AN ABSTRACT OF THE THESIS OF
Master of Arts
for the degree of
Betty Lou Carman
Interdisciplinary Studies in English, Reading
in
August 12, 1977
presented on
and Psychology
Title: AN ATTEMPTED STANDARDIZATION OF READING AND
WRITING SKILLS TEST USED AT OREGON STATE
UNIVERSITY'S SKILLS CENTER
Abstract approved:
Redacted for Privacy
Dr. Faith Norris
The objective of the investigator in carrying out this
of
study is: to evaluate the reliability and validity
ENGLISH
SKILLS
TEST-REVISED
(EST-R);
the
to construct local
norms based on a small sample of subjects from Oregon State
University's
WR. 40X; to investigate the statistical terms
and methods used in educational test standardizations;
review
standardized
tests
English programs; to probe
communication
for placement in college
used
the
to
general
issue
campus
of
skills centers; and to study the application
of reading instruction in remedial English programs.
The EST-R is a revision of the test currently used
Oregon
several
departments
at
English
proficiency
test.
Communication
Skills
composition classes.
Reading,
Center
It
State
is
for
University
also
used
placement
The test consists of
five
as
by
in
by
an
the
English
subtests:
Spelling, Grammatical Usage, Vocabulary and Essay
Correction.
This study outlines administration and scoring
procedures and provides information regarding assessment of
EST-R scores.
The
standardization
is
based
on
EST-R
three
as a pre-test to 33 WR. 40X students;
administrations:
as
a post-test to 16 WR. 40X students; and as a pre-test to 45
WR. 230
Test scores, frequencies, mean, median,
students.
given
standard deviation and range for each group are
for
the Reading Test and for the combined Spelling, Grammatical
Usage, Vocabulary and Essay Correction
scores
test.
Transformed
(percentile rank, z- and Z-scores) are provided for
the Wr. 40X Reading pre-test and combined pre-test.
The Kuder-Richardson 20 (K-R 20)
subtest
reliability
for the tested groups ranges from -0.059 to 0.891.
The K-R
Grammatical
Usage,
20
reliability
of
the
Spelling,
Vocabulary and Essary Correction tests combined ranges from
0.750 to 0.849.
Criterion-related validity was explored using validity
coefficients and group separation.
EST-R is a valid
instruction.
indicator
of
the
This data indicates the
effects
of
WR.
40X
AN ATTEMPTED STANDARDIZATION OF
A READING AND WRITING SKILLS TEST
USED AT OREGON STATE UNIVERSITY'S
COMMUNICATION SKILLS CENTER
by
Betty Lou Carman
A THESIS
submitted to
Oregon State University
in partial fulfillment of
the requirements for the
degree of
Master of Arts
Completed August 12, 1977
Commencement June 1978
APPROVED:
Redacted for Privacy
Professor of English
Redacted for Privacy
head of Department English
Redacted for Privacy
bean of Graduate School
Date thesis is presented
August 12, 1977
Typed by BLou and George Carman for Betty Lou Carman
ACKNOWLEDGEMENTS
This thesis represents a focusing of both interest and
energy for one who was once without an organizing direction
many
So
me
in
have supported my explorations in so
My thanks to Dr. Faith Norris who,
many ways.
saw
people
potential
a
did
I
not
years
ago,
see in myself.
example, her valuable friendship, and her assistance,
Her
have
been, and are, exceedingly important to me.
Thanks,
too,
Ted
Dr.
to
Madden
whose
patience,
generosity and expert guidance have been essential to
study
this
and whose approach to teaching the difficult subject
of psychometrics has
greatly
benefited
me
and
many
so
others.
My
appreciation,
too,
to Dr. Ken Ahrendt who,
goes
back when I was an undergraduate,
impressed
upon
me
the
responsibility that I, as an English teacher, have to build
reading
students'
invaluable,
as
skills.
has
his
His
assistance
classes
finding
in
been
have
source
material for this study.
Tim Perkins is another person to whom I
owe
great
a
deal; he gave me the opportunity to try myself as a teacher
and offered his encouragement and support
He
and
all
this
to
study.
my other friends--students and co-workers--at
the Communication Skills Center have
shared
so
contributed so to my confidence and competence.
much
and
My
sister, and best friend, Alice Ann Daily, has been
a vital source of warmth, support and commiseration.
I
am
looking forward to helping her get through her own master's
program.
And my amazing
contributed
and
wonderful
George,
has
to my life and to the present endeavor in more
ways than I could possibly name.
his
husband,
His
unflagging
support,
love, and his brilliance have given me great strength.
The many hours he spent teaching me and helping me to use a
computer for text editing and layout have added much to the
quality (and decreased the painfulness of
the text of this study.
preparation)
I am so very fortunate.
of
TABLE OF CONTENTS
I.
II.
III.
IV.
INTRODUCTION
1
REVIEW OF LITERATURE
A.
Skills Centers
B.
Reading in Remedial English Programs
C.
College-Level English Placement Tests
D.
Statistical Terms and Methods
5
5
9
13
18
THE ENGLISH SKILLS TEST-REVISED
A.
Administration
B.
Scoring
1.
Reading
a.
Speed Score
b.
Comprehension Score
c.
Evaluation
2.
Scoring Spelling, Grammatical
Usage, Vocabulary, and Essay
Correction
C.
Description of Test Results
D.
Standardization
1.
Validity
a.
Validity Coefficient
b.
Group Separation
2.
Reliability
3.
Norms
31
38
38
38
39
42
47
SUMMARY
50
BIBLIOGRAPHY
52
APPENDIX
55
24
24
27
27
27
28
28
29
LIST OF TABLES
Table
I.
II.
III.
IV.
V.
VI.
VII.
VIII.
IX.
X.
XI.
XII.
XIII.
XIV.
XV.
Page
Reading Test Speed Chart
27
WR. 40X Reading Pretest
32
WR. 40X Pre-test
33
WR. 40X Reading Post-test
34
WR. 40X Post-test
35
WR. 230 Reading Pre-test
36
WR. 230 Pre-test
37
Validity Coefficients (r)
41
Group Separation
41
Reliability
46
Score Transformation Table,
WR. 40X Pre-test: Reading
48
Score Transformation Table,
WR. 40X Pre-test: Spelling,
Grammatical Usage, Vocabulary, and
Essay Correction Combined
49
EST-R Test Scores, WR. 40X Pre-test
55
ESR-R Test Scores, WR. 40X Post-test
56
EST-R Test Scores, WR. 230 Pre-test
57
AN ATTEMPTED STANDARDIZATION OF
A READING AND WRITING SKILLS TEST
USED AT OREGON STATE UNIVERSITY'S
COMMUNICATION SKILLS CENTER
INTRODUCTION
I.
One of the developments in higher education during the
last two decades is the recognition that not
students
are accepted into universities are adequately prepared
who
to survive the challenge.
institutions
variety
prompted
This dilemma has
many
to set up learning centers to serve the needs
of their open-enrollment constituency.
a
all
services,
of
including
These usually offer
diagnosis
of
skill
weaknesses, auto-instructional skill lessons, tutoring, and
placement counseling.
There
facilities:
which
those
are
major
two
subdivisions
those which give mathamatics
offer
of
learning
instruction
English/study skills assistance.
and
Both
mathamatics and language skills centers are common, but, it
likely
seems
that,
because every student must be able to
read, write and study (and
college
students
seem
because
to
English
skills
among
be declining more rapidly than
mathamatics skills), English skills centers may be found on
more
community
college
and
university campuses than can
mathamatics centers.
This study
is
an
outgrowth
of
the
investigator's
two-year experience as a graduate teaching assistant at one
Page 2
such
learning
Communication
Oregon
center,
University's
State
Skills Center (CSC). The CSC offers a number
of services to students including: study skills lectures to
groups
living
well
as
individual
as
tutoring;
individualized reading instruction by graduate students
writing
reading;
self-instructional
punctuation,
help
with
WR. 40X,
packages
a
papers,
theses,
remedial
English
supplemental instruction prior
composition
(WR. 121).
and
usage,
individual
reports;
and
class for students who need
to
enrolling
Center
The
classes;
grammar,
for
and paper organization;
spelling
resumes,
vocabulary
tutoring;
in
also
freshman
in
offers
to the
university an English proficiency test (the ENGLISH
SKILLS
TEST).
Several schools, including Forestry, Health, P.E.,
and Agriculture, require that all their students pass
this
examination before they may graduate.
The
ENGLISH
SKILLS TEST (EST) is also used as a tool
to determine placement in some
English
classes.
At
the
beginning of the term, each freshman composition student is
given the test, and, if
work
is
appropriate,
instructor
the
the
decides
remedial
student is then transferred to
WR. 40X.
The EST is also
effectiveness
of
used
WR. 40X
to
help
the
instruction.
CSC
assess
the
Students take the
test at the beginning and the end of the course and
scores
Page 3
Since one objective of 40X instruction is to
are compared.
enable
raise the student's skills to a level sufficient to
him
pass
to
the
test
pass the post
improvement
EST, it is hoped that each student will
very
figure into the student's 40X grade;
an
some
whether
receives
he
N is dependent solely on his performance on
or
P,
I,
show
least,
The test score does not, however,
score.
in
the
at
or,
regular 40X coursework.
The English Skills Test was developed by the staff
the
CSC
with
been
Department and has
consists
and
Each of these
independent
the
subtest
interpreted as being broadly diagnostic.
a
student's
on
any
subtest,
in
that
if
non-compensatory;
passing
level
skill-deficient
It
1976.
fall,
Grammatical. Usage, Spelling,
Vocabulary and Essay Correction.
separately,
since
use
in
subtests:
four
of
of members of the English
cooperation
the
of
area,
The
is
scored
are
scores
scoring
is
score falls below the
he
is
considered
and some sort of remedial
work is required--depending upon the particular
use
being
made of the EST.
Because
new,
no
is
standardization
information
validity
test
the
has
questioned.
is
available
not
These
been
locally written and relatively
has
been
performed.
Little
regarding reliability, and test
systematically
shortcomings
have
and
approached
proved
to
be
Page 4
disadvantageous in the investigator's work, both
of
tutor
skills and as 40X teacher at the Communication
writing
The purpose of this study, therefore, is to
Skills Center.
provide
as
first step toward standardization of the EST,
the
i.e., to analyze the test and
construct
set
a
of
local
norms.
conducting
In
revision
major
of
reading
ten-minute
study, the investigator made one
this
previous
the
was
subtest
form
of
the
skill.
Lime 1imiL, the vocabulary Lest was shortened (from
10
items)
In
the test within the 50-minute class period
keep
to
a
added in the belief that
reading skills are an integral part of language
order
EST;
15
to
as was the spelling test (from 24 to 10 items).
The format of the latter was also changed, however, so that
40
spelling
words
could
be
tested rather than just 24.
original
Hundreds of students have already taken the
EST,
so the questions used on both vocabulary and spelling tests
(These are the easiest to memorize and
were changed.
pass
on Lo a friend.) The Grammatical Usage and Essay Correction
sections were left unchanged.
The testing population used for this
f
two
groups:
the
and
a
Writing
consisted
spring 1977 WR. 40X class to whom a
pre-test and post-test of
given,
study
the
230
revised
EST
(EST-R)
were
class, i.e., students who had
successfully completed regular freshman composition.
Page 5
REVIEW OF LITERATURE
II.
Skills Centers
A.
Most
skills
language
seem
centers
response to the fact that more students than
institutional
ever before are entering post-secondary
without
education
There
the English skills prerequisite to academic success.
is a great deal of disagreement regarding
the
decline
students;
an
be
to
of
the
reason
for
language proficiency among post-secondary
quality
some complain of a deterioration in the
of English instruction at the grade and high school levels,
attribute
while others
open-enrollment
the
skills
lower
standard
standard
dialects
English.
demands
It
to
seems
read,
write
and
likely
that
both
as
well
ineffective skill instruction and open enrollment
as
other
other
English to college campuses where they have
much difficulty meeting the
speak
the
to
policy adopted by many universities.) Such
policies, it is argued, entice speakers of
than
range
factors
(e.g.,
the
rise
in
popularity
television viewing at the expense of recreational
of
reading)
contribute to the problem.
KirT, "Notes on Remedial English at the
'Arthur H.
Higher Education Level, ' ENGLISH LANGUAGE TEACHER, XXVII
(June, 1973), 245.
Page 6
Although
goal of every skill center is to assist
the
students requiring remedial
English
there
training,
are
inter-campus differences in the constituent teaching
great
staff, the philosophical organization,
materials
the
and
methods used to meet this goal.
Some
learning
centers are staffed by regular members
of the school's English department;
as
Arthur
does
are
adequately
enthusiastic
H.
2
many
however,
argue,
King, that few university instructors
trained
deal
to
with,
sufficiently
or
about coping with, the level of remedial work
required by many students.
3
One alternative is
to
hire
a
team of specialists, but this incurs higher costs than most
colleges can sustain.
problem
this
is
A
the
more
commonplace
employment
Oregon
(e.g.,
through teaching
but
State
University)
assistantships
student
of
supervised by one or more specialists.
resolution
In
most
of
teachers
colleges,
student tutors are paid
or
work-study
programs;
some schools, like Miami-Dade Community College 4 staff
their
centers
2
Course,"
with
supervised
student
volunteers
who
Ada
Y.
Hatch,
"Reading-Centered Composition
JUNIOR COLLEGE JOURNAL,
XXVI
(March,
1956),
395-99.
3King, "Notes on Remedial English," pp. 245-50.
4
Grace W.
Welch, "Organizing a Reading and Writing
Laboratory in which Students Teach,
COLLEGE COMPOSITION
AND COMMUNICATION, XXV (December, 1974), 437-39.
Page 7
receive academic
credit rather than remuneration for their
work.
The employment of peer tutors makes
than
merely
a
standpoint.
fiscal
much
Since
instruction consists of individualized
from
sense
learning
more
of
the
packages,
the
demands
of
teaching are limited to tasks well within
the
ability
of
most
An
tutors.
extra
advantage
of
student-operated
centers
is that such facilities are much
more
provide
the
likely
to
relaxed,
non-threatening
atmosphere so essential to attracting remedial students.
Centers
on
different campuses also differ in the way
students enter remedial
voluntary,
drop-in
instruction.
programs
6
Some
while
constituency from referrals by English
referrals
may
come
as
a
result
testing or informal diagnosis.
with
5
others
run
strictly
draw
professors.
7
their
These
of classroom placement
Most educators would
agree
Arthur King that "English is not simply a skill to be
learned in English periods but
5
Welch,
Laboratory," p.
6
7
"Organizing
a
a
means
of
Reading
teaching
and
and
Writing
438.
Ibid., p. 43.
John W. Gregory,
"An Approach to
Functional
English in a Four-Year Junior College," JUNIOR COLLEGE
JOURNAL XXIX (December, 1958), 203-5.
Page 8
learning
subjects;"8
other
centers, however,
departments
other
responsibility
for
are
only
of many learning
the
staffs
now
successfully
than
English
that
the
English
skill
convincing
they,
too,
share
of
their
levels
students.
addition
In
to
inter-campus
differences
in
procurement of learning center students, there are
also
variety of ways in which instruction is carried out.
centers group students into
regularly
are
remedial
usually
supplemented
by
a
A few
which
classes
and are traditionally structured.
the
meet
These classes
individualized
(usually
self-instructional) skill lessons in spelling, agreement or
9
other areas of
particular
organized
for
strictly
packages.
Some facilities, like Oregon State
weakness.
Other
centers
are
individualized tutoring or lesson
University's
Communication Skills Center, offer both remedial coursework
for those with broad-based skill deficiencies and
or
tutoring
individualized instruction packages for those with more
specific handicaps.
8King, "Notes on Rc-itedial English," p. 245.
9
205.
Gregory, "An Approach to
Functional
English,
Page 9
Reading in Remedial English Programs
B.
Although most educators would agree that reading is
skill, some colleges still do not include reading
language
In
instruction in their remedial language skills centers.
instances
some
a
the
reason
for this exclusion may be the
added expense of instructional materials and personnel.
In
cases, however, reading is not taught on campuses for
many
years
i.e.,
ago;
reading
is
appropriate college subject.
and
recognize,
not
It has
by many as an
deemed
taken
twenty
there
the same reason that spelling was not taught
some
time
answer to, the needs for remedial grammar,
punctuation and spelling at the college level, and
take
the
to
it
may
best arguments of reading proponents, as well as
the evidence offered by existing programs, to convince some
academicians
to
include
reading
instruction in remedial
English.
One remedial program incorporating reading and writing
skills
instruction is that offered at Miami-Dade Community
College's Skill Center.
its
volunteer
constituency
instructor's referrals.
is
documented
This student-taught
primarily
center
from
The success of the
draws
composition
Center's
work
in one study by an 8% decrease in drop rate
from composition classes among students who participated in
Page 10
the multiple skills remedial work.
voluntary program of "concentrated work in
similar
A
writing
and
reading
is
Center.
received
who
students
that
skills"
Skills
Basic
College's
10
11
offered
Brooklyn
by
Here, studies indicate
instruction
reading
showed
appreciable gains above those of a control group in reading
to
success
in
College,
reading
is
rate and vocabulary, both skills important
English and other academic coursework.
Staten
At
incorporated
regular-track
into
Community
Island
not
English,
only
as
remedial
well.
The
instruction
but
objectives of the
English staff are to:
Teach the student, regardless of his background,
ability,
or
purpose,
to develop to the limits of
his capacity his use of the various communication
skills, particularly reading, writing, speaking and
listening. 12
Both
staff
and
students
report
satisfaction
with
the
results of remediation.
Another school which emphasizes reading instruction in
all levels of freshman English is Boise Junior College (now
10
Welch,
"Organizing
Laboratory," p.437-39.
a
Reading
and
Writing
11
Shirley W. Wedeen, "Two-Year Basic Skills Study,"
JOURNAL OF READING, X (January, 1967), 231-37.
12
Sanford Radnor, "A Three-Track Community College
English Program," JUNIOR COLLEGE JOURNAL, XXIX (October,
1958), 97.
Page 11
Boise
The
College).
State
composition courses there are
described as "reading-centered," and
goal
that
is
skills
helping
of
effective
in
communication. "13
stated
school's
the
"the student acquire the basic
reading
and
in
oral
written
and
In addition to the basic reading skills,
the Boise approach stresses reading
the sciences, and engineering.
techniques
for
math,
This program, too, is rated
successful by students and staff.
A similar outline for
Quinsigamond
includes
Junior
reading,
four-hour
remedial
English
is
used
at
College.
The
skills
course
there
composition
and
speech
skills
in
single semester course.
a
Instruction has been so
successful that GPA increases and a decrease
in
attrition
are reported among students completing the course. 14
staff
The
of
Emory-at-Oxford
"English should now include the
arts
(reading,
writing,
also
broad
speaking
areas
believes
of
that
language
and listening)."15 The
reading
instruction
English
program includes content area skills applicable to
13
offered
in
the
college's
remedial
Hatch, "Reading-Centered Composition,"p.395.
14
James W.
Brown,
Margaret Watson, and Robert
Bowden,
"Building Basic Skills at the Community College
Level:
A New Approach," JOURNAL OF THE READING SPECIALIST,
IX (May, 1970), 144-50, 58.
15
203.
Gregory,
"An Approach to Functional English," p.
Page 12
the individual student's academic concentration.
A
similar
development
curriculum
incorporating
reading
skill
into remedial English instruction has also met
with success at Wingate Junior College.
16
What is the rationale behind all these programs
include
reading
which
instruction in English skill remediation?
Perhaps John Weber best
sums
up
the
arguments
of
most
reading proponents when he says:
Reading is such an important factor for success in
English, not to mention in other work and perhaps
life after college,.
.that reading improvement
should be an integral part of remedial English.17
.
16
Ethel K.
Smith, "Remedial Work in English at
Wingate Junior College," JUNIOR COLLEGE JOURNAL, XXVI
(March, 1956), 400-4.
17
John Weber, "Recommendations for Better English
Instruction," JUNIOR COLLEGE JOURNAL, XXXVIII (February,
1960, 34.
Page 13
College-Level English Placement Tests
C.
Freshman
customary
English
placement
testing
long
has
been
on many college campuses, and yet there is still
no consensus about which test, or what kind of test, should
be
administered.
One
national
study of higher learning
institutions which use such examinations revealed
different tests were used.
used standardized
locally
tests
27
Of the 142 schools studied, 130
available
print;
in
six
used
devised tests and six employed tests which, though
printed, were used by only one university.
Aside from the controversy
choice
that
of
placement
test,
disagreement about whether an
regarding
there
is
objective
sample, or both, should be administered.
18
the
particular
also
widespread
test,
a
writing
Those who favor a
written test argue:
Freshman English
is
a
course
of
writing;
therefore,
a speciman of writing is the only valid
evidence of ability to cope with the course.19
Detractors contend that written tests create
problems
both because of the time required to score them and because
18
19
Goodlad, "Diagnosis and Prescription in
Practice," EDUCATION DIGEST, XXXI (May, 1966),
John I.
Educational.
8-11.
Kenneth L.
Knickerbocker, "Placement of Freshmen
First-Quarter English," JOURNAL OF HIGHER EDUCATION,
XXII (April, 1951), 211.
in
Page 14
the grading of such tests is
unreliable.
therefore,
necessarily
subjective
A well-written objective test, it
is argued, can not only test writing ability, but can
time
money
and
consistency.
well
as
provide
as
a good compromise, but
the
problems
be
scoring
for
given
equal
weight?
not
so
both
objective
This seems to be
of
grading written work remain.
of
correlation of the two tests is
they
using
and writing samples for placement.
expense
save
20
Still other educators advocate
tests
and,
reliability
and
In addition, the
simple.
21
Should
What if one test predicts
success and the other predicts failure?
One professor addressed this dilemma by
accuracy
composition
of
objective test
objective
and
tests.,
a
He
grade
studying
prediction provided by an
combination
concluded
of
both
written
This
of
22
conclusion is perhaps supported by the producers
of standardized tests.
20
and
that "The objective test
ralone] is a more reliable and vastly simpler predictor
performance.
the
Of the five
Knickerbocker, p. 212.
21Ibid., p. 211.
22Ibid., p. 214.
college-level
English
Page 15
placement
enough
review
to
two
24
in
the
YEARBOOK,
MEASUREMENTS
Buros considers important
which Oscar K.
tests
are objective
seventh
three
tests
edition
MENTAL
of
are solely objective,
with
23
and
unstandardized
optional,
writing sample subtests.
CLEP
The
SUBJECT EXAMINATION IN ENGLISH COMPOSITION,
objective
although it offers both
writing
and
subtests,
provides neither scoring advice nor standardization for its
writing test.
questions:
The objective test consists of four types of
error-recognition in which sentence errors must
be picked out but not
paragraphs
identified
(essentially
language
and
analysis
name;
of
the
history
of modern English grammar;
group of sentence-conversion items wherein the
of
a section
a reading skill test);
testing "the students' knowledge
English
by
the
of
25
and a
student
is
directed to perform such tasks as "substitute an infinitive
for the first gerund. .26
Because
of
the
heavy
emphasis
23
COLLEGE PLACEMENT TEST IN ENGLISH COMPOSITION,
MISSOURI COLLEGE ENGLISH TEST, and WRITING TEST:MCGRAW-HILL
BASIC SKILLS SYSTEM.
24
CLEP SUBJECT EXAMINATION IN ENGLISH COMPOSITION
and COLLEGE ENGLISH PLACEMENT TEST.
25
Oscar K.
Buros, THE MENTAL MEASUREMENTS YEARBOOK
(Highland Park: Gryphon Press, 1972), p. 186.
26
Ibid.
Page 16
and language history, the CLEP test
terms
technical
upon
is appropriate only for determining placement in
the
most
tightly traditional English courses.
COLLEGE
The
ENGLISH
PLACEMENT TEST also includes an
standardized
unstandardized writing test with a
objective
The latter largely ignores grammar, mechanics and
subtest.
spelling and concentrates on higher level skills such as:
judgements relating to the selection of a subject
identification of dominant,
for a composition;
subordinate and irrelevant topics from unorganized
distinguishing which sentence in a group is
data;
most eEfectively expressed.27
Since these skills go
structure
sentence
beyond
to
focus on paragraph development and style matters, this test
might work
regular
well
for
separating
advanced
track composition students.
placement
and
It would, however, be
of little use with remedial students.
The COLLEGE
offers
only
sentence
PLACEMENT
TEST
objective items.
structure
are
IN
ENGLISH
COMPOSITION
Faulty diction, grammar and
adequately
covered,
but
punctuation and spelling are almost ignored,
organization,
leaving gaps in information regarding student skills.
The MISSOURI COLLEGE ENGLISH TEST is a power (speeded)
test
which
27
is
applauded for the "technical excellence of
Buros, p. 189.
Page 17
its standardization. "28 It also covers a wide skills range,
but
effectiveness of the test is somewhat undercut by
the
the format of two-thirds of its questions.
told
determine
to
contain errors
spelling
grammar,
punctuation,
In a number of questions, however,
or no errors.
one or more
are
underlined segments of themes
whether
capitalization,
of
Examinees
responses
may
be
automatically
eliminated,
leaving as few as two possible answers.
WRITING TEST:MCGRAW-HILL BASIC SKILLS SYSTEM is a
The
better-written
exercises
test
The
n29
but it is
publishers
norms.
28
"offers
many
interesting
and represents a commendable effort at measuring
important skills;
validated.
which
Buros, p. 201.
29Ibid., p. 214.
only
advise
casually
normed
and
construction of local
Page 18
Statistical Terms and Methods
D.
A number of statistical formulas
study.
One
of
were
used
this
in
most basic expressions is that of the
the
mean (the arithmetic average).
)
n
Where:
X = mean
>_X = the sum of all test scores
n = the number of scores
The standard deviation is
a
measure
variability
of
which indicates the dispersion of scores.
1
rnI-X 2
s = n
(2.)
Where:
-
(-2iX) 2
s = standard deviation
n = the number of scores
7X 2
=
the
sum
of
squares
the
of all
scores
= the sum of all test scores
(3.) The variance (s2) is a measure of total variability in
a
group
of test scores.
It is the square of the standard
deviation.
A correlation coefficient
relationship
between
two
is
used
sets
coefficient, in this case showing the
of
to
indicate
data.
correlation
One
the
such
between
Page 19
two
tests,
is
secured
application
by
of
the
Pearson
product-moment technique.
nfXY - (`LX) (Y)
(4.)
r
n>
Where:
-
Jn=Y2
(2:
(; Y)
r = coefficient of correlation
n = the number of scores
XY = the sum of
each
pair
score
on
cross-products
the
of
one
of
scores
(X represents
test;
Y
represents
score on another.)
X = the sum of all X-test scores
X
2
=
sum
the
of
the
squares
of all
X-test scores
Y = the sum of all Y-test scores
Y
2
= the
sum
of
the
squares
of
all
Y-test scores
The
point-biserial
correlation
coefficient designed to show
continuous
measure
the
is
a product-moment
relationship
between
(such as test score) and a dichotomous
measure (e.g., success or failure).
Xa - Xt
(5.)
rpbWhere:
st
r
X
X
a
pb
a
t
q
= point-biserial correlation
= mean of group A
= mean of total group
Page 20
= standard deviation of total group
s
p = number of people in
group
A
(e.g.,
number of people who succeeded
q =
number
of
people in group B (e.g.,
number of people who failed.
The Kuder-Richardson formula
yields
a reliability coefficient.
(K-R
20
20)
procedure
It provides an index of
consistency or homogeneity within a
test,
indicating
the
relationship among the test items.
2
k
(6.)
rkk
- Lpi qi
s
2
k-1
Where:
r
s
s
kk = K-R 20 reliability coefficient
2
= the variance of total test scores
pi =
the proportion of people passing an
item
qi = 1 - pi = the
proportion
of
people
failing an item
The
standard
error
of
measurement
standard deviation of a test's errors
provides
an
index
of
the
of
anticipated
represents
the
measurement
and
variability
obtained scores around the true score.
(7.)
SEm = sll
Where:
rkk
SEm = standard error of measurement
s = standard deviation
= the reliability
of
Page 21
A
significance
means.
provides
t-test
criteria
determining
for
the
an obtained difference between two sample
of
As applied to the present study:
Xs - Xu
(8.)
t
(s
s
Where:
2/n ) + (s
s
2
u
In )
u
t = the ratio of mean difference
standard
error
to
the
the
difference
of
successful
(passing)
of
unsuccessful
(failing)
of
t = D/SE
mean
students
mean
=
students
2
= variance of successful students
s
s
s
n
n
2
= variance of unsuccessful students
u
= number of successful students
s
u = number of unsuccessful students
To secure a value of t for the difference between
means
for
the
the same test administered to the same group on
two different occasions, we must first
find
the
standard
error of the difference between correlated means:
(9.)
SED =
Where:
NI(SEm1)2 + (SEm2)2 - 2(r12)(SEm1)(SEm2)
SE
=
D
standard
error
of
difference
between correlated means
SEm
1
= standard error of measurement
first test administration
for
Page 22
SEm
2
=
standard error of measurement for
second test administration
r
12
= Pearson product-moment
coefficient
first
for
correlation
second
and
administrations of test
In this particular application:
(10.)
SEm =
\f n
Where:
SEm
1
= standard error of meansurement
s = standard deviation
n = number of test scores
Standard scores provide
scores
a
norm-referenced
to
score is the z-score which
means
scores.
expresses
of
converting
raw
One type of standard
scores
on
a
scale
having a mean of zero and a standard deviation of one.
X - X
z =
(11.)
Where:
z = z-score
X = the individual's score
X = the test (entire group) mean
s = the standard deviation of the test
To
eliminate negative numbers we may convert z-scores
to Z-scores.
The latter express individuals' scores
on
scale with a mean of 50 and a standard deviation of 10.
a
Page 23
Z = 50 + 10z
(12.)
Where:
Z = the individual's Z-score
z = the individual's z-score
30
30
Formulas 5,
9
and 10 above are from Henry E.
Garrett's STATISTICS IN PSYCHOLOGY AND EDUCATION,
sixth
edition.
All other formulas are from Frederick G. Brown's
PRINCIPLES OF EDUCATIONAL AND PSYCHOLOGICAL TESTING, second
edition.
Page 24
THE ENGLISH SKILLS TEST-REVISED
III.
A.
Administration
Preparation
The ENGLISH SKILLS TEST-REVISED (EST-R)
The Reading subtest (questions 1-30)
limits.
test
31
and
is
timed
at 10 minutes.
Lest is timed at 40 minutes.
32
has two time
is
speed
a
The remainder of the
This should allow ample
time
for most students to complete subtests two through five.
Prior
to
giving the test the administrator should be
sure to have sufficient supplies of
and
#2
pencils
with
good
tests,
erasers.
answer
sheets
He must also have a
reliable watch or clock with which to time the test.
examination
The
testing
environment.
room
should
afford
a
comfortable
When large groups are being tested,
enough room should be provided so
that
students
may
sit
with one empty desk separating them.
31
The EST-R is
currently in use and, therefore,
will not be published herein.
Qualified persons may
receive copies of the test upon inquiry to Betty Lou
Carman, 1104 NW 29th Street, Corvallis, Oregon, 97330.
32
Reading subtest contains
seven
reading
selections which range from 16 to 312 words long.
Included
are passages from science, English, and social studies
texts, as well as excerpts from poetry and newspapers.
The
Page 25
should
Students
about the test and what it
told
be
will be used for prior to administration.
Directions for Administration
When all students are seated, say:
"Clear
your
will
pencil
A
completely.
desks
be
provided."
Now
pass
each student.
out
#2 pencil and one answer sheet to
one
Say:
"Fill in the area at the top
answer
your
of
Be sure to
in your name and social security number.
Write
name
your
and
Are there any questions about how
security number.
social
below
spaces
also darken the appropriate
sheet.
to fill out your answer sheet?"
Give students
Answer questions.
complete this task.
"As
you
sheet.
answer
thoroughly.
given
30.
If
time
to
Say:
the
distinct, dark marks
answer
sufficient
in
you
test questions, be sure to make
the
wish
appropriate
to
spaces
on
the
change an answer, erase
Make no marks on the test itself.
You will be
10 minutes to do the Reading subtest, questions 1 to
Are there any questions?"
Answer questions.
Pass out tests.
"You may begin the Reading Test."
Say:
Page 26
After 10 minutes say:
Now turn ahead in your
"Stop.
Spelling
Test.
sure
Be
sheet to question 31.
Reading
Test;
it
is
Do
a
test
booklet
to
the
also move up on your answer
to
not
go
back
to
work
on
the
special speeded test and we want
everyone to have had the same amount of time
on
it.
You
will have 40 minutes to complete the remainder of the test.
In that time you will do the Spelling Test, the Grammatical
Usage
Test,
Test.
Are there any questions?"
Answer
the Vocabulary Test, and the Essay Correction
appropriate
administration,
e.g.,
questions
there
is
concerning
test
no penalty for incorrect
responses. Say:
"When you have finished, turn in your
sheet and test.
Ready?
Begin."
After 40 minutes say:
"Sto p.
"
Collect all test materials.
pencil,
answer
Page 27
B.
Scoring
1.
Reading
Speed Score
a.
An
40%
ample
of
reading
test
is allowed to
time
re-scan reading selections in response to questions, and to
find
mark
and
desired
choice
on
the answer sheet; the
reading rate figures, then, are calculated on the basis
six
minutes
actual
reading time.
of
The Reading Test Speed
Chart (table I) shows the approximate rate
(in
words
per
minute) achieved according to the last response made on the
Reading Test.
student
is
For example, if the last item answered by
number
13,
his approximate Reading Test rate
would be 136 words per minute (wpm).
TABLE I.
Last
Question
Answered
1
2
3
4
WPM
72
74
74
77
69
5
6
84
7
87
8
96
102
103
9
10
a
READING TEST SPEED CHART
Last
Question
Answered
WPM
11
12
13
14
15
16
17
18
19
20
131
133
136
137
145
184
189
193
196
199
Last
Question
Answered
21
22
23
24
25
26
27
28
29
30
WPM
214
218
219
222
224
274
276
278
280
284
Page 28
b.
Comprehension Score
The comprehension score is computed
by
dividing
the
number
of the last Reading Test question answered into the
number
of
correct
multiplied
by
comprehension
The
responses.
The
100.
quotient
resulting
percentage
score,
figure
i.e.,
then
is
represents a
percent
the
of
questions which were attempted (or skipped) by that student
at that speed.
For example, let us say
that
the
which the student responded was number 16.
rate would then be 184 wpm.
he
answered
question
last
to
His approximate
If, out of the
16
questions,
12 correctly, his comprehension percentage is
12/16 X 100 or 75% at 184 wpm.
c.
Evaluation
This test is
broadest
sense.
meant
be
to
will
It
diagnostic
not
delineate
weaknesses, but will provide an approximate
and
comprehension
The
score.
intent
only
in
particular
reading
here
the
speed
to make
is
students aware of their skill
deficiencies
them
may receive from the reading
to
the
services
they
and
alert
to
improvement staff at the Communication Skills Center.
It is impossible to set universally agreeable
scores;
standards
differ
and
it
is
likely
cutting
that many
Page 29
teachers will choose for themselves a level at
which
believe
instruction
remedial
developmental
or
should be recommended.
reading
These recommendations may
they
made
be
using a percentile rank cutting point (e.g., everyone whose
score
comprehension
on
percentile
speed
or
falls
below
given
a
referred for instruction), or by setting an
is
arbitrary comprehension rate and
speed
score
as
cutting
points.
Another
approach
is
approximate reading rate and
him
of
to
inform
the
comprehension
student of his
advise
score,
his percentile rank in the class (this information
will be given for tests scored by Oregon State University's
Computer
Center)
and
leave to him/her the decision about
whether or not he needs reading instruction.
Students should, however, be advised that anyone whose
reacting
speed
is less than 150 wpm or whose comprehension
rate is less than 60% is likely to find
his
college
work
severely handicapped by his reading ability.
2.
Scoring
Spelling,
Grammatical Usage, Vocabulary, and
Essay Correction
The
scoring
Vocabulary
that of the
and
of
Essay
Reading
the
Spelling,
Grammatical
Usage,
Correction subtests is simpler than
Test.
Computer
scoring
gives
the
Page 30
number
each student answers correctly on each subtest.
hand score, use the key obtainable from the CSC
and
To
count
the correct responses in each subtest.
If
scores
are used for placement in English classes,
or as an indication of English proficiency as
the
required
by
schools of Agriculture, Forestry, and Health and P.E.,
the cutting scores agreed upon by
observed.
these
must
schools
The minimum passing scores are:
be
Spelling Test,
15 correct; Grammatical Usage Test, 20 correct;
Vocabulary
Test, 7 correct; Essay Correction Test, 25 correct.
These
same standards may apply if the test is used to
find areas of
skill
weakness
so
that
students
may
referred to the Communication Skills Center for tutoring.
be
Page 31
Description of Test Results
C.
This
study
is
based on three administrations of the
EST-R: one as a pre-test to 33 WR. 40X students; one
as
a
post-test to the 16 students remaining in 40X at the end of
the term; and one as a pretest to 45 WR. 230 students.
were
administrations
made
during
spring
All
term, 1977, at
Oregon State University.
Table
frequencies
II
and
for
table
the
Spelling,
Grammatical
Correction
tests
instruction.
III
Reading
show
Test
Usage,
administered
test
and
and
for the combined
Vocabulary
to
scores
students
and
Essay
prior to 40X
Page 32
TABLE II.
WR. 40X READING PRE-TEST
Cumulative
Score Frequency Frequency
Cumulative
Score Frequency Frequency
2
2
2
13
1
3
2
14
2
4
0
4
6
1
7
6
4
7
1
11
12
8
2
14
9
3
17
10
11
12
4
21
22
22
15
16
17
18
19
20
21
22
0
5
1
0
Mean = 10.18
Median = 9
2
2
1
1
1
0
1
Standard Deviation = 5.54
Range = 21
23
25
25
27
29
30
31
32
32
33
Page 33
TABLE III.
WR. 40X PRE-TEST
SPELLING, GRAMMATICAL USAGE,
VOCABULARY AND ESSAY CORRECTION COMBINED
Cumulative
Score Frequency Frequency
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
1
1
0
0
0
0
0
0
0
0
0
0
1
1
2
2
2
2
0
0
0
0
0
0
1
1
1
1
1
1
1
1
1
2
2
1
2
3
1
4
2
0
6
6
1
7
1
8
Mean = 51.88
Median = 51
Cumulative
Score Frequency Frequency
46
47
48
49
50
51
52
53
54
55
56
0
8
1
9
4
13
14
14
17
18
18
57
58
59
60
61
62
63
2
64
65
0
66
67
68
1
0
3
1
0
1
0
1
2
1
1
1
1
1
1
2
0
1
Standard Deviation = 10.18
Range = 47
19
19
20
22
24
25
26
27
28
29
29
30
32
32
33
Page 34
Tables IV and V show Reading Test
scores
frequencies
and
for
combined
and
test
EST-R administered as a
the
post-test to 40X.
TABLE IV.
WR. 40X READING POST-TEST
Cumulative
Score Frequency Frequency
5
2
2
6
7
8
9
1
2
3
3
5
8
10
11
12
13
14
0
2
1
1
1
0
Mean = 10.38
Median = 8
10
11
12
12
13
13
Cumulative
Score Frequency Frequency
15
16
0
17
18
19
20
21
22
23
0
0
0
0
1
1
0
1
Standard Deviation = 5.37
Range = 19
13
14
14
14
14
14
15
15
16
Page 35
TABLE V.
WR. 40X POST-TEST
SPELLING, GRAMMATICAL USAGE,
VOCABULARY AND ESSAY CORRECTION COMBINED
Cumulative
Score Frequency Frequency
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
4
0
0
4
4
4
1
5
5
5
6
6
6
6
0
0
1
0
0
0
6
0
0
1
0
0
0
0
0
0
Mean = 56.81
Median = 56
12
12
12
13
13
13
13
13
13
13
Cumulative
Score Frequency Frequency
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
0
0
0
0
0
1
0
1
0
0
0
0
0
0
0
0
0
0
0
1
Standard Deviation = 11.01
Range = 40
13
13
13
13
13
14
14
15
15
15
15
15
15
15
15
15
15
15
15
16
Page 36
Tables VI and VII contain Reading
Test
and
combined
test scores and frequencies for the EST-R administered as a
pre-test to WR. 230 students.
TABLE VI.
WR. 230 READING PRE-TEST
Cumulative
Score Frequency Frequency
5
6
1
1
1
2
7
4
6
8
9
1
7
3
3
10
13
17
19
10
11
12
13
14
15
16
4
2
2
3
3
0
Mean = 15.49
Median = 14
21
24
27
27
Cumulative
Score Frequency Frequency
17
18
19
20
21
22
0
23
24
25
26
27
3
1
1
1
27
28
29
30
4
4
33
33
36
40
44
0
44
1
45
3
0
Standard Deviation = 6.69
Range = 23
Page 37
TABLE VII.
WR. 230 PRE-TEST
SPELLING, GRAMMATICAL USAGE,
VOCABULARY AND ESSAY CORRECTION COMBINED
Cumulative
Score Frequency Frequency
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
1
0
0
0
2
0
0
0
0
0
1
0
0
1
3
0
3
1
2
Mean = 59.18
Median = 59
Cumulative
Score Frequency Frequency
1
56
57
58
59
3
3
3
3
60
61
62
63
3
64
2
3
65
66
67
68
69
70
71
72
73
0
0
1
1
1
4
4
4
5
8
8
11
12
14
4
0
3
2
2
2
1
4
2
1
1
18
18
21
23
25
27
28
32
34
34
34
36
37
38
1
40
41
3
44
1
45
2
Standard Deviation = 8.64
Range = 37
Page 38
Standardization
D.
1.
Validity
Test validity is determined by ascertaining how
well,
or to what extent, the test measures what it is designed to
measure.
of
There are several ways to quantify
the
validity
a test; this investigation utilized two such approaches
to study the validity of
the
EST-R,
i.e.,
the
validity
coefficient and group separation.
a.
Validity Coefficient
To
find
validity
a
coefficient
for
the EST-R, we
correlated success (a "pass") or failure (a withdrawal,
incomplete
or
pass") on WR. 40X with pre-test and
"no
a
an
with post-test scores using the point-biserial calculation.
The point-biserial correlation of pre-test EST-R score
and success/failure on WR. 40X is r=+.16.
is
not significant at the 0.10 level.
no
evidence
administered
an
for
as
a
inference
40X
that
This correlation
We have, therefore,
the
EST-R,
as
pre-test, is a valid predictor of
success in the course.
The
post-test
r=+.57.
Thus,
point-biserial
EST-R
score
correlation
and
coefficient
for
success/failure on WR. 40X is
This coefficient is significant at the 0.05 level.
we
have evidence of a real relationship between the
Page 39
content of the EST-R and the content of WR. 40X.
Why
such
there
is
a
difference
validity
the
in
coefficients for the pre-test and post-test?
One reason is
that 40X instruction intervened between the two tests.
validity
The
coefficient
shows
that
profited most from the course were
those
whose
post-test
best correlated with success in WR. 40X;
those who
significant
scores
profited least
correlated
were
with
whose
those
failure
post-test
in WR. 40X.
those
and
adequacy
the
of
best
scores
This validates both
the EST-R post-test as an indicator of the effects
instruction
who
procedures
of
40X
used
for
evaluating WR. 40X students.
b.
Group Separation
Group separation examines test validity
the
means
of
two
groups to determine whether there is a
significant difference between them.
studied
were
comparing
by
The two EST-R
groups
those who were successful and those who were
unsuccessful in WR. 40X.
The method
used
was
a
t-test,
performed first on the two groups' pre-test scores and then
on their post-test scores.
The pre-test t-value of 1.1151 is not
the
0.10
at
level; the post-test t-value of 3.4553, however,
is significant at the
post-test
significant
is
0.01
level,
indicating
the
EST-R
likely to be a valid indicator of separation
Page 40
between those who fail and those who
after
WR. 40X,
pass
the instruction is completed.
Thus the group separation and the validity coefficient
methods
both
indicator
of
support
EST-R
the
post-test
success or failure in WR. 40X.
attribute the greater validity of the
with
as
a
valid
Again we may
post-test
(compared
of the pre-test) to the effects of instruction;
that
profited
those who
most
instruction
from
(passed
40X)
scored significantly higher on the post-test than did those
who did not succeed in WR. 40X.
One method of directly testing the EST-R's validity as
an
indicator of the effects of instruction is to perform a
t-test on both successful and
pre-test
and
post-test
unsuccessful
40X
EST-R scores to see if there is a
significant difference between the pre-test
group
means.
0.01 level.
The
There
and
post-test
t-value=3.16 and is significant at the
is,
then,
a
high
probability
difference between post- and pre-test means.
our earlier findings and underscores the
EST-R
students'
of
a
This supports
validity
of
the
as an indicator of the improvement that results from
WR. 40X instruction.
One further t-value is calculated to determine whether
there
is a significant difference between the means of the
40X EST-R pre-test
t-value
is
and
of
the
WR. 230
pre-test.
This
1.52 and is not significant at the 0.10 level.
Page 41
Again we find that the EST-R
is
not
a
valid
predictive
instrument when administered prior to instruction.
The above data are summarized in tables VIII and IX.
TABLE VIII.
VALIDITY COEFFICIENTS (r)
Groups Correlated
WR. 40X Pre-test:
Successful x Unsuccessful
+0.16
WR. 40X Post-test:
Successful x Unsuccessful
+0.57*
*Significant at 0.05 level
TABLE IX. GROUP SEPARATION
Group
40X Pre-test x Post-test
3.116**
40X Pre-test x 230 Pre-test
40X Pre-test:Successful x
Unsuccessful
40X Post-test: Successful x
Unsuccessful
**Significant at the 0.01 level
Page 42
2.
Reliability
When
we
attempting
to
investigate
a
determine
what
test's
reliability,
proportion
of
we are
the
total
variance of test scores is attributable to true (not error)
variance.
over
We are also interested in how stable the test is
time;
i.e.,
if
no instruction intervenes would the
same person score similarly on a test adminstered and
re-administered
after a period of time?
impractical, however, to
since
it
may
be
administer
Since it is often
the
test
twice
another
means
approach involves
consistency,
and
difficult to insure that no learning of
test content has occurred between administrations,
use
then
of
finding
determination
test
of
we often
reliability.
the
test's
This
internal
of the consistency of performance on the test
items.
To determine the internal consistency
the
Kuder-Richardson
used.
20
of
the
EST-R,
(K-R 20) reliability formula was
The K-R 20 is a type of split-half reliability which
allows
us
other item.
to
correlate
each item on the test with every
The reliability coefficients thus obtained are
shown on table X.
Table
X
shows
the
K-R
20
individual test for all three groups:
post-test
groups
are
and
230
also
pre-test.
given
for
reliability
40X
of
pre-test,
each
40X
Reliabilities for all three
the
combined
Spelling,
Page 43
Grammatical Usage, Vocabulary and Essay Correction tests.
A combined test reliability for the pooled 40X and 230
pre-test
scores
included
because
also
is
given.
This
calculation
comes closer to representing a mixed
it
group of students, i.e., students with a broader
abilities,
than
the combined Spelling, Grammatical
Essay
Correction
range
the 40X or 230 students alone.
do
is
Since
Vocabulary
Usage,
of
and
test is often administered as an English
proficiency
test
information
regarding
mixed
to
the
groups
students,
of
reliability
some
of these tests is
likely to be useful.
Several other considerations should be noted regarding
the
figures reported on table VI.
although the Reading Test
shows
First is the fact that,
highest
the
reliability
figures, that test is speeded, and, because the K-R 20 is a
type of split-half reliability formula, these
coefficients
are spuriously high.
Another
important
consideration
the
is
reliability coefficients are affected by test
else
being
shorter
equal,
This
one.
reliability
that
length;
all
longer test is more reliable than a
explains
coefficients
individual test
Reading Test.
a
fact
except
are
those
why
the
higher
of
the
combined
than
those
artificially
tests'
of any
high
Page 44
range
The
scores (the heterogeneity of the group
of
tested) also contributes to the
more the range is reduced, the more the
The
coefficient.
reliability
the
of
size
Since, on the 129-item
reliability coefficient is reduced.
combined Spelling, Grammatical Usage, Vocabulary, and Essay
post-test
and
pre-test
only
were
WR. 40X
pre-test,
Correction test, the ranges for WR. 230
and
59
50,
57
respectively, the reliability coefficient is not as high as
it would be if the range were more extended.
reliability
A
educational
standardized
achieve
0.90
of
this
higher
or
tests.
33
The
expected
is
EST-R
of
not
does
in the present study, indicating that
level
deletion
further revision, addition and
of
items
is
in
order.
standard
The
error
Spelling,
Grammatical
Correction
tests
measurement for the combined
of
Vocabulary
Usage,
Essay
seven, table X.
The
standard error of measurement may be used to calculate
the
is
shown
in
row
and
expected variability (error) in an individual's test score.
Using the standard error of measurement
score
band
can
figure
a
or range (confidence interval) within which we
can state, with a given
degree
of
individual's true score lies.
33
we
Frederick G. Brown, p. 78.
confidence,
that
the
Page 45
CIp = X + (z)(SEm)
Where:
CIp =
confidence
interval
for
a given
probability
X = an individual's test score
z = a factor. When:
p (probability) = 0.68, z = 1.00
p = 0.85, z = 1.44
p = 0.90, z = 1.64
p = 0.95, z = 1.96
p = 0.99, z = 2.58
For example, then, to compute an individual's
score
range
in which we can be 95% confident we use the formula for the
limits:
X + 1.96 SEm
If a student's score is 50 and the SEm = 4.426
50 + 1.96 (4.426)
50 + 8.675
50 +9
41 to 59
Therefore, the student's score range at the 95%
level is 41 to 59.
confidence
Page 46
TABLE X.
RELIABILITY
Pre-test
WR. 40X
and
Pre-test Post-test Pre-test WR. 230
WR. 40X WR. 40X
WR. 230
pooled
1) Reading Test
30 items
.851.
.854
.891
2) Spelling Test
20 items
.622
.735
.502
3) Grammatical Usage
35 items
.526
.676
.352
-.059
.552
.360
5) Essay Correction
34 items
.778
.576
.798
6) Tests 2-4 Combined
99 items
.811
.849
.750
.803
4.426
4.278
4.320
4.412
4) Vocabulary Test
10 items
7) SE Test 2-4
Combined
Page 47
3.
Norms
The
following
tables
XI and XII represent norms for
the EST-R administered as a pre-test to Wr. 40X.
population
State
consisted
University's
transformed
of
students
33
WR. 40X
spring
The
test
enrolled in Oregon
term,
The
1977.
scores given are percentile ranks (PR), z- and
Z-scores.
Percentile ranks convert raw scores to class rank; the
percentile
represents
score
percent
the
of
scoring at, or below, a particular raw score.
if
a
student's
the
group
For example,
PR is 83, he scored as well as, or better
than, 83% of those in his group.
A z-score is a
conversion
of
the
raw
score
to
a
standard scale with a mean of zero and a standard deviation
of one.
A z-score of two indicates a
score
two
standard
deviations above the group mean.
Z-scores
represent
raw
converted to a standard scale with
standard
deviation
of
10.
A
scores
a
which
mean
of
have
50
been
and
a
Z-score of 65 indicates a
score 1.5 standard deviations above the mean.
Separate norms are given for the Reading
XI)
and
for
the
combined
Spelling,
Test
(table
Grammatical Usage,
Vocabulary and Essay Correction tests (table XII).
Although the EST-R administered as a
be
pre-test
cannot
considered a valid predictor of 40X success or failure,
Page 48
it
should
useful
be
instructors
to
compare
to
WR. 40X
both
their
and
to
EST-R
students'
scores to these 40X norms.
TABLE XI.
SCORE TRANSFORMATION TABLE
WR. 40X PRE-TEST:
X
PR
----
2
3
2
5
4
11
18
26
33
38
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
Mean = 10.18
Median = 9
45
56
64
65
67
71
74
77
83
88
91
94
95
97
READING
z
Z
---
--35
-1.48
-1.30
-1.21
-0.94
-0.76
-0.57
-0.39
-0.21
-0.03
0.15
0.33
0.51
0.69
0.87
0.05
1.23
1.41
1.59
1.77
1.95
2.13
37
39
41
42
44
46
48
50
51
53
55
57
59
61
62
64
66
68
70
71
Standard Deviation = 5.54
Number of Scores = 33
WR. 121
pre-test
Page 49
TABLE XII.
SCORE TRANSFORMATION TABLE
WR. 40X PRE-TEST
SPELLING, GRAMMATICAL USAGE,
VOCABULARY, AND ESSAY CORRECTION COMBINED
X
PR
z
-22
23
--
---
24
25
26
27
2
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
1
2
2
2
2
2
2
2
2
2
3
5
5
5
5
5
5
6
9
14
17
18
21
-2.94
-2.84
-2.74
-2.64
-2.54
-2.44
-2.35
-2.25
-2.15
-2.05
-1.95
-1.86
-1.76
-1.66
-1.56
-1.46
-1.36
-1.27
-1.17
-1.07
-0.97
-0.87
-0.77
-0.68
Mean = 51.88
Median = 51
Z
--
X
PR
z
--
---
21
22
23
24
25
26
27
46
47
48
49
50
51
52
53
54
55
56
57
-23
28
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
58
59
60
61
62
63
64
65
66
67
68
24
32
39
41
45
52
53
55
56
58
62
68
73
76
79
82
85
86
88
92
95
97
-0.58
-0.48
-0.38
-0.28
-0.18
-0.09
0.01
0.11
0.21
0.31
0.41
0.50
0.60
0.70
0.80
0.90
0.99
1.09
1.19
1.29
1.39
1.49
1.58
Standard Deviation = 10.18
Number of Scores = 33
Z
--
44
45
46
47
48
49
50
51
52
53
54
55
57
57
58
59
60
61
62
63
64
65
66
Page 50
IV.
SUMMARY
It is the conviction of this investigator
reading
possible,
instruction
remedial English courses.
reading
individual's
To
should
encourage
deficiencies,
added to the ENGLISH SKILLS TEST,
a
be
an
when
that,
included
in
awareness
of
a reading subtest was
placement/proficiency
test used at Oregon State University (OSU).
This new test,
Grammatical
consisting
of
Vocabulary
and Essay Correction subtests, was entitled the
Reading,
Spelling,
Usage,
ENGLISH SKILLS TEST-REVISED (EST-R).
Standardized administration and scoring procedures for
the
EST-R
test were made.
WR. 40X
defined, and three administrations of the
were
The EST-R was given as a
students,
as
and as a pre-test to 45
pre-test
to
33
a post-test to 16 WR. 40X students,
WR. 230
students.
Test
scores,
frequencies, mean, median, standard deviation and range for
each group were given for the
other
four
subtests
Reading
combined.
(percentile rank, z- and Z-scores) were
Test
and
Transformed
provided
for
the
scores
for
the
WR. 40X Reading pre-test and combined test pre-test. 34
34
Raw
data
used in this study is available
Betty Lou Carman, 1104 NW 29th Street,
Corvallis,
Oregon, 97330.
from:
Page 51
Criterion-related
validity
studies
indicated
EST-R scores, when the test is administered as a
to
WR. 40X,
post-test
have concurrent validity with WR. 40X grades.
Subtest reliability ranged from
good
that
educational
tests
have
-0.059
a
0.891.
to
reliability
better, the EST-R must undergo more
Since
0.90 or
of
research
to
increase
the reliability of the test.
The
researcher's
next step will be to administer the
EST-R to a much larger and more
e.g.,
a
representative
sample
heterogeneous
of
population,
incoming Oregon State
University freshmen.
Such an administration
higher
both range and reliability.
figures
for
should also be re-appraised, and
item revision begun.
the
task
should
of
yield
Validity
systematic
Page 52
BIBLIOGRAPHY
Baron, Denis and Harold Bernard. EVALUATION TECHNIQUES FOR
CLASSROOM TEACHERS. New York: McGraw-Hill, 1958.
Bossome, Richard M., ed. TEACHING BASIC ENGLISH
New York: Van Nostrand Reinhold, 1971.
COURSES.
Kenneth J.,
"A Reading Program for Freshmen,"
JUNIOR COLLEGE JOURNAL, X (September, 1939), 16-20.
Brough,
James W., Margaret Watson, and Robert Bowden.
"Building Basic Skills at the Community College Level:
A New Approach," JOURNAL OF THE READING SPECIALIST, IX
(May, 1970), 144-150, 158.
Brown,
Frederick
G.
PRINCIPLES OF EDUCATIONAL
PSYCHOLOGICAL TESTING. New York: Holt, Rinehart
Winston, 1976.
Brown,
AND
and
Oscar K., ed.
THE MENTAL MEASUREMENTS YEARBOOK.
Highland Park: Gryphon Press, 1972.
Buros,
Dabbs, Lowell.
"A Report of Remedial English in Colleges,"
JUNIOR COLLEGE JOURNAL, XXVII (March, 1957), 381-387.
Diederich, Paul B.
"Testing in the New English
ENGLISH RECORD, III (Spring, 1953), 11-17.
Program,"
Doyle, Marvyl.
"Remedial English in the Junior College:
Students,
Programs,
Plans," Selected Papers,
46th
Annual Convention, American Association of Junior
Colleges (1966), 77-78.
Ebel, Robert L., Ed. ENCYCLOPEDIA OF EDUCATIONAL RESEARCH.
Toronto: Macmillan, 1969.
Fleischer, Ernest.
"Reading Surveys and Their Implications
in Freshman English Programs," JOURNAL OF THE READING
SPECIALIST, V (May, 1966), 160-164.
Forst,
Inez.
JUNIOR
"An English Testing and Guidance Program,"
COLLEGE JOURNAL,
XVII
(February,
1947),
234-243.
Garrett, Henry E.
STATISTICS IN PSYCHOLOGY AND EDUCATION.
New York: David McKay, 1970.
Page 53
Stanley.
Glass, Gene V. and Julian C.
IN
EDUCATION
AND PSYCHOLOGY.
Prentice-Hall, 1970.
STATISTICAL METHODS
Englewood Cliffs:
Golden, Ruth L.
IMPROVING PATTERNS OF LANGUAGE
Detroit: Wayne State University Press, 1960.
USAGE.
"An Approach to Functional English in a
Gregory,
John W.
Four-Year Junior College," JUNIOR COLLEGE JOURNAL,
XXIX (December, 1958), 203-205.
"Diagnosis
John
I.
and
Prescription in
Educational Practice," EDUCATION DIGEST, XXXI (May,
1966), 8-11.
Goodlad,
Ada Y.
"Reading-Centered Composition Course,"
JUNIOR COLLEGE JOURNAL, XXVI (March, 1956), 395-399.
Hatch,
Higgins,
John A.
"Remedial Students'
Needs," COLLEGE
COMPOSITION
AND COMMUNICATION,
XXIV (May,
1973),
188-192.
Johnson, Roy Ivan and Hugh McCammon.
"Language Instruction
in
the Junior College," COLLEGE ENGLISH, II (March,
1941), 584-592.
King, Arthur H.
"Notes on Remedial English at the Higher
Education Level," ENGLISH LANGUAGE TEACHER, XXVII
(June, 1973), 245-250.
Knickerbocker,
Kenneth L.
"Placement of Freshmen
in
First-Quarter English," JOURNAL OF HIGHER EDUCATION,
XXII (April, 1951.
Litsey, David M.
"Trends in College Placement Tests in
Freshman English," THE ENGLISH JOURNAL XLV (May,
1956), 250-256.
McDaniel, Marjorie C.
"University Student Aid Center,"
EDUCATION LEADERSHIP, XXVIII (March, 1971), 641+.
Otto,
Wayne,
Richard A. McMenemy and Richard J. Smith.
CORRECTIVE AND REMEDIAL TEACHING.
Boston:
Houghton
Mifflin, 1973.
Popham, W.
James.
EDUCATIONAL
Harper and Row, 1967.
STATISTICS.
New
York:
Page 54
Radnor, Sanford.
"A Three-Track Community College
Program," XXIX (October, 1958), 97-100.
English
Ethel K.
"Remedial Work in English at Wingate
(March,
Junior College," JUNIOR COLLEGE JOURNAL, XXVI
1956), 400-404.
Smith,
Sutton, Doris G. and Daniel S. Arnold.
"The Effects of Two
Methods of Compensatory Freshman English," REASEARCH
IN THE TEACHING OF ENGLISH,
VIII (Summer, 1974),
241-249.
Weber,
John.
"Recommendations
for
Better
Instruction,"
JUNIOR
COLLEGE
JOURNAL,
(February, 1968), 32-36+.
English
XXXVIII
Wedeen, Shirley U. "Two-Year Basic Skills Study,"
OF READING, X (January, 1967), 231-237.
JOURNAL
Grace
W.
"Organizing a Reading and Writing
Laboratory
in
Which
Students
Teach,"
COLLEGE
COMPOSITION AND COMMUNICATION, XXV (December, 1974),
437-439.
Welch,
APPENDIX
Page 55
TABLE XIII.
EST-R TEST SCORES
WR. 40X PRE-TEST
Student
Reading
Score
Spelling
Score
Gram.
Usage
Score
Vocab.
Score
1
2
6
18
5
2
3
4
10
4
8
6
4
11
5
6
7
8
9
6
9
5
19
10
18
20
15
12
13
10
13
16
16
15
15
18
15
12
18
17
10
15
15
4
10
11
12
13
9
8
22
14
9
11
10
54
14
15
6
16
17
18
19
20
21
22
23
14
24
25
26
27
28
29
30
31
32
33
7
6
10
17
16
10
8
8
4
4
5
3
3
2
4
7
4
14
23
20
25
11
12
10
6
4
4
10
6
19
19
6
2
4
16
14
16
3
11
13
4
13
4
9
8
4
9
7
5
12
12
4
14
14
9
17
12
18
11
21
11
20
17
23
18
20
10
20
13
17
3
5
6
6
5
6
8
Essay
Correctn.
Score
4
15
21
23
18
26
24
26
25
16
25
16
21
22
20
23
22
24
24
22
7
18
15
25
26
19
21
19
5
24
5
2
26
3
2
29
20
4
23
4
3
6
6
Page 56
TABLE XIV.
EST-R TEST SCORES
WR. 40X POST-TEST
Gram.
Student
Reading
Score
Spelling
Score
1
7
5
2
3
13
16
14
4
5
6
7
9
5
8
9
8
21
9
11.
10
11
12
13
10
14
15
16
5
23
8
6
7
8
11
12
12
Usage
Score
19
16
15
18
21
6
14
15
26
15
16
12
28
16
14
4
15
17
23
9
12
15
14
10
10
16
11
Vocab.
Score
2
7
4
2
1
5
7
5
6
5
7
9
7
6
4
6
Essay
Correctn.
Score
20
22
26
20
22
18
22
25
21
25
30
32
22
20
21
29
Page 57
TABLE XV.
EST-R TEST SCORES
WR. 230 PRE-TEST
Gram.
Student
Reading
Score
Spelling
Score
4
25
11
20
15
5
6
7
7
7
24
8
9
7
10
11
12
13
18
19
12
17
13
13
13
13
10
19
14
14
7
14
15
16
12
25
25
13
10
11
16
16
11
1
2
3
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
5
24
15
25
23
11
21
24
11
13
23
21
15
10
12
9
9
27
24
13
14
14
11.
11
16
13
12
14
15
15
13
13
13
12
5
16
14
10
13
11
16
17
Usage
Score
25
23
22
20
16
20
21
19
19
18
20
19
17
15
22
16
17
17
16
12
16
21
19
19
17
21
25
16
11
17
20
16
17
13
23
23
Vocab.
Score
5
2
8
3
2
5
5
5
4
4
5
2
2
5
8
3
8
6
7
4
7
4
5
8
6
6
5
6
2
5
7
4
5
4
6
7
Essay
Correctn.
Score
29
17
26
28
26
25
29
24
18
20
6
28
21
25
26
23
17
26
28
24
24
24
25
22
23
29
13
22
19
25
27
26
20
13
27
23
Page 58
EST-R TEST SCORES
TABLE XV.
WR. 230 PRE-TEST, CONTINUED
Gram.
Student
37
38
39
40
41
42
43
Reading
Score
Spelling
Score
14
13
Usage
Score
23
6
10
18
14
21
14
23
11
12
11
44
9
8
45
10
23
16
16
19
17
21
20
9
8
9
9
Vocab.
Score
5
5
9
4
5
3
6
3
4
Essay
Correctn.
Score
33
14
25
13
25
28
24
18
25