Noesis
The Journal of the Mega Society
Number 83
July 1993
EDITORIAL
Rick Rosner
5139 Balboa Blvd #303
Encino CA 91316-3430
(818) 986-9177
IN THIS ISSUE
MORE POMFRIT
LETTER FROM RON HOEFLIN ON ANALOGIES, ULTRA, OTHER TESTING ISSUES
MORE ON THE SHORT FORM TEST BY CHRIS COLE
TWO PROBLEMS FOR SHORT FORM TEST FROM PETER SCHMIES
CHALLENGING TEST FROM ALAN AAX
HEAVY ICE VS. LIGHT WATER BY ROBERT DICK
LETTER FROM CHRIS HARDING CONCERNING CTMU
RELATIVITY ARTICLES FROM ROBERT HANNON
POMFRIT'S ANALOGIES
|
61. INITIALS |
NAME |
MONOGRAM |
|
62. 26 |
62 |
PSEUDORHOMBICUBOCTA-HEDRON |
|
63. PLEASURE |
HEDONIA |
PLEASURE FROM ANTICIPATING SUCCESS |
|
64. WATCHING |
VOYEUR |
BODILY SMELLS |
|
65. RICH |
POOR |
WEALTH |
|
66. MAN |
WOMAN |
ANULOMA |
|
67. DEDUCTIVE |
INDUCTIVE |
APRIORISTIC |
|
68. 45 |
63 |
ASANKA |
|
69. EARTH |
NAVIGATION |
SPACE |
|
70. GRANDMOTHER'S SISTER |
GRANDFATHER'S BROTHER |
ATMATERTERA |
|
71. ROMAN CATHOLIC |
ROSARY |
MOHAMMEDAN |
|
72. ITALY |
GONDOLA |
MALTA |
|
73. DIFFICULT |
EASY |
DYSANIA |
|
74. OX |
CAT |
BOANTHROPY |
|
75. MARRIAGE |
STAYING SINGLE |
GAMOPHOBIA |
|
76. INDIVIDUAL |
EGOTISM |
GROUP |
|
77. WAX IMAGE |
INVULTATION |
PICTURE |
|
78. CHILDREN |
GIFTED CHILDREN |
PEDIATRICS |
|
79. WOMAN |
MAN |
SQUAW |
|
80. STEPMOTHER |
STEPFATHER |
NOVERCAPHOBIA |
|
81. MARS |
JUPITER |
AREOGRAPHY |
|
82. PERSON |
PASSPORT |
CAR |
|
83. DOWN |
UP |
PITCH |
|
84. HOSPITAL ADMINISTRATION |
NOSECONOLOGY |
MEDICAL REMEDIES |
|
85. INCENSE-BURNING |
KNISSOMANCY |
BUBBLES RISING IN A FOUNTAIN |
POMFRIT'S SERIES
|
BB. 7, 14, 23, 36, 57, 94, 163, ? |
CC. 1, 18, 108, 270, 500, 924, 1638, ? |
|
DD. 10, 190, 1710, 1710, 9690, 38760, 116280, 271320, ? |
EE. 0 3 2 2 5 8 0 ?
|
|
FF. -9, -18, -24, -16, 70, 660, 4970, ? |
GG. 1 7 7 2 4 5 3 8 ? |
|
HH. 7 3 8 9 0 5 6 0 ? |
II. 1, 7, 37, 175, 781, 3367, 14197, ? |
|
JJ. 3 3 5 4 4 3 5 ? |
KK. 3 6 8 4 0 3 1 4 ? |
|
LL. 1 1 4 4 7 2 9 8 ? |
MM. -7, 7, 45, 119, 241, 423, 677, ? |
|
NN. 5, 6, 19, 31, 330, 942, 107958, ? |
OO. 9, 22, 24, 12, 3, 4, ? |
|
PP. 1, 3, 7, 13, 21, 31, 43, ? |
QQ. 11, 31, 71, 91, 32, 92, 13, 73, ? |
|
RR. 5 4 6 9 7 5 8 1 9 ? |
SS. 999, 1329, 1725, 2193, 2739, 3369, 4089, ? |
|
TT. 8, 2, 3, 14, 15, 6, 16, ? |
UU. 89, 13, 95, 51, 18, 33, 85, ? |
|
VV. 20, 23, 5, 14, 20, 25, 0, 20, 23, ? |
WW. O, 32, 10, 50, 20, 68, 30, 86, 40, ? |
|
XX. 8 6 6 0 2 5 4 ? |
YY. 9 8 6 9 6 0 4 ? |
A LETTER FROM RON
HOEFLIN
[Ed's comment: Congratulations to Ron Hoeflin on the dismissal of the nuisance suit against him for reprinting an uncopyrighted article.]
MORE ON THE SHORT FORM TEST BY CHRIS COLE
I first want to apologize to Ron Hoeflin for publishing the answers to several of his questions. This resulted from a misunderstanding on my part. I thought that Ron was donating all these problems to the Short Form Test, and that he did not intend to do the Ultra Test. I published the problems (and regrettably the answers) because I needed some examples of what is wrong with current hi-end tests, and it is exceedingly difficult to come up with good questions for examples. Since I thought these problems were basically "retired," I did not think it was wrong to publish them. I hope I have not compromised them beyond repair. Sorry, Ron.
As should be clear from Ron's preceding letter, the plan to consolidate the Short Form Test and the Ultra Test is defunct. Ron will go forward with the Ultra Test on his own, and I still hope one day to put together the Short Form Test with the help of the rest of you. By the way, we are now in need of a name, as the name "Short Form" is a misnomer. The test will not be any shorter in form than the Mega, Titan or Ultra, although hopefully it will be shorter in time. Since I note that members are rather good at coming up with creative names, perhaps one of you will be inspired. If so, please pass along the result.
With that said, let me rush to my own aid and attempt to resurrect my reputation as an empiricist. Ron argues that we should let discrimination value alone determine our choice of test questions. I am of course in basic sympathy with this statement, but I have two systematic objections:
1. The population on which the trial questions are being tried is not being randomly sampled from the world's population as a whole. This leads to a systematic bias, which is the bane of all statistics. I am reminded of the story of the student defending his Master's thesis. Part of the thesis was a statistical survey conducted by the student. One of the professors on the thesis defense committee asked the student how he had conducted his survey. The student answered that he had randomly selected locations in the city to stand and interview people. The professor asked when the student did this. The student answered that he did it Tuesdays and Thursdays at 4:00 p.m. The professor asked how these times had been chosen. The student answered that they were the only times that he did not have classes. The student's thesis was rejected, because the survey was systematically biased by the times the student chose to conduct the interviews.
Now, how can we correct for the sampling bias of the trial test process? The complete answer to this question is probably quite difficult to formulate. But we can at least avoid some of the obvious problems of culture-bias, such as questions relying on knowledge of Greek mythology, English etymology, etc. Lest anyone think this is academic, I have personally spoken with people who object to the Mega test and cite this as their main reason.
2. Some questions can be answered well by a computer which has access to a large dictionary,
encyclopedia, almanac, etc. Not everyone has access to such things, but as time passes more and more people will, and the ability to do simple information retrieval is not a test of intelligence. We already implicitly acknowledge this when we exclude questions requiring specialist knowledge from the test. Ron, for instance, will not use a math question that requires calculus. (Ironically, I do not agree with this, since I think virtually every high school in the country now teaches calculus). We all agree that questions requiring, say, specialist knowledge of archaeology are inappropriate. Why? Because we are trying to measure intelligence, not knowledge acquisition. And we are trying to measure it in an unbiased way. So, a question about the social mores of the Etruscans is biased in favor of a specialist, who would come by such knowledge not by way of superior intelligence, but rather by way of making a living.
This is a rather fine line to draw. We like analogies like:
A, AB, B, BO, O : BO :: A, C, E, G, T: E
because we think that people "should" know about blood types and DNA bases, but of course not too long ago this was specialist knowledge, and not too long before that it was nonexistent knowledge.
Perhaps like the Wizard of Oz we want to test what people have learned when they weren't trying to learn.
It is easy to criticize. Can I come up with some problems that (1) are not culturally biased, and (2) cannot be solved by computer? I have used the resources of the Internet to try the following set of problems out on our very small population of networked-connected Mega members (currently, seven of us). Every one of these problems was solved by at least one Mega member. Also, most of the problems have very good near solutions, which may allow us to do something that Rick has argued for: assign partial credit. So, try these out and send in your answers: Also, let us know which of Peter Pomfrit's problems meet the above criteria, and try out Peter Schmies' problems below.
32. backbone : tailbone :: letter : ?
33. purple : orange :: child : ?
34. mirror : mercury :: balloon : ?
35. queen : knight :: telescope : ?
36. horse : saddle :: wind : ?
37. nail : screw :: musket : ?
38. stereo : monaural :: drumsticks : ?
39. mop : evaporate :: shovel : ?
40. grass : trees :: bacteria : ?
41. chameleon : mocking bird :: circus : ?
TWO PROBLEMS FROM PETER SCHMIES
42. 6 20 14
4 15 7
10 39 ?
9 53 26
43. B-V H-N P-R C-?
LETTER AND TEST FROM ALAN AAX
HEAVY ICE VS. LIGHT WATER BY ROBERT DICK
LETTER FROM CHRIS HARDING CONCERNING CTMU
Dear Rick:
I have read Chris Langan with considerable interest and I believe I have understood the point he is driving at. As others have pointed out, THERE STILL REMAINS A DIFFICULTY. Using CTMU can he provide us with the NEW IDEAS that can lead to verification through the experimental process? Does it point to things as yet unknown? A theory of everything would through expansion lead to everything that we now don't know of and be able to tell us everything about everything. True, such an "expansion" may still be some time off and any failure in this regard may not necessarily prove fatal. However, interesting results may be close at hand given the rightness of his "frame."
I believe the failure of the General Theory of Relativity may be found in the fact that the form of these equations are those of the simple partial differential. This as much applies to Quantum Mechanics. No wonder they are both in conflict! The reality is, such equations cannot deal with infinities. Reflecting on this suggests that this failure can similarly be found in the structure of language logic and the very conceptual process on which all of our human culture rests and is a profound social (external) limitation of the human mind [cf. the paradox].
Can I then issue a challenge to Chris Langan to solve the N. Chomsky puzzle and give us the underlying structure on which language rests? and give us a tool for dealing with the conceptual difficulties we are beset with.
Best Regards,
Chris Harding
RELATIVITY ARTICLES FROM ROBERT HANNON