[mod.ai] Nonsense Quiz

Wedekind.ES@XEROX.COM (10/22/86)

A couple of years ago, on either this list or Human-Nets, there appeared
a short multiple-choice test which was written so that one could deduce
"best" answers based on just the form, not the content, of the questions
(in fact there wasn't much content, since almost every word over 3
letters long was a nonsense word).

If anyone has this test, I would very much like to see it (along with
any "official" answers you may have saved).  If you want to see what I
receive (or, better yet, if you have any original questions to add to
the test), just let me know.

thanks,
Jerry 

Laws@SRI-STRIPE.ARPA (Ken Laws) (10/30/86)

Here's a copy of the quiz taken from Human-Nets.  Those interested
in such things should get a copy of

  R.M. Balzer, Human Use of World Knowledge, ISI/RR-73-7,
  Information Sciences Institute, March 1974, Arpa Order
  No. 2223/1.

It contains fairly detailed analysis of text such as "Sooner or later
everyone runs across the problem of pottling something to a sprock
inside the lorch."



Date: 9 Sep 1981
From: research!alice!xchar   [ Bell Labs, Murray Hill ]
Reply-to: "research!alice!xchar care of" <CSVAX.upstill at Berkeley>
Subject: test-taking skills

    In HUMAN-NETS V4 #37, Greg Woods pointed out that high scores on
multiple-choice tests may (as in his case) reflect highly developed
test-taking skills rather than great intelligence.  The test below
illustrates Greg's thesis that one can often make correct choices that
are "not based at all on...knowledge of the subject matter."  I got
this test from Joseph Kruskal (Bell Labs), who got it from Clyde
Kruskal (NYU Courant Institute), who got it from Jerome Berkowitz
(Courant Institute).  Unfortunately, Prof. Berkowitz is currently out
of town, so I cannot trace its origin any farther back.

    I will supply the generally accepted answers, and perhaps some
explanations, later.


           --Charlie Harris

* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *

The following is a hypothetical examination on which you could get
every item correct by knowing some of the pitfalls of test
construction.  See how well you can do!  (Circle the letter preceding
the correct response.)

1.  The purpose of the cluss in furmpaling is to remove
     a.cluss-prags            c. cloughs
     b. tremalis              d. plumots
2.  Trassig is true when
     a. lusps trasses the vom
     b. the viskal flans, if the viskal is donwil or zortil
     c. the begul
     d. dissles lisk easily
3.  The sigia frequently overfesks the trelsum because
     a. all sigia are mellious
     b. sigias are always vortil
     c. the reelsum is usually tarious
     d. no trelsa are feskable
4.  The fribbled breg will minter best with an
     a. derst                 c. sortar
     b. morst                 d. ignu
5.  Among the reasons for tristal doss are
     a. The sabs foped and the foths tinzed
     b. the dredges roted with the orots
     c. few racobs were accapted in sluth
     d. most of the polats were thonced
6.  Which of the following is/are always present
    when trossels are being gruven?
     a. rint and vost         c. shum and vost
     b. vost                  d. vost and plone
7.  The mintering function of the ignu is most
    effectively carried out in connection with
     a. razma tol             c. the fribbled breg
     b. the grosing stantol   d. a frally slush
8.   a.                       c.
     b.                       d.




Date: 15 Sep 1981 15:14:39-PDT
From: ihuxo!hobs at Berkeley (John Hobson)
Reply-to: "ihuxo!hobs in care of" <CSVAX.upstill at Berkeley>
Subject: test-taking skills

Charlie--
The hypothetical exam on test-taking skills that you submitted to
HUMAN-NETS Digest V4 #46 has been an object of much interest here at
Indian Hill.  A number of us have taken the test and we would like to
see just how well we did.  The answers and reasons for those answers
are as follows:

1.  The purpose of the cluss in furmpaling is to remove
     a. cluss-prags           c. cloughs
     b. tremalis              d. plumots
1--a. The cluss is mentioned in the question and in the answer.

2.  Trassig is true when
     a. lusps trasses the vom
     b. the viskal flans, if the viskal is donwil or zortil
     c. the begul
     d. dissles lisk easily
2--a.  The word trassig in the question and the verb trasses in the
answer.

3.  The sigia frequently overfesks the trelsum because
     a. all sigia are mellious
     b. sigias are always vortil
     c. the reelsum is usually tarious
     d. no trelsa are feskable
3--c. The key word here is "usually", along with "frequently" in the
question.  Anyway, it is often best to give a non-absolute answer in
case there is an exception.

4.  The fribbled breg will minter best with an
     a. derst                 c. sortar
     b. morst                 d. ignu
4--d.  The giveaway here is the article "an" since "ignu" is the only
answer staring with a vowel.

5.  Among the reasons for tristal doss are
     a. The sabs foped and the foths tinzed
     b. the dredges roted with the orots
     c. few racobs were accapted in sluth
     d. most of the polats were thonced
5--a.  This is a bit more subtle, but we think that since the question
calls for "reasons" in the plural and (a) is the only answer with more
than one reason, that the answer is (a).

6.  Which of the following is/are always present when trossels are
    being gruven?
     a. rint and vost         c. shum and vost
     b. vost                  d. vost and plone
6--b.  Vost is mentioned in all possible answers, so vost must always
be present.

7.  The mintering function of the ignu is most effectively carried out
    in connection with
     a. razma tol             c. the fribbled breg
     b. the grosing stantol   d. a frally slush
7--c.  Since in question 4 (above), the fribbled breg was mintering
with an ignu, the thing mintering with the ignu is, of course, the
fibbled breg.

8.   a.                       c.
     b.                       d.
We haven't the foggiest.  Perhaps "all of the above".

I once took a multiple-guess test in English History where the last
question was:

The only British Prime Minister ever assassinated was:

a. Clement Atlee                e. None of the above
b. Spencer Perceval             f. One or more of the above
c. The Duke of Wellington       g. Don't know
d. All of the above             h. Don't care

b, f, g and h were accepted as correct answers.

                                John Hobson
                                ihuxo!hobs
                                Bell Labs -- Indian Hill



Date: 18 Sep 1981 12:13 PDT
From: Kolling at PARC-MAXC
Subject: test-taking skills

About that test.....

I think the answer to 2 is b, not a.  Either a or b is possible (not c
because it isn't grammatically correct, and not d because it's fuzzy
due to "easily".  Looking at the answers as follows:  1. a 2. a or b
3. c 4. d 5. a 6. b 7. c 8. ?  Note the pattern a,b,c,d, so I think 2
is b and 8 is d.

Karen (Now you know how I got through school.)



Date: 29 September 1981 0858-EDT (Tuesday)
From: Mary.Shaw at CMU-10A
Subject: Test-taking skills

I agree with Karen on the answers: a, b, c, d, a, b, c, d.  John's
reasons are correct except for #s 2 and 8.  Karen is right about 8
(it's the pattern).  The reason #2 is b rather than a is that option b
is markedly dissimilar from all the others.  (One of the rules of
test-writing is to avoid making the right answer stand out because
it's much longer or shorter than the others, especially if it's longer
because of a qualifying clause as in b here.)

Mary