u
r: n; 2.. 9."

”uh. Alh” 3 r. H”!
:0‘ V , .ouc
On". ”a. “60»
I... “*3 to. .— . .
.oo.‘ ‘ 1 ...

z an 3.. . ; .. .. p. ..
.2... .v... “J U3" 1 . u)...
r. .. ... : a... u .35 .
.VI‘ .0 V as

u.. _ . .
.. ..: 3..
u or. "in. ““0“ .3... u .o. . u (.36
”n... ‘3'.” ‘u. no... . :5”. ‘ . ..

”'74 N". In“ Jmfit ‘OQI. to ... ~.

' to. O ‘. .Q .“hr. .9. ’Q‘ to\ o
a. n. .M g: .3
p. Q nu.” a J 5...: 3.... u.s .r . V.
s ..r .c on a: .a m. u .a .q . an.)
I . «o I. - . a ..
qna “I.& c ’0 PW 4? 4m r ‘ OH- . .b
a, M 0 wk?‘ .u. &~06 V.v & .o. «.1 ....n u»
9... .. ”mu J . .Z. x u 3
«I. Ilnf “no .w... 5.. 3 t. . 6' r2
. I . .
3k. _..;V 5 a... mu ,3. ., ..
f» luv” 0; "VD-o 1.1. S.» u . \ fw—
I .0 .or . a D ‘ .1,- . i.
I o Iufi.‘ Wu (k
1- p” It. . 5... “Nam .05: .4 1....“ 5. .
a a. ... in, .z.,. . » 3...;
l f t.‘ t t “V U .f\ .- . (J
,PD.“ Iv .I ’9‘... . c o .s
V 'a van. Fflu . w I“ (0».
a. o. O ‘ A .no I .' ‘9
a; . . 3 . .. (k
“or' I». 05.”. '10- .I a A ‘o ’1‘.

a
J
5
-I .
o
l.
‘3 '.
$3.:
0
2/.
. ..\.
"1.:
.§ I
1. \‘Z
,.
-
k

. n, s

Pr . on. . in .. .

' K‘... “an.- ~.\~ - H.

C. b "O Q

“or . Wl.“ .90.... ”NJ”.
Om ... O n..... a
"3’ w“. m a 51'
r...“ Vroln t!‘ I”,

_e__:__;_____: a

 

.v‘dfid

A STUDY OF THE LQ. CCXx‘SISTENCY, INTERNAL CONSISTENCY,
AND CORRELATIONS WITH CERTAIN AOADEEOE GRADES OF THE
CALIFORNIA TEST OF MENTAL.MATURITT - SHORT FORMS

By
Stephen Orne grebs

A THESIS

Submitted to the College of Eflueation
Michigan State University of Agriculture and
Applied Science in partial fulfillment of
the requirements for the degree of

MASTER OF ARTS

Department of Foundations of Education

1957

lilillIlll‘Il'l

ACKNOWLEDGEMENTS

The investigator wishes to express sincere appreciation
for the patient and persistent guidance of Dr. Victor H. Noll,
a sine qua non of any value or scholarliness to be found in
this report. Much appreciation is also due the administrative
personnel of the Lansing, Michigan, public schools whose
cooperation was essential in the gathering of data.for this
study. The scoring office of Michigan State University is
due an accolade.for its assistance in a very difficult and

lengthy item count operation.

 

lllllll" .lll‘ll‘
[[l‘ul [II

TABIE OF CONTENTS

CHAPTER PAGE
I. THEINVES‘I‘IGATIONANDmHmaFSTLHJY....... 1
II.RINIEVCFLITMTURE................ S
III. CONSISTENCYQFTHEI.Q. 13
Iv. INTI'RMLCONSISTmCIOTTNOCTmIFORm....... 23
v. SUWARYOFTEIBINVIBTIGATION............ 30

BIBLICERAPHI........................33

TABLE

II.

III .

LIST OF TABLES

PAGE

Sample I .Q . Means and Population I .Q . Medians from

Administration of the CTMM Intermediate and

Advanced - S-Forms, 19142 Edition, to the Same

Pupils with a Three Year Interval . . . . . . . . . 15
Correlations of the I.Q.'s from Administration of

the CTMM Intermediate and Advanced - S-FormS,

l9h2 Edition, to the Same Pupils with a Three

YearInterval................... 16
Standard Deviations of I.Q.'s from CTW Inter-

mediate and Advanced - S-Forms, 19342 Mition,

Administered to the Same Pupils with a Three

Iearlnterval.............fi...... 18
Correlations of I.Q.'s from Administration in 1951

of the CTMM Intermediate - S-Form, 19h2 Edition,

With Grade Averages Obtained in the Next Three

Years.........................20
Correlations of I.Q.'s from Administration in 1951;

of the CTMM Advanced - S-Form, 19142 Edition, with

Grade Averages Obtained in the Previous Three

Years 0 O O O O O O O O O O O O O O O O O O O O O O 21

TABLE
VI .

VII.

VIII .

Frequency of Correlations of Item Scores to Total
Factor I.Q. Scores from Administration of the
CTMM Elementary and Intermediate - S-Forms,
1950 Edition, to Five Hundred Pupils for Each
Fom,inLansing,hfi.chigan.............
Item Count of Answers Chosen by the Upper and Lower
27% of Total Factor I.Q. Scores from Administration
of the CTMM Elementary S-Form, 1950 Edition, to
Five Hundred Pupils in Lansing, Michigan . . . . . .
Item Count of Answers Chosen by the Upper and Lover
27% of Total Factor I.Q. Scores from Administration
of the CTMM Intermediate S—Form, 1950 Edition, to
Five Hundred Pupils in Lansing, Michigan . . . . . .

PAGE

25

36

h?

CHAPTEI I
THE INVfiTIGATION AND METHGD OF STUDY

The purpose of this investigation was to make a statistical
analysis of the scores obtained from the administration of the
California Test of Mental Maturity - Short Form1 (referred to as the
CTMM for the remainder of this paper in the interest of brevity) to
students in the public schools of Lansing, Michigan. This study ex-
amined the consistency of the I.Q.'s as measured by two forms of the
test when administered to the same students with a three year interval
between administrations of the forms. The relationship of these I.Q.'s
to certain academic grade averages earned during the three year inter-
val was also examined. It was a further purpose of this study to
measure and evaluate the internal consistency of "0 forms of the test.
- The CTMM was used throughout the Lansing public school system
and was regularly administered to all students as follows: Elementary
form - fourth and sixth grades 5 Intermediate form - ninth grade; and
Advanced form - twelfth grade. The tests were administered by teachers
and scored by machine. The I.Q.'s were calculated by trained personnel
and sent to the teachers to be recorded in the pupil's permanent record
folder.

 

Imitabeth T. Sullivan, mm: W. Clark, and Ernest w. Tiegs,
California Short-Form Test gt; Mental Maturity (Los Angeles: California
Test Bureau, ISM—53 I935) .

 

l‘lllll-[Illl

The basic approach of this study is the normative survey
method involving processes of measurement and evaluation . Chapter II
is concerned with the background of studies and literature preceding
this investigation. Chapter III covers the first part of the study
itself. This part of the study is based on the results of administer-
ing the 19h2 edition of the CTMM forms. The intermediate form of the
CTMM was administered to 169 pupils in the ninth grade in 1951 and the
advanced form to these same students in twelfth grade in 1951:. The
I .Q. '8 thus obtained were compared by product moment correlations .
Also, their correlations with grade averages were obtained in four
academic subject areas, covering courses taken during the three-year
period between the administrations of the two tests. The latter
correlations were biserial coefficients. Chapter IV covers the second
part of the study which involved measurement of the internal consistency
of the elementary and the intermediate forms, 1950. Successes on each
item for upper and lower 27% were used as a basis for determining co-
efficients of discrimination by means of a Flanagan chart. This index
gives a measure of their demonstrated ability to discriminate between
the highest and lowest scorers on the total test . Chapter V consists
of a emery and conclusion of this report.

The data for Part 1 of the study (Chapter III: Consistency of
the I. Q.) were taken from permanent record folders of the twelfth grade
pupils in the three public high schools in Lansing, mchigan. All of
the CTMM forms referred to in this study were administered on the basis

of one hundred percent coverage of the grades involved. The data for

[[ll‘lllll'l‘l [II
I [I‘ll Il'lllllllllll III. all I

3.

Part 2 (Chapter IV: Internal Consistency of Two CTMM Forms) were
taken from the individual answer sheets for the elementary and the
intermediate forms .

The CTMM purports to be a test of mental ability consisting of
seven subtests each of which is purported to be a test of an aspect
of an element of mental ability. These subtests combined to make four
groups of subtests purport to measure four major factors involved in
intelligence.

The first group, called "Spatial Relationships ," contains two
subtests - ”Sensing Right and Left" and "Manipulation of Areas". The
first consists of twenty items each with a picture of an object from
a symetrical pair. The subject is to discriminate between left and
right. The second consists of fifteen items each with pictures of
geometrical forms in various positions. Given a figure the subject
is to select the same figure in another position from among three views
of similar figures.

The second group, called "Logical Reasoning," contains two sub-
tests - "Similarities" and "Inference". The first consists of fifteen
items each with seven picture situations . The subject is to determine
how the first three are alike then select one more with this character-
istic from among the other four. The second consists of fifteen items
each with a major and a minor premise and three conclusions. The
subject is to select the correct conclusion.

The third group, called I‘Numerical Reasoning" contains two sub-

test - l'Number Series" and "Numerical Quantity". The first consists of

h.

fifteen items each with an increasing or decreasing number series.
The subject is to select the number that does not belong in the series
from five offered for each of the first ten items and is to select
three numbers missing from the series from five offered groups for
each of five items. The second contains fifteen items each with an
arithmetical story problm. The subject is to solve the problem and
select the correct answer from four offered.

The fourth group, called "Verbal Concepts ," contains one sub-
test - "Total Verbal Concepts". This subtest consists of fifty items
each with a given word and four offered synonyms . The subject is to
select the correct synonym.

These seven subtests are also regrouped to obtain a ”Language
Factors" score which is the sum of the scores of three subtests
(Inference, Numerical Quantity, and Total Verbal Concepts) and to
obtain a "Non-language Factors" score which is the sum of the rmining
four subtests (Sensing Right and Left, Manipulation of Areas, Similar-
ities, and Number Series). The sum of all seven subtest scores is
called the "Total Mental Factors" score. An I.Q. score is calculated

from each of these factor group scores.

 

||l l‘i

 

CHAPI‘EI II
REVIEW CF LITERATURE

The CTMM short forms were devised as one period tests from a
longer parent test . The 1950 editions were prepared on five levels:
preprimary, primary, elementary, intermediate, and advanced . Subtests
purport to offer specific measures in spatial relationships, logical
reasoning, numerical reasoning, and verbal concepts. The seven sub-
tests involved in these measures are rearranged to offer two composite
scores. One group purports to measure language factors of intelligence
and the other purports to measure non-language factors of intelligence.
Finally a total score is obtained by adding the language factor and
non-language factor scores . Three I.Q.'s are computed; one each from
the language factor, non-language factor, and total factor scores and
are so named . These various scores are claimed useful for prognosis
and diagnosis in education, guidance, personnel administration,
clinical diagnosis, and research.1

The test was published in 1937 and subsequent revised editions
in 19142, 19147, and 1950. Cursory comparison of the 191;? and 1950 forms
indicated changes in norms, femat, color, timing, examples, stems,
and foils. The vocabulary subtest, for example, had twelve changes in

 

lJames 0. Coleman, m of Investi ations Number Three:
California Test 2f; Mental t z—(Los rig-oi?” Ema Il‘est
fiureeu, I953), p. .

6.

foils. Comparison of the l9h2 am 1950 forms revealed changes in the
order of the subtests, and the edition of one entire subtest, namely
"Number Series".

Though information from one study using a particular edition of
the forms may be quite similar to information from a comparable study
using another edition of the forms, there was little evidence found ttat
such similarities of information could be consistently expected. This
is important in the evaluation of material concerning the CTMM because,
although more than one hundred studies have been published concerning
the CTMM, only a small percentage used the 1950 (to date the latest)
edition of the forms.

Correlations in the .90 's for language factor and total factor
I.Q.'s and in the .70's and .80's for non-language factor I.Q.'s between
the long and short foms were reported in an early study using the first
edition.2

Reliability coefficients in the .90 's for the CTMM elementary,
intermediate, and advanced short form LQJe were published in the
prospectus for the 19147 edition and in more complete form by Coleman .3
These coefficients were based on from 250 to 1,000 cases using the
split-halves method corrected by the Spearman-Brown formula .h Sections
“Weiting Daniel Belden, Factor An is of the California

Tests of Mental Maturit (an unpublisha at y y he-I'oTEngEIes
may-5cm 1938771

3mm., p. 10, Table 13.

hElisabeth T. Sullivan, willie W. Clark, and Ernest W. Tiegs,
California Test Bureau Manual California Short-Form Test of Mental

Maturit Advanc ed": 'G—rades 9 to Adult - 19' 50' '§-_ E (B? EgEI'ee'T
EMT-this Test Bureau, I950)Tp. E.

 

 

 

llI!|lllllll llll I I'll II I

7.

of this same table appear on page four of each of the manuals as re-
liability coefficients for these three forms of the newer, 1950,
edition of the test. On this same page a correlation of .521; was
given for the CTMM - Advanced S-Form language factor scores with non-
language factor scores for 600 cases in grades nine to twelve. Several
early studies involving the first edition were referred to as having
found evidence that these scores can distinguish between reading de-
ficiency and mental deficiencyfi’ 6’ 7

Sheldon and Manolakes with a 1:22 pupil sample found a correla-

8To

tion of .702 of the CTMM with the Stanford—Binet, Revised Form L.
date other studies correlating these two tests have used older editions
of the CTMM or the parent long form and have used considerably smaller
samples.

Plumb, using a sample containing 370 males in Nebraska State
Reformatory for Men, correlated I.Q_.'s from a CTMM - 191;? form with
I.Q.'s from the Wechsler Bellevue Intelligence Scale. His correlations

ranged up to .725 (Wechsler Bellevue total scale with CTMM total

 

5Arthur 3. Trader, "A Study of the California Test of Mental
Maturity, Advanced Battery,“ Journal of Educational Research, 32:
329-335, January, 1939.

6Ruth Strang, 'Relationships Between Certain Aspects of Intelli-
gence and Certain Aspects of Reading," Educational £99. Psychological
Measurement, 33355-359 , Winter, 19143 .

7Esther Grace Nolan , "Reading Difficulty vs. How Mentality,“
California Journal 9_f_ Secondary Education, 17:31i-35, January, 19142 .

8William D . Sheldon and George Manolakes, "A Comparison of the
Stanford-Binet, Revised Form L, and the California Test of Mental Matur-
ity (S-Form)," 2.13.2 Journal of Educational Ps chol , IIN (December,
19510, 500. '—

 

 

 

I. ll'l'.'l'lll‘ll I11\ I11! 11 11. 11" III 111 l I

8.

factors) and to .726 (Wechsler Bellevue verbal scale with CTMM total
factors)? This study also states that Dr. Willis W. Clark of the
California Test Bureau reported to Plumb that a seventy-two case study
in a private school population in Devereant, Pennsylvania,.found cor-
relations of .71 (Wechsler Bellevue verbal scale with CTMM language
factors), .69 (Wechsler Bellevue performance scale with.CTMM non-
language factors), and .81: (Nechsler Bellevue total scale with CTMM
total factors) .10

Numerous studies have reported that the CTMM correlates as well
with other group tests of intelligence as they do with each other.
This information is of no particular importance to this study so the
interested reader is referred for further information to the manuals
for the various forms, Coleman;1 and Buros.12

Coleman reported intercorrelations of test sections on the CTMM
advanced short form ranging between .178 and .h87 based on 600 high

school students .13 These intercorrelatione lend some support to the

 

9Galen R. Plumb, The Evaluation of the Mental Characteristics
of _a_ Reformato Population as_ReveaIed E 255 mer-Beflevue
Intel ence ale and the C-aIifomia Tests of Mental Maturit ,
actor's thesis, Umersity 5f NebrasWoIr-l', Nebraska, 1951),

Po 149’ Table 33e
101bid., Appendix A, Table 8.

 

 

uColeman, 22. gig” p. 7, Table 7, and pp. 27-30.
120scar K. Buros (Editor), The Nineteen Fifty-Three Mental

Measurements Yearbook (Highland Park: New Jersey: The Mental Measure-
ments Ear‘nooE, I953).

Ecclm, 22s 9359’s, P. 8, Table 8.

1111'"!

9.

claim that the subtests measure different factors, but identification
of the factors involved has not been settled.
In an extensive statistical analysis in this area, Nelson
based a study of scores from administration of an early CTMM primary
form to 183 third grade pupils and of an early CTNM elementary form
to 193 fifth grade pupils correlated with scores on the Progressive
Achievement Tests (Reading and Arithmetic Batteries). He suggested
that the memory and spatial relationships subtests of the CTMM might
be eliminated from the test since they do not contribute greatly either
to diagnosis or to prognosis of achievement .111 For all types of
achievement involved, higher critical ratios were found for the CTMM
language mental age than for the non-language mental age .:LS
Schellenberg found the language components of the CTMM most
useful in predicting scoreson the Progressive Achievement Tests .16
Sister Mary Rosilda found a correlation of .112 between the CTMM
advanced form and the Cooperative Elementary Algebra Test Forms R, S,

and Q, Revised Series, when administered to 635 high school freshmen.17

 

114A. Gordon Nelson, The Value of Certain Lests of Differentiated
Abilities for Di osis and-P170“ osis of Eflev ement -—(Unpublished
Doctor's Thaiis, fiew Yorn vers y, N'Ew York, I 9113) p. 153,and
p. 125, Table 36.

151bid., p. 102-106, Table 29 .

“Ernest D. Schellenberg, A St of the Relationshi s of the
Pro ressive Achievement Tests and the al'fiornia Tests of Mental—— Maturit
Subtests Master's thesis, UniversT—y of §uthernmomiaz sInTgeIes,
I958), p. 120-121.

173ieter u. Reeilde, "Is an 1.0. an Index to Algebra Ability?"
Journal of Educational Re__s____earch, 1111: 391-393, January, 1951.

10.

Dejmek reported a correlation of .327 of the CTMM non-language scores
with scores on the Minnesota Paper Form Board - Revised Series AA.18
Heston described the success of scores of twenty-three non-verbal test
variables in predicting academic success. He reported that a correla-
tion of .5160 of the Ohio State Psychological Examination (a verbal
test) scores with ratios of grade points to credit hours was increased
to a maximum "shrunken" multiple correlation (using the Wherry shrinkage
formula) of .6583 by the addition of six non-verbal variables, namely

R and L Hands (Time), Minnesota Rate of Manipulation-Placing, Knox
Cubes, Feature Profile, R. and L Hands (Rights), and Dot Estimation
(Rights).19 Though this reference to Heston does not involve CTMM
non-language scores it does indicate a way in which the non-language
elements of the CTMM may be increasing the correlation of the CTMM total
factors scores with academic success. Lanier reported that students
dropping out of high school scored significantly higher on the CTMH
non-language score than on the language score. Non drop-outs scored
significantly higher on CTMM language scores than on non-language

scores .20 The evidence from these studies appears to indicate that

 

18Frank W. Dejmek, "A Study of Relationships Among Scores on
the Minnesota Paper Form Board - Revised Series AA, the Non-Language
Section of the California Tests of Mental Maturity, and Test Two of
the Non-language California Test of Mental Maturity," Journal g_f
Educational R_e_____search, 112: 307-311, December, 19148.

 

19Joseph C. Heston, "The Use of Non-Verbal Tests in the
Prediction of Academic Success ," T_h_e J_o____urna1 of Educational P cholo
33:609-612, 19112.

20.1. Armand Lanier, "A Guidance Faculty Study of Student
Withdrawals," Journal 9_f_ Educational Research, 143:205-212, November, 19149.

11.

the use of CTMM scores for prediction of achievement test scores is of
value primarily in determining group levels and relationships .

Kenesevich found the correlation between the CTMM intermediate
and advanced forms and grade point averages was .147 for twenty-eight
sophomores and twenty seniors in a high school. Upward shifts in the
I.Q.'s from the sophomore to the senior year for the same students
were noted but the difference of the means was not statistically
significant.21

Bolton reported the coefficients of correlation of CTMM I.Q.'s
with marks (over two semesters) in various courses, including those
covered in this study, and found the standard errors of estimate all
too high for prediction of grades .22

In reference to an elementary form of the CTW, Ordahl said,
"Item validities were established by the method of biserial correlation."
She also stated that items were arranged in each section by difficulty,

the percent passing each item.23

 

21Stephen J. Kenezevich, "The Constancy of the I.Q. of the
Secondary Pupil,“ Journal of Educational Research, 39 :506-516,
March, 1916. I

22Floyd B. Bolton, "Value of Several Intelligence Tests for He-
dicting School Achievement," Journal of Educational Research, 141:
133-138, October, 1914?.

23Vida Ellison Ordahl, “An Intercomparison of the Otis Self
Administering Test of Mental Ability, Intermediate Form A; the Stanford-
Binet, Form L5 and the California Test of Mental Maturity, Elementary
S-Form' (Unpublished Master's thesis, University of Southern California:
Los Angeles, 19116), p. 9-10, citing "a personal interview with the co-
author, Dr. Willis '11. Clark, Director of Research and Guidance, Los
Angeles County Schools, (November, 19143)."

12.

Many of the early studies concerning the CTMM involved the
parent long form. The studies referred to in this chapter were
selected because most of them dealt with the short forms. They were
the most extensive studies found or were the only ones found concerning
a particular area of the investigation. It may have been noted that
usually the only clue to the particular edition of the forms involved
has been the date of the study itself.

In all cases the letters CTMM have been used to refer to the
short forms unless specifically stated otherwise.

In summary, the CTMM forms in addition to the total I.Q. give
a language factor I.Q. and a non-language factor I.Q. which have been
shown to measure different factors. Scores of subtests and scores of
groups of subtests have been purported to measure certain factors of
intelligence. Studies have been inconclusive concerning support of
the factor identification. The reliability coefficients from CTMM
I.Q. scores have'been as high as for other group tests of intelligence.
More recent and comprehensive studies have shown correlations of the
CTMM total factor and language factor I.Q.'s'with.Stanford-Binet and
'Iechsler Bellevue individual intelligence test I.Q.'s have ranged in
the low seventies. The CTMM total factor and language factor I.Q.'s
have been shown to correlate well'with.measures, such as school grades
or test scores, that depend upon reading ability as an important variable.
The CTMM non-language factor I.Q.'s have not been shown to have con-
sistently high correlations with anything but have not been thoroughly

investigated .

CHAPTER III
CONSISTENCY CF THE I.Q.

The first part of the study involves I.Q. Obtained.from
administration of the CTMM Intermediate and Advanced S—Forms, l9h2
edition, to secondary pupils in the public schools df Lansing,
Michigan. It was standard procedure in the Lansing public schools
for CTMM forms to be administered to all pupils as they reached
certain appropriate grade levels. In the l95h-l955 school year the
advanced form was administered to 936 twelfth.grade pupils in the
city's three public high schools. These pupils constitute the
population for this part of the study.

A sample was drawn.from this population by examining the
records1 of every third pupil starting with the first in each school.
Of these only 169 had records sufficiently complete to be used in this
study. The completeness of the records depended upon the record
keeping procedures of each school and of various teachers. The pupils
'were assigned to teachers in alphabetical.order so no selective
factors were identified.

Language factor, non-language factor, and total.factor I.Q.'s
'were found in the records of these 169 pupils from the advanced.form.

These pupils had been administered the CTMM Intermediate S—Form, l9h2

 

18cc pages 1, 2 of this report.

1h.

edition, when they had been in the ninth grade in the 1951-1952 school
year three years earlier. The records of all 169 pupils contained the
total factor I.Q.'s from the intermediate form but only eighty of the
records contained the language factor and non-language factor I.Q.'s
from the intermediate form.

Means of the three I.Q.'s for both forms administered to the
sample cases and the three population medians for the advanced form
are given in Table I. The difference of sample means for non-language
factor I.Q.'s was not statistically significant but the differences of
sample means for the language factor and total factor I.Q.'s were
significant at the one percent level of confidence. The data in this
Table indicate a general increase in I.Q. from the intermediate to the
advanced fem.

The I.Q.'s obtained from the administration in 1951 of the CTMM
Intermediate S—Form, 1914.2 edition, were correlated by the product-
moment method with the I.Q.'s obtained from the administration in 1951;
of the CTW Advanced S-Form, l9h2 edition. These correlations are
given in Table II.

These correlations were considered high enough to be useful in
the evaluation of group levels and differences but low for prediction
and evaluation based on individual scores . The diagnostic and prog-

nostic powers claimed for the various subtest scores2 relative to the

 

zElizabeth T. Sullivan, Willis 1. Clark, and Ernest w. Tiegs,
California Test Bureau Manual California Short-Form Test of Mental
Haturit Advanced - Grease 9—22 Adult - 1935 _S-F'orm (rs; EgeIes:
We Test Bureau, I955), pp. 3—8.

 

15.

TABLE I

SAMPLE I.Q. MEANS AND POPULATION I.Q. MEDIANS
FROM ADMINISTRATION OF THE CTW INTERMEDIATE
AND ADVANCED S-FORLB, 19142 EDITION
TO THE SALE PUPILS AND WITH A
THREE YEAR INTERVAL

 

w-.——-———-— -.—— —_ ._—_..~ -._-~ -—__- ._~ _‘_ -—.-»-——-—-‘ ___. _.._--"____. *“_w——-—¢.—*

 

 

Pop . Sample Sample Significance
N° ~ °f Median Mean Mean Level
1°Q' Cases Adv. Form Int. Form Adv. Form Difference
“9510* (1951) (1951;) of Means
Total
Factor 169 101: 102 106 1%
Lang.
Factor 80 10b, 98 105 1%
Non-Lang.
Factor 80 103 103 106 Nil

 

J__
r

 

*From an unpublished report by Bernard W. Ansley, Director
Tests and Measurements, Lansing Public Schools, Lansing, Michigan, 1951;.

TABLE II

CORREIATIONS or THE I.Q.'S FROM ADMINISTRATION OF THE
CTMM INTERMEDIATE AND ADVANCED - s-Foms,
19142 EDITION, TO THE SAME PUPILS
WITH A THREE YEAR INTERVAL

 

 

 

_Factor I.Q. N r
Total Factor 169 .66
Language Factor 80 .57

Non-Language Factor 80 .58

 

17.

mental ability factors for which they have been named have not been
adequately supported or refuted by the evidence in the studies
mentioned.

Inconsistencies in the I.Q.'s from one form to another may have
involved one or more of several possible variables . The obvious
possibility is that the two forms do not measure the same factors in
the same amounts with equal accuracy. A strong possibility, especially
for this and older age levels, is that the basic assumption in the
construction of the CTMM that the average terminal age of mental growth
is sixteen years was causing measurable inaccuracies in those individual
cases which varied from this average. Another possibility is that in-
dividual rates of mental growth changed resulting in changes in the
true I.Q.'s. This study discloses evidence that these differences may
exist but makes no attempt to discover ww.

It is interesting to note that not only were there some signifi-
cant differences of the means but also there were statistically sig-
nificant differences of dispersion. The standard deviations of the
three factor group I.Q.'s, their differences, and the significance
levels of those differences are given in Table III.

The three CTMM I.Q.'s from each of the forms were correlated
with the grade averages in four academic subject areas covering course
taken during the three year period between administrations of the CTMM
forms. Courses in other subject areas were elected by too few pupils
in the sample for correlations to be meaningful. Coefficients were
obtained from biserial correlations using I.Q . 's as the primary con-

18.

TABLE III

STANDARD DEVIATIONS CF I.Q.'S FROM cnm INTERNEJIATE
AND ADVANCE) — S-Foms, l9h2 EDITION,
ADMINISTmm TO THE SAME PUPILS
AND WITH A THREE YEAR INTERVAL

 

 

Significance
Inter- Advanced Difference Level of
mediate Form Difference
I .Q . N Form 1951;
1951
Total
Factor 169 ll '2 ll '6 * J4 Nil
Lang .
New 80 1h .8 11.14 -3 .h 1%
””4““ ' 80 13 .9 19 .2 *5 .3 15‘

Factor

 

l9.

tinuous variable and grade averages as the divided continuous variable.
The upper portion of grade averages included A and B averages and the
lower portion included C, D, and E averages and those who failed to
elect courses in the subject area. The subjects included were English,
mathematics, social studies, and science. English and social studies
included courses required of all pupils.

The coefficients obtained when grade averages were correlated
with I.Q.'s from administration of the intermediate form in 1951 are
shown in Table IV. Coefficients from correlation of grade averages
with I.Q.'s from administration of the advanced form in 19511 are shown
in Table V.

For both CTMM forms correlations of grade averages with total
factor I.Q.'s and language factor I.Q.'s were high enough to indicate
that these I.Q.'s might be quite useful in evaluation of group levels
and comparisons. The non-language factor I.Q.'s did not appear to be
measuring the same factor or factors measured by these grade averages .
This further supported the fairly well established observation that
the so-called language factor and non-language factor groups of tests
measure different factors .3

The difference of the 1951 and 1951; correlations of grade
averages in science with total factor I.Q.'s was .214. No other cor-

relation differences were this large. Because chance may have caused

 

3Ibid., p. 1;.

TABLE IV

CORRELATIONS OF I.Q.'S FROM ADMINISTRATION IN 1951
OF THE CTMM INTERMEDIATE - S—FORAB, 19142 EDITION,
WITH GRADE AVERAGES OBTAINED

IN THE NEXT THREE TEARS

20.

 

 

 

Factor I . Q .
subject
Total Language Noananguage
English .60 .51 .10
mathematics .55 .hh -.05
Social Studies .63 460 .11
Science .h? .hh .08

 

21.

TABLE V

CORRELATIONS OF I.Q.'S FROM ADMINISTRATION IN 1951;
OF THE CTMM ADVANCE) S—FORM, 19142 EDITION,
WITH GRADE AVERAGE OBTAINED
IN THE PREVIOUS THREE YEARS

 

 

 

 

 

Factor I . Q.
Subject
Total Language Non-Language
filglish .58 .59 .06
Mathematics .52 .53 -.03
Social Studies .66 .72 .30
Science . .71 .65 .13

N 169 80 80

22.

such a change, to attempt to infer from this single sample would be
unjustified. It is interesting to note, however, that forty-six pupils
in the high science grade series averaged a 7.1 I.Q. score increase in
the three year interval as opposed to a h.6 I.Q. score increase for
forty-seven pupils who did not take science and a 2.7 I.Q. score in-
crease for seventy-three pupils who had low science grade averages.
Evidence found in this part of the study involved CTMM forms in
the EM edition and, therefore, can at best be considered roughly
indicative of what might be found with forms from the latest (1950)

“ltd-0n e

CHAPTm 1V

INTmNAL CONSISTENCI OF TWO FORTE OF THE CAIJI‘ORNIA
TEST (1“ MENTAL MATURITI, ELEMENTARY
AND INTERMEDIATE MORE

The forms involved in this part of the study were the CTMM
Elementary and Intermediate S—Forms, 1950 edition. The elementary
form was administered on a one hundred percent basis to the pupils in
the fourth and sixth grades of the public elementary schools of
Lansing, Michigan. In these grades the test was administered to
2,51l7 pupils. The answer sheets were machine scored in the system's
scoring office then were returned to the schools for recording. A-
request from the scoring office that these answer sheets be returned
to that office for the purposes of this study resulted in the return
of approximately 1500 answer sheets. By selecting every third sheet
from this group (arranged in class groups) a sample of five hundred
was obtained for this study.

Intermediate forms were administered to 1323 pupils in the
city 's public junior high schools. These were machine scored and
processed in the manner described for the elementary form. (If these
about one thousand were administered on a one hundred percent basis to
the ninth grade classes. These answer sheets were returned to the
scoring office by the respective schools as requested. By selecting
every other sheet from this group (arranged by class groups) a sample
of five hundred was obtained for this study.

2h.

Each sample of five hundred answer sheets was divided into
three groups according to total factor I.Q. score. Tvm groups contained
the upper and lower twenty-seven percents of these scores. The middle
group had no further purpose in the study. For the upper and lower
twenty-seven percent groups the answers selected for each item were
counted by machine in the scoring office of Michigan State University.
The counts obtained for each choice with the scores for the correct
answers underscored are recorded in Tables VII and VIII in the Appendix.
The percentage of each group selecting the correct answer and the
average of these percentages have also been recorded in these Tables.
In the last column of these Tables has been recorded a correlation co-
efficient interpolated from a Flanagan Table to measure the relation-
ship of the correct choices to the total factor I.Q.'s. A frequency
count of these correlations is given in Table VI.

If a given item samples one factor of intelligence it should not
be expected to correlate perfectly with a composite measure of all
factors of intelligence nor does greater item complexity guarantee a
better correlation with the composite measure. It was observed,
however, that many items on both forms had high correlations with the
grouped total factor I. Q. scores. 0n the other hand, assuming that
correlation coefficients below twenty may reflect indifference to the
total measure, there were twenty-seven items found on the elementary
form and forty-six items found on the intermediate form that possibly
are not contributing to the test except by chance effect. The large

number of omits for these items would seem to be a major factor in

25.

TABLE VI

FREQUENCI OF CORRELATIONS CF ITEM scans) TO TOTAL
FACTOR I.Q. sacrum FROM ADMINISTRATION (1“ THE
cum mm AND INTERMEDIATE S-FORLB,
1950 EDITION, T0 FIVE HUNDRH) PUPILS FOR
EACH FORM, IN LANSING, MICHIGAN

 

 

 

 

Frequency
Flanagan
r Elementary Intermediate
Form Form

60 - 69 7 O
50 - 59 2h 11
L0 - 1.0 28 21
30 - 39 3h 31
20 - 29 20 36
10 - 19 10 28

0 - 9 11 15
Negative 6 3

 

26.

these low correlations.

kamining the indexes of difficulty, the average percent
answering the itan correctly, it was observed that in this study the
pupils found the first three subtests very easy and the last four
subtests progressively more difficult. A check of the Flanagan
correlations of the items in the elementary form subtest seven (a
vocabulary test) with indexes of difficulty below thirty (very dif-
ficult) showed that this group of items included sixty-three percent
of the total number of items on the form with correlation coefficients
below twenty. Further exaNdnation of the items in this subtest dis-
closed that of the thirty-six foils in the last twelve items eighty-
three percent worked negatively in discriminating between the upper
and lower twenty-seven percent groups in the sample .1

This was at least in part a result of the much larger number of
pupils in the upper group attempting each of these items. This could
be expected to cause a higher index of discrimination for the correct
choice than might be expected with an equal number in each group at-
tempting the item.

For comparison an index of discrimination was calculated for
each of these twelve items using the method proposed by Davis. This
method accounts for the number of testees who did not reach the item

and also accounts for the number of testees who reached the item but

 

13ee Table VII, Appendix, PP.

27.

did not select an answer. Machine item count data includes these “0
types of omits in one total. This could cause the index of discrimina-
tion to be spuriously high. After conversion of these indexes to
product-moment correlation coefficients comparable to the Flanagan co-
efficient, using Davis' Table I (page 13), eleven of the twelve items
had coefficients lower than obtained from the Flanagan chart ard one
had a coefficient only slightly higher than the Flanagan coefficient.
In this case the proportion of successes in the lower group was negative
after correction for chance and was replaced by a low positive pro-
portion, from Davis'sTable II (page 35), because if the signs of the
proportions are different there is no straightforward way of using

Davis's chart .2

 

2Frederick E . Davis, Item—Anagsis Data - Their Computation,

Interpretation, g Use in Test Cons ruc ion “lamb-flags, flassactmsetts:
Harvard University, I936)? pp. 3 .

 

The formula used to determine the estimated proportions of
successes in each group is as follows:

11- W
P. K—I

N-NR

Where P - estimated proportion of the sample that lmows the answer
to an itan

N - number of testees in 27% portion of sample

B - number of testees in 27% portion of sample that answer
the item correctly

I - number of testees in 27% portion that answer incorrectly

NR - number of testees in 27% portion that do not reach the

item in the time Limit
K - number of choices in the item

A chart is used to determine the discrimination index from the
P's calculated for the highest and lowest 27% groups.

28.

These differences found between the coefficients determined by
Davis's method and the corresponding coefficients determined by
Flanagan's method appeared to indicate that the large number of omits
on these items did not cause the Flanagan coefficients to be spuriously
low. These differences also appeared to be of too low a magnitude to
be significant themselves.

Further investigation would be needed to examine the possibility
that items such as these are effective discriminators at very high
levels of mental capacity but are rendered ineffective in a test of
this type because of chance effects from the guessing of the much more
numerous pupils with less mental ability. This weakness is avoided in
the individually administered I.Q. tests, such as the Stanford-Binet,
because the subject is stopped when it is clear that he has reached ani
exceeded his capacity level. The alternative to scaled item diffi-
culties of limiting the range of item difficulties offers the prospect
of measuring higher levels of capacity by volume of average difficulty
answers rather than by power discriminating answers . This method
might suffice for purposes of general classification but would be less
useful for prognostic and diagnostic use at higher mental capacity
levels.

These and other problems that rise from the data of this study
must be held with some reservations because of limitations in the
study itself. For example, the study is limited to one city so its
data can not be applied to other cities with equal confidence. The

test differentiations between intelligence levels are less valid than

for the population as a whole because the samples are more home-
genous as a result of their being drawn from specific educational

levels.

29.

CHAPTER V
SUMITIARY OF THE INVESTIGATION

The first part of the study, covered in Chapter III, involved
comparisons of I.Q.'s from administration of CTMM forms to the same
pupils before and after three years of high school. The CTMM Inter-
mediate and Advanced S-Forms, 19h2 edition, were used respectively
for the two administrations. Contained in the sample were the LC).
scores of 169 pupils in the public schools of Lansing, Michigan.

The means of the language factor and total factor I.Q.'s in-
creased significantly. The mean of the non-language factor I.Q.'s
increased but not significantly.

Correlations of the language factor, non-language factor, and
total factor I.Q.'s on the intermediate form with the corresponding
I.Q.'s on the advanced form ranged from .57 to .66. The total factor
I.Q.'s from both forms were correlated with grade averages accumulated
in four academic subject areas during the three years of high school.
Language factor and total factor I .Q. correlations with grade averages
were moderately high. Non-language factor I .Q. correlations with grade
averages were not statistically significant. The greatest difference
in correlations of grade averages with the I.Q.'s of one form and with
those of the other form was .214, in science.

The difference between dispersions of the total factor I.Q.'s

from the two forms was not significant. The decrease in dispersion of

31.

language factor I.Q.'s and the increase in dispersion of non-language
factor I.Q.'s were significant at the one percent level of confidence.

The second part of this study involved the CTMM Elementary and
Intermediate S-Forms, 1950 edition. A sample of five hundred elementary
form machine scored answer sheets was obtained from administration of
the test to fourth and sixth grade pupils in the public schools of
Lansing, Michigan. Another sample of five hundred intermediate form
machine scored answer sheets was obtained by administration of the test
to ninth grade pupils in the same system. The upper and lower twenty-
seven percent of the answer sheets in each sample were chosen on the
basis of total factor I.Q. scores. The answers selected for each of
the items were machine counted and tabulated. The scores obtained
were compared.

These comparisons included the following: comparison of answer
selections of the upper and lower twenty-seven percent groups on each
item, calculation of difficulty indexes for all the items, and cal-
culation and sumarization of correlations of the item count scores
with the grouped total factor I.Q. scores.

Most of the items in both forms were shown to be effective
discriminators but twenty-seven items in the elementary form and forty-
six items in the intermediate form were shown to be poor discriminators.
Closer examination of the elementary form showed that a majority of the
poor discriminators were among the more difficult items in subtest
seven (vocabulary). It was noted that these items had been omitted by

large numbers of pupils .

32.

Examination of the indexes of difficulty for the items on both
forms indicated that both samples found the first three subtests in
their respective forms very easy and the last four subtests progress-
ively more difficult.

The CTMM forms examined were shown in this study to be relatively
consistent in their measurements of group levels and differences in
mental ability. There were definite irxiications that the language
factor I.Q. 's and non-language factor I.Q.'s involve measurement of
different factors. Moderately high correlations were found of total
factor and language factor I.Q.'s with certain academic grade averages.
Certain changes in the dispersion of I.Q.'s were found when the inter-
mediate and advanced forms were administered to the same subjects with
a three year interval. Though most individual items on the elementary
and intermediate form (1950 edition) were shown to be effective dis-
criminators between high and low total scores, a substantial proportion
of the items were shown to be poor discriminators. Further item

analysis appears constructively promising .

BIBLICERA PHI

Bolton, Floyd B. "Value of Several Intelligence Tests for Hedicting
School Achievement," Journal of Educational Re__§____earch, 1:1:133-138,
October, 19147.

 

Buros, Oscar K. (Editor). The Nineteen Fiij Three Mental Measurements
Yearbook. Highland Park, Neerrsey: The MentaI Measurements
ye ”5001‘; 1953 0

Coleman, James C. Sumngry_ of Investigations Number Three: California
Te___s__t of Mental Matur urity. Los Angeles: California Test Bureau, I956.

Davis, Frederick B. Item Analysis_____ Data - Their Com%tation, Inter-
etation, and Use In est onstruction. amb ge, sachusetts:
arv niversIty [19116.

 

 

 

Dejmek, Frank W. "A Study of Relationships Among Scores on the
Minnesota Paper Form Board - Revised Series AA, the Non-Language
Section of the California Tests of Mental Maturity, and Test Two
Of the Non-Language California Test of Mental Maturity,“ Journal
of Educational Research, 142: 307-311, December, 191:8.

Heston, Joseph C. "The Use of Non-Verbal Tests in the Prediction of
Academic Success ," Th__e_ Journ____a_l of Educational P cholo , 33:608-
611» 191:2.

Kenezevich, Stephen J. "The Constancy of the I Q.. of the Secondary
Pupil," J_o_______urnal of Educational Res___e______arch, 39:506-516, March, 191:6.

 

Lanier, J. Armand. “A Guidance Faculty Study of Student Withdrawals ,"
Journal of; Educational Research, 1:3:205-212, November, 19149.

 

Nelson, A. Gordon. "The Value of Certain Tests of Differentiated
Abilities for Diagnosis and Prognosis of Achievement ." Unpublished
Doctor's thesis, New York University, New York, 19143.

Nolan, Esther Grace. “Reading Difficulty vs Low Mentality," California
Journal of Secondary Education, 17:314-39, January, 191:2.

Ordahl, Vida E. "An Intercomparison of the Otis Self Administering
Test of Mental Ability, Intermediate Form A; the Stanford-Binet,
Form L; 8111 the California Test of Mental Maturity, Elementary
S-Form,‘ Unpublished Master's thesis, University of Southern
California: Los Angeles, 19146.

31:.

Plumb, Galen R. "The Evaluation of the Mental Characteristics of a
Reformatory Population as Revealed by the Wechsler Bellevue
Intelligence Scale and the California Tests of Mental Maturity.“
Doctor's thesis, University of Nebraska, Lincoln, Nebraska, 1951.

Rosilda, Sister M. "Is an I.Q. an Index to Algebra Ability?"
Journal 9_f_ Educational Research, 1414:391-393 , January, 1951.

Schellenberg, Ernest D. "A Study of the Relationships of the Progressive
Achievement Tests and the California Tests of Mental Maturity
Subtests." Master's thesis, University of Southern California,

Los Angeles, 191:8.

Sheldon, William D . and George Manolakes. "A Comparison of the Stanford-
Binet, Revised Form L, and the California Test of Mental Maturity
(S-Form)," The Journal of Educational Ps cholo , 1:5:1499-501T,
December, 1931;. '—

 

Strang, Ruth. “Relationships Between Certain Aspects of Intelligence
and Certain Aspects of Reading," Educational 2'29. Psychological
Measurement, 3:355-360, Winter, 1913.

Sullivan, Elizabeth T., Willis W. Clark, and finest VI. Tiegs.
California Test Bureau Manual California Short-Form Test of Mental
Maturit Advanced - Grade 9 _to Adult ~193morm. Es IngeIes:
Cam ommurm930 .

Traxler, Arthur E. "A Study of the California Test of Mental Maturity,
Advanced Battery," Journal gf_ Educational Research, 32:329-335,

January, 1939 .

 

APPENDIX

36.

TABLE VII

ITEM COUNT OF ANSFFRS CHOSEN BY THE UPFFR AND LONER 27% OF
TOTAL FACTOR 1.0. SCORES FROM ADMINISTRATION OF THE 0mm
ELEMENTARY S-FORM, 1950 EDITION, T0 FIVE HUNDRED
PUPILS IN LANSING, MICHIGAN

 

% Average Correlation
Item No. Of EaCh Response Correct % Item to
No. ~ 1R 2L 3 1.: 5 Omit Choice Correct U%-L% Total Test
(Flanagan)
Test 1 *
1 U 122 12 1 90.1; 78.2 21:.5 .3h
L 9.2 38 8 65.9
2 U 11 12A 0 91.9 80.1 23.7 .37
L 37 g 6 68.2
3 U 11 12A 0 91.9 76.3 31.2 .h3
L 1:6 8g 7 60.7
N U 125 10 0 92.6 75.6 3h.1 .147
L _7_9_ N6 10 58.5 .
5 U 12 122 1 92.1: 75.5 33.9 .h6
L 1:9 _7_9_ 7 58.5
. 6 U 129 6 0 95.6 76.7 37.8 .56
L. _7_8_ he 9 57.8
7 U 11 123 1 91.1 73.1; 35.5 .175
L 53 _7_5_ 7 55.6
8 U 12 123 0 91.1 77.1 28.1 .39
L 142 _8_5_ 8 63.0
9 U 10 125 0 92.6 71: .8 35.6 .h?
L 1:8 11 10 57.0
10 U 122 13 0 90.1: 77.1: 26.0 .37
I- 81 33 15 611.11
11 U 121 13 1 89.6 67.0 1:5 .2 .51
L 69 52 13 1:11.17

 

I

Correct choices are underscored.

37.

TABLE VII (66:11,.)

 

 

% Average Correlation
Item No. Of Each Response Correct % Item to
No. l 2 3 h S Omit Choice Correct U%~I$ Total Test
(Flanagan)

12 U 11 122 2 90.1: 71.9 37.1 .147
L to _7_2_ 13 53.3

13 U 122 12 1 90.1: 72.3 36.3 .116
I- 13 31: 28 5h.1

11: U 33 102 0 75.6 55.2 80.8 .112
L 19 .117. 39 3h.8

15 U 15 118 2 87.1: 68.9 37.0 .171:
L 30 68 37 50.1:

16 U 3h 97 h 71.9 53.8 36.3 .37
L 35 1.1.3 52 35.6

17 U 1.18 1h 3 87.1: 59.3 56.3 .59
L g 36 57 31.1

18 U 26 105 h 77.8 52.6 501 .50
L 37 37 61 27.1;

19 U 116 12 7 85.9 61.1 89 .6 .53
L 52 23 63 36.3

20 U 13 113 9 83.7 60.0 117.1: .50
L 21 1,2 65 36.3

Test 2

210 0 1 0 131: 0 99.3 93.8 11.1 .171:
L 2 h 6 _112 h 88.2

22 U 0 0 135 0 0 100.0 90.8 18.5 .58
L. 6 7 _1_1_g 10 2 81.5

23 U 10 1 121 3 o 89.6 77.1: 211.1; .31;
L 25 11 _8_§ 10 1 65.2

21; U 11 122 2 0 0 90.1: 78.6 23.7 .35
L 27 go 12 h 0 66.7

25 U 0 0 131: 0 1 99.3 90.8 17.1 .52
L 8 7 _111 9 0 82.2

 

TABLE VII (cont .)

38.

 

 

No. of Each Response % Average Correlation
Item Correct % Item to
No. l 2 3 14 S Omit Choice Correct Dig-1% Total Test
(Flanagan)

26 U 135 0 0 0 0 100.0 90.0 20.0 .60

L 1_0_8_ 9 6 7 5 80.0
27 U 2 2 3 128 0 9h.8 80.8 28.1 .175

L 7 26 10 go 2 66.7
28 U 2 132 1 0 0 97.8 83.7 28.2 .51;

L 19 21, 8 8 6 69.6
29 U 2h 8 13 90 0 66.7 50.7 31.9 .33

L 53 16 11 51 8 81:8
30 U 1 1: 1 129 0 95.6 78.9 33 .1: .52

L 11 15 15 £3.14. 10 62.2
31 U 37 73 21 5 -1 511.1 38.9 30.14 .32

L 39 _33 33 19 12 23 .7
32 U 16 1:2 68 10 -1 5011 311.9 31.1 .31:

L 32 M. £6 20 17 19.3
33 U 13 93 1: 21 1: 68 .9 55.2 27.1: .28

L 29 56 13 16 21 111.5
31: U 11 9 13 97 5 71.9 55.6 32.6 .314

L 17 15 26 _5_3_ 21: 39.3
35 U 9 7 56 55 8 111.5 27.1 28.9 .36

L 7 16 _1_7_ 65 30 12.6
Test 3
36 U 0 O 3 132 0 97.8 914.1 7.1: .28

- L 2 6 5 _1_2_2 0 90.1:

37 U l O 130 1 3 96.3 88.5 15.6 .36

L 10 8 $92 8 0 80.7
38 U 1 131 3 0 0 97.0 88.9 16.3 .140

L h _139 12 3 7 80.7
39 U 0 1 Bu 0 0 99 .3 91.9 117.9 .50

L h 8 _1_1_h h 5 811.1:

 

TABLE VII (cont .)

39.

 

 

% Average Correlation
Item No. Of EaCh Response Correct % Item to
No. 1 2 3 h 5 Quit Choice Correct U%-L% Total Test
(Flanagan)

to U 132 1 3 O -1 97.8 91.1 13.1; .38
L 11:, 8 9 5 -1 811.11

1:1 U 11 12 1: 115 0 85.2 72.6 25.2 .32
L 19 19 16 _81 O 60.0

1.2 U 2 0 121. 9 0 91.9 80.1: 23.0 .36
L 6 7 _9_3_ 20 9 68.9

113 U 0 0 2 133 o 98.5 86.7 23.7 .Sh
L h 6 111 3.91 10 711.8

M: U 2 121. 5 2 2 91.9 78.6 26.7 .39
L 12 _8_§ 8 17 10 65 .2

115 U o 3 O 131 1 97.0 811.1 25.9 .50
L 1. 15 t 2.2 16 71.1

116 U 0 0 138 0 1 99 .3 87.1 211.5 .60
L o 5 _1_0_1 11 18 711.8

117 U 0 1 0 131: o 99.3 811.1 30.1: .611
L 6 10 h 93 22 68.9

he U 3 131 o o 1 97.0 81.5 31.1 .51.
L 9 82 6 7 211 65.9

119 U 0 133 o O 2 ‘ 98.5 83.11 30.3 .59
L t 2g 6 s 28 68.2

50 U 3 2 101: 23 3 77.0 611.8 211.1: .27
L 11 5 11 17 21 52.6

Test 14 .
51 U 21: 23 86 2 63.7 511.5 18.5 .19
27 115 61 2 1:5 .2

52 U 7 119 9 o 88.2 70.8 3h.9 .112
L 26 _7_2_ 33 I: 53.3 70.8

53 U 2 8 126 -1 93.3 78.2 30.3 .141:
L 16 26 _85 8 63.0

 

ho.
TABLE VII (cont .)

 

 

% Average Correlation
Item No. of Each Response Correct % Item to
No. 1 2 3 h 5 Omit Choice Correct U%-L% Total Test
(Flanagan)
5h U 16 113 6 0 83.7 69.3 28.9 .31;
L 27 lg 28 6 5h .8
55 U 16 9 108 2 80.0 61.1 37.8 J40
L 29 35 51 lb 1.2 .2
56 U M 89 3 1 65 .9 ha .9 3h .0 .35
L 55 g; 22 15 31.9
57 U 8 9 120 .2 88.9 75.2 117.1. .52
L 21 38 _56 20 1.1.5
58 U 87 10 39 -1 611.14 h7.8 33.3 .31;
L 52 30 39 214 31.1
59 U 22 101 11; -2 7h .8 55.6 38.5 .ho
L 22 _lg 33 31 36 .3
60 U . 26 1:1 63 5 116.7 37.1 19 .3 .21
L 23 39 31 36 27.1:
61 U 99 21 11 h 73.3 1.7 .8 51.1 .51
1. 39 36 3o 39 22.2
62 U 1.2 73 10 10 511.1 1.1.5 25.2 .26
L 21 _3_9_ 2h 51 28.9
63 U 26 78 17 11; 57.8 116 .3 23 .0 .28
I- 12 14], 19 57 3h .8
6h U 12 7 102 1b 75 .6 118.6 511.1 .5h
L 19 23 22 6h 21.5
65 U 13 21. 76 22 17.8 117.9 5.9 .11
L 23 lé 26 70 11.9 _
Test 5
66 U 0 1 1 0 132 1 97.8 90.1; 111.8 .ho
L 3 1 7 12 _11_g 0 83.0
67 U 2 98 16 11 5 3 72 .6 56 .7 31.9 .33
L 1!. 55 22 13 25 6 80.7

 

hl.
TABLE'VII (cont.)

 

 

No of Each Response % Average Correlation
Item ' Correct % Item to
No. l 2 3 h 5 Omit Choice Correct Ui-H% Total Test
(Flanagan)
68 U 17 10 6h 30 7 7 117.1. 33.0 28.9 .32
L 35 22 35 31 13 9 18.5
69 U 39 15 13 1.1. 18 6 32 .6 25 .6 11..1 .18
L 25 29 21. _2_5 15 17 18 .5
70 U 5 9 1. 23 81 13 17.0 15 .6 2.9 .05
L 23 7 16 _1_9_ 1.8 22 11..1
71 U 36 12 30 7 2h 26 8.9 10.1. -3 .0 -.07
L 39 16 22 10 25 23 11.9
72 U 11 11. 20 11 50 29 11.8 13.7 2.2 .01.
L 21 16 11 8 36 37 12.6
73 U 12 20 12 10 51 30 37.8 30.1. 1h.8 .17
L 11 31 15 12 31 35 23 .o
71. U 15 21. 19 15 26 36 11.1 11.9 -1.5 -.03
L 11 21. 18 15 18 1.3 12.6
75 U 10 38 28 1. 17 38 28.2 21.3 111.9 .22
L 11. E 26 11. 18 1.5 13.3
Test 6
76 U 0 0 135 0 0 100.0 97.8 M. .35
L 3 0 _1_2_9_ 2 1 95.6
77 U 125 o 9 1 0 92 .6 71.9 1.1.5 .53
L 69 6 57 3 0 51.1
78 U 11 1. 1 119 0 88.2 59.3 57.8 .60
L 68 9 13 £13; 1. 30.11 ’
79 U 29 76 16 8 6 56.3 39 .7 33 .3 .35
L 25 _3_;._ 27 27 25 23 .0
80 U 97 22 6 5 5 71.9 119.7 1111.5 .111.
L 31 13 55 17 19 27.1.
81 U 1.6 18 11 1.7 13 311.8 25.6 18 .5 .21.
L 22 1.1. 20 22 27 16.3

TABLE VII (cont .)

 

 

Z Average Correlation
Item No. of Each Response Correct 2 Item to
No. l 2 3 l1 5 Quit Choice Correct U%-L% Total Test
(Flanagan)
82 U 51. 6 53 12 10 39 .3 27.8 23 .0 .29
L 1.5 17 22 10 1.1 16.3
83 U 8 77 19 15 16 57.0 1.1.1 31.8 .33
L 16 31. 21. 11 50 25 .2
814 U 18 58 16 21 22 1.3.0 30.0 26.0 .30
L 7 £2 25 25 55 17 .0
85 U 51. 15 21 12 33 1.0.0 26 .0 28.1 .36
L 16 11 23 23 62 11.9
86 U 38 23 26 7 1.1 19 .3 111.1 10.1. .20
L 17 21 _1_.2_ 12 73 8.9
87 U 16 31 29 9 50 11.9 8.2 7.5 .21
L g 23 17 9 80 11.1.
88 U 1 20 17 22 61 16.3 11.1 10.1. .21.
L 7 17 16 Q 87 5.9
89 U 16 15 21 15 68 15 .6 13 .1. 11.5 .09
L h 111 15, 9 93 11.1
90 U 19 12 11. 21 69 11..1 12.7 3.1 .06
L 15 1. 6 15 95 11.1
Test 7
91 U 0 130 1 1. 0 96 .3 76.3 1.0.0 .58
L 13 16 8 11. 21. 56 .3
92 U 126 6 3 0 0 93 .3 69.3 1.8 .1 .58
L 6_1_ 26 22 11. 12 1.5 .2
93 U 5 127 1 2 0 911.1 72 .3 113.7 .56
L 21 _6_8 17 18 11 50.1.
911 U 3 l1 1 126 1 93 .3 70.0 116 .6 .57
L 18 19 11. §_3_ 21 1.6.7
95 U 6 121 1 5 2 89 .6 58.2 62.9 .61.
L 30 36 25 22 22 26.7

 

TABLE VII (cont .)

113 .

 

 

% Average Correlation
Item No. Of EaCh Response Correct % Item to
No. l 2 3 h 5 Quit Choice Correct 014$ Total Test
(Flanagan)
96 U 6 2 122 1. 1 90.1. 59 .3 62 .2 .61.
L 30 19 2?. 21 27 28.2
97 U 1 86 1.2 6 0 63.7 1.3 .7 1.0 .0 .111
L 7 32 21. 38 31. 23 .7 .
98 U 1 120 10 1. 0 88 .9 60.8 56 .3 .59
L 12 1_.1._ 26 16 37 32 .6
99 U 1. 2 121. 3 2 91.9 60.1 63 .7 .66
L 1.1. 15 29. 6 32 28.2
100 0 1. 3 20 101. 1. 77 .0 51.1.8 111. .115
L 8 12 27 1.11 111. 32.6
101 U 19 21. 11. 711 14 511.8 37.8 311.1 .37
L 17 15 31 _2__8 1.11 20.7
102 U 17 70 8 32 8 51.9 31..1 35.6 .110
L 19 22 17 29 1.8 16.3
103 U 22 6 21. 76 7 56 .3 35.2 1.2 .2 1.7
L 33 11. 22 _12 1.7 111.1
101. U 12 55 31. 29 5 1.0.7 28.2 25.1 .31
L 17 31 32 22 1.3 15 .6
105 U 10 17 55 112 11 110.7 27.0 27.11 .35
L 29 20 ii! 18 50 13.3
106 U 17 21 9 78 10 57.8 1.2 .3 31.1 .32
L 8 1.2 17 36 32 26 .7
107 U 17 1.5 33 23 17 33 .3 21.1 21.1. .35
L 13 _1_2_ 22 36 52 8.9
108 U 1.5 16 36 25 13 11.9 10.8 2.3 .01.
L 28 1_3_ 18 18 58 9.6
109 U 11 57 19 29 19 1.2 .2 25 .6 33 .3 .1.3
L 22 1g 20 16 65

8.9

 

TABLE VII (cont.)

 

 

% Average Correlation
Item NO’ Of EaCh Response Correct % Item to
No. l 2 3 h 5 0.1.11. Choice Correct U%-L% Total Test
(Flanagan)
110 U 17 19 76 11 12 56 .3 37.1. 37.8 .1.1
L 9 21 _2_5 11. 66 18 .5
111 U 19 1.11 3h 15 23 32 .6 23 .11 18 .5 .25
L 10 _19 25 10 71 111.1
112 U 17 1.3 1.1 20 11. 30.1. 21.2 18 .5 .27
L 11 27 16 11. 67 11.9
113 U 28 60 13 17 17 Ml. 27.0 31..8 1.1.
L 12 1_3_ 25 17 68 9.6
111. U 9 61. 30 11 21 1.7.11 28.2 38.5 .119
L 23 12 17 9 71. 8.9
115 U 72 12 18 15 19 53 .3 311.5 37.7 J42
L _2_1 12 21 10 71 15 .6
116 U 11. 31. 1.3 18 26 31.9 23.1. 17 .1 .23
L 10 20 29 11 71. 111.8
117 U 1.9 38 11 7 30 28.2 19.7 17 .1 .26
L 18 _1_5 22 6 75 11.1
118 U 15 26 32 22 1.0 19 .3 111.5 9.7 .18
L 11 13 17 11. 80 9.6
119 U 15 36 30 13 1.1 26 .7 18.6 16.3 .26
L 5 _1_1._ 17 18 81 10.1.
120 U 5 18 30 1.3 39 31.9 21..1 15 .6 .20
L 8 7 11 _2_2_ 87 16.3
121 U 12 5 58 16 1.1. 1.3.0 25 .2 35 .6 .117
L 16 9 _l_9 15 85 7.1.
122 U 13 19 119 7 117 36 .3 23 .0 26.7 .37
L 13 16 13 8 85 9.6
123 U 31. 20 19 8 51. 25 .2 16.0 18 .5 .33
L _9_ 15 17 7 87 6.7

 

1.5 .
TABLE VII (cont.)

 

 

S Average Correlation
Item No. of Each Response Correct % Item to
No. l 2 3 l. 5 Quit Choice Correct U%-L% Total Test
(Flanagan)
1211 U 15 19 31 13 57 23 .0 16.1 111.1 .211
L 10 13 _1_2_ 12 88 8.9 ‘
125 U 12 22 17 21. 6o 17 .8 17.8 .0 .00
L 5 7 10 _2_1_. 89 17 .8
126 U 11 112 13 12 57 31.1 23.7 111.8 .19
L 8 23 7 6 92
127 U 11. 16 21. 11. 67 17.8 16.0 3.7 .06
, L 5 6 _12 7 98 11..1
128 U 17 8 13 25 72 9.6 7.0 5.2 .16
L 12 11 .6. 13 93 11.1.
129 U 15 111 13 21 72 15.6 11.9 7.1. .15
L 8 9 8 _11_ 99 8.2
130 U 26 10 13 12 71. 7.1. 7.8 -.8 -.02
L 19 _11_ 5 6 91. 8.2
131 U 211 13 12 7 79 8.9 6.7 11.5 .15
L 11. 6 _6_ 8 101 1..1.
132 U 7 6 29 9 81. 6.7 11.1 5.2 .211
L 3 10 16 _2_ 101. 1.5
133 U 11 12 10 12 90 7.1. 6.3 2.2 .07
L 7 10 1 8 103 5.2
131. U 9 13 21 7 85 5.2 5.2 .0 .00
L 3 10 10 _7_ 105 5.2
135 U 15 1. 18 8 90 5.9 5.2 1.5 .06
L 12 5 6 6 106 11.11
136 U 10 6 15 12 92 M. 1..8 -.8 -.03
L 7 1 9 8 101. 5.2
137 U 1. 11. 13 13 91 9.6 7.1. 14.1. .13
L 5 9 7 8 106 5.2

 

116 .
TABIE VII (cont .)

 

 

% Average Correlation
Item No. Of Each Response Correct % Item to
No. l 2 3 h ‘ 5 Quit Choice Correct U%-L% Total Test
(Flanagan)
138 U 11 8 12 11 93 5.9 5.9 .0 .00
L 9 _8_ 9 3 106 5 .9
139 U 18 9 5 10 93 3 .7 6.3 -5 .2 -.18
L 5 12 _12_ 2 101. 8.9
11.0 U 20 3 6 15 91 2 .2 3.0 -1.5 -.09
I. 10 g 5 8 107 3.7

TABLE VIII

117.

ITEM COUNT OF AW CHOSEN BY THE UPPER AND IONFR 27% OF
TOTAL FACTOR I.Q. SCORES FROM ADMINISTRATION OF THE CTMM

PUPILS IN LANSING, MICHIGAN

INTERMEDIATE S-FORM, 1950 EDITION, TO FIVE HUNDRED

 

 

 

 

Item

No. of Each Response

%

Correct

Average

%

Correlation
Item'to

 

 

*Correct choices are underscored.

No. l 2 3 h S (knit Choice Correct [Lt-Li Total Test
R L (Flanagan)
Test 1
1 U 8 126 1 93 .3 83.3 20.0 .35
2 U 122 12 1 90.1. 80.7 19.3 .30
L 26 38 1 71.1
3 U 127 6 2 914.1 88.5 11.1 .25
L _11_g_ 22 1 83.0
1. U 13 122 0 90.1. 78.5 23 .7 .31.
L 1.1. 29 1 66.7
5 U 126 9 0 93 .3 88.5 9.6 .21
L _113 21 1 83.7
6 U 133 0 2 98.5 91.1 111.8 .hS
L 11.3 22 0 83.7
7 U 25 108 2 80.0 71..8 10.1. .13
L 1.1 21. 0 69 .6
8 U 126 8 1 93.3 8h.8 17.0 .31
L 123 32 0 76 .3
9 U 6 128 1 91..8 81..8 20.0 .37
L 32 191 2 711.8
10 U 116 18 1 85 .9 77.7 16.3 .23
L 11. 1.0 1 69 .6
11 U 18 116 1 85.9 72.9 25.9 .33
L 50 _8; 1. 60.0
12 U 10 121. 1 91.9 79.3 25.2 .38
L 1.1 a; 1. 66.7

TABLE VIII (cont .)

1.18.

 

 

% Average Correlation
Item N0. of EaCh Response Correct % Item to
No. l 2 3 11 5 Quit Choice Correct U%-I.% Total Test
(Flanagan)
13 U 115 12 8 85 .2 711.8 20.8 .27
L 82 37 11 611.1.
11. U 31. 96 5 71.1 63.3 15.5 .17
L 51 15 9 55 .6
15 U 121 13 1 89 .6 80.3 18 .5 .28
L 2‘; 26 13 71.1
16 U 118 16 1 87.1. 81.1 12.6 .19
L 3.31 22 12 711.8
17 U 66 51. 15 1.0.0 36.3 7.1. .08
L 61. £1. 27 32.
18 U 1.1. 86 5 63 .7 58.5 10.1. .11
L 52 1g 11 53.3
19 U 32 91 12 67.1. 55.5 23 .7 .25
L 1.9 _5_9_ 27 1.3.7
20 U 30 99 6 73 .3 6o .0 26.6 .28
L 51 £2 21 1.6.7
Test 2
21 U 13 9 9h 17 2 69 .6 55.9 27.1. .28
L 21. 32 _5_7 19 3 1.2.2
22 U 128 1 6 0 0 91..8 88.1 13.3 .30
L 113 6 10 8 1 81.5
23 U 1 2 21 111 0 82.2 71.1. 21.5 .27
L 5 6 1.2 _83 0 60.7
21. U 0 12 120 3 0 88.9 83 .7 10.1. .18
L 5 15 3.96 9 0 78 .5
25 U 12 55 11 55 2 1.0.7 28.8 23 .7 .29
L 23 2_3_ 18 69 2 17.0
26 U 19 29 79 5 3 58 .5 1.3.3 30.3 .31
L 1.3 1.0 38 12 2 28.2

 

1.9 .
TABLE VIII (cont .)

 

 

76 Average Correlation
Item No. of Each Response Correct % Item to
No. 1 2 3 l. S (knit Choice Correct [Ti-1% Total Test
(Flanagan)
27 U 17 70 5 1.2 1 31.1 22.9 16.3 .22
L 26 77 10 22 2 11..8
28 U 12 19 31. 1.5 25 33 .3 25 .1 16.3 .21
L 23 13 1.8 g_3_ 28 17 .0
29 U 36 16 31 1.0 12 23 .0 11..8 16.3 .30
L 1.2 11 g 60 13 6.7
30 U 21 15 2 93 1. 68 .9 60.1. 17 .0 .19
L 1.5 8 7 19 5 51.9
31 U 70 1.7 5 8 5 311.8 28.1 13.3 .16
L 72 _2_9 9 16 9 21.5
32 U 6 28 27 61. 10 1.7.1. 32.9 28.9 .33
L 30 1.2 22 £5 16 18 .5
33 U 29 30 21. 23 29 22.2 20.7 2.9 .01.
L 36 3.2 21. 33 16 19 .3
31. U 1.0 20 28 16 31 20.7 11.1. 12.5 .21.
L 1.8 26 _1_1 17 33 8.2
35 U 19 9 37 1.8 32 35 .6 311.1 3.0 .03
L 9 15 37 1.1. 30 32 .6
Test 3
36 U 0 1. 90 36 5 66.7 57.8 17.8 .19
L 3 11 66 51 1. 1.8.9
37 U 108 15 11 1 0 80 .0 69.2 21.5 .26
L 19 33 9 11. 0 58.5
38 U 35 18 9 69 1. 51.1 314.8 32 .6 .36
L 79 13 11. 2_5_ 1. 18 .5
39 U 16 36 19 51 13 26.7 21.5 10.1. .15
L 17 23 1.3 1.9 1. 16.3
1.0 U 11. 56 51 10 1. 37.8 28.5 18 .5 .23
L 18 66 26 18 7 19 .3

 

50.
TABLE VIII (cont.)

 

 

% Average Correlation
Item No. of Each Response Correct % Item to
No. 1 2 3 1. 5 Omit Choice Correct 115-1.: Total Test
(Flanagan)

1.10 12 116 0 7 0 85.9 83.3 5.2 .09
L 16 _1_9_9 2 5 3 80.7

1.2 U 11 51 7o 3 0 51.9 37.0 29.7 .33
L 22 75 32 1. 1. 22 .2

1.3 U 25 67 10 21. 9 1.9.6 1.3.7 11.8 .12
L 32 51 17 21 11. 37.8

1.1. U 31 81 11 9 3 60.0 58.1 11.8 .12
L 30 65 15 15 10 1.8.2

1.5 U 21. 61. 12 17 18 117.1. 35.5 23 .7 .26
L 1.3 32 21 17 22 23 .7

1.6 U 26 1.8 13 32 16 23 .7 17 .0 13.3 .21.
L 1.3 30 30 _1_1. 18 10.1.

1.7 U 56 26 8 21 11. 19 .3 18.9 .8 .01
L 37 25 19 15 39 18.5

1.8 U 9 69 15 12 30 51.1 1.3.3 15.5 .17
L 5 1.8 17 17 1.8 35 .6

1.9 U 1. 1.5 6 1.3 37 31.9 20.7 22.3 .33
L 7 55 11. 13 1.6 9.6

50 U 11 31 21. 28 1.1 17.8 11.8 11.9 .26
L 21 30 _8 22 51. 5.9

Test 1

51 U 10 115 10 0 85.2 70.0 30.1. .36
L 36 11. 23 2 511.8

52 U 121 8 5 1 89.6 71.1 37.0 .1.6
L 11 53 11 0 52 .6

53 U 0 126, 8 1 93 .3 72.2 1.2.2 .51.
L 0 62 1.6 20 51.1

51. U 20 20 95 0 70.1. 55.9 28.9 .30
L 29 1.5 56 5 1.1.5

51.

TABLE VIII (cont.)

 

 

Z Average Correlation
Item No. Of EECh Response Correct % Item to
No. l 2 3 1. 5 Omit Choice Correct 05-18 Total Test
(Flanagan)
55 U 1 103 29 2 76.3 55.2 1.2.2 .1.3
L 13 1._6_ 76 0 31..1
560 2 129 2 2 95.6 81.1 28.9 .1.8
L 26 9_9_ 17 2 66.7
570 13 12 110 0 81.5 61.1. 31.1 .38
L 33 32 61. 6 1.7.1.
580 7 126 1 1 93.3 87.1. 11.8 .25
L 13 113 9 3 81.5
59 U 121 7 6 1 89.6 80.0 19.2 .29
L 25 15 23 2 70.1.
60 U 12 8 115 o 85 .2 62.6 1.5.2 .1.9
L 23 1.6 51. 12 1.0.0
610 1. 121 8 2 89.6 70.3 38.5 .117
L 13 62 1.0 13 51.1
62 U 126 1. 3 2 93 .3 73.7 39.2 .52
L 13 22 17 23 51..1
63 U 16 39 73 7 5h.1 h3.3 21.5 .23
L 21. 27 1.1. 1.0 32 .6
61. U 9 13 106 7 78.5 51.1. 511.1 .51.
L 17 31. _3_3_ 51 21.1.
65 U 35 11. 71. 12 511.8 18.0 21.5 .23
L 16 15 1.5 59 33 .3
Test 5
66 U 0 0 0 0 129 6 95 .6 88 .1 111.9 .31.
L 1. 5 5 7 1g 5 80.7
6711 1 101. 19 2 8 1 77.0 61..1. 25.1 .28
L 11 19 21 13 15 5 51.9
68 U 110 1 2 2 11. 6 81.5 61.1 1.0.8 .1.3
L 55 13 16 18 26 7 1.0.7

 

TABLE VIII (cont.)

52.

 

 

% Average Correlation
Item No. of Each Response Correct % Item to
No. l 2 3 h 5 Omit Choice Correct U%-L% Total Test
(Flanagan)

69 U 37 19 11. 1.2 9 11. 31.1 20.3 21.5 .32
L 1.8 26 26 13_ 10 12 9.6

70 U 11. 11 66 13 11 10 1.8.9 37.8 22 .2 .21.
L 27 21 _3_§_ 5 27 19 26 .7

71 U 11. 35 11. 16 21. 32 10.1. 12.6 -1..1. -.09
L 19 30 20 8 31. 23 11..8

72 U 11 26 25 25 13 35 18.5 16.6 3.7 .07
L 15 32 22 39 20 26 11..8

73 U 19 28 19 22 7 1.0 20.7 11..8 11.8 .22
L 23 1g 35 19 12 31. 8.9

71. U 25 11 13 9 38 39 28.2 21.1 11..1 .20
L 28 19 11 22 19 36 11..1

75 U 13 17 22 27 17 39 20.0 11..8 10.1. .19
L 9 31. 26 _1_3_ 16 37 9.6

76v 28 1.6 11. 2 3 1.2 31.1 20.7 26.7 .1.0
L 39 1g 16 8 7 55 7.1.

77 U 10 8 7 1.6 9 55 31. .1 23 .0 22 .2 .30
L 11. 12 17 16 8 68 11.9

78 U 13 3 25 1. 10 80 18.5 M. 8.1 .15
L 7 9 111 6 11. 85 10.1.

79 U 16 7 h 12 5 91 3.7 11.1 -.7 .0
L 11 15 8 10 g 85 1..1.

80 U 11. 8 10 1. 3 96 10.1. 8.1 1.5 .12
L 8 19 7 6 8 87 5 .9

Test 6

81 U 0 0 121. 9 2 91.9 85 .2 13.1. .21.
L 9 7 196 12 1 78.5
L 15 _§_ 15 21. 3 57.8

53 .
TABLE VIII (cont.)

 

Correlation

 

% Average
Item Correct Item to
No. l 2 3 h 5 Wit Choice Correct U%-L% Total Test
(Flanagan)

83 U 0 0 29 103 3 76 .3 511.1 111.1. .1.5
L 9 11. 69 1.; 0 31.9

81. U 1. 128 3 0 0 91.8 72 .2 1.5.2 .58
L 30 51 17 11. 7 1.9.6

85 U 20 1. 109 0 2 80.7 59.2 1.2.9 .1.5
L 58 17 51 6 3 37.8

86 U 7 0 108 17 3 80.0 59.6 1.0.7 1.3
L 21 11 51 1.6 1. 39 .3

87 U 59 19 36 18 3 113.7 31.1 25 .2 .29
L 15 31 1.1 27 11 18 .5

88 U 6 19 29 69 12 21.5 16.3 10.1. .18
L 11. 21. 15 71 11 11.1

89 U 81. 11 35 2 3 25 .9 21. .0 3.7 .05
L 1.6 20 _3_9_ 23 16 22 .2

90 U 29 50 7 112 7 37 .0 21. .8 214.11 .32
L 1.9 17 22 21. 23 12.6

91 U 102 1. 16 8 5 75 .6 50.7 1.9 .7 .50
L _35 30 17 18 35 25 .9

92 U 1. 107 10 1 13 7.1. 7.0 .7 .02
L 10 71 g 9 36 6.7

93 U 17 81. 22 7 15 62 .2 38.1 1.8.1 .51
L 16 151 29 20 51 11.1

91. U 37 11. 23 1.0 21 17 .0 12.9 8.1 .16
L 1.0 15 12_ 15 53 8.9

95 U 27 21 19 29 39 21.5 21.5 .0- -0-
L 19 11. 17 32

Test 7

96 U 0 6 0 128 1 91.8 91.5 6.6 .18

L 3 3 5 119 5 88 .2

 

51. .
TABLE VIII (cont .)

 

 

% Average Correlation
Item No. Of Each Response Correct % Item to
No. l 2 3 1.. 5 Wit Choice Correct U%-L% Total Test
(Flanagan)
97 U 0 127 0 7 1 91.1 86.7 11..8 .29
L 10 . 191 10 6 2 79 .3
98 U o 6 129 0 0 95 .6 89.3 12 .6 .30
L 11 1. 111 7 1 83 .0
99 U 2 126 7 0 0 93 .3 80.0 26 .6 .1.1
L 13 29 17 12 3 66.7
100 U 0 11 119 1. 1 88 .2 75 .6 25 .2 .33
L 11 11. 15 25 0 63 .0
101 U 39 67 5 18 6 13 .3 12.9 .7 .01
I. 30 1.6 37 11 5 12.6
102 U 23 1.1. 21 1.0 7 29.6 21. .0 11.1 .15
L 13 1.5 33 25 19 18.5
103 U 1 1. 121 8 1 89.6 69 .2 1.0.7 .1.8
L 23 15 _6_§_ 19 12 1.8 .9
101. U 0 120 2 11 2 88.9 71.1. 28.9 .38
L 11 _81 16 18 9 60.0
105 U 3 126 1 5 0 93 .3 71.1. 1.3.7 .55
L 15 51 23 21 9 1.9.6
106 U 0 0 130 5 0 96 .3 86.6 19 .3 .1.1
L 8 10 191 9 1. 77 .0
107 U 26 18 7 79 5 58.5 38.5 1.0.0 1.1.
L 37 31. 23 15 4.16 18 .5
108 U 7 107 8 10 3 79 .3 58.1 1.2.3 .111.
L 15 59 3o 32 8 37.0
109 U 12 18 30 59 16 22.2 22 .2 -0- -0—
L 17 29 _3_g 36 23 22 .2
110 U 23 68 19 17 8 50.1. 35.5 29.7 .33
L 11. 28 21 50 22 20.7

 

TABLE VIII (cont.)

55.

 

 

% Average Correlation
Item No. Of Each Response Correct % Item to
No. l 2 3 1.1 5 Wit Choice Correct 06-13% Total Test
(Flanagan)
111 U 1. 6 123 0 2 91.1 68 .9 1111.1. .511
L 30 22 _6_5 9 11 1.6.7
112 U 8 6 2 119 0 88.2 66.3 1.3.8 .1.8
L 28 20 10 59 17 1.1.1.
113 U 16 5 112 o 2 83 .0 65.6 38.8 .39
L 18 26 Q5 15 11 1.8 .2
111. U 31. 1.5 11. 29 13 33 .3 26.3 11.0 .17
L 29 g6_ 20 36 21. 19 .3
115 U 6 l1. 9 101 5 71.8 51.1. 1.0.7 1.2
L 20 33 21 1.5 15 31..1
116 U 5 12 112 1. 2 83.0 57.8 50.1. .51
L 11. 27 g 26 21. 32.6
117 U 11 119 1. 0 1 88.2 61.8 1.6.7 .52
L 16 55 20 21 22 1.1.5
118 U 10 13 h 1011 1. 77.0 1.9 .2 55 .5 .51
L 27 23 25 19_ 31 21.5
119 U 23 71 18 10 13 52 .6 27.1. 30.1. .33
L 21. 59 21 23 37 22 .2
120 U 36 35 37 13 11. 27.1. 21. .0 6.7 .09
L 13 36 211 29 28 20.7
121 U 9 112 8 1. 2 83 .0 59 .3 1.7.1. .19
L 26 1_8_ 29 7 25 35 .6
122 U 1.0 16 3 75 1 55 .6 1.1.9 27.1. .28
L 29 23 19 311 26 28.2
123 U 21. 10 78 21 2 57.8 1.2.2 31.1 .32
L 17 18 _3_§ 22 1.2 26 .7
121. U 23 23 50 26 13 37.0 26 .3 21.1. .28
L 18 39 21 22 35 15 .6

 

TABLE VIII (cont .)

 

 

% Average Correlation
Item No. of Each Response Correct Z Item to
No. l 2 3 1. 5 Wit Choice Correct Uz-Lz Total Test
(Flanagan)
125 U 17 31 30 1.5 12 23 .0 19 .3 7.1. .11
L 22 _21 32 25 35 15 .6
126 U 21 10 89 5 10 66.0 1.3.0 1.6.0 .1.7
L 19 31 11 13 1.5 20.0
127 U 8 56 l. 59 8 1.1.5 28.9 25 .2 .30
L 20 1'5 16 38 39 16.3
128 U 13 35 31 31 25 23 .0 l7 .0 11.9 .20
L 16 21 33 15 50 11.1
129 U 20 58 18 21 18 1.3.0 314.8 16.3 .18
L 17 16- 22 15 1.5 26 .7
130 U 30 11. 19 1.7 25 31. .8 21. .0 21.5 .29
L 21 17 21 18 58 13.3
131 U 36 13 1.1. 22 20 32 .6 21.1. 16.3 .22
L 22 20 11 11. 57 16 .3
132 U 19 23 51 23 19 37.8 25.5 21. .5 .32
L 10 18 18 21. 65 13.3
L 12 1.0 9 11 57 12.6
131. U 16 30 17 1.1 31 22 .2 19.6 5.2 .08
L 11. 15 12 18 68 17 .0
135 U 15 23 1.2 21. 31 31.1 22.6 17 .0 .21.
L 12 22 19_ 7 75 111.1
136 U 19 29 39 10 38 21.5 17.8 7.1. .11
L 7 19_ 18 13 78 111.1
137 U 16 20 53 13 33 39 .3 27.1. 23 .7 .30
L 5 20 11 12 77 15 .6
138 U 1.0 11 13 26 1.5 29 .6 22.9 13 .3 .18
L 22 9 12 10 82 16.3

 

57.
TABLE VIII (cont.)

 

 

% Average Correlation
Item No. Of EaCh Response Correct % Item to
No. l 2 3 1. 5 Wit Choice Correct U%-L% Total Test
(Flanagan)
139 U 21. 11 27 25 1.8 20.0 15.2 9.6 .17
L 10 20 11 10 81 10.1.
11.0 U 11. 32 29 13 1.7 10.1. 7.1. 6.0 .18
L g 9 23 13 81. 1..1.
11.1 U 22 16 27 21. 1.6 17.8 12.6 10.1. .21
L 13 15 10 19 87 7.1.
11.2 U 35 10 23 13 51. 17 .0 12.9 8.1 .16
L 15 15 12_ 3 90 8.9
11.3 U 8 11 17 52 1.7 8.2 8.2 —0— -0-
L 1. 11 11. 17 89 8.2
1141. U 11. 17 39 15 50 12.6 12.6 -0- -0-
L 6 _11 17 10 85 12.6
1115 U 3h 20 5 23 53 111.8 11.5 6.6 .111
L 16 11 17 7 81. 8.2

 

11001.1 USE 81111

Demco-293

Date Due