EDUCATIONAL RECORDS SUPPLEMENTARY BULLETIN H
Functions of the Educational Records Bureau
in Comparable Measurement
r 763 f
EDUtATIONAL RECORDS BUREAU
437 West 59th Street
New York 19, N. Y.
EDUCATIONAL RECORDS SUPPLEMENTARY BULLETIN H
FUNCTIONS OF THE EDUCATIONAL RECORDS BUREAU
IN COMPARABLE MEASUREMENT
Arthur E. Traxler
EDUCATIONAL RECORDS BUREAU
437 West 59th Street
New York, N. Y.
Introduction. . . . . .
Early Objective Tests . . .
State-Wide Testing Programs .
National Testing Programs . .
Establishment of the Educational
How the Bureau Functions. . .
Bureau Testing Programs . .
Publications. . . . . .
Test Program Bulletins . .
Teacher Education in the Use
Research . . . . .
Committee Reports. . . .
Cumulative Records. . . .
Services to Public Schools. .
Services to Colleges. . . .
. . ....... . ..
. . . . . . . . .
Records Bureau ........
. .. .
Cooperation with Other Examining 'Agencies
Educational Conferences. . .
Test Construction. . . . .
Use of New Measurement Devices ,
'Advisory Services . . . . .
Services to Government 'Agencies .
U.S. Office of Indian Education.
U.S. Merchant Marine Academies .
Qualifying Test for the Army and
College Training Program . .
SConclusion. . . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . .
. . . .
THE FUNCTIONS OF THE EDUCATIONAL
RECORDS BUREAU IN COMPARABLE MEASUREMENT
An understanding of the work of the Educational Records Bureau can be had
only by viewing it against the background of the wnole measurement movement
Like a number of similar organizations, it is the product of a philosophy and a
method of studying school children that are relatively new in America, that are
not founded on tradition, and that have no counterpart outside this country.
In the early history of examinations, attempts at the measurement of the
ability and achievement of students were sporadic and unorganized Up to the
second decade of the present century, measurement was almost entirely an in-
stitutional affair and it has remained just that for the great majority of schools
and colleges almost to the present time Examipations have traditionally been
made by classroom teachers who have had neither training for nor special interest
in this aspect of their work Except for such advanced thinkers as Horace Mann,
the idea that there could be a theory and technique of measurement apart from the
process of instruction itself seems not to have occurred to anyone until recent
Because examinations were for the most part made carelessly and without
reference to the experience of other persons, the typical early examination was
very inadequate Examinations were at first entirely oral and were not infre-
quently concocted on the spur of the moment to confound the unhappy candidate
The written essay examinations which began to be introduced in America early in
the nineteenth century were an improvement over the oral ones, but their lack of
reliability and validity was amazing--all the more so because their inferiority
was so seldom suspected by their complacent authors.
So great sas the faith of each individual school and college in its own
particular examination system at the turn of the last century that it required
- 2 -
much work and persuasion on the part of such outstanding leaders as President
Eliot of Harvard and Nicholas Murray Butler of Columbia to bring some order
out of the chaos. Finally, however, their efforts prevailed and the College
Entrance Examination Board was established. Later, the Secondary Education
Board was set up to perform for private secondary schools a function similar
to the work of the College Board at a higher level. The services of the College
Entrance Examination Board and the Secondary Education Board in improving essay
examinations and in bringing about uniformity in examining from one institution
to another are too well-known to require comment here.
Early Objective Tests
Probably brief-answer tests have occasionally been used by teachers for
generations, but apparently it was less than fifty years ago that anyone
realized that there was inherent in the brief-answer question a technique on
which a science of education might be built Once the idea was expressed,
however, it took hold very rapidly From 1897, when Rice made his now historic
speech, "The Futility of the Spelling Grind," to the Department of Superintend-
ence of the National Educational Association, up to the middle 1920's, there
was a great development of objective tests of all kinds without much conscious
attempt at organization. Shortly after 1900, the idea of objective measurement
first expressed by Rice caught the interest of Thorndike, Stone, Freeman,
Courtis, Kelley, and a host of other rising young educational leaders and they
in turn set numerous students to work, The result was a multiplicity of
objective tests in nearly every subject field
Many of these new tests were very inferior. For years, the concept of
measurement ran far ahead of technological development in the field, and the
sudden interest of commercial organizations in objective tests was not altogether
salutary The work of Thorndike and some of the other.earlier leaders in test
- 3 -
construction was monumental, but many other persons did not have the same genius
in test making. The result of all the influences at work was that many hastily
constructed and poorly validated measuring instruments were thrown on the market
to be seized upon by hundreds of administrators and teachers who were entirely
unprepared-to use test results intelligently but who had been told that these
new instruments would work wonders in their schools. Confusion, misuse, and
disillusionment followed in rapid succession in many places. It is true that
here and there schools under intelligent and trained leadership were able to
work out excellent guidance programs based on measuring instruments that had
been carefully selected from the mass of available tests, but these schools were
the exception rather than the rule.
The great need was for centralized leadership to organize and direct both
test construction and use. To meet this need, a variety of agencies has arisen
within the last twenty years.
State-Wide Testing Programs
One of the first and most influential of the efforts directed toward the
organization and unification of testing efforts is found in various state
testing programs. It appears that the first state-wide testing program was
begun in Minnesota shortly after the first World War, but other states quickly
followed this lead until testing has been undertaken on a state-wide basis
in at least sixteen states, including Alabama, Colorado, Connecticut, Georgia,
Indiana, Iowa, Kansas, Kentucky, Michigan, Minnesota, Ohio, New Hampshire,
North Carolina, South Carolina, Washington, and Wisconsin. One of the oldest,
best organized, and best known of these is the Iowa State Testing Program,
under the direction of Dr. E. F. Lindquist. The tests constructed for use in
that state are also employed in several other programs.
- 4 -
Under the leadership of the state university or one of the other institu-
tions of higher learning, some of these state programs began with the testing
of the high-school seniors. Test scores secured in this way have served to
provide colleges with objective data for the admission and placement of students,
It soon became apparent, however, that the testing of high-school seniors,
valuable though this procedure proved to be, was not enough. It became in-
creasingly evident that what was needed was a program of annual comparable
measurements that would serve as a basis for guidance in the elementary and
secondary schools and would provide highly reliable information for selective
admission and placement in college. The long-time study of the Relations of
Secondary and Higher Education in Pennsylvania, conducted by the Carnegie
Foundation under the direction of Dr. William S. Learned and Dr. Ben D. Wood,
greatly added impetus to the movement for comparable measurements. We find,
therefore, that to an increasing degree the state testing programs have tended
to be placed on an annual basis and to cover a range of.grades.
The state testing leaders have not been satisfied to rest with the giving
of the tests. In many instances, they have provided scoring services and have
made up state-wide norms each year, and all of them have furnished expert
advice in testing and guidance, information on the use of the test results,
and various summary reports to the schools in their respective areas.
National Testing Programs
Although state-wide testing programs have been of inestimable help to
the guidance programs of schools in a number of states, a need has quite
naturally grown up for the services of organizations that would transcend state
boundaries and that would provide testing services on a nation-wide scope The
most ambitious venture in the production of tests to serve secondary schools
and colleges throughout the country has been the establishment of the Cooperative
- 5 -
Test Service, under the auspices of the American Council on Education, through
a subvention from the General Education Board. The broad purposes of this
service were to help coordinate testing and guidance in America and to co-
operate with institutions and individuals interested in adjusting educational
procedures to the needs of individual students. Under the direction of Dr. Ben
D. Wood, the Cooperative Test Service undertook, starting in 1932, to produce
annually an extensive series of comparable forms of tests covering all academic
fields for both the junior college and the secondary school. After having
issued six forms of a series of tests, most of which were ninety minutes in
length, the Cooperative Test Service began in 1938 to produce a new series of
forty-minute tests which are comparable in difficulty with the longer ones
For most high-school and junior-college subjects there are now several forms
of the shorter tests. These tests, which are made with the assistance of many
subject matter and testing specialists, are each year distributed very widely
throughout this country. National norms for various types of institutions are
now available so that it is possible for any secondary school or college to
evaluate achievement and.accumulate comparable measurements for individuals in
all subjects ordinarily included in the academic curriculum.
Establishment of the Educational Records Bureau
In addition to the need for a non-commercial test-making agency whose
activities would be nation wide, such as the Cooperative Test Service, a need
began to be felt several years ago for an organization that would not publish
tests but rather would assist schools in scoring, interpreting, and using tests
in individual guidance. Therefore, in 1927, about five years before the
Cooperative Test Service began to issue its tests, the Educational Records
Bureau was formed by a small group of independent, or private,schools to serve
- 6 -
as a research and service agency and to assist them in obtaining annual comparable
measurements on their pupils. The Bureau was chartered under the Board of Regents
of the University of the State of New York as a nonprofit-making service and
research agency to schools and colleges. It has grown steadily until now it in-
cludes about three hundred fifty institutions. Most of these are independent
elementary and secondary schools, although about twenty public schools and ap-
proximately the same number of colleges are included in the membership. The
Bureau is in no sense a closed corporation; rather, its services are available
to every school, and new schools are constantly being added to its list of members.
How the Bureau Functions
The policy of the Bureau and the broad features of its work are determined
by a Board of Trustees elected by the member institutions. Functioning under
the Board, there are four standing committees, known as the Independent Schools
Advisory Committee, the Committee on Tests and measurements, the Public Schools
Advisory Committee, and the Committee on School and College Relations. The
members of these committees are elected by the schools belonging to the Bureau
There is a permanent staff of twenty persons, which is augmented at certain
times of the year by a part-time staff of from fifty to two hundred individuals.
Although the work of the Educational Records Bureau has been aided materially
by a number of small grants from different organizations, the Bureau may properly
be regarded as a self-supporting organization. Its existence is dependent en-
tirely upon its member schools. The annual membership fee is fifteen dollars for
each institution, but by far the largest proportion of its income is derived from
its services in connection with scoring and reporting the results of tests for
the different schools belonging to the Bureau. Without the income from this
source, the Bureau could not exist. Its success, therefore, is-due to the loyal
1For a core complete history of the Bureau see: Eleanor Perr Wood, *The Bureau--Ten Years Old,* The
1937 Achievement testing Program of the Educational Records bureau. Educational Records Bulletin No. 20,
- 7 -
support of its member institutions and to the highly personalized service to
these schools that were developed from the beginning by Mrs. Eleanor Perry
Wood, who was for several years the Associate Director of the Bureau
Bureau Testing Programs
It was the conviction of the schools that originally banded together to
establish the Educational Records Bureau that comparable measurements based on
comprehensive objective tests are of great importance in a continuing study of
pupils These schools recognized such measurements as an indispensable sup-
plement to other types of information used in constructive educational guidance
and in administering transfers, promotions, and admissions to secondary schools
and to colleges. The members of the Bureau have consistently adhered to this
conviction. Consequently, the conducting of testing programs has from the
beginning been the most important activity in which the Bureau has engaged
Two testing programs are held annually by the Bureau's member schools
The fall program consists of academic aptitude and reading tests, supplemented
by some achievement tests for placement and diagnosis In the spring, a com-
prehensive achievement testing program is conducted in all academic subjects
at the elementary and secondary school levels The tests to be used are
selected by the Committee on Tests and Measurements in consultation with the
members of the Executive Staff Each school is free to decide the extent to
which it will participate in the testing programs The schools obtain the
tests and complete directions for giving them from the Bureau They administer
the tests themselves, usually within a specified period, although they may
utilize the services of the Bureau for testing at other times if they so desire.
Some schools score their tests locally, but the majority return them for
central scoring by the Bureau-s staff at a per-test-per-pupil cost.
The services of the Bureau in connection with the tests consist of much
more than merely scoring them Distributions of the scores are made and the
- 8 -
necessary statistical work, such as the computation of medians and quartiles,
is done. Class lists are prepared and sent to the schools showing part scores
and total scores and percentile ratings corresponding to the scores. The
tables from which the percentile ratings are derived are based on the results
of tests scored by the Bureau's staff. Locally scored test results are not
included in the norms A special technique for insuring accuracy in the
scoring is carefully followed in connection with all tests centrally scored.
When the test results are returned to a school, they're accompanied by
an extensive descriptive and interpretative report. It is customary for the
person preparing this report to take up the scores of each class on each of
the tests and to discuss for the school the indications in the results for
both groups and individuals. Sample distributions, a class list and an il-
lustrative letter explaining and interpreting the scores of a ninth-grade class
on the Cooperative English Test are shown on pages 9 to 14.
As already indicated, nearly all the schools participating in the annual
testing programs of the Educational Records Bureau are independent elementary
and secondary schools. It is therefore possible for the Bureau to develop
norms for independent schools, and this .is highly desirable, Since the in-
dependent schools are mainly college preparatory in nature, they enroll a
selected group of students whose median .intelligence quotient on a test such
as the Otis is about 115 The norms for public schools are, therefore, much
too low and norms for this selected college preparatory group.are essential
to effective guidance in these schools During the last fifteen years, the
Educational Records Bureau has developed by far the most comprehensive body
of normative data for college preparatory pupils available anywhere In fact,.
aside from Koo's study Private and Public Secondary Education,1 the data in
the Educational Records Bureau files provide almost the only comprehensive
objective appraisal of the academic aptitude and achievement of independent
'Leonard V. Koos. Private and Public Secondary Education. Chicago: The University of Chicago Press, 1931.
School Park Country Day
COOPERATIVE ENGLISH TESTS, FORM R
Date April 29, 1943
Scaled A7 MECHANICS B17 EFFECTIVENESS Cl: READING TOTAL
Score OF EXPRESSION OF EXPRESSION COMPREHENSION A+ B-C
72 1 1
64 1 2 ....3...................... 1
62 _.... A. ................... ..!. 2
60 2 2 3 .... ..................... 2...................
58 ... "................... ---....-2 .-................ 2.. -.2 -
58. 1------------------------- --------------------------
56 ....3 ................... ,, 1 ... ........... .... .- 1
54 1 ........................ '.................-
52 :' -' -. .. 1 ...... .. .. 17:-,." .... 7 7.................
4 1 -- ...._1................. ,- 2 ... ..................
40 ................. i ________________-...............
38 --- --1- ----- -
Total 21 21 21 21
Q3 57 8 62 9 .59,8 60 8
Md 52,5 .59Z5 .55 5 55 3
01 48 8 50 3 .50 6 .50 8
Ran e 36-64 39-72 39-74 37-72
-- ----End-of-year public-school sean
.......... Independent-school median
COOPERATIVE ENGLISH TEST CI: READING COMPREHENSION, FORM R
School Park Country Day Grade 9 Date April :29, 1943
------ End-of-year public-school mean
.......... Independent-s hool median
School Park Country Day City New York State 1--
Date of Date
Grade 9 ____ Report April 28, 1943 Adm. April 7, 1943
COOPERATIVE ENGLISH TESTS, Form R
CHRON. B1 C1 TOTAL
Names of Pupils A0 j MECH. SF EFFECT. OF READING COMPREHENSION ENGLISH
S EXPRESSION PRESS V b. peed Level Tot
rr-s 1r1 r-rr~dL'"
1. Barton, Barbara A.
2. Bradley, John W.
3. Denton, Robert D.
4. Dickinson, Samuel J.
5. Duncan, George W.
6. Elkins, George H.
7. Fenton, Frederick B.
8. Frost, Virgil C.
9. Haynes, Julie A.
10. Hunt, Barbara A.
11. Kerr, Mary J.
12. Livingstone, Martin H.
13. McKean, Sarah A.
14. McNaughton, Jean
15. Prescott, Elsie
16. Royer, Laurence
17. Simpson, Martha F.
18. Smith, Marie A.
19. Swanson, John S.
20. Thompson, Carol E.
21. Warren, Gertrude W.
Or.da No. of C.Hss
60 188 63
- 12 -
EDUCATIONAL RECORDS BUREAU
437 West 59th Street
New York City
April 29, 1943
Mr. Harold W. Stetson
Park Country Day School
New York, New York
Dear Mr. Stetson:
We are sending herewith the report of the results of the Cooperative
English Test, Form R, which was administered recently to the pupils in the
ninth grade of the Park Country Day School. The report consists of dis-
tributions and a class list showing the scores in mechanics of expression,
effectiveness of expression, vocabulary, speed of comprehension, level of
comprehension, total reading comprehension, and total English score. We
have entered independent-school percentiles on the class list, which we are
sending in duplicate.
The class list is probably the most interesting and useful part of the
report, since it shows the scores and percentile ratings of the individual
pupils. However, when one is interpreting the results for the class as a
whole, it is helpful to study the distributions, for they show how the
scores compare with the medians or norms for independent- or private-school
The general form of the distribution sheet may need a word of explana-
tion, although the distributions are quite easy to read when one understands
how they are arranged. The column closest to the left-hand margin of the
distribution sheet shows the score scale arranged in intervals of two. The
dotted lines across the distribution columns indicate the independent-school
medians for the various grades and the heavier broken lines show the public-
school medians. The medians and ranges of the scores of your ninth-grade
pupils are reported at the bottom of each distribution sheet. The median
is also shown graphically by the short horizontal.line adjacent to each
distribution, and the range.of the middle 50 per cent of the cases is in-
dicated by the vertical line parallel to the distribution.
The results of the Cooperative English,Test, and of most of the other
Cooperative tests, are expressed in terms 9f a kind of derived score, known
as the Scaled Score. The Scaled Score system is based on a mean of 50 and
a standard deviation of 10 for a defined group. A Scaled Score of 50 is
the score that the average pupil with an'I.Q.' of 100 would be expected to
make at the end of the course if he had had the usual kind and amount of
instruction. The Scaled Scores tend to range from 20 to 99, but the scores
of beginning students are sometimes below 20 on certain tests, and very
superior advanced students may sometimes make Scaled Scores slightly above
100 in certain aspects of the Cooperative tests. In English, or any other
subject which is studied throughout the secondary school, a Scaled Score of
- 13 -
Mr. Harold W. Stetson -- 2 April 29. 1943
50 is the score that the average pupil would be expected to make at the end
of his high-school course. For one-year subjects, such as plane geometry,
the end of one year of study is, of course, the reference point.
The independent schools, as you know, tend to enroll pupils who are
above average in academic aptitude. Because of this fact, the Scaled Score
of the average independent-school pupil near the end of the course tends to
be approximately 60.
Persons who are well acquainted with tests have found the use of Scaled
Scores very helpful, but teachers who are not accustomed to thinking in sta-
tistical terms may feel that they are unnecessarily technical. The Scaled
Scores are preferable to raw scores for purposes of comparing the results of
different tests and of the several parts of the same test. However, teachers
who are not interested in working out the implications in Scaled Scores may
employ them in the same way that they have always used raw scores. It should
be made clear to the teachers that these scores are not raw scores and do not
show directly how many questions each pupil has answered correctly. Teachers
who are just beginning to work with objective tests usually are able to grasp
the meaning of percentiles more readily than Scaled Scores.
You will note that there are two distribution sheets for the English
test. The first sheet shows the distributions of the Scaled Scores on Test A,
mechanics of expression; Test Bl, effectiveness of expression; Test Cl, reading
comprehension; and on these three tests combined. The second sheet contains
the distributions for vocabulary, speed of comprehension, level of comprehension,
and the total reading comprehension score. It is apparent, therefore, that
the distributions given in the right-hand column of the second sheet and in
the third column of the first sheet are identical
If you will refer to the last column of the first sheet, you will note
that the median total Scaled Score made on the English test by your ninth
grade, 55.3, is almost identical with the independent-school median for Grade X.
Fourteen of the twenty-one pupils have total Scaled Scores above the independent-
school median for the ninth grade.
The total scores of this class are distributed over a rather wide range.
The lowest total Scaled Score, 37, made by Frederick Fenton, is slightly below
the end-of-year public-school mean for Grade VIII, and corresponds to an in-
dependent-school ninth-grade percentile rating of 5. At the top of the class,
Samuel Dickinson stands out with a total-Scaled Score of 72, which is equivalent
to a percentile rating of 98. This pupil's score is approximately ten Scaled
Score units, or one standard deviation, ab6ve the independent-school median
for Grade XII. Three other pupils, Gertrude Warren, Martha Simpson, and John
Swanson, have total Scaled Scores which reach or surpass the independent-school
Perhaps a short statement concerning the interpretation of percentiles
would be helpful. As you no doubt know, a percentile shows the per cent of
the scores in a distribution which are exceeded by .a given score- Samuel
Dickinson's percentile, 98, means that his total English score is above those
- 14 -
Mr. Harold W. Stetson -- 3 April 29, 1943
of 98 per cent of the independent-school ninth-grade pupils, and is surpassed
by those of 2 per cent. The lowest total score percentile, 5, made by
Frederick Fenton, means that this pupil's total score exceeds those of only
5 per cent of the independent-school pupils in Grade IX, and is surpassed by
those of 95 per cent of the pupils. Percentiles range from 1 to 100, and a
percentile of 50 is, of course, the median or average.
Turning to the distributions for the three main parts of the test, one
sees that the median Scaled Score of your ninth grade in effectiveness of
expression, .'59.5, is slightly above the independent-school median for Grade XI.
The medians for mechanics of expression and reading comprehension fall between
the independent-school medians for Grades IX and X. The level of comprehension
median, as shown on the second distribution sheet, is almost identical with the
independent-school eleventh-grade median, while the vocabulary.and speed of
comprehension medians are close to the medians for independent-school pupils
in Grade IX.
In general, it appears that your ninth grade is, .as a group, considerably
advanced in effectiveness of expression and level of reading comprehension,
is slightly above average for independent-school ninth-grade classes in
mechanics of expression and total reading comprehension, and is almost exactly
average for the independent-school group in vocabulary and speed of reading
comprehension. It is .apparent that nearly all the pupils are .above average
in comparison with the public-school norms for ninth-grade English.
The Scaled Scores and percentiles on the parts of the English test are
useful in the diagnosis of strengths and weaknesses of individual pupils as
well as classes. For example, Julie Haynes is rather low in effectiveness of
expression, is not far below the independent-school ninth-grade median in
mechanics of expression, and is above the median in total reading comprehension
Martin Livingstone, on the other hand, is high in effectiveness of expression,
is close to the independent-school median in mechanics of expression, and is
considerably below the median in total reading comprehension. It seems probable
that these two pupils need different kinds of special help in English.
Please let us hear from you if there are questions concerning this report,
or if we can be of assistance to you or your teachers in the interpretation
and use of the scores on the test.
- 15 -
In addition to the testing programs, the Educational Records Bureau carries
on a number of other activities, some of which have increased in importance
within the last few years.
Test program bulletins.-Following each testing program, the Bureau issues
a bulletin summarizing the results for the whole group of independent schools
whose tests were scored centrally. These bulletins customarily show the dis-
tribution of scores for each independent-school grade or class on all the tests
included in the program, together with medians and quartiles, and also public-
school medians for comparative purposes. The distributions are frequently
accompanied by descriptive and interpretative material. A typical distribution
page is reproduced on page 16.
The bulletins dealing with the achievement tests also contain comparative
charts in which the median scores made by every class in each of the schools
are shown by code numbers on a scale of percentiles arranged according to unit
of sigma on the normal probability curve. Through observing the position of
its code number on the scale, any school can trace out the median achievement
of all its classes. The chart for literary acquaintance and spelling in
Grades IX-XII, which appeared in a recent bulletin, is reproduced on page 17,
and the position of one school (239) is shown graphically by the solid line
across the chart. It will be seen that the medians for this school were up to
or above the independent-school medians except in the case of spelling at the
Teacher education in the use of tests.-From the beginning, a great deal of
the energy of the Bureau staff has been directed toward assisting teachers in
the interpretation and use of test results The Bureau has reinforced its em-
phasis on this important aspect of testing and guidance through the publication
'A price lhot of the Bure.au val.lsble publications a..y be obtained on request.
- 16 -
TEST C: READING COMPREHENSION
COOPERATIVE ENGLISH TEST, FORM T
DISTRIBUTIONS OF PUPIL TOTAL SCALED SCORES AND CLASS MEDIANS OF
INDEPENDENT SCHOOLS PARTICIPATING IN ACHIEVEMENT TESTING PROGRAM, APRIL, 1943
SCALED PUPIL SCORES CLASS MEDIANS SCALED
SCORE 7 8 9 10 11 12 7 8 9 10 11 12 SCORE
92+ .5 4
90 1 2 4 3
88 4 1 1 .5
86 2 1 6 10
84 1 3 8 9 19
82 2 2 4 8 20
80 3 12 11 16 18
78 2 5 5 28 24
76 3 11 11 26 32
74 1 4 15 17 21 38
72 10" 24 24 48 62
70 12 16 28 50 ,59
68 25 43 39 64 91
66 3 16 43 55 97 100
64 2 33 53 63 107 115
62 5 42 86 64 86 114
60 18 .47 86 111 153 157
58 4 44 77 116 179- 150
56 23 54 109 111_ 156 1221
54 31 70 114~ 164 164--121..
52 20 59- 91 151 138 101
50 20 90 91 107 '117 ... 80
48 38 82 94 ....79... 64 45
46 33 69 101 66 52 24
44 33 55 ......77.. 43 23 23
42 31 45 50 41 16 12
40 33 32 35 13 9 5
38 19 13 20 12 10 3
36 12 17 9 8 3 2
34 ..-....18..... 5 10 6 1 2
32 11 4 5 1
30 6 5 7
26 2 2 1
24 2 1
22- 3 3 1
Total 370 853 1298 1359 1661 1561 27 54 74 68 79 74 Total
Q3 .53,5 .59 5 61 9 61 8 65.4 67.9
MId 46.8 .521 55.0 55.9 .58.9 61.2
Q1 41.1 46.9 48.2 .51 3 .53.7 55,5
Range 20-75 17-90 19-91 33-88 34-93 34-97
Schools 30 .54 65 58 .59 54
47 2 .527 .55.7 .55.6 .58.7 61 6 Md
40,.5 45.0 43.0 46..5 48.0 45 0 Range
.57 0 67.0 66.0 68.0 69.0 73 0
led-of-year public-school aediss C used in Gredis 7, 8, 8; C in Grades 10 11, 12
C used In Grad3s 7, 8, 9; C2 in Grades 10 11, 12
............ End-of-year public-school median
1 3 4 68
2 1 1 6 10 66
2 1 5 8 64
1 6 1 6 10 62
2 7 6 11 9 60
1 12 13 13-ld 12 .58
1 4 9.-d 8d 17 9 .56
9 12 16 10 ......5 ...... 54
3 9--d 9 12 3 52
3 7 7 4 --4-" 1 50
4 9 .5 2..... 1 2 48
6 5 3 3 46
2 3 ...2.. 1 44
2 1 42
FIGURE 2 1942
ACQUA I NTANCE LINCOLN DIAGNOSTIC
Cooperative Test-Form Q Form 2
GRADE 9 10 II 12 8 9 10 II 12
+1.5Sa-95e -241S Se +
-92 -234 -92
-90 -134A -924 -234
-87 -121 -134A -87
+ I o -84 -84 +Io
-3 -108 -34
-81 -226 34 -81
-77 132 -108 -35 234 77
-75 --------2134 i15E------- ---
-73 134A -121 24 21 -126 -73
-241 -41G ,-239
+ .5 --69 -230 -203 -42 -126.* '42 5 -69
-226 -239 35
-65 -203 -23A 20 -126 52 -65
-103 239 B 132 55
: .45*241S -2328 .* .-4116B- -1 -
:60 232 -50A -226Y -239 16A 60
S* -103* *.243 -115V i34B-46 5 35 -203 c 55
0 2 2328 123 239 213 23412 -58
-50 1 34B - - - -203-- -108 - 16A -- 50 0
1239 108 -109 215 -55 -5811 2026
w A tc 528o"115 -202 -33 202 -45
215 23 A 13 -115W 203 239
-40 IIg2 241G 58
-40 215 46 -204 -3C -204 40
-234 134 2l -33C :
-35 -222 L67 IIA 23B 33C -116A 35
-M1 5 -204 6 31 5-
1-.5- 1 -226R" -115B o_012 -1 16C 116c
-27 -115 -46 -27
-25 ------115M- iii --------- ------------------37-- -25
-23 23 3 -50B -204 -I16B -23
-i -213 203 134 -330 -19
-i -16 23 11 -222 3G -37 -33G -16 -
-23B -213 -33G -42A -33G
-13 -37 -33G 37 -13
-10 -II S -23B -204S -0
8 -33G -8
- 1.5o- -6.5 -65 -1.50
-5 -33G -33G -37B -5
*ABOVE GRADE 'BELOW GRADE
PUBLIC-SCHOOL END-OF-YEAR MEDIANS TRANSMUTED INTO INDEPENDENT-SCHOOL
.... SCHOOL 239
- 18 -
of several bulletins and articles dealing specifically with the uses of tests and
test results Among these bulletins are No. 18, The Use of Test Results in
Diagnosis and Instruction in the Tool Subjects, No. 23, The Use of Tests and
Rating Devices in the Appraisal of Personality; No 25, The Use of Test Results
in Secondary Schools; and No. 34, The Nature and Use of Reading Tests. Bulletins
Nos 18 and 23 were recently revised and reprinted. The Bureau has also issued
in multigraph form, a "Primer" on measurement entitled, Introduction to Testing
and the Use of Test Results The member schools are supplied with all bulletins
published by the Bureau.
Because of the wide-spread interest in remedial reading, and also because
of the close relationship between reading ability and the whole guidance problem,
the Bureau has attempted to keep its member schools informed concerning research
in the reading field After issuing a series of mimeographed bibliographies on
reading over a period of years, the Bureau published in 1942, a bulletin entitled,
Ten Years of Research in Reading, which summarized reading studies during the
Research -The files of the Educational Records Bureau contain data for an
almost unlimited number of research studies, and a certain amount of research is
carried on by the members of the staff notwithstanding their heavy service load
A considerable number of studies involving the use of some of the newer tests
have been published in the bulletins summarizing the results of the fall and
spring testing programs.1 Members of the staff have also contributed articles
1Bulletin 19- "Comparable Tests and School Marks;," Bulletin 20: "The Interpretation and Use of Scaled
Scores," Bulletin 22: *A Study of the California Test of Mental Maturity:; Bulletin 24: 'The Relation
between Speed and Level of Literary Comprehension;- Bulletin 26 'The Reliability and Validity of the
American Council Psychological Examination,' "A Study of the Reaults of Reading Tests Administered a
Year Apart; "Soae Correlation Date on the California Test of Mental Maturity;' Bulletin 27:-"A Cor-
relation Study of the Cooperative Mathematics Test for Grades 7, 8, and 9;, -A Report on the Thurstone
Test for Primary Mental Abilities;' *A Study of the Junior Scholastic Aptitude Test," Bulletin 29
"The Reliability of the American Council on Education Psychological Examination, 1939 Edition;' "A
Study of the New Edition of the Iowa Silent Reading Test for Bigh Schools and Colleges: Bulletin 30
"The Cooperative English Test, Fore Q Correlations with School Marks and Intercorrelatlons, "-Group
Scales versus Occupational Scales for the Strong Vocational Interest Blank;" "The Academic Aptitude
of French and Latin Classes in Six Independent Schools, Bulletin 31: 'Correlation between IQ s an
the Nen Edition of the Kuhblann-Andereon Test and Binet IQ-s, "A Study of the Van Wagenen-Dvorek
Diagnostic Examination of Silent Reading Abilities;; Bulletin 33 "Some Data on the Difficulty and
Validity of the Cooperative Tests in Biology, Chealstry, and physics, Form ETB-R;* "The American
- 19 -
based on Bureau data to educational journals 2 Reprints of some of these have
been obtained for distribution to member schools.
Committee reports.-The Committee on School and College Relations, with
Dr. Eugene R Smith as chairman, has issued four reports surveying entrance
requirements of colleges These reports were published in 1932, 1933, 1935,
and 1943 The last report, entitled Fourth Report of the Committee on School
and College Relations of the Educational Records Bureau, contains a detailed
analysis and summary of the replies given by four hundred colleges to the
recommendations of the committee. It is believed that this 'is one of the most
important reports ever made available on the relations between secondary and
Council Peychological Exsaination Intercorrelatlone of Scoree on the 1938, 1939, and 1940 College
Freshman Editions;' Bulletin 35. -Some Data on the Reliability and Validity of the Cooperative Teet
of Social Studies Abilities,. -Intelligence Quotients Derived from the Amarican Council Psychological
Examination, College Freehban Edition;* *A Study of the Revised Edition -* the Stanford Achievement
Teet.,* Bulletin 38: *Goma Data on the Results of the Cooperative Tests in French, Latin, and Secondary
School Mathematics, Form S;' Bulletin 37, ,The Speed of Rending of r.plls in Independent Secondary
Schoole' 'The Reliability and Validity of the Seavis-Brealich Arithmetic Teste,* Bulletin 38: 'A
Study of the Lincoln Diuenoutlc Spelling Test;''Soes Data on the Cooperative Amerlcan Bietory Test.
2*Correlation of Achievement Tests and School Marke,t School Review (December, 1937); *The Relationship
between the Length and the Reliability of a Test of Rate of Reading,* Journal Of Educational Research
(September, 1938); *A Comulatlve Record Form for the Elesentary School,' Elementary School Journal
(Septesber, 1939); 'An experimental Study of a New Mathea&tics Test for Grades 7, 8, and 9,' Mathematics
teacher (October, 193i); 'The Correlation between Achievement Scores and School Marks In an Independent
School for Boya, Journal of Applied Psychology (February, 1940); 'One Reading Test Serves the Purpose,.
The Clearing Rouse (asrch, 1840); 'What is a Satisfactory IQ for Admislaalon to College?y School and
Society (April, 1940), 'The Reliability and Vslidity of the American Council Psychological Exaslnetion,
1938 Edition,' Journal of Educational Research (October, 1040); 'Sooe Data on the Kuder Preference
Record, educational and Psychological Measurement (July, 1941); *Cumulative Teet Records: Their Nature
and Uses," Educational and Psychological Measurement (October, 1941); -The Reliability of the Bell
Inventories and Their Correlation with Teacher Judgment,. Journal of Applied Psychology (December, 1941),
"A Note on the Wrenn Study Habits Inventory,- Journal of Genetic Psychology (1842), -Effect of Type of
Desk on Resulte of Machine-Scored Teste,' School and Society (September, 1942), 'The Independent School
and Education Teeterdoy, Todey, and Tomorrow,* School and Society (November C2, 1942); 'Relationship
of Elementar -School AchieveMset Tests to Achievmen. t Tests Teken in the Secondary School, Journal of
Educational Research (November, 1842); *81Spllfled Method for Scoring the Strong Vocational Interest
Blank Applied to a Secondary-School Croup,' Journal of Educational Psychology (November, 1942); 'Selping
the Individual Pupil Adjust to the Probleme of War and Peace,* School Review (Noveaber, 1842); 'ProgressIve
Methods :e Related to Knowledge of Amerlcan 81ecory,. School and Society (1ay 29, 1943).
For years various leaders in schools and colleges have maintained that
the most fundamental instrument in a long-time guidance program is an in-
dividual cumulative record of objective measurement data and other information.
They have pointed out that a cumulative record makes it possible to study the
status and growth of each pupil and to adjust his program in accordance with
all available information about him, and that it should also contribute in an
important way to the transfer of the pupil to other schools and should serve
in college admission and placement The development, about 1930, of standard
cumulative record forms for secondary schools and colleges, by Dr. Ben D Wood
and a committee working under the auspices of the American Council on Education,
was one of the most important steps in spreading the practice of maintaining
cumulative records In 1933, a special adaptation of the American Council
cumulative record form was prepared for independent elementary and secondary
schools by Eleanor Perry Wood and Winston B. Stephens of the Educational Records
Bureau. A sample copy of the form is shown on pages 21 and 22. In this par-
ticular record, only the test portion of the form has been filled out
Many of the member schools of the Educational Records Bureau now keep
cumulative records of their test results Some schools maintain the cards
locally; others send them to the Bureau to have the test portion of the card,
including the graphic record, filled out, after which the records are returned
to the school. When the pupil applies for admission to another school or to
college, the Bureau, on request from the pupils present school, sends a copy
of the pupil's cumulative record to the institution to which he is going and
accompanies the record with an interpretative letter
In 1941, the original American Council cumulative record folder for secondary
schools was revised by a committee under the chairmanship of Eugene R. Smith.
The committee is now revising the college form and the one for elementary school
TENTT -I SJ%.I. I TE.T
fi''"' ,,\A.. TI | ti^-"/>/ AT? C'
z+ s r -r Tx"T ]r '.4 II
- :... S'I.C
+ ; I
CUMULATIVE REC S
4 I iI I
-EST S I TEST ETTA I ,1
^.5. --c,/ /'A. | o|T
_ L r _L
St-AT St n I-$ .
I, Z1/ I ,
I, t. I r DI 1 +l i "1 ul
.1~~~ j -' C
EDUCATIONAL RECORDS BUREAU
NAME "/o,, -A' n .
- I l 4 II
TEST ST0"E 1 TEA
C<'^-/-*.rc.a T, // -*.
|w /s _a3
I ~-. j
437 WEST 591- STREET, NEW YORK Nv
Sz TEST T ST a
-zZ C aeef.6"^. ~ ~ 7
..&t&aCSA ____ U,' & ^r
5 fl.A 4C
-dZ_ 'gi, t._ Z izs Z- ^.iZ .^^aM^H ar.
Pt ATC^~ SCf^ W t- ^__.' .2 .S ^a.LJCAA flLT-g
a kl c 6
... . . .
M. F, BIRTHOATE-PLACE ENL HEALTH RELIGION RACERNAT'L'TY OCCUPATION ADDRESS
STEP PARENT RES
on GUARDIAN BUS.
LANGUAGE SPOKEN IN HOME TYPE OF COMMUNITY IF PARENTS SEPARATED
BEFORE 10- AFTER 10- BEFORE 10- AFTER 10- GIVE DATE
YEAR ANO AGE
ATTENDANCE A T A T A T A T A T A
The Bureau has been active in bringing the new revision of the American Council
forms1 to the attention of high schools throughout the country.
There are, of course, many types of cumulative record cards in addition to
those published by the American Council on Education and the Educational Records
Bureau, and there is an almost endless 'variety of other record and report forms.
Schools have, in recent years, become conscious of the fact that their educational
theory and practice have oftentimes run far ahead of their antiquated methods of
recording and many of them have recently been studying the problem of record
keeping and have produced new experimental record forms. Different organizations,
such as the National Education Association and the Progressive Education
Association, have been very active In making and trying out new records. Because
of the multiplicity of the records, it has appeared desirable to have a central
agency for the collection and subsequent distribution of those that seem most
promising. Some years ago, the Educational Records Bureau therefore began to
collect samples of record forms from various school systems. The Bureau was able
to make up several duplicate sets of carefully selected record and report forms,
which it has loaned to schools for a period of one month, at a nominal charge to
cover mailing costs. These sets are available to non-member as well as member
institutions. At least 200 school systems and colleges have taken advantage of
this loan service.
Services to Public Schools
Partly for financial reasons, testing programs have developed much more
slowly among public schools than among independent schools. Another reason for
the difference in'the rate of development of testing in the two types of in-
stitution is that the independent schools, by their very nature, have long been
concerned with the individual pupil, while many public schools, being much larger,
have necessarily directed their attention to groups and have not until recently
'Published by the Azerican Council on Education, 744 Jackson Place, Washington, D.C.
- 24 -
become concerned with the study and guidance of the individual student.
Because of the fact that the individual is so easily lost in the large
groups of pupils who pass through our urban public schools, the value of
cumulative records based on objective measurements is potentially even greater
in public schools than in private schools. There has been a need to call the
attention of public schools to the values of cumulative comparable measurement
and to demonstrate what public schools can do when they attack this problem
intelligently, The main contribution of the Educational Records Bureau to
this problem has been the Public School Demonstration Project, which was begun
in 1934 and was carried on for a period of five years under a special grant
from the Carnegie Foundation for the Advancement of Teaching. The major features
of the project were comparable tests, cumulative records, a program of teacher
education, and a series of case studies of individual pupils.
A report on the entire project entitled Guidance in Public Secondary Schools
was published in 1939. Among the chapters in the report were, "Five Years of
Guidance in the Plainfield High School," "The Cumulative Record and Its Uses,"
"Developing a Cumulative Record Card for Local Use," "An Experiment in Marking
and Reporting," "In-Service Education of Teachers in Guidance," "How a Testing
Program Contributes to Guidance," "An Experiment with Anecdotal Records," and
"Contributions to Guidance Through Case Studies." It is believed that the work
of the seven demonstration centers, which cooperated in the project has in-
fluenced measurement and guidance in many other schools.
The Bureau provides a scoring service for public schools which is utilized
not only by the member public schools, but also by a considerable number of
public schools which are not members of the Bureau. Two types of machine-
scoring service have been developed for public schools One type, known as
Plan A, includes not only scoring but also distributions and class lists pre-
senting the results and a letter of explanation and interpretation. The other
type, called Plan B, includes scores and distributions of scores only and is
considerably less expensive than the first plan.
Services to Colleges
The Educational Records Bureau has no regular testing program for colleges,
but it does a large amount of scoring, particularly by machine, for certain
colleges in connection with the local testing programs of these institutions.
Much of this work is concerned with the scoring of tests which the colleges
administer to entering freshmen.
The Bureau also helps secondary schools plan and carry out entrance testing
programs and scholarship tests for candidates for admission. Extensive scoring
and interpretative services are provided in connection with some of these
entrance and scholarship tests.
Cooperation with other Examining Agencies
If the many difficult problems of measurement are to be met successfully,
it is imperative that there be close cooperation among all examining and
guidance agencies The Educational Records Bureau has been very fortunate in
that its relations with all other agencies engaged in similar work have, from
the beginning, been very cordial. The Bureau's relationship with the Cooperative
Test Service has been especially close. Since 1933, these two organizations have
been working under one director, Dr. Ben D. Wood, in such close harmony that even
persons well acquainted with the activities of both groups have frequently been
under the mistaken impression that they were one and the same organization, For
more than ten years, the tests produced by the Cooperative Test Service have
annually formed the basis for the measurement of achievement in the secondary
schools holding membership in the Bureau. There are several other ways in which
the Bureau has engaged in activities along with other organizations. These in-
clude educational conferences, test construction, and the use of new measuring
Educational Conferences.-For ten successive years, a series of educational
conferences was conducted each fall in New York City under the auspices of the
Committee on Personnel Methods and on Educational Testing of the American
Council on Education, the Commission on the Relation of School and College of
the Progressive Education Association, the Cooperative Test Service, and the
Educational Records Bureau. The high quality of the programs planned for these
conferences has attracted educational leaders from schools and colleges through-
out the country. The published proceedings of this series of conferences
represent some outstanding contributions to contemporary thought on educational
problems in the United States. The conditions created by the war, especially
those pertaining to travel, have necessitated the temporary abandonment of the
annual conference, but it is planned to resume this type of activity as soon as
Test Construction.-Although the Educational Records Bureau is not a test-
making agency and while there is no intention of setting up an elaborate
organization for the production of tests, the Bureau is in an especially favor-
able position to contribute.to the evaluation and construction of tests through
cooperation with such test-making organizations as the Cooperative Test Service.
Leaders in the testing movement have pointed out that the ultimate hope in
achievement-test construction is to have the tests represent the combined
thinking and work of subject-matter specialists and test technicians. Member
schools of the Bureau can contribute significantly to the making of tests through
stating their objectives, indicating the areas in which the tests should be made,
and appraising tests after they are developed in an effort to meet the needs of
these schools. Three forms of tests in each of the following fields have been
constructed by committees of teachers appointed for this purpose: mathematics
for grades 7, 8, and 9, biology, chemistry, and physics. All of these tests have
been published by the Cooperative Test Service. A Bureau committee on social
studies tests has worked with members of the Cooperative Test Service staff in
the preparation of two forms of a general achievement test in the social studies
Another Bureau committee has participated in the preparation of a secondary
school mathematics test, which was published by the Cooperative Test Service
It is hoped that this kind of cooperative effort may eventually be extended to
all subject fields
Use of New Measurement Devwces.-The Bureau can render service to its own
members and to test-making bodies through helping to spread the use of new
measuring instruments that are especially promising. This type of service is
illustrated by cooperation between the Bureau and the Secondary Education Board
in the distribution and use of the Secondary Education Board Junior Scholastic
Aptitude Test. The various forms of the Junior Scholastic Aptitude Test are pre-
pared by the Bureau of Research of the Secondary Education Board and the
Educational Record Bureau serves as the sole distributing and scoring agency for
A somewhat similar endeavor, which the Bureau has undertaken very recently,
is the distribution and scoring of the Yale Aptitude Tests, prepared by Dr A B
Crawford of the Department of Personnel Study of Yale University. The schools
should ultimately benefit from such cooperation through enlarged services and
improved measuring instruments.
The Educational Records Bureau provides an advisory service in testing,
record keeping, and guidance mainly for the benefit of its member schools, but
representatives of schools not holding membership in the Bureau frequently call
on this organization for advice both by correspondence and in personal interviews
In consonance with its desire to be of the largest possible educational service,
the Bureau meets these requests for information and advice as far as its re-
sources will permit Consequently, hundreds of advisory letters are written to
non-members each year and many hours are spent in interviews with representatives
of schools outside the Bureau membership. Since these advisory services do not
directly benefit the member schools of the Bureau, they may be regarded as a
contribution made to education generally by the institutions to which the
Bureau owes its existence.
Services to Governmental Agencies
U.S. Office of:Indian Education.-For several years the Educational Records
Bureau has distributed, scored, and reported the results of tests administered
to Indian students who were applicants for government scholarship loans to
enable them to attend college. Percentile norms have been prepared for these
tests on the basis of the scores of the Indian students. At the request of the
Office of Indian Education, the Bureau recently made a study on the value of
test scores, school marks, and other criteria for the prediction of success of
Indian students in college.
U.S. Merchant Marine Academies.-In the spring of 1942, the Bureau helped
to Plan a testing program for entering cadets in three academies of the U.S.
Merchant Marine. For more than a year, the Bureau supplied, scored, and re-
ported the results of four tests employed in that program with each new cadet.
The aptitude and achievement of several thousand cadets were evaluated in this
Qualifying Test for the Army and Navy College Training Program.-The
Educational Records Bureau is serving as one of the regional offices for the
Qualifying Test for the Army and Navy College Training Program. Eleven civilian
agencies are cooperating in that program under instructions from the armed
services. Ten regional offices are working under the direction of the national
office, which is located at the College Entrance Examination Board. The region
served by the Educational Records Bureau is New York State and the Ne 'Ingland
States. Nearly two thousand schools in this region cooperated in the Qualifying
Test which was administrated on April 2, 1943, and about sixty-two thousand
boys in this area took the test. The tests were processed by the regional
office and the results reported to the Army and Navy in accordance with direc-
All organizations participating in the Qualifying Test for the Army and
Navy College Training Program are working on a strictly cost basis and not
one cent of profit is being derived by any of these agencies. The staff members
of the Educational Records Bureau, like those of the other participating organi-
zations, welcome especially this opportunity to make this small .contribution to
the war effort.
In summary, one may say that the service functions of the Educational
Records Bureau are well established and are at present the most important of
all its functions. Both because of financial limitations and of emphasis on the
service phase, the research aspect of the Bureau is not highly developed,
although much interest in research is present, as is shown by the production of
studies of limited scope whenever time has permitted. The remarkable expansion
of the activities of this organization, within a period of approximately fifteen
years, furnishes evidence of the effectiveness of the Bureau's work and the
loyalty of its member schools. Moreover, it is a testimonial to the soundness
of the fundamental idea on which the Bureau is founded the promotion of the
education and adjustment of individuals through a cooperative and continuous
program of comparable measurements, interpreted from the viewpoint of a broad
educational philosophy and in the light of a comprehensive cumulative record
for very pupil.