. I am pretty close to sure the answer is 1/27
I get 2/27 (each of the three distinct cycles is
traversable either forwards or backwards).
> 47. 3 (The list is of the digits of pi/4)
48. pi^2 r^4 / 2 The hyper-volume of a 4-dimensional hyper-sphere.
(I don't know why the first term isn't zero.)
mensa...@aol.com
gregw...@yahoo.com wrote:
> Some very conceited people claim to be among the very smartest in the
> world and have formed a large number of pretentiously-named clubs to
> revel in their supposed intellectual superiority. This is despite
very
> little (or rather absolutely nothing) in the way of actual
intellectual
> accomplishments among them to substantiate their incredible claims of
> intelligence.
Really?
>
> Their main club is the "Mega Society" which they say is reserved for
> those with one-in-a-million IQ's, ie above 176. Of course for good
> reason conventional IQ tests do not have a scale that reaches such a
> height, so they have created what they claim is a test that can
measure
> the intelligence of those with very high IQ's.
Don't those who are actually intelligent realize that the only
thing such tests measure is how high a score you can get on
the test?
>
> When I first heard about this four years ago, I decided to put a stop
> to this by publishing the answers to their current test, the Titan
Test
> (notice again their extreme vanity in choosing this name).
That'll show 'em!
> This had
> been done before, to their Mega Test, which they had to then abandon
as
> a criterion of membership.
But "Meganator" just doesn't have the same ring to it.
Woody Brison
number 25 in eleven steps
In the following diagram, on the left I show how I can
place one large (blue) square on top of another (yellow)
square.
http://www.geocities.com/woody_brison/mega25.gif
On the right I show how I follow that with nine more
squares.
If it's required to make the pattern exactly then two
more large squares can be placed at the top and at the
left to give the overall boundaries, two more squares
totals 13.
If the rules don't allow anything to stick out as in the
diagram on the left, then I don't think it can be done at
all with the pieces given. Is the essence of the question
being able to think outside the box?
Also, it looks to me like this organization, whatever it
is, just collects your fee of $30 and gives you membership
in return. It would be up to you to decide whether it's
worth $30. But someone coming along and compromising the
test... how would that hurt them? Now they'd have more
people who could pass the test and be eligible for
membership, i.e. more $30 fees paid. If you are trying
to deflate them, I suspect you need to think of some
other strategy
Wood
gregw...@yahoo.com
journal they seem to be concerned that their tests have been
compromised. I don't think that they are in this for the money, but out
of vanity.
gregw...@yahoo.com
I get 2/27 (each of the three distinct cycles is
traversable either forwards or backwards)."
I figured out 1/27 four years ago, so I can't tell you how I got it.
How sure are you? I wrote in my notes I was 95% sure but I didn't write
how I got to it or the cause of my doubt.
---
< 48. pi^2 r^4 / 2 The hyper-volume of a 4-dimensional hyper-sphere.
(I don't know why the first term isn't zero.) >
I should have seen that one myself. it comes after the area of a circle
and the volume of a sphere.
Keep the answers coming!
Michael Mendelsohn
> Some very conceited people claim to be among the very smartest in the
> world and have formed a large number of pretentiously-named clubs to
> revel in their supposed intellectual superiority.
Some very conceited people claim to be among the very fastest in the
world and have organized a quatrannual event to revel in their supposed
athletic superiority.
> This is despite very
> little (or rather absolutely nothing) in the way of actual intellectual
> accomplishments among them to substantiate their incredible claims of
> intelligence.
This is despite very little (or rather absolutely nothing) in the way of
actual physical accomplishments among them to substantiate their
incredible claims of athleteship.
> Their main club is the "Mega Society" which they say is reserved for
> those with one-in-a-million IQ's, ie above 176.
Their main event are the Olympic games, which they say are reserved for
one-in-a-million runners who do the 100m dash in less than 10 seconds.
> When I first heard about this four years ago, I decided to put a stop
> to this by publishing the answers to their current test, the Titan Test
> (notice again their extreme vanity in choosing this name). This had
> been done before, to their Mega Test, which they had to then abandon as
> a criterion of membership.
Wow! What an achievement!
May I call you "spoilsport extraordinaire" now?
> I did not attempt to have my answers scored by the author of the test
> for his $30 fee, some questions have more than intelligent answer, and
> I have no way of knowing the intended answer so I put what I thought
> was the best.
I wasn't able to run the 100m dash that fast, so I'll just post how far
I could go in 10 seconds.
> Here are my answers along with a few parenthetical comments and
> explanations
At least two of them are wrong.
Danny Kodicek wrote:
> What I find interesting about this is how little
> any of these questions have to do with IQ -
> they're almost all tests of vocabulary and knowledge.
Well, the test certainly isn't culture-independent.
It assumes that intelligence is proportional to the ability to learn,
and a motivation to do so; by this reasoning, more intelligent people
ought to have a greater store of general knowledge and vocabulary
pertinent to the culture they live in. Recall ability plays a role, too.
Cheers
Michael
--
It's silly talking about how many years we will have to spend
in the jungles of Vietnam when we could pave the whole country
and put parking stripes on it and still be home by Christmas.
-- Ronald Reagan, October 10, 1965
Danny Kodicek
"Michael Mendelsohn" <inv...@msgid.michael.mendelsohn.de> wrote in message
news:424E4DE0...@msgid.michael.mendelsohn.de...
> gregw...@yahoo.com schrieb:
> Danny Kodicek wrote:
> > What I find interesting about this is how little
> > any of these questions have to do with IQ -
> > they're almost all tests of vocabulary and knowledge.
>
> Well, the test certainly isn't culture-independent.
> It assumes that intelligence is proportional to the ability to learn,
> and a motivation to do so; by this reasoning, more intelligent people
> ought to have a greater store of general knowledge and vocabulary
> pertinent to the culture they live in. Recall ability plays a role, too.
Ah yes, the standard explanation to why success on intelligence tests
correlates with education: 'because people with high intelligence are more
likely to stay in school'. Isn't it at least plausible that the causality
goes the other way?
Danny
Mark P
> Here are my answers along with a few parenthetical comments and
> explanations
Here are a few more...
>
>
> 1. STRIP : MึBIUS :: BOTTLE : KLEIN
>
2. THOUGHT : ACTION :: OBSESSIVE : compulsive
37. This can be done via Bayes. Let allW be the event that all marbles
are white, let 10W be the event that we draw 10 white marbles in our
first ten draws. We want to know:
P(allW | 10W) = P(10W | allW) * P(allW) / P(10W)
= 1 * 2^-10 / P(10W)
= 2^-10 / P(10W)
That only tricky part is figuring out P(10W) which is not, btw, simply
2^(-10) because there are correlations. This can be done by considering
the 11 possible distributions of numbers of each color marble, the
corresponding probability of each such distribution occurring, and the
probability of 10W in each case. Summing these gives:
P(10W) = 2^(-10) [10c10 * (10/10)^10 {all marbles white} +
10c9 * (9/10)^10 {9 white marbles} + ...
10cK * (K/10)^10 {K white marbles}]
where k = 10, 9, ... 1 though smaller k values make negligible
contributions.
I punched this into a calculator and, barring any mistakes, I get 7% as
the probability that all marbles were white.
> 38. I am pretty close to sure the answer is 1/27
I get 2/27. Each ant has 3 choices for its path. The first ant's
choice is arbitrary. The ant towards which the first ant crawls has 2/3
viable choices. The remaining two ants have 1/3 viable choices each.
39. I get 4 / 3^7 = 4/2187
Michael Mendelsohn
> > Danny Kodicek wrote:
> > > What I find interesting about this is how little
> > > any of these questions have to do with IQ -
> > > they're almost all tests of vocabulary and knowledge.
> >
> > Well, the test certainly isn't culture-independent.
> > It assumes that intelligence is proportional to the ability to learn,
> > and a motivation to do so; by this reasoning, more intelligent people
> > ought to have a greater store of general knowledge and vocabulary
> > pertinent to the culture they live in. Recall ability plays a role, too.
>
> Ah yes, the standard explanation to why success on intelligence tests
> correlates with education: 'because people with high intelligence are more
> likely to stay in school'. Isn't it at least plausible that the causality
> goes the other way?
It's a common misconception, but learning doesn't actually have all that
much to do with school.
IQ scores are pretty predictive of school performance, except at the
high end - which is where these societies supposedly operate, so yes,
it's a bit of a contradiction. But there are IQ tests (read: puzzle
compilations) that are culture neutral, and they can (AFAIK) be used to
gain entry.
Lastly, consider that these societies try to gather members that have
some things in common; they select for whatever the test selects (puzzle
solving skill, probably), so even if you are as intelligent as they are,
if you can't (or don't want to) solve these puzzles, you might not have
all that much in common with them anyway.
Being a member of a high-IQ society is no cause for arrogance: the
membership numbers don't approach the number of equally gifted
individuals in the population by some orders of magnitude, so if you're
a member, any non-member you meet on the street could be just as
intelligent as you are.
Danny Kodicek
> Danny Kodicek schrieb:
> > "Michael Mendelsohn" <inv...@msgid.michael.mendelsohn.de> wrote in
message
> > > gregw...@yahoo.com schrieb:
> > > Danny Kodicek wrote:
> > > > What I find interesting about this is how little
> > > > any of these questions have to do with IQ -
> > > > they're almost all tests of vocabulary and knowledge.
> > >
> > > Well, the test certainly isn't culture-independent.
> > > It assumes that intelligence is proportional to the ability to learn,
> > > and a motivation to do so; by this reasoning, more intelligent people
> > > ought to have a greater store of general knowledge and vocabulary
> > > pertinent to the culture they live in. Recall ability plays a role,
too.
> >
> > Ah yes, the standard explanation to why success on intelligence tests
> > correlates with education: 'because people with high intelligence are
more
> > likely to stay in school'. Isn't it at least plausible that the
causality
> > goes the other way?
>
> It's a common misconception, but learning doesn't actually have all that
> much to do with school.
:) Point taken. But I think my argument still stands.
>
> IQ scores are pretty predictive of school performance, except at the
> high end - which is where these societies supposedly operate, so yes,
> it's a bit of a contradiction. But there are IQ tests (read: puzzle
> compilations) that are culture neutral, and they can (AFAIK) be used to
> gain entry.
>
> Lastly, consider that these societies try to gather members that have
> some things in common; they select for whatever the test selects (puzzle
> solving skill, probably), so even if you are as intelligent as they are,
> if you can't (or don't want to) solve these puzzles, you might not have
> all that much in common with them anyway.
I think this is the real point, and you're right: if you consider these
tests as essentially a screening device to ensure that the 'right' people
get into the club, it's fair enough. The only problem I have with it is
dignifying the result with the word 'intelligence'. It's sheer arrogance.
The whole concept of a 'general intelligence' is flawed. I speak from the
perspective of one who happens to do very well at IQ tests, because my brain
happens to work that way, but I can think of literally dozens of other
mental faculties which could equally well be included in the term 'general
intelligence' but usually aren't, many of which I happen not to have - an
ability with languages, memory for names and faces, social intuition (the
ability to understand people's motivations, something my wife's particularly
good at), financial acuity, observation, hand-eye coordination, management
ability, organisation, visual design - and many others I do to some extent
but still don't count for anything - musical appreciation, acting,
storytelling, etc. Why should logical reasoning, verbal dexterity and one or
two others take precedence? For that matter, why should the ability to use
these skills under time pressures also be so important? The whole exercise
is misguided. Of course you can 'measure' IQ - make up some tests and fire
away. But does it actually have any real meaning? I doubt it.
>
> Being a member of a high-IQ society is no cause for arrogance: the
> membership numbers don't approach the number of equally gifted
> individuals in the population by some orders of magnitude, so if you're
> a member, any non-member you meet on the street could be just as
> intelligent as you are.
I applied for Mensa once as a child, just out of interest (I'd tried one of
their puzzles in the paper). I was invited to join (as I say, I've always
done well with IQ tests, as I imagine most people on this group would do),
but when I read their magazine I realised these *really* weren't the kind of
people I could imagine spending time with.
Danny
Michael Mendelsohn
Yeah, right.
They're driving around with vanity license plates that read "IQ 170".
The test represents an investment in tie and effort: to get a puzzle
collection to yield meaningful IQ results, you ought to ideally run it
past a couple thousand members of the general population. That is of
course not practicable for an amateur organization, so they're making do
with a few dozen people from their number and a comparison to how these
people did on regular tests. Still, doing the statistics on that and
getting the results accepted seems to be an arduous process. By
providing answers to the test, you've changed a big part of it (because
the challenge is now not to solve the questions, but to remember the
answers), and all that effort was in vain as the results cannot now be
compared with the previous results.
Real IQ tests are not published to the general population, and those
that get them need to keep even the questions secret. If an IQ society
did that, you'd decry them for having an elitist secret test that you'd
have to pay for and couldn't disclose, and there'd be less people
inclined to actually look at the test and attempt taking it.
So what you've done is that you've destroyed something of value to a
group of people you don't even know personally (I assume). What a great
intellectual achievement!
Michael Mendelsohn
> "Michael Mendelsohn" <inv...@msgid.michael.mendelsohn.de> wrote in message
> > Danny Kodicek schrieb:
> > > "Michael Mendelsohn" <inv...@msgid.michael.mendelsohn.de> wrote in
> > > > gregw...@yahoo.com schrieb:
> > > > Danny Kodicek wrote:
> > > > > What I find interesting about this is how little
> > > > > any of these questions have to do with IQ -
> > > > > they're almost all tests of vocabulary and knowledge.
> :) Point taken. But I think my argument still stands.
IQ is what an IQ test tests - if you think that the cross-norming
procedure holds up, then the Titan test measures IQ.
> get into the club, it's fair enough. The only problem I have with it is
> dignifying the result with the word 'intelligence'. It's sheer arrogance.
Well, tell that to Binet. The concept of general intelligence tests is
about 100 years old by now.
> The whole concept of a 'general intelligence' is flawed.
And indeed there are theories of intelligence that propose a
multi-factored model of intelligence.
> I speak from the
> perspective of one who happens to do very well at IQ tests, because my brain
> happens to work that way, but I can think of literally dozens of other
> mental faculties which could equally well be included in the term 'general
> intelligence' but usually aren't, many of which I happen not to have - an
> ability with languages, memory for names and faces, social intuition (the
> ability to understand people's motivations, something my wife's particularly
> good at), financial acuity, observation, hand-eye coordination, management
> ability, organisation, visual design - and many others I do to some extent
> but still don't count for anything - musical appreciation, acting,
> storytelling, etc. Why should logical reasoning, verbal dexterity and one or
> two others take precedence? For that matter, why should the ability to use
> these skills under time pressures also be so important?
A main commercial (and military) use of IQ tests is to screen
prospective employees, and these seem to be the sort of skills that come
in handy.
> The whole exercise
> is misguided. Of course you can 'measure' IQ - make up some tests and fire
> away. But does it actually have any real meaning? I doubt it.
Depends on how you define "real meaning". A well-made and validated IQ
test certainly measures a part of personality in a somewhat stable way
that can be seen to correlate with some other observations of people, so
it does have scientific value.
> I applied for Mensa once as a child, just out of interest (I'd tried one of
> their puzzles in the paper). I was invited to join (as I say, I've always
> done well with IQ tests, as I imagine most people on this group would do),
> but when I read their magazine I realised these *really* weren't the kind of
> people I could imagine spending time with.
I know some Mensans I avoid, and some I love to spend time with.
One more argument in favor of the notion that IQ is not a measurement of
personal worth, either way. ;)
gregw...@yahoo.com
Michael Mendelsohn wrote:
> Some very conceited people claim to be among the very fastest in the
> world and have organized a quatrannual event to revel in their
supposed
> athletic superiority.
Well no, the analogy fails because the winners of Olympic events
probably really are the fastest people in the world. These people are
accomplishment-less geeks who claim to be among the smartest 300 people
in America. And athletes do something for society as a whole to merit
their awards, namely entertain us.
> This is despite very little (or rather absolutely nothing) in the way
of
> actual physical accomplishments among them to substantiate their
> incredible claims of athleteship.
I'm not really sure what the form of your argument is supposed to be.
Olympic atheletes do have real accomplishments, Mega Society members
don't. If you are trying to say the form of my argument would allow one
to say great athletes are undeserving of their medals, you are clearly
wrong. They are the best, not pretenders like Mega Society members.
If you look at their journal, in fact, much of it is self-pitying crap
where they explain their lack of success in school and work because
they were so smart "nobody understood them." Poor dears!
> > When I first heard about this four years ago, I decided to put a
stop
> > to this by publishing the answers to their current test, the Titan
Test
> > (notice again their extreme vanity in choosing this name). This had
> > been done before, to their Mega Test, which they had to then
abandon as
> > a criterion of membership.
>
> Wow! What an achievement!
> May I call you "spoilsport extraordinaire" now?
The only thing spoiled is something already rotten, namely unfounded
and pretentious claims of extraordinary intelligence.
gregw...@yahoo.com
that get them need to keep even the questions secret. If an IQ society
did that, you'd decry them for having an elitist secret test that you'd
have to pay for and couldn't disclose, and there'd be less people
inclined to actually look at the test and attempt taking it.>>
I wouldn't decry them, it would be nothing more than what ETS does. In
fact I've taken a number of standardized tests and I think they are in
fact much better measures of IQ than the silly and poorly written Titan
test, even they don't purport to be IQ tests.
<<So what you've done is that you've destroyed something of value to a
group of people you don't even know personally (I assume). >>
Perhaps of value to them, but of negative value to society as a whole,
which should reserve its respect for intellectuals who actually do
something useful for us all.
They can have all the silly tests they want, but they ought not claim
to all be smarter than the average faculty member at Harvard and MIT,
which they do. If they want people to thing of them as smart, which I
don't blame them for, then they ought to go do something that indicates
they really are smart.
Canon
I'm a member of the Mega Society, so I guess I should comment on this.
>From the title of your post I infer that your main objection to the
Titan Test is that you do not think it is a legitimate intelligence
test. But you do not seem to be arguing, as others have, that
intelligence does not exist or that it cannot be measured. Some of the
responders have touched on these subjects, implying that this is at
least partially your point, but I'm going to assume otherwise.
Your main point is that if these tests were a good measure of
intelligence, then the members of the Mega Society would be
distinguished by actual intellectual accomplishments. I think you mean
by "intellectual accomplishments" things like earning a Nobel Prize.
You then make a number of specific criticisms of individual test
questions, but let's start with the main point. I assert that it is a
logical fallacy. You assume that being intelligent tells you what is
important to work on. But in fact, values are subjective, and no
amount of experimentation will tell you what is good or beautiful or
worth spending time on. Beauty is in the eye of the beholder.
You might argue that high intelligence at least should evidence itself
by deep understanding of some particular area of study. But even this
is a non sequitur. Perhaps high intelligence evidences itself in
broadness of inquiry. At least one member of the Mega Society reads
voraciously, consuming several books a day. But this is not a
necessary consequence of high intelligence. The simple truth is that
you cannot deduce from the facts of the world what you ought to do.
So, your main argument that if the Titan Test were a valid measurement
of high intelligence then the Mega Society would contain people of high
intellectual accomplishments is simply invalid.
Before discussing objections you lodge to specific questions, let me
note in passing the related sophistry that some people fall into when
arguing that intelligence tests are invalid. The argument goes
something like this: intelligence tests are invalid because look at all
the people who score poorly who do well in life (and vice versa). But
this assumes that there is a generally agreed upon notion of "doing
well in life." I think that's a lot less generally agreed upon than
what intelligence is.
So on to your objections to specific questions. I think there is a
general consensus among Mega Society members that Mega and Titan Test
questions are too culturally biased and too research oriented. We have
started to make some tentative steps to fix this (see the very
preliminary online test at http://www.mental-testing.com), but these
objections should not be overblown. Even if the test is culturally
biased, this means only that certain test takers are not going to score
highly. Given that these are the first tests that can distinguish IQs
in the 150 - 180 range, it isn't a critical flaw that they exclude some
people. We have to start somewhere.
Also, it matters what culture we're biased towards. For example, you
say:
>22. LANGUAGE GAMES : LUDWIG :: PIANO CONCERTI FOR THE LEFT HAND : PAUL
>(So obscure it is silly. Ludwig and Paul Wittgenstein were brothers.
>One was a pianist, the other was the famous philosopher who used the
>phrase "language games" in his book Philosophical Investigations.
>Knowing this meaningless bit of trivia can hardly be an indication of
>great intelligence.)
One man's trivia is another man's profound insight. I don't think it's
trivial to know about the work of Ludwig Wittgenstein. I don't think
it's trivial to know what a language game is. You are implying that
the test is replete with pop culture trivia akin to "Jeopardy!" But
this is not the case. Some questions on the test are culturally
biased, but the culture in question is the culture of late twentieth
century philosophy, literature, mathematics, and so forth. Such
knowledge is correlated with intelligence.
-Chris Cole
rec.puzzles archivist
www.rec-puzzles.org
mensa...@aol.com
Canon wrote:
> Hi Greg,
>
> I'm a member of the Mega Society, so I guess I should comment on th
is.
And I do not speak for the OP, but would still like to comment.
>
> From the title of your post I infer that your main objection to the
> Titan Test is that you do not think it is a legitimate intelligence
> test. But you do not seem to be arguing, as others have, that
> intelligence does not exist or that it cannot be measured. Some of
the
> responders have touched on these subjects, implying that this is at
> least partially your point, but I'm going to assume otherwise.
>
> Your main point is that if these tests were a good measure of
> intelligence, then the members of the Mega Society would be
> distinguished by actual intellectual accomplishments. I think you
mean
> by "intellectual accomplishments" things like earning a Nobel Prize.
> You then make a number of specific criticisms of individual test
> questions, but let's start with the main point. I assert that it is
a
> logical fallacy. You assume that being intelligent tells you what is
> important to work on. But in fact, values are subjective, and no
> amount of experimentation will tell you what is good or beautiful or
> worth spending time on. Beauty is in the eye of the beholder.
Ok, high intelligence doesn't necessarily mean great
accomplishments. So what is the Mega Society looking for?
Potential great accomplishments or simply the $30.
>
> You might argue that high intelligence at least should evidence
itself
> by deep understanding of some particular area of study. But even
this
> is a non sequitur. Perhaps high intelligence evidences itself in
> broadness of inquiry. At least one member of the Mega Society reads
> voraciously, consuming several books a day. But this is not a
> necessary consequence of high intelligence. The simple truth is that
> you cannot deduce from the facts of the world what you ought to do.
Doesn't that prove the point that the Titan test cannot
measure intelligence? If neither deep knowledge nor broad
knowledge is a necessary consequence of intelligence,
can you enlighten me on just what exactly _is_ a consequence
of intelligence and how the Titan test measures it?
>
> So, your main argument that if the Titan Test were a valid
measurement
> of high intelligence then the Mega Society would contain people of
high
> intellectual accomplishments is simply invalid.
Does the Mega Society have some sort of statement as to
what its purpose is? Other than manufactured elitism.
>
> Before discussing objections you lodge to specific questions, let me
> note in passing the related sophistry that some people fall into when
> arguing that intelligence tests are invalid. The argument goes
> something like this: intelligence tests are invalid because look at
all
> the people who score poorly who do well in life (and vice versa).
But
> this assumes that there is a generally agreed upon notion of "doing
> well in life." I think that's a lot less generally agreed upon than
> what intelligence is.
So IQ scores have no correlation to anything? Aren't you
making the same argument that I made when I said the only
thing an IQ test measures is how high a score you can get
on an IQ test? Why do you flaunt your membership in a society
whose requirements have no meaning? Or better still, why don't
you tell us what exactly makes all you Mega Society people
so special.
>
> So on to your objections to specific questions. I think there is a
> general consensus among Mega Society members that Mega and Titan Test
> questions are too culturally biased and too research oriented. We
have
> started to make some tentative steps to fix this (see the very
> preliminary online test at http://www.mental-testing.com), but these
> objections should not be overblown. Even if the test is culturally
> biased, this means only that certain test takers are not going to
score
> highly. Given that these are the first tests that can distinguish
IQs
> in the 150 - 180 range,
Has that been accepted by anyone other than the Mega Society?
> it isn't a critical flaw that they exclude some people.
Really?
> We have to start somewhere.
Why do you have to start somewhere? Why don't you document
those who you are intending to exclude. Is your club only
for those who have a Liberal Arts education?
>
> Also, it matters what culture we're biased towards. For example, you
> say:
>
> >22. LANGUAGE GAMES : LUDWIG :: PIANO CONCERTI FOR THE LEFT HAND :
PAUL
>
> >(So obscure it is silly. Ludwig and Paul Wittgenstein were brothers.
> >One was a pianist, the other was the famous philosopher who used the
> >phrase "language games" in his book Philosophical Investigations.
> >Knowing this meaningless bit of trivia can hardly be an indication
of
> >great intelligence.)
>
> One man's trivia is another man's profound insight. I don't think
it's
> trivial to know about the work of Ludwig Wittgenstein. I don't think
> it's trivial to know what a language game is.
I guess that answers my question on who you're trying to exclude.
> You are implying that the test is replete with pop culture
> trivia akin to "Jeopardy!" But this is not the case.
> Some questions on the test are culturally biased,
Trivia is trivia whether its source is pop culture or
Liberal Arts. If you're so smart, tell us how that correlates
to intelligence.
> but the culture in question is the culture of late twentieth
> century philosophy, literature, mathematics, and so forth.
So, basically, you're admitting that the Titan test does
_not_ measure intelligence but knowledge of late twentieth
century trivia.
> Such knowledge is correlated with intelligence.
Exactly what is the correlation? Are you trying to say a
philosopher who can't program a computer is more intelligent
that a computer programmer who doesn't know Wittgenstein
from a hole in the ground?
Andrew Bull
DARKNESS OF THE
NIGHT SKY IN AN INFINITE UNIVERSE) is OLBERS (the former is Russell's
paradox, the latter is Olbers'), not KEPLER.
A couple of others:
5. TIRE : RETREAD :: PARCHMENT : ? is possibly PALIMPSEST
7. SWORD : DAMOCLES :: BED : ? is presumably PROCRUSTES
Lash Rambo
@g14g2000cwa.googlegroups.com:
> logical fallacy. You assume that being intelligent tells you what is
> important to work on. But in fact, values are subjective, and no
> amount of experimentation will tell you what is good or beautiful or
> worth spending time on. Beauty is in the eye of the beholder.
1. High intelligence does not tell you what is worthwhile.
> You might argue that high intelligence at least should evidence itself
> by deep understanding of some particular area of study. But even this
> is a non sequitur. Perhaps high intelligence evidences itself in
> broadness of inquiry. At least one member of the Mega Society reads
2. High intelligence is evidenced not by specialization, but by broad
inquiry.
> this is not the case. Some questions on the test are culturally
> biased, but the culture in question is the culture of late twentieth
> century philosophy, literature, mathematics, and so forth. Such
> knowledge is correlated with intelligence.
3. The Titan Test, (supposedly) measuring extremely high intelligence,
focuses on specialized knowledge that lots of people with high
intelligence would have found worthwhile (enough for said knowledge to be
correlated with high intelligence).
How do you reconcile assertions 1 and 2 with 3?
Further, it's unclear why knowledge acquisition and retention are
evidence of "intelligence" at all! A CD can be burned with the contents
of an encyclopedia--does this mean the CD is highly intelligent?
(Certainly, the CD doesn't know if the contents are worthwhile, yet the
contents represent a broad inquiry into human knowledge!)
gcrh...@yahoo.com
> "> 38. I am pretty close to sure the answer is 1/27
>
> I get 2/27 (each of the three distinct cycles is
> traversable either forwards or backwards)."
>
> I figured out 1/27 four years ago, so I can't tell you how I got it.
> How sure are you? I wrote in my notes I was 95% sure but I didn't
write
> how I got to it or the cause of my doubt.
Each ant independently has a choice of 3 edges to traverse.
There are 4 ants, hence there are 3^4 total possibilities.
Consider the graph of an arbitrary polyhedron. A cycle-cover
is a set of disjoint simple cycles such that every vertex is
in some cycle. Now no ant will run into another ant if and
only if the set of traversed edges form a cycle-cover.
(If two ants traverse the same edge, then that edge cannot
be part of a cycle because the number of edges in a cycle
equals the number of vertices and there is one ant per vertex)
So we need to find all cycle-covers of the tetrahedron graph.
A cycle contains at least three vertices. The tetrahedron
graph has only 4 vertices. If we use a cycle of only three
vertices, then we cannot cover the remaining vertex with a
cycle. Hence, the only cycle-covers consist of a single
cycle (i.e. a hamiltonian cycle). How many are there?
There are 4 vertices and each vertex is adjacent to all other
vertices. Hence if we label the vertices a, b, c, d, then
any permutation of the vertices corresponds to a cyclic path
on all the vertices. There are 4! = 24 permutations but
each path is counted multiple times. Any two permutations
that are cyclic shifts of each other count the same path.
Every permutation of length 4 has 4 cyclic shifts, hence
we are counting each path 4 times. So the total number
of paths is 24/4 = 6. (You could also get this by
fixing a start vertex and then noting that there are 3!=6
ways to permute the remaining vertices).
So the probability is 6 / 3^4 = 2 / 27.
Since you came up with 1 / 27. I presumed that you must
have noted that there were three distinct cycles and forgot
that each cycle corresponds to two distinct cyclic paths --
i.e. you can traverse a cycle in either direction.
For completeness, the six distinct paths listed in pairs
corresponding to the same cycle are as follows.
a b c d <---> a d c b
a b d c <---> a c d b
a c b d <---> a d b c
All other permutations are a cyclic shift of one of the
above. So the short answer your original question is,
I'm pretty sure.
For the cube, I get a probability of 8 / 3^7
There are two distinct types of cycle covers. You can have
a cycle of the vertices on one face combined with a cycle of
the opposite face. There are three pairs of opposite faces
and hence three cycle covers of this form. Both cycles in
a cover can be independently traversed in one of two
directions. Hence, there are 4 paths for each cycle cover
and 12 paths total for covers of this form. The other
type of cover is a hamiltonian cycle. Assuming I counted
everything correctly I get 12 more paths for a total of
24 non-collision paths. Each of the 8 ants has an
independent choice of three edges, hence there are 3^8
total possibilities. 24 / 3^8 = 8 / 2187.
So that you can verify (hopefully!) that I counted the
hamiltonians correctly, I'll label the vertices of the
cube graph as follows.
(view in a fixed font)
1--------------2
|\ /|
| \ / |
| \ / |
| \ / |
| 4----3 |
| | | |
| | | |
| 5----6 |
| / \ |
| / \ |
| / \ |
|/ \|
8--------------7
The 12 paths on the 6 distinct hamiltonian cycles are as follows
(again listed in pairs).
1 2 3 4 5 6 7 8 <---> 1 8 7 6 5 4 3 2
1 2 3 6 7 8 5 4 <---> 1 4 5 8 7 6 3 2
1 2 7 6 3 4 5 8 <---> 1 8 5 4 3 6 7 2
1 2 7 8 5 6 3 4 <---> 1 4 3 6 5 8 7 2
1 4 3 2 7 6 5 8 <---> 1 8 5 6 7 2 3 4
1 4 5 6 3 2 7 8 <---> 1 8 7 2 3 6 5 4
Did I get them all?
For the icohedron, I'ld want a computer program to search
for all the cycle covers.
If chairMAN is a sexist term, then so is chairperSON
and for the exact same reason.
Canon
posters.
Intelligence is usually defined as problem solving ability, or
something like that. I'm not proposing a new definition.
The Mega Society is looking for high intelligence. We do not
administer the Titan Test and we do not get any money from it.
The Constitution of the Society
(http://www.megasociety.org/constitution.html) states that the Society
exists to facilitate interaction among its members and to assist them
in gaining access to resources to accomplish their individual purposes.
I can only speak to why I joined. I've been an AI researcher all my
life and I think one reason for the slow progress of AI is that we
don't know much about the nature of intelligence. Sometimes the best
place to look for underlying rules is extreme cases. This is why we
build particle accelerators and this is why Oliver Sacks seeks out the
patients he does. I think I've learned a lot by being a member.
The Mega Test is composed of questions that either require a high level
of intelligence to solve or require knowledge that tends to be
correlated with high intelligence, or both. We would prefer if there
were more of the former than the latter, but creating high range
problems is very hard. It's an evolving art. If you think it's easy,
try it some time.
I don't like being called elitist. I've explicitly disclaimed that the
Mega Society is about potential greatness. The Mega Society is no more
elitist than a club soccer team. You need to make the cut, but that
doesn't make you a good person.
gcrh...@yahoo.com
> Here are my responses to most of the issues raised by the last few
> posters.
>
> Intelligence is usually defined as problem solving ability, or
> something like that. I'm not proposing a new definition.
I've never heard it defined as problem solving ability.
That's only one limited type of intelligence. There really
is no such thing as general intelligence, mentally people
are an amalgamation of many different types of intelligence.
I work in the algorithms branch of computer science which
is essentially problem solving. So naturally I do well on
problem solving tasks and also on intelligence tests but
there are other types of mental tasks that I don't do so
well on such as learning a foreign language or painting a
picture. This almagamation is easier to see on people
for whom some type(s) of intelligence is at a completely
different level than other types. Consider Mozart for
example. He wasn't noted for being adept at problem
solving. His abilities were pretty ordinary in every way
except one. Ahh but that one! He was the greatest musical
genius ever. To be able to write his masterful Linz
symphony in only four days borders on the miraculous.
But I suppose his incredible musical intelligence doesn't
count on your scale because he couldn't have passed the
meaningless Titan Test.
mensa...@aol.com
Canon wrote:
> Here are my responses to most of the issues raised by the last few
> posters.
>
> Intelligence is usually defined as problem solving ability, or
> something like that. I'm not proposing a new definition.
And yet, you endorse culturally biased questions as being
indicative of intelligence. Can you please explain what
problem solving ability that measures?
>
> The Mega Society is looking for high intelligence.
And yet, you have no way of properly measuring it.
> We do not
> administer the Titan Test and we do not get any money from it.
Is the person who administers the test a member of the
Mega Society?
>
> The Constitution of the Society
> (http://www.megasociety.org/constitution.html) states that the
Society
> exists to facilitate interaction among its members and to assist them
> in gaining access to resources to accomplish their individual
purposes.
> I can only speak to why I joined. I've been an AI researcher all my
> life and I think one reason for the slow progress of AI is that we
> don't know much about the nature of intelligence.
And yet, you presume to be the arbiters of who is and who is
not intelligent. You're the genius here, doesn't that strike you
as hypocritical?
> Sometimes the best
> place to look for underlying rules is extreme cases. This is why we
> build particle accelerators and this is why Oliver Sacks seeks out
the
> patients he does. I think I've learned a lot by being a member.
>
> The Mega Test is composed of questions that either require a high
level
> of intelligence to solve or require knowledge that tends to be
> correlated with high intelligence, or both. We would prefer if there
> were more of the former than the latter, but creating high range
> problems is very hard. It's an evolving art. If you think it's
easy,
> try it some time.
Yes, it is difficult for a blind man in a dark room to find a black
cat that isn't there. But unlike you, I don't claim to have found it.
>
> I don't like being called elitist.
elite: n. 1. the choicest part, as of a social group.
If you don't like fleas, maybe you shouldn't sleep with dogs.
Whether you like it or not, the Mega Society is an elitist group.
> I've explicitly disclaimed that the
> Mega Society is about potential greatness. The Mega Society is no
more
> elitist than a club soccer team.
Do you let fatties join your soccer club? What about gimps,
are they welcome? How did you pass the test if you don't
understand analogies?
> You need to make the cut,
"There is no elitism in the Mega Society. And when I say
none, I mean there is a certain amount."
>but that doesn't make you a good person.
And that wasn't an issue. The issue is whether culturally biased
trivia questions measure intelligence. Your reply has not addressed
that issue at all.
Mike
@l41g2000cwc.googlegroups.com>, gregw...@yahoo.com says...
...
> 31. The answer is 22.
Canon
There are many kinds of intelligence. Musical intelligence such as
exhibited by Mozart is not the kind of intelligence measured by the
Titan Test. The Titan Test measures the kind of intelligence measured
by what psychometricians call tests of "general" intelligence.
There is evidence that the Mega and Titan tests are good measures of
general intelligence. At the low end of their range they are well
correlated with other generally accepted, standardized tests of
intelligence. The data is available on this Web site:
http://www.eskimo.com/~miyaguch/
If a test is culturally biased, it means that people who are not
members of a certain culture will find the questions on the test very
difficult. This does not imply that people who are a member of the
culture will find the test easy. The Mega and Titan tests are
culturally biased toward English-speaking people who are educated to a
certain degree. Thus the Mega Society does not contain any members
from the non-English speaking world. It also does not contain any
people who have not graduated from high school. We recognize that this
is because the admissions tests are culturally biased. As I've said
before, we're working on trying to eliminate this bias, but it will
take some time.
Rather than simply criticizing the Titan Test, how about helping to fix
it? We are looking for good problems for our experimental site
If you have ideas for good high range problems, please let us know.
gcrh...@yahoo.com
>
> There are many kinds of intelligence. Musical intelligence such as
> exhibited by Mozart is not the kind of intelligence measured by the
> Titan Test. The Titan Test measures the kind of intelligence measured
> by what psychometricians call tests of "general" intelligence.
>
> There is evidence that the Mega and Titan tests are good measures of
> general intelligence. At the low end of their range they are well
> correlated with other generally accepted, standardized tests of
> intelligence. The data is available on this Web site:
The recognition that the Titan test doesn't measure musical
intelligence is progress but you missed the larger point.
THERE IS NO SUCH THING AS *GENERAL INTELLIGENCE*.
That's an old view that's been pretty much discredited. The
Titan test cannot possibly be an accurate measure of something
that doesn't exist. I don't care how well it correlates with
other IQ tests; they can't measure "general intelligence" either.
I believe the only thing the standard IQ test has been highly
correlated with is performance in school. The IQ test can be
another tool to help identify students who may need more help
in school. I don't see any other practical benefit of the IQ test.
Ed Murphy
> gregw...@yahoo.com schrieb:
>> Some very conceited people claim to be among the very smartest in the
>> world and have formed a large number of pretentiously-named clubs to
>> revel in their supposed intellectual superiority.
>
> Some very conceited people claim to be among the very fastest in the world
> and have organized a quatrannual event to revel in their supposed athletic
> superiority.
>
>> This is despite very
>> little (or rather absolutely nothing) in the way of actual intellectual
>> accomplishments among them to substantiate their incredible claims of
>> intelligence.
>
> This is despite very little (or rather absolutely nothing) in the way of
> actual physical accomplishments among them to substantiate their
> incredible claims of athleteship.
>
>> Their main club is the "Mega Society" which they say is reserved for
>> those with one-in-a-million IQ's, ie above 176.
>
> Their main event are the Olympic games, which they say are reserved for
> one-in-a-million runners who do the 100m dash in less than 10 seconds.
Oh, you were doing so well, until this part. Running 100m in less than
10 seconds *is* an actual physical accomplishment; a damned impressive
one, in fact. Congratulations, you've just contradicted yourself.
Where you *do* have a point - or would, if you weren't so busy being
smug, and if furthermore you didn't mind substantially agreeing with
the original poster - is this:
There are many forms of physical ability - strength, endurance, speed of
movement, speed of changing directions, accuracy, timing, experience
with the rules of a specific form of athletic competition, and probably
some other things. The 100m dash is a fine test of speed of movement,
but a lousy test of most of the others.
Similarly, the Titan Test is a fine test of certain forms of intelligence,
while overlooking other forms. Canon, the test's other defender in this
thread, said this himself. The disagreement is over the following points:
* Which forms of intelligence are part of "general intelligence"?
* Which forms of intelligence does an IQ test directly measure?
* Say there are forms of intelligence {A,B} and an IQ test only
directly measures A. Can its failure to directly measure B
be defended? You can argue that possession of B is strongly
correlated with possession of A, and/or that "general intelligence"
does not include B. Both are liable to be widely disputed.
In particular, you mention in a later message that the Titan Test is
good at identifying suitable candidates for commercial and military
operations. However, you also mention that the forms of intelligence
directly measured by the Titan Test are the forms that those operations
usually need; they don't necessarily need "general intelligence",
whatever that is.
Michael Mendelsohn
> On Sat, 02 Apr 2005 09:46:40 +0200, Michael Mendelsohn wrote:
> > gregw...@yahoo.com schrieb:
> >> Some very conceited people claim to be among the very smartest in the
> >> world and have formed a large number of pretentiously-named clubs to
> >> revel in their supposed intellectual superiority.
> >
> > Some very conceited people claim to be among the very fastest in the world
> > and have organized a quatrannual event to revel in their supposed athletic
> > superiority.
> >
> >> This is despite very
> >> little (or rather absolutely nothing) in the way of actual intellectual
> >> accomplishments among them to substantiate their incredible claims of
> >> intelligence.
> >
> > This is despite very little (or rather absolutely nothing) in the way of
> > actual physical accomplishments among them to substantiate their
> > incredible claims of athleteship.
>
> 10 seconds *is* an actual physical accomplishment; a damned impressive
> one, in fact. Congratulations, you've just contradicted yourself.
Actually, solving a substantive number of questions on the Titan test is
an actual mental accomplishment; a damned impressive one, in fact
(provided we believe you didn't cheat, which is a problem with
unsupervised tests).
The problem is: how much is that accomplishment worth? We assume it's
worth something to the people doing it.
Running 100m up Omaha Beach on D-Day is a lot more of an accomplishment
than running 100m in any Olympics, if you ask me. Ask someone else and
you might get a different answer - it's a free world. ;)
> Where you *do* have a point - or would, if you weren't so busy being
> smug, and if furthermore you didn't mind substantially agreeing with
> the original poster - is this:
The original poster didn't show any reasoning I would agree with.
I haven't seen any supporting evidence for the grief he seems to have
with the Mega Society, and I can't see that a minor club uses a puzzle
test to select its members is a threat to society at large. ;)
> * Which forms of intelligence are part of "general intelligence"?
Look up Stanford-Binet; for another take, maybe Raven's matrices and
Cattell. These have been around for ages, and they've defined what "IQ"
stands for. Disagreeing with them is fine, but you should then not use
the term "IQ".
> * Which forms of intelligence does an IQ test directly measure?
That depends on the test.
> * Say there are forms of intelligence {A,B} and an IQ test only
> directly measures A. Can its failure to directly measure B
> be defended?
Well, you have a ruler which directly measures length. Can its failure
to directly measure volumes be defended?
What a silly question!
> In particular, you mention in a later message that the Titan Test is
> good at identifying suitable candidates for commercial and military
> operations.
I did not mention that.
I mentioned that "real" IQ tests, administered by psychologists under
supervision, and normed by running it past a large population sample
(this procedure costs 6 figures or more), are employed for that purpose.
> However, you also mention that the forms of intelligence
> directly measured by the Titan Test are the forms that those operations
> usually need; they don't necessarily need "general intelligence",
> whatever that is.
--
Michael Mendelsohn
> Why do you flaunt your membership in a society
> whose requirements have no meaning? Or better still, why don't
> you tell us what exactly makes all you Mega Society people
> so special.
Haven't seen him flaunt it, haven't seen him opine that Mega Society
people are special in any way.
You're building a straw man.
Cheers
Michael
Michael Mendelsohn
> Michael Mendelsohn wrote:
> > Some very conceited people claim to be among the very fastest in the
> > world and have organized a quatrannual event to revel in their
> supposed
> > athletic superiority.
>
> Well no, the analogy fails because the winners of Olympic events
> probably really are the fastest people in the world. These people are
> accomplishment-less geeks who claim to be among the smartest 300 people
> in America. And athletes do something for society as a whole to merit
> their awards, namely entertain us.
Well, what I've seen of Mensa meetings, the potential for entertainment
is certainly there - all that's missing are the cameras and a snappy
commentator! ;)
Chess may be another good analogy - are chess grand masters
"accomplishment-less geeks"? Does the answer to this depend on your
personal value system or on the fact how much of them you can see on TV?
> I'm not really sure what the form of your argument is supposed to be.
> Olympic atheletes do have real accomplishments, Mega Society members
> don't.
I have been trying to get you to state by what criteria you divide
"real" accomplishment from fake ones. You avoid getting to that point.
> If you are trying to say the form of my argument would allow one
> to say great athletes are undeserving of their medals, you are clearly
> wrong. They are the best, not pretenders like Mega Society members.
Does one have to be "the best" to have accomplished anything?
> If you look at their journal, in fact, much of it is self-pitying crap
> where they explain their lack of success in school and work because
> they were so smart "nobody understood them." Poor dears!
If you want to criticise a specific article, please point me to one and
we can do so by email. I'v enot been interested enough to peruse their
journal, so far.
> > Wow! What an achievement!
> > May I call you "spoilsport extraordinaire" now?
>
> The only thing spoiled is something already rotten, namely unfounded
> and pretentious claims of extraordinary intelligence.
Sadly, you are not spoiling these claims.
Michael Mendelsohn
> <<So what you've done is that you've destroyed something of value to a
> group of people you don't even know personally (I assume). >>
>
> Perhaps of value to them, but of negative value to society as a whole,
> which should reserve its respect for intellectuals who actually do
> something useful for us all.
You claim that society respects the members of the Mega Society.
This sounds a bit far-fetched to me, can you back this claim up?
> They can have all the silly tests they want, but they ought not claim
> to all be smarter than the average faculty member at Harvard and MIT,
> which they do. If they want people to thing of them as smart, which I
> don't blame them for, then they ought to go do something that indicates
> they really are smart.
So you think being a faculty member at one of these institutions is
indicative of smartness? Or would you admit that an ability to do
networking and to fit in help a long way towards getting one of these
posts?
And would you not admit that people in those positions are endowed by
society with more means to be useful to society than those who are not?
Successful smart people don't need a society - they have regular faculty
meetings (or similar, if they're working in the industry etc.). It is
those smart people who don't fit in who need a society like this,
because they'll feel alienated anywhere else.
If you can't hold down a steady job, it's hard to exchange thoughts with
smart people on a regular basis - well, unless you read rec.puzzles, of
course. :)
(Actually you can find smart people everywhere on the net!)
Cheers
Lash Rambo
> Ed Murphy schrieb:
>> * Say there are forms of intelligence {A,B} and an IQ test only
>> directly measures A. Can its failure to directly measure B
>> be defended?
>
> Well, you have a ruler which directly measures length. Can its failure
> to directly measure volumes be defended?
Not if the ruler was advertised as a device for measuring "general
quantities."
Canon
poster other than you was hurting some work that I was trying to get
done. I think people have understood this and politely stopped posting
answers to Titan Test questions. I appreciate that and am thankful.
If you want to discuss the nature of intelligence, I'm interested in
the subject, but I suggest we take it offline. My email address is
ch...@questrel.com.
mensa...@aol.com
Michael Mendelsohn wrote:
> "mensa...@aol.com" schrieb:
> > Why do you flaunt your membership in a society
> > whose requirements have no meaning? Or better still, why don't
> > you tell us what exactly makes all you Mega Society people
> > so special.
>
> Haven't seen him flaunt it,
He did say he's a member, didn't he? And you have to be genius
to be a member, don't you? He certainly could have given his
opinions without mentioning his membership. But nooooooo, he
wanted everyone to know that _he_ has an IQ of 175 and _his_
opinions carry more weight than anyone else's.
> haven't seen him opine that Mega Society people are special
> in any way.
Are you a member of a High-IQ society? People who join such
societies think they are special. I've seen it first hand at
Mensa. And here's a quote from someone at the Mega Society:
"Humanity in the collective produces only the odd great man
or two in any century. We must see ourselves in light of this.
Your work in test construction has been championed by two of
psychology's luminaries and doing a straw poll of our Mega
membership would, I'm sure, produce a few more comparables
and many big achievers in general. So we are already out of
proportion to our numbers."
It may not be written in their constitution, but it's there.
>
> You're building a straw man.
I like how you picked the least important comment I made to
talk about a strawman.
Ted Schuerzinger
news:1112662245.3...@l41g2000cwc.googlegroups.com:
> Are you a member of a High-IQ society? People who join such
> societies think they are special. I've seen it first hand at
> Mensa.
This implies that you joined Mensa at one point, and that you think you're
special, too. :-p
--
Ted <fedya at bestweb dot net>
TV Announcer: It's 11:00. Do you know where your children are?
Homer: I told you last night, *no*!
<http://www.snpp.com/episodes/4F06.html>
mensa...@aol.com
Ted Schuerzinger wrote:
> Somebody claiming to be "mensa...@aol.com" <mensa...@aol.com>
wrote in
> news:1112662245.3...@l41g2000cwc.googlegroups.com:
>
> > Are you a member of a High-IQ society? People who join such
> > societies think they are special. I've seen it first hand at
> > Mensa.
>
> This implies that you joined Mensa at one point,
Sorry, Sherlock, but my experience with Mensa consisted of
attending a few of their parties as the guest of a member.
I have never been a member, nor would I want to be. Like most
members, I was there for the socializing and the fun activities
(old sci-fi movies, games, "How to Strip for Your Man", etc.).
I just happen to be the Chicago Mensa Quiz Bowl Champion of 1996.
Yep, beat them at their own game.
But then there was that infandous little poster I saw. It was
by a member who was trying to organize a study to determine
"why we're so special".
> and that you think you're special, too. :-p
I'm not special, just better than the Mensa.
Ed Murphy
> Ed Murphy schrieb:
>> On Sat, 02 Apr 2005 09:46:40 +0200, Michael Mendelsohn wrote:
>> > This is despite very little (or rather absolutely nothing) in the way
>> > of actual physical accomplishments among them to substantiate their
>> > incredible claims of athleteship.
>>
>> Oh, you were doing so well, until this part. Running 100m in less than
>> 10 seconds *is* an actual physical accomplishment; a damned impressive
>> one, in fact. Congratulations, you've just contradicted yourself.
>
> Actually, solving a substantive number of questions on the Titan test is
> an actual mental accomplishment; a damned impressive one, in fact
> (provided we believe you didn't cheat, which is a problem with
> unsupervised tests).
I didn't say it wasn't; I was pointing out that you stepped on your
own "nothing in the way of actual physical accomplishments" comment.
>> * Which forms of intelligence are part of "general intelligence"?
>
> Look up Stanford-Binet; for another take, maybe Raven's matrices and
> Cattell. These have been around for ages, and they've defined what "IQ"
> stands for. Disagreeing with them is fine, but you should then not use the
> term "IQ".
Fair enough. I view the claim "the Titan Test's results correlate
with the results of many other widely-accepted IQ tests" with healthy
suspicion, though - doesn't the Titan Test claim to measure a range
in which most of those other tests are fuzzy? Anyway, if the Titan
Test's determinations of *both* over-150 and under-150 substantially
correlate with the same determinations by those other tests, then
they've at least got empirical evidence on their side.
>> In particular, you mention in a later message that the Titan Test is
>> good at identifying suitable candidates for commercial and military
>> operations.
>
> I did not mention that.
> I mentioned that "real" IQ tests, administered by psychologists under
> supervision, and normed by running it past a large population sample (this
> procedure costs 6 figures or more), are employed for that purpose.
My mistake, then. My other points (snipped here) stand, though, to the
extent that they consider IQ tests in general and not the Titan Test in
particular.
Michael Mendelsohn
> On Mon, 04 Apr 2005 20:40:58 +0200, Michael Mendelsohn wrote:
> > Ed Murphy schrieb:
> >> On Sat, 02 Apr 2005 09:46:40 +0200, Michael Mendelsohn wrote:
> >> > This is despite very little (or rather absolutely nothing) in the way
> >> > of actual physical accomplishments among them to substantiate their
> >> > incredible claims of athleteship.
> >>
> >> Oh, you were doing so well, until this part. Running 100m in less than
> >> 10 seconds *is* an actual physical accomplishment; a damned impressive
> >> one, in fact. Congratulations, you've just contradicted yourself.
> >
> > Actually, solving a substantive number of questions on the Titan test is
> > an actual mental accomplishment; a damned impressive one, in fact
> > (provided we believe you didn't cheat, which is a problem with
> > unsupervised tests).
>
> I didn't say it wasn't; I was pointing out that you stepped on your
> own "nothing in the way of actual physical accomplishments" comment.
Thanks, that means the analogy fulfilled its purpose: Greg claimed
athletes accomplish something, while puzzle solvers don't, and I think
that's absurd.
At that point he got into the "useful to society" argument.
Cheers
Michael
Michael Mendelsohn
> >> * Say there are forms of intelligence {A,B} and an IQ test only
> >> directly measures A. Can its failure to directly measure B
> >> be defended?
> >
> > Well, you have a ruler which directly measures length. Can its failure
> > to directly measure volumes be defended?
>
> Not if the ruler was advertised as a device for measuring "general
> quantities."
I'd advertise a ruler as "suitable to measure spatial quantities", i.e.
lengths in whatever orientation in space.
ObPuzzle: what is the most general way to describe a ruler?
Cheers
Michael
gregw...@yahoo.com
> Hi Greg,
>
> I'm a member of the Mega Society, so I guess I should comment on
this.
Thank you for your intesting answer. I won't assume that as a member
you necessarily agree with all the claims of the organization, so I
will ask you separately. Do you think you are one of the most
intelligent 300 people in the USA, and 6000 most intelligent people in
the world? Don't hedge here, define intelligence any way you think it
is commonly understood. What do you estimate your IQ to be? Other than
the Mega Test score, what evidence can you proffer in support of this?
In other words, in other instances where you competed against smart
people, such as in standardized tests, college exams, and equity
investing, what were the results?
> From the title of your post I infer that your main objection to the
> Titan Test is that you do not think it is a legitimate intelligence
> test. But you do not seem to be arguing, as others have, that
> intelligence does not exist or that it cannot be measured.
Yes, that's correct.
> Your main point is that if these tests were a good measure of
> intelligence, then the members of the Mega Society would be
> distinguished by actual intellectual accomplishments.
No, I think the test is not a good measure of IQ (or rather not as good
a measure as its extraordinary claims) independent of whether Mega
Society members have made great intellectual accomplishments. Rather
the seeming lack of intellectual accomplishments is merely evidence of
the test not being what it claims. An examination of the content of the
test itself would also reveal this, as would the poor quality of
writing of the Mega Society's journal.
> You might argue that high intelligence at least should evidence
itself
> by deep understanding of some particular area of study. But even
this
> is a non sequitur. Perhaps high intelligence evidences itself in
> broadness of inquiry. At least one member of the Mega Society reads
> voraciously, consuming several books a day. But this is not a
> necessary consequence of high intelligence. The simple truth is that
> you cannot deduce from the facts of the world what you ought to do.
>
> So, your main argument that if the Titan Test were a valid
measurement
> of high intelligence then the Mega Society would contain people of
high
> intellectual accomplishments is simply invalid.
It is invalid only if I make an extreme claim that such a club *must*
contain people of "high intellectual accomplishments." My argument,
however, is that the Titan Test has flaws obvious to those who have
studied intelligence tests such that it cannot be what it claims, that
is a means of indentifying the very smartest people in the world. And
secondly as evidence of this, those who have scored very high on the
test do not have the intellectual accomplishments that one would
reasonably (but not with 100% certainty) expect of such a group.
> Before discussing objections you lodge to specific questions, let me
> note in passing the related sophistry that some people fall into when
> arguing that intelligence tests are invalid. The argument goes
> something like this: intelligence tests are invalid because look at
all
> the people who score poorly who do well in life (and vice versa).
But
> this assumes that there is a generally agreed upon notion of "doing
> well in life." I think that's a lot less generally agreed upon than
> what intelligence is.
Well I don't think intelligence tests are invalid, you correctly
surmised this above. In fact many years ago I defended the use of one
such test, the LSAT, in graduate school admissions in a column for my
school newspaper. The mere fact that so many competitive and
intellectually demanding organizations, from universities to investment
banks, require that applicants disclose their standardized test scores
is very strong evidence for the usefulness and validity of standardized
tests as a measure of intelligence.
> So on to your objections to specific questions. I think there is a
> general consensus among Mega Society members that Mega and Titan Test
> questions are too culturally biased and too research oriented.
Actually the problem of cultural bias is the least of its problems.
Were it the only problem then it would at least be valid for us Western
native English speakers. In fact saying that the problem with a test
that includes any verbal questions is that it is culturally biased is
really praising with faint damnation, even the best tests that include
verbal problems will have such a flaw.
The Titan Test may be a fun set of analogies and puzzles, but compared
to REAL intelligence tests such the major IQ tests and the ETS tests
(SAT, GRE, LSAT, GMAT, etc.) it is worthless crap. The questions are
not well written, there is insufficient question variety, the format is
not one that can reliably measure IQ as it is commonly understood, and
the test conditions of the takers are extremely variable. Yet it claims
to be *better* than real intelligence tests in measuring the IQ's of
very smart people.
> We have
> started to make some tentative steps to fix this (see the very
> preliminary online test at http://www.mental-testing.com), but these
> objections should not be overblown. Even if the test is culturally
> biased, this means only that certain test takers are not going to
score
> highly.
As discussed above, my problem is not cultural bias. I have looked at
this website and while the result if everything goes according to plan
will be an improvement on the Titan and Mega tests it will still not
measure IQ in the 150-180 range. I also observe that as someone who is
particularly good at one method of estimating intelligence (puzzles)
you and the other designers of the test are biased in favor of such
measures, and this bias shows on both the untimed Mega and Titan test
as well as the questions on your website. Part of being smart is also
ingesting information and processing it very quickly, this is not
tested at all. Any good intelligence test ought to test this so at
least part of it must be timed.
> Given that these are the first tests that can distinguish IQs
> in the 150 - 180 range, it isn't a critical flaw that they exclude
some
> people.
No, they can't distinguish IQ in the 150-180 range. I do think such a
test is possible, the Titan Test, however, is so poorly made that it
does not come even close to doing so. I don't know about tests that go
all the way up to 180, but tests that measure IQ reliably up the 160's
exist. My quick estimate is only one person in 60,000 could score 180
on two consecutive LSAT exams. 25 in 100,000 get an 180 per exam, and
about 1/4 of those who get a 180 on one exam get another 180 if they
retake the test. So that is 8 in 100,000. I then multiply the
denominator by 5 to account for the fact that the sample of those who
take the test are generally college graduates with GPA's above 3.0 to
get 8/500,000 or 1/60,000. So the LSAT can measure IQ up to 161 if
taken once and 166 if taken twice.
While not a perfect test, the LSAT is a very good test of IQ and much
more a high-end IQ test than anything on Miyaguch's site. Getting a
perect score on the GRE general test and several of the subject tests
is also a feat only a few people in a million could do.
> One man's trivia is another man's profound insight. I don't think
it's
> trivial to know about the work of Ludwig Wittgenstein. I don't think
> it's trivial to know what a language game is. You are implying that
> the test is replete with pop culture trivia akin to "Jeopardy!"
So you're saying that knowledge of the history of philosophy is a valid
thing to test on an intelligence test. To some extent I agree, but the
question is an example of how poorly the test is written. It isn't a
question about W's major ideas, but a bit of trivia about his life. I
bet most of the greatest and most intelligent philosophers in the
Western world could not answer that question. Those who have happened
to have read a biography of Wittgenstein, however, probably would know
the answer. A better question would not be of this silly bit of trivia
but one that tested in a straightforward way if one understood, for
instance, Berkeley's argument against substance, which is important to
the history of philosophy, and would make a good question because those
with a broad knowledge of philosophy would probably know it, and
someone without one probably would not know it.
gregw...@yahoo.com
Chris Canon wrote:
> I don't like being called elitist. I've explicitly disclaimed that
the
> Mega Society is about potential greatness. The Mega Society is no
more
> elitist than a club soccer team. You need to make the cut, but that
> doesn't make you a good person.
Bad analogy. The soccer club only says that its members are better
players than those who tried out and didn't make the cut. And it might
not even go that far. It may only say its members are the best fits for
their teams.
On the other hand the Mega Society claims to be an organization of
those with IQ's higher than 999,999 out of 1,000,000 people. Big
difference.
gregw...@yahoo.com
allow people to claim that they have one in a million IQ's and are
smarter than most faculty at Harvard and MIT, an arrogant and silly
thing to do, especially since it is very unlikely that they do in fact
have one in a million IQ's or are smarter than most faculty at Harvard
and MIT. Posting answers is one way of subverting this arrogance. Plus
some of the questions are fun.
mensa...@aol.com
gregw...@yahoo.com wrote:
> I hope people continue to post answers. The purpose of the
> test is to allow people to claim that they have one in a
> million IQ's
If you search through their web site, you'll find that they
acknowledge that there is NO test that can distinguish IQs
in the one in a million range.
But it's in very small print so as not to affect sales.
Canon
The ceiling of the various standardized tests available is around 150.
This is because they contain problems designed to accurately spread the
testing population within three standard deviations of the mean. For
example, the SAT, or at least the old SAT, reports scores on each
subtest with a mean of 500 and standard deviation of 100, in the range
200 to 800. The reason for this is simple: economics. 99+% of the
population occurs within three standard deviations of the mean. On the
16 point standard deviation scale, this means that a perfect score on
the SAT corresponds to an IQ of 148.
About one in a million high school students each year scores a perfect
score (1600) on the SAT. Does this mean that a perfect score on the
SAT corresponds to an IQ of one in a million? No. In order to
distinguish between 3 standard deviations and 4.75 standard deviations
(which is about one in a million rarity) the test would need to contain
items with difficulties in that range. For example, it would need to
contain several questions that only one person in, say, 100,000 could
answer. But in fact it does not contain any such questions, simply
because there is no market for measuring IQs above 3 standard
deviations.
There are various theories about what are the aspects of intelligence.
Some that have been proposed are speed of thinking, size of working
memory, size of associative store, and specialized abilities like 3D
visualization. I am ignoring here other abilities like musical talent
that figure in a wider definition of intelligence, although I would
include the ability to exhibit insight that is frequently associated
with creative thinking. I include this because I know from computer
science that a better algorithm trumps brute force.
Each question on an IQ test will probably exercise several aspects of
intelligence. To avoid systematic bias it is necessary to hedge the
test with questions that have complementary biases. This is why most
tests have verbal and mathematical sections, and further subsections
within these. It is not a defect for a problem to require specialized
knowledge, as long as the specialist would be disadvantaged on other
questions.
Mid-range standardized IQ tests tend to emphasize those aspects of
intelligence that are easy to test. The simplest of these is speed.
Chronometrics, the study of the relationship between speed of thinking
and intelligence, has show that the correlation is poor at higher IQs.
This is true even in sub-150 range, and the effect is believed to be
negligible at higher IQs.
A test that can measure IQs in the range of 150 - 180 must contain
problems that can only be answered by a very small fraction of the
population. It is not enough to simply come up with questions that are
hard to answer. For example, it is not enough to ask someone to do
long division by hand, because this might just be a test of
perserverence. Aside from the all the other aspects mentioned above,
ideally the high range question should require some kind of insight,
because again this is the most powerful aspect of intelligence. The
Mega and Titan Tests tend to be rich in insight-requiring problems.
mensa...@aol.com
So did you guys ever get this issue straightened out?
<quote>
Concerning Mega Admission Standards
Kevin Langdon
(reprinted from Noesis #125)
Therefore, Mega's de facto cutoff is either 43/172 (approximately
the 99.9997th percentile, one in 300,000, or 44/173 (approximately
the 99.99975th percentile, one in 400,000), and this is about as
high as the tests currently in use can reasonably be claimed to
measure.
We must face the question of the limits to discrimination of the
available selection instruments. We must either make a serious
attempt to accept members at the 99.9999th percentile and raise
our standards to the one-per-million level or accept that our
qualifying level is 2.5- or 3-per-million and cannot be higher
(due to the state of the art of high-range psychometrics) and
drop our qualifying scores.
</quote>
Are you still claiming the Titan Test discriminates at the
one-in-a-million level?
Canon
who has scored at the one-in-a-million level on a test of general
intelligence that is credibly claimed by its authors to be able to
discriminate at this level. Ron Hoeflin is the auhtor of the Mega and
Titan Tests, and he claims the one-in-a-million level on the Titan Test
is at 43 correct out of 48. Kevin Langdon has published criticisms of
some details of this norming, as you cite above. A majority of the
membership voted to accept the Titan Test with a cutoff of 43. Since
the Constitution of the Society also explicitly states that no member
may claim to speak for the Society as a whole, I will leave the meaning
of that vote up to others to interpret. As for myself, I think the
issue deserves further study, but I'm willing to accept Dr. Hoeflin's
assessment until a new approach is investigated. One of the things we
are doing with mental-testing.com is investigating various ways to norm
high range tests, such as new Bayesian techniques for statistical
inference.
gregw...@yahoo.com
> Here are my answers to some of the points raised in the last few
posts:
>
> The ceiling of the various standardized tests available is around
150.
Chris Canon does not know what he is talking about. In a previous post
I showed how the LSAT can discriminate up to IQ 161 if taken once and
IQ 166 if taken twice. Standardized IQ tests measure IQ up to 164, or
SD 4.
> This is because they contain problems designed to accurately spread
the
> testing population within three standard deviations of the mean.
Chris Canon does not know what he is talking about. He is wrong when he
says that such tests only attempt to spread the test population within
three standard deviations of the mean. In fact they can, and do
discriminate above 3 SD. I not only doubt his self-flattering pretend
IQ of 170+ but his basic reading comprehension when it comes to the
scoring scales for standardized tests. He also do need to be able to
discriminate at the level above 3 SD, and the standardized tests they
make their applicants do just that.
For
> example, the SAT, or at least the old SAT, reports scores on each
> subtest with a mean of 500 and standard deviation of 100, in the
range
> 200 to 800.
No, you don't know what you are talking about. The old scale of the SAT
was never mean 500 SD 100.
> About one in a million high school students each year scores a
perfect
> score (1600) on the SAT.
Completely false, by orders of magnitude. In 2004 939 out of 1.42
million, or about 1 in 1500 got a score of 1600 on the SAT I. Chris
Canon once again shows he does not know what he is talking about.
> Does this mean that a perfect score on the
> SAT corresponds to an IQ of one in a million? No. In order to
> distinguish between 3 standard deviations and 4.75 standard
deviations
> (which is about one in a million rarity) the test would need to
contain
> items with difficulties in that range.
False. A test could have no questions of that difficulty and still
discriminate in that range if there were groups of questions that are
this difficult to answer all correctly. Chris Canon once again shows he
does not know what he is talking about.
> For example, it would need to
> contain several questions that only one person in, say, 100,000 could
> answer.
False, per above.
> But in fact it does not contain any such questions, simply
> because there is no market for measuring IQs above 3 standard
> deviations.
Wrong again, there is a such a market and standardized tests meet it.
You also fail to consider that the SAT is more than just the SAT I. In
combination with SAT II subject tests the SAT can further discriminate
among very high scorers and between the many students who score 1600 on
SAT I.
Please not that Chris Canon failed to answer all of the following
points:
How the the Titan Test can measure IQ between individuals if test
conditions are not the same and not controlled for?
Why does the Titan Test not test speed? To not test speed he must
maintain that to answer a question in one hour is not a sign of greater
intelligence that answering the question in one week. Nonsense. I call
bullshit on his vague appeal to "chronometrics" that speed is not a
good measure of intelligence at a high ranges. Where is your evidence?
Chris Canon also fails to answer these questions to his ambiguous
assertion that he does not actually believe that he is smarter than
999,999 out of a million people despite his membership in a group that
claims to be a high-IQ society and claims to only accept members with
purported one-in-a-million scores on certain purported IQ tests.
For his convenience, I will repeat this part of my pervious post:
Canon
I will take this issue offline with him.
Here is a chart from the College Board site
(http://www.collegeboard.com/prod_downloads/about/news_info/cbsenior/yr2003/pdf/table_3b.pdf).
At the bottom you will see that the mean of the 2004-5 SAT Composite
V+M is 1026 and the standard deviation is 210. Actually, I understated
my argument. The mean of each subtest is over 100, so the full range
of the SAT is less than 3 standard deviations. This is a well known
fact and is not disputed by anyone in the psychometric community.
The figure of one 1600 per million is based on the old SAT. Since the
SAT was "recentered" this figure has gone up. Since this
recentering no high IQ society accepts scores from the SAT. Again, I
understated the argument. The fact that many more people now get
perfect scores indicates that the test is even less able to
discriminate at the high end than it previously was.
The LSAT has the additional complication that it is a test of
specialist skills. As we've already discussed, if a test of general
intelligence contains a question that requires specialist skills, it
must also contain questions that disadvantage the specialist in order
to hedge against systematic bias. The LSAT obviously does not contain
any questions that disadvantage the legally trained testee.
Here is a statistics question for you: How many questions in the 0 -
3 standard deviation range would you need to ask to get an accurate
assessment of IQ at 4.75 standard deviations? If you do this analysis,
you'll understand why the SAT is not a high range test.
Perhaps this argument will help too. Do you think the SAT makes a good
low range test? For example, do you think it should be used to
distinguish what kind of learning disability a person with IQ of 50
has?
You were the one who stated that speed is an important component of
high intelligence. What is your evidence? Since you've apparently
never heard of cognitive chronometrics, this looks like an assertion
you made without anything to back it up.
The test conditions for the SAT are not a positive feature of the test.
They do not approximate the conditions for decision making in real
life. For most people the SAT is one of the worst experiences of their
lives.
My name is Chris Cole, not Chris Canon.
Ed Murphy
> You were the one who stated that speed is an important component of high
> intelligence. What is your evidence? Since you've apparently never heard
> of cognitive chronometrics, this looks like an assertion you made without
> anything to back it up.
Speed indicates that you've internalized certain patterns of knowledge and
can apply them instinctively, without having to apply conscious thought to
each step of the process. Do you assert that this internalization and
instinctive application is not an important component of high intelligence?
Canon
My comment quoted about was in response to a criticism that the Titan
Test needs to include some timed questions. As far as I can tell,
there is no evidence that differences in the seconds can be used to
measure high intelligence.
But even if that is wrong, it is not relevant because a test can't use
every technique. The Titan Test passes the statistical significance
norms that are accepted for all the other usually cited tests of
intelligence. It is the only unspolied test normed in the 150 - 180
range.
What I am saying is that it's a shame to spoil it by posting answers.
gregw...@yahoo.com
Chris Cole wrote:
> At the bottom you will see that the mean of the 2004-5 SAT Composite
> V+M is 1026 and the standard deviation is 210.
And you said it was mean 1000 and SD 100. You were off by 2.6% and
110%. You simply don't know what you are talking about. You open your
mouth without knowing if the crap you spew is true or false.
> Actually, I understated my argument.
"I said something that was false, but know I know the truth I am even
more right" is a really pathetic excuse for confidently posting false
statistics.
> The figure of one 1600 per million is based on the old SAT.
But you didn't say that. You said the one in a million score was for
"each year." You were wrong by several orders of magnitude.
Lesson for everyone: Chris Cole know does not the first thing about
psychometric testing. His claim to be one of the smartest 300 people in
America based on his phoney IQ Test is a conceited lie.
> Again, I
> understated the argument.
No, again you made false statements regarding an intelligence test. You
can't be trusted on this topic, untruths just pour out of your keyboard
and onto the Internet.
> The LSAT has the additional complication that it is a test of
> specialist skills.
False again. Anyone trusting Chris Cole's claim that the Titan Test and
the Mega Society can distinguish people of high IQ need to remember
that Chris Cole thinks he knows many things that just ain't so. Lord
knows how many other such untruths pollate his conceited
"one-in-a-million" brain.
> as we've already discussed, if a test of general
> intelligence contains a question that requires specialist skills, it
> must also contain questions that disadvantage the specialist in order
> to hedge against systematic bias. The LSAT obviously does not
contain
> any questions that disadvantage the legally trained testee.
Except the LSAT is a test of general intelligence, and contains
precisely zero questions that advantage or disadvantage a legally
trained testee, and is not designed to test legal knowledge in any way.
Again the lesson is: don't trust anything Chris Cole has to say about
intelligence tests. He just doesn't know what he is talking about, and
will say false things to support his position.
> The test conditions for the SAT are not a positive feature of the
test.
Yeah, because of all that cognitive chronometrics research that all the
major universities of America foolishly ignore, but YOU Chris Cole,
despite your ignorance of the most basic elements of the SAT, are smart
enough to point out as a source of a flaw in the SAT. You are not only
smarter than most of the faculty of Harvard and MIT, you also smarter
than their admissions staff!
Get real. You're not "Mega" anything, and your phoney IQ tests and
psuedo-scientific babble about "normalizing" your phoney IQ tests
doesn't mean they are better *in any way* than the SAT, the LSAT, or
the major IQ tests in measuring IQ, in the high range or otherwise.
> They do not approximate the conditions for decision making in real
> life.
Decisions don't have to be made quickly in everyday lives? Of course
they do. Intelligence ests are timed for a reason, getting X answers
right in 10 minutes indicated intelligence better than X answers right
in two weeks.
>For most people the SAT is one of the worst experiences of their
lives.
Which proves nothing, other than many people can't accept their SAT
test scores so feel compelled to write and or take their own tests that
cater to their own stregths (speed not being one of them) to confirm
their supposed "MEGA" intelligence.
Ed Murphy
> Ed Murphy wrote:
>> On Sat, 16 Apr 2005 04:53:00 -0700, Canon wrote:
>>
>> > You were the one who stated that speed is an important component of
> high
>> > intelligence. What is your evidence? Since you've apparently
> never heard
>> > of cognitive chronometrics, this looks like an assertion you made
> without
>> > anything to back it up.
>>
>> Speed indicates that you've internalized certain patterns of
> knowledge and
>> can apply them instinctively, without having to apply conscious
> thought to
>> each step of the process. Do you assert that this internalization
> and
>> instinctive application is not an important component of high
> intelligence?
>
> My comment quoted about was in response to a criticism that the Titan Test
> needs to include some timed questions. As far as I can tell, there is no
> evidence that differences in the seconds can be used to measure high
> intelligence.
People have been discussing differences between minutes and
weeks. Your "differences in the seconds" is a straw man.
> But even if that is wrong, it is not relevant because a test can't use
> every technique. The Titan Test passes the statistical significance norms
> that are accepted for all the other usually cited tests of intelligence.
> It is the only unspolied test normed in the 150 - 180 range.
>
> What I am saying is that it's a shame to spoil it by posting answers.
Then it's *not* unspoiled.
Canon
Ed Murphy wrote:
> On Sun, 17 Apr 2005 17:19:38 -0700, Canon wrote:
>
> >> Speed indicates that you've internalized certain patterns of
> > knowledge and
> >> can apply them instinctively, without having to apply conscious
> > thought to
> >> each step of the process. Do you assert that this internalization
> > and
> >> instinctive application is not an important component of high
> > intelligence?
> >
> > My comment quoted about was in response to a criticism that the
Titan Test
> > needs to include some timed questions. As far as I can tell, there
is no
> > evidence that differences in the seconds can be used to measure
high
> > intelligence.
>
> People have been discussing differences between minutes and
> weeks. Your "differences in the seconds" is a straw man.
The October 2004 SAT had 138 questions to be finished in 2 hours and 30
minutes, for an average of 65 seconds per question. The difference
between answering the question and not answering the question could be
measured in seconds, not minutes. The cognitive chronometirc work I
was citing measures response times in seconds or fractions of a second.
Thus the phrase "differences in seconds" was not a straw man.
So that this post is not a mere argument over terminology, let me point
out a major problem with timed tests. As you stated above, much of
what goes on in thinking occurs in the unconscious mind and we are not
consciously aware of it. An intelligence test is a benchmark of human
thinking, and a benchmark is a simulation of realistic conditions.
Most people taking the SAT find it to be stressful. Some people report
that they do not remember what happened, and others report that they
missed simple questions because they "couldn't think." When the
unconscious mind is occupied warding off incipient panic, it is not
available to process information in the way that it normally would.
Thus the SAT is a lousy benchmark. It is more aptly termed the "Stress
Aptitude Test."
mensa...@aol.com
> let me point out a major problem with timed tests. As you
> stated above, much of what goes on in thinking occurs in
> the unconscious mind and we are not consciously aware of it.
> An intelligence test is a benchmark of human thinking, and
> a benchmark is a simulation of realistic conditions. Most
> people taking the SAT find it to be stressful. Some people
> report that they do not remember what happened, and others
> report that they missed simple questions because they
> "couldn't think." When the unconscious mind is occupied
> warding off incipient panic, it is not available to process
> information in the way that it normally would. Thus the SAT
> is a lousy benchmark. It is more aptly termed the "Stress
> Aptitude Test."
So simply removing the time stress makes your tests
"a simulation of realistic conditions"? It does not.
There's one aspect of "real" problem solving you're
overlooking: feedback. In the real world, what's
important is getting the right answer. When I write
a database query, the boss doesn't stand over me with
a stopwatch. And he doesn't care how many times I had
to revise it. The ONLY thing he cares about are the
correct numbers appearing in the report. And there
are plenty of real world situations where simply
knowing that an answer is wrong is sufficient feedback.
For example, on that silly mental-testing website
there's a problem about refueling a plane in flight.
I worked out the math, checked and rechecked, and
confidently posted an answer. I was promptly informed
that the answer was wrong. Didn't give the correct
answer, didn't provide any hint as to why my answer
was wrong, but just knowing that it was made me
immediately recognize what I had overlooked. Had I a
do-over, I would have gotten it right. But there's no
provision for that on yours or anyone else's test.
And it would be messy to try to incorporate that,
wouldn't it? Instead of having a nice clean black and
white picture of intelligence, you would have various
shades of grey. And although less realistic, the black
and white picture is easier to score and makes
pigeon-holing people easier than the grey image.
"I'm looking over here because the light's better"
- classic crackpot science.
Ed Murphy
> Ed Murphy wrote:
>> People have been discussing differences between minutes and weeks.
>> Your "differences in the seconds" is a straw man.
>
> The October 2004 SAT had 138 questions to be finished in 2 hours and 30
> minutes, for an average of 65 seconds per question. The difference
> between answering the question and not answering the question could be
> measured in seconds, not minutes. The cognitive chronometirc work I was
> citing measures response times in seconds or fractions of a second.
> Thus the phrase "differences in seconds" was not a straw man.
Fair enough, as far as it goes, but you've failed to defend the Titan
Test's decision to not test speed at all.
> So that this post is not a mere argument over terminology, let me point
> out a major problem with timed tests. As you stated above, much of what
> goes on in thinking occurs in the unconscious mind and we are not
> consciously aware of it. An intelligence test is a benchmark of human
> thinking, and a benchmark is a simulation of realistic conditions. Most
> people taking the SAT find it to be stressful. Some people report that
> they do not remember what happened, and others report that they missed
> simple questions because they "couldn't think." When the unconscious
> mind is occupied warding off incipient panic, it is not available to
> process information in the way that it normally would. Thus the SAT is a
> lousy benchmark. It is more aptly termed the "Stress Aptitude Test."
Stress *is* a realistic condition. The ability to think clearly despite
stress is, in many situations, an important component of intelligence.
Tim S Roberts
"Ed Murphy" <emur...@socal.rr.com> wrote in message
news:pan.2005.04.19...@socal.rr.com...
I've got a great counter-argument to that, but I can't think of it at the
moment.
Tim
Canon
> For example, on that silly mental-testing website
> there's a problem about refueling a plane in flight.
> I worked out the math, checked and rechecked, and
> confidently posted an answer. I was promptly informed
> that the answer was wrong. Didn't give the correct
> answer, didn't provide any hint as to why my answer
> was wrong, but just knowing that it was made me
> immediately recognize what I had overlooked. Had I a
> do-over, I would have gotten it right. But there's no
> provision for that on yours or anyone else's test.
>
> And it would be messy to try to incorporate that,
> wouldn't it? Instead of having a nice clean black and
> white picture of intelligence, you would have various
> shades of grey. And although less realistic, the black
> and white picture is easier to score and makes
> pigeon-holing people easier than the grey image.
> "I'm looking over here because the light's better"
> - classic crackpot science.
Actually, we do have a mechanism for allotting partial credit (see the
hangman problems, for example), so we could allow people to submit
second guesses. I'm assuming from your post that you think that would
be a good idea. What do other people think?
Canon
Ed Murphy wrote:
> Fair enough, as far as it goes, but you've failed to defend the Titan
> Test's decision to not test speed at all.
Actually, the Titan Test does test speed, though not in the seconds.
It tests speed in the way you mentioned earlier, namely, the ability to
search a large body of knowledge looking for a pattern. For that to
happen in a feasible time frame, you need to be able to think quickly.
It does not test speed in the seconds because research has shown that
this is not well correlated with intelligence at high levels. Although
I don't know why that is for sure, one model is that people have a
limited range fo thinking speeds, and that people max it out at, say,
the 90% percentile.
> Stress *is* a realistic condition. The ability to think clearly
despite
> stress is, in many situations, an important component of
intelligence.
Fair enough. Returning to the benchmark analogy, there are different
benchmarks for different purposes. There are benchmarks for gaming
systems and there are benchmarks for office systems. This relates to
another poster's statement that there is no such thing as general
intelligence. There's a kernel of truth to that statement. Just as
there is no one universally accepted benchmark for computers, there is
no one definition of general intelligence.
So perhaps we can agree to disagree on this point. I think being able
to think under stress is not an important component of intelligence.
Nick Wedd
gregw...@yahoo.com writes
>Some very conceited people claim to be among the very smartest in the
>world and have formed a large number of pretentiously-named clubs to
>revel in their supposed intellectual superiority. This is despite very
>little (or rather absolutely nothing) in the way of actual intellectual
>accomplishments among them to substantiate their incredible claims of
>
>Their main club is the "Mega Society" which they say is reserved for
>those with one-in-a-million IQ's, ie above 176. Of course for good
>reason conventional IQ tests do not have a scale that reaches such a
>height, so they have created what they claim is a test that can measure
>the intelligence of those with very high IQ's.
>
>When I first heard about this four years ago, I decided to put a stop
>to this by publishing the answers to their current test, the Titan Test
>(notice again their extreme vanity in choosing this name). This had
>been done before, to their Mega Test, which they had to then abandon as
>a criterion of membership.
I can kind of understand why people might want to join such a society,
though I would have no wish to join one myself. But I cannot understand
why you object so strongly to it, and why you dismiss their claims as
"incredible".
If some people set up a club for the five thousand fattest people in the
world, would you also be offended by that, and set out to sabotage their
scales?
Nick
--
Nick Wedd ni...@maproom.co.uk
mensa...@aol.com
> "incredible".
What's "incredible" is that the claims that these tests measure
intelligence is disputed. Everyone knows that, even the Mega Society
acknowledges that the tests are dubious. And yet, they voted to make
those tests the standard for admission. Since when did scientific
truth become a democracy? You don't find it "incredible" that someone
uses crackpot science as the basis for their society?
>
> If some people set up a club for the five thousand fattest people in
the
> world,
How is "fattest" defined? Weight? Girth? Regardless, the criteria
chosen is scientifically verifiable.
> would you also be offended by that,
Suppose the club was for the five thousand "handsomest" people.
Would you be offended if, by the dubious definition of "handsome",
you noticed that only Caucasians were members of the club?
> and set out to sabotage their scales?
And what if the scales were biased? And what if the people operating
the scales knew they were biased but voted to keep using them anyway?
And what of the rights of the victims of this fraud?