A FACE? QESEGN T0 MEASLéRE

i:

1
¥

ﬁgiiée

as}?

:5:
(9

3‘2.
l ff
0"

gaﬁéar'resz

inn.

   
     

  

LIBRARY ‘

Michigan State
University

 
 

is to certify‘ _
"a' ' a. ..
thesis entitled

A Facet Design to Measure
Reading Comprehension

presented by 5

Warren De loy Wilde

has been accepted towards fulfillment
of the requirements for

Ph . D . degree in Elementary
Education

7%. 4/, (A, KM

/ / Major professor

Date May 18,. 1973

 

0-7639

WQ/QVQFQW'
high hovel
BOOK BINDERY LTD.
10372 - 60 Ave_, Edmonton

"THE HIGHEST LEVEL OF
CRAFTSMANSHIP"

ABSTRACT
A FACET DESIGN TO MEASURE
READING COMPREHENSION
by

Warren Deloy Wilde

This study attempted to measure reading compre-
hension of grade ten students using a multiple choice test
constructed by a facet design. The facet design concen-
trates on the relation of the items to the information
presented in the text or elsewhere rather than on skills
and abilities which are presumed to be involved in compre-
hension. The study also compared comprehension as measured
by the Davis Reading Test and a cloze test with the results
of the facet test.

The sample consisted of 186 students chosen from
a grade ten population in two large high schools in Edmonton,
Alberta, Canada. All students were enrolled in a basic
English course. The students received the Davis Reading
Test and the facet test on one day and the cloze test one
week later.

The quality of the facet test was determined by
means of an item analysis which yielded an item difficulty

index, a bi-serial correlation, and an item reliability

 

index. Kuder-Richardson 20 was used to compute the reliabil—
ity of the facet test. Test results on the facet test were
compared with those on the Davis Reading Test and the cloze
teSt by means of a Pearson product-moment correlation
coefficient.

It was concluded that: l. the facet test was a
reliable measure of reading comprehension; 2. the facet
test discriminated among achievement levels; 3. there was
a significant relationship between the facet test and other
measures of reading comprehension, i.e. the cloze test and

the Davis Reading Test.

 

 

 

A FACET DESIGN TO MEASURE

READING COMPREHENSION

by

Warren Deloy Wilde

A THESIS

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Department of Elementary
and Special Education

1973

 

 

Dedicated to my loving family,

Ruth, Leslie, Renee, Jana, Russell, Marlys.

ii

ACKNOWLEDGMENTS

The writer wishes to express appreciation to his
major advisor, Dr. B. H. Van Roekel, for his guidance and
assistance throughout the graduate program and especially
in the writing of this thesis.

The interest of my committee members, Dr. Keith
Anderson, Dr. Ruth Useem, and Dr. Gerald Duffy, as well as
one member who retired from the faculty, Dr. Harold Byrum,
is sincerely appreciated.

Appreciation is expressed to my colleagues at the
University of Alberta who encouraged me to complete the
thesis and gave able counsel in time of need. Especially
a thanks to Dr. Marion Jenkinson for her support.

A very special thanks to my wife and family for
their support and encouragement throughout my graduate
program and especially during the summer when this thesis

was written.

iii

TABLE OF CONTENTS

Page

LIST OF TABLES ......... ...... ...... ..... ............. vi
LIST OF FIGURES . ..... . ....... ........................ vii

Chapter

I. THE PROBLEM ................ ............... ... 1
What is reading? ...... .......... ........... 1
Measurement in education ................... 2

Need for this study ........................ 4

Purpose of this study ...................... 4

Facet design ............................... 5
Hypotheses .................. ..... .......... 7
Additional considerations .... ...... ........ 7
Limitations of the study ................... 7
Definition of terms ........................ 8

Overview ................................... 8
II. REVIEW OF RELATED LITERATURE ................. 10
Measurement in reading ..................... 10

Common complaints about tests 13

What is reading comprehension? ............. 14
Need for new theory ........................ 19
Facet test ................................. 23
Cloze technique . ......... .................. 29

Summary ................. ...... ............. 31

iv

III. THE DESIGN AND PROCEDURES ...... ..... .

Population and sample ..... ...... .. .......

Measuring instruments ..............
Pilot Study. ..... ......OOOOOOOOO...

Administration of test instruments .

Analysis of data ......... ....... .........

Summary ......... ...... ..... ........... ...
IV. RESULTS ............... ...... ..............

Introduction ..... . ......... . ...... . ......

Results concerning hypothesis one ..
Results concerning hypothesis two ..

Results concerning hypothesis three

Additional considerations ................

Summary ........ ..........................

V. SUMMARY AND DISCUSSION .... ....... ..........
Summary of study ............ ..... ........

Discussion ..................... ..........

Further research .............. ..... ......
BIBLIOGRAPHY ..... ....................... ..... .......

APPENDIX

A Passages for the facet test and cloze test.

B Facet test directions and questions .......

C Cloze test directions and questions .......

50
52
52
52
53
58
60
63
66
66
70
74
77

83
94
108

Table

2.1

 

LIST OF TABLES

Eight distractors of a test item in which
figures vary in three facets, A, B, and C ..

Title and source of test passages .
Readability of facet test passages
Facet test mappings into the range
Example of facet test question ....

Facet test statistics ........ .....

Frequency distribution of facet test

000......

..;_..~ ",;—‘_ *_‘.‘_ _ _ _ '_'

Cutting points for five groups on facet test ..

Item reliability on facet test for all subjects

Results on facet test of low, mid, high groups.

Correlation coefficients for facet test,

cloze
test and Davis Reading Test ...................

Per cent response for all distractors on facet

test items for all examinees ......

Range of mapping sentence matched to

distractor statement ...............

Tabulation of error response for students
in lower 40 per cent of facet .................

vi

Page

24
37
38
42
44
53

54

59

61

64

65

LIST OF FIGURES
Figure

2.1 Test item from Guttman and Schlesinger
(1967) .......... . ...........................

vii

CHAPTER I

THE PROBLEM

What is reading?

 

The teaching of reading has been emphasized more
than any other curriculum area. Reading is the only part
of the elementary school curriculum which is customarily
preceded by a systematic readiness program. Reading is
usually accorded a greater portion of instructional time
than any other subject in the elementary school. Ability
to read is acknowledged by many as a prerequisite to learn-
ing in most subject matter fields. Reading has been the
most persistently investigated of the receptive communi—
cation processes. (Gray, 1960, estimated that nearly 4,000
scientific studies of the sociology, psychology and teaching
of reading are available.) That there has been no abatement
in interest among researchers in reading is evident from
Harris (Ebel, 1970, p. 1140), who reports that the number
of scientific investigations completed in the 1960's was
nearly double that of the previous decade and many of these
were conducted by authorities from a variety of fields
outside of education.

In spite of the attention received and in spite

of the importance which our society accords this basic

1

2

skill, the process of reading is poorly understood. The
amount of research and the inconclusiveness as to a univers-
ally acceptable definition and methodology may be the reason
i that most recent textbooks of reading instruction have
chapters explaining the components of the reading process.
However, a comparison of these books reveals disagreement

as to the components of reading, the sequence in which they
should be taught and the method of instruction.

While disagreement exists as to definition and
methodology in reading, there are also aspects of agreement.
Few would argue that two basic components of the reading
curriculum are word recognition skills and comprehension
skills. Jenkinson (1968) says that most studies of reading
comprehension have identified four factors, i.e. vocabulary,
interrelationship among ideas, abstract reasoning, and fac-
tors specific to the special subject areas. There also
seems to be a general acceptance that measurement of reading

performance should accompany the instruction.

Measurement in reading

 

The measurement of achievement in reading has re-
ceived attention since early in the twentieth century, even
though an absolute concensus of the process and components
of reading has been lacking. The primary role of measure-
ment in reading should be to determine if learning has taken
place and to determine specific strengths and weaknesses of

the learner in order to facilitate the planning of future

 

3

instruction. Many authors of textbooks on reading instruction
discuss evaluation and recommend the assessment of student
achievement before planning a program of instruction. They
argue there is very little value in planning a reading lesson
if a child cannot profit from the instruction. Glaser

and Nitko (Thorndike, 1971) suggest that measurement pro—
cedures need to be designed with the information requirements
of a specific instructional system in mind.

Though few educators question the importance of
measurement, issues involving the kinds of test items (mult—
iple choice, true-false, essay, etc.) that should be used
and the kinds of achievement that should be measured remain
unresolved. Objective tests of the multiple choice variety
are most common on the standardized test market, yet they are
criticized by proponents of essay-type tests. Critics of
the multiple choice test argue that there seems to be no
absolute method of ensuring that the foils (incorrect re-
sponses) are less correct than the correct responses. Hence,
producing adequate distractors may be more of an art than a
science. The decision as to whether an answer to a multiple
choice type question is correct or not may depend more upon
who is making the choice than anything else. Bormuth (1970)
points out that a score on an achievement test constructed
by currently used procedures must be interpreted as the
student's response to how the test writer perceives the
instruction. Because so little is known about the factors

that determine test writers' behavior, perhaps we should

 

4

be more concerned with the nature of what is to be measured
than with the form of the measurement.

The second issue concerns the kind of achievement
which ought to be measured. Lindquist (1951) notes that
achievement generally has been based upon textbooks and
that educators have tended to teach what is in the textbook
and test for recall of textbook facts. Ebel (1969) points
out that we have moved from the measurement of facts to
measuring other domains such as the cognitive and affective
ones but he seems unsure as to whether much has been accomp-
lished by this change. He even forecasts a move back to

measurement of the factual information.

Need for this study

 

Educational measurement is primarily used to pro-
vide direction and purpose to educational effort and to
report the degree of success in learning. The foregoing
has pointed out two problem areas: (a) how can something
be measured if it is not clearly defined, and (b) if there
is no assurance of content validity, how can we be sure of
what is being measured? A new approach to the resolution
of this dilemma seems to be appropriate and that is the

purpose of this study.

Purpose of this study

 

This study was an attempt to employ a new method

of multiple choice test construction for measuring reading

 

 

 

5

comprehension. The approach endeavored to ensure consistency
within every test item while Operationally building in the
validity of certain aspects of reading comprehension and

relating this specifically to passage content.

Facet design

 

This study was patterned after the work of
Schlesinger and Weiser (1970) who proposed that a EEEEE
design be used to construct a test of reading comprehension.
Facet design concentrates on the relation of the item to
the information presented in the text or elsewhere, rather
than on skills and abilities which are presumed to be in—
volved in comprehension. What a given item measures is
defined operationally in terms of how it is related to the
information presented in a passage and the reader's abilities
in comprehending it are measured only indirectly.

In facet testing the respondent is probably influ-
enced as much by the distractors of the multiple choice item
as by its stem. This approach presumably acts to systematize
the operation of generating the distractors and ensures that
the correct answer is related to the text. The facets of
reading comprehension defining the relationship of the
distractor—statement to the text are formulated first and
then related to various parts of the distractor. While the
correct distractor—statement will have to be correct in all
aspects, the incorrect distractor-statements will contradict

the text in some aspect or be related to information not

6

included in the text. Even though a distractor-statement
contradicts the text it may still be attractive to a
respondent since one of its constituent clauses agrees with
the text or could even agree with information obtained from
some other source but not stated in the text.

This factor approach provides many advantages over
the intuitive approach normally used in item construction.
Systematic construction of item distractors should not only
provide consistency but ensure that a parallel form of the
test is identical since the items are drawn from the universe
of possible questions. Diagnostic information may be avail-
able and future responses should be predictable if an analysis
of test errors reveals a consistent pattern among the re-
spondent's choices.

Little previous work had been done with the facet
design in the area of reading comprehension1 and this study
endeavored to determine the usefulness of this procedure.

To judge the usefulness of the facet design it was decided
to analyze the relative effectiveness of the distractors
for each item in the test and to compare the facet test
results with achievement based upon other currently used

tests which have been statistically validated.

 

1A research report was published: Schlesinger, I. M. and
Zehavit Weiser, "A Facet Design for Tests of Reading
Comprehension," Reading Research Quarterly, V, 4:566-580,
1970. Upon corresponding with the researchers it was
found that little work had been done with this design in
reading but most of the work had been done in intelligence
testing.

 

Hypotheses

 

l. A test constructed by facet design will not be statist-
ically reliable.

2. There will be no difference among achievement levels
when a test is constructed by facet design.

3. There will be no statistically significant correlation
between the facet test and other measures of reading
comprehension, i.e. between (a) the facet test and
Cloze test, and (b) the facet test and the Davis Reading
Test.

Additional considerations

 

1. Will an analysis of responses to items provide additional
information?

(a) Does the form of information facet (whether in-
formation is stated in exact words or in paraphrase)
have any effect upon examinee response?

 

(b) Does the source of information facet (whether
information comes from the text or some other
source) have any effect upon examinee response?

 

Limitations of the study

 

This research was limited to the extent that the
time allotted for the administration of the Cloze test was
not sufficient. This resulted in this test becoming a
speed test rather than a power test, which had the effect
of reducing the intercorrelations between items on the four
passages and introducing an undetermined error factor into the
correlations between (a) the cloze test and the Davis Reading

Test, and (b) the cloze test and the facet test.

 

Definition of terms

 

l.

facet test: This is a multiple choice test where each

 

item is constructed on a priori definitional grounds
after Schlesinger and Weiser (1970). The items are
constructed in such a way as to relate only to the
content of information being tested and as such are not
dependent upon previously acquired information.

facet: "...two sets of elements, A and B, are called
facets, and their Cartesian space is the set of all pairs
of elements ab where a is an element of A and b is an
element of B. A Cartesian space may consist of any
number of facets, or sets of elements; with n facets,

any one point in the Cartesian space has n component
elements." (Guttman and Schlesinger, 1967, p. 3)

Facets are arbitrary sets of elements within a space

such as comprehension. These can be thought of as sets
of ideas rather than constants. In this study the facets
of comprehension decided upon were (a) form of informa—
tion, and (b) source of information.

text: This refers to the passages on which the test

 

items are based and which students read to answer the

questions.

distractor: This refers to all distractor statements of

 

an item, whether the correct answer or incorrect response.

Overview

A review of the literature pertaining to this study

 

9
is outlined in Chapter II. There is reference made to edu-
cational measurement, tests of reading comprehension, and a
definitive explanation of facet design.

A description of the research design and test
instruments, along with the results of a pilot study are
provided in Chapter III.

A detailed analysis of the data and summary of
the findings are delineated in Chapter IV.

The findings of the study related to the problem

are discussed in Chapter V together with suggestions for

 

further research.

CHAPTER II

REVIEW OF RELATED LITERATURE

The skill of reading is not well defined and
methods to measure reading achievement are currently being
questioned. Though many tests of reading comprehension are

available, their construction may be based more on the

 

intuition of the author than on a definite, predetermined
set of criteria. The facet design provided a plan for the
compilation of a test which could be built upon an a priori
basis. Items for the test are based directly upon content
purposely exposed to the examinee rather than upon informa-
tion which the examinee might have been exposed to through
normal experience. The validity of the test relies upon
the method of item construction to the extent that the
components of the test are considered facets of reading
comprehension. It was felt necessary, however, to make
comparisons among other tests of reading comprehension which

were considered to be adequate measures of this skill.

Measurement in reading

 

An important aspect of our educational system is
the measurement of achievement. Lindeman (1967, p.5) states
that, "Measurement in education is essential if the evaluation

10

11

process is to be accurately and effectively carried out."
E. L. Thorndike at the turn of the century initiated the
concept of comprehension in reading which has resulted in
the construction of various devices and instruments to measure
reading comprehension. Thorndike (1917) said that compre-
hension in reading was much the same as reasoning in mathe-
matics. Cook (Lindquist, 1951) refers to measurement in
education as a means to more accurately observe the individ-
ual, and that tests merely act to quantify certain aspects of
behavior. Lindquist (1951) noted that nearly all of the
standardized achievement tests and informal school examin-
ations constructed to date have been designed to measure
achievement in established school subjects.

Common sense suggests that mere measurement of
behavior is futile unless the information obtained can be
put to good use. It is generally agreed that test data
should be used to help plan educational programs for individ-
uals or foster adjustment of instruction to individual
capacity. Cook (Lindquist, 1951, p.35) believes that "the
determination of pupil status in a given area of achievement
and the adjustment of instruction to status should be a
continuing process in every classroom."

The measuring instrument yields quantitative
information about the individual but the evaluative aspect
of measurement is really qualitative in nature. Even though
the teacher receives a number of correct and incorrect re-

sponses from an achievement test, he must relate these to

 

12

the quality of individual progress. Sometimes this is done
by comparison to what the individual has previously achieved
but often it is a comparison with peers. When tests are
designed to yield specific information of progress, future
instruction is more easily planned. Instruction based upon
exact needs of the individual should allow for more efficient
progress than instruction based uponrelative information.
Heilman (1967) said the only justifiable purpose for using
reading tests is to obtain data about an individual's reading
ability so that a reading program can be built from the data
secured.

In this regard, Glaser and Nitko (Thorndike, 1971)
list four activities of instructional design that influence
measurement requirements, i.e., 1. analysis of the subject-
matter domain under consideration, 2. diagnosis of the
characteristics of the learner, 3. design of the instruction-
al environment, and 4. evaluation of learning outcomes. The
learning must be analyzed in terms of subject-matter content
as well as in terms of the behavioral processes that are
being learned. One must also take into account the entering
behavior of the student, including the extent to which he
has already acquired what is to be learned along with the
necessary prerequisites for further learning. Thorndike
(1971) suggests that if there is lack of a theoretical frame-
work upon which to base instruction, that measurement and

evaluation will be seriously affected.

 

13

Common complaints about tests

 

There have been many complaints made about tests
constructed to measure educational skills including reading
comprehension. One general limitation, not necessarily
inherent in the test, is that knowing the correct answer is
often equated with education. Nonetheless we do find that
what is tested for becomes important and what is not tested
for becomes less important. The weakness could be eliminated
if there were specified objectives known to be essential to
the content or skill being measured. We usually find that
tests are constructed without adequate objectives, which
results in the testing of information that may or may not
be releVant. Lindquist (1951) observed that all standard-
ized tests for school examination have been constructed to
measure achievement but that these were based upon textbooks
as a course of study, resulting in a test for recall of
what was contained in the textbook rather than being based
upon clear cut objectives of what should be taught.

In examining currently available tests many
limitations are evident. Hoffmann (1962) said one limitation
of multiple choice tests is the introduction of ambiguities
resulting from assumptions made by the test author. This
stems from the author's failure to develop a plan from which
to construct the distractors. Hoffmann believes that tests
favor the nimble-Witted, quick reading candidate who forms
fast, superficial judgments. It has been pointed out by

Bean (1953) and Vernon (1962) that reading is a decided

 

14

factor in any multiple choice test and that there is depend-
ence upon the individual's facility with instructions and
his ability to c0pe with the item format. It was stated by
Hunt (1955) that most standardized high school reading tests
measured few reading skills. Holland (1967) found that even
in programmed instruction there was much irrelevant informa-
tion in relation to the test items.

It is interesting to note the opinions of teachers
regarding tests as delineated in the survey done by Austin
and Morrison (1963). The study revealed that much testing
was being done in the schools but that all too frequently
the results were not used to determine present status or
plan for future progress. Test results were often misused
because of a basic misunderstanding of test limitations.
There was dissatisfaction with standardized tests because:
1. tests lacked diagnostic value, 2. tests overestimated
real reading ability, 3. tests did not measure the same
skills taught in the classroom, 4. there was too much
emphasis placed on vocabulary and not enough on comprehension
and other skills, and 5. the tests did not measure higher
level reading skills. While these Opinions may stem from a
misunderstanding of test usage, there are serious implica-
tions for the authors whose major aim is to construct instru-

ments that measure reading comprehension.

What is reading comprehension?

 

It was previously stated that a clear understanding

 

15

of what is to be measured is essential before one attempts
to measure it. Even though numerous studies have been
”conducted with regard to various aspects of comprehension
there still appears to be little agreement as to the nature
and scope of the components which make up the reading comp—
rehension act.

Thorndike (1917) examined mistakes made in para-
graph reading and concluded that the potency of any word or
word group in a question may be far less than its appropri-
ate value in relation to the words which make up the rest of
the question. He said that to understand a paragraph implied
that the weightings of words must be kept in proper prepor-
tion in order to evoke a response which would satisfy the
purpose of the reader.

Berry (1931) summarized the skills of a good
reader. These individuals read for mastery of the general
outline. They are able to relate subordinate details to the
general outline, select key sentences or determine the
topics of selection, visualize what is being read, note new
or difficult terms and concepts, and group major issues and
their implications. Berry noted that good readers are able
to infer the meanings of unknown words from their contexts
and relate the parts of a selection to its whole.

Hunt (1957) used 204 multiple choice items with
370 examinees in an effort to measure six skills related to
reading comprehension. He concluded that only the vocabulary

items were measuring a skill in comprehension that was

 

16

significantly different from the others.

If we assume that comprehension is what comp-
rehension tests measure then a study by Lennon (1968) is
important. He found through a survey of thirty or more
studies of reading comprehension that the following compon-
ents of reading ability can be measured reliably.

l. A general verbal factor related to vocabulary
knowledge. -

2. Comprehension of explicitly stated material,
understanding literal meaning and ability to
follow directions.

3. Comprehension of implicit or latent meaning,
such as the ability to draw inference, predict
outcomes and perceive a hierarchial arrangement
of ideas within a selection.

4. Appreciation which includes the ability to see
the intent or purpose of the author, judge the
mood or tone of a selection and detect the
literary devices which the author uses to
accomplish his purposes.

Davis (1941) surveyed the literature to identify
the comprehension skills deemed most important. He compiled
a list of several hundred skills, many of which pertained
to the mechanics of reading, many overlapped and others
seemed incapable of measurement by objective test items.
After careful scrutiny and removal of some skills (for
reasons noted above) the remainder were grouped into nine
clusters of operational skills thought to be employed in
reading comprehension.

l. Remembering word meanings;

2. Selecting appropriate word meanings in the
light of context;

3. Following the organization of a passage, as
in identifying antecedents and references;

4. Identifying the main thought of a passage;

5. Answering questions for which explicit answers
are given;

 

l7

6. Weaving together the ideas in a passage;

7. Drawing inferences about the content of a
passage;

8. Recognizing literary devices and identifying
the author's tone and mood;

9. Drawing inferences about the author's purpose
and point of View.

Davis (1964) endeavored to measure the skills in
reading using multiple choice tests. He concluded that
there were two general abilities, i.e. memory of word mean-
ings and verbal reasoning; three specific abilities, i.e.
following the organization of a passage, as in identifying
antecedents and references, recognizing literary devices
and identifying the author's tone and mood, and drawing
inferences from content; one general category, i.e. answer-
ing questions for which explicit or paraphrased answers are
given.

Davis (1967) found that word knowledge and reason-
ing seemed to account for virtually all of the variance of
comprehension.

In a subsequent study, Davis (1968) endeavored to
obtain an estimate of the percentage of non-chance unique
variance in the reliable variance of each of the most
important skills of comprehension among mature readers. He
selected eight skills for measurement: 1. remembering word
meanings, 2. finding answers to questions answered explic-
itly or merely in paraphrase in the context, 3. weaving
together ideas in the content, 4. drawing inferences from

content, 5. recognizing a writer's purpose, attitude, tone

or mood, 6. drawing inferences about the meaning of a word

18

from context, 7. identifying a writer's techniques, and

8.. following the structure of a passage. He concluded that
comprehension among mature readers is not a unitary mental
skill or operation. He decided that substantial parts of
the mental abilities used in the eight skills judged to be
of importance in comprehension are independent of one
another.

One can plainly see that reading comprehension is
not a single skill and other investigators have elaborated
on the topic. Jenkinson (1968) suggests that most studies
have identified four factors in reading comprehension,
namely: vocabulary, interrelationship among ideas, abstract
reasoning, and specific content field factors. Schubert
and Torgerson (1968) say that most tests of comprehension
place a high value on memory and further suggested that
standardized tests of reading may not say so much about
comprehension as about the speed of answering questions.
While Langer (1969) says reading is really a thinking process,
Zintz (1970) says comprehension skills are classified as
literal comprehension and interpretative skills, and Harris
(1970) believes that a minimum essential of reading compre-
hension is an understanding of the words used by the author.

Auerbach,(l971),after an extensive review of
comprehension tests and studies of reading comprehension,
concludes that "...there is no clear understanding yet of

what these tests are measuring." (p.45)

19

Need for new theory

 

Disagreement concerning the nature and measurement
of reading comprehension has resulted in a variety of
approaches to evaluation of pupil achievement. The basic
consideration in our schools should be the individual person
and his errors on tests must be given equal consideration with
his correct responses if adequate program planning is to
result. Some specialists believe test errors are the fault
of the test instrument and others suggest that errors are
the result of inadequate psychological processing of informa—
tion. A third group thinks that a definite sequence of steps
must be prescribed for reading instruction and each step
should be criterion referenced for easier evaluation.

Osburn (1968) believes that we must be just as
interested in the wrong answer choice as the correct answer
if we are to provide help to the student. Unfortunately,
as Osburn points out, in practice we have done statistical
analyses of test data in an attempt to generalize to a collec—
tion of test items. But ideally the basis of generalization
should come from the operational definition of the procedures
used in generating and sampling items that will make up the
test. To accomplish Osburn's goal one would have to specify
in advance all items that could possibly appear in the test
and then sample randomly from the universe of items which
are based upon the content. This is theoretically possible

but would be very difficult to accomplish in practice.

Simons (1971) believes that theory based research

 

20
is necessary if basic psychological processes in reading
comprehension are to be understood. The process of compre-
hension has often been confused with the product of compre—
hension since we usually measure the latter and infer the
former. The fundamental problem with traditional reading
tests, according to Simons, is the confusion about what is
being measured. Reading tests could be measuring several
skills and abilities but we are not sure because tests may
lack construct validity and this is difficult to determine.
He advocates more research on the processing of information
since improved instruction depends upon an understanding of
the basic psychological processes.

More recently, criterion-referenced tests have
received more attention. Johnson and Kress (1971) indicate
that in the case of reading tests, a careful analysis of the
task being evaluated is a true prerequisite to the building
of these tests. They suggest that this type of test requires
a careful survey of all the skills and abilities which com-
prise reading. Once the skills and abilities are identified,
each must be defined in terms of both its end result and the
learning it requires and then placed in a hierarchial sequence.
Test items are drawn from each level of the sequence, result-
ing in the determination of whether or not a particular in-
dividual can complete successfully a specified task and at
what level of proficiency an individual is operating in pro-
gressing toward mastery of a particular skill or ability.

Prescott (1971) feels that the common multiple choice test

21

item is inadequate because the correct answer alternatives
"...are possible choices an individual might make in his
ignorance." (p. 351) He further states that multiple choice
test items as they are currently constructed function only
when the individual has some knowledge on which to make his
choice, but if he guesses when having no information the item
measures nothing even if the guess is correct. With criterion—
referenced tests there is an attempt to evaluate performance
in terms of whether an individual has achieved or has failed
to achieve specified objectives. This seems to have merit
since the examiner would be concerned about comparing the
performance of the testee with some concrete standard of
reading performance rather than with other individuals. In
this case one would be just as interested in an analysis of
errors as in the correct answer. The errors would have
diagnostic value to the extent that they indicated specific
reading difficulty.
Bormuth (1970) has recently expressed his Views

on achievement test writing. He favors the construction of
tests by operational definition. He holds the fundamental
premise that:

An achievement test item cannot produce meaningful

results unless it can be related to and derived

from the instruction by a set of operations. The

operations themselves describe a set of publicly

observable manipulations by which a test item is

derived from the instruction. (p.3)

Bormuth is critical of traditional methods of item

writing and points out that no constraints are placed on the

22
writer and no distinction has been made between content and
behavior. There is no limitation on how content is to be
sampled or the number of test items. There is no control
exercised over the relevance of test items to the content
or instruction. Bormuth advocates a method of test con-
struction which would allow the student's score to be compared

with a criterion of mastery of the content and be relevant to

instruction.

The items in standardized achievement tests must
be relevant to the students' instruction. There
is no way to interpret a test score if an unknown
proportion of the items test content not included
in that student's instruction. The score cannot be
compared with the scores of students in the test's
norm group, because many of the students in the
norm group would have been subjected to many
different instructional programs. Part of the
variability in their scores, then, would be due to
the fact that greater or lesser proportions of the
items test content contained in their instruction.
Thus, unless it can be demonstrated that the items
in the test were relevant to the instruction of
both the student being evaluated and the students
in the norm group, the score must be regarded as
confounded. (pp.23-24)

 

Traditional item writing techniques prevent
inspection by others. There is no formula or specific
instructions to ensure consistency which others might follow.
No one working independently can actually duplicate the item
except by chance, nor can they specify what must be measured
by each item.

According to Bormuth's proposal, the phrasing of
an item as well as other aspects of the test would be de-
termined by definition. The content of each item would be

specified beforehand and item writers working independently

23
could replicate each other's work. The rules for relating
items to the instruction would ensure that a test is validly
measuring passage content.

The domain of item writing has been opened to
questioning. There are demands for improved methods. Any
method of item writing must endeavor to ensure validity and
items must be randomly selected from a pool which has defineable

limits. The facet design is an attempt to meet these criteria.

Facet test

 

 

Schlesinger and Weiser (1970) following the work
of Guttman and Schlesinger (1966, 1967) and earlier work by
Guttman (1965) attempted still another approach for measuring
comprehension which makes possible the systematic construction
of test items. It concentrates on the relation of the item
to the information presented in the text or elsewhere but
not on the skills and abilities that are presumably involved
in comprehending the text information. The authors point out
that since we are unable to specify the abilities necessary
for reading comprehension and we cannot measure the processes
but only infer them, then we should concentrate on making
items which are closely associated with the information ex-
posed to the examinee. This design can be applied to multiple
choice type questions and what the item measures is defined
operationally in relation to the information upon which it is
based.

Guttman and Schlesinger (1966) were originally

24

concerned with intelligence testing and reasoned that a
new approach to test construction was necessary to increase
the validity of intelligence tests. Prior work had been done
by Guttman (1965) in defining intelligence. The result was
a new concept labelled facet design. The facets appear to
be sets of ideas rather than constants. Guttman and
Schlesinger (1967, p.3) defined the concept of facet as
follows:

...two sets of elements, A and B are called facets,

and their Cartesian space is the set of all pairs

of elements ab where a is an element of A and b is

an element of—B. A Cartesian space may consist of

any number of facets, or sets of elements; with g

facets; any one point in the Cartesian space has 5

component elements.
An example is provided in Table 2.1 below.
Table 2.1 Eight distractors of a test item in which figures

vary in three facets, A, B, and C (after Guttman
and Schlesinger, 1967)

 

 

 

Distractor Facet A Facet B Facet C

1 al bl cl
2 al bl c2
3 al b2 cl
4 al b2 CZ
5 a2 bl Cl
6 a2 b1 c2
7 a2 b2 Cl
8 a b

N
N
O
N

 

 

25

There can be only one correct answer in which all
three facets represent the facts. Then there are three
other distractors which differ from the correct answer on
one of the three facets, three distractors which differ
from the correct one on two of the three facets, and one
distractor which differs on all three facets. If distractor
l is the correct answer then distractor 2 differs from it
in facet C only, distractor 3 in facet B, and distractor 5
in facet A. Distractors 4, 6, and 7 differ from it on two
facets each and distractor 8 is wrong on all three facets.

An example of a test item taken from Guttman and
Schlesinger (1967) is shown in Figure 2.1 on page 26. In
this example the missing figure of the ninth square must be
provided by the subject. Completing the item requires
examinees to consider the facets, (a) shape, (b) size,

(c) orientation, and (d) place.

The function of a distractor in any item is to be
attractive to the respondent who does not know the correct
answer, according to English and English (1958). For most
test authors, the compilation of adequate distractors is
not only basic but difficult, and yet it is mostly decided
upon by intuition which suggests that item writing may be
described as an art (Davis, 1964).

In the facet design it is argued that distractors
can be constructed in a systematic manner on an a priori
basis. This approach provides at least three useful

features not supplied by less systematic approaches.

 

26

Figure 2.1 Test item from Guttman and Schlesinger (1967)

O Q d

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

27

1. There is prediction of the relative empirical difficult-
ies of distractors, 2. there is less variation in test
results from undesired factors, and 3. there is the possi—
bility of classifying the examinees or determining their
reading difficulties according to the types of answers to
which they were attracted.

The facet design is a way of devising questions
systematically, thus keeping arbitrary decisions to a mini—
mum. Then too the systematic construction of distractors

for all questions is facilitated by the facet design. This

 

assumes particular importance in the construction of parallel
forms of a test. Facet design leads to a more complete
definition of the universe of test items from which only a
sample is drawn for a specific test. Unless some such
design is employed, there could be no way of definitively
predicting whether items included in two parallel forms are
of comparable difficulty, even though they may belong to the
same universe. In the construction of parallel forms of a
test, control over the selection of distractors is of
particular importance. In facet design the distractors are
produced by identical criteria.

Facet design applied to a test of reading compre-
hension concentrates on the relationship between the test
item and the text upon which it is based. The facets of
comprehension are defined and then related to the text
content. Schlesinger and Weiser (1970) chose four facets as

follows:

28

l. the form of information (explicit, implicit);

2. the source of information (the text, formal
instruction, informal experience);

3. the context in which the information appears
(apprOpriate, not appropriate);

4. the frequencijith which the information appears
in the source. (p.571)

 

Guttman and Schlesinger (1967) point out, however, that egg
facet can be applied as well as several facets to this
design as long as the test writer specifies beforehand how
the distractor will be written. The facet design can best
be summarized by means of a "mapping sentence" which is a
formal statement of the research design. (Schlesinger and
Guttman, 1969; Schlesinger and Weiser, 1970)

The mapping sentence used by Schlesinger and

Weiser (1970, p.577) is as follows:

The relation of part (x) of statement (y) to

information appearing {one } time(s) in {explicit}

several implicit
form in the appropriate place in the text 1
not appropriate formal instruc-[
tion I
informal exper—
ience
agreement
-—————9 no information
contradiction

The domain of the mapping sentence (the part that
precedes the arrow) contains the facet by which each dis-
tractor is analyzed. The range of the mapping sentence is
found to the right of the arrow and contains three basic

categories of "agreement", "no information", and "contra-

diction". The sentence is read from left to right with one

29

element being selected from each facet and these are mapped
into the range. Used in this manner the mapping sentence
provides for each part of a distractor-statement (y), as
many mappings into the range as there are combinations of
elements in the domain.
While this design is still experimental, this
study endeavored to assess its usefulness.1 A prediction
of the facet design made by Jordan (1968) seems apprOpriate
at this point.
Facet theory promises to be extremely useful in
determining the content of research instruments.
As this technique becomes more known and available
to the social scientist, it is likely that attitude
research instruments will contain the kind of con-
tent that will allow systematic quantitative pre-
dictions from qualitative data.

The facet test proposes a new approach for measuring reading

comprehension but there is need of comparison with other

existing tests.

Cloze technique

 

The "Cloze" procedure was first introduced by
Taylor (1953) as a measure of readability. Later Taylor
(1956) recognized that the Cloze procedure was also suitable
for use as a measure of the ability of readers to understand
printed materials. Passages to be used for the cloze tech-
nique are prepared by deleting words according to an object-

ively specified procedure. It is customary for words to be

1

 

The facet design as used in this study is explained more
fully in Chapter III.

 

3O

deleted in one of two ways: either all of the words of a
particular type, i.e., semantic or syntactic type are deleted;
or every nth word is omitted. Deleted words are replaced by
blanks which are of uniform length and it is the task of the
subject to place in each blank the word which he thinks best
fits the immediate context and the passage theme. Subjects
may either read the unmutilated passage first and then fill

in the blanks of the cloze test or, in most cases of read-
ability studies, the subject is required to fill in the blanks

without having first been introduced to the unmutilated pass-

 

age. The completed tests are normally scored by counting as
correct only those words which exactly match the deleted
words. An acceptable alternative is to also count as correct
the synonyms of the deleted words.

Research subsequent to Taylor (Jenkinson, 1957;
Rankin, 1959; Ruddell, 1963; Bormuth, 1964, 1965, 1967,

1969) has established the usefulness of the cloze procedure
as a measure of comprehension and has provided data which
attests to its validity and reliability as a comprehension
test.

The cloze procedure has been compared with multiple
choice tests to check on the measurement of comprehension.
The research cited above has been accomplished through this
means to a large extent. However, though cloze has been
compared with some standardized tests (Rankin, 1965), it
has seldom been compared with the Davis Reading Test, as

noted by Weaver (1965). This has been done in terms of a

31

correlation coefficient in this study and is reported in

chapter IV.

Summary

While measurement has been a part of the education-
al scene for many years and has had great importance since
E. L. Thorndike's work in 1914, there is still much dis-
agreement as to whether many tests, especially in reading,
are valid for their stated purposes. Tests have been used
by classroom teachers often as an end result rather than to
promote better instruction, which should be the goal. A
lack of understanding of test limitations on the part of
the users may be the reason for this. It also appears that
many tests in reading comprehension have been constructed
without a clear set of objectives since there is not con-
census as to the definition of reading and all of its com—
ponents. This latter fact has led to criticism of reading
comprehension tests and some question about their validity
and usefulness.

Recently, test conscious educators and theoreti—
cians have proposed new models of test construction based
upon a new theory base that purports to build validity into
the instrument because of its operational definition.
Bormuth, Guttman, Schlesinger, Weiser and others see this
as an essential feature of tests.

The facet design is just one model of test con-

struction which can be used to devise a test of reading

 

32

comprehension. With this model the test items are constructed
systematically on a priori basis or from a preplanned defin—

ition, which acts to ensure validity and internal consistency.

 

CHAPTER III

THE DESIGN AND PROCEDURES

This research study was designed to examine the
use of a facet design to measure reading comprehension. A
forty item multiple choice test based on the principle of
facet design was prepared by the researcher and was admin-
istered, along with two other measures of reading compre—
hension, to a sample of grade ten students. Scores on the
facet test were correlated with scores on the two other

measures of reading comprehension.

ngulation and sample

 

The sample consisted of 186 students chosen from
a grade ten pOpulation who had just entered high school two
weeks previous to the study. These students had completed
junior high school in several feeder schools and represented
a large area of the city surrounding the high school.

Three grade ten English classes were selected from
each of two public high schools in Edmonton, Alberta. Since
all students are required to complete an English course,
these students appeared to be a representative sample of grade
ten students in Edmonton. Some students were enrolled in

courses leading to University entrance and others were not.

33

 

34

Measuring instruments

 

The three measures of reading comprehension used
in this study include: (1) the Davis Reading Test, form 2B,
designed for use in grades eight through eleven, (2) a facet
test, constructed by the writer, consisting of a series of
four passages with accompanying multiple choicequestions and (3)
Cloze test based upon the identical passages contained in
the facet test.

1. Davis Reading Test

 

The Davis Reading Test is a group test designed

 

for grades eight through eleven and is useful in assessing
the overall reading ability of individuals. The test is
composed of eighty items constructed to measure the ability
to use nine operational skills necessary in reading compre-
hension (Davis Reading Test Manual, 1962). The test items
are based on the content of several passages of varying
topics and lengths which high school students might be
required to read. The test consists of eighty items and

provides two scores: A Level of Comprehension score, com-

 

puted on the first forty items, which indicates the depth of

understanding displayed by the student and a Speed of Compre-

 

hension score based on all eighty of the items in the test,
which indicates the rapidity and accuracy with which he
understands this self—same material.

Davis computed the reliability of this test using
parallel forms with a two—to-four week interval between test

administrations. The reliability coefficients for grade

35

ten are reported as .78 for Lgygl and .86 for Speed.

Reviewers (Buros, 1968) say that the method of
construction of the Davis Reading Test seems to ensure test
validity. (Buros, 1968) ‘ Davis surveyed the literature and
compiled a list of several hundred skills which educators
believed necessary for effective reading. Skills pertaining
to the mechanics of reading and those incapable of being
measured by objective test items were deleted. The remaining
skills were grouped into nine clusters. After careful ex-
perimental study, Davis concluded that each of the nine oper-
ational skills is closely related to a basic ability under-
lying accurate comprehension in reading. He reasoned that
memory for word meaning was so pervasive an ability that it
was decided not to measure it separately. The eight remaining
operational skills were grouped into the five categories
listed below. (Davis Reading Test Manual, 1962)

1. Finding the answers to questions answered
explicitly or in paraphrase in a passage;

2. Weaving together the ideas in a passage
and grasping its central thought;

3. Making inferences about the content of a
passage and about the purpose or point of
view of its author;

4. Recognizing the tone and mood of a passage
and the literary devices used by its author;

5. Following the structure of a passage.
Davis reports that this test correlated highly with other
measures of verbal skills and has a predictive validity of

about 0.5 when compared with high school and college English

grades.

 

36

2. Choice of passages for facet and Cloze tests
Passages for these tests were selected from a

variety of materials that were deemed suitable for grade ten
pupils. The Department of Education of the Province of
Alberta has compiled, for each grade level, a list of recom—
mended reading materials for each of the subject matter areas
in the curriculum. Four books which teachers agreed would be
most widely used in grade ten, were chosen from this list.
These books corresponded to the subject areas of Social
Studies, Science and English. One passage was selected from
each of the four books as reported in Table 3.1. Each passage
was self—contained in the sense that it could be read with
understanding without having read preceding or subsequent
paragraphs. The passages ranged in length from 586 words to
614 words. In addition, one article was chosen from a current,
popular news magazine1 that would pertain to either the Social
Studies or Science curricula. (See Table 3.1) The titles
of the passages and their sources are provided in Table 3.1.
Some problems appearing in the pilot study (discussed later)

resulted in the deletion of one passage. (See Table 3.1)

3. Readability study

 

The Fry formula (Fry, 1968) was used to assess the
readability of the passages. The formula is reported as a

reliable measure of readability by Maginnis (1969) and Pauk

(1970).

 

lTIME, May 17, 1971; see Ap endix A.

An informal survey of teac ers.in high school revealed that
Time magazine had often been Cited for related information
dealing with Social Studies.

37

Table 3.1 Title and source of test passages

 

Source No. of

Passage title

words

Curriculum
area

 

Upgrading
Neanderthal
Man

Types of
Municipal
Government

TIME Canada Edition 619
May 17, 1971,
pp. 50-51.

Greason, George K. 586
and Roy C. King;
The Citizen and Loc—

 

al Government, The

 

Macmillan Company of
Canada, Limited,
Toronto, 1964.

 

 

 

Social Studies
and Science

Social Studies

 

How Snow Helps Hardy, W. C., 614 Science
and Hinders Alberta: A Natural
Survival History, M.C. Hurtig

Edmonton, Alberta,

1968.
The Balance Renewable Natural 600 Science
of Nature Resources in Alberta,

Dept. of Education,

Edmonton, Alberta,

September, 1962.

Deleted from study

Across the Heyerdahl, Thor 604 English

Pacific: I

The Kon~Tiki Ex—
pedition, pp. 39-41.

 

 

 

38

 

 

e.w wma Hm
m.m mma oma Asesum gone emumamec
n mpmuw m.m N.m omH baa NHH H ”camﬂomm map mmouom
m.m mwa nma
m.m hma mHH cos
mmmaaoo s.e m.s sea ASH ma Hmenumecmmz mcﬁemnmdp
m.v mma mHH
m.v mma mma Hm>w>mom mumpsﬂm
A means m.s m.m mmH Asa was was mdamm 30cm Rom
m.m mwa oHH
o.o vva em
m momuo o.m v.m ova wma NHH ouswmz mo oosmHmm one
o.m mwa mm
v.m Hma mm DQTEQHT>00
mamaaoo H.m o.m NEH ems ems Hmmnoﬂqsz Lo mmdsn
ommmmmm mamamm omcmmmm 9H onEMm
QH monoz some cw mpuos ooa some cw THQEMm
ooa \mmosm mouo3 ooa \mmanma mpuoa ooa comm
Hm>ma lucmm mo \mmosmunom lawn mo \mmHQmHHmm CH mpHOB
muaaﬂnmpmom .0: new: mo .02 .0c :mmz mo .02 mo .02 mmmwmmm

 

 

momMmmmm Dump umomm mo %uﬂaﬂnmpmmm N.m canoe

39

Three samples of approximately 100 running words
each were taken from each of the passages. The mean number
of words per sentence and the mean number of syllables per
100 running words was determined for each of the passages.
This information, along with the readability levels computed

therefrom, is reported in Table 3.2.

4. Facet test

 

This test was constructed in accordance with the
work of Schlesinger and Weiser (1970) and Guttman and
Schlesinger (1966, 1967). While Guttman's original approach
was aimed at increasing the validity of intelligence tests,
it has since been suggested that this approach is applicable
to evaluating reading comprehension (Schlesinger and Weiser,
1970). The facet design concentrates on the relationship
between the content of the test item and the information from
the text upon which it is based, rather than on the skills
and abilities which are presumed to be involved in choosing
the correct response to the item. Although Schlesinger and
Weiser (1970) suggested that four facets could be applied
this study concentrates on two facets only, namely, form of

information and source of information.

 

 

The form of information facet pertains to whether

the test item distractor contains information explicitly

 

stated or implicitly stated from the text or some other source.

 

By explicit is meant that the information must be stated in

the exact words used in the passage or some related passage.

 

40

Implicitly stated information is defined as a paraphrase of

 

the information stated in the passage or some related source.

The source of information facet refers to where the

 

information for a particular test item distractor was obtained.
The test items were constructed in such a way that each dis-
tractor statement contained either information presented in
the passage or information from other sources and this was
expressed either in the exact words of the source or in para-
phrase form.

Each distractor statement was constructed so that
it conforms to one of three conditions, all of which are
referred to as the range: (a) it agrees with information
found in the passage, (b) it contradicts (disagrees) with
information found in the passage, or (c) there is no infor-
mation in the passage relating to it. To summarize the

foregoing, a mapping sentence such as used by Guttman and

 

Schlesinger (1967), London (1968), and Schlesinger and Weiser

(1970), illustrates this approach.

The variable portion of each distractor
_ (Facet One)
statement Will be a paraphrase
exact words

(Facet Two)
of information found in the text
in some other source

(Range)
Agreement
3 No information
Contradiction

 

The "domain" of the mapping sentence, that part which precedes

41

the arrow, contains the facet according to which each dis-
tractor will be analyzed. The "range" of the mapping
sentence is found to the right of the arrow. The sentence
is read from left to right with one component being selected
from each facet. Thus the facet design provides that each
variable part of the distractor statement will have as many
possible "mappings" into the range as there are combinations
of elements in the domain.

The facet test in this study was constructed
according to the design of Schlesinger and Weiser (1970)
except that only two facets are built into this test. If
one were to let a = paraphrase, a = exact words, b =

l 2 1

text (passage), and b2 = other source of information, the

possible combinations of the variable component of the

statement would be a b (paraphrase from the text), a b

l l l 2
(paraphrase from some other source), a2 bl (exact words
from the text), a2 b2 (exact words from some other source).
Thissxﬂuﬂmeis shown in Table 3.3. Each distractor (variable

component of the statement) thus includes two facets. The
variable part of the domain can be mapped into the range in
such a way that the total statement, including both non-vari-
able (stem) and variable (distractor) components, is (a) in
agreement with information found in the passage, (b) contra-
dicts information found in the passage, or (c) includes
information not found in the passage.

Two distractor statements for each item contain

the subfacet b2 where information is taken from a source

 

42

other than the passage. One distractor is closely related
to the theme of the passage which the student reads, while
the other distractor is taken from a passage somewhat un-
related to the theme of the passage. As a result, in the
combinations al b2 (paraphrase, other source) or a2 b2
(exact words, other source), one distractor contains
information related to the passage theme and the other

distractor comes from a passage with little or no relation-

ship to the theme of the original passage.

Table 3.3 Facet test mappings into the range

 

 

 

Facet Parts Symbol Variable Symbol Range
Combinations
1. Form of paraphrase al paraphrase albl agree-
information (implicit) text ment
exact words a2 paraphrase alb2 contra-
(explicit) other source diction
2. Source of text bl exact words a2bl agree-
information (passage of text ment
test)
other source b2 exact words a2b2 contra-

other source diction

 

For example, question number one from the passage,
"Upgrading Neanderthal Man" (Time, May 17, 1971), is given

below as it appeared in the test.

43

From the first discovery of Neanderthal's bones

a. there has been a connection with Dr. Ralph
S. Solecki, who led an expedition to Lebanon
last summer.

b. his name has been associated with crudely
primitive in manner or conduct.

c. his name has been synonymous with brutishness.

d. Darwin's theory has withstood a century of
criticism.

There were ten questions constructed for each of
the four passages (see Appendix A) so that the entire test
contained forty items. Within each item only two kinds of
distractors could possibly be accepted as correct responses.

Only those distractors where the facet - source of information

 

- came from the passage which the students read could be
correct, since by definition the correct answer was either a
paraphrase or exact words from the passage.

Table 3.4 explains the construction of the items

used in the facet test.

44

 

.uwop may maﬂxmu mucopoum map an coon ommmwmm can on usowsoo ca pmumHmH
umn3oEOm maze mmB 50H£3 Haﬁuopmﬁ Eoum @800 psmEmpMDm mwsu uwnp mamme pwumawu #02 ««

 

.ummu mnu maHxMD mpswpsum was ho pooh ommmmmm Hmsuom
may on oEocp HmHHEHm QDH3 ommmmwm m EOHM oEmo psoETDMDm wasp umzu mange pmumaom «

 

 

«kpmumamu #02
.nas .d .smma .xuo» smz
.mcmmﬁoo cmaHHEomz one
.NmoaooN mmoaaou ..4 .M

 

cOHpm .EmHOHDﬂHo mo muspsmo

 

.maﬂom can .3 .m .Hmsmmm IEuomcH oz m coonmsuﬂs mm: muoonu m.cH3uma mg «m
.mmoccmﬂpsun
oommmmm momum< ADHB mSOESEOQSm soon we: mama was HQ mm
«pmumamm .uoopcoo
mm .m .mapmmoaomosm soaps Mo HmcsmE EH m>ﬂpﬂEHHm mampsuo
xoom pauoz .vH mEdHo> IEuomcH oz EDHB UTDMHOOmmm soon mm: mam: was Nb Hm
.Hoaﬁom pmma cosmnoq on COHDH
coon ou upmmxm so Ema 0:3 .ﬂxooaom .m smamm
mucmpowm 0D sm>Hm mmmmmmm mpowpmuusou .HQ EDHB cospomccoo m sown mm: mumnu Ho Ho
sowumcﬂnaoo
soauman0mcﬂ mo condom omscm Ausoaopmpm THQMHHM>V Houomupmﬂo umomm

 

meson m.Ho£#Hmpsmmz mo mmo>oomﬂp pmuﬂm may Eonm

OOH—90m HOSUO I Q

"AEmDmv COHDHOQ THQMHHM>GD

N a N H

.pxop I Q .mUHOB pooxo i m .omonnmmumm I m

 

mco .oz coﬂummso so: ngunmpcwmz mcﬁpmanD Homemmmm

soﬂumosv umou Doomm mo onmem v.m magma

45
5. Cloze test

 

The cloze test, first constructed by Wilson Taylor
(1953), is an instrument designed to measure comprehension
and readability of materials. In recent reviews, Rankin
(1965) and Bickley, et_al. (1970), conclude that Cloze tests
are valid and reliable measures of reading comprehension.

A cloze test was constructed using the passages from
the facet test previously discussed. Counting from the first
word in the passage, every tenth word was deleted. Hyphenated
words were counted as one word. In order to provide adequate
context, Culhane (1970) recommends that the first sentence
in a Cloze passage remain unmutilated. In this instance,
however, the unmutilated passages had previously been read
during the administration of the facet first and that appeared
to provide sufficient context to permit the mutilation of
passages, counting from the beginning of the passage.

Each deleted word was represented by a line equal in length

to ten typewriter spaces. Taylor (1956) recommends that Cloze
tests contain at least 50 items (deletions). Each of the
subtests (passages) contained from 58 to 62 items (deletions)
for a total of 240 items in the test.

The scoring of these passages was modified slightly
from the usual exact word system. Proper nouns with only one
letter missing from the exact order, such as Soleck for Solecki,
were counted as correct. Plurals were accepted if it appeared
difficult to distinguish from sentence context whether a plural

or singular form was required, provided number (singular and

46

plural) did not influence the meaning of succeeding sentences.
Words at the beginning of sentences were counted correct re-
gardless of capitalization. The administration of the Cloze
test followed by one week the administration of the Davis

Reading Test and the facet test.

Pilot studyl

 

Twenty-three pupils enrolled in a grade ten math—
ematics course during summer school 1971 were chosen as
subjects for a pilot study. The summer school was adminis-
tered by the Edmonton Public School Board. There were both
male and female pupils registered in the course and regular
academic (grade placement ranged from ten to twelve.

Equal numbers of facet and Cloze tests were distributed in a
random order so that twelve students had the facet test and
eleven students completed the Cloze test.

Five passages were selected from materials that
grade ten students might be expected to read. A Cloze test
and a facet test were constructed for each passage. The
cloze test for the five passages consisted of 302 items.
Each passage in the facet test was followed by ten items for
a total of 50 items.

The cloze tests and facet tests were organized in
a similar manner. In the Cloze test each unmutilated passage
was followed by a mutilated version with every tenth word

deleted. The five passages with corresponding cloze tests

 

1The choice of passages and test construction have previously
been described.

 

47

were then put in a folder in the following order: Across the
Pacific: I, Types of Municipal Government, How Snow Helps and
Hinders Survival, Upgrading Neanderthal Man, The Balance of
Nature. The facet test was assembled in the same order ex-
cept that facet questions followed the respective passages.
Seventy minutes were allotted for the tests and
students were permitted to leave when finished. All students
wereable to complete the facet test but three students
failed to complete the Cloze test within the time allotment.
The pilot study showed that some modification was
in order. Because testing time to fit the school schedule,
was restricted to forty-five minutes, it would be necessary
to shorten the tests. Weaknesses revealed by item analysis
showed that revisions in one of the passages would be too
extensive to be practical. Accordingly the passage, entitled,
"Across the Pacific: L was deleted from both the Cloze test
and the facet test. Because some students proved to be slow
workers, the passages and test items in the Cloze test and
the facet test were arranged in rotational order to avoid
bias which might stem from the inability of some students
to read the latter parts of the tests within the time limits.
The pilot study revealed some weaknesses among the
distractors used in the facet test. It showed that distractors

dealing with information taken from another source, although

 

unrelated to the theme of the test passage, should appear
highly plausible if they were to be attractive to the examin-

ee. Sometimes it was necessary to revise the distractor

48

containing information from another source to make it more

 

explicit, at other times it was deleted in favor of a more
attractive alternate that would better fit the context ident-

ified by the invariable component of the statement.

Administration of test instruments

 

The tests used in the study were administered in
the third week of the fall term of the academic year 1971-72.
Both high schools from which subjects for the study were
selected operate on the semester system. Class schedules
provide for ninety minute periods with short rest periods
at the mid point. Two testing sessions were scheduled for
each school, one week elapsing between sessions.

In the first session students were given the Davis
Reading Test during the first half of the period and the
facet test in the last half of the period. Time limits for
the Davis Reading Test are forty-five minutes. The mid-point
rest period was used to collect and distribute testing
materials.

Directions for administering the facet test spec-
ified that students were to read the passage and then respond
to the multiple choice items which followed without referring
again to the passage. Students were to proceed with the
passages in the order in which the passages were assembled in
the test folder.

One week hence the cloze test was administered to

these self-same students. Students were directed to read the

49

first unmutilated passage and then respond to the cloze test
for that passage, in that order, and then proceed to the next
passage in the same manner until the test was completed or
time was called, whichever came first. Students were not
permitted to leave the room until the time limit expired.

The passages for both the facet and cloze tests were
assembled so that each unmutilated passage was followed by
the test for that passage. The sequential arrangement of
the materials with each test was systematically varied so
that not all students dealt with the passages in the same
order. The testing materials were randomly distributed and
students were instructed to proceed according to the order

in which the materials were received.

Analysis of data

 

Data processing was done at the University of
Alberta computer center.

The Kuder—Richardson Formula 20 was used to
determine the reliability of the facet test. Thorndike
and Hagen (1969) suggest that there is no universally
accepted minimum value of this coefficient but that the higher
the value the better satisfied is the test author. Davis
(1964) says that many researchers set 0.50 as the minimum
acceptable value of a reliability coefficient. When groups
exceed 100 in number, a reliability coefficient of 0.60 was
shown acceptable by Thorndike and Hagen (1969). Accordingly

a reliability coefficient of 0.60 was set as the minimum

50

acceptable value for the facet test.

An item analysis provided measures of the difficulty
and the discriminatory power of each of the items in the
facet test. The difficulty index is expressed as the per cent
of students choosing the correct distractor for an item.

The biserial correlation coefficient and the item reliability
index are used to show the degree to which an item discrimin—
ates between high and low achievers.

The Davis Reading Test was used as a criterion
measure for grouping students according to achievement levels.
Students whose scores exceeded a point one-half standard
deviation above the mean were classified as high achievers.
Students classified as low achievers earned scores which
deviated more than one-half standard deviation below the
mean. Students with scores falling between one-half standard
deviation below the mean and one-half standard deviation
above the mean were designated as the mid group. An item
analysis was completed on the facet test for each group.

Finally, a Pearson product-moment correlation
coefficient was computed to determine the relationship among

the three measures of reading comprehension.

Summary

This study was concerned with the development of a
test based on the principle of facet design for measuring
achievement in reading comprehension. The test, along with

two other measures of reading comprehension, was administered

51

to a sample of grade ten students enrolled in two of the
public high schools in Edmonton, Alberta. This chapter
describes these tests and how they were constructed as well
as how they were administered. A pilot study which aided

in the research is also described.

CHAPTER IV

RESULTS

Introduction

 

A multiple choice test based on facet design was
prepared and administered to a sample of tenth grade stu-
dents. The test was constructed on an Operational plan to
ensure that each item was related to information in the
text. Test results were analyzed and are reported in this

chapter.

Results concerning hypothesis one

 

A test constructed by facet design will
not be statistically reliable.

A Kuder-Richardson Formula 20 reliability coeffic-
ient was computed on respondent's scores from the facet test.
Table 4.1 lists the number Of subjects, the test mean,
standard deviation and correlation coefficient of 0.80. An
acceptable value for this coefficient was previously set at
0.60 .

It was concluded that this test was a reliable

measuring instrument.

 

lSee Analysis of data, Chapter III.

52

53

Table 4.1 Facet test statistics

 

 

 

Number of test mean standard Kuder-Richardson 20
subjects deviation reliability
185 25.67 6.13 0.80

 

Results concerning hypothesis two

There will be no difference among achievement

levels when a test is constructed by facet

design.

Student responses to the facet test were analyzed
by means Of item analysis. Table 4.2 shows the raw score
distribution. A further clarification Of this information
is provided in Table 4.3.

Five groups were formed, each representing a
twenty per cent segment Of the subjects taking the test.
Group one is the twenty per cent with the lowest scores on
the facet test and group five is the twenty per cent with
the highest scores. The other groups represent similar
divisions within the range. Table 4.3 summarizes inform-
ation tO show that each group contains substantial numbers

Of scores and indicates a distribution Of scores according

to achievement.

54

Frequency distribution of the scores on the facet

test

Table 4.2

 

 

Frequency

*
*1.* *in* *1.* *

*
*

30
26
22

*

18

14

10

 

16 20 24 28 32 36 4O

12

Class interval

 

55

Table 4.3 Cutting scores for five groups on facet test

 

 

Group
1 2 3 4 5
Score 21.5 21.5-24.5 24.5-27.5 27.5-30.5 30.5
number of
subjects 41 33 40 32 39

 

Table 4.4 summarizes information computed about
each item. The item difficulty index is the per cent of
students choosing the correct item response. High values
indicate many students selected the correct response. How
well the item discriminated between higher achievers and
lower achievers is indicated by the bi-serial correlation
coefficient and the item reliability index. Values for the
bi-serial correlation coefficient are seldom greater than
1.0, while values for the item reliability index do not
exceed 0.50.

The item reliability index combines both the item
difficulty index and bi-serial correlation into one number
that expresses the adequacy Of an item. It has been found
by researchers at the University of Alberta that a reliabil-
ity index less than 0.12 indicates extremely weak items.

There were thirteen items with an item reliability
index less than 0.20 (Table 4.4). Six Of these had an item

difficulty index less than 0.50, indicating that for these

56

 

xmpcﬁ

 

 

 

 

 

HN. NN. MN. NN. mN. nN. «N. mN. ma. ON. mN. ma. ON. NH. no. ma. ha. Ha. «a. HN. muHHHQMHHOH
smua
DGOHOHMMOOO
me. we. me. om. Nm. em. mm. av. Hm. Ne. m5. Nv. mm. mN. ea. om. em. om. mm. mv. COHDMHOHHOO
Hmﬁnmmiﬂn
xmpsﬂ
Hm. mm. no. em. mm. mm. mm. em. em. mm. mm. Nm. mm. om. mN. mv. mm. mm. mm. mm. muHSOHmMHp
Emuﬂ

ow mm mm mm mm mm vm mm Nm Hm om mN mN SN wN mN «N MN NN HN
xmpca
HN. mN. ma. mN. Nm. «N. mN. mN. «a. HN. ma. mN. mN. MN. ON. mN. oN. ha. ma. «N. mpﬂaﬂnmﬂamu
amps
Dcmﬂowmmooo
me. am. mm. mm. vb. Nm. mm. mm. as. vv. mm. Hm. mm. we. ov..mm. «m. em. Nm. mm. COHDmHOHHOO
HMﬂHmmIHQ
xmpcﬂ
mm. mm. ON. mm. mm. Nm. mm. mm. mm. mm. Nm. av. mm. mm. mm. mm. mm. mm. Nm. Nb. NDHSOHMMHO
Eoua

ON ma ma 5H ma ma ea ma NH Ha 0H m m n m m w m N H

HODESG Emuﬂ

muomngsm Ham mom umou Doomm so muﬂaﬂnwﬂamm EODH v.w canoe

57

six items the correct distractor was not a popular choice
with the majority Of students. Four items had an item
difficulty index greater than 0.80 but an item reliability
index less than 0.17. These latter four items, while chosen
by most students, did not discriminate between the better
and poorer students and must be classed as the weaker items.

- Three groups were formed from the scores on the
Davis Reading Test Level of Comprehension scorel. These
groups were labelled low, mid, and highz. An item analysis
Of the facet test was completed for each of these groups.
The number of students, mean score, and standard deviation
for each group are listed in Table 4.5. The groups have
well defined limits and indicate that the facet test does
discriminate among achievement levels.

A Kuder-Richardson 20 reliability coefficient was
computed for each group and is reported in Table 4.5. The
acceptable criterion Of 0.60 for this value was surpassed
for each group even though the variance was considerably
reduced as compared with the total group.

The conclusion drawn from the data summarized in
Tables 4.2, 4.3, 4.4, and 4.5 was that test items construct-
ed by facet design discriminate among achievement levels

and leads tO a rejection Of hypothesis two.

 

l The Level Of Comprehension score was chosen because almost

all students completed the first forty items of the Davis
Reading Test upon which this score is based. This would
seem to provide a more accurate measure Of reading compre-
hension than if all eighty items were used where many Of
(the students did not complete the test.

2 See Chapter III, Davis Reading Test.

58
Table 4.5 Results on facet test Of low, mid, high groups*

 

 

group no. mean standard K-R 20
deviation reliability

low 54 20.22 5.06 0.685

mid 66 26.12 4.34 0.617

high 65 29.74 4.99 0.762

 

*
These groups were determined by scores on the Davis Reading
Test.

Low group - 1/2 s.d. below the mean

Mid group - between _ 1/2 s.d. from the mean

High group - 1/2 s.d. above the mean

Results concerning hypothesis three
There will be no statistically significant
correlation between the facet test and other
measures of reading comprehension, i.e., between
(a) the facet test and cloze test, and (b) facet
test and the Davis Reading Test.
A Pearson product-moment correlation coefficient
was computed to determine the relationship between (a) the
facet test and the cloze test, and (b) the facet test and

the Davis Reading Test. The results are reported in Table

4.6.

59

Table 4.6 Correlation coefficients for facet test, cloze
test and Davis Reading Test

 

 

N cloze test Davis Reading Test a.

 

facet test 150 .460* .664* (level)

.712* (speed)

 

* correlation coefficients were significant beyond the .01
level Of confidence

a. The Davis Reading Test yields two scores: (a) level,
items 1 - 40, (b) speed, items 1 - 80.

The Davis Reading Test yields two scores. The
score on the first forty items is designated as the level Of

comprehension, and the score on the total test as the speed

 

of comprehension. The facet test correlations with both

 

scores on the Davis Reading Test were respectively .664 and
.712. These were significant beyond the .01 level of
confidence.

The correlation coefficient Of .460 between the
facet test and the cloze test was significant beyond the
.01 level Of confidence.

It was concluded that a significant relationship
exists between (a) the facet test and the Cloze test, and
between (b) the facet test and the Davis Reading Test, thus
rejecting hypothesis three. These tests appear to be

measuring similar aspects Of reading comprehension.

60

AdditiOnal considerations

 

There seemed to be more information that could be
gleaned from the study that might be Of value. The questions
to be answered are related to the facets of the test.

a. Does the form Of information facet (whether
information is stated in exact words or in
paraphrase) have any effect upon examinee
response?

Information from the item analysis was useful in
answering this question. The item difficulty index for the
facet test summarized in Table 4.4 was compared with the

per cent response for each distractor found in Table 4.7.

The twenty items having exact words from the text

 

(a2 bl) as the correct answer had an item difficulty index
greater than 0.50 and eight Of these items had an item
reliability index less than 0.20. This is to say that all
items with a2 bl as the correct answer had over 50 per cent
response from the students taking the test, with eight items
being poor discriminators between better and poorer students.

The twenty items having paraphrase from the text

 

(al bl) as the correct answer had only eleven items with an
item difficulty index greater than 0.50 and five of these
items had an item reliability index less than 0.20.

The paraphrase items appear to be more difficult
for examinees. Test items with exact words from the text
seem to be easier to detect than a paraphrase from the text.
This may be due to the difference in language structure used

in paraphrase questions or it may be that exact words from

 

61

Asm>wm mum

haco mumaﬁss 0H0£3 mocﬂm 00H Hmuou no: Op mnEsHoo oﬁomv

Emuw may now nozmcm uoouuooa

 

 

 

 

 

thsom
a ma m 4 NH ea ca 4 ma oa N A m m mm mm H o H N NnNm umnuo .mcuoz.uomxm
mm as mm a a «N m am am No.6m 6H mm m 6N a mm «a mm mm HnNm uxmu .meuoa nomxm
# c a a a a a a a s
0H m 4H m AN 0H m m m NH m H N mm m N m m A NH Nada mousom Lasso .mmmuaamuma
ma am on Na mm am as mm NN a 6 Am N as mN as am N A NN HnHm axon .mmmunamuma
yo an ¥ ¥ ¥ % ﬂ k. #5 *0
ow mm mm km mm mm «m mm Nm am On mN mN NN 6N mN 4N mN NN AN
Aummu ca
mcﬁummmmm Hopuo can Dosv
QUHSOm
H 8 AN m ma m m m m m mN NH N A N o m ma A m NnNm umnuo .meuoz Lumxw
mm 4N mN m a HN.NN mm mm No mm ma ma mm as as No mm a m HQNE uxmu .meuoa Lomxm
¥ ¥ ¥ .1 ¥ ¥ k. ¥ 3.. ¥
4 HH.Nm NH m mm NH N N MN m 4N m m 6 oH an N No HA Nana mUAEOm nonuo .mmmunmmuma
m mm 0N ms me am a m m 6 am He NA m 6H m AN ma Hm Na Anam Exmu .mmmuaamuaa
.4. u? 9% um .4. * k. 30 ¥ 3..
oN ma NH NH 6H ma 4H NH.NH HA oH.m m a o m a m N A coaumaansoo
nonﬁdc Eouﬂ uouomuumwn

%

mmwsﬂﬁmxm Ham Mom mamas ummu uwomw so muouomuumwp Ham mom mmaommmu usmo mom b.¢ magma

62
the text are much more easily distinguished from any other
distractor.
b. Does the source of information facet (whether
information came from the text or some other
source) have an effect upon examinee response?

Test responses from ten students scoring in the
lowest 40 per cent Of scores on the facet test were analyzed.
Only the incorrect responses were chosen for analysis. The
correct distractors for any item would always have informa-
tion from the text as their source of information facet but
the incorrect distractors would have information from another
source as the source Of information facet two-thirds of the
time.

The answer for each distractor statement had a
range of agreement, contradiction and no informationl. The
range for each distractor, the facet map for each distractor
and the distractor choice for each item are reported in
Table 4.8. By matching student responses to the facet

combinations on each item it was possible to determine if a

student was constantly selecting a particular source of

 

information. The error responses Of the ten students are

 

categorized in Table 4.9.

Every student's responses are different, with few
apparent patterns developing. It appears that the better
achieving students from among this group (students falling
at the 40th percentile or below on the facet test) choose

distractors where the source of information was related to

 

 

1 See Chapter III

63

the passage theme. Better students can probably detect
information that has come from another source.

An analysis of responses from grouped data and
from individual responses does appear to provide additional
information about the facet. There is an indication that
paraphrased information is more difficult to comprehend than
exact words and also that information not closely associated
with the passage theme is more easily detected in the item

distractors.

Summary
The results Of the data analysis for this study

were presented in this chapter. It was concluded that:

1. the facet test was a reliable measure of reading
comprehension; 2. the facet test discriminated among
achievement levels; and 3. there was a significant relation-
ship with other measures Of reading comprehension, i.e. the
cloze test and the Davis Reading Test. A further analysis

Of responses from group data and for selected individuals
suggests that paraphrased information is more difficult to
comprehend than exact words and that information not closely
associated with the passage theme is more easily detected

in the item distractors.

64

.mEmﬁu OmMmmmm ou

woumHou

umou o:u.ao sowumcmmep uouowuumﬁp I ease
mEozu mmmmmmm on poumHOH no: sowumﬁnomsw I z

cowumﬁuomcw I m .muoapmuusoo I O .mooumm I m .omcmu I as

 

mZ

ﬂﬁ

UN

00

pm

UN

0%

.mU

@Z

00

cm

00

NZ

usoﬁmumum nouomuumwp on ponouwﬁ moaousom msﬁmmmﬁ mo

©Z
DU

mm

.um
3

92

pm

Qﬁ

NZ

pm

0<

QZ

ha

Um

HH

OZ

mm

Gd

UZ

pm

$0

CZ

00

Md

CZ

90

mm

UN

6m

00

NZ

02

QZ

MO

6%

Cd

Um

Qﬁ

NZ

00

|||||||||||| *¥Ioﬂloﬂ¥ﬂ

mN

Um

MU

mN

Um

8

FN

Um

mﬂ

OZ

0N

UZ

ha

Um

mN

Um

04

«N

CZ

Um

8N

OH

nonﬁdc Bond

MN

0%

UN

GZ OZ
mm wﬂ
om AM
no mo
NN AN
QZ NZ
00 U0
mm cm
mm 02
Isswﬂaﬂ
N H

NQ Nm
HQ NM
NQ am

an am

an an

ace umomm

wmcmm m.¢ OHQmB

65

Table 4.9 Tabulation of error response for students in

lower 40 per cent of facet test

 

 

v1

 

Student Error response on facet items

c":l Rb NC

1 _ 12 8 8
2 9 10 4
3 l3 5 8
4 6 9 7
5 9 6 7
6 8 7 6
7 ll 7 2
8 8 8 2
9 ll 5 0
10 7 6 2

 

the answer is a contradiction of information stated in
the text ‘

the information came from another source related in
theme to the passage content

the information came from another source with no
relationship in theme to the passage content

CHAPTER V

SUMMARY AND DISCUSSION

Summary of study

 

The purpose Of this study was to measure reading
comprehension by means of a test prepared from a facet
design. There were four major questions from which three
hypotheses were formulated and the fourth question was
considered to the extent that information was available.
The questions are stated as follows:

1. Will a test constructed by facet design be a
statistically reliable measure of reading
comprehension?

2. Will a test constructed by facet design dis-
criminate among achievement levels?

3. Does a test constructed by facet design show
significant relationship tO other measures Of
reading comprehension?

4. Will analysis of responses to the items provide
additional information about the examinee?

(a) Does the form of information facet (whether
information is stated in exact words or in
paraphrase) have any effect upon examinee
response?

(b) Does the source of information facet (whether
information comes from the text or some other
source) have any effect upon examinee
response?

 

 

A search of the literature revealed a lack of
agreement as to the definition Of reading. Many authorities

had written about it and much research had been conducted
66

67

but there was still no concensus on the definition.

Literature pertaining to measurement in education
contained controversy about testing. At question was the
validity Of tests as well as the type Of measurement instru-
ment to be adopted. More recently, authors were advocating
a new approach to measurement that would assure a valid
testing of content and instruction.

A facet design for test construction was advocated
by Schlesinger and Weiser (1970) and adapted by the investi-
gator for this study. Through this design it is possible
to systematically construct items to measure reading compre-
hension. The facet test concentrates on the relation of the
item to the information presented in the text, rather than
on skills and abilities presumed to be involved in reading
comprehension. The items in this test consisted of a stem,
or non variable part, and four variable (distractor) portions
made up of two parts. Each part of the distractor stands in
the relation Of agreement, contradiction, or no information
to the passage.

The tests were based upon four passages taken from
information that students could have been exposed to or
might have been required to read. The passages were approx-
imately 600 words in length and varied in readability from
seventh grade to college level. There were ten questions
prepared from each passage, yielding forty items for the
test.

The facet test was administered to 185 tenth grade

68

students from two high schools in Edmonton, Alberta, Canada,
and the results were compared to results from two other tests
Of reading comprehension administered to the same sample.

A cloze test was constructed on the identical
passages as those of the facet test. Every tenth word was
deleted from the passage. Students were given the original
passages to read first and then required to fill in the
blanks with the exact word in the mutilated passage. The
results were compared to the facet test scores.

The Davis Reading Test was administered to all
students in the sample and the results compared to the facet
test scores.

The Kuder—Richardson Formula 20 reliability co-
efficient was calculated for the facet test.

The scores of the facet test were analyzed by
means of item analysis. A frequency distribution was pre-
pared and an item difficulty index for each item was cal-
culated as well as the bi-serial coefficient and item
reliability index. The item analysis provided evidence as
to the quality Of the facet test based upon respondent's
scores.

A Pearson product-moment correlation was computed
between the scores Of the facet test and cloze test and
between the facet test and the Davis Reading Test to
determine the degree of relationship. It was decided that
additional insights might be gained through a non—statistical

analysis.

69

An additional consideration was whether an analysis
of responses would provide further information about the
test. The item difficulty index was compared to the item
reliability index for each item to determine if the form of
information affected examinee response. Ten students were
selected from the lower forty per cent Of scores on the
facet test. The error responses Of these students were
analyzed in terms of the distractor combination chosen to
determine if the source of information affected the examin-
ee's response.

The analysis of data led to the following con-
clusions:

1. A Kuder—Richardson Formula 20 reliability coefficient
disclosed the facet test to be a statistically reliable
instrument for measuring reading comprehension.

2. The facet test did discriminate among achievement levels
of examinees.

3. There was a significant relationship between other
measures of reading comprehension. The Pearson product-
moment correlation coefficient indicated a significant
relationship between the facet test and cloze test and
between the facet test and the Davis Reading Test.

4. The test data suggested that paraphrased information was
more difficult to comprehend than explicitly stated
information and information from another source not
closely related to the passage theme was more easily

detected in item distractors than information closely

70

related in theme to the text.

DisCussion

 

The reliability of a test can be considered in
two ways. The first way is to think of it as consistently
measuring whatever it measures. This was computed by the
Kuder-Richardson Formula 20 and reached an acceptable level
for this research. Another concept of reliability is that
related by Thorndike (1971), indicating that a test must
reliably measure the content upon which it is based. This
was built into the test by means Of Operations which system-
atically put passage content into the item distractors.
This ensures that the test items are related to the text.
The close association between item content and passage
content was accomplished by means Of an Operational approach
called £3332 design.

One facet Of reading comprehension applied in this

test was the form Of information. The information in a

 

distractor was either exact words (explicit statement) or a
paraphrase (implicit statement) of information. Two
distractors in every item had to contain information from the
text stated in one Of these ways. One Of these distractors
was the correct answer. The other two distractors also had
either a paraphrase or exact words, though the information
was taken from another source.

It was suggested from the data that paraphrased

information might have caused distractors to be more diffi—

71

cult than when information was in exact words. Carroll (1927)
provided information that was in agreement with this finding
as he concluded from a study dealing with detailed directions
that material explicitly stated caused less errors than
material implicitly stated. However, the form of information
was only one facet considered in this test.

The second facet was source Of information. The

 

information could have come from the text or some other
source. Only when information came from the text was the
answer correct. InfOrmation from another source was by
design either closely related to passage theme or unrelated
to passage theme. The data suggested that information closely
related to passage theme rendered the distractor more accept-
able to examinees.

The pilot study provided worthwhile information
concerning the attractiveness Of items. It was discovered

that if information from another source were used in an item

 

distractor it had to be closely associated with the ideas in
the passage. One could paraphrase information from another
source only if it seemed to bear a close relationship to the
stem Of the item or be closely associated to something said
somewhere in the passage. The researcher found that an idea
from part of the passage, not related to the item stem, might
be attractive to the examinee even if the exact information
about the idea came from another source. It was also dis-

covered that if the facet combination was exact words from

 

another source (a2 b2) there had to be a close relationship

 

72

of theme between the information in the other source and
information in the test passage. Item compilation was an
onerous task using the two facets of this study.

The item distractors, in order to be attractive
to the examinee, needed to be plausible answers. The only
way that an a2 bl combination (exact words from the text)
could be an incorrect answer was to select information
unrelated to the stem of the item. This proved to be very
difficult tO accomplish and is probably reflected in the
data, as this combination was chosen a high percentage of
the time. The al b2 combination (paraphrase from another
source) was always an incorrect answer since information did
not come from the test passage. This distractor combination,
if related to the passage theme and the stem of the distract-
or, could only be false if it were made so on purpose or if
it was lifted from its original context so that it lost
original meaning. To make such an item too attractive would
defeat the purpose of the test.

The distractors had to be constructed in such a
manner that there was only one correct answer. If a dis—

tractor combination Of exact words from another source were

 

true and almost identical to information found in the passage,
students had no possible method by which to discriminate it
from the correct distractor statement. This problem could
have led to some distractors discriminating negatively

between higher and lower achievers.

 

73

The facet test, as a whole, did discriminate among
achievement levels but some items did not discriminate well
between the higher and lower achievers: None Of the items
achieved a high discrimination index as described by Ebel
(1965). Some Of the items with a discrimination index less
than 0.20 may need revision. It must be noted, however,
that these items are valid in terms of test content and to
dispense with the item is to reject the whole item class.
Osburn (1968, p. 102) says that "any decision to exclude
items based upon item analysis data must result in a redefin-
ition of the universe of content." The item distractors,
when reviewed, need to be carefully analyzed and rewritten
in conjunction with the operational rules Of facet design.
The generalization as to the validity Of the test cannot come
through item analysis but only through an Operational defin-
ition Of procedures used in generating the items that go into
the test.

The facet test was compared to the Davis Reading
Test which was validated mainly on a statistical basis and
the cloze test which claims Operational validity. The
Pearson product—moment correlation coefficient showed a
significant relationship among these instruments. This
indicates that the tests are measuring similar aspects of
reading comprehension.

The Cloze test was constructed from the same
passages as the facet test and was administered to the same

sample Of students. The correlation coefficient, while

 

74
significant, was lower than expected since both tests are

Operationally defined. The cloze test proved to be more
lengthy than expected, resulting in only one third of the
students completing it. It was thus a speeded test and the
correlation Of a speeded measure with a power test where all
students finished introduced an unknown error factor. This

was a limitation of the studyl.

Further research

 

Further research is necessary to judge the worth

Of facet design. Some possible areas are delineated below.

1. More facets could be employed in the design. Schlesinger
and Weiser (1970) suggest four facets to be used but
there are more which can be applied. The generating Of
items, however, may become more difficult with increased
facets. Use Of a computer could prove helpful.

2. The scoring Of responses could be changed from counting
only the correct distractor to letting the student de-

cide whether the distractors: (l) have no information

 

related to the text, (2) contradict information stated

 

in the text, or (3) whether the distractor agrees with
information stated in the text.

3. One could determine if the facet design can be used to
diagnose reading comprehension difficulties.

4. Individual respondents might be questioned on their in-

correct responses tO determine at what point the student

 

1 See Chapter I

 

75
made his error.

A readability formula could be applied to the distractor
statements to determine if the readability level is
similar to the text information when the author is

allowed to paraphrase information.

 

 

BIBLIOGRAPHY

BIBLIOGRAPHY

Auerbach, I. An Analysis Of Reading Comprehension Tests.
Harvard College, Graduate School Of Education,
Cambridge, Mass., Unpublished Thesis, 1971.

Austin, M. C. and Morrison, C. The First R. New York:
The Macmillan Company, 1963.

 

Bean, K. L. Construction of Educational and Personal Tests.
New York: McGraw-Hill Book Company, Inc.,l953.

Berry, B. T. "Improving Freshman Reading Habits," English
Journal, 20 (1931), pp. 824-28.

Bickley, A. C., Ellington, B. J. and Bickley, R. J. "The
Cloze Procedure: A Conspectus," Journal of
Reading Behavior, II, 3 (Summer, 1970I, pp. 232-49.

 

 

Bormuth, J. R. On the Theory of Achievement Test Items.
Chicago: The University of Chicago Press, 1970.

Buros, O. K. (ed.) Reading Tests and Reviews. Highland
Park, New Jersey: The Gryphon Press, 1968.

 

Buros, O. K. (ed). Seventh Mental Measurements Ygarbook.
Highland Park, New Jersey: The Gryphon Press, 1970.

 

Carroll, R. P. An Experimental Study Of Comprehension in
Readin . New York: Bureau of Publications,
Teachers College, Columbia University, 1927.

 

Culhane, J. W. "Cloze Procedures and Comprehension," The
Reading Teacher, 23, 5 (February, 1970), pp. 410-13.

 

Davis, F. B. Fundamental Factors Of Comprehension in Read-
ing. Unpublished doctor's dissertation. Harvard
University, 1941.

Davis, F. B. and Davis, C. C. Davis Reading Test (Manual),
New York: The Psychological Corporation, 1962.

 

Davis, F. B. Educational Measurements and Their Interpre-
tation. Belmont, California: Wadsworth Publishing
CO., 1964.

 

77

 

 

78

Davis, F. B. Identification and Measurement of Reading
Skills of High-School Students. Co-operative
Research Project NO. 3023, U.S. Department Of
Health, Education, and Welfare, 1967.

Davis, F. B. "Research in Comprehension in Reading,"
Reading Researchyguarterly, III, 4 (Summer, 1968),
pp. 499-545.

Ebel, R. L. Measuring Educational Achievement. Englewood
Cliffs, New Jersey: Prentice-Hall, Inc., 1965.

Ebel, R. (ed). Encyclopedia Of Educational Research.
(4th ed.), American Educational Research Associa-
tion, The Macmillan Company, 1969.

 

English, H. B. and English, A. C. A Comprehension Dictionary
of Psychological and Psychoanalytical Terms.
New York: Longmans, Green and CO., 1958.

 

Fry, E. B. "A Readability_Formu1a That Saves Time, "Journal
Of Reading, 11 (April, 1968), pp. 513-16.

 

 

Fry, E. B. "The Readability Graph Validated at Primary
Levels” The Reading Teacher, 22 (March, 1969),
pp. 534-38.

Gray, W. 8. "Reading." In C. W. Harris (ed.). Encyclopedia

 

of EducationalIResearch. (erd ed.). New York:
Macmillan, 1960, pp. 1086-1135.

 

Guttman, L. "A Faceted Definition Of Intelligence," Scripta
Academica Hierosolymitana, 14 (1965), pp. 166-81.

Guttman, L. and Schlesinger, I. M. Development of Diagnostic
Analytical and Mechanical Abilipy Test Througp
Facet Design and Analysis. (Report of Project
NO. OE - IS - I - 64). U.S. Department Of Health,
Education, and Welfare, 1966.

 

 

 

Guttman, L. and Schlesinger, I. M. "Systematic Construction
of Distractors for Ability and Achievement Test
Items," Educational and Psychological Measurement,
27 (1967), pp. 569-80.

 

Harris, A. J. How to Increase Reading Ability (5th Ed.).
New York: David McKay Company, 1970.

Heilman, A. W. Principles and Practices of Teaching Reading
(2nd Edi). Columbus, Ohio: Charles E. Merrill
Books, Inc., 1967.

 

 

79

Hoffman, B. The Tyranny of Testing. London: Collier -
Macmillan, Ltd., 1962. ,

Holland, J. G. "A Quantitative Measure for Programmed
Instruction," American Educational Research
Journal, 4 (1967), pp. 87-102.

Hunt, J. T. "Selecting a High School Reading Test," High
School Journal, 39 (October, 1955), pp. 49-52.

Jenkinson, M. D. "Basic Elements of Reading Comprehension,"
Proceedings Of the Second World Congress on
Reading, Copenhagen, Denmark, (August), 1968,
pp. 41-47.

Johnson, M. S. and Kress, R. A. "Task Analysis for Criter-
ion—Referenced Tests," Tpe Reading Teacher, 24, 4
(January, 1971), pp. 355-59.

Jordan, J. Attitudes Toward Education and Physically Dis-
abled Persons in Eleven Nations, East Lansing,
Michigan: Michigan State University, 1968.

Langer, J. H. "Vocabulary and Concepts: Essentials in the
Reading - Thinking Process," Elementary School
Journal, 69 (April, 1969), PP. 381-85.

 

Lennon, Roger T. "What Can Be Measured." The Role of Tests
in Reading, Proceedings of Annual Education
Conferences (Russell Stauffer, Ed.). Neward:
University of Delaware, 9 (March, 1960), pp. 67-80.

 

Lindeman, R. H. Educational Measurement. Glenview, 111.:
Scott, Foresman and Company, 1967.

Lindquist, E. F. (ed.). Educational Measurement, American
Council on Education, Washington, D. C. Mensha,
Wisconsin: George Barton Publishing Company, 1951.

 

Maginnis, G. H. "The Readability Graph and Informal Reading
Inventories," The Reading Teacher, 22 (March,
1968), PP. 516-18.

 

Osburn, H. G. "Item Sampling for Achievement Testing,"
Educational and Psychological Measurement, 28,
(1968): pp. 95-104.

 

Pauk, W. "Another Practical Note on Readability Formulas,"
Journal of the Reading Specialist, IX (1970),
pp. 141-43.

 

 

80

Prescott, G. A. "Criterion - Referenced Test Interpretation
in Reading," The Reading Teacher, 24, 4 (January,
1971), PP. 347-54.

 

Rankin, E. F. "The Cloze Procedure - Its Validity and
Utility," The Eighth Yearbook Of the National
Reading Conference, Milwaukee, Wisconsin: The
National Reading Conference, Inc., 1959.

 

 

Rankin, E. F. "The Cloze Procedure - A Survey Of Research,"
Fourteenth Yearbook of the National Readipg
Conference, 1965, pp. 133-50.

 

Schlesinger, I. M. and Guttman, L. "Smallest Space Analysis
Of Intelligence and Achievement Tests," Psychol—
ogical Bulletin, 71 (1969), pp- 95-100.

 

Schlesinger, I. M. and Weiser, Z. "A Facet Design for Tests
of Reading Comprehension," Reading Research
Quarterly, V, 4 (Summer, 1970), pp. 566-80.

 

 

Schubert, D. G. and Torgerson, T. L. (ed's.). Readings
in Reading: Practice Theory Research, New York:
Thomas Y. Crowell Company, 1968.

 

Simons, H. D. "Reading Comprehension: the need for a new
perspective," Reading Research Quarterly, VI,
3(Spring, 1971): pp. 338-63.

 

Taylor, W. L. "Cloze Procedure: A New Tool for Measuring
Readability," Journalism Quarterly, 30, 1953.

 

 

Taylor, W. L. "Recent Developments in the Use Of Cloze
Procedure," Journalism Quarterly, 33, 1956.
Thorndike, E. L. "The Understanding of Sentences,"

Elementary School Journal, 18 (1917), pp. 98-114.

 

Thorndike, R. L. and Hagen, E. Measurement and Evaluation
in Psychology and Education, (3rd ed.). New York:
John Wiley & Sons, Inc., 1969.

 

 

Thorndike, R. L. Educational Measurement (2nd ed.).
American Council on Education, Washington, D.C.,
1971.

 

Traxler, A. E. "Critical Survey of Tests for Identifying
Difficulties in Interpreting What is Read,"
Promoting Growth Toward Maturity in Interpretating
What is Read, Supplementary Educational Monograph,
NO. 74, University of Chicago Press, Chicago, 1951,
pp. 195-200.

 

81

Vernon, P. E. "The Determinants Of Reading Comprehension,"
Educational and Psychological Measurement, 22
(1962): pp. 269-86.

Weaver, W. W. "Theoretical Aspects Of the Cloze Procedure,"
Fourteenth Yeagpook Of the National Reading
Conference, 1965, pp. 115-32.

 

Zintz, M. V. The Reading Process: The Teacher and the
Learner, Dubuque, Iowa: Wm. C. Brown Company, 1970.

 

 

 

APPENDIX A

PASSAGES FOR THE FACET TESTS

AND CLOZE TESTS

 

 

83

The Province of Alberta has six distinct types of municipal
corporations - four urban and two rural. Cities, towns, villages, and
urban counties are urban municipalities; municipal districts and rural
counties are those found in rural areas.

As you will be aware, Alberta has a number of urban centres such
as Edmonton, Calgary, Lethbridge, and Medicine Hat, whose populations
have grown sufficiently to enable them to become cities. They are the
largest urban municipal corporations in the province.

We are going to outline the city in detail as the basis for under-
standing municipal corporations.

Prior tO 1951, cities were given separate charters by the.provin-
cial government. Calgary received its charter in 1893, Edmonton in 1904,
Medicine Hat and Lethbridge in 1906. Each charter included such details
as the name of the city and the powers itscouncil would exercise.

The practice Of issuing individual charters to cities created a
number of problems. If a city's charter was lost temporarily, there was
difficulty in deciding what laws or statutes governed its operation.
Moreover, the separate charter system discouraged closed relations and
co-Operation between Officials of one city and another. Realizing the
weaknesses of the existing charter system, the provincial legislature
passed the City Act which came into effect in January, 1952.

This statute, by ensuring that the charter of cities would be
basically uniform, laid the groundwork for improved local government in
large urban centres. Cities, as well as other municipal corporations,
are, of course, under the control of the Department of Municipal Affairs,

a department of the provincial government.

 

84

How does an urban area become a city? The answer to this question
may be found in the City Act. To acquire that status a town must have a
population of over six thousand people and its council must apply to the
Minister of Municipal Affairs for legal recognition as a city. If the
Minister so recommends, the Lieutenant Governor in Council (the cabinet)
may, by proclamation, grant the town the status of city. This proclama-
tion will, among other particulars, state the name Of the new city, its
area, and the date on which it is Officially to become a city.

The City Act of 1952 did much to establish on a uniform basis the
system of government for large urban centres. Executive and legislative
powers were granted to councils composed of a mayor and aldermen. The
major, elected for a two-year term by all the voters, is assisted by a
deputy mayor who is selected by the council from among its members. If
the mayor is absent from meetings, the deputy mayor assumes his duties.

If neither Of these municipal Officers is present at a council meeting,
the councillors present may appoint an acting mayor to preside over
deliberations.

In addition to the elected mayor, a city council includes aldermen.
By law a council must not have fewer than six nor more than twenty alder-
men, and the number Of aldermen must be an even one. One-half of the
aldermen are elected each year for a term of two years by all the voters.

Alberta's largest cities, Edmonton and Calgary, show a few variations
from the general practice in electing their councils. In Edmonton, the
mayor and twelve aldermen are elected every two years on a city-wide basis.
In Calgary, six wards have been established, with two aldermen elected to
represent each ward. One-half of the aldermen retire each year. The mayor,

however, is chosen by all the voters Of the city for the usual two-year term.

85

Almost from the moment his bones were first discovered in
Germany's Neander Valley a century ago, his name has been synonymous
with brutishness: a squat, shambling creature who wooed his women with
a club and sometimes ate his fellow men when he was hungry. Scientists
have long doubted this harsh popular image of Homo neanderthalensis, or
Neanderthal man. Now, as the evidence accumulates, Neanderthal man is
rapidly being rehabilitated into a more attractive ancestor of modern man.

From remains found in Europe, archaeologists have already concluded
that Neanderthals were skilled hunters and toolmakers, held formal burial
rites that indicated a belief in an afterlife, and even practiced a
primitive form of Social Security for their aged and infirm. More recently,
paleontological examination of skeletons has suggested that Neanderthal
man's stooped appearace may have been the result of disease rather than low
evolutionary status. According to this theory, he was plagued by a dietary
deficiency Of vitamin D. This deficiency was aggravated by the diminished
sunlight Of the ice age, and eventually caused rickets. New, the most
detailed and sympathetic picture yet of Neanderthal man comes from exten-
sive diggings by an American-led expedition in a mountain cave near the
village of Shandar in Iraqi Kurdistan.

In an article in the current Smithsonian magazine, and in a forth-
coming book, Shanidar: The First Flower People (Knopf: $8.95), the
expedition's chief archaeologist, Dr. Ralph S. Solecki, reports that at
least one of the nine Neanderthal skeletons uncovered in the Shanidar cave
was buried with flowers. Another skeleton was that of a man about 40
(equivalent to an age of 80 by modern life-spans) who had been born with

a withered right arm. The limb had apparently been amputated above the

86

elbow by a Neanderthal "surgeon." The man's age and physical condi-
tion indicated to the scientists that he had been unable to fend for
himself. They surmised that his fellows kept him alive until he met
his death in an accidental rockfall inside the cave, a common peril for
these communal hunters who lived from 100,000 to 40,000 years ago.
Comments Anthropologist Carleton S. Coon: "On the grounds of behavior
alone, the Shanidar folk merit the title of Homo sapiens.

Chimp or Philosopher. Neanderthals conducted other elaborate
rites besides funerals. Clues to one Of these were uncovered in
Lebanon last summer when an expedition led by Solecki, who is a professor
Of anthropology at Columbia University, found the dismembered skeleton Of
a small deer in a cave overlooking the Mediterranean. The 50,000 year-
Old bones had apparently been arranged in an orderly way and sprinkled
with red ocher, a substance used for symbolic purposes by Neanderthal man.
Reporting on the discovery last week, Solecki said: These men were trying
to ensure a successful hunt by the ceremonial treatment of one of the
animals." In other words, Neanderthal man resorted to a form of hunter's
magic.

How did so sophisticated a creature acquire such an unwarranted
reputation? For one thing, the first Neanderthal bones were dug just about
the time that Darwin astonished the world with his announcement that man
and ape were descended from a common ancestor. Neanderthal's apish image
was further enforced by the writings early in this century Of the respected
French paleontologist Pierre Marcellin Boule. His portrait of Neanderthal

as a stunted, beetle-browed creature who walked with bent knees and arms

87

dangling in front of him served as the model for several generations of
artists and cartoonists. While certain coarse features in Neanderthal
man are undeniable, on physical considerations alone he deserves far
better treatment. As the late Harvard anthropologist Earnest Hooton

once commented: "You can, with equal facility, model on a Neanderthaloid

skull the features of a chimpanzee or the lineaments of a philosopher.

88

Cold and snow are twins in the boreal forest. For many sorts of
mammals, as well as for grouse, snow is a protection against a plummet-
ing temperature. The short-legged, stout-bodied, short—tailed voles,
for instance, spend almost the entire winter in the air space between
the soil and the snow. SO do the shrews and a few mice.

We know very little as yet about this world between the ground
and the snow, except that its temperature seldom gets very much below
freezing, and that the light is extremely dim. We can guess that its
relative humidity is high and that the sounds Of the world above are as
muffled as the light. The group of voles living in this sort of quarter-
world includes the meadow and red-backed voles which are also found in
the rest of the province, the phenacomys vole and the bog lemming.
Occasionally the voles build runways through the snow but usually,
because of the air space, there is no need to do this. However, they
Often drive ventilation tunnels to the surface of the snow cover. They
also bore through the snow itself from time to time, apparently in search
Of food particles fallen from the trees or trapped in the snow—cover.
Their habitat has holes, like those in Swiss cheese, wherever there is
qaminiq with snow too thin to give insulation.

While the voles use snow for survival, beavers and muskrats avoid
the extreme cold of winter by living under the ice of ponds and rivers.
This is the warmest place for them, since in water the temperature cannot
fall below freezing. Both species move about as they do in summer, seldom
coming onto the surface Of the land unless their ponds are frozen to the
bottom. Sealed Off by snow and ice and protected by the frozen walls of

their houses, they are relatively safe. But every Eden, even a frozen

 

89

one, has a handicap. Winter does not prevent mink and otters from
invading their habitat.

Snow can hinder as well as help. In Alberta everyone knows from
experience how difficult it is to struggle through knee-deep snow. The
animals of the boreal forest have solved this difficulty in various ways.

One group of mammals simply wades through the snow as we do. The
best adapted animal in this group is the moose. It moves easily through
two feet of snow because of its long legs and a special arrangement of
muscles and bones that permits the legs to work almost straight up and
down. The toes of the moose seldom leave drag marks in the snow so that
it is usually difficult to tell from its tracks in which direction it
was moving. The wapiti and smaller deer run into trouble at somewhat
lesser depths. The predator of the moose, the wolf, also wades through
the snow and, as long as it is soft and fluffy, the wolf is at a consid-
erable disadvantage. But if a crust forms, the wolf may be able to stay
on the surface while the moose keeps on wading, but with difficulty.
Hence, the depth and structure of the snow mean life or death to both
the wolf and the moose. Animals living in herds reduce the energy needed
for ploughing through snow by playing follow-the-leader. This is the
technique used by the bison of Wood Buffalo National Park, by caribou,
and in deep snow by deer, when they tramp down an area to form what is

known as a yard.

90

The relationships that exist among the plants and animals that
have been described are almost unbelievably complex. Each species has
its food requirements, its space requirements and its requirements of
temperature, light and water. It is common to refer to "competition"
among organisms for these requirements. Organisms which are at any
disadvantage in the competition, for any reason, decline in numbers
and may become locally or universally extinct. Of course conditions
may change (for instance, moist years may follow drought) and the trend
to decline may be reversed. Because natural conditions are constantly
changing, the numbers of living things fluctuate. As examples, for a
few years there may be many hares, then, for some years, hares may be
scarce. Or, around a slough willows may predominate for a period of
time, to be drowned when the water level rises, and replaced by cattails.
In some years sparrow hawks are abundant, in some grasshoppers thrive.
In some years Canada thistle is a greater nuisance than in others. We
are not always sure what causes the fluctuations.

The relationships between organisms are often called together
"The Balance Of Nature". This phrase means that living things occur as
best they can in the face Of the competition of all other living things
and the changes of physical conditions (weather, soil conditions, etc.).
The Balance of Nature is not something that is the same year in and year
out. The balance changes all the time because new and different influ-
ences are always affecting one or another member of the natural community.
This, in time, affects neighbouring members of the community so that
living conditions might be bettered or worsened for different members of

the community. The balance will have shifted.

91

Let us examine an imaginary case, to see how a balance might
shift over a period of time. We shall look at a grassy area at the
margin of the aspen parkland zone. The grasses there might be chiefly
a rough fescue, porcupine grass and June grass. The grasses generally
prevent the establishment of aspen: under the proper conditions, grasses
are more successful competitors for soil than aspen. But the situation
can be reversed if bare areas of soil develop. Such bare areas can be
formed by burrowing Richardson's ground squirrels, badgers, pocket gophers
or other animals, or by overgrazing and trampling by cattle, or in days
past by bison. Bare soil areas give shrubs, such as snowberry, a chance
to become established. Perhaps sharp-tailed grouse, using the bare area
as a dusting ground, left seeds there in droppings. Other seed-eating
birds or wind dispersal can also seaibare areas. If snowberry and wolf
willow become established, their shade inhibits grass growth. Coyotes
tend to den in shrubby sites further disturbing the soil and preventing
the growth of grass. Some grasshoppers, preferring grass, become scarce,
as do some ants. Different species of ants and grasshoppers are found
which prefer shrubby habitat. Aspen seedlings appear. As aspen develops,
it begins to shade and restrict snowberry and wolf willow. Shrubs come to
form a fringe around the edge of an enlarging grove of aspen. Songbirds
are attracted to nesting sites. Flickers and crows inhabit the woods.
Snowshoe hare and Franklin's ground squirrels prefer the woods. So does
the great horned owl.

As the aspen ages, spruce seedlings may begin to appear. Spruce

may be held back if a lot Of hares are present to feed on the tender young

92

plants. Fire will stop the development of spruce, while after a fire
aspen can regenerate by means of suckers. But these aspen suckers can
be held down by the foraging activities of rabbits or other browsers.

In such cases, grass will reappear, once again to be the main vegetation,

with its own distinctive population of animals.

 

 

APPENDIX B

FACET TEST DIRECTIONS

AND QUESTIONS

Research Project

NAME

 

TYPE OF SCHOOL PROGRAM (e.g., Matriculation, Diploma, etc.)

 

DIRECTIONS

This package of materials contains four passages for you to read
and a set of 10 multiple—choice questions about each passage.

1. Read the first passage through thoroughly. Turn to the questions
that follow and answer each one by circling the correct answer.

DO not look back at the passage you have just read when answering
the questions. Answer all of the questions even though you may
have to guess.

2. Choose the best possible answer to the questions based only upon
information contained in the related passage.

3. When you have read the first passage and answered the questions,
go on to the next passage and questions. Continue through the
passages and questions in the order assembled until you are finished.

 

4. Do not go back to check your answers.

94

 

Questions

95

1. In order for an urban area to acquire city status

a.

a resolution containing this idea is passed by the cabinet
and initialled by the Prime Minister.

the Lieutenant Governor in Council may grant it upon the
recommendation of the Minister of Municipal Affairs.

it must have thousands of people, much housing, stores,
schools, offices, theaters and parks.

the mayor and twelve aldermen are elected every two years
on a city-wide basis.

2. Choose the central theme of the passage.

The city in Alberta.

The laws that group communities into classes in relation
to their population.

Alberta's largest cities, Edmonton and Calgary.

Certain rural areas of the province are more thickly
populated than others.

3. The city act

a.

allowed for separate charters to be granted by the
provincial legislature.

was passed to solve the differences between municipal
bodies and school boards.

came into effect in January 1952.

outlines the regulations concerning elections in cities.

 

96

4. The most recent city act

a.

b.

granted separate charters to every city.

gave the mayor the right to be an ex officio member of
every board, commission, or organization where the council
has jurisdiction to appoint members.

 

did much to establish on a uniform basis the system of
government for large urban centers.

allowed that at no time can any individual or group in
government make a "blanket" claim-to-fame for having
achieved the perfect solution.

5. The proclamation granting city status would

a.

b.

6. According

a.

provide six types of municipal corporations.

recognize, by law, the rights and privileges of citizens.
provide police and fire protection, street and traffic
control, health and welfare services and other services to
maintain community life .

state the name of the new city, its area, and the date on
which it is officially to become a city.

to law, the city council

issues separate charters that provide details of the
city's name and powers of its council.

must not have fewer than six nor more than twenty aldermen.
must be Canadian citizens or British subjects.

are pleased when people in the community take a genuine
interest in the business of the municipality.

7. In Edmonton

a.

the mayor and twelve aldermen are elected every two years
on a city-wide basis.

two aldermen are elected from each ward.

one of the powers of government is to create and establish
the municipal level of government.

aldermen receive a regular yearly allowance.

 

97

8. The deputy mayor

a.

has a threefold function to perform which is legislative,
executive and judicial.

is chosen by all the voters of the city for the usual
two-year term.

may receive pay for his work in this capacity.

assists the mayor and is selected by the council from
elected aldermen.

9. The city act laid the groundwork for better local government in
urban centers because

a.

cities were given separate charters by the provincial
government.

it made sure that the city charters were basically uniform.

it established the eligibility of candidates and persons
to vote as well as other matters related to municipal elections.

it outlines clearly the mandatory and optional powers of
the council.

10. One could describe the municipal corporations in Alberta as

a.

a body or unit whose inhabitants ( permanent residents and
ratepayers ) have been united by law.

being similar to the strong~mayor form of government
with the mayor having broad powers.

a number of urban centers such as Edmonton, Calgary, Lethbridge
and Medicine Hat whose populations have grown sufficiently
to enable them to become cities.

being either urban of rural.

 

 

98

Questions

1. From the first discovery of Neanderthal's bones

a. his name has been associated with crudely primitive in
manner or conduct.

b. his name has been synonymous with brutishness.

c. there has been a connection with Dr. Ralph S. Solecki, who
led an expedition to Lebanon last summer.

d. Darwin's theory has withstood almost a century of criticism.

2. From remains found in Europe, archeologists have concluded that

a. Neanderthals were skilled hunters and toolmakers.

 

b. Neanderthals made excellent flint tools.

c. Earnest Hooton was correct in saying that you could, with
equal ease, put the features of a chimpanzee on the

Neanderthal skulls.

d. vitamin D is a group of about 10 fat-soluble vitamins that
prevent rickets.

3. Recent studies suggest that Neanderthal Man's stooped appearance may
have been
a. the result of sprinkling red ocher on animal bones.
b. the result of disease rather than low evolutionary status.

c. due to the fact that his brain was somewhat larger than
modern man.

d. essential to determine the origin of inherited variations
and the effects of the environment upon them.

99

4. Dr. Solecki reports that

a. bones from a human skeleton were discovered over a
century ago in Germany's Neander Valley.

b. Neanderthal people probably lived between 30,000 and 60,000
years ago.

c. rickets is a vitamin deficiency that prevents the bones
.from hardening properly thus resulting in deformations.

 

d. at least one of the nine Neanderthal skeletons uncovered
in the Shanidar cave was buried with flowers.

5. On a recent expedition to a mountain cave in Iraqi Kurdistan

a. scientists discovered a Skullcap.

 

b. Darwin astonished the world with his announcement that
man and ape were descended from a common ancestor.

c. there were several scientists led by an American.
d. it was found that making the stone axe required a number
of simple skills.
6. The main idea expressed in this passage is that

a. the ice age has been present during the majority of man's
existence.

b. Neanderthal, is the name of a race of prehistoric men
who lived in caves in EurOpe, North Africa and western and
central Asia.

c. Neanderthal's apish image was further enforced by the
writings early in this century of the respected French
palenntologist Pierre Marcellin Boule.

d. Neanderthal Man may not have been as primitive as originally
speculated.

100

7. The Neanderthal villages probably had someone comparable to
a. a medical doctor.

b. men about five to five and one-half feet tall who did not
walk erect because of curved thigh bones.

c. the respected French paleontologist Pierre Marcellin Boule.

d. shamans who speak with the voices of the dead.

8. Dr. Solecki, a professor of anthropology,
a. would be concerned about the origin of Shanidar folk.

b. astonished the world with his announcement that man and
ape were descended from a common ancestor.

c. would be concerned about the culture of Shanidar folk.

 

d. views people as highly innovative creatures, organizing
and reorganizing their field of experience at every moment.

9. Neanderthal Man got his original reputation concerning his
life-style because

a. scientists discovered a Skullcap in the Neander Gorge near
Dusseldorf, Germany.

b. of extensive diggings by an AmeriCan-led expedition in a
mountain cave near the village of Shanidar in Iraqi Kurdistan.

c. the Neander Valley discovery was made near the time Darwin's
theory was announced.

d. they produce, through cultivation, most of their own food,
but produce also for a market.

10. Scientists have concluded that a Shanidar man of forty would be
a. a stooped, squat and shambling creature.

b. equivalent to an age of 80 by modern life-spans.

c. suffering from a vitamin deficiency that prevents the bones
from hardening properly, thus resulting in deformations.

d. as erect as any modern man, he did not have a bull neck, and
he was not knock-kneed.

101

Questions

1. The world between the ground and the snow
3. is protection for the grouse against cold temperatures.

b. is the home for many organisms who survive the harsh
winters in Alberta.

c. has lichens and other low growing plants.

d. seldom gets very much below freezing.

2. Beavers avoid the extreme cold of winter by
a. living under the ice of ponds and rivers.

b. building runways through the snow.

 

c. storing an ample supply of willows, aspen poplar and birch.
d. becoming very fat and yielding to hibernation.
3. The moose is one of the best adapted animals for wading through snow
because
a. it has rounded, wide-spread hoofs on the hind feet.

b. of special arrangement of muscles and bones that permits
the legs to work almost straight up and down.

c. it seldom leaves drag marks.
d. it makes daily treks through the wintering area to keep the
snow in pathways at a minimum level.
4. A yard is

a. the technique used by the bison of Wood Buffalo Park, by
caribou and in deep snow by deer.

b. an area set aside for a particular business or activity.
c. an area where snow has been tramped down by deer or caribou.

d. a small enclosed area open to the sky.

5.

102

You could describe the appearance of the vole which lives in the space
between the snow and the ground as

a. dainty and for the most part richly colored.

b. a plump animal with stubby legs.

c. twins in the boreal forest.

d. a relatively plump rodent with small ears.
"Their habitat has holes, like those in Swiss Cheese, wherever there
is qaminiq with snow too thin to give insulation!‘ The word qaminiq

a. is an Eskimo word referring to snow that reaches the ground.

b. means runways through the snow.

c. most likely means snow covering the ground.

d. is a hollow conduit or recess.

The wapiti is a member of the deer family
a. smaller than the moose.
b. known as an elk.

c. seldom coming onto the surface of the land unless their
ponds are frozen to the bottom.

d. with big, heavily-furred feet, which with long legs, create
a rather bizarre appearance.
Deep, crusted snow hinders survival of the moose because
a. even a frozen Eden has a handicap.
b. he is unable to forage sufficiently well.

c. snow crystals often cling together to form snow pellets
over an inch thick.

d. the wolf may be able to stay on the surface while the
moose keeps on wading.

 

103

9. Choose the best theme for the passage.

a. The greater the variety of food the better is the chance of
survival.

b. Cold and snow are twins in the boreal forest.

c. In some years, it snows only a little, but in others it
snows a great deal.

d. How snow helps and hinders survival.

10. Voles living in the space between ground and snow

a. can reduce the energy needed for survival by playing
follow-the-leader.

b. often drive ventilation. tunnels to the surface of the
snow cover.

c. have a dark, glossy-brown pelage on top and slatey grey below.

d. of course can exist only in small numbers, since for
support they require so many of the lesser creatures.

Questions

104

1. Which of the following would be most correct about plants and
animals according to the author?

a. Shrubs can grow in bare areas.

b. The food and space requirements of each species varies
with climatic conditions.

c. Each species has its own food requirements and its space
requirements.

d. Predatory animals subsist on prey which in turn must capture
other animals or eat plants.

2. "The Balance of Nature"

a. refers to the distribution of animals in Alberta.

b. is not something that is the same year in and year out.

c. means that for a few years hares may be scarce but then
increase in number.

d. includes the change of simple chemical substances to living

things and living things back into simple chemicals.

3. Under certain conditions

the relationships among all plants and animals is complex.
pocket gophers do not live in some parts of the extreme
southeastern part of the province even though they inhabit

surrounding areas.

willows planted on a riverbank are good protection against
soil erosion.

grasses are more successful competitors for soil than aspen.

 

4.

105

Which of the following could be considered the best theme of the passage?
a. The balance of nature is related to nutrition.
b. Grasses generally prevent the establishment of aspen.

c. The relationships among plants and animals are subject to
many varied conditions.

d. All sorts of limiting factors can be found and often their
results are unexpected.
The author of this passage agrees that plants which do not normally

grow in one area

a. are helpful in building up the soil with minerals and
root fibres.

b. may flourish if competitive plants are removed.

c. are always affecting one or another member of the natural
community.

d. were transported and placed there by a number of different
agencies.
According to this author, every living situation
a. is undergoing change either slowly or rapidly.
b. could be affected by climate under certain conditions.
c. begins to shade and restrict snowberry and wolf willow.
d. is the essential first step that prepares the way for all
the life that exists on earth.
Cattail is a plant that may grow
a. at the margin of the aspen parkland zone.
b. in the sandy areas where jack pine is common.
c. to a height of 5 or 6 feet, and have long, broad leaves.

d. where willows once predominated but were drowned out by
high water.

 

106

8. According to this passage, coyotes

a.

9. Browsers,

form bare areas which allow plants, other than grasses,
to begin growing.

serve to illustrate the potential danger in introducing
new species which might compete with and overcome native species.

tend to den in shrubby areas.

live in regions of western North America from Panama to Alaska.

as referred to by the author, are
animals, such as the rabbit, that eat tender shoots and twigs.
predators that avoid man.

shrubs that come to form a fringe around the edge of an
enlarging grove of aspen.

subject to the same basic principles of adaptation and
balanced control that apply to all living communities.

10. Bare areas are necessary in the balance of nature because

a.

small animals, especially insects, mites and other arthropods,
dwell in the soil in very large numbers.

they provide a home for many living things such as bacteria,
earthworms, etc., which add to the organic content.

conditions are constantly changing which results in
fluctuations of living things.

these give shrubs, such as the snowberry, a chance to
become established.

 

 

APPENDIX C

CLOZE TEST DIRECTIONS

AND QUESTIONS

108

NAME

 

DIRECTIONS

 

This package of materials contains four passages for you to read
and a fill-in-the-blanks test for each passage.

1. Read the first passage carefully, then proceed to the pages
immediately following marked "Questions." The passage you
have just read is rewritten with blanks at certain intervals.
You are to fill in the blank with the exact word that was in
the original passage.

DO NOT LOOK BACK AT THE ORIGINAL PASSAGE WHEN FILLING
IN THE BLANKS.

2. Try to complete every blank even if you must guess at the
answer.

 

3. When you have read the first passage and completed filling
in the blanks with the appropriate word for the accompanying
test, you should continue on to the next passage and test and
complete it in the same way.

4. Continue through the passages and questions are they are as—
sembled until you are finished.

5. Work as quickly as you can. Do not spend too much time on any
one blank.

F—

109

 

 

 

 

 

 

Questions

Almost from the moment his bones were first discovered
Germany's Neander Valley a century ago, his name has synonymous
with brutishness: a squat, shambling creature who wooed women
with a club and sometimes ate his fellow when he was hungry.
Scientists have long doubted this popular image of Homo
neanderthalensis, or Neanderthal man. Now, the evidence
accumulates, Neanderthal man is rapidly being rehabilitated a

 

more attractive ancestor of modern man.

 

 

 

 

 

 

 

 

From remains in Europe, archaeologists have already con-
cluded that Neanderthals were hunters and toolmakers, held formal
burial rites that indicated belief in an afterlife, and even
practiced a primitive of Social Security for their aged and
infirm. More , paleontological examination of skeletons has
suggested that Neanderthal man's appearance may have been the
result of disease rather low evolutionary status. According to
this theory, he was by a dietary deficiency of vitamin D. This
deficiency aggravated by the diminished sunlight of the ice age,

 

eventually caused rickets. Now, the most detailed and sympathetic

 

yet of Neanderthal man comes from extensive diggings by

 

 

American-led expedition in a mountain cave near the village

 

Shandar in Iraqi Kurdistan.

 

 

 

 

 

In an article in the Smithsonian magazine, and in a forth-
coming book, Shandiar: The Flower People (KnOpf: $8.95), the
expedition's chief archaeologist, Dr. Ralph S. , reports that at
least one of the nine Neanderthal ' uncovered in the Shandiar cave
was buried with flowers. skeleton was that of a man about 40
(equivalent an age of 80 by modern life—spans) who had

 

 

born with a withered right arm. The limb had been amputated above

 

 

 

 

 

110

 

 

 

the elbow by a Neanderthal "surgeon." man's age and physical
condition indicated to the scientists he had been unable to
fend for himself. They that his fellows kept him alive until
he met death in an accidental rockfall inside the cave, a

 

peril for these communal hunters who lived from 100,000

 

 

40,000 years ago. Comments Anthropologist Carleton S. Coon: "On

grounds of behavior alone, the Shanidar folk merit the

 

 

of Homo sapiens.

Chimp or Philosopher. Neanderthals conducted other rites

 

besides funerals. Clues to one of these were in Lebanon last

 

 

summer when an expedition led by , who is a professor of

 

anthropology at Columbia University, the dismembered skeleton of

 

a small deer in a overlook the Mediterranean. The 50,000 year-

 

old bones had apparently arranged in an orderly way and sprinkled

 

with red , a substance used for symbolic purposes by Neanderthal

 

man. on the discovery last week, Solecki said: These men

 

trying to ensure a successful hunt by the ceremonial

 

 

of one of the animals." In other words, Neanderthal resorted to

 

a form of hunter's magic.

 

 

How did sophisticated a creature acquire such an unwarranted
reputation? For thing ,the first Neanderthal bones were dug just
about time that Darwin astonished the world with his announcement

 

man and ape were descended from a common ancesror.

 

 

apish image was further enforced by the writings early this century

of the respected French paleontologist Pierre Marcellin . His

lll

portrait of Neanderthal as a stunted, beetle-brewed creature

 

 

 

 

walked with bent knees and arms dangling in front him served

as the model for several generations of and cartoonists. While
.certain coarse features in Neanderthal man undeniable, on
physical considerations alone he deserves far better . As the

 

late Harvard anthropologist Earnest Hooton once commented: "
I

can, with equal facility, model on a Neanderthaloid skull

 

 

features of a chimpanzee or the lineaments of a ."

 

 

 

 

 

 

 

 

112
Questions

Cold and snow are twins in the boreal forest. many sorts

 

of mammals, as well as for grouse, is a protection against a

 

plummeting temperature. The short-legged, , short-tailed voles,

 

for instance, spend almost the entire winter the air space

 

between the soil and the snow. do the shrews and a few mice.

 

We know little as yet about this world between the ground

 

the snow, except that its temperature seldom gets very

 

 

below freezing, and that the light is extremely dim. can guess

 

that its relative humidity is high and the sounds of the world

 

above are as muffled the light. The group of voles living in

 

this of quater-world includes the meadow and red-backed voles

 

which also found in the rest of the province, the vole

 

 

and the bog lemming. Occasionally the voles build through the

 

snow but usually, because of the air , there is no need to do

 

this. However, they drive ventilation tunnels to the surface of

 

the snow . They also bore through the snow itself from time

 

time, apparently in search of food particles fallen from

 

 

trees or trapped in the snow-cover. Their habitat has , like those

 

in Swiss cheese, wherever there is qaminiq snow too thin to give

 

insulation.

While the voles snow for survival, beavers and muskrats

 

avoid the extreme of winter by living under the ice of ponds

 

rivers. This is the warmest place for the, since water

 

the temperature cannot fall below freezing. Both species about as

they do in summer, seldom coming onto surface of the land unless

 

their ponds are frozen the bottom. Sealed off by snow and ice

 

 

 

 

 

 

113

 

 

 

 

 

 

and by the frozen walls of their hoses, they are
safe. But every Eden, even a frozen one, has handicap.
Winter does not prevent mink and otters from their habitat.
Snow can hinder as well as help. Alberta everyone
knows from experience how difficult it is struggle through knee-
deep snow. The animals of the boreal have solved this difficulty

 

in various ways.

One group mammals simply wades through the snow as we do.

 

best adapted animal in this group is the moose. moves

 

 

easily through two feet of snow because of long legs and a special

 

arrangement of muscles and that permits the legs to work almost

 

straight up down. The toes of the moose seldom leave drag

 

in the snow so that it is usually difficult tell from

 

 

its tracks in which direction it was . The wapiti and smaller deer

 

run into trouble at lesser depths. The predator of the moose, the

 

wolf, wades through the snow and, as long as it soft

 

 

and fluffy, the wolf is at a considerable . But if a crust forms,

 

the wolf may be to stay on the surface while the moose keeps

 

wading, but with difficulty. Hence, the depth and strutture

 

the snow mean life or death to both the and the moose.

 

 

Animals living in herds reduce the needed for ploughing through

 

snow by playing follow—the-leader. This the technique used by

 

the bison of Wood Buffalo Park, by caribou, and in deep snow by

 

deer, they tramp down an area to form what is as a

 

 

yard.

Questions 114

The Province of Alberta has six distinct types of

 

corporations — four urban and two rural. Cities, towns, villages,

 

urban counties are urban municipalities; municipal districts and rural

are those found in rural areas.

 

 

 

 

 

As you will aware, Alberta has a number of urban centres
such Edmonton, Calgary, Lethbridge, and Medicine Hat, whose
populations have grown to enable them to become cities. They
are the urban.municipal corporations in the province.

We are going outline the city in detail as the basis for

 

municipal corporations.

 

Prior to 1951, cities were given separate by the provin-

 

cial government. Calgary received its charter in , Edmonton in

 

1904, Medicine Hat and Lethbridge in 1906. Each included such

 

details as the name of the city the powers its council would

 

exercise.

The practice of individual charters to cities created a

 

number of problems. a city's charter was lost temporarily, there

 

was difficulty deciding what laws or statutes governed its operation.

 

Moreover, separate charter system discouraged closed relations and

 

co-operation between of one city and another. Realizing the

 

weaknesses of existing charter system, the provincial legislature

 

passed the City which came into effect in January, 1952.

 

This statute, ensuring that the charter of cities would be

 

basically laid the groundwork for improved local government in

 

large centres. Cities, as well as other municipal corporations,

 

are, course, under the control of the Department of Municipal

 

, a department of the provincial government.

 

 

 

 

 

 

 

115

 

 

 

 

 

How does an area become a city? The answer to this question

be found in the City Act. To acquire that a town must

have a population of over six people and its council must apply to
the Minister Municipal Affairs for legal recognition as a city. If

 

Minister so recommends, the Lieutenant Governor in Council (the

 

) may, by proclamation, grant the town the status of . This

 

 

proclamation will, among other particulars,state the name the new city,

 

its area, and the date on it is officially to become a city.

 

The City of 1952 did much to establish on a uniform

 

 

the system of government for large urban centres. Executive legisla-

 

 

tive powers were granted to councils composed of a and aldermen. The

 

major, elected for a two-year term all the voters, is assisted by a

 

deputy mayor is selected by the council from among its members.

 

the mayor is absent from meetings, the deputy mayor his

 

 

duties. If neither of these municipal officers is at a council

 

meeting, the councillors present may appoint acting mayor to preside

 

over deliberations.

In addition to elected mayor, a city council includes aldermen.

 

By Law council must not have fewer than six nor more twenty

 

 

aldermen, and the number of aldermen must be even one. One-half of

 

the aldermen are elected each for a term of two years by all the

 

 

Alberta's largest cities, Edmonton and Calgary, show a few

 

from the general practice in electing their councils. In . the mayor

and twelve aldermen are elected every two on a city—wide basis. In

 

Calgary, six wards have established, with two aldermen elected to

 

represent each ward. of the aldermen retire each year. The mayor,

 

however, chosen by all the voters of the city for usual

 

two-year term.

116
Questions

The relationships that exist among the plants and animals

 

have been described are almost unbelievably complex. Each species

 

its food requirements, its space requirements and its requirements

 

 

 

 

 

 

 

 

 

temperature, light and water. It is common to refer "competition"
among organisms for these requirements. Organisms which are any
disadvantage in the competition, for any reason, decline numbers
and may become locally or universally extinct. Of conditions

may change (for instance, moist years may follow ) and the trend
to decline may be reversed. Because conditions are constantly
changing, the numbers of living things . As examples, for a

few years there may be hares, then, for some years, hares may be
scarce. around a slough willows may predominate for a period

 

time, to be drowned when the water level rises,

 

 

replaced by cattails. In some years sparrow hawks are , in some

 

grasshoppers thrive. In some years Canada thistle a greater

 

nuisance than in others. We are not sure what causes the

 

fluctuations.

The relationships between organisms often called together

 

”The Balance of Nature." This phrase that living things occur as

 

best they can in face of the competition of all other living

 

things the changes of physical conditions (weather, soil

 

conditions, etc._: Balance of Nature is not something that is the

 

year in and year out. The balance changes all time

 

 

because new and different influences are always affecting or

 

another member of the natural community. This, in' , affects

 

neighbouring members of the community so that living might be

 

 

 

117

bettered or worsened for different members of community. The

 

balance will have shifted.

Let us.examine imaginary case, to see how a balance

 

might shift a period of time. We shall look at a

 

 

area at the margin of the aspen parkland zone. grasses there

 

might be chiefly a rough fescue, porcupine and June grass. The

 

grasses generally prevent the establishment aspen: under the

 

proper conditions, grasses are more successful for soil than

 

aspen. But the situation can be if bare areas of soil develop.

 

Such bare areas be formed by burrowing Richardson's ground

 

squirrels, badgers, pocket or other animals, or by overgrazing

 

and trampling by , or in days past by bison. Bare soil areas

 

shrubs, such as snowberry, a chance to become established.

 

sharp-tailed grouse, using the bare area as a dusting ,

 

 

left seeds tnere in droppings. Other seed-eating birds or

 

 

 

 

 

 

 

 

 

disperal can also seed bare areas. If snowberry and willow become

established, their shade inhibits grass growth. Coyotes to den in

shrubby sites further disturbing the soil preventing the growth

of grass. Some grasshoppers, preferring grass, scarce, as do

some ants. Different species of ants grasshoppers are found

which prefer shrubby habitat. Aspen seedlings . As aspen develops,

it begins to shade and restrict and wolf willow. Shrubs come to

form a fringe the edge of an enlarging grove of aspen. Songbirds
attracted to neating sites. Flickers and crows inhabit the

 

Snowshoe hare and Franklin's ground squirrels prefer the woods.

 

does the great horned owl.

 

 

 

118

As the aspen ages, seedlings may begin to appear.

 

Spruce may be held if a lot of hares are present to feed

 

the tender young plants. Fire will stop the development

 

spruce, while after a fire aspen can regenerate by

 

 

 

 

of suckers. But these aspen suckers can be held by the foraging
activities of rabbits or other browsers. such cases, grass will
reappear, once again to be main vegetation, with its own

 

distinctive population of animals.

 

 

MTITI'ITIQINHLRMITIIﬁﬂlﬂlﬂlﬁﬂﬂﬂjﬂjﬂiﬂl?“