ﬁfs'“ ~ -' - V '
1" ‘
I ,3 , r'. V 1
.\'._ f "I T ‘ I) .k
- 7% ‘7’
wk’Wd‘w'Ji‘W‘W’mWw f
Q
“mem 1' 3g
'.)’.,./:'/"- kg». i 6"":
W____/
ABSTRACT
A STUDY OF THE RELATIONSHIPS AMONG ACADEMIC MOTIVATION,
LEVEL OF ASPIRATION. LEVEL OF EXPECTATION, AND GAIN
OR LOSS IN ACHIEVEMENT IN A COLLEGE READING
IMPROVEMENT SITUATION
by Jack 0. Anderson
The study was concerned with relationships between
(1) degree of academic motivation as measured by the M—Scales,
(2) setting of levels of aspiration and expectation. and
(3) gain in reading rate and comprehension. The population
was divided by sex and then by the independent variable.
academic motivation. Male N was 99 and female N was 34.
Three major null hypotheses were advanced.
1. There are no significant relationships between
academic need—achievement scores and gain or loss
in achievement.
2. There are no significant relationships between
academic need-achievement scores and the discrepancy
between final levels of aspiration and final levels
of expectation.
1Developed by William W. Farquhar and associates.
A Comprehensive Study of Motivational Factors Underlying
Achievement of Eleventh Grade High School Students (ResearCh
Project No. 846 (8458): Supported by the U. S. Office of
Education, in cooperation with Michigan State University, 1959).
Jack 0. Anderson
3.] There are no significant relationships.among
academic need-achievement scores, final levels of
aspiration, final levels of expectation. and gain or
loss in achievement. I
Two analysis procedures were employed: 5(1) the
significance of the difference between High Need-Achievement
means and Low Need-Achievement means. using Student's t—
distribution: and (2) product-moment zero order, first order.
and second order partial correlation coefficients.
Students completed the MrScale during the first class
and wrote the Iowa Test of Silent Reading, Ag during the
second class. During the fourth class they set initial levels
of aspiration and expectation for their performance on the
final Iowa Teet, as (1) words per minute. and (2) compre—
hension score (number correct of‘a possible 35). Final levels
of aspiration and expectation wereset during the tenth class:
these were used in the study. in the effort to achieve realism.
Iowa Test form Cm was the post test and the differences between
Am and Cm scores furnished the gain in achievement variables.
MeScale scores served as significant negative predictors
for male gain in comprehension. Levels of aspiration were
significant predictors of female gain in rate, and were highly
significant negative predictors for male and.female gain in
comprehension. Levels of expectation served as highly
Jack 0. Anderson
significant positive predictors of male gain in rate and
female gain in comprehension, and as highly significant
negative predictors of male gain in comprehension.
For males. all three major null hypotheses were
rejected: for females, only major null hypothesis three was
rejected.
A STUDY OF THEnRELATIONSHIPS AMONG ACADEMIC MOTIVATION:
LEVEL OF ASPIRATION, LEVEL OF EXPECTATION, AND GAIN
OR LOSS IN ACHIEVEMENT IN A COLLEGE READING
IMPROVEMENT SITUATION
BY
Jack 0: Anderson
A THESIS
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
DOCTOR OF EDUCATION
College of Education
1962
.ACKNOWLEDGMENTS
'Grateful acknowledgment is made to the following
people:
To Dr. William W.Farquhar, director of this research,
for his advice, assistance, and confidence expressed throughout
the course of this study.
To Dr. Charles Blackman, my major professor, for his
advice, assistance, and patience.
To Dr. Elizabeth Rusk for her advice and encouragement.
To Dr. Ernest Melby for his exemplary leadership.
To Dr. John Useem for his suggestions and confidence.
To Dr. David Payne for his assistance.
To Dr. Leroy Olson for his advice and assistance.
ii
TABLE OF CONTENTS
Chapter Page
I. FORMULATION AND DEFINITION OF THE PROBLEM . . 1
Need for the Study 1
Background of the Study 2
Purpose of the Study 4
The Hypotheses 5
II. REVIEW OF THE LITERATURE . . . . . . . . . 6
Pertinent Studies 6
Application of the Literature to This
Study 14
Summary 16
III. DESIGN AND METHODOLOGY . . . . . . . . . . . 17
Operational Definitions 17
The Population 19
The Sample 19
Chronology of Experimental Procedures 21
The Null Hypotheses and Analysis
Procedures 23
Statistical Treatment 24
Summary 25
IV. ANALYSIS OF THE DATA . . . . . . . . . . . . 26
Major Null Hypotheses 27
Subsidiary Null Hypothesis ‘28
Analysis of Differences of Means, Male 28
Rate Gain 28
Comprehension Gain 28
Grade Point Average 29
Analysis of Differences of Means, Female 30
Rate Gain 30
Comprehension Gain 31
Grade Point Average 31
iii
Chapter Page
Correlation Coefficients, Male, Rate
Gain 32
Correlation Coefficients, Male.
Comprehension Gain 35
Correlation Coefficients for Subsidiary
Null Hypothesis 37
Correlation Coefficients, Female, Rate
Gain 38
Correlation Coefficients, Female, -
Comprehension Gain 40
Summary 42
V. DISCUSSION OF THE RESULTS . . . . . . . . . . . 44
M-Scale and Its Use in a College
Reading Improvement Situation 44
Male, Rate Gain 44
Male, Comprehension Gain 46
Female, Rate Gain 46
Female, Comprehension Gain 46
Level of Aspiration and Its Use in
a College Reading Improvement
Situation 47
Male, Rate Gain 47
Male, Comprehension Gain 47
Female, Rate Gain 47
Female, Comprehension Gain 48
Level of Expectation and Its Use in a
College Reading Improvement Situation 48
Male, Rate Gain 48
Male, Comprehension Gain 49
Female, Rate Gain 49
Female, Comprehension Gain 49
Summary 49
iv
Chapter Page
VI. SUMMARY, CONCLUSION, AND RECOMMENDATIONS . . 51
Summary 51
The Problem 51
The Design and Procedure of the Study 51
The Analysis 52
Conclusions 55
Recommendations 56
BIBLIOGRAPHY . . . . . . . . . . . . . . . . . . . . 58
Table
4.10
4.11
LIST OF TABLES
Bipolar theory of high and low academic
need achievement motivation . . . . . . .
Population, sample, and sample losses . . .
Tests of significance of difference between
means: male, rate gain . . . . . . . . .
Tests of significance of difference between
means: male, comprehension gain . . .
Tests of significance of difference between
means: male GPA and M-Scale . . . . . .
Tests of significance of difference between
means: female, rate gain . . .
Tests of significance of difference between
means: female, comprehension gain .
Tests of significance of difference between
means: female GPA and MrScale . . .
Product-moment correlation coefficients:
male, rate gain . . . . . . . . . . . . .
First order partial correlation coefficients:
male, rate gain . . . . . . . .6. . . . .
Second order partial correlation coefficients:
male, rate gain . . . . . . . . . . . . .
Product-moment correlation coefficients:
male, comprehension gain . . . . . . .
First order partial correlation coefficients:
male, comprehension gain . . . . . . . .
Second order partial correlation coefficients:
male, comprehension gain . . . . . . . .
vi
Page
20
29
29
30
31
32
33
33
34
35
36
.37
Table Page
4.13 Product-moment correlation coefficients:
female, rate gain . . . . . . . . . . . . 38
4.14 First order partial correlation coefficients:
femalel rate gain 0 o o o o o o o o o o o 39
4.15 Second order partial correlation coefficients:
female! rate gain 0 o o o o o o o o o o o 39
4.16 Product-moment correlation coefficients:
female, comprehension gain . . . . . . . 40
4.17 First order partial correlation coefficients:
female, comprehension gain . . . . . . . 41
4.18 Second order partial correlation coefficients:
female, comprehension gain . . . . . . . 42
5.1 Frequency distribution--M-Scale, female . . . 45
5.2 Frequency distribution—-M~Scale, male . . . . 45
5.3 Variables which served as significant
predictors of gain in achievement . . . . 50
6.1 Major null hypotheses decisions by sex . . . 54
vii
CHAPTER I
FORMULATION AND DEFINITION OF THE PROBLEM
. Need for the Study
Variance in learning as reflected in tests of achieve-
ment remains an enigma. Why, for instance, will some pupils
who are able, according to tests of aptitude and ability, fail
to achieve at the anticipated level? Obviously, each individual
is the product of his unique experiences, and some experiences
may be intensive, even traumatic, to the extent of preventing
adequate learning. Since, however, average and below average
achievement is the rule rather than the exception, there must
be other common reasons. Among those reasons individual
academic motivation is certainly one of the most important.
Academic motivation has been considered only relative-
ly recently. In 1958, Allportl summarized the existing moti—
vational research and simultaneously offered guidance and
encouragement for future research.
The search for the units that comprise motivation
and compose personality is very ancient. Not until
the past generation or two has appreciable progress
been made. During recent years, however, we have
.1Gordon Allport, "What Units Shall We Employ?" Gardner
Lindzey, The Assessment of Human Motives (New York: Rinehart
& C091 InC., 1958), pp. 239—60.
1
followed a bewildering array of approaches, many of
them fresh and imaginative, and resulting in more
measured aspects than anyone can conveniently
compute. . . .
We have to accept the fact thatlupto now rela-
tively little agreement has been achieved. It seems
that each assessor has his own pet units and uses
a pet battery of diagnostic devices. But it is
too early to despair. Instead of discouragement,
I hope that our present disagreement will lead to
continuous and wholesome experimentation. Essential
to continued progress is a firm belief in the
"outer reality" of personal and motivational systems.
The fact that the units we seek are invisible should
not deter us. Nor should we yield to the destructive
skepticism of extreme methodologists who hold that
search is chimerical. Finally, while we must admit
the variabilities of the structure we seek, which
are caused by changing situations without and continual
growth and change within, we should take this fact into
our design and theory, not surrendering our belief that
reasonably stable personal and motivational structures
exist.
Background of the Study
The concerns of this study are the possible relation-
ships of (1) degree of academic motivation (high or low
academic need-achievement), (2) setting of levels of aspiration
and expectation (goal-setting), and (3) gain (or loss) in
reading rate and comprehension. Reviewers of motivational
study, specifically Allport?’ 3 and Lindzey,4 encourage
2Allport, op. cit., pp. 239-59.
3Gordon Allport, "The Trend in Motivational Theory,"
Understanding Human Motivation,.Stacey.& DeMartino, eds. ’
(Cleveland: HOward Allen, Inc., 1958), p. 64.
4Gardner Lindsey, The Assessment of Human Motives.(New
York: Rinehart & Co., Inc., 1958), p. 15.
experimentation which does not depend wholly upon projective
techniques but assumes that in a healthy personality, moti-
vation can be taken at its expressed value. The use of
levels of aspiration and expectation reflects the basic
philosophy that expressed aims are acceptable criteria of
motivation. I
Farquhar5 and associates have developed a motivational
task theory. It is an extension and modification of McClelland's6
theory and the research reported by Atkinson7 on the achievement
motive. A summary of the Farquhar g£_al. theory is given in
Table 1-1. The chief modification included in this theory is
a trend away from the use of projective devices and techniques.
On the basis of this theory, the Farquhar team developed
four instruments which comprise the Mdchiggn State Motivational
Scales (Mr-Scales).8 Scores derived from the MrScales are
5WilliamW‘. Farquhar, A Comprehensive Study of
Motivationgl Factors Uthglying Achievement of Eleventh Grade
High School Students (Research Project No. 846 (8458):
Supported by the U.S. Office of Education, in cooperation with
Michigan State University, 1959).
6D. McClelland and J. Atkinson, et al. The Achievement
Motive (New York: aAppleton-Century-Crofts, 1953).
7John Atkinson (ed.), Motives in Fantasy, Action and
Society (Princeton, N.J.: D. Van Nostrand Co., 1958).
8M-Scales consist of.four instruments: (1) Generalized
Situational Choice Inventory, (2) Preferred Job Character-
istics Scale, (3) WOrd Rating List, and (4) Human Trait Inventory.
used as the independent variable in the present study.
Table 1-1. Bipolar theory of high and low academic need
achievement motivation.
Motivational Situation:
High Academic Need-Achievement Low Academic Need-Achievement
1. Long Term Involvement 1. Short Term Involvement
2. Unique Accomplishment 2. Common Accomplishment
3. Competition with a Maximal 3. Competition with a Minimal
Standard of Excellence Standard
Every published study of the results of college level
reading improvement programs shows gain made by the class.
No study has yet reported reading improvement data and related
it to academic motivation. Results of this study of academic
motivation in a reading improvement situation may provide
help for an instructor to predict individual success and to
plan further for the expected failures and low-degree
successes .
Purpose of the Study
Essentially, the study describes the individuals'
academic motivation as reflected in the M-Scales, records
their levels of aspiration and expectation,and reports
their gain (or loss) in achievement. These data are studied
to determine the interrelationships among them.
The Hypotheses
To investigate possible relationships this study was
designed to provide answers to three major null hypotheses:
1.
There are no significant relationships between
academic need-achievement scores and gain or loss
in achievement.
There are no significant relationships between
academic need—achievement scores and the discrepancy
between final levels of aspiration and final levels
of expectatiOn.
There are no significant relationships among academic
need-achievement scores, final levels of aspiration,
final levels of expectation, and gain or loss in
achievement.
CHAPTER II
REVIEW OF THE LITERATURE
This chapter presents a review of the pertinent
studies and then relates their application to the present
study.
Pertinent Studies
Intention, as a psychological concept, passed into
disuse when reflex theory and its derivatives became~the
foundation for scientific theories of behavior. At least two
prominent psychology scholars refused, however, to abandon
intent as a variable. Lewis,1 in 1946, said, "All acts
have in common the character of being intended or willed. .
There is no obvious way in which we can say what act it is
which is thought of or done except by specifying this intent
of it."
Lewin2 had, in 1938, used the concept of intent
to combat the overly simple associationistic theory that
lClarence I. Lewis, An Analysis of Knowledge and
Valuation (LaSalle, Illinois: Open Court, 1946), p. 367.
2Kurt Lewin, "Intention, Will, and Need," Ch: 5,
Organization_§nd Pathology of Thought, David Rappaport,
(New York: Columbia University Press, 1951).
6
actions arestrengthened whenever they are successful.
Miller §£_al.3 have provided a useful bridge between these
early theories of intent and today's developing motivational
theories. They ask, "What does it mean when an ordinary man
_has an ordinary intention?" and answer, "He has begun the
execution of a Plan [their construct] and this intended
action is a part of it." They believe that the expression
of an intent refers to the uncompleted parts of either a
conscious or "unconscious" Plan, modeled after Tolman's4
cognitive learning theory. Conscious and "unconscious"
Plans are nearly always extensions of or alterations of
previous P1ans-—se1dom is a new Plan originated. Thus,
in short, we have Miller's Plans or Tolman's sign-signif-
icate r.e1ations-,, or a choice of several other constructs;
they all involve goals. If we admit goals we must admit
the inherent setting of goals:_goalesetting in the present
study consists of setting levels of aspiration and expecta-
tion.
Numerous unique variables affect each action since
each action is unique. This is the principal reason that
3 a
George A. Miller, Eugene Galanter, and Karl Pribram,
Plans and the Structure of Behavior,(New Ybrk: Henry Holt
& Company, 1960), pp. 59-71.
4E. C. Tolman, "Psychology Versus Immediate Experience,'
Philosophy of Science, Vol. 2 (1935), pp. 356—80.
the reported studies on aspiration and on expectation are
concerned always with a single type of action. A study
of reading improvement, for example, cannot produce results
inferable to other actions because reading background, in-
telligence, initial reading rate, mental health, physical
health, and emotional state are just some of the-variables
which are unique to each person for a particular action.
At the same-time, this is why results of aspiration studies
of subjects in a gambling situation, for instance, are non-
inferable to reading situations. Sutliffe5 expressed this
point in detail in 1955. The few studies which have been
reported are reviewed here, mainly to show the need for
such studies in specific academic areas. Most use only
the level of aspiration.
I
Atkinson 7 has twice reported on gambling situations.
In 1957 he concluded that people with a strong motive to
achieve prefer immediate risk whereas those with a strong
‘ 5J. P. Sutliffe, "Task Variability and the Level Of
Aspiration," Australian Journal of Psychology, Monograph
Supplement,_No. 2 (1955), 85 pages.
6John Atkinson, Jarvis Bastian, Robert Earl, and
George Litiven, "The Achievement Motive, Goal Setting, and
Probability Preferences," Journal of Abnormal Sociaersychology,
Jan., 1960, 60:27-36.
7John Atkinson, "Motivational Determinants of Risk-
taking Behavior," Psychological Review, 1957, 64:359-72.
motive to avoid failure prefer either easy taSks or extremely
difficult and risky tasks. The Farquhar gt_al. theory of
academic motivation includes strong motivation to avoid
failure with strong motive to achieve, illustrating once
again that motivation studies, to be meaningful, need to be
specific. Atkinson's 1960 study produced results similar
to his 1957 study, with the extension of that part of the
theory to include not only games of chance but also games
of skill.
DeSoto §t_§1,8 asked 96 college students to predict
their success on a many-valued achievement scale. The task
consisted of word association, and the actual sequence of
successes and failures had been already programmed by the
experimenters. They found (1) that the subjects predicted
their successes proportionately to their previous successes
and (2) that the subjects persisted in predicting more suc—
cesses than they actually attained. Even though the study
was set in an academic situation, the first two results
have little or nothing to add to a theory of academic moti-
vation. Their final finding, however, carried an implication
extremely helpful in designing the present study: "since
the subjects were admonished to be realistic in their
8Clinton DeSoto, Edmund Coleman, and Peter Putnam,
"Predictions of Sequences of Successes and Failures," Journal
of Experimental Psychology, Jan., 1960, 59:41-46.
pl
1'!
10
predictions, it was concluded that realism in predictions of
personal successes and failures is not as easily obtained as
realism in the setting of levels of aspiration on a mere
finely differentiated scale."
Predictions by individuals of their probable success
for a fixed goal were studied by Diggory §t_al.9 The goal was
to sort 40 of 70 cards into 10 sets within 25 seconds. They
found that individual prediction varied with individual rate
of progress, proximity to the goal, and distance from the
deadline (number of trials allowed) when this information was
supplied preceding each trial. The results of the Diggory
study substantiate empirically, long-standing educational
principles included in teaching people as individuals. The
subjects for the Diggory experiment were called High Motivated
because they volunteered: however, they were excused from
producing a term—paper, so the appellation of High Motivated
appears questionable.
10,
Kausler has contributed two studies involving
9James Diggory, Eugene Riley and Ruth Blumenfeld,
"Estimated Probability of Success for a Fixed Goal," American
Journal of Psychology, March, 1960, 73:41-55.
10Donald Kausler, "The Effects of a Qualitative Frame
of Reference on Level of Aspiration," Journal of Social
Psychology, Nov., 1958, 48:217-21.
11Donald Kausler, "Aspiration Level as a Determinant
of Performance," Jgurnal of Personality, Sept., 1959,
27:346-51.
11
level of aspiration in the area of academic motivation,
In the 1958 study three groups were administered a percep-
tual learning scale which was called an intelligence test.
One group was told that the test was very easy, a second
group that it was very difficult, and the third group was
told nothing regarding difficulty. The qualitative infor-
mation on difficulty did not significantly affect the levels
of aspiration but did increase the variabilities of the
performance scores.
In 1959, Kausler studied.the motivational properties
inherent in setting a level of aspiration. Three groups
performed on a simple arithmetic test. Group A was instructed
to express an aspiration level, group B did not express an
aspiration level, and group C was informed of the performance
scores of group B and then instructed to express levels of
aspiration. Performance scores of groups A and C were
significantly higher than for group B
For group A, Kausler partialledcnnzthe differences
in task aptitude and found no correlation between magnitude
of aspiration level and magnitude of performance score.
Group C, which set levels of aspiration after being informed
of the performance of group B, set higher levels of aspiration
than did group A but its mean performance score was not
significantly different from that of group A. However, in
12
group C the partial correlation between aspiration level
and performance level was significant at .01 level. Kausler
concluded that expression of a level of aspiration evokes
a set for optimal performance. This set, in turn, is in-
creased by inclusion of a frame of reference.
Of McClelland's12 considerable contributions to the
field of motivation study, the following quotation has had
the utmost significance to the design of the present study.
He states that the level of aSpiration, alone, is likely to
reflect unrealistic goal choices. "The trick seems to be
to get the subject to make a realistic aspiration statement
in a performance context without allowing it to be too much
influenced by the actual performance in that context. If
it is too much influenced, then the leVel of aspiration
becomes not a goal statement but something of an expectation
statement or perhaps an ego defense against failure."
Sivertsonl3 published a qualitative study of goal-
setting, level of aspiration, and social norms in 1957.
He found that success tended to be associated with low goal-
setting and failure with high goal-setting. Because his
12David McClelland, Personality (New York: Holt,
Rinehart & Winston, 1951).
3Dagfinn Sivertson, "Goal—Setting, the Level of
Aspiration, and Social Norms," Acta Psychologica, 1957,
13:54-60.
13
study was concerned with 11 and 12—year old boys, his findings
reflect only the norms of a sub-culture, but the sociological
findings may have significance for other groups. He found
that high goal categories were associated to a high degree
with various expressions of parental pressure and that
parental attitudes were more important to the subjects than
were material rewards.
Robayel4 arrived at similar qualitative conclusions
the same year. Her significant findings were that (1) level
of aspiration involves hope of success and issues from the
internalization of parental love, and (2) level of expectation
involves confidence in the result and emerges from a feeling
of acceptance anchored in secure family relationships.
Studying the discrepancy between level of aspiration
and achievement, Schultz and Ricciuti15 attempted to relate
the discrepancies to previous college achievement. They
found the correlation to be low and not significant. In
a negative manner their study supports Robaye's.
4 . . . .
Franc1ne Robaye, N1vaux d'Asp1rat1on et c'Expec-
tation (Levels of Aspiration and Expectation),(Paris:
Presses Universitaires de France, 1957).
5Douglas Schultz and Henry Ricciuti, "Level of
Aspiration Measures and College Achievement," Journal of
General Psychology, 1954, 51-267-75.
14
Finally, Robaye'sl6 1954 study provided the basis
for definition of levels of aspiration and expectation.
She stated that the level of aspiration is the goal that
the subject desires to attain and the level of expectation
is the goal which he expects to attain:»-Using 50 eleven
and 12-year-old girls she found the correlation between
performance and level of expectation not statistically
significant. According to Robaye, the subjects did not show
objective judgment regarding their capacities.
Application of the Literature to this Study
Prominent in the literature is the conclusion that
realism in setting goals is elusive and difficult to obtain.
Realism in setting goals is attempted in the present study
by having_the students set their goals twice but using only
the final expressions of aspiration and expectation. The
initial expressions of aspiration and expectation combined
with ten class hours of practice achievement are intended
to furnish a realistic frame of reference for the students'
1 . . . . .
6Franc1ne Robaye, "Sur La Dist1nct1on Entre Niveau
d'Aspiration et d'Expectation et sur la Valeur Diagnostique
de la Comparaison dans l'etude de la Personalité d'un Groupe
d'Adolescents." (Concerning the Difference between Level
of Aspiration and of Expectation and the Diagnostic Value
of Their Comparison in the Study of Personality in‘a Group
of Adolescents). Cahiers de Pedagogie de l'Université de
Liege, 1954, 13:78—86.
15
setting of final levels of aspiration and expectation.
Kausler'sl7 eonclusion that a set for optimal performance
is evoked by the expression of an aspiration level is a
principal reason for eliminating a control group: his findings
are qualitatively and quantitatively significant.
McClelland's18 advice is taken into account by
including both level of aspiration and level of expectation
in the study, thus eliminating the aspiration statement
reflecting expectation. Despite Schultz and Ricciuti's1
finding of low correlation between aspiration and past
academic achievement, the students are asked to report their
grade point average in the present study. Krumboltz20 has
shown that MSU students will reliably report their grade
point averages and these averages are inspected for possible
relationships in the present study because this study is
more elaborate than was Schultz and Ricciuti's. Finally,
Robaye?l has provided the definitions of aspiration and
expectation.
l7Kausler, op. cit., p. 15.
18McClelland, op. cit., p. 18.
19Schultz and Ricciuti, op. cit., p. 20.
20John D. Krumboltz, "A the on the Accuracy of
Self-Reported Grade Point Averages," M§U College-of Education
Quarterly, Winter, 1960, p: 26.
21Robaye, op. cit., p. 21.
16
Summary
The review of literature pertinent to the present
investigation has pointed up the diversity of specific in-
terests of the investigations. Goal-setting, under a myriad
of names, has.replaced intention as a psychological concept._
Research in motivation is plentiful but general.
Few studies of academic motivation in a truly academic setting
were found and none of these were concerned with the area
of reading improvement.
Of those-academic motivation studies found, definitions
of levels of aspiration and expectation have the widest
applicability. Most studies deplore the lack of realism in
setting levels of aspiration and expectation. Attempts
to overcome this problem have been included in the design of
the present study through two devices: (1) the setting of
initial and final levels of aspirati n and.expectation,
using only the final levels, and (2) the furnishing of a
further frame of reference, achievement of previous reading
improvement classes.
CHAPTER III
DESIGN AND METHODOLOGY
Operational Definitions
High academic need-achievement and low academic
need—achievement are defined in Table 1-1. These definitions
are scores derived from the MrScales responses, placed
along two continua, one contimuum for each sex. Arbitrary
division of each continuum into three parts obtains high,
medium, and low academic need—achievement. For the male
population scores 120 to 133 inclusive became high N-ach.,
scores 108 to 119 inclusive became medium N-ach., and 71 to
107 inclusive became low N-ach. For the female population
scores 111 to 127 inclusive were designated high Neach., 104
to 110 inclusive were designated medigmfN-agh., and 82 to
103 inclusive were designated low N-ach.
Level of aspiration is the quantitative indication
which an individual publicly makes (by recording it on a
sheet of paper for the instructor to keep) concerning his
desired future performance in an activity. Levels of
aspiration were recorded as words read per minute (wpm) and
total number of comprehension questions correctly answered
18
of a possible 35.
Level of expectation is the quantitative indication
which an individual publicly makes (by recording it on a
sheet of paper for the instructor to keep) concerning his
expected future performance in an activity. Levels of
expectation were recorded as words read per minute (wpm) and
total number of comprehension questions correctly answered
of a possible 35.
Initial level of aspiration and initial level of
expectation were recorded by the subjects when they had
(
just learned their performance scores (wpm and total correct
answers) on the Iowa Test of Silent Reading, Form Am, and
before they had become involved in actual reading improve-
ment procedures.
Final level of aspiration and final level of expec—
tation were recorded by the subjects during the tenth class
hour, when they had become involved in reading improvement
procedures, and after seeing again their initial performance
scores (Iowa Test, Am) and their initial levels of aspiration
and expectation.
Gain or loss in achievement consists of two differences:
(1) initial words read per minute (Iowa Test, Am) and final
words read per minute (Iowa Test, Cm), and (2) total number
of comprehension questions answered correctly of a possible
19
35 on the initial test and total number of comprehension
questions answered correctly of a possible 35 on the final
test.
The Population
The population consisted of all students (107 males,
38 females) enrolled in Reading Improvement (092) in the
University College, Michigan State University, during Winter
term, 1962.1 Although Reading Improvement is offered
primarily for freshmen and sophomores in the University
College, anyone who hasgraduated from high school may enroll
in the course, upon payment of a $15.00 fee. Past records
indicate that approximately 70%»of the enrollees have been
freshmen, and that upperclassmen, graduate students,
faculty members, and visitors have made up the remaining 30%.
Five instructors taught the nine sections.
The Sample
All members of the population completed the MrScales
during the first class period. Dividing the group into
sexes was necessary since the MrScales have proved to be
more discriminating for males than for females. These groups
were then divided into high academic need-achievement,
lSeven foreign students excluded.
20
medium academic need-achievement, and low academic need-
achievement, according to arbitrary division of the range
of scores into approximate thirds. Five males and two
females were excluded from the population because they were
foreign students and their lack of familiarity with American
English language and idiom resulted in spurious MrScales
scores. Of the 107 males in the population, eight dropped
the course, and of the 38 females, four dropped it. Popu-
lation and samples are depicted in Table 3—1. The entire
population participated in all experimental procedures.
I
Table 3—1. Population, sample, and sample losses.
Population Dropped Final
Course Sample
Males, High~ 36 2 34
Males, Med. 35 4 31
Males, Low 36 2 34
Total 99
Females, High 14 l 13
Females, Med. 10 2 8
Females, Low 14 1 13
Total 34
21
Chronology of Experimental Procedures
The first class period was devoted to completion of
the M—Scales. This class period was chosen for this activity
to avoid possible student-instructor identification being
reflected in the M—Scales answers.
During the second class meeting the students wrote
the Iowa Test of Silent Reading, Form.Am. For all forms
of the Iowa Test, test 1A measures rate and comprehension
on physical science material and test'lB measures rate and
comprehension on social science material. The rate was
computed in words read per minute (wpm) and is the average
of the students' rates for the two sub-tests. There are
35 possible correct answers to these sub-tests (10 physical
science and 25 social science). The comprehension score
was the total number of these 35 questions which were cor-
rectly answered.
Their Iowa Test results were shown to the students
during the fourth class meeting. After they had seen their
scores, as explained above, in order to furnish them a frame
of reference, the following paragraph was read to them:
In this course we will be concerned with im-
proving (1) rate of reading (wpm), (2) comprehension,
and (3) vocabulary. Past records indicate that the
average reading rate of a Reading Improvement gigss
doubles and the average comprehension improves
slightly.’ Individual improvement has not yet been
thoroughly studied: what you are going to be asked
22
to do is part of an attempt to study improvement
by individuals in a scientifically measurable manner.
When you are asked these questions, please make
your answers an accurate reflection of your own
feelings. '
Then the students were asked to record, at the side
of their Iowa Test Answer Sheets, their answers to these
four questions:
1.
What do you hope to read in me on another form
of the Iowa Test (similar material) at the end
of the term?
What do you expect to read in wpm on another
form of the Iowa Test (similar material) at
the end of the term?
Of the 35 possible (10 science, 25 social
studies) on another form of the Iowa Test at
the end of the term, what is the total number
of comprehension questions you hope to answer
correctly?
Of the 35 possible (10 science, 25 social
studies) on another form of the Iowa Test at
the end of the term, what is the total number
you expect to answer correctly?
The fifth, sixth, seventh, eighth, and ninth class
meetings were devoted to explanation and application of
SQBR,2
recognition and use of transitional words and phrases,
timed readings with comprehension checks, vocabulary study,
and numerous other procedures for developing rapidity and
2This method of reading and study developed by F. P.
Robinson in Effective Study (New York: Harper & Brothers,
1946),
p. 28.
23
efficiency of reading. By this time the students had seen
tangible evidence of reading improvement in their daily
work, and they were asked, during the tenth class meeting,
to record again their levels of aspiration and expectation.
They were shown again their first Iowa Test results and
their initial levels of aspiration and expectation, and
then asked to record their answers to the same four
questions which they were asked during the fourth class
meeting. These answers were labeled final levels of aspir-
ation and expectation.
The Iowa Test of Silent Reading, Form Cm, was
administered during the eighteenth class hour (to allow for
test correction time) and the results were shown the students
on the last day of the class. The average rate in words per
minute and the total number of correct comprehension answers
comprise the final achievement data.
The Null Hypotheses and Analysis Procedure‘
The basic null hypotheses of the investigation were
stated in Chapter I. Specific formulation of these
hypotheses for the independent variable, high or low academic
need-achievement are:
1. There are no significant differences in mean gain
in achievement.
24
2. There are no significant differences in the mean
discrepancy between final level of aspiration and
final level of expectation.
3. There are no significant differences among mean
final level of aspiration, mean final level of
expectation, and mean gain in achievement.
It was decided to use the upper and lower thirds of
each sex continuum of M—Scales scores to define high and
low academic need-achievement in order to optimally discrimi-
nate on the independent variable. When further investigation
of the major null hypotheses necessitated the computation of
correlation coefficients, the entire population data was included.
Statistical Treatment
To test all three null hypotheses the significance
of the difference between High N—ach. means and Low N-ach. means
was determined by entering the table of Students' t-distribution.
For the purpose of testing two groups in which n1 = n2, the
difference in the variance of the groups may be ignored
provided the t-table is entered with n-l degrees of freedom.3
The level of significance for rejecting or failing to reject
3George W. Snedecor, Statistical Methods (Ames,
Iowa: Iowa State College Press, 1946), p. 83.
25
the null hypotheses was set at the five per cent level of
confidence.
Product-moment correlation coefficients were computed
for the variables in null hypothesis three. Partial correlation
between each pair of variables was then computed, to reflect
the relationship of each pair after removing the influence
of the others.
Summary
The population was divided by sex and the independent
variable (high, medium, or low academic need-achievement)
was assigned to each approximate third of each sex, according
to M-Scale score: thus these people comprised the sample.
Operational definitions of high and low academic need-achieve-
ment, level of aspiration, level of expectation, and gain or
loss in achievement were presented.
The sample gain or loss in achievement was derived
from two forms of the Iowa Test of Silent Reading. The
students were asked to record levels of aspiration and
expectation.
These data were compared statistically by entering
Students' table of t-distribution and by correlation.
CHAPTER IV
ANALYSIS OF THE DATA
In this chapter the analysis of the data is presented
in two main sections: the first deals with tests of signif-
icance of differences between means and the second investi-
gates relationships among the variables through the product-
moment, first-order partial, and second-order partial
correlation coefficients. For the first section the high
and low academic need—achievement (MrScale) scores determine
the N: male N - 68 and female N = 26. Because of statistical
procedure inherent in the definition of the product-moment
correlation coefficient, for the second section malewa= 99
and female N = 34.
Males and females are considered separately throughout
the study because the M+Sca1e has been more discriminating
for males than for females. The analysis of data is
further subdivided to consider the variables in their .
relations to gain in words per minute (rate) and gain in
comprehension.
0‘
27
Major Null Hypotheses
Three major null hypotheses were advanced to investi—
gate relationships in the study:
1.
There are no significant relationships between
academic need-achievement scores and gain in achieve—
ment.
There are no significant relationships between
academic need—achievement scores and the discrepancy
between final levels of aspiration and final levels
of expectation.
There are no significant relationships among academic
need—achievement scores, final levels of aspiration,
final levels of expectation, and gain in achievement.
Major null hypothesis two was concerned with the
possibility that the expresSion of a level of aspiration does
not differ significantly from the expression of a level of
expectation. All three major null hypotheses were tested
using Students' t-distribution.
Complete investigation of major null hypothesis three
required the compilation of product—moment and partial cor-
relation coefficients.
28
Subsidiary Null Hypothesis
A subsidiary null hypothesis was investigated. The
students' reported MSU grade point averages were inspected
for relation to academic need—achievement scores. In the
Ip-test analysis, male N = 57 and female N = 20 and in the
correlational analysis male N = 85 and female N = 27. The
difference in N's for this hypothesis and the N's in the
major null hypotheses was due to those students who were first—
term freshmen without MSU grade point averages. The subsidiary
null hypothesis was: ‘
There are no significant relationships between
academic need—achievement scores and student-
reported grade point averages.
Analysis of Differences of Means, Male
Rate Gain
There were no significant differences between means
for any of the dependent variables considered in the major
null hypotheses and all three major null hypotheses were
accepted.
Comprehension Gain
There were no significant differences between the
means for any variables considered in the.major null hypotheses
29
Table 4.1. Tests of significance of difference between means:
male, rate gain.
Variables N t
Level of Aspiration 68 .946 Accept
Level of Expectation 68 1.540 Accept
Discrepancy between
Aspiration and
Expectation 68 -.543 Accept
Gain in Rate 68 .067 Accept
M1 = M2 reject if t > i 2.00 (.05. 66 d.f.).
Table 4.2. Tests of significance of difference between means:
male, compnehension gain.
Variables N t '
Level of Aspiration 68 .049 Accept
Level of Expectation 68 1.453 Accept
Discrepancy between
A8piration and
Expectation 68 1.642 Accept
Gain in Comprehension 68 .390 Accept
M1 = M.2 reject if t >.i 2.00 (.05, 66 d.f.).
Grade Point Average
For the subsidiary null hypothesis the difference
between the means was significant and this null hypothesis
was rejected.
30
Table 4.3. Tests of significance of difference between means:
male GPA and MeScale. "
Variables N t
Grade Point Average 57 10.80"?"b Reject
MrScale Score 68 19.093a Reject
M1 = M2 reject if t >.: 2.00 (.05, 66 d.f.).
aSignificant at .0005 level.
bReject if t >.: 2.02, (.05, 55 d.f.).
Analysis of Differences of Means, Female
Rate Gain
There were no significant differences between means
for any of the dependent variables considered in the major
null hypotheses and all three major null hypotheses were
accepted.
Table 4.4. Tests of significance of difference between means:
female, rate gain. '
Variables N t
Level of Aspiration 26 -.615 Accept
Level of Expectation 26 .988 Accept
Discrepancy between
Aspiration and
Expectation 26 .485 Accept
Gain in Rate 26 .745 Accept
M1 = M2 reject if t >.: 2.06 (.95, 24 d.f.).
31
Comprehension Gain
There were no significant differences between the
means for any variables considered in the major null hypotheses
and all three major null hypotheses were accepted.
Table 4.5. Tests of significance of difference between means:
female,comprehension gain.
Variables N t
Level of Aspiration 26 -.847 Accept
Level of Expectation 26 .671 Accept
Discrepancy between
Aspiration and
Expectation 26 —l.8003 Accept
Gain in Comprehension 26 2.0478 Accept
M1 = M2 reject if t.) i 2.06 (.05, 24 d.f.).
aSignifiCant at .10 level.
Grade Point Average
For the subsidiary null hypothesis the difference
between the means was not significant and this null hypothesis
was accepted.
32
Table 4.6. Tests of significance of difference between means:
female, GPA and M-Scale.
v . V‘ " 1‘
Variables N t
Grade Point Average 20 .318a Accept
M-Scale 26 2.760b'C Reject
M1"= M2 'aReJGCt if t > i 2.10 (.05. 18 d.f.).
bReJeCt if t >.: 2.06 (abs, 24 d.f.).
cSignificant at .02 level.
Correlation Coefficients: Male, Rate Gain
None Of the product-moment correlation coefficients
between the independent variable and the dependent variables
in the major null hypotheses were significant at the .05
level. Consideration of the discrepancy'between level of
aspiration and level of expectation as a variable was
discontinued at this point because the partialling out of the
various variables' influences accomplished the desired
information.
Academic motivation as measured by the MrScale, as
seen in Table 4.7, appeared to exert little influence on the
gain in rate, while the level of aspiration and.1evel of
lDiscrepancy between level of aspiration and level
of expectation as a variable was discontinued at this point
of the analysis also for male, comprehension: female, rate:
and female, comprehension, because all the r's were highly
significant.
33
expectation were highly significant variables affecting the
gain in rate. Levels of aspiration and expectation were
closely related.
Table 4.7. Product-moment correlation coefficients: male,
rate gain.
Lev. Asp. Lev. Exp. Gaianch.t. DiscnzAspt EXP.
M—Scale .119 .002 .019 .034
Lev. Asp. .8933'b .4098'b
Lev. Exp. .491a’b
aSignificant at .05 level. bSignificant at .01 level.
Table 4.8. First order partial correlation coefficients:
male, rate gain.
Lev. Aspiration, 3 = Lev.
Gain in Achievement.
Subscripts: l = M—Scale, 2
Expectation, 4
r 12.3 = .259a r 14.3 = .021
r 12.4 = .122 r 23.4 = .869a'b
r 13.2 = -.233a r 24.3 = -.074
r 13.4 = -.008 r 34.2 = .3073'b
r 14.2 = —.o33
aSignificant at .05 level. b Significant at .01 level.
The correlation between academic motivation as
measured by the M—Scale and level of aspiration was significant
after the influence of the level of expectation variable was
34
removed; however, the level of aspiration correlation with
gain in achievement was negative, small, and not significant
(Table 4.8). The correlation between academic motivation and
level of expectation was negatively significant after the
influence of level of aspiration was removed, and yet the
correlation between level of expectation and gain in achieve—
ment was positive and highly significant.
At this point, then, the M—Scale appeared to measure
positively male level of aspiration for rate of reading and
measure negatively male level of expectation for rate of
reading. The only variable which predicted gain in rate
of reading was the level of expectation.
Table 4.9. Second order partial correlation coefficients:
male, rate gain.
_———r
Subscripts: 1 = MrScale, 2 = Lev. of Asp., 3
4 = Gain in Ach.
Lev. of Exp.,
r 12.34 = .2613 r 24.13 = -.082
r 13.24 = -.232a r 34.12 = .3083'b
r 14.23 = .042
3Significant at .05 level bSignificant at .01 level.
After partialling out the effects of the other
variables, the resulting second-order partial correlation
coefficients showed little relation between academic
35
motivation as measured on the M—Scale and gain in reading
rate for males (Table 4.9). The effect of level of aspiration
on gain in male reading rate was also insignificant.
Academic motivation was significantly related to
level of aspiration (positive) and to level of expectation
(negative). The level of expectation was related to male
gain in reading rate positively and was highly significant.
On the basis of correlational analysis, major null
hypothesis one was accepted and major null hypotheses
two and three were rejected.
Correlation Coefficients: Male,
Comprehension Gain
None of the product-moment correlation coefficients
between the independent variable and the dependent variables
in the major null hypotheses were significant at the .05 level.
Table 4.10 Product-moment correlation coefficients:
male, comprehension gain.
Lev. Asp. Lev. Exp. Gain Ach. Disc., Asp.Exp.
a
M-Scale .113 .155 -.127 .059
Lev. Asp. .6448“b -.018 .
Lev. Exp. ‘—.4O9a'b
aSignificant at .05 level.
bSignificant at .01 level.
36
Level of aspiration and level of expectation were significantly
related, but only the level of expectation was significantly
related to gain in comprehension and it was negative (Tablew
4.10).
Table 4.11. First order partial correlation coefficients:
male, comprehension gain.
Subscripts: 1 = MeScale, 2 = Lev. Aspiration,
3 = Lev. Expectation, 4 = Gain in Comprehension.
r 12.3 = .017 r 14.3 = -.2113.
r 12.4 = .112 t 23.4 = .6983'b
r 13.2 = .108 t 24.3 = --.403E=1Ub
r 13.4 = .114 r 34.2 = -.ssoa’b
r 14.2 = -.130
aSignificant at .05 level. bSignificant at .01 level.
As seen in Table 4.11, when the influence of the level
of expectation variable was removed there was a significant
negative correlation between MsScale score and gain in
comprehension. Both level of aspiration and level of
expectation were highly significantly related to gain in
comprehension but, as shown in Table 4.11 the correlation
between level of aspiration and gain in comprehension cor-
relation was positive and the correlation between level of
expectation and gain in comprehension correlation was
negative.
37
Table 4.12. Second order partial correlation coefficients:
male, comprehension gain.
Subscripts: l = M—Scale, 2 = Lev. Aspiration,
3 = Lev. Expectation, 4 = Gain in Comprehension.
r-12.34 = .045 r 24.13 = --4083'b
r 13.24 = .051 r 34.12 = —._544alb
r 14.23 = -.223a
aSignificant at .05 level. bSignificant at .01 level.
As seen in Table 4.12 there was virtually no corre-
lation between MrScale score and either level of aspiration
or level of expectatiom. The correlation between MrScale
score and gain in comprehension was negative and significant
when the influence of the other two variables were removed.
The correlations between level of aspiration and gain in
comprehension and level of expectation and gain in comprehension
were also both negative but they were also highly significant,
after the other variables' influences had been removed.
All three major null hypotheses were rejected.
Correlation Coefficients for
Subsidiary Null Hypothesis
The product-moment correlation coefficient between
M-Scale score and grade point average for males was .013
and for females it was .077, neither of which was significant.
38
The subsidiary null hypothesis was accepted.
Correlation Coefficients, Female, Rate Gain
As was true with males, for females none of the
product-moment correlation coefficients between M-Scale
scores, the independent variable, and the dependent variables
in the major null hypotheses were significant at the .05
level (Talbe 4.13). Female level of aspiration correlated
positively with gain in rate and was highly significant;
level of expectation was positively correlated with gain in
rate but was significant only at .05 level.
Table 4.13. Product-moment correlation coefficients:
female, rate gain.
Lev. Asp. Lev. Exp. Gain, Ach. Disc. Asp. Exp.
M—Scale -.047 -.059 .135 .239
Lev. Asp .9293!b .481‘="b
Lev. Exp. .391a
aSignificant at .05 level. bSignificant at .01 level.
Females were apparently realistic in expressing a
level of aspiration,because the correlation coefficient
between level of aspiration and level of expectation, for
rate, was .929 and highly significant (Table 4.13).
39
Table 4.14. First order partial correlation coefficients:
female, rate gain.
Subscripts: l - MeScale, 2 = Lev. Aspiration,
3 = Lev. Expectation, 4 = Gain in Ach.
r 12.3 = -.276 r 14.3 = .172
r 12.4 = -.129 r 23.4 = .9183'b
r 13.2 = .041 r 24.3 = .347a
r 13.4 = -.123 r 34.2 = -.173
r 14.2 = .180
aSignificant at .05 level. bSignificant at .01 level.
When first order partial correlation coefficients
were computed only two coefficients were significant (Table
4.14). Level of aspiration continued to correlate at the
.01 level with level of expectation. Level of aspiration
correlated at the .05 level with gain in rate, after the
influence of expectation was removed, but when the aspiration
variable was partialed out, level of expectation did not
correlate significantly with gain in rate.
Table 4.15. Second order partial correlation coefficients:
female, rate gain.
Subscripts: l = MeScale, 2 = Lev. Aspiration,
3 = Lev. Expectation, 4 = Gain in Ach.
r 12.34 =- -.041 r 24.13 = .348a
r 13.24 = -.013 r 34.12 = .183
r 14.23 =' .175
aSignificant at .05 level.
40
As seen in Table 4.15, after second order correlation
coefficients were computed, only one variable, level of
aspiration, was significantly related to gain in rate for
females. Major null hypotheses one and two were accepted
and major null hypothesis three was rejected.
Correlation Coefficients, Female
Comprehension Gain
None of the product—moment correlation coefficients
between the independent variable, M-Scale scores, and the
dependent variables in the major null hypotheses were signif-
icant at the .05 level (Table 4.16).
t d ‘“
Table 4.16. Product-moment correlation coefficients:
female, comprehension gain.
Lev. Asp. Lev. Exp. Gain, Ach. Disc. Asp. Exp.
MrScale -.l44 .216 .232 -.201
Lev. Asp. .726a,b _.923a,b
Lev. EXP. _.493a,b
aSignificant at .05 level. bSignificant at .01 level.
Three highly significant coefficients are shown in
Table 4.17: aspiration and expectation were positively
related; both level of aspiration and level of expectation
correlated negatively with gain in comprehension.
41
Table 4.17. First order partial correlation coefficients:
female, comprehension gain.
MrScale, 2 = Lev. Aspiration,
Subscripts: l
3 = Lev. Expectation, 4 = Gain in Comprehension
r 12.3 = -.449a r 14.3 = .398a
r 12.4 = .187 r 23.4 = .809a'b
r 13.2 = .5243'b r 24.3 = —.943a'b
r 13.4 = .120 r 34.2 = .668a'b
r 14.2 = .289
aSignificant at .05 level. bSignificant at .01 level.
As shown in Table 4.18, there was a negative corre-
lation between MeScale and level of aspiration, significant
at the .05 level, after the expectation influence had been
removed. After removing the influence of aspiration the Mr
Scale scores correlated positively with level of expectation
significant at the .01 level. MrScale correlated significantly
with gain in comprehension after removing the influence
of expectation, but did not correlate significantly when
only aspiration was removed. The two most important
variables affecting gain in comprehension for females were
the level of aspiration (negative and highly significant)
and level of expectation (positive and highly significant).
42
Table 4.18. Second order partial correlation coefficients:
female, comprehension gain.
Subscripts: l = MrScale, 2 = Lev. Aspiration,
3 = Lev. Expectation, 4 = Gain in Comprehension
r 12.34 = .154 r 24.13 = 4.9323"b
r 13.24 = -.054 r 34.12 = .6355"b
r 14.23 = -.084
aSignificant at .05 level. bSignificant at .01 level.
As shown in Table 4.19, after partialling out the
effects of the other variables, there was little correlation
between MrScale scores and any of the three dependent
variables for comprehension. An almost perfect negative
correlation existed between level of aspiration and gain in
comprehension and a highly significant positive correlation
existed between level of expectation and gain in comprehension.
Major null hypotheses one and two were accepted and major
null hypothesis three was rejected.
Summary
The analysis of data was presented in two main sections
according to the statistical procedures used for the investi-
gation. The first section dealt with tests of significance
of differences between means, using only the high academic
need achievement data and the low academic need achievement
43
data. The second section investigated relationships among
the variables through product-moment, first order, and second
order correlation coefficients. Each of these main divisions
was subdivided to treat the data according to sex and
these subdivisions dealt with (1) reading rate gain and
(2) comprehension gain. No significant differences between
means for the major null hypotheses were found. For the
subsidiary null hypothesis the resulting difference between
means for male M—Scale score and grade point average was
significant at the .0005 level. The difference between means
under the same subsidiary null hypothesis for females was not
significant. When the larger N's were used for calculation
of the product-moment correlation coefficients for the
subsidiary null hypotheses, no significant correlations were
found.
For further investigation of the major null hypotheses,
28 product-moment correlation coefficients, 36 first order
correlation coefficients, and 20 second order partial cor-
relation coefficients were calculated. On the basis of these
coefficients, major null hypotheses 1, 2, and 3 were rejected
for males and only major null hypothesis 3 was rejected for
females. Discussion of these results will be presented in
the next chapter.
CHAPTER V
DISCUSSION OF THE RESULTS
In Chapter V the results derived from the analysis
of data are interpreted and discussed. There are three
principal dimensions of the discussion: (1) M—Scale and
its use in a college reading improvement situation, (2) level
of aspiration and its use in a college reading improvement'
situation, (3) level of expectation and its use in a college
reading improvement situation. Each of these principal
divisions deals separately with sex and achievement variable.
M—Scale and Its Use in a College
Reading Improvement Situation
The frequency distributions of M-Scale scores are
depicted in Table 5.1. Although the female N is small,
the distribution appears more normal than does the distri—
bution of male scores, which appear to be negatively
skewed.
Male, Rate Gain
When the influence of the other variables had been
partialled out, the M-Scales scores correlated significantly
44
45
Table 5.1. Frequency distribution--M~Scale, female.
70- 75— 80- 85- 90- '95- 100- 105- 110- 115- 120- 125- 130- I
74 79 84 89 94 99 104 '109 114 119 124 129 134
34
2
ll
Table 5.2. Frequency distribution--M~Scale, Male
1 0 l 3 5 . 10 8 12’ 8 17 19 8 7
70- 75- 80- 85- 90- 95— 100- 105— 110- 115- 120- 125- 130-
74 79 84 89 94 99 104 109 114 119 124 129 134
N=99
46
with both level of aspiration (.261) and level of expectation
(-.232). But the correlation between MeScales scores and gain
in rate for males (.042) was low; decidedly the M—Scales
did not predict gain in rate for males.
Male, Comprehension Gain
The second order partial correlation between M-Scale
scores and gain in comprehension for males was (-.223) and
significant at the .05 level. This indicates that a high-
motivated male would or could gain little in comprehension and
a low-motivated male would gain much in comprehension,
because of regression toward the mode occurring whenever the
total possible is limited.
Female, Rate Gain
No significant relation was found between MrScales
and female rate gain (.175).
Female, Comprehension Gain
No significant relation was found between MrScales
scores and female comprehension gain (-.084).
1It should be noted, however, that Aspiration and
Expectation r's were even higher negatives and significant
at the .01 level.
47
Level of Aspiration and Its Use in a
College Reading Improvement Situation
Level of Aspiration correlated negatively, generally,
with gain in achievement for males. This would indicate that
the male student is prone to overestimate his ability to
improve and this is consistent with previous findings.2
For females both correlations were significant.
Male, Rate Gain
No significant relation was found between Level of
Aspiration and gain in rate for males (—.082).
Male, Comprehension Gain
The relation between level of aspiration and gain in
comprehension for males was significant but negative (-.408).
A lack of realism in setting the level of aspiration is
indicated, as was true for MrScale scores and level of
expectation for males.
Female, Rate Gain
Level of aspiration was the only one of the three
variables investigated which showed any significant relation
to female gain.in.rate (.384). It would appear that many
'2DeSoto et al., Diggory et al., Kausler and
McClelland, op. cit.
48
females realistically state their hopes for rate gain and
then proceed to gain that for which they hoped.
Female, Comprehension Gain
Generally, females expressed either high levels of
aspiration and then gained little in comprehension or low
levels of aspiration and then gained much in comprehension.
The correlation (-.932) was significant at the .01 level.
An extreme lack of realism was operating in female expression
of level of aspiration, but it serves as a useful predictor.
Level of Expectation and Its Use in a College
Reading Improvement Situation
Expression of a level of expectation-literally
separated the men from the women. Wherever the level of expec-
tation served as a positive predictor of gain in achievement
for males it did not for women.
Male, Rate Gain
The correlation here was .308 and was significant at
the .01 level. Where the level of aspiration had shown ex-
tremely little correlation with male rate gain, the-level of
expectation was the only significant predictor of gain in
rate for_males.
49
Male, Comprehension Gain
All three variables investigated for their predictive
ability were significant and negative for this variable;
however, level of expectation showed the highest correlation
(-.544). Low achieving males apparently expected a high gain
in comprehension and high achieving males expected little
gain.
Female, Rate Gain
The correlation between level of expectation and gain
in rate for females was low (-.183) and not significant. For
female rate gain, level of aspiration was the sole predictor.
Female, Comprehension Gain
Level of expectation was the best positive predictor
of female gain in comprehension (.635). Most females gained
in comprehension near to what they expected. Interestingly,
female level of expectation correlated almost a perfect
negative with level of aspiration.
Summary
Interpretation and discussion of the results of the
analysis of the data were presented in Chapter V. The three
predictive variables were investigated to determine their
50
usefulness in a college reading improvement situation. Table
5.3 shows those second order partial correlations which were
significant. For this sample, and by inference, for the reading
improvement population at Mﬂchigan State University, the M-
Scales were of little predictive value.
Table 5.3 Variables which served as significant predictors
of gain in achievement.
Male Female
Rate Comp Rate Comp
M-Scale -.223a
Level of Aspiration -.4083'b .348a —.932a'b
Level of Expectation .3083!b -.5443Ib .6353!b
aSignificant at .05 level. bSignificant at .01 level.
Level of aspiration predicted positively for females
on rate gain and negatively for both sexes on comprehension
gain. Level of expectation predicted positively for male
rate gain and female comprehension gain, and it predicted
negatively for male comprehension gain.
CHAPTER VI
SUMMARY, CONCLUSIONS, AND RECOMMENDATIONS
Summary
The summary is presented in three sections: the
problem, the design and procedure of the study, and the analysis.
The Problem
The major concern of the study was to investigate the
relationships among (1) level of academic need—achievement as
measured by the MrScales, (2) level of aSpiration, (3) level
of expectation, and (4) gain (or loss) in reading rate and
comprehension. All students from Continental United States
homes who were enrolled in Reading Improvement (092), University
College, Michigan State University, during Winter term, 1962,
comprised the sample.
The Design and Procedure of the Study
All students completed the MrScales during the first
class meeting. Because the MrScales have, to date, proved
more discriminating for males than for females, the entire
analysis was conducted by sub-sample for each sex. There were
99 males and 34 females. Arbitrary division of these two
51
52
ranges of M—Scales scores into approximate thirds defined high
academic need-achievement, medium academic need-achievement,
and low academic need-achievement.
During the second class meeting the students wrote the
Iowa:gest of Silent Reaging (Form Am). Subtests 1A and 1B
yielded reading rates and comprehension scores. After the
students had inspected their Iowa Test scores, during the fourth
class meeting, they were read a paragraph of general information
regarding achievement gains by past reading improvement classes.
They were then asked to record their initial levels of aspir-
ation and expectation in relation to their future achievement
on another form of the Iowa Test. Both level of aspiration and
level of expectation were recorded as (1) words to be read per
minute, and (2) number of correct answers of a possible 35.
After five hours of actual involvement in reading
improvement procedures, the students recorded final levels of
aspiration and expectation. These were the measures used in
the analysis because logic dictated that they should be more
realistic than the initial ones. The Cm form of the Iowa Test
was administered during the eighteenth class hour. The two
forms of the Iowa Test produced the variables, gain in achieve-
ment, in the form of rate and comprehension differences.
The Analysis
Three major null hypotheses were advanced:
53
1. There are no significant relationships between
academic need-achievement scores and gain or loss
in achievement.
2. There are no significant relationships between
academic need-achievement scores and the discrepancy
between final levels of aspiration and final levels
of expectation.
3. There are no significant relationships among academic
need-achievement scores, final levels of aspiration
final levels of expectation, and gain in achievement.
The level of confidence for rejection of the major
null hypotheses was set at the five per cent level.
A subsidiary null hypothesis was investigated:
There are no significant relationships between
academic need-achievement scores and student-reported
grade point averages.
Two statistical procedures were employed for the
analysis: (1) the significance of the difference of means
of each variable for high and low academic need-achievement
scores was tested entering Students' t-table: (2) using the
entire range of scores, product-moment, first order partial,
and second-order partial correlation coefficients-were com-
puted.
For the three major null hypotheses (both sexes)
54
no significant differences between means were found. For the
subsidiary null hypothesis the difference between means was
significant at the .00005 level for males but was not signifi-
cant for females.
The major null hypotheses, on the basis of significance
at the .05 level of the second order correlation coefficients
were rejected or accepted as reported in Table 6.1.
Table 6.1. Major null hypotheses decisions by sex.
Major Null Hypotheses:
There are no signif-
icant relationships
between (or among) Male Female
academic need—achieve—
ment scores and: Rate Comp Decision Rate Comp Decision
#1 gain or loss in Not Not Not
achievement Sig. Sig. Reject Sig. Sig. Accept
#2 the discrepancy
between final
levels of aspir-
ation and final Not Not Not
level of expec- Sig. Sig. Reject Sig. Sig. Accept
tation
#3 final levels of
aspiration, final
levels of expec-
tation, and gain
or loss in
achievement Sig. Sig. Reject Sig. b
a
Sig. Reject
aOnly aspiration and rate gain were significant.
b I I I 0
Only asp1rat1on, eXpectation, and comp. gain were
significant.
55
Conclusions
The conclusions of this study can be generalized
only to the Winter term, 1962 population of Reading Improvement
(092), University College, Michigan State University. Because,
however, the entire population was used, randomness would have
been superfluous, so the conclusions may safely be generalized
to any winter term population of Reading Improvement (092)
at the University College, Michigan State University.
Variables which were not taken into account during the
design of the study may have confounded the results. Using
the final levels of aspiration and expectation undoubtedly
brought these variables closer to reality than would have
occurred with the use of the initial ones, but the "set" which
Kauslerl proposed is evoked by the setting of aspiration levels
undoubtedly was operating here without control. Secondly,
inherent in the successful teaching of a reading improvement
course is the motivating of students to succeed: it is
possible that the entire population would have been rated
high in academic need—achievement had the M—Scales been
administered on the last day of the course. The assessment
of the $15.00 fee for Reading Improvement probably intervened
to exclude some low motivated students from enrolling in the
lKausler, op. cit., p. 11.
56
course, and is likely responsible for the skewed male distri-
bution. Finally, it remains to be proved that the MrScales
provide valid measures of academic motivation in college
students. The M-Scales were originally valided on eleventh
grade public school students.
The conclusion reached is that the M—Scales, alone,
provided little predictive validity for males and none for
females. In conjunction with the setting of levels of aspir-
ation and expectation the M—Scales were of some value for
males and of very little value for females. Consistently
throughout the study, either the level of aspiration, level of
expectation, or a combination of both were generally high pre-
dictors of achievement gain.
Recommendations
A similar study with one exception is desirable: one
in which the M—Scales are administered at the end of the course.
Another investigation might be an expanded version of the
present study but having one cell set levels of aspiration, one
cell set levels of expectation, one cell write the MrScales,
and one serve as the control group.
Academic motivation in a college reading improvement
situation needs to be studied in conjunction with some measure
of intelligence. Another useful study would investigate the
57
relationships between academic motivation and growth in various
aspects of reading skill such as vocabulary, ability to infer,
visualize, criticize, and summarize.
Finally, based upon the conclusions of the study, it
appears that experimentation with levels of aspiration and
expectation might prove beneficial in almost every academic
area. There lies in these goal-setting devices the evocation
of the set toward a maximal effort on the part of the student.
The effectiveness of the various approaches to programmed
learning could be immeasurably increased by inclusion of the
students' setting of levels of aspiration and expectation.
BIBLIOGRAPHY
Atkinson, John (ed.). Motives in Fantasy, Action, and
Society. Princeton, N.J.: D. Van Nostrand Co.,
1958.
Atkinson John. "Motivational Determinants of Risk-taking
Behavior," Psychological Review, 1957, 64:359-72
Atkinson, John et al. "The Achievement Motive, Goal Setting,
and Probability Preferences," Journal of Abnormal
Social Psychology, Jan., 1960, 60:27-36.
DeSoto, C., Coleman, E., and Putnam, P. ."Predictions of
Sequences of Successes and Failures," Journal of
Experimental Psychology, Jan., 1960, 59:41-46.
Diggory, J., Riley, E., and R. Blumenfeld. "Estimated
Probability of Success for a Fixed Goal," American
Journal of Psychology, March, 1960, 73:41-55.
Edwards, Allen. Experimental Design in Psychological Research.
New York: Rinehart and Company, 1960.
Farquhar, William W. A Comprehensive Study of Motivational
Factors Underlyinngchievement of Eleventh Grade
High School Students. (Research Project No. 846
(4858): supported by the U.S. Office of Education,
in cooperation with Michigan State University, 1959).
Hall, C., and Lindzey, G. Theories of Personality. New
York: John Wiley, 1957.
Hilgard, Ernest. Theories of Learning. New York: Appleton-
Century-Crofts, 1956.
Kausler, Donald. "Aspiration Level as a Determinant of
Performance," Journal of Personality, Sept., 1959,
27:346-51.
. "The Effects of a Qualitative Frame of
Reference on Level of Aspiration," Journal of Social
Psychology, Nov., 1958, 48:217-21.
58
59
Krumboltz, John. "A Note on the Accuracy of Self-Reported
Grade Point Averages," Michigan State University
College of Education Quarterly, Winter, 1960.
Lewis, Clarence. An Analysis of Knowledge and Valuation.
LaSalle, 111.: Open Court, 1956.
Lindzey, Gardner (ed.). The Assessment of Human Motives.
New York: Rinehart and Co., Inc., 1958.
McClelland, David. Personality. New YOrk: Holt, Rhinehart
and Winston, 1951.
McClelland, D., and Atkinson, J. et al. The Achievement
Motive. New YOrk: Appleton-Century-Crofts, 1953.
Miller, G., Galanter, E., and Pribram, K. Plans and the
Structure of Behavior. New York: Henry Holt and
Company, 1960.
Rappaport, David. Organization and Pathology of Thought.
New Yerk: Columbia University Press, 1951.
Robaye, Francine. Nivaux d'Aspiration et d'Expectation
(Levels of Aspiration and Expectation). Paris:
Presses Universitaires de France, 1957.
, . "Sur La Distinction Entre Niveau d'Aspiration
et d'Expectation et sur La Valeur Diagnostique de
>la Comparaison dans l'etude de la Personalité d'un
Groupe d'Adolescents (Concerning the Difference between
Level of Aspiration and of Expectation and the Diag-
nostic Value of Their Comparison in the Study of
Personality in a Group of Adolescents)." Cahiers
de Pedagogie de l'Université de Liege, 1954, 13:78-86.
Robinson, F. P. ngective Study. New Ybrk: Harper and
Brothers, 1946.
Schultz, D. and Ricciuti, H. "Level of Aspiration Measures
and College Achievement," Journal of General
Psychology, 1954, 51:267-75.
Sivertson, Dagfinn. "Goal-Setting, the Level of Aspiration,
and Social Norms," Acta Psychologica, 1957, 13:54-60.
60
Snedecor, George. Statistical Methods. Ames, Iowa: Iowa
State College Press, 1956.
Stacey, C. and DeMartina, M. (eds.). Understanding Human
Motivation. Cleveland: Howard Allen, Inc., 1958.
Sutliffe, J. P. "Task Variability and the Level of
Aspiration," Australian Journal of Psychology,
Monograph Sppplement, 1955, No. 2.
Tolman, E. C. "Psychology Versus Immediate Experience,"
Philosophy of Science, 1953, Vol. 2, pp. 356-80.
Walker, H. and Lev, J. Statistical Infsrence. New Yerk:
Henry Holt and Company, 1953.
,.1:(
r
"1.
... -‘7 ‘4
t8 3:: 1" “
soar (.3399
C., at) .541":- !\ lq/érl
, ﬂ.“
,1“
"Illlllllllﬂllllll“