University of Southern California Dissertations and Theses

The interactive effects of occupational therapy students' learning style with teaching methods (lecture vs. group-discussion) on their problem-solving skills, achievement, study time and attitude...

(USC Thesis Other)

The interactive effects of occupational therapy students' learning style with teaching methods (lecture vs. group-discussion) on their problem-solving skills, achievement, study time and attitude...

PDF

Download

Open document

Flip pages

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content THE INTERACTIVE EFFECTS OF
OCCUPATIONAL THERAPY STUDENTS' LEARNING STYLE
WITH TEACHING METHODS (LECTURE VS. GROUP-DISCUSSION),
ON THEIR PROBLEM-SOLVING SKILLS,
ACHIEVEMENT, STUDY TIME AND ATTITUDE:
AN APTITUDE-TREATMENT INTERACTION (ATI) STUDY
by
Noomi Katz
A Dissertation Presented to the
FACULTY OF THE GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(Education)
August 1981
UMI Number: DP24781
All rights reserved
INFORMATION TO ALL USERS
The quality of this reproduction is dependent upon the quality of the copy submitted.
In the unlikely event that the author did not send a complete manuscript
and there are missing pages, these will be noted. Also, if material had to be removed,
a note will indicate the deletion.
Dissertation Publ sh*ng
UMI DP24781
Published by ProQuest LLC (2014). Copyright in the Dissertation held by the Author.
Microform Edition © ProQuest LLC.
All rights reserved. This work is protected against
unauthorized copying under Title 17, United States Code
ProQuest LLC.
789 East Eisenhower Parkway
P.O. Box 1346
Ann Arbor, Ml 48106-1346
UNIVERSITY OF SOUTHERN CALIFORNIA
T H E G R A D U A T E S C H O O L
U N IV E R S IT Y P A R K
LO S A N G E L E S . C A L IF O R N IA 9 0 0 0 7
This dissertation, written by
............. Noomi Katz
under the direction of h.^C... Dissertation Com
mittee, and approved by all its members, has
been presented to and accepted by The Graduate
School, in partial fulfillment of requirements of
the degree of
t vi C *
Ph ,T>.
'%\
D O C T O R O F P H I L O S O P H Y
Dean
DISSERTATION COMMITTEE
Chairman
ii
ACKNOWLEDGEMENTS
I am very grateful to Hadassah Organization and
to the School of Occupational Therapy, Hebrew University
Jerusalem, Israel, which enabled my graduate studies and
made this research project possible. Special thanks to
the faculty members of the Occupational Therapy School
in Jerusalem, especially to its chairperson Bela Efrati,
for their friendly support and encouragement.
I wish to express special gratitude to Claudia
Allen for introducing me both to her theory "Cognitive
Disability: Occupational Therapy Assessment and Manage
ment," which served as a cornerstone in my master's
thesisf and to the content area of the present Aptitude-
Treatment Interaction (ATI) study. Her friendly support
and encouragement throughout my graduate studies and
especially her active collaboration In preparing the
materials and instruments, as well as her assistance in
teaching the course In this study, made this research
project possiblef and helped me all along the way.
I would like to thank also Dr. Stephen Abrahamson,
my committee chairperson and Dr. Elizabeth Yerxa for
their guidance and understanding.
iii
Finally, I wish to express special thanks to my
family— my husband Chaim and my children Yuval, Chagai,'
and Ayelet--for their support, understanding and willing
ness to participate in my educational adventure in the
United States.
iv
TABLE OF CONTENTS
v „ . . -
Page
ACKNOWLEDGEMENTS. ii
LIST OF TABLES. . . .................................... vi
LIST OF FIGURES . ........... ix
LIST OF ABBREVIATIONS................................. xi
Chapter ' T „ >
I. THE PROBLEM........... 1
Introduction ......... 1
Background of the Problem . . ........... 1
Statement of the Problem and Hypotheses . 10
Purpose and importance of the Study . . . 12
Assumptions . ........................... 13
Definitions of Terms ...................... 16
Delimitations ............................. 2 3
II. LITERATURE REVIEW............................ 2 5
Introduction ............................... 25
The Teaching/Learning Process ........... 25
Aptitude-Treatment Interaction ........... 32
Individual Learning Style . ............... 4 0
Teaching Methods (Lecture, vs. Group
Discussion)............................... 48
Problem-Solving Skills in Health •
Professions ...................... 58
Achievement, Attitude Toward Instruction,
and Amount of Study T i m e ............... 5 8
Summary.................................... 71
V
Chapter Page
III.: METHOD........................................... 73
Introduction . ............................. 73
Research Design........... ............... 73
Subjects and Sampling Strategy 7 6
Instrumentation............... 78
Data Gathering . . . ...................... 89
Data Analysis . ............... 93
Methodological Limitations ............... 96
IV. RESULTS.................. 98-
V. SUMMARY, CONCLUSIONS, DISCUSSION, AND
RECOMMENDATIONS . . . . . . . 153
Summary of the Study ................. 153
Major Results ........................... 158
Hypotheses . . . . . ......................16 2
Conclusions and Discussion........... 164
Recommendations........... ...............191
REFERENCES................................ 195
APPENDIXES. . . . . . . . . . . . . 208
A. Informed Consent Form.........................20 9
B. Student Ratings of Situations That Facilitate
Their Learning....................... 211
C. Patient Management Problems (PMP)................ 215
D. Student Evaluation............... , 222
E. Correlation Matrix. ........................... 226
F. Learning Style Inventory ..................... 230
vi
LIST OF TABLES
Table Page
1. Features of the Two Teaching Methods
Investigated in This Study.............. 58
2. Test-Retest Reliability of the L S I ..... 80
3. Correlations Between LSI Scales and Two
Items of the Students' Rating Scale .... 83
4. Comparison of Achievement Scores Between
Experts and Students........................ 8 6
5. Comparisons of PMP Overall Competence
Scores Between Experts and Students .... 87
6. Correlations Between the PMP Scores and
the Multiple-Choice Exam................ 88
7. Descriptive Statistics of Independent
Variables, Means, Standard Deviations, and
Frequencies for Each Treatment......... 99
8. Means and Standard Deviations of Outcome
Variables, Achievement and Problem-Solving,
for Each Treatment....................... 100
9. Means and Standard Deviations of Outcome
Variables, Study Time and Attitude Toward
Instruction, for Each Treatment....... 101
10. Summary of Stepwise Regression Analysis on
Overall Course Score (z) and Achievement . . 10 3
11. Summary of Stepwise Regression Analysis on
PMPs Overall Competence . . .105;
12. Summary of Stepwise Regression Analysis on
PMP Proficiency and Efficiency (first case)' 106,
13. Summary of Stepwise Regression Analysis on
PMP Proficiency and Efficiency (second case) -.113
vii
Table Page
14. Summary of Stpewise Regression Analysis on
Study Time for Every Class and Exams .... 114
15. Summary of Stepwise Regression Analysis on
Attitude Toward Instruction-Benefit and
Confidence............. ......................
121
16. Summary of Stepwise Regression Analysis on
Attitude Toward Instruction-Enjoyment and
Overall Score . . . .........................
122
17. Regression Equations for Outcomes Showing
the Unstandardized Partial Regression
Coefficients for the Aptitudes in Each
Treatment Group ............................. 129
18. Regression Equations for Outcomes Showing
the Unstandardized Partial Regression
Coefficients for the Aptitudes in Each
Treatment Group .............................
130.
19 . 2-Way Analysis of Variance, Treatment by
Learning Stype Type on Achievement and
Problem-Solving ............................. 131
•
o
CN
2-Way Analysis of Variance, Treatment by
Learning Style Type on Study Time and
Attitude Toward Instruction ...............
132
21. Descriptive Statistics of Independent Variables
Means, Standard Deviations, and Frequencies
for Two Educational Levels ..................
r
14 3
22. Means and Standard Deviations of Outcome
Variables, Achievement and Problem-Solving
for Two Educational Levels ...............
14 4,
23. Means and Standard Deviations of Outcome
Variables, Study Time and Attitude Toward
Instruction, for Two Educational Levels . .
14 5.
24 . 2-Way Analysis of Variance, Treatment by
Educational Level on Study Time and
Attitude Toward Instruction ...............
14 6
viii
Table Page
25. 2-Way Analysis of Variance, Treatment by
Educational Level on Achievement and
Problem-Solving .........................148
26. 3-Way Analysis of Variance, Treatment x
Educational Level x Learning Style Type on
Achievement and Problem-Solving ............. 14 9
27. 3-Way Analysis of Variance, Treatment x
Educational Level x Learning Style Type on
Study Time and Attitude Toward Instruction 150
28. Summary of Variables Which Show Significant
Contributions in Explaining Variance for
All Outcomes . . . . . . . . . . . ......... 185
29. Trends Combining the Three Dimensions:
Learning Type, Educational Level, and
Treatment for All Outcomes..................187
ix
LIST OF FIGURES
Figure Page
1. Major Variables in the Theory of School
Learning . . ................................. 26’
2. The Relations Between Events of Learning,
Outcomes of Learning, and Conditions of
Learning...................................... 2 7
3. Comparison of the Experimental Learning
Model and the Problem-Solving Process . . . 42
4. The Study's Design and Matching Hypothesis 74
5. Regression of Overall Course Score on
Perceived Benefit from Lecture Showing ATI 107
6. Regression of Achievement on Perceived
Benefit from Lecture Showing ATI ...... 108
7. Regression of PMP First Case Overall
Competence, Efficiency and Proficiency on
Learning Style Type Showing A T I 10 9
8. Regression of PMP Efficiency Second Case
on Verbal Ability Showing ATI ............. 115
9. Regression of Overall Course Score on
Learning Style T y p e ........................ 116
10. Regression of PMP Overall Competence
Second Case, Proficiency and Efficiency on
Learning Style T y p e ........................ 117.
11. Regression of Study Time for Every Class
Session on Perceived Benefit from
Lecture Showing A T I ;i2 3
X
Figure Page
12. Regression of Study Time for Every Class
Session on Individual Learning Style
(AE-RO) and Type Showing A T I ............... 124
13. Regression of Study Time for Exams on
Individual Learning Style (AG-RO) and Type
Showing A T I ........................ 126
14. Regression of Attitude Toward Instruction,
Benefit and Enjoyment on Perceived Benefit
from Lecture, Showing A T I .................. 133
15. Regression of Study Time for Every Class
and Exams on Educational Level ............. 136
16. Regression of Attitude Toward Instruction-
Benefit on Educational Level . ................ 13 8
17. Summary of Trends for All Outcomes as a
Function of Learning Type, Educational
Level, and Treatment......................... 186
xi
LIST OF ABBREVIATIONS
ATI Aptitude-Treatment Interaction
PMP Patient Management Problem
PMP I PMP first case
PMP IP PMP first case Proficiency
PMP IE PMP first case Efficiency
PMP IOC PMP first case Overall Competence
pmp 11: PMP second case
PMPIIP PMP second case Proficiency
PMPIIE PMP second case Efficiency
PMP I IOC PMP second case Overall Competence
GPA Grade Point Average
0. T» Occupational Therapy
LSI. Learning Style Inventory
1
CHAPTER I
THE PROBLEM
Introduction
The present study deals with the effects of
teaching methods, lecture and group-discussion, and indi
vidual learning style on occupational therapy students'
achievement, problem-solving, amount of study time outside
class, and attitude toward instruction. The appropriate
approach for the investigation of this problem is the
Aptitude-Treatment Interaction model which focuses mainly
on the interaction Effect between instructional methods
and learner characteristics; it was utilized in this
study conceptually and methodologically.
Chapter I consists of the following sections:
background of the problem, statement of the problem and
hypotheses, purpose and importance of the study, assump
tions, definitions of terms, and delimitations.
Background of the Problem
Research in occupational therapy is still very
limited, but its importance is acknowledged and emphasized
strongly today (Yerxa, 1978). In general, the area of
research, in occupational therapy can be divided into two;
1) clinical studies dealing with clients in various
environments, and 2) studies dealing with education for
the occupational therapy profession. In both situations
research is urgently needed. The topic of the present
research falls in the second category of studies in
education.
Problem-solving skills, as applied to treatment
of clients in various settings and life situations, are
emphasized as one of the most important outcomes or
objectives of education in occupational therapy (Hopkins
& Tiffany, 19 78). The ultimate goal of any education
program for health professionals is to prepare its
students to become skilled in treating clients in real-
life situations. There is usually a wide gap between
classroom learning and performance in the field. Evalua
tion of the effects of education for health professions
is based on two types of evidence: direct and indirect.
Direct evidence comes from observations and records of
performance in actual practice. Indirect evidence comes
from simulation of life situations, like the "Patient
Management Problem" (PMP) in medical education (Barro,
1973; McGuire, Solomon, & Bashook, 19 76). This latter
approach to testing attempts to close the gap between
testing knowledge in traditional ways and observations
of clinical performance.
3
The technique of written simulation (PMP) is
widely used in medical education and it comprises a . :
section on the National Board of Medical Examiners' tests
(Hubbard, 1978). In other health professions its use
is still limited. First attempts to construct PMPs
and use them in testing are reported in nursing education
(De Tornyay, 1968; Gover, 19 72; McIntyre, McDonal, Bailey,
and Claus, 1972; Me Laughlin, 1979, 1980; and Page, 1978).
In occupational therapy case simulations were first
published by Briggs, Duncombe, Howe, and Schwartzberg,
(1979) but only for teaching purposes and not as a testing
device.
Theories and models of instruction concern
themselves with three major variables; (1) teaching
methods of instruction, (2) learner characteristics,
and (3) learning outcomes (Bloom, 1975); or in other
terminology: (1) events of learning, (2) conditions
of learning (internal and external), and (3) learning
outcomes (capabilities) (Gagne, 1977) .
Learner characteristics interact with instruction
to accomplish a particular outcome (Bloom, 1976; Joyce,
1978). This crucial postulate is also the premise of
the Aptitude-Treatment Interaction between personal and
environmental variables which will result in a desired
outcome. "'Treatment' is seen as any manipulable vari
able of instruction, while 'aptitude' is defined as any
4
characteristics of a person that forecasts his probability
of success under a given treatment" (Cronbach & Snow,
1977, p. 6). These broad definitions enable the investi
gation of a wide spectrum of individual variables, and
instructional components.
As mentioned earlier, if the most important
learning outcome in health professions education is
problem-solving skill, which takes the form of managing
patients1 problems, then the interaction between instruc
tion and individual learner characteristics should be
studied in order to determine what is the best way to
achieve this goal.
Two teaching methods, lecture and group-discussion,
were chosen for study because of their frequent use in
occupational therapy education, as well as in other health,
professions and in higher education in general. Findings
of studies comparing lecture with group discussion are
generally confusing and contradictory. Research on the
subject started as early as 1925 and continued with a
very similar approach through the 1960s. In a frequently
cited review written by Dubin & Taveggia (1968) in which
the authors reanalyzed 91 studies, among them 45 on
lecture vs. group discussion, the overall conclusion was
that there is "no significant difference" among various
methods of college teaching as measured by final exams.
McKeachie, in his continuing reviews over the years
5
(1954, 1963, 1970, 1976), agrees that an overall generali
zation was apparent: there is usually no difference
between methods when knowledge on a final exam was
measured. But, group discussion was found in many
studies to be more effective than lecture in developing
concepts, problem-solving, attitude change, and moti
vation for learning (Berliner & Gage, 1976, Gall & Gall,
1976; McKeachie, 1978). Similar results were found
in studies performed in medical, dental, and nursing
education. One of the first, and perhaps one of the
most classic studies is the Zimmerman-and-King (1963)
study, ''Evaluation of Student Centered Group." They
found group-discussion significantly superior to lecture
on measures of achievement, critical thought, and
problem solving ability, but only equally effective
on measures of knowledge.
The conclusions drawn from the findings may
point out the importance of matching the teaching
method to the course objective of learning outcome.
While this conclusion seems reasonable, many of the
research findings are inconsistent, perhaps reflecting
the lack of methodological rigor (Gall & Gall, 1976) ,
and poor criterion tools (McKeachie, 1963). In addition,
one other important variable which probably contributes
to the lack of consistent results is that of the
6
individuals who participated as subjects in the studies.
In the conventional comparative research, learner
characteristics were rarely considered, the investigators
assuming equal effects of treatment for every person.
As Abrahamson 119 76) points out,
What is hidden in this experimental design, however,
is the fact that there is always a significant
overlap of socres, some of which may be directly
atttibutable to the fact that certain teaching
methods are particularly helpful to certain indi
vidual students while other students are helped
by other techniques (p. 1025).
In order to correct this flaw in the theoretical base,
as well as the methodological approach, the notion
of Aptitude-Treatment Interaction (ATI) was developed.
This, appraoch is presumed to provide more appropriate
answers to research questions dealing with instructional
effectiveness (Cronbach, 1975; Cronbach & Snow, 1977).
The basic equation presented by Kurt Lewin (cited in
Hunt, 1973), Behavior=Person x Environment, is defined as
Learning Outcome-Aptitude x Treatment. The focus in
both definitions is on the interaction between the
person's internal variables and outside variables.
There are those like Hunt (.1975) who suggest that the
ATI approach limits the broader Lewinian view by using
the term aptitude. Nevertheless, the ATI approach
succeeded in focusing attention on individual differences
among learners, and on the Important interactions of
7
those characteristics with instructional methods.
Thus, the third major variable in any investigation
of the teaching/learning process should lead to the
learner himself.
Individual differences in learning are a long
acknowledged fact but a vary difficult one to investi
gate (Gagne, 196 7). Individual characteristics were
studied in various contexts f like verbal learning,
problem-solving, motor learning, etc. The major
question remaining is how to classify these differences.
The differences most frequently studied are those of
mental ability, creativity, cognitive styles and per
sonality types.
Recently f more emphasis has: been given to the
concept of "learning style" which seems to encompass
cognitive and perceptual, as well as personalityf
factors (Kolb, 1974, 1979; Llorens & Adams, 1978) .
Individual learning style appears to be a relatively
stable trait which may affect the person's problem
solving skills, his preference for teaching methods,
and his career choice (Kolb, 29 76, Plovnick, 19 75).
Kolb developed a model of experiential learning that
". . .conceptualize the learning process in such a
way that differences in individual learning styles
and corresponding learning environments can be identi
fied" (1974, p. 2). The model is conceived as a
8
four-stage cycle along two dimensions ("abstract-concrete"
and "active-reflective") which seem to correspond also
to stages in a problem-solving sequence (Kolb 1974, 1976).
(See Figure 3), .
Based on the theoretical model, a "Learning Style
Inventory" (LSI) was devised by Kolb, Rubin, & McIntyre,
(1976)Jwhich enables the measurement of individual styles
along the two dimensions, and produces two composite
scores classifing four learning types. The concept
of "learning style" as postulated by Kolb relates to
the problem-solving process, and as such to the learning
outcome emphasized in education for health professions.
On the other hand, as a learner characteristic it is
assumed to interact with the teaching method used in
the education process. The specific dimensions on which,
the LSI is based appear to correspond to the demands
of the two methods, lecture and group discussion, in
terms of the amount of activity and/or reflection required
from the learner. At the same time, the dimension of
abstract vs. concrete can be manipulated in both methods,
thus providing varied combinations of method and learning
style.
This correspondence between the dimensions of
the teaching methods (outlined in Chapter II)f and the
active/reflective dimension of the learning style led
9
to the matching hypothesis of this study using the
capitalization model of the ATI approach. This model
assumes that the treatment builds on the student's
strengths, and instructional components are in accordance
with students' capabilities, strategies or styles
(Cronbach & Snow, 19 77; Salomon, 19 71).
According to the ATI approach followed in this
study emphasis is given to the measurement of various
aptitude variables in order to understand the possible
interaction between learner characteristics and instruc
tional components. These variables may all be of inter
est, or, may be used as coveriates beyond which the
variable of interest can be investigated.
In this study, four such variables were Identi
fied and taken in account. ’ (.1) Verbal ability, as noted
by Matarazzo (.19 72) 11. . . there is a correlation of
approximately .50. between measured intellignece and
performance in school." (p. 285) Furthermoref in many
ATI studies the G factor seem to explain and interact
with the treatment more than any other variable (Snow,
1976br Snow, Federico, & Montague, 19 80). (2) Prior
achievement seems consistently to show significant :
nteraction (Tobias, 1976), when defined as prior familiar
ity with the subject matter. In the present study this
variable refers to level of achievement in the occupa
tional therapy curriculum as defined by GPA and is
10
assumed to have main effect more than interaction effect.
(3) Perceived benefit from lecture, or (4) Perceived bene
fit from small group, are assumed to interact with
actual benefit. Salomon (1981) elaborates on the notion
of "perceived demand characteristics" (PDC) postulating
that one's perception of the situation or task require
ments may influence the "amount of invested mental
effort" (AIME) and in turn influence acquisition of
information, and achievement.
Another intervening variable in these relation
ships may be indicated by the negative correlations
between enjoyment and achievement found in some recent
studies and reviewed by Clark (1981) , or in the findings
of Strom and Hocevar (1981) relating attitude to course
grade. Thus, for exploratory reasons, two additional
outcomes were included in the study, "amount of study
time outside class" and "attitude toward instruction"
as reported by students.
Statement of the Problem and Hypotheses
The main problem which this study investigated
may be stated as follows; Does an interaction exist
between Individual Learning Style (as a learner charac
teristic) , and different teaching methods ( . . lecture- or
group-discussion as the treatment manipulation), when
11
outcome variables like "achievement," "problem-solving
skills," "amount of study time outside class," or
"attitude toward instruction" are considered?
Orginally, four specific hypotheses relating
to problem-solving skills and achievement were formulated.
The first and main one relates to the interaction.
1. Students whose learning style "matches"
the teaching method will score higher on measures of
problem-solving and achievement than students whose
learning style and teaching method are "mismatched"
(see design in Chapter III).
2. Students whose learning style "matches" the
group-discussion method will score the highest on measures
of problem-solving.
3. Students participating in group-discussion
will score higher on measures of problem-solving than
lecture students. No significant difference will be
found in achievement (as measured on a multiple-choice-
question exam for knowledge and comprehension).
4. Students whose learning style emphasizes
the active dimension over the reflective (AE-RO), will
score higher on measures of problem-solving. No signi
ficant difference will be found in achievement (knowledge
and comprehension).
12
In relation to the two additional outcomes, study
time and attitude, only questions were raised in terms
of exploring the effects of learning style and instruction
on variables which are assumed to influence cognitive
outcomes.
Purpose and Importance of the Study
The main purpose of this study was the investi
gation of instructional effectiveness when learner charac
teristics are taken in account, and essential learning
outcomes such as "achievement," "problem-solving," "study
time," and "attitude toward instruction" were assessed.
The study followed the ATI model which enables the inves
tigation of the interactive effects of learner variables,
instructional components, and learning outcomes.
Investigation of instructional effectiveness in
occupational therapy education has been rarely done (or
published), and experimental studies are almost non
existent. The study which was primarily designed for
theory testing and not necessarily for immediate practical
application, suggests some preliminary interrelations
among the various components or variables investigated,
and may add information to a "theory of instruction"
which is, in a developmental state (Snow, 1980). As an
ATI study, its main purpose is testing the interaction
hypotheses and adding to the understanding of instructional
13
effectiveness. As such, supporting the matching hypothe
ses between individual learning style and the teaching
method components has important implications for theory,
as well as possible adaptation between these two variables
in planning instruction in occupational therapy.
Another important aspect of this study refers
to the problem-solving outcome and its measurement through
the Patient Management Problem (PMP) technique. The
use of this technique in occupational therapy is new
(Briggs et al., 1979), thus development of PMPs and
testing out their actual qualities is an important con
tribution of this study, but much more needs to be done
in this area.
The present study combines a powerful experimental
design in a real educational situation, including essen
tial variables for the understanding of the teaching/
learning process in occupational therapy, and in general.
Assumptions
The major conceptual assumptions underlying this
study are related to (1) the notion of individual differ
ences between learners and their interaction with instruc
tional methpds.f and to ( . 2 ) . problem-solving skill snd its
measurement with the PMP technique. Individuals differ
in many ways. This basic assertion can hardly be refuted.
It is assumed that these differences among people cause
14
differential interactions between individual learners
and instructional methods, leading to varying degrees
of learning outcomes. Cronbach and Snow (1977) state that
"Aptitude x Treatment interaction exists. To assert
the opposite is to assert that whichever educational
procedure is best for johnny is best for everyone else
in Johnny1 s school1 1 (p. 462).. The method of instruction
is assumed to have different impact on various students,
and students may benefit from different teaching methods
(Abrahamson, 1976) .
Learning style as an individual's characteristic
is assumed to be a more-or-less stable trait which
relates to the individual's perceived benefit from a
learning situationf and to the actual ability to benefit
from certain teaching methods (Kolb, 1976; Wolf & Kolbf
1979). . Moreover, the "Learning Style Inventory" is
assumed to measure adequately this trait, thus enabling
the investigation of its interaction with instructional
manipulation in education.
The process underlying performance of an occupa
tional therapist, a physician or any other health
professional, is assumed to reflect the ability of this
person to inquire, judge, and make decisions, all of which
are steps in the problem-solving process required in
a clinical situation (Hopkins & Tiffany, 1978; Ways,
Loftus, & Jones, 1973). Any encounter between a
therapist and a client or patient involves a problem- 15
solving process (Andrew, 1972; Elstein, Shylman & Sprafka,
1978). Thus, problem-solving skill which takes the form
of managing patients' problems is assumed to be of crucial
importantance and is the main goal of education for health
professions.
The concept of simulation consists of 11. . . placing
an individual in a realistic setting where he/she is con
fronted by a problematic situation that requires a sequence
of inquiries, decisions and actions" (McGuire et al.,
1976, p. 1). This situation is assumed to be similar to
the one which a person will face in actual practice.
Therefore, it is postulated that a student who does well
in successfully solving a written simulation exercise,
will also do well in solving actual problems in a
clinical situation. The focus in this process is on the
cognitive domain or on cognitive aspects of the therapist's
problem-solving ability, rather than on communication and
interpersonal relations (Elstein et al., 1978). A positive
correlation is assumed between performance on the PMP
exercise and actual performance on the job.
The following is a summary of the main assumptions:
1. Individual differences exist, and therefore
individual learners interact differentially with instruc
tional methods. This assumption underlies the basic
Aptitude-Treatment-Interaction conceptualization.
16;
2. Learning style is a more-or-less stable trait
which relates to individual's perceived and actual bene
fit from a teaching method. Furthermore, the LSI is
assumed to measure this trait accurately.
3. The process of problem-solving underlies
the health professional's function in clinical situation,
and as such is assumed to affect it considerably. The
PMP as a simulation technique is assumed to depict
this process and a positive correlation is assumed to
exist between both.
Definitions of Terms
Aptitude-Treatment Interaction refers to any
learner characteristic (aptitude) which is assumed to
interact with Instructional components (treatment) in
affecting learning outcomes. This term implies a certain
methodological approach and an emphasis on interaction
effects as opposed to main effects of either aptitude
or teaching methods ( Cronbach, 19 75; Cronbach & Snow,
1977; Snow, 1976b, 1980.) .
Teaching methods or instructional methods.
According to Gagne and Briggs (1979), "Instruction is a
set of events which affect learners in such a way that
learning is facilitated." (p. 3) They conceive 'instruc
tion' as ". . .all the events which may have a direct
effect on the learning of a human being. . 'teaching'
17
according to this view is only one form of instruction.
(p. 3) Nevertheless, in most texts, instructional
methods and teaching methods are used interchangeably,
and this approach is adopted also in this study.
Lecture is defined as "A more or less continuous
oral presentation of information and ideas by the teacher,
with little or no actiye participation by the members of
the class." (.Stoval, 1958, p. 255)
Group-discussion. According to Holcombe and
Garner (1973) , "A discussion involves the verbal inter
action of a number of individuals who perceive one another
as participants in a common activity." (p. 48) Or in
other words,
. . . purposeful, systematic, oral exchange of ideas,
facts., and opinions by a group of persons who share
in the group’s: leadership. This definition of the
discussion method postulates five attributes: the
discussion method occurs when (1) a group of persons,
usually in the roles of moderator-leader and parti
cipant, ( . 2) assembles at a designated time and place,
(3) to communicate interactively, (4) using speaking,
nonverbal, and listening processes, (5) in order to
achieve instructional objectives. (Gall & Gall,
1976, pp. 168-9).
Student-centered group is an approach to groups
discussion which is usually contrasted with the instructor-
centered method. Its goal is to ". . . encourage greater
student participation and responsibility." (McKeachie,
1978, p. 52) In studies comparing both methods, the
general conclusion is that when the learning objective
is problem-solving, student-centered approach is more
18
beneficial. But, as McKeachie (19 78) indicates it may
have . . some serious weaknesses, at least in achiev
ing lower level cognitive goals," (p. 52) Therefore,
the method used in this study, while based on the student-
centered approach, had some control and input from the
instructor to insure "coverage" and understanding of the
subject matter. (See the treatment manipulation descrip
tion in the data-gathering section, Chapter III).
Individual Learning Style. According to Kolb
(.1974, 1976, 198JL) learning style encompasses cognitive
and perceptual, as well as personality, factors. Most
people develop a learning style that emphasizes some
learning abilities oyer others. Kolb believes that
each individual resolves conflicts in a characteristic
way, and each, person develops a learning style that has
some strong points and some weak points. Consequently,
each person may benefit from a particular learning
environment as conceptualized by .Randolph and Posner
(19 79) in their grid for matching an individual's learn
ing style with learning situations.
Learning Style Inyentory (.LSI) is a self-
descriptive forced-choice questionnaire developed by Kolb
(1976) to measure an individual’s relative strengths
and weaknesses as a learner. The inventory is based qn
an experiential learning model (.see Chapter II) which is
19
conceived as a four-stage cycle along two dimensions:
(1) the degree to which a person emphasizes abstract-
conceptualization vs. concrete-experience in choosing
and ranking words that characterized his/her learning
style, and (21 the degree to which a person emphasizes
active-experimentation vs. reflective-observation. Six
scores are obtained from the LSI, one for each 'pole1
and two 'composite' scores. In addition, two adjacent
pole scores or the two composite scores are used to place
an individual on the learning-styl'e-type' grid, thus indi
cating one of four learning types (accomodator, diverger,
assimilator, and converger).
The composite score on the active-reflective
dimension (AE-RO). and the learning type were employed
in this study as the operational definition for the
individual learning style variable.
Problem-solving skill. According to Gagne (.19 77)
". . . problem-solving depends upon previously learned
rules. It also depends upon a type of intellectual
skill governing the individual's own thinking processes;
a cognitive strategy." (p. 156) Problem-solving is
defined by Gagne as a ". . . process by which the learner
discovers a combination of previously learned rules
which, can be applied to achieve a solution for a novel
situation." Cp. 155) Similarly, Dewey (1938) suggested
20.1
that problem-solving consists of three components;
problem-sensing, problem-definition and formulation
and problem-resolution.
In the health, professions, problem-solving takes
the form of a clinical management process and "... can
be thought of in terms of six independent.and sequential
components: problem-sensing, problem-hypothesizing,
problem-searching and definition, problem-identification,
problem-resolution, and problem-verification.1 1 (Andrew,
1972, p. 953) In more practical terms the process includes
the following steps: (1) sensing that a problem exists
via patient presentation, ( . 2) . gathering data from
patient and relatives, as well as laboratory tests, etc,
(3). defining and formulating the problem, (.4) deciding
upon a course of action based on initial hypothesis,
(5) observing the consequences, formulating, and revising
the treatment (Andres, 19 72; Ways/ 19 73).
The focus in education for the health professions
and also in this study is on a problem-solving process
similar to the one outlined above and performed by an
individual rather than by a small group. Neverthelessf
learning through small group-discussion is hypothesized
to enhance the acquisition of this skill (skill means
the ability to solve problems successfully).
21
Patient Management Problem (PMP) , . The opera
tional definition of problem-solving skill in this study
is scores on two PMPs devised for occupational therapy
students calculated according to the McGuire et al. ,
(19 76), scoring system. The scores include proficiency,
efficiency, and overall competence. A PMP is a written
simulation exercise which depicts a clinical situation
and requires the examinee to work through the various
steps of the test similar to a real-life situation
(McGuire et al., 1976). Two basic types of PMP exist
in medical education, a "linear" format like the one
used by the National Board of Medical Examiners, and
a "branching" format developed by McGuire at the University
of Illinois. In the linear format, each examinee follows
the same sequence, while in the branching format each
one may follow his/her own sequence according to decisions
in each step of the exercise. In this study a version
of the branching format was used (McGuire et al., 1976,
Model 5: The single-major complication model, sche
matically represented on p. 280) tsee Appendix C for
diagrams of the PMPs).
Achievement was defined operationally as a score
on a final exam of the multiple-choice type, (140 items)
developed to measure knowledge and comprehension of the
content area taught during the instructional period of
the study. Thus, achievement is conceived of as mastery
of the subject matter in terms of demonstrating the
ability to recall and understand the materials.
Amount of study time outside class was divided
into two variables: (1) amount of study time for every
class session, and (2) amount of study time for exams.
These variables were operationally defined as a score
on a 5 point Likert scale type questionnaire. Subjects
had to check an option on the scale, where options were
given in terms of time in hours (See Appendix D, items
one and two).
Attitude toward instruction was operationally
defined as a combined score on a 5 point Likert scale,
including items related to benefit, enjoyment, and
confidence, (see Appendix D, ’ .items 3, 4, & 5) which
were also analyzed separately. The attitude measure
employed refers to what Strom and Hocevar (19 81) call
"transitory course-related variable" based on self-report
through, course evaluation, "It may be that this type of
variable also moderates the effects of course structure
on student achievement because of its potential influ
ence on a student's motivation." (p. 2)
Overall course score was operationally defined
as a combined z score calculated from achievement score
and the two scores for overall competence on the PMPs.
23
Cognitive outcomes refers to the achievement
and problem-solving variables when considered together.
Delimitations
The scope of this study was delimited by the
investigator in several ways,
-1. The study was delimited to one_occupational
therapy curriculum at the University of Southern
California.
2. The subjects were all the students enrolled
in the academic year 19 80-19 81, consisting of two classes;
seniors and first-year entry level graduate students.
All of them were females except for one.
3. All subjects signed an informed consent
form in order to participate in the study.
4. Only two teaching methods, lecture and group-
discussion, were employed in the study.
5. individual differences were investigated
with heavy emphasis on Learning Style as defined by
Kolb and measured with the LSI.. Other aptitude measures
were used mostly as covariates.
6. Problem-solving skill was measured only with
the written simulation technique prepared as a Patient
Management Problem especially for this study.
24;
Hence, care s h o u ld be taken in generalizing the results
of this study to other populations or settings f to
other teaching methods, and to other aspects of indivi
dual differences. In addition, the fact that subjects
volunteered to participate and knew that they were
involved in a study needs to be taken in' account when
generalizing results.
25
CHAPTER II
LITERATURE REVIEW
Introduction
The literature review presented in this chapter
consists of two main parts, including subsections. The
first part is comprised of two sections: (1) a general
conceptualization of the teaching/learning process, fol
lowed by (2) the Aptitude-Treatment Interaction approach.
The second part consists of three main sections pertaining
to the specific variables chosen to be investigated in
this study: (1) individual learning style, (2) teaching
methods: lecture versus group discussion, and (3) problem
solving in the health professions. In addition, a brief
discussion of achievement, attitude toward instruction and
amount of study time outside class is presented and at
the end a short summary is provided.
The Teaching/Learning Process
Theories and models of instruction, in contrast to
learning theories, concern themselves with aspects of
teaching or instruction, learner characteristics, and the
learning outcome. Bloom (19 76) proposes a theory of school
26
learning, presented as a model of the three major vari
ables contributing to the process (Figure 3. ) .
Student
Characteristics
Cognitive entry
behavior
Affective entry
characteristics
Instruction
Learning
Task(s)
-----
Quality of
instruction
Learning
Outcomes
Level and
^ type of
achievement
Rate of
learning
Affective
outcomes
Figure 1.
Major Variables in the Theory of School Learning
Note. From Human Characteristics and School Learning
by B. S. Bloom. McGraw-Hill, 1976.
According to this model, students' characteristics
are supposed to interact with the instructional components,
and this interaction will cause a particular outcome.
Bloom emphasized the task itself and the quality of
instruction as important factors within the instruction
component. Joyce (1978) postulates three theses concerning
models of teaching. First, many alternate approaches to
teaching exist. Second, teaching methods make a differ
ence in what is learned. Third, students are a powerful
part of the learning environment, and they react differ
ently to various teaching methods. His approach highlights
27
the flexibility and adaptability of teaching methods to
the learner's style, while at the same time pointing out
that students can learn to increase their abilities to
profit from most methods. "Putting the learner into the
equation" in investigating instruction is more important
(Joyce, 1978, p. 13).
A more elaborated theory of instruction is provided
by Gagne (1977), Briggs et al. (1979) and Gagne and Briggs
(19 79) in a continuous effort to explain the process of
learning. Gagne presents a schematic model (Figure 2).
EVENTS OF LEARNING
CONDITIONS
OF
LEARNING
LEARNING
OUTCOMES
(CAPABILITIES)
Figure 2.
The Relations Between Events of Learning, Outcomes
of Learning, and Conditions of Learning
Note. From The Conditions of Learning by R. M. Gagne.
Holt, Rinehart and Winston, 1977.
28
Events of learning are the actual components inclu
ded in an instructional program. In order to plan these
events, the necessary conditions for learning are divided
into (1) conditions within the learner— internal charac
teristics or previously acquired skills, and (2) conditions
in the learning situation— external factors in the environ
ment leading to a particular outcome. For each of the
outcomes, or categories of learning capabilities, Gagne
analyzed the necessary internal and external conditions
and related them to the events of learning. Thus, planning
of instruction should be a function of the learning out
come, and the optimal match between the learner charac
teristics with the external conditions necessary to reach
the particular outcome. As Gagne (1977) states:
In the most general sense, instruction is intended
to promote learning. This means that the external
situation needs to be arranged to activate, to
support, and to maintain the internal processing
that constitutes each learning event. (p. 24)
According to Gagne (19 77), in the hierarchy of the
intellectual-skills capabilities (learning outcomes),
problem-solving is placed at the top of the ladder:
". . . problem solving depends upon previously learned
rules. It also depends upon a type of intellectual skill
governing the individual's own thinking processes: a cog
nitive strategy"(p. 156). Problem-solving is defined by
Gagne as 11. . .a process by which the learner discovers a
29
combination of previously learned rules which can be
applied to achieve a solution for a novel situation" (p.
155). It is conceived as the most complex learning out
come, as the final step in a hierarchical sequence of
learning intellectual skills. In order to achieve this
skill, the optimal external condition is "discovery learning"
in which "the learner is confronted with an actual, or a
represented, problem situation not previously encountered"
(Gagrie & Briggs, 1979 , p. 71) . This concept is similar to
Bruner's "discovery approach."
Bruner (1966) differentiates between a theory of
instruction which is prescriptive and normative in nature,
to a theory of learning and development which is descrip
tive. But, all three aspects, development, learning and
instruction are related and have to be-congruent in order'to
reach a particular outcome. "A theory of instruction seeks
to take account of the fact that a curriculum reflects not
only the nature of knowledge itself but also the nature of
the knower and of the knowledge process" (Bruner, 1966,
p. 72). Thus, the same emphasis on the interaction between
instruction (subject matter and process), and the indivi
dual learner is presented. "The fact of individual differ
ences argues for pluralism and for an enlightened opportun
ism in the materials and methods of instruction" (p. 71).
30
A theory of instruction according to Bruner (1966)
has four major features: predisposition to learn, struc
ture and the form of knowledge, sequence of materials, and
finally, form and pacing of reinforcement. Learning is
conceived as a problem-solving process which depends upon
the exploration of alternatives. This process is affected
by the four features of instruction, and therefore in plan
ning the events of learning all four aspects have to be
taken into account.
The teaching/learning process as conceptualized in
the models presented or seen as a theory of instruction
have basically similar features which can be integrated
under the framework and methodological approach of Aptitude
Treatment Interaction between learner characteristics—
his/her predisposition to learn— and internal conditions,
all of which refer to individual differences (aptitudes),
and instruction— teaching methods, events of learning,
structure and sequence (treatment). This interaction,
(and not the separate components of the models), is seen
as the main factor in determining a particular learning
outcome. Efforts toward a more comprehensive understanding
of all the variables involved in the teaching/learning
process, which would comprise a theory of instruction, are
presented in Snow et al. (Eds.) (19 80) Aptitude, Learning,
and Instruction. According to Glaser (1980):
31 -
Putting all three of the words in the title . . .
together . . . brings us to the notion of adaptive
instruction, [which means] . . . that the actions
taken in an instructional setting . . . vary as
a function of past and present information about
a student. (p. 322)
The same emphasis on adaptive instruction is given by
Shuell (1980) who discusses the various ways matches
between the learner and the environment can be made based
on Salomon's (1971) three heuristic models. These models
are: the remedial approach, the compensatory model, and
the preferential or capitalization model. When using the
capitalization model the match builds on the strengths of
the learner, similarly defined by Glaser's (19 77) model 3
which essentially accommodates to different individual
styles and abilities. Atkinson (19 72) indicates that
manipulation of instruction variables has minimal effects
compared " . . .to the impact that is possible when
instruction is adaptive to the individual learner" (p. 9 29).
But, Atkinson adds a fourth criteria for a theory of
instruction which deals with costs and payoffs of instruc
tional actions. This component of a theory of instruction
as well as various other practical ones are included by
Glaser (1980) under a "macrotheory of teaching and
instruction."
A comprehensive theory of instruction is a very
complex one; as Snow (19 77) indicates, ATI findings make
generalization difficult and therefore he suggests "that
32 1
instructional theories need at least initially to be local
theories by subject-matter as well as local by locale"
(p. 11). Opposing this view, Shuell (1980) argues that
1 1 . . . an integrated theory is both desirable and possi
ble" (p. 298), and Gehlbach (1979) emphasizes the differ
ence between applied research which is prescriptive and
may lead to local theories, in contrast to basic research
leading to a general theory and underlying relationships.
He sees the possibility and necessity to work on both
levels toward a theory of teaching and instruction. Simi
larly, Glaser (19 80) concludes that the development of
adaptive instructional systems should rely on field
research, as well as "research and theory construction on
individual differences, learning, and cognitive performance
as these relate to the acquisition of complex knowledge and
skill" (p. 324).
Aptitude Treatment Interaction (ATI)
"The search for interactions is an idea whose time
has come" (Cronbach & Snow, 1977, p. 491). The main pre
mise of the ATI approach is to correlate and integrate two
schools of research: the study of the external conditions
versus individual differences (Cronbach, 1975). The main
goal is finding interactions between person variables and
environmental variables (Snow, 19 76a). "The substantive
problem before us is to learn which characteristics of the
33
person interact dependably with which features of instruc
tional methods" (Cronbach & Snow, 1977, p. 493).
These quotations from two researchers, who are the
leading authorities in this line of investigation, illus
trates the underlying theoretical approach to their metho
dology. They define 'aptitude' as "any characteristics of
a person that forecast his probability of success under a
given treatment, while 'treatment' is seen as any manipu-
lable variable in instruction" (p. 6). This definition
leaves broad limits to investigate multitraits and multi-
methods within the ATI frame of reference. Cronbach and
Snow (19 77) provide a comprehensive review of research
in the area of ATI which illustrates the variety of apti
tudes and treatments chosen for investigation. There are
those who criticize ATI and say that its main focus is only
on research methodology: it becomes a specific area of
inquiry and therefore limits the broader view of person-
environment interaction (Hunt, 1975). On the other hand,
supporters emphasize the elaborate research methodology and
its statistical analysis which provide means for "...
gradually constructing a matrix of learning situations and
learner characteristics which may facilitate the develop
ment of a theory of instruction" (Salomon, 1971, p. 328).
Salomon suggests three heuristic models of ATI: the reme
dial approach, the compensatory model and the preferential
34
model. For each model, an outline of the treatment's
functions, the aptitudes measured, and predictions are
presented (Salomon, 1971, p. 340).
The general approach of ATI research proceeds from
empirical studies, hypothesis formation, choice of vari
ables and statistical analysis toward possible generaliza
tions which may converge into a theory of instruction. As
Cronbach (1975) mentioned "... the research methods in
common use are inefficient and produce misleading results"
(p. 509). They do not take in account individual differen
ces, and they focus on main effects of teaching methods
rather than on interaction effects. Explaining the many
contradicting and inconsistent results summarized by Dubin
and Taveggia (1968) , Cronbach and Snow (1977) indicate that
1 1 . . . the investigator should try to specify the learner
skills or styles important to each treatment, the processes
each treatment calls for, and the prerequisite information
each calls upon" (p. 171).
Stated briefly, the typical ATI study is a two-
group experiment with two independent variables (treatment
and aptitude) and one dependent variable. The treatment
variable is manipulated while the aptitude variable is
measured before the instruction starts. At the end of the
instructional period an outcome measure is obtained which
is the dependent variable. This outcome score is regressed,
35
in the statistical analysis, on the aptitude score for each
treatment group, and both regression lines are compared for
evidence of aptitude-treatment interaction. The preferred
design is a factorial design: in the simple case described,
a 2x2 design with random assignment of subjects or random
ized block design (blocking on the aptitude first and then
randomly assigning to treatments). A more complex ATI
study consists of combining more aptitudes with more para
meters of the instructional situation, thus constructing
different levels of factorial designs. In ATI research,
the hypothesis of interest is that of the interaction
effect and only secondarily the main effects of treatment
and aptitude. In every study, each of the three hypotheses
or any combination of them may be found significant or
non-significant.
Studies which utilize the ATI approach compare
variables of the treatment, while taking in account indi
vidual differences and their interaction on the outcome
measures. For example, Domino (1971) found interaction
between student achievement orientation and teaching style.
Those who were taught in a manner consonant with their
achievement orientation (capitalization model)Lhad signifi
cantly higher scores, while no difference in main effects
of teaching style were found. Similarly, Dowaliby and
Schumer (19 73) found a disordinal interaction between
36
teacher-vs-student-centered instruction and manifest
anxiety. Students with high anxiety did better in the
teacher-centered mode, while low anxiety students did bet
ter in the student-centered mode. The Johnson and Neyman
technique for region of significance was used, thus allow
ing more accurate conclusions. In an analysis of main
effects, no difference between the groups was found.
These two examples seem to indicate that:
Perhaps there are no "main effects" to be found
and significant treatment effects will be evidenced
only when successful attempts are made to account
for individual differences rather than simply dis
missing them as error. (Dowaliby & Schumer, 19 73,
p. 130)
The authors even suggest that perhaps the studies which
have indicated no difference in the past, may be replicated
as ATI studies. Following Cronbach and Snow suggestions,
Porteus (1976) integrated the two studies and added addi
tional aptitudes to examine multiple effects of student
variables with the same teaching methods. Domino (1974)
replicated Dowaliby's study and found the same results.
While some of the studies following ATI approach
show interaction effects, others do not. Cronbach and Snow
(1977) summarize hundreds of studies in their book Apti
tudes and Instructional Methods and suggest implications
for research strategy:
37
We could recast what has been said above as follows:
The premise of ATI research to date is that general
izations about main effects are too simple; one can
account for more variance by identifying relevant
first-order ATI. The inconsistency among first-order
ATI persuades us that higher-order ATI, combining
aptitudes with many parameters of the instructional
situation, are operative. (p. 391)
In addition, the authors point out that:
Some instructional treatments seemed to be beneficial
to one subgroup while at the same time having negative
effects for another subgroup. This dark side of
instructional psychology has long been neglected,
probably because main effects are so rarely negative,
but it is forced upon us by ATI. (p. 391)
Based on these and other implications the authors suggest
that educational research should move to apply and observe
alternative kinds of instruction in naturalistic settings
for various students.
In a series of more recent studies, Peterson (1977,
1979) found statistically significant ATI between teaching
approaches and students' ability, anxiety, and conformity.
In addition, Peterson, Janicki, and Swing (1980) report
findings of two studies: (1) a naturally occurring ATI,
and (2) a short-term experimental study, both utilizing
the same variables. The results indicate a partial repli
cation of ATI scores across the two studies. This outcome
is emphasized by the authors as evidence that ATI can be
replicated, and generalization may be possible.
. . . of ATI from a situation where treatments are
experimentally manipulated to a situation where
treatments are naturally occurring. This suggests
that ATI findings can be generalizable to an actual
school setting and that such findings may have impli-
cations for improving school practice. (p. 359)________
38
In another recent ATI study, Green (19 80) followed
the suggestion of Cronbach and Webb (1975) to "examine
between-group and within-group regression lines separately"
(p. 724) when studying intact classes. Green found that:
"... the significant interaction was not with individual
student aptitudes, but rather with the social effects of
being a member of a class that was relatively high or low
in ability or academic confidence" (p. 300). This finding
supports the Cronbach and Snow (19 77) elaboration of the
statistical analysis required in ATI research, namely when
classrooms are used between classes as well as within class
analyses have to be conducted. This suggestion becomes
almost a requirement when intact classes are used and stu
dents are not randomly assigned to classes.
In a progress report provided by Snow (19 76b) on
"research on aptitude for learning" the main conclusion is
that "G [general ability] has been the most widely studied
aptitude construct . . . and thus has produced the most
ATI" (p. 68). In a summary table relating treatment vari
ables to general ability, the general hypothesis is:
"instructional treatments differ in the information-
processing burdens they place on, or remove from, the
responsibility of the learner, and the regression slopes of
cognitive outcomes on G become steeper or shallower
accordingly" (Snow, 1976b, p. 69). This relationship
39
was found also in a more recent study by Yalow (1981) where
the treatment with minimal supplements shows the steepest
slope. Interestingly, in a recent study with college stu
dents, Strom and Hocevar (19 81) did not find the same
relationship. Although they found a significant inter
action between verbal ability and achievement, it shows
that as structure is higher more able students do better.
This is in opposition to Snow's summary, as high structure
is associated with less burden on the learner.
In another recent study Janicki and Peterson (1981)
found only main effect of ability on achievement while
attitude/locus of control factor had a significant inter
action showing a positive correlation between attitude and
achievement in a small group in contrast to direct instruc
tion. The same variables show in the Strom and Hocevar
(1981) study the opposite relationship, as high structure
method is correlated with high locus of control, and moder
ate structure with low locus of control. The contradic
tions between these findings are not easy to explain
especially because the measures used are rarely the same.
A possible reason in this case is the sample, as Janicki
and Peterson worked with elementary school students and
Strom and Hocevar investigated college students. These
among many other variations are probably the reasons for
Snow (1977) and Cronbach and Snow (1977) to suggest the
40 J
development of local theories as opposed to a general
theory of instruction.
In spite of the difficulty in generalizing across
many studies the ATI approach provides a rigorous research
methodology and elaborate statistical analysis. Following
this methodology research in education is more feasible
and .inferences can more readily be drawn. Its main contri
bution is in focusing attention on individual differences
of learners, and providing a way to take them in account
while studying instructional effectiveness.
Individual Learning Style
The notion of individual differences is a long
acknowledged and obvious fact (Gaghe, 196 7). Individual
differences were classified for example, in terms of intel
ligence and creativity, cognitive styles (Witkin, 1976;
Goodenough, 1976), or personality types (Myers-Briggs,
1970). Under these broad categories sub-classifications
and various specific traits or aptitudes were investigated.
More recently the notion of "learning style" was
introduced by different authors (Kolb, 1974, 1979; Canfield
& Lafferty, 1974). This trait seems to be closely related
to the notion of cognitive style and personality attributes.
Learning styles are defined as relatively constant attri
butes or preferences of an individual which interact with
instructional circumstances (Irby, 1979). As such,
41..
investigation of individual learning style and its impact
on the teaching/learning process may contribute to under
standing of learning outcomes.
Kolb (1974, 1979) developed a model of experiential
learning that "... conceptualized the learning process
in such a way that differences in individual learning
styles and corresponding learning environment can be iden
tified" (1974, p. 2).
According to Kolb (19 76),
As a result of our hereditary equipment, our particu
lar past experience, and the demands of our present
environment, most people develop learning styles that
emphasize some learning abilities over others.
Through socialization experience in family, school
and work we come to resolve the conflicts between
being active and reflective and between being immediate
and analytical in characteristic ways. (p. 4)
In order to measure the individual learning style, Kolb
devised a Learning Style Inventory (LSI), which'is based
on the experiential learning model. The model is conceived
as a four stage cycle along two dimensions: abstract-
concrete and active-reflective.
The LSI measures an individual's relative emphasis on
four learning abilities— Concrete Experience (CE),
Reflective Observation (RO), Abstract Conceptualiza
tion (AC), and Active Experimentation (AE)— plus two
combination scores that indicate the extent to which
an individual emphasizes abstractness over concreteness
(AC-CE) and the extent to which an individual empha
sizes action over reflection (AE-RO). (Kolb, 1976,
p. 1)
42
O io o s e a
Model or Goa!
Identify
D ifference*
ACCOMMODATION
CONVERGENCE
ASSIMILATION
Figure 3.
Comparison of the Experiential Learning Model
and the Problem-Solving Process
Note. From Organizational Psychology an Experiential Approach by
Kolb et al., Copyright 1979 in Englewood Cliffs, New Jersey,
by Prentice Hall.
4 3 <
From the six scores generated by the inventory, Kolb iden
tified four statistically prevalent types of learning
styles which he named: converger, diverger, assimilator,
and accommodator. Each of these types falls in one of
the quarters of the LSI grid, thus comprised from a combi
nation of two adjacent poles.
Kolb (1979) relates the two major dimensions to the
directions of cognitive development identified by Piaget
and also emphasized by Bruner. The concrete-abstract
dimension parallels also Witkin's research on cognitive
styles in terms of global versus analytical functioning,
or field-dependence versus field independence. The active-
reflective dimension seems more similar to Kagan's
reflection-impulsivity cognitive style dimension.
The Learning Style Inventory (LSIj * is used primar
ily in two ways: first, as- a learning devise helping an
individual or a group to understand its particular learning
style, and second, for research purposes.
Most of the research studied by Kolb and others is
in the area of career choice, relating majors in college
or professions to the individual styles. Testing college
students resulted in a distribution of undergraduates'
majors quite consistent with the theory that each field of
study has its unique characteristics and people choose
areas that are consistent with their learning styles
44
(Kolb, 1974, 1976). In a study of medical students Plov-
nick (1975) found this suggested relationship with medical
specializations. In a later study Sadler, Plovnick and
Snope (1978) found that "family practice residents tend to
learn through a more active and concrete approach" (p. 849).
Contrary to the previous findings, Wunderlich and Gjerde
(19 78) found no association between learning style and
career choice in medicine. Still, one of their results
indicates that the composite score (active-reflective)
differentiates between psychiatry and family medicine.
Family doctors seem significantly more active than psychi
atrists who are more reflective.
Grochow (1973) in a study of managers* problem
solving skills found a strong relationship between their
learning style and the way they solve problems and make
decisions. This finding supports Kolb's conceptualization
that "... learning and problem-solving are not different
processes but the same basic process of adaptation viewed
from different perspectives" (Kolb, 19 74, p. 34). Kolb
demonstrates this point by overlaying Pounds' (1965) model
of the problem-solving process on the experiential learning
model. It is shown that "... the stages in a problem
solving sequence generally correspond with the learning
style strengths of the four major learning styles described
earlier" (p. 34). (See figure 3))
45 ;
In another study correlating LSI scores and "stu
dents 1 rating of situations that facilitate their learning"
(Table 8, p. 33, Kolb, 1976), it seems that the learning
style corresponds to characteristics of the learning situa
tion. For example, lecture correlates positively (p < .05)
with Reflective-Observation (RO), and negatively (p < .001)
with Active-Experimentation (AE) and (AE-RO), a composite
score of this dimension (Appendix ©). In contrast, small
*
group discussion correlated positively with Active (AE)
and (AE-RO) both (p < .05). Thus, it appears that the
active/reflective dimension of the LSI corresponds to the
demands of the two teaching methods— lecture and group-
discussion— in terms of the amount of activity and/or
reflectivity required from the learner.
Randolph and Posner (1979) provide a "conceptual
grid of learning styles and pedagogical techniques" (Figure
1, p. 463), which presents a framework for planning
learning situations that match the individual’s style. In
their grid the active/reflective dimension corresponds to
more student-centered instructional methods (active-
accommodator and converger) versus teacher-centered methods
(reflective-diverger and assimilator). Stated in other
words, "This dimension illustrates the difference between
learning-by-doing and learning-by-observing" (McMullan &
Cahoon, 1979, p. 455).
46
In occupational therapy few studies concerning
individual students' learning style exist. Llorens and
Adams (1976, 1978) used the Canfield-Lafferty Learning
Style Inventory (19 74), and the Myers-Briggs Type Indicator
(1962), to describe occupational therapy students' learn
ing styles. Their findings indicate students' preferences
for: informal teaching conditions, working with others,
and setting one's own objectives, all of which point toward
involvement in the learning process. In another study by
Rezler and French (1975), who identified learning prefer
ences and personality types of students in six allied
health professions, occupational therapy students seem to
prefer concrete learning situations, interpersonal rela
tions and teacher-structured methods. The same results
were found by Rogers and Hill (19 79) who investigated
learning style preferences of occupational therapy students
at USC using the Learning Preference Inventory (LPI)
devised by Rezler and French (19 75).
This investigator, in a pilot study using Kolb's
LSI and testing 44 occupational therapy students at USC,
found that approximately 6 3 percent of the sample scored
high on concrete and were equally divided between the
active and reflective poles (accommodators and divergers).
The other 37 percent were divided into 20 percent assimila-
tors, and 17 percent convergers. Comparing the other
47
scales mentioned above to Kolb's LSI, the occupational
therapy students' learning preferences (on the Canfield
inventory) seem to correspond to the diverger type (high
concrete and reflective), while their personality type
(Myers-Briggs indicator) correspond in part to the accommo-
dator (high concrete and active) and the diverger type.
All of these findings support Kolb's postulate
about the relation between career choice and individual
learning style. But, as suggest by Llorens and Adams
(1978) ,
The value of studying individual differences of
students within a profession as being of more import
ance for teachers who wish to adjust their teaching
styles to students preferences in learning, than
identifying differences among the professions.
(p. 161)
Moreover, the concept of learning style seems to encompass
cognitive and perceptual, as well as personality factors
(Kolb, 1974, 1979; Llorens & Adams, 1978); thus it appears
to be a relatively stable trait which goes beyond mere
preferences and beliefs. As such, this individual aptitude
may account for differences in instructional effectiveness,
in particular when conceptualized on the basis of the
capitalization model--matching teaching methods to indi
vidual learning style, building upon the learner's assets
and strengths as Cronbach and Snow (19 77) describe it.
According to their review 1 1 . . . basing instructional
adaptations on student preferences does not improve
4 8
learning and may be detrimental. But, a capitalization
strategy need not rest on Ss's stated preferences"(Cronbach
& Snow, 1977, p. 170). The learning strategy or style a
student has may differ from his/her preference and still
be a better predictor of certain outcomes. They recommend
further investigation of these typical strategies or styles.
The individual learning style and the specific
dimension of active/reflective chosen to be investigated
in this study, is hypothesized to match the two teaching
methods— lecture and group-discussion (review in the next
section)--in terms of student versus instructor centered
methods or learning-by-doing versus learning-by-observing.
Teaching Methods: Lecture Versus Group-Discussion
This section focuses on the traditional way inves
tigation of teaching methods was conducted and the results
found. The two methods lecture and group-discussion, were
chosen because of their frequent use in occupational ther
apy education, as well as in higher education in general.
As Dietrick (1960) mentioned in his review, a major
difficulty in determining relative effectiveness of group
discussion is the many existing variations of this method.
In addition, the same method is identified in different
terminology: student-centered group discussion, student-
led discussion, non-directive method, democratic, group-
centered discussion, etc. In contrast, the lecture method
49
is identified as: traditional or conventional method,
instructor-centered (when performed in small classes)
directive, authoritarian, etc. For the purpose of this
review comparing lecture to group discussion, the different
variations within a method and the different terminology
are jointly considered.
The most simple and appropriate definition of
lecture, according to an early source (Stovall, 1958) is
". . . a more or less continuous oral presentation of
information and ideas by the teacher, with little or no
active participation by the members of the class" (p. 255).
In contrast, "A discussion involves the verbal interaction
of a number of individuals who perceive one another as
participants, in a common activity" (Holfeomb & Garner, 1973,
p. 4 8).. This common activity refers to learning through
discussion with an emphasis on learning in terms of cog
nitive outcomes (Hill, 1969). By accepting these defini
tions/ those predominately affective goals and emotional
growth, or groups with focus on social interaction, whose
main goal is attitude change, have been excluded from the
review groups. Integrating various sources, which- are
only a small representative sample of the conceptual and
empirical literature (Miller (Ed.)., 1961; McKeachie, 1963,
1970, 1976, 1978; Holcomb & Garner, 19 73; Stanford & Roark,
1974; Olmstead, 1974; Gall & Gall, 1976; McLeish, 1976;
Gage, 1976; Sharan & Sharan, 1976; and Gagne & Briggs, 1979).
50
the following summary charts presenting the advantages and
limitations of both methods are provided. It is important
to emphasize that the following listing includes the
possible attributes of the methods, but not always neces
sarily the practiced ones.
Advantages and Limitations of the Lecture Method
ADVANTAGES
Instructor can;
Vitalize facts and ideas
Supplement material of
textbook
Clarify difficult concepts
Emphasize significant
information
Synthesize from many
sources
Introduce a new subject
Demonstrate how to organize
material and give learning
guidelines
Explain material through
different avenues
Incorporate most recent
research
Add demonstrations and
use audio-visual aids
Pace students' rate of work
LIMITATIONS
The method:
Provides limited feedback
for students and teacher
Increases students depen
dence
Requires the student to
hold a passive role
Is not appropriate for
objectives like:
problem solving
learning motor skills
attitude change
social interaction
Results in low rate of
retention
Provides one way communica
tion
Is overused and misused for
inappropriate goals
Provide opportunity to
raise questions
51
ADVANTAGES (cont.) Lecture Method (cont.)
The method is:
Economical in use of manpower
and can save student time
Can be given to any size class
Better for lower ability
students
Most appropriate for
achieving knowledge and recall
of facts
Advantages and Limitations of the
Group Discussion Method
ADVANTAGES
The method:
Provides opportunity to
interact with instructor
and peers
Enables to raise questions
and pursue ideas and
problems
Enables synthesis of
various experiences
Provides active partici
pation and involvement in
the pearning process
Enables to practice skills
Requires from the student
responsibility for his own
learning
Allows free interaction
Provides feedback
LIMITATIONS
It is difficult to predict
adequate coverage of subject
matter
Students1 preparation in
advance is necessary
Size is important--large
numbers do not provide same
opportunity for each member
It is time consuming
There are students who domi
nate the discussion
Difficulty in eliciting
minority opinions
Students' privacy is some
times disturbed
Members who are unwilling
to respect intellectual
skills of peers interfere
in the group process
52 .
Group Discussion Method (cont.)
ADVANTAGES
Fosters independence and
cooperation to achieve a
common goal
Provides verbalization
of ideas which helps to
learn and clarify own
points of view
Requires listening to
others, sharing with
others, and compromising
with others in order to
achieve a common goal
Enables active decision making
Gives control over learning
activities
Provides a relaxed atmosphere
for learning
Allows cooperation in planning
Is a stimulating experience
Fits the following objectives:
application, problem-solving,
critical thinking, changing
attitudes, social interaction,
delayed retention, motivation,
leadership roles, skills in
group problem-solving, abstract
levels of thought.
In general, the findings of traditional comparative
research in this area are confusing and contradicting.
Research on the subject started as early as 1925, and con-
tinued in a very similar approach through the 1960s.
53 "
In a frequently cited review written by Dubin and Taveggia
(1968), (in which the authors reanalyzed 91 studies, among
them 4 5 on lecture versus group discussion), the overall
conclusion indicates that there is "no significant differ
ence" among various methods of college teaching as measured
by final exams. The authors emphasize, in a very provoca
tive way, that there is no use in further studies of the
same kind. They suggest that research should take on new
directions, and researchers should ask new questions and
try to answer them in new ways.
Many authors argued against these conclusions,
pointing out that they have not considered other kinds of
learning outcomes or higher cognitive levels (McPherson,
1976). Gall and Gall (1976) disagree on two major issues:
(1) they emphasize methodological problems recurring in
this specific area of research, and (2) they criticize
Dubin and Taveggia who "... did not critically review
the methodology used in each study" (p. 197). They
" . . . recommend that efforts be turned to conducting new
studies characterized by methodological rigor" (p. 198).
McKeachie, in his continuing reviews on teaching methods
effectiveness over the years, concludes that while there is
no difference on measures of knowledge, group discussion is
more effective in developing concepts, problem-solving
skills, attitude change and motivation. Berliner and Gage
(1976), as well as McKeachie (1976) emphasize the
54
importance of specifying objectives and measuring each of
them with appropriate tools. McKeachie and Kulik (1975)
provide a summary of research on effective college teach
ing, in which they emphasize once more that small classes
and the discussion method are more effective for developing
problem-solving ability. Moreover, studies comparing
student-centered discussion with instructor-centered, indi
cate that the former is superior on measures of critical
thinking and problem-solving. It seems that increased
activity and interaction within a group improves students'
performance. The method of "group investigation" developed
by Thelen is an example of a strategy which focuses on the
inquiry aspect of learning and the importance of active
participation in the process (Joyce & Weil, 1972), "Activi
ties cease to be inquiry when the teacher is the sole
source of the problem identification and formulation of
plans ..." (p. 41). According to Thelen, the group size
should range from 10 to 15 students which is similar to
the size of a small class indicated in various studies.
"This is large enough for a diversity of reactions and
small enough for individual participation" (Joyce & Weil,
1972, p. 43).
Reviewing studies in medical, dental and nursing
education which pertain to the same comparison of lecture
versus group discussion, very similar findings were
55/'
reported. Investigation of teaching methods effectiveness
in medical education is relatively new, dating back to the
early 1960s.
Teaching and Learning in Medical Education (Miller
[Ed.], 1961) was among the first major theoretical publica
tions dealing with the issue of teaching/learning in medi
cal education, emphasizing the importance of the instruc
tional process. An early and still classical study was
Zimmerman and King (196 3) "Evaluation of Student Centered
Group" in medical education. They found the group discus
sion to be significantly superior to lecture on measures
of critical thinking and problem-solving, while equally
effective on knowledge measures. The group discussion in
their study was organized such that the students led the
group, while the instructor was viewing the discussion on
a videotape, during the first hour. For an additional half
hour he joined them and provided feedback, clarified diffi
cult points in the material or facilitated further dis
cussion. This method seemed to require full students'
participation and at the same time allowed the instructor
some intervention. The results of this study are supported
by others in medical education (Netsky, Banghart & Hain,
1964; McCarthy, 1971; Barrows & Tamblyn, 1976; Mayau,
1978; Horne, 1979; Huckabay, Cooper & Neal, 1977, 1979),
and in general support the conclusions stated previously
56
that no significant difference was found on measures of
knowledge, but groups are superior for acquiring problem-
solving skills and increasing critical thinking. In con
trast, in a more recent study, Shumway and Donahue (19 80)
found no difference on PMP performance between the two
methods.
The research findings suggest some directions for
education which have to be implemented in order to evaluate
their actual effectiveness in real life. Ways, Loftus and
Jones present an approach to medical education which inte
grates problem-solving as a teaching mode, utilizing the
small group format, thus focusing on problem-solving as
the learning process as well as learning outcome which is
similar to Gagne's (1977) conceptualization. Such a pro
gram has to be studied longitudinally and thus, will pro
vide in the future more accurate answers. This approach
to research, while very important, is only one way to
improve understanding of instructional effectiveness. Tra
ditional research as reviewed here attempted to find the
"best method" for a particular learning outcome, without
taking into consideration the individual learner. The
effects of a particular teaching method were assumed to be
equally effective for each individual, or the individual
differences were cancelled out by the use of means in the
analysis. This ommission of the third major variable in
57
the model of the teaching/learning process may explain the
inconsistent and contradicting results of many studies.
Furthermore, research which attempts to deal with all
three variables--the individual learner, the teaching
method and the learning outcome— enables better understand
ing and more accurate conclusions.
This line of reasoning led to the area of research
on instruction referred to as Aptitude-Treatment Inter
action (ATI) or Trait-Treatment Interaction (TTI), which
stresses the importance of combining the study of indivi
dual differences and instructional effectiveness (Berliner
& Cahen, 1973; Cronbach & Snow, 1977) . As discussed
earlier in the section on ATI, and emphasized by Cronbach
and Snow (1977), treatments need to be analyzed according
to the amount of burden placed on the student.
Referring back to the summary charts in this"
section and taking in account basic features of the two
methods as manipulated in this study (Chapter III), the
main differences among them are presented in Table 1.
These components of the two methods are hypothe
sized to match with students' learning style such that,
active learning style will profit more from the group-
discussion method, while reflective learning style will
match better the lecture method.
58
Table 1
Features of
Methods
the Two Teaching
in This Study
Lecture Group-Discussion
instruetor-dentered student-centered
large class size small class size
limited feedback high degree of feedback
limited involvement high degree of involvement
only reading assignments assignments requiring
active experimentation with
materials, and preparation
of issues to discuss
limited responsibility
for class function
high responsibility for
class function
individual learning cooperation required to
achieve a common goal
Considering, in addition, the findings summarized
from traditional studies concerning the problem-solving
outcome, students in,the group discussion method were
hypothesized to do better on measures of problem-solving
like the Patient Management Problem (PMP) discussed in the
next section.
Problem-Solving in the Health Professions
In the most recent and comprehensive textbook of
occupational therapy, Hopkins and Tiffany (1978) state
that:
59 <
The occupational therapy process is one which utilizes
problem solving methods for finding the best, most
appropriate means of helping those individuals requir
ing occupational therapy intervention reach to their
highest potential for function in their own roles
within their own environments. Occupational therapists
must approach each new situation as an opportunity or
challenge to find meaningful solutions to the problems
facing the patients or clients with whom they are
working.
Occupational therapists must become skilled Vin
problem identification, using astute observation and
the many evaluation procedures available to them. In
this way a data base can be established which will
identify the real problems. Once general problems are
identified, they can be broken down into subproblems
that guide the therapist in determining the goals of
treatment or intervention. Alternative approaches
for solving the problems must be identified and a plan
of action is chosen that seems to be promising for
resolution of the problem. (p. 109)
The problem-solving process outlined above is
identical with its description in medical education, and
as defined by Vu (1980) involves !identification of a
problem, collecting data about it, interpreting the infor
mation, and integrating the data in order to plan the
treatment. Andrew (1972) and Ways et al. (1973) outline
in more detail the steps involved in the problem-solving
process (as presented in Chapter I under definitions of
terms), which takes the form of a clinical management pro
cess in which the therapist is involved during the treat
ment of his clients. Actual clinical practice is conceived
in terms of a problem-solving process in all health pro
fessions, and therefore the main emphasis is given to the
acquisition and measurement of problem-solving ability.
60
There are two aspects of problem-solving with which
educators in the health professions are confronted: (1)
how to teach the problem-solving process, and (2) how to
evaluate the skill or ability acquired. In the past, and
in most traditional schools problem-solving was not expli
citly taught in school; students were supposed to acquire
it mostly in their clinical affiliations through experi
ence. There is some change in this approach in medical
education as seen for example in McMaster University which
developed a problem-based learning program (Neufeld &
Barrows, 1974), as well as in a few other schools (Berner,
1976; Shumway & Donahue, 1980; Ways et al., 1973). In
addition to the few comprehensive teaching programs, spe
cific teaching techniques were developed to allow students
to practice and work through various types of clinical
problems, among them are the written simulations such as
Patient Management Problem (McGuire et al., 1976), card
decks, computer based simulations, simulated patient, etc.
In occupational therapy a first attempt to teach
psychological treatment planning through case simulations
is reported by Briggs, et al (19 79), and as stated by the
authors their goal is " . . . to add in development of
basic clinical skills as preparation for actual practice"
(p. 1) .
61 \
The second aspect, evaluation of the problem
solving ability, is comprised of two parts, as Barro (1973)
states:
Two major approaches to the evaluation of performance
are the assessment of process and outcome. Assessment
of process is the evaluation of the activities of
physicians and other health professionals in the
management of patient; . . . assessment of outcome is
the evaluation of end results in terms of health and
satisfaction. (p. 1053)
Traditionally, the main focus in medical education was
centered on the evaluation of the outcome aspect of problem
solving and not its process. The newly developed simula
tion techniques provide a way of evaluating to some extent
the process involved.
Patient Management Problem (PMP)
Programmed testing of Patient Management Problems
(PMP) or written simulations was introduced for the first
time in 1961 by the National Board of Medical Examiners.
The aim was " . . .to test aspects of clinical competence
dealing with the ability to identify, to resolve, and to
manage patient problems in a method which simulates a real
istic clinical situation" (Hubbard, 1978, p. 40). The
written simulation as an educational technique, had advan
tages for instruction, diagnostic testing and formal
examinations. It can be viewed on a continuum between con-
ceptional testing methods, and evaluating field performance
in a clinical setting, which are usually the two aspects
62
for assessing students' success. For the examinee the
PMP provides realistic feedback in a practical time frame;
it gives a sense of relevance and responsibility for action
without danger to a person's life (McGuire et al., 1976).
The concept of simulation consists of "... pla
cing an individual in a realistic setting where (s)he is
confronted by a problematic situation that requires a
sequence of inquiries decisions and actions" (McGuire et
al., 1976, p. 1). The simulation techniques are assumed
to provide better indication of performance and particu
larly of problem-solving skills essential in the treatment
process. Furthermore, it is postulated that the process
involved in the simulation is similar to the one which a
person will face in actual practice. Thus, it is assumed
that a student who is able to solve a written simulation
(PMP), will succeed also in solving actual problems on the
job.
Description of the instrument. The PMP is a paper-
and-pencil exercise. It starts with a brief paragraph
describing the situation, the patient and the problem
(opening scene), as well as the role the participant has
to play (Appendix G). The description reveals only those
aspects of the situation which the participant (whose role
the examinee takes) would have immediately available.
Following the presentation of the initial data, a first
63
section of choices is offered, and directions of how many
options are allowed as choices is given (Appendix E).
After reading the options the examinee chooses among them
and records his/her decision by developing a latent image
on a specially treated answer sheet (practically, it is
done using a special pen and rubbing it across the number
of item chosen). The response received may direct to
another section, provide feedback or additional informa
tion, or present results of actions taken, all of which may
be taken in account and followed during the exercise
(Appendix E: ) • Each PMP consists of a series of sections
not necessarily in the right order. Two basic types of
PMP exist in medical education, a "linear" format like the
one used by the National Board of Medical Examiners, in
which each examinee follows the same sequence of sections,
and a "branching" format developed by McGuire at the Uni
versity of Illinois, which allows the examinee to follow
his/her own sequence according to decisions in each step
of the exercise. Every PMP is built in such a way that an
optimal route exists, and if taken by the examinee results
in the highest proficiency and efficiency scores. McGuire
et al. (1976) describes various models for the construction
of PMPs. In this study (model 5) the single-major compli
cation model was used (schematically represented in
McGuire's book on p. 280 which includes few branching
64
options and may result in one wrong alternative at the
end (See Appendix E for diagrams of the two PMP's con
structed for this study).
The scoring system used in this study followed
McGuire et al. (19 76) . Thus, every PMP yielded 5 scores:
proficiency, error of ommission, error of commission,
efficiency, and overall competence (which is derived from
the proficiency and efficiency scores). In the data
analysis the two error terms were not used, thus three
scores for each PMP are reported in the study.
Proficiency is defined as " . . , the degree to
which an individual's route through the simulation . . .
correspond with the one judged to be optimal by the author
and other experts in the fie Id" (p. 2000* i : It-Reflects the abil
ity to choose the helpful options and avoid the wrong or
harmful ones.- Efficiency is defined as the ratio between
the number of helpful choices and the total number of
options. The overall competence score reflects "...
the appropriate balance between proficiency and efficiency
in the resolution of the particular problem presented in a
single simulation" (p. 201) (for actual development of the
scoring system, coding and weighting of items, see McGuire
(1976, Chapter 13). Additional ways to weight items were
reported by some authors, but as suggested by Donnelly's
(19 76) results, there were no differences between the way
65 .
the items were weighted; simple procedures seemed as ade
quate as more complex ones.
Studies Related to the PMP1s Measurement Properties
Page and Fielding (1980) have said, "Written simu
lations of clinical encounters or PMP's, are becoming
increasingly prominent in the testing programs of medical
colleges, licensing, and certifying bodies" (p. 529). As
such, the main question which investigators study is their
reliability and validity. Reliability is difficult to
establish in the traditional ways, especially internal
consistency because items are interrelated and dependent
on each other, and subjects who choose various routes are
not comparable in terms of the items they choose. Consis
tency over various PMP's was reported by Berner, Hamilton
and Best (1974) and McGuire and Babbott (196 7) between .73
to .90 depending on the number of PMP1s and their length.
These results which seem to suggest high consistency among
different PMPs are not supported by most later research
(Harasym, 1979; Sprafka & Scheifley, 1980). On the con
trary, much discussion is devoted to the phenomenon of
"case specificity" or inconsistency in performance of
students across cases, showing low correlations among PMPs.'
Skakun (19 79), who used a computerized version of the PMP,
found the same pattern of low correlations between the
various PMPs. Another, more basic way to determine relia-
bility, namely the interjudge agreement is used in most
66
studies. Correlations of .71 to .85 are reported by
Sedlacek and Nattress (1972) and Page and Fielding (1980)
referring to agreements about decisions made throughout
the exercise. The same method is used also to determine
content validity of the PMPs which is the major strength
of the'instrument as judged by experts and by students who
use them in their training. Besides content validity,
construct validity is demonstrated in many studies-by cor- ,
relating the scores on the PMP' s with scores on multiple-
choice exams, showing usually low to moderate correlations
(Berner et al., 19 74, McGuire & Babbott, 196 7; Palva, 19 74;
Skakun, 1979; Sprafka & Scheifley, 1980). This is also
reported from the National Board of Medical Examiners
(Hubbard, 1978). As Berner indicates "the low positive
correlation may mean that different areas of competence
were being tapped by the two examinations" (19 74,p. 671).
Another approach to construct validity or "the known group"
is to correlate scores of experts and students, or students
at different educational levels. These coefficients are
usually low, indicating the discrimination power of the
test (Hefler & Slater, 1971; Barro, 1974; McGuire et al,,
1976; Page & Fielding, 1980). The same approach is also
used in nursing education (DeTornyay, 196 8; Gover, 19 72;
McLaughlin, 1979, 1980).
6 7 ■
As discussed earlier, performance on the PMP is
assumed to correlate positively with performance in clini
cal settings. To investigate this relationship some stu
dies attempted to compare performance on PMPs with per
formance in practice using simulated patients (patients
who are trained to role play various cases) (Page &
Fielding, 1980; Feightner & Norman, 1976), or actors who
played the patient's role (Goran, Willianson & Gonnella,
1973; Elstein et al., 1978). In all these studies only
low positive correlations were found. Performance on PMP
was usually inflated, particularly many more "must do"
items were used, more behaviors were displayed, and greater
number of encounters were selected. These findings led the
investigators to conclude that PMPs test ability to per
form and not performance per se. Elstein et al. (1978),
who undertook the most extensive study, reported a series
of investigations where the main purpose was to describe
cognitive processes used by physicians to reach diagnostic
decisions. The simulations were played by actors and were
called high-fidelity in contrast to low-fidelity of the
PMP. In general they found that diagnostic performance is
a case specifically related to the medical problem; as
mentioned before, the process seemed to be hypothetico-
deductive, and no statistically significant difference was
found between groups of physicians chosen for their
6 8
proficiency, but in comparison to medical students their
findings indicate differences in performance. McGaghie
(1980) reanalyzed and critiqued the high-fidelity simula
tions studies and found that "... the research contains
several basic defects," (p. 920) mainly because they used
a small and restricted sample size, and analyzed the data
with factor analysis which is not appropriate.
Taking research findings in this area together, it
seems that PMPs provide indication for inadequate perfor
mance, but have not enough discriminating power on the
higher end of the problem-solving ability. All researchers
recommend further investigation of PMPs because of their
high content and construct validity, their instructional
usefulness, and their advantage as a testing instrument
for large groups when only short time is available and an
objective score is needed.
Achievement, Attitude Toward Instruction,
and Amount of Study Time Outside Class
Achievement is the most commonly used outcome in
traditional as well as in ATI research, especially achieve
ment as measured by final exams requiring the student to
demonstrate knowledge acquisition. Advantages of consider
ing interactions between students' characteristics and
instructional components in explaining achievement were
presented in many studies, and reviewed in detail by
6 9' ;
Berliner and Cahen (1973), and Cronbach and Snow (1977).
But, as Strom and Hocevar (1980) state, only few studies
have considered satisfaction as an ATI outcome.
Affect, as a variable in ATI research is used both
ways, as an aptitude and/or an outcome. Under this general
area various terms were used, like preferences and beliefs,
attitudes, satisfaction, enjoyment, etc. Differences are
also made in terms of preferences or attitudes toward the
model--the subject matter or general attitude and the like
--or in terms of satisfaction, enjoyment or benefit from
structure versus freedom in the instructional method
(Cronbach & Snow, 1977; Clark, 1981; Strom & Hocevar, 1980,
1981). The few studies which investigated these variables
provide inconsistent findings; for example, Shaw and Hunt
(1979) report that satisfaction in terms of preference for
structure did not mediate between course structure and
achievement, and as an outcome had no significant results.
In contrast, Strom and Hocevar (1980) "suggest that indi
vidual differences moderate the influence of course struc
ture on student satisfaction" (p. 11).
Another relationship which shows some consistency
is between achievement and enjoyment. Clark (19 81)
reviewed various studies which show negative correlations
between students * achievement and enjoyment from the
methods in which they participate.
70 >
Similar relationships were found also in some studies
between preference for teaching approach and performance
(Dowaliby & Schumer, 19 73; Peterson & Janicki, 19 79) which
led Cronbach and Snow (19 77) to the conclusion that stu
dents' preferences or beliefs do not generally interact
with method of teaching. Snow (1980) goes further to
suggest that "learner control" as an overall concept of
the above mentioned affective variables, does not compen
sate for the effects of individual differences. In this
view he is also supported by Atkinson (1972) who states
that " . . . the learner is not a particularly effective
decision maker . . . of what to study, when to study, and
how to study" (p. 9 30).
Attitude variables were investigated as outcomes in
two recent studies reported by Corno, Mitman and Hedges
(1981), and Janicki and Peterson (1981). In both studies
attitude toward school or toward math showed significant
ATI with ability, while attitude toward approach in Janicki
and Peterson's study had no significant ATI and was mainly
influenced by teacher main effect.
Although some relationships are suggested by the
literature, no conclusive findings can be presented, and
many more studies are needed in order to understand the
contributions and interrelationships among the various
cognitive and affective variables.
ir
Salomon (19 81) postulates that the Amount of
Invested Mental Effort (AIME) may provide the mediating
variable between Perceived Demand Characteristics (PDC)
and performance, or learning.
Such perceptions are influenced by three related
factors: (a) the nature of the encountered message-
event, (b) one's past experience with and future
anticipations from the event and (c) one's per
ceived self-efficacy in processing such a message-
event "in depth," (Draft, Chapter 5, no page no.)
Thus, amount of study time invested as part of AIME may
suggest some directions and relationships among aptitudes
and instruction.
Summary
The literature reviewed in this chapter outlines
the framework for this investigation. The teaching/'1 -
learning process is conceptualized as an interactive pro
cess between a person's characteristics and the environ
mental manipulation. These two factors while interacting
may affect differentially every particular outcome sought.
Thus, the extent to which goals or objectives of any edu
cational process are reached depend largely on the inter
action between the individual and the situation he/she is
in. This is also the main premise of the Aptitude-
Treatment Interaction approach which is followed in this
study conceptually and methodologically.
The main variables which were chosen for investi
gation in this study--individual learning style (aptitude),
72
and teaching methods or lecture versus group-discussion
(treatment)— were hypothesized to interact, and as such
explain the outcome-problem-solving skills of occupational
therapy students. A review of concepts and studies perti
nent to these three major areas is provided in the chapter,
as well as a discussion of additional outcomes measured.
The ATI matching hypothesis based on the capital
ization model was presented and used as the underlying
design conceptualization and hypothesis predicted. The
active/reflective dimension of the individual learning
style was postulated to correspond to the two teaching
methods in terms of the amount of participation and
involvement required from the student. Thus, matching
these two variables was considered more beneficial for
students, allowing them to utilize their strengths which
would result in higher performance, especially on an out
come like problem-solving.
73
CHAPTER III
METHOD
Introduction
This chapter consists of the following sections:
the research design, subjects and sampling strategy,
instrumentations, data gathering plan, data analysis,
and methodological limitations.
Research Design
The research methodology of this study follows
an experimental design, the posttest-only control group
design (Campbell & Stanley, 1963), and as recommended
by these authors ". . . if appropriate antecedent variates
are available, they should certainly be used for blocking
or leveling, or as covariates" (p. 26). This design is
basically the same design utilized in most ATI research
(described briefly in Chapter II under the ATI heading)
in which one independent variable, the treatment, is
manipulated while the aptitude („s) used as independent
variables or covariates are measured before the experi^
mental period starts. The dependent variables or outcomes
are measured at the end of the experiment (posttest only),
thus controlling for any pretest effects of interaction
between pretest and treatment.
74
The study is conceived as a 2x2 factorial design
with one main independent variable, "teaching method,"
manipulated, and the other main independent variable,
"individual learning style," measured before the instruc
tion started.
Figure £ may represent the design and the hypothe
sized match between the teaching method and individual
I
[learning style or type.
Individualized Learning S
Reflective (
Diverger & Assimilator Acommodator ; &
Active
Teaching
Method
(Treatment)
Group
Discussion
Note. + match
- mismatch
Figure : 4
The Study's Design and Matching Hypothesis
■ 75 •
This design is considered to be one of the more
powerful designs for experimental purposes in education
because of its adequate control for internal validity,
expecially due to the random assignment and lack of
pre-testing (Kerlinger, 1973; Campbell & Stanley, 1963).
The question of generalizability is more difficult to
answer because subjects and schools are not randomly
selected. The best solution to this problem, as suggested
by Kerlinger (19 73) is replication of the study. (The
investigator is planning to repeat the study in Israel,
thus trying to resolve some of the generalizability
problem at a later date.)
The study includes the following variables: Two
main independent variables— (1) teaching method with two
levels lecture and group discussion, and (2) individual
learning style, defined as a composite score on the LSI
of the Active/Reflective dimension (AE-RO), or categorized
into Learning Types on the same dimension (Accommodator
& Converger/Diverger & Assimilator).
Five additional aptitude variables were obtained:
(1) prior achievement, subjects1 GPA in occupational
therapy; (2) verbal ability, defined as a score on a
group measure of verbal ability— Terman1 s Concept Mastery
Test (CMT); (3) perceived benefit from lecture; and (4)
perceived benefit from small group, both of these being
76. '
defined as a score on a 7 point Likert scale; and (5)
Educational level--seniors or graduate students.
Four kinds of outcomes or dependent variables
were assessed in the study. Two were cognitive outcomes;
Cl) achievement, defined as a score on a forty item
multiple-choice-question-exam; and (2) problem-solving
skill, defined as a score on two "Patient Management
Problems" CPMP). For these two outcomes specific hypothe
ses were outlined in Chapter I. The two additional
outcome variables were added in an exploratory way and
only questions were raised regarding them. /The other
outcomes were; (.3) amount of study time for every class
session, as well as study time for exams defined on a
5 and 4 point Likert scale respectively; and (4) attitude
toward instruction, which divided into four scores—
Benefit, Enjoyment, Confidence, and Overall Attitude,
all of which are measured on a 5 point Likert scale.
Measures are outlined in the instrumentation
section and appropriate Appendixes.
Subjects and Sampling Strategy
The subjects in this study were occupational
therapy students (seniors and first-year entry-level
graduates), enrolled at the University of Southern
California (USC) in the academic year 1980-1981 who
signed an informed consent form (see Appendix A).
77
Forty-four students participated from a class of forty-
seven (2 students refused to sign the consent form and
1 was sick half of the period) among them 43 students
were females and 1 was male. The scores obtained from
the only male were inspected to determine their equality
to the females' scores. No significant differences were
found; thus, he was included in the study's data analysis.
Subjects were randomly assigned, using a table of
random numbers, to either of two teaching methods:
lecture or group-discussion. Following a randomized block
design procedure they were first divided on the basis
of their LSI scores into two groups; according to reflec
tive or active learning styles, and then randomly assigned
within these groups to the two teaching methods, half
to the lecture and half to the group-discussion method.
Within the group—discussion method subjects were once
more randomly assigned to two equally conducted classes.
Thus, the sample size of 44 was divided into 22 subjects
in one lecture class and two small group-discussion
classes, of 12 each.
Subjects received a general explanation about the
study two months before instruction started and at that
time they signed the informed consent form. They were
asked not to study at home with members of the other
classes during the experimental period and not to exchange
78
materials. The investigator was aware of the threat to
external validity occuring in a situation where subjects
know that they are involved in a research study, (the
known "Hawthorne effect") but it was assumed that it
would not have a substantial effect, because the course
was essentially conducted in a similar manner to other
courses in the program in terms of schedule, requirements,
grades, etc.
instrumentation
Six instruments were employed in this study;
1. Learning Style Inventory (Kolb et al., 19 76) -(.LSI)
2. Students Ratings, of Situations that Facilitate
their Learning
3. Concept Mastery Test (CMT), Terman (1956)
4. Multiple-choice-question exam
5. Patient Management Problem (PMP)
6. Student Evaluation questionnaire
The Learning Style Inventory (LSI) was published
in a technical manual (Kolb, 1976) including its scoring,
profile, and grid. The model upon which it is based is
provided in Chapter II. The instrument was developed
by Kolb et at., (19 74), and data on its reliability and
validity are presented in the technical manual (Kolb,
19 76, updated 1978). Split-half reliability for various
samples and for each of the six scores range between .40
79
to .86. The composite scores exhibit values in the
higher range .78 to .86 for the different samples, while
their range for test-retest reliability is between .43
to .71 for the various samples. Geller (1979) presents
additional test-retest data which are very similar to the
previous one; scores ranged between .52 and .70.
The instrument was criticized by various authors
because the reliability values are generally too low
under classical measurement theory (Freedman & .Stump1,
1978; Geller, 1979). In addition, the instrument construc
tion as a forced—rank choice seems to introduce bias in
favor of the bi-polar dimensions (Lamb & Certo, 19 78;
Wunderlich & Gjerde, 1978). In a recent article Kolb
(.19 81) rejected the critics by emphasizing that split-half
and test-retest reliability are most appropriate for
independent variables, while
The basic learning modes assessed by the LSI, however,
are theoretically interdependent (i.e., any action,
including responding to the test, is determined in
varying degrees by all four learning modes) and vari
able (i.e., the person * s interpretation of the situa
tion should, to some degree, influence which m o d e s
are used). (Kolb, 1981, p. 291)
Thus, Kolb expected the reliability coefficients to be
lower than usually agreed upon. According to Geller
(19 79) these reliability values might be satisfactory
with small groups who have a narrow range of differences
similar to this study, but not for large and disparate
groups.
80
For the sample of this study, test-retest reliability
was calculated between two administrations of the inventory
in a two-month interval (the first two months prior to
instruction and the second time a week before). These
reliability coefficients are low, especially for the com^
posite scores. They are much lower than reported in the
literature. It is not clear why there is such a large
difference between the values when each sub-group (n=22).
is considered. The subjects who comprised the sample
for the group-discussion method have much higher reliabil
ity values than the lecture group or the combined sample,
Table 2
Test-Retest Reliability of the LSI
CE RO AC
LSI
AE
SCALES
AC-CE AE-RO
Learning
Type
All subjects
Cn-44) .46 .60 .40 .56 .30 .31 .53
Lecture (n=22) .37 .53 .09 .62 .23 .24 .40
Groups (n=22). .62 .72 .71 .52
•7.3
.77 .82
Note. Coefficients larger than .30 are significant at .05
31
In regard to validity, various kinds are discussed
and provided by Kolb (19 76, 1981). A discussion of
content validity based on the theoretical framework and
the model of experiential learning is presented. On this
basis Kolb (1981) argues also that the forced-choice rank
ing format of the LSI is in accordance with the theory's
postulate 11. . . that a learning response to any life
situation requires the resolution of conflicts among the
four learning modes, . . .thus, a test of learning styles
should be constructed so that it also required a similar
conflict among choices" (p. 293). Concurrent validity
of the LSI scores and performance tests, personality
tests, preferences for learning situations and for teach-;
ers . (Kolb, 1976, pp. 28, 30-34) . The most significant
correlations exist between the dimensions of concrete/
abstract and feeling/thinking on the Myers-Briggs Type
Indicator (MBTI), as. well as between active/reflective
and extrovert/introvert.
Interesting, in terms of this study, are the correla
tions between the LSI and students' rating of situations
that facilitate their learning (see Appendix B for table
of correlations) (Kolb, 1976, p. 33), because the instru
ment was used in this study as a measure of Perceived
Benefit from lecture or small group (two of the items
on the scale).
82
Students' Ratings of Situations that Facilitate
Their Learning is a 7-point Likert scale adopted in part
from Kolb (1976) (Appendix B). These ratings seem to
indicate students' attitudes toward different learning
situations, or their perceived benefit from a certain
mode of instruction and thus are presumed to correlate
also with their actual performance under a given teaching
method.
Test-retest reliability for the sample of this
study was calculated between two administrations of the
test (in a two month interval). The following coefficients
relate to the two items used in the study:
Perceived Benefit from Lecture r=.60 (p=.0001), and
Perceived Benefit from Small-Group r=.46 (p=.001).
Pearson r correlations were also calculated between
the LSI scores and the rating scale, parrallel to the table
presented by Kolb (1976) and shown in Appendix B.
The correlation values found are very similar to
Kolb's data and in some cases even more pronounced. Per
ceived benefit from lecture correlated positively with
reflective (RO) and negatively with active (AE) or (AE-RO),
while Perceived Benefit from small-group correlated
negatively with (RO) and positively with (AE) and (AE-RO),
except for the groups' sample.
83
Table 3
Correlations Between LSI
of the Students1
Scales and Two Items
Rating Scale
CE RO AC AE AC—CE AE-RO
Perceived Benefit
from Lecture
Al1 subj ects (n=4 4) .15 .19 .08 -.23* -.03 -.23*
Lecture (n=22) .24 .15 .11 -.13 -.07 -.15
Groups (n=2 2) -.03 .42** .11 -.51** .05 -.51**
Perceived Benefit
From Sma11-Group
All subjects (n=44) -.17 .04 .17 .03 .14 .04
Lecture (n=22) .01 .29* .01 .44** .01 .39**
Groups (n=22) -.31* .12 . 29* -.26 .25 -.21
*p < . 10
**p< .05
Concept Mastery Test (CMT)— Form T was developed
by L. M. Terman (1956) and is defined in its manual as:
. . . A measure of ability to deal with abstract
ideas at a high, level. It is suitable for admin
istration to college juniors or seniors and to
graduate students. The test consists of two parts:
I, the identification of synonyms and antonyms/ and
II, the completion of analogies." (Terman, 1956, p. 3)
The manual and its supplement (19 56, 1973)_ provi de
data on the test's norms, reliability and validity.
84 '
Parallel forms' reliability ranges for various samples
from .86 to .94. Intercorrelations between the two parts
of the test range from .62 to .76. For the sample of this
study both parts correlated .77. These coefficients
indicate that the parts are sufficiently similar to justify
the use of a total score or a separate one. The analogies
part of the CMT was also correlated with the Miller
Analogies Test and a coefficient of .89 was found (Wallen
& Campbell, 1967).
Validity of the test is mainly presented by
correlaging it with various other aptitude or intelligence
tests, like the SAT (r~.66 to .68), or the WAIS (r=.55).
While these data provide evidence of concurrent validity,
additional data for construct validity may be seen in
the ranking of samples according to their educational
level and presenting a similar order in the mean CMT
score (a known group approach to construct validity), The
mean of the sample in this study was 64.75, which is low
according to the data in the manual, but when compared
to another recent study with college students (Strom &
Hocevar, 198.1) almost equal values were found (52.33/65.40/
60.35) for various groups.
The next three instruments were constructed by
the two instructors participating in this study and com
prise the measures of the dependent variables.
8 5
A Multiple-Choice-Question-Exam was constructed
to measure knowledge and comprehension of the content
area and is referred to as the conventional achievement
outcome in this study. The test consisted of 40 items
and was administered to all subjects at the end of the
instructional period. Internal consistency was determined
to be KR20=.5J_; a bit higher value was found for the sub
jects in the group method (KR20=.67). Although these
values are not very high it was considered satisfactory
for research purposes. Validity of the test was deter
mined in three ways,. First, content validity was assured
through a table of specification and by experts in the
area (staff members of the occupational therapy depart
ment at LAC/USC psychiatric hospital who carry out their
treatment in occupational therapy according to the theory
of Cognitive Disability which comprised the content
area):. Second, construct validity was established through
the known-group approach. The test was given to the
expert group, to a group of 0.T. students which studied
the subject-matter a year ago, and to the experimental
group. (Table 4)
Patient Management Problem (PMP) is a paper-and
pencil exercise which has the characteristics of a written
simulation. It deals with managing a clinical situation
in sequential order of a problem-solving process. The
36
Table 4
Comparisons of Achievement Scores
Between Experts and Students
Group
No.
Item Mean . . .
t
. SD Mean-percent :one-tailed p
Last Year
students 35 17.8 2.86 50.8
(_n=19)
7.19 .0005
Experts (n=6) 35 26.5 1.64 75.7
Students, in
2.48 .02
the s tudy
(n=44) 40 28.82 3.58 72.0
two PMPs for this study were developed on the model of
the test in medical education. The actual test construc
tion followed format detailed by McGuire et al., (1976)
in which the whole process is outlined in detail, including
the concept of written simulations, their construction,
scoring and interpretation. A brief description about
the test format is presented in Chapter II under the
section related to problem-solving skills in health
professions, as well as; findings about its reliability
and validity. Diagrams of this study's PMPs, directions
and an example page are presented in Appendix C.
- 87
The McGuire et al (.1976) scoring system was used,
and three of the five scores— proficiency, efficiency,
and overall competence-— were entered in the data analysis
for each PMP. Content valididy was assured with the
same group of experts mentioned above. Construct validity
was determined in similar ways as discussed in the liter
ature, i.e., the known group approach (shown in Table 5),
and correlations with the multiple-choice exam (shown
in Table 6).
Table 5
Comparisons of PMP Overall Competence
Scores Between Experts and Students
Group Score Mean SD
Last year
students
(n=l2)
Experts f (n-6)
Students in ,
the s tudy
(n=44)
PMP IOC 20
57
4 3
10
17
20
2.11
2.6 49
,05
.10
Last year
students
Experts;
Students in
the study
PMPIIOC 40
76
42
20
23
26
3.3 4
3.08
. 005
.005
Note. Range of score 1-100%, accepted level 80%
88
Table 6
Correlations Between the PMP Scores and
the Multiple-Choice-Exam
Multiple-choice
exam FMPIP PMP IE PMPIOC PMPIIP ' ..PMPIIE PMPIIOC
Experts .19 .09 .18 .37 .39 .38
Students in the
study .25 .25 .26 .15 .23 .18
The findings presented in the last two tables
indicate that experts, score significantly higher than
students, on the PMPs (as well as on the multiple-choice-
exam shown in Table 4) as expected. In addition, corre
lation coefficients between the two measures are low to
moderate which, indicates that they measure different
constructs.
A Student Evaluation Questionnaire was developed
to explore relationship between the independent variables
and the amount of time students studied outside of class,
as well as their attitude toward the instructional .method
they participated in. The instrument is presented in
Appendix D and includes 8 items constructed as a Likert
5-point scale. The first two items refer to the time
variable and were used separately in the analysis. Items
89
3 to 7 refer to the attitude variable and were used in a
combined attitude score and also seperately: item 3 —
benefit, item 4— enjoyment, item 5— confidence. Internal
consistency for the attitude items was calculated with
coefficient alpha and found to be .82.
Data-Gathering Plan
The data gathering consisted of three main phases.
Cl), Subjects were pre-tested two months before the start
of the instructional period with :the following instruments:
Learning Style Inventory (LSI), students* ratings of
situations that facilitate their learning, and the Concept
Mastery Test (CMT). In addition, . their occupational
therapy grade point average was obtained from the depart
ment secretary. One week prior to instruction, the LSI
and the rating scale were administered a second time for
the purpose of calculating test-retest reliability as
reported in the instrumentation section.
(2) At the beginning of the spring semester 1981
(during which the course was taught) students were random
ly assigned to classes as outlined in the section on
sampling strategy. The course topic for all classes was
the same--''Cognitive Disability: Occupational Therapy
Assessment and Management." The subject matter consisted
of a theoretical approach, its application to treatment
in a clinical setting, and evaluation tools developed
90
under this framework. The topic comprised one section
of a year-long course in Psychiatric Occupational Therapy.
The period of instruction for this section continues
usually for three weeks, twice a week for an hour-and-a-
half each session; thus, six class periods are devoted
to this topic in the USC curriculum. Therefore, the same
/
schedule and amount of time were employed in this/study,
thus keeping the instructional period in its natural
parameters.
C3) In the week following the last session, three
instruments were administered to all students at the same
time, in the following order: a student evaluation ques
tionnaire, a multiple-choice exam, and two PMPs.
Social security numbers were used in all the
data collection; thus, confidentially of subjects was
maintained. In order to record a course grade which was
based on the score of the multiple-choice exam only, the
responsible faculty member received a list of numbers
with grades and he transferred them to the appropriate
names of students.
Treatment manipulation. The two teaching methods
dealt with the same topic, the same reading materials
were used, the same audio-visual aids were employed and
the same sequence in content was kept. The main differ
ence was in the mode of instruction, in the size of
91
classes, and the amount of individual preparation required
from students through homework assignments.
Two instructors participated in the study; both
of them were equally involved in preparation of course
materials, methods of instruction and in developing the
instruments to measure the outcome variables. The lecture
was given mostly by one instructor with the exception of
one session which was conducted by the other, while the
two small groups were conducted at the same time, in
paralell manner, by both instructors. In order to create
a control for differences among them, they changed the
groups in which they participated at each session.
The lecture method was conducted in a traditional
manner. The instructor presented the important information
during the session in a formal way with some occasional
audio-visual aids. Only reading assignments were required
outside class. The method was highly instructor-centered;
students had opportunity to ask questions during class,
but were not specifically encouraged to do so. They were
provided with objectives at the beginning of the course
and with some directions of areas emphasized in the
exams in the last session (the same objectives and direc
tions were given also to the group-discussion classes).
The group-discussion method was planned as a
student-centered group. Participants had to read assigned
92
materials before each session. In addition, they were
provided with case studies and questions to explore, as
well as a list of main issues to prepare and discuss
in class at the next session. No formal presentation
was given in class by the instructors; their role was
that of facilitator.
The groups met for the first half period (45
minutes) of every session without an instructor. The
instructor followed their discussion through a TV monitor
in the next room, and thus had first hand information
about the class conduct without interferring in the stu
dents ' discussion itself. This procedure enabled a more
free discussion in which students had to rely on them
selves to explore the material, while at the same time it
allowed the instructor to follow their progress and take
notes for the second half.
The instructor joined the group immediately
following the first 45 to 50 minutes, and group discussion
then continued according to the group needs or wishes.
The second half depended largely on what was achieved
in the preceding discussion and varied each session.
Students continued to clarify certain points, asked
questions in areas in which they were not sure or needed
better understanding. The instructor provided feedback,
corrected occasional mistakes or extended the discussion
of important issues. The main goal in instructor
9 3 .
participation was to insure that the main points of the
subject matter were conveyed correctly and understood
by the students. (In a short preview to the course in
which the materials for the first session were handed
out, the instructor role and the procedures of the method
were explained to the students.)
The group-discussion method followed in this
study had the advantage of a real student-centered group,
but at the same time it allowed the instructor's input
when necessary for students' understanding. The method
was partially adapted from Zimmerman and King (1963)
where it was employed successfully with medical students.
In their study the group had a break after the first half
in which the instructor viewed the taped session, and
then reconvened with the students for the second half.
The reason for it seems to be practical (less equipment
needed), as speculated on by this investigator, but the
difference does not seem to change the method drastically.
Data Analysis
The principal method used for the statistical
analysis of the data was multiple regression analysis
which is considered a more efficient and powerful approach
for detecting interactions (Cronbach & Snow, 1977), and
which allows the use of continuous and categorical vari
ables in the same analysis without loss of information
94 ^
(Kerlinger & Pedhazur, 1973). The procedure used was
the increment method stepwise regression, in which
successive increments of the variance are added, and
calculations of semipartial correlations for the least-
square solution with the interaction are entered last
into the equation. This procedure is also appropriate
with unequal cell frequencies, as stated by Kerlinger
and Pedhazur (1973): ". . .one notes the increment in
the proportion of variance due to the interaciton when
it is entered last in the analysis" (p. 190).
Following the ATI model, a stepwise forward
inclusion was performed in conjunction with a preestab
lished hierarchy among sets of variables: ". . . the
variable that explains the greatest amount of variance
unexplained by the variables already in the equation
enters the equation at each step" (Nie, 1975, p. 345).
For every outcome variable the stepwise regression was
performed in the following order:
(1); .4 Aptitude scores were entered first as a group:
prior achievement
verbal ability
perceived benefit from lecture
percevied benefit from group
(2) 2 Scores from the LSI were entered both second:
active/reflective composite score (AE-RO)
learning type
(3) Treatment, coded as (-1,+1) for the two teaching
methods, lecture and group-discussion, was entered
third.
95
(4) Aptitude-Treatment-Interaction
terms for each aptitude x treatment were entered
together in the fourth step (6 terms).
Using this procedure, a more precise and accurate
analysis of the amount of variance accounted for by the
interaction of individual learning style and teaching
method is possible. Whenever the multiple regression
indicated significant interaction, simple regression of
these variables in each method was performed (a between
group comparison using class means), and a graphic repre
sentation of the regression lines is presented in Chapter
IV.
There is one problem with the use of multiple
regression analysis in this study and that is the small
sample size (_n—44), which causes larger error terms of
the regression coeffients. Therefore additional analysis
was performed with 2-way ANOVA— (1) learning type by
teaching method, and (2) educational level by teaching .
method— as well as 3-way ANOVA--learning type by teaching
method by educational level. Educational level as an
independent variable was included for exploratory reasons.
The hierarchial approach for analyzing data when unequal
cell frequencies in the factorial design are present was
used (Nie, 19 75, p. 406).
The data were inspected mainly for trends and
therefore tables of the various cell means are presented
, 9 6' ,
in Chapter IV and discussed in Chapter V. For all vari^
ables, descriptive statistics, means, standard deviations,
and frequencies by treatment groups and by educational
levels are provided in Chapter IV. In addition, tables of
intercorrelation (Pearson r) among all variables are
presented in Appendix E.
Methodological Limitations
The main limitations of this study refer to
instrumentation, short time of the instructional period,
and small sample size. All the instruments used to measure
the dependent variables were new, instruetor-made measures.
Although they were pilot-tested and verified with a group
of experts, their measurement qualities were still less
than ideal, and therefore the findings and interpretations
based upon them should be viewed with caution as prelimi
nary indications or trends in an exploratory theory-testing
study.
Another limitation was the short duration of the
instructional period, as Cronbach and Snow (19 77 indicate
"• . .a period of habituation is probably necessary
before the student is working with full effectiveness"
(p. 44); thus, they recommend a period of at least ten
class sessions, although "from a formal point of view, an
ATI experiment can be of any duration" (p. 42)... Practical
97.
considerations did not allow for more time in the USC
curriculum but in a replication study it may be interesting
to test students after the same number of sessions, and
if possible extend the number of sessions and test them
again. In any case the short duration could have caused
the treatment manipulation to be less effective, especially
for the novel one— -the group-discussion.
The third main limitation was sample size.
Cronbach and Snow (19 77) (who are very strict in this
regard) recommend having 100 subjects in each treatment,
but most ATI research employed far fewer subjects for
obvious cost and practical reasons. The main reason
for their recommendation is the loss of power in the :
design which ". . . made Type-II errors highly probable.
That is to say, the hypothesis of no interaction has often
been accepted when an important interaction was present1 1
(p. 46). This may be the case also in this study and
therefore a lower significance value was used (.10) and
trends in the data were examined. But, Cronbach and Snow
(.1977) further assert that their 11. . . . . rule of thumb
can be weakened when especially powerful experimental
designs are used1 1 (p. 46) . They refer to blocking cases
before random assignments to treatments, and statistical
control of a variable that predicts outcome and factorial
designs, both of which were undertaken in this studyf and
therefore may increase confidence in the findings.
9 8
CHAPTER IV
RESULTS
The findings of the study are presented in this
chapter in a series of tables and figures. The accepted
level of significance for all statistical analysis was
set a p < .10 to allow for trends to show in a small sam
ple as utilized in this study (n=44).
Tables 7, 8, and 9 present descriptive statistics,
means and standard deviations for all variables in the
study. Table 7 presents data for the aptitude and inde
pendent variables. Means and standard deviations are pro
vided for continuous variables for each treatment, and
frequencies are given for categorical variables for each
treatment. No statistically significant differences (in
a t test) were found between the two treatment groups. In
Table 8 means and standard deviations of outcome variables,
achievement and problem-solving are presented for each
treatment group. Table 9 presents the same statistics for
outcome variables amount of study time and attitude toward
instruction. The only significant difference between the
two treatment groups was found on amount of study time for
every class session (t=-2.08, p=.05), which is mainly due
Table 7
Descriptive Statistics of Independent Variables, Means,
Standard Deviations and Frequencies for Each Treatment (n=44)
Variables Lecture (n=22) Group Discussion (n = 2 2)
Mean SD Mean SD t
Verbal Ability 65.09 23.16 64.41 27.99 NS
Prior Achievement (GPA in O.T.) 3.56 .24 3.41 .39 NS
Perceived Benefit from Lecture 6.09 .75 6.41 .59 NS
Perceived Benefit from Group 6.09 .81 5.82 .91 NS
Individual Learning Style
Active/Reflective (AE-RO) 2.59 6.34 5.64 5.96 NS
Learning Style Type:
Active/Accommodator and Converger
Reflective/Assimilator and Diverger
n=13
n= 9
n=12
n=10
Educational Level:
Seniors
Graduates
n=14
n= 8
n=16
n= 6
' i
KD
v o
Table 8
Means and Standard Deviations of Outcome
Variables, Achievement and Problem-Solving,
for Each Treatment (n='44)
Variables Lecture (n = 2 2) Group-Discussion (n=22)
Mean SD Mean SD t
Achievement 29.05 2.61 28.59 4.39 NS
Problem-Solving on PMP:
first case Proficiency
Efficiency
50.05
66.59
21.18
10.44
50.86
67.68
20.40
13.84
NS
NS
Overall Competence 42.59 19.86 43.68 20.25 NS
second case Proficiency 53.59 30.13 44.14 24.06 NS
Efficiency 63.82 20.21 58.27 20.37 NS
Overall Competence 46.27 28.39 36.68 22.70 NS
Overall Course Score
(combined z-score of achievement
and overall competence of both PMP's) 0 1.88 0 2.05 NS
H
O
O
Table 9
Means and Standard Deviations of Outcome
Variables, Study Time and Attitude Toward Instruction,
for Each Treatment (n=44)
Variables Lecture (n=22) Group-Discussion (n = 2 2)
Mean SD Mean SD t
Amount of Study Time Outside Class
For Every Class Session 2.73 .98 3.32 .89 -2.08*
For Exams 2.64 1.09 2.41 1.14 NS
Attitude Toward Instruction
Benefit 4.04 .72 3.95 .99 NS
Enjoyment 3.86 .99 3.50 1.34 NS
Confidence 3.54 .96 3.32 1.13 NS
Overall Attitude (combined score) 3.82 .76 3.59 1.02 NS
Note. ap=.05
H
o
________________________________________ i —1
102
to the way the teaching methods were constructed. This
result indicates that at least practically the method func
tioned as it was expected.
The next seven tables (10-16) and their respective
figures (5-14) provide summaries of stepwise regression
analysis for each outcome variable (two in a table). For
every significant interaction result, a figure showing the
ATI is presented on the following page.
In Tables 10-13, significant main effect of prior
achievement (GPA in O.T.) is indicated in relation to
achievement and problem-solving. The overall course score
shows the cumulative effect (F=21.07, p <.01) and GPA
accounts for 3 3.4 percent of its variance. When taken
separately, prior achievement accounts for more variance in
2
achievement (R % = 26.84, F = 5.41, p < .01) then m the
2
PMP1s (R % = 5.0% to 11%, all of which are significant at
the .05 or .10 levels).
Another variable which showed significant main
effect is perceived benefit from small group, but only for
achievement (F = 6.85, p < .01) and the second PMP (PMP II
OC, F = 2.32, p < .10; PMP II P, F = 4.04, p < .05; PMP II
E, F = 2.33, p < .10).
Inspection of the interaction terms show some sig
nificant ATI's with the cognitive outcomes. In Table 10
an interaction with perceived benefit from lecture was
10 3
Table 10
Summary of Stepwise Regression Analysis on
Overall Course Score (z) and Achievement
Variables in
Regression df
Overall
Score^s
R2%
Course
d
b F
Achievement3
R2% b F‘
Aptitude 6
Prior Achievement (GPA) 1 33.4 3.77 21.07*** 26.84 6.91 5.41***
Verbal Ability
Perceived Benefit
1 2.0 -.19 1.29 .7 -.27 <1
from Lecture
Perceived Benefit
1 .6 -.22 <1 .1 -.95 <1
from Small Group
Individual Learning
Style Active/
1 .4 .35 <1 10.46 1.46 6.85***
Reflective (AE-RO) 1 .6 .53 <1 .6 .11 <1
Learning Style Type 1 .8 -.52 <1 2.1 -.92 1.23
Treatment
Aptitude Treatment
Interactions
1
6
0 4.3 <1 .3 - -11.3 <1
GPA x Treatment 1 .2 -.62 <1 .2 -.84 <1
Ability x Treatment
Perceived Benefit
from Lecture x
1 2.3 .12 1.51 .9 .15 <1
Treatment
Perceived Benefit from
1 8.1 1.05 5.20*** 12.17 2.36 9.16***
Group x Treatment 1 .4 -.20 <1 .3 -.40 <1
(AE-RO) x Treatment
Learning Type x
1 1.3 .89 <1 3.1 .13 <1
Treatment
Full Model
Residual
1
13
30
1.0
51.4
48.6
-.62 <1 0
58.11
41.89
-.19 <1
Note. Variables entered the regression not necessarily in the fixed
order presented in all tables.
R2%=percent variance account for; b=unstandardized partial regression
coefficient adjusted for all other terms in the regression; F=F ratio
adjusted for all variables preceding in the regression
^ constant3 -12.36
c
constant= 4.17
*£ < .10
**£ < .05
***p < .01
104
found (overall score: F = 5.20, p < .01, and achievement:
F = 9.16, p < .01). These relationships are depicted in
Figure 6 and indicate a disordinal interaction: The less a
student perceived to benefit from lecture the higher his/
her achievement in the group method, and the lower in the
lecture method.
In Tables 11 and 12, ATI's for the first PMP and
learning type are indicated and presented in Figure Q (PMP
I OC, F = 2.42, p < .05; PMP I P, F = 1.87, p < .10; PMP I
E, F = 2.35, p < .05). Figure 7 presents a graphic repre
sentation of a regression on a categorical variable coded
as -1 and +1. The same relationship will be referred to in
Table 19 (a 2-way ANOVA). The interaction shows 4 sub
groups: first, the reflective-type students who perform
highest in the lecture method; second, the active-type stu
dents in the group method; third, the reflective-type stu
dents in group method; and fourth, the active-type students
in lecture method. The two lower subgroups change places
in Figure 7 but the relationship for the two higher achiev
ing subgroups is consistent. These two subgroups are the
match groups between instructional method and individual
learning style presented in the study's design and hypo
thesized to score higher. In all three figures, the match
between the lecture and the reflective type seems to result
in the highest performance which is in contrast to the
Summary of Stepwise
Table 11 105
Regression Analysis on PMPs Overall Competence
PMP: I oc a,d
PMP 11 OC a»
c
Variables in
(first case)
(second case.)
Regression df R H b F R % b F
Aptitudes 6
Prior Achievement
(GPA) 1 11.3 14.3 5.37** 8.9 29.48 4.11**
Verbal Ability 1 0 -.14 <1 2.6 -.10 1.25
Perceived Benefit
from Lecture 1 .7 .22 <1 . 1 .63 <1
Perceived Benefit
from Small Group 1 .4 2.16 <1 4.86 -3.9 2.32*
Individual Learning
Style Active/
Reflective (AE-RO) 1 2,6 1.28 1.16 1.3 -1.1 <1
Learning Style Type 1 2.1 -9.1 <1 1.5 4.9 <1
Treatment 1 0 1.28 <1 .9 -118.1 <1
Aptitude-Treatment
Interactions 6
GPA x Treatment 1 3.2 -13.4 1.48 1.7 7.94 <1
Ability x Treatment 1 0 .24 <1 1.8 .19 <1
Perceived Benefit
from Lecture x
Treatment 1 .4 2.58 <1 1.3 6.75 <1
Perceived Benefit
from Group x
Treatment 1 2.8 2.16 1.28 .9 5.51 <1
(AE-RO) x Treatment 1 0 - <1 .2 1.22 <1
Learning Type x
Treatment 1 5.3 -5.5 2.42** 1.7 -7.1 <1
Full Model
13
29.35 28.14
Residual 30 70.65 71.86
Note. *p < .10
**£ < .05
***p < .01
constant = -13.04, constant = -29 .11
Variables entered the regression not necessarily in the fixed order
presented in all tables .
R %=percent variance account for; b=unstandardized partial regression
coefficient adjusted'for all other terms in the regression;-F=F— ratio
adjusted for all variables preceding in the-regression.
* - —-
106
Table 12
Summary of Stepwise Regression Analysis on
PMP Proficiency and Efficiency (first case)
Variables in
Regression df
PMP I
R2 %
P a-,d
b F
PMP I
R2 %
E
b F
Aptitudes
Prior Achievement (GPA)
6
1 10.5 13.28 4.95*** ‘ 9.6 11.56 4.47**
Verbal Ability 1 0 -.14 <1 .2 -.96 <1
Perceived Benefit
from Lecture 1 .3 -1.36 <1 3.3 3.16 1.55
Perceived Benefit
from Small Group 1 .2 1.38 <1 1.9 2.52 <1
Individual Learning
Style Active/
Reflective (AE-RO) 1 2.3 1.28 1.01 3.9 .86 1.83
Learning Style Type 1 2.9 -9.8 1.33 .4 -4.70 <1
Treatment 1 0 56.8 <1 .1 44.71 <1
Aptitude Treatment
Interactions
GPA x Treatment
6
1 2.8 -12.5 1.28 1.3 -6.04 <1
Ability x Treatment 1 0 .31 <1 0 - <1
Perceived Benefit from
Lecture x Treatment 1 .8 3.56 <1 0 -.47 <1
Perceived Benefit from
Group x Treatment 1 3.0 -6.0 1.34 2.2 -3.34 1.0
(AE-RO) x Treatment 1 0 - <1 0 - <1
Learning Type x
Treatment 1 4.2 -4.96 1.87* 5.0 -3.31 2.35**
Full Model
Residual
13
30
27.43
72.57
28.21
71.79
Note. *£ < .10
**£ < .05
< #oi
Q — Q
constant = 12.30, constant = -3.67
Variables entered the regression not necessarily in the fixed order
presented in all tables.
£ 2 c # # • •
R %=percent variance account for; b=unstandardized partial regression
coefficient adjusted for all other terms in the regression; F=F ratio
adjusted for all variables preceding in the regression.
Overall
Course
Score (z)
+3
+2
+ 1
0
'•j.
7t1
-2
-3
\e
1 2 3 4 3 6 77
’ -Perceived Benefit from Lecture
Figure 5
Regression of Overall Course Score on Perceived
Benefit from Lecture Showing ATI
40 -
30 - •
Achievement
Lecture TT^T
20 -■
10 -•
Perceived Benefit from Lecture
Figure 6
Regression of Achievement on Perceived
Benefir from Lecture Showing ATI
PMPIQC
109
(1.54)
Groups-type AE
Groups-type R0
Eecture-
(a)
-1
RO
TT
AE
TYPE
Figure 7
Regression of PMP first case Overall Competence,
Efficiency and Proficiency on Learning
Style Type Showing ATI
110
Group-type AE (.31)
PMPIE
-1 + 1
RO AE
TYPE
Figure 7
J
111
100
80
60
(,971
•type AE
Lecture-tvpp a f ( _ 8 < 9 3 )
s-type Ro
20
-1 +1
R0 AE
TYPE
Figure 7
112
hypothesis related to problem-solving which predicted
higher scores for the active type in the group method.
One additional interaction for cognitive outcome
was found and it is presented in Table 13 between the
second PMP efficiency score and verbal ability (F = 4.12,
p < .01), while the proficiency and overall competence
scores show only a similar trend. The relationship in
Figure 8 indicates a disordinal interaction in which high-
ability students perform better in the lecture and lower
in group while the opposite holds for less verbally able
students. Inspection of regressions of the other cognitive
variables with verbal ability shows an opposite trend (even
though nonsignificant) in which more verbally able stu
dents do better in the group method and less verbally able
to better in lecture.
Table 14 provides data for two time variables:
amount of study time outside class for every session, and
for exams. For both outcomes a significant main effect on
verbal ability was found (F = 9.46, p < .01; F = 6.8, p <
.01 respectively): the higher the verbal ability, the
less time a student reported to study for every session
and for exams, regardless of teaching method.
In addition, perceived benefit from lecture showed
main effect as well as interaction with amount of study
time for every session (F = 10.23, p < .01? F = 1.93, p <
.10, respectively). The ordinal interaction presented in
113
Table 13
Summary of Stpewise Regression Analysis on
PMP Proficiency and Efficiency (second case)
Variables in
Regress ion df
PMP II
r2%
P a»d
b F
PMP II
r2%
E a,c
b F
Aptitudes t-6
Prior Achievement (GPA) 1 5.0 28.9 2.39* 5.4 31.15 2.56*
Verbal Ability
Perceived Benefit
1 1.9 -.89 <1 4.1 -.18 1.89
from Lecture
Perceived Benefit
1 0 .63 <1 .1 -1.79 <1
from Small Group
Individual Learning
Style Active/
1 8.7 -5.03 4.04** 5.2 -2.62 2 .33*
Reflective (AE-RO) 1 1.3 -1.23 <1 1.5 -1.53 <1
Learning Style Type 1 1.4 5.24 <1 4.1 10.33 1.93
Treatment
Aptitude-Treatment
Interactions
1
6
.7 -107.7 <1 1.2 -79.2 <1
GPA x Treatment 1 .2 5.90 <1 1.2 9.44 <1
Ability x Treatment
Perceived Benefit from
1 1.7 .19 <1 8.2 .27 4.12***
Lecture x Treatment
Perceived Benefit from
1 1.0 1.64 <1 1.1 3.85 <1
Group x Treatment 1 1.1 5.00 <1 0 - <1
(AE-RO) x Treatment
Learning Type x
1 3.3 1.64 1.48 2.1 .57 <1
Treatment
Full Model
Residual
1
13
30
.7
27.60
72.40
-9.75 <1 0
34.55
65.45
<1
Note. *p < .10
**p < .05
^***p < .01
constant = -12.94, constant = -3.28
Variables entered the regression not necessarily in the fixed order
presented in all tables.
S L 0 o • •
R %=percent variance account for; b=unstandardized partial regression
coefficient adjusted for all other terms in the regression; F=F ratio
adjusted for all variables preceding in the regression.
114 ,
Table 14
Summary of Stepwise Regression Analysis on
Study Time for Every Class and Exams
Variables in
Regression df
For Every Class a»
R2% b F
For Exams a
R2% b F
Aptitudes
Prior Achievement (GPA)
6
1 2.2 -.56 1.4 1.9 -.21 <1
Verbal Ability 1 15.0 -.11 9.46***13.9 -.16 6.8***
Perceived Benefit
from Lecture 1 19.59 .39
1
10.23*** 4.0 .43 2.06
Perceived Benefit
from Small Group 1 .2 -.77 <1 3.7 .10 1.85
Individual Learning
Style Active/
Reflective (AE-RO) 1 .5 t .27 <1 1.5 .71 <1
Learning Style Type 1 2.1 -.61 1.29 .3 -.35 <1
Treatment 1 3.5 2.10 2.22* 3.7 3.53 1.88*
Aptitude-Treatment
Interactions
GPA x Treatment
6
1 .1 -.21 <1 1.0 -.47 <1
Ability x Treatment 1 0 -.88 <1 1.5 -.89 <1
Perceived Benefit from
Lecture x Treatment 1 2.8 -.57 1.93* 0 .53 <1
Perceived Benefit from
Group x Treatment 1 .6 -.15 <1 2.1 -.28 1.1
(AE-RO) x Treatment 1 3.7 .71 2.44** 1.9 .10 <1
Learning Type x
Treatment 1 .8 -.61 <1 4.1 -.51 2.12**
Full Model
Residual
13
30
51.65
48.35
39.93
60.07
Note . *2. < •
**p < .05
.01
d c
constant = 3.97, constant = 1.06
Variables entered the regression not necessarily in the fixed order
presented in all tables.
cl 2 o • • • •
R %=percent variance account for; b=unstandardized partial regression
coefficient adjusted for all other terms in the regression; F=F ratio
adjusted for all variables preceding in the regression.
115/
PMPIIE
100
80
60
40
20
40 60 80 100 120 20
Verbal Ability
Figure 8
Regression of PMP Efficiency (Second Case)
on Verbal Ability Showing ATI
Note, PMPIP and PMPIOC both show the same trend,
116
Overall
Course
Score
+2
+ 1
Gr oup s-me an
typ?
(-.65)
-1
-1 + 1
TYPE
Figure 9
Regression of Overall Course Score on Learning Style Type
Note. ATI Not statistically significant
117
100
80
I
PMPIIOC 60
- ^ *
^ — - —
40
Groups-type R O -----(a)
Groups-type AE
20
/v.
0
-1 +1
TYPE
Figure 10
Regression of PMP Overall Competence (Second Case)
Proficiency and Efficiency on
Learning Style Type
Note. ATI no statistical significance
118
PMPIIP
100
Vec
80
60
Lecture-type AE
Group-type AE
40
20
0
-1 + 1
TYPE
Figure 10
119/
PMPIIE
100
80
Lecture mean
60
Groups-type AE
40
20
+ 1
TYPE
Figure 10
(l20>
Figure 11 indicates that students in the group overall
studied more time for every session (as indicated also pre
viously by the significant t test), but the interaction
shows that the students in the lecture who had low percep- j
tion of benefit studied significantly less than those who
had high perception of benefit from the method (treatment
main effect: F = 2.22, F = 1.88, p < .10).
The other two significant interactions with study
time relate to individual learning style and type (F = 2.44,
p < .05; F = 2.12, p < .05). As shown in Figure 12, the
ordinal interaction indicates that reflective learning style
students studied more time in the group and less in the
lecture, while active students studied almost the same
amount of time for every class. (The same relationships
are shown in the second figure with learning type.) Figure
13 shows a disordinal interaction with study time for exams
in which active learning style students in the lecture
studied more time for exams than those in groups, and
reflective students in groups needed more time to study
than those in lecture. These relationships are considered
also in Table*20 (a 2-way ANOVA) using treatment and learn
ing type as independent variables.
Tables 15 and 16 provide data for the outcome vari
ables attitude toward instruction. For all four analyses,
verbal ability and perceived benefit from lecture show
significant main effect at the .05 or .10 level; the higher
121
Table 15
Summary of Stepwise Regression Analysis on Attitude
Toward Instruction-Benefit and Confidence
Variables in
Regression df
Attitude-
R2% b
Benefit
F
a, d
Attitude
R2% b
-Confidence3
F
Aptitudes
Prior Achievement (GPA)
6
1 5.39 .92 2.
89*
3.2 .74 1.5
Verbal Ability 1 10.0 .15 4.67** 5.2 .10 2.89*
Perceived Benefit
from Lecture 1 10.0 .70 5. 13** 5.4 .51 2.48*
Perceived Benefit
from Small Group 1 1.7 .13 .3 . 18 <1
Individual Learning
Style Active/
Reflective (AE-RO) 1 0 .16 <1 .1 .10 1.6
Learning Style Type 1 0 .75 <1 0 .43 <1
Treatment 1 .2 -1 .36 <1 1.45 -1 .0 <1
Aptitude-Treatment
Interactions
GPA x Treatment
6
1 1.7 .59 <1 .4 .42 <1
Ability x Treatment 1.6 .50 <1 1.9 .71 <1
Perceived Benefit from
Lecture x Treatment 1 4.0 .24 2.04* 1.4 .22 <1
Perceived Benefit from
Group x Treatment 1 .8 .19 <1 .4 .11 <1
(AE-RO) x Treatment 1 . 1 . 13 <1 2.0 .35 <1
Learning Type x
Treatment 1 .4 .13 <1 0 .48 <1
Full Model
Residual
13
30
36.29
63.71
23.44
76.56
Note. *p < .10
**£ < *05
***p < .01
d — r
constant = -5.47, constant = -3.33
Variables entered the regression not necessarily in the fixed order
presented in all tables.
C I 2 « • • •
R %=percent variance account for; b=unstandardized partial regression
coefficient adjusted for all other terms in the regression; F=F ratio
adjusted for all other variables preceding in the regression.
122
Table 16
Summary of Stpewise Regression Analysis on Attitude Toward
Instruetion-Enjoyment and Overall Score
Variables in
Regression df
At t i tude-En j oymenta* ^
R 2Z b F
Attitude
R2% b
-Overall Sc
F
Aptitudes 6
Prior Achievement (GPA) 1 3.0 .93 1.57 4.8 .87 2.57*
Verbal Ability
Perceived Benefit
1 8.7 .20 4.00** 10.1 .15 4.72**
from Lecture
Perceived Benefit
1 8.0 .81 3.95*'* 9.9 .67 5.02**
from Small Group
Individual Learning
Style Active/
1 1.8 .75 <1 .5 .79 <1
Reflective (AE-RO) 1 .2 .60 <1 .3 .34 <1
Learning Style Type 1 .1 .96 <1 .1 .56 <1
Treatment
Aptitude-Treatment
Interactions
1
6
3.0 .99 1.45 1.88 -.53 <1
GPA x Treatment 1 1.3 .51 <1 1.0 .51 <1
Ability x Treatment
Perceived Benefit from
1 — — —
0 .87 <1
Lecture x Treatment
Perceived Benefit from
1 6.7 -.42 3.46**** 5.3 -.29 2.78**
Group x Treatment 1
- - -
.5 .11 <1
(AE-RO) x Treatment
Learning Type x
1 1.6 .30 <1 0 .53 <1
Treatment
Full Model
Residual
1
13
30
34.67
65.33
0
34.93
65.07
.54 <1
Note. *p < .10
**p < .05
-k-k~kp < ^oi
Q Q
constant = -6.54, constant = -5.16
Variables entered the regression not necessarily in the fixed order
presented in all tables.
cl 2 c r • • •
R %=percent variance account for; b=unstandardized partial regression
coefficient adjusted for all other terms in the regression; F=F ratio
adjusted for all variables preceding in the regression.
Groups (.97)
Study
Time for
Every
Class
Perceived Benefit from Lecture
Figure 11
Regression of Study Time for Every Class
Session on Perceived Benefit from
Lecture Showing ATI
124
Study
Time
for
Every
Class
2
1
U I 2 3 5 5 6 17
R0
Individual Learning Style
(AE-RO)
Gr°ups
Lecture (.11)
Figure 12
Regression of Study Time for Every Class
Session on Individual Learning Style
(AE-RO) and Type Showing ATI
125
Study
Time
for
Every
Class
5
'----
Lecture-tyPe &
4
3
2
1
-1
+ 1
TYPE
(b)
Figure 12
Lecture (.68)
Study
Time
For
Exams
17
ae:
-10
RO
Individual Learning Style
(AE-RO)
Figure 13
Regression of Study Time for Exams on Individual
Learning Style (AE-RO) and Type Showing ATI
127.
Study
Time
For
Exams
4
RO
Lecture-type AE
3
Groups-type AE
(-.26)
2
1
+.1
AE
RO TYPE
(b)
Figure 13
128
the attitude in both methods. The same trend seems to
occur for GPA; especially significant were effects upon
overall scores and benefit scores.
Perceived benefit from lecture shows also interac
tions with three of the outcomes, except for confidence
(benefit: F = 2.04, p < : .10; enjoyment: F = 3.46, p < .01;.
overall score: F = 2.78, p < .05). in Figures 14a, b, c,
the ordinal interactions seem- to indicate that high percep
tion of benefit from lecture is associated with high (or
favorablejL attitude in both methods (an unclear result for
those in groups), but on the lower end, group people show
a significantly less favorable than lecture people espe- '
cially for the variable of enjoyment (a finding which needs
more discussion in Chapter V)r.
Tables 17 and 18 present regression equations for
outcome variables showing the unstandardized partial regres
sion coefficients (b) and constant (a) for all aptitudes
in each treatment.
I.n addition to regression analysis, the data were
analyzed using 2-way Analysis of Variance, treatment by
learning-style-type, as shown in Tables 19 and 20. Table
19 presents, the means: for all cognitive outcomes, achieve
ment and problem-solving in the four cells. The only sig^
ni.ficant result was found for the overall competence score
of the first PMP (F = 2.9.14, p = .09). . But, inspection of
129
Table 17
Regression Equations for Outcomes Showing the
Unstandardized Partial Regression Coefficients
for the Aptitudes in Each Treatment Group
Outcomes
and
Treatment Constant
b for
GPA
b for
Verbal
Ability
b for
Perceived
Benefit
Lecture
b for
Perceived
Benefit
Groups
b for
(AE-RO)
b for
Type'
Achievement
Lecture -7.18 6.04 -.12 1.14 1.05 .25 -1.11
Groups 15.43 7.73 -.42 -3.31 1.86 -.20 -.82
PMP I P
Lecture 72.20 .46 -.11 2.20 -4.80 1.37 -15.25
Groups -44.04 26.06 -.17 -5.01 7.40 1.17 -4.35
PMP I E
Lecture 39.04 5.91 -.95 2.69 -.72 . 81 -7.72
Groups -48.72 17.51 -.98 3.68 5.89 .93 -1.69
PMP I OC
Lecture 55.44 .31 -.12 2.80 -3.93 1.35 -15.08
Groups -78.79 27.88 -.16 -2.43 8.13 1.20 -3.27
PMP II P
Lecture -121.06 34.81 .10 7.11 .29 .41 -4.51
Groups 94.77 23.00 -.28 -5.84 -10.03 -2.88 15.00
PMP I E
Lecture -81.02 40.23 .88 2.06 -2.66 -.93 10.69
Groups 75.58 21.81 -.45 -5.62 -2.61 -2.14 10.59
PMP II OC
Lecture -150.19 38.04 .91 7.68 1.38 .82 -1.69 )
Groups 89.05 21.54 -.29 -5.82 9.78 -2.37 12.05
Overall
Course
Score
Lecture -20.15 3.66 -.76 .95 .25 .16 -1.26
Groups -6.74 4.07 -.31 -1.12 .40 .48 .18
Note. The coefficients reflect adjustments for all other terms in the
regression.
130
Table 18
Regression Equations for Outcomes Showing the
Unstandardized Partial Regression Coefficients
for the Aptitudes in Each Treatment Group
Outcomes
and
Treatment Constant
b for
GPA
b for b for
b for Perceived Perceived
Verbal Benefit Benefit
Ability Lecture Groups
b for
(AE-RO)
b for
Type
Study Time
for Every
Class
Lecture 6.08 -.78 -.20 .33 -.22 .44 -.27
Groups 1.88 -.35 -.23 .45 .73 -.98 .14
Study Time
for Exams
Lecture 4.59 -.68 -.25 .48 -.18 .18 -.86
Groups -2.47 .26 -.77 .37 .38 -.34 .15
Attitude-
Benefit
Lecture -6.83 1.52 .10 .45 .33 -.29 .20
Groups -4.13 .32 .20 .94 -.66 .32 -.72
Attitude-
Enjoyment
Lecture -5.91 1.46 .20 .39 .10 .37 .79
Groups -7.45 .43 .19 1.24 .66 -.27 .11
Attitude-
Confidence
Lecture -4.35 1.17 .18 .29 .13 -.34 .44
Groups -2.40 .34 .35 .73 -.95 .28 .52
Attitude
Overall
Lecture -5.39 1.38 .16 .38 .19 -.87 .11
Groups -4.55 .36 .14 .97 -.33 .19 .27
Note. The coefficients reflect adjustments for all other terms in the
regression.
- 131
Table 19
' 2-Way Analysis of Variance, Treatment by
Learning Style Type on Achievement and Problem-Solving
Learning Style Type
. Reflective (RO)
Divergent & Assimilator
Active (AE)
Accommodator & Converger
(n=9) (n=13)
Achievement 29.11(+) 29.00
PMP I P 60.67(+) 42.69(-)
N
PMP I E 71.22(+) 63.68(-)‘
4->
O PMP I OC 52.78(+) 35.54(-)
a PMP II P 61.00 (+) 48.46
PMP II E 63.56 64.00(+)
PMP II OC 52.78(+) 41.77
4-1
Course score (z) 1.00(+) -.32 (-)
IS
(n=10) (n=12)
£
Achievement 29.00 28.25(-)
PMP I P 49.80 51.75
PMP I E 66.80 68.42
PMP I OC 42.00 45.08
PMP II P 42.70(-) 45.33
w PMP II E 61.40 55.67(-)
p - f
a
PMP II OC 36.00(-) 37.25
Course score (z) -.22 -.22
Note. Numbers in cells are means for the respective outcomes
variables.
(+) = highest score
(-) = lowest score
132
Table 20
2-Way Analysis of Variance, Treatment
by Learning Style Type on Study Time and Attitude Toward
Instruction
Learning Style Type
Active (AE)
Accommodator & Converger
Reflective (RO)
Diverger & Ass imi labor
(n=13) (n=9)
Study Time:
for every class
for exams
2.67(-)
2.44
2.87
2.70
Attitude:
4.11(+)
3.78
3.89(b)
Benefit
Enjoyment
Confidence
4.00
3.92(-)
3.31
Overall Attitude 3.93(b) 3.74
(n=10 (n=12)
Study Time:
3.90(b)
2.77(b)
for every class
for exams
2.83
2.17(-)
Attitude:
3.92(-)
3.42(-)
3.56(-)
Benefit
Enjoyment
Confidence
4.00
3.60
3.30(-)
Note. Number in cells are means for the respective outcomes variables.
Study time for every class— Treatment main effect F=4.976, p=.03
Type main effect Fj=3.344, p=.07
Interaction effect F=4.769, p=.03
133
Attitude
Benefit
5
4
3
2
1
7 6 5 4 3 1 2
Perceived Benefit from Lecture
Figure 14
Regression of Attitude Toward Instruction, Benefit
and Enjoyment on Perceived Benefit from
Lecture, Showing ATI
134
Attitude—
Enjoyment
5
4
3
2
1
(b)
Perceived Benefit from Lecture
Figure 14
Attitude
Overall
Score
136
- S e ’ c'^°"C
Study Time
for Every
Class
Lecture-Senior
Lecture-Grad
+ 1
Grad
-1
Senior
EDUCATION
Figure 15
Regression of Study Time for Every Class and
Exams on Educational Level
Note. ANOVA significant treatment effect F=3.829 £=.05
137
Study Time
For Exams
4
3
2
1
+ 1 -1
^Senior Grad
EDUCATION
(b)
Figure 15
Note. ANOVA significant education effect F=3.99 p <.01
138
5
4
3
Attitude
Benefit
2
----Groups-Senior
1
-1 +1
Senior Grad
EDUCATION
Figure 16
Regression of Attitude Toward Ijistruction-^Benefit
on Educational Level
Note, ANOyA significant interaction F=3.745 p?=,06
13 9,
the means show some important trends which are in accord
ance with the regression results and are worth consider
ing. All the highest scores indicated by (+), except for
PMP II E, are found in cell I (reflective-type-in-lecture).
The ranking in the other cells is not always consistent
but many of the scores rank second in cell IV (active-type-
in group). More of the lowest scores (-) seem to exist in
cells II and III (active-type-in-lecture and reflective-
type-in-group). The same relationship was presented in
Figure 7 for the first PMP and is shown also in Figure 10.
These trends are also consistent with the first hypothesis
that a match between individual learning style and the
teaching method will result in higher performance, but, is
in contrast to the second hypothesis predicting that the
match of active-type-in-groups will perform the highest.
Table 20 provides the same analysis for the time
outcomes and attitude. Significant results were found for
amount of study time for every class (treatment effect:
F = 4.976, p = .03; type effect: F = 3.444, p = .07; inter
action effect: F = 4.769, p = .03). Scheffe's test of
contrasts shows that cell III (reflective-type-in-group) is
significantly higher than each of the other cells (F =
12.13, p < .05); this group reported studying much more
time for every class session.
14 0.
In contrast, the reflective-type-in-lecture reported
the least amount of time, but as seen in Table 19 performed
the highest. (The same relationship is presented in Figure
13.)
The pattern for study time for exams is the same
for the highest value, but the least amount of time was
reported by the active-type-in-group, presumably because
they were involved more with the method which fitted them
all along (Figure 14).
The trend for the outcomes of attitude toward
instruction indicates higher attitude for reflective-type-
in-lecture except in the area of enjoyment, which is high
est for active-in-lecture, while the lowest scores seem to
occur in cell IV for active in groups.
Iri order to explore the relationship between atti
tude and the cognitive outcomes— achievement and problem-
solving--directly, as suggested by Strom and Hocevar (1981),
Pearson r correlations and simple regressions for each
treatment were performed using the attitude variables as
predictors.
For the achievement outcome, significant positive
correlations were found in the lecture method (r = .40
benefit, .20 enjoyment, .31 confidence, and .34 overall
attitude), while no correlation was found in the group-
discussion method. In contrast, correlations with the PMP
scores were found to be significantly positive in the group
141'
method (overall attitude with PMPIOC: r = .27, and with
PMPIIOC: r = .36, F = 3.05, p < .10) (see other correla
tions in Appendix E), while almost no correlations or very
low ones were found in the lecture method. Inspecting the
regression lines, it seems that an ordinal-like interaction
is indicated, where students with high attitude do almost
as well in both methods, but those with low attitude per
form much lower in the group-discussion method. When the
separate attitude scores were considered, significant F
values were found for the confidence variable (PMPIP: F
= 3.00, PMPIE: F = 3.67, p < .10; PMPIIP: F = 6.66;
PMPIIOC: F =*5.98, p < .05) in the group-discussion method
(correlation ranging mostly between .35 to .49), while in
the lecture the highest correlations are with the benefit
variable. This last trend is evident in the combined
score for overall course score: students in the group-
discussion method seem to be more confident (r = .40, F =
4.03) while those in lecture seem to consider it more bene
ficial (r = .36, F = 3.88) . Overall attitude and the
course score correlate positively in both methods (lecture:
r = .30; group discussion: r = .23).
In summary, it seems that attitude plays a signi
ficant role in the lecture method for the achievement out
come, and in the group-discussion method for the problem
solving outcome. In addition, lecture students score higher
on benefit, and group students, on confidence.
142
To explore the influence of educational level
(seniors vs. graduate students) on the various outcome
measures, a further data analysis was performed. Tables
21, 22, and 23 present descriptive statistics, means,
standard deviations, and frequencies for all aptitudes and
independent variables as well as all outcome variables.
In addition, significant t-tests are shown in Table 21 for
verbal ability (t = 2.33, p < .0 5), prior achievement (t =
-3.34, p < .01), individual learning style (t = -1.96, p
< .05). The direction indicated shows that graduates have
higher scores on verbal ability, have higher GPA, and are
of the more active type.
In Table 23, significant t-tests exist for both
time variables (t = 1.89, t = 2.06, p < .10). Even though
this relationship is weaker, it indicates that graduate
students— -regardless of method--reported less time spent
on studying. In Table 24, these same results are shown
in a 2-way ANOVA (education by treatment), but here the
additional information about the method points out that
the seniors in groups reported more study time for every
class than any other subgroup (Scheffe's test for con
trasts: F = 6.79, p < .10). Study time for exams is
reported highest by seniors in lecture.
In the same table (24), the results for the atti
tude variables indicate that graduate students in groups
Table 21
Descriptive Statistics of Independent Variables,
Means, Standard Deviations, and Frequencies for Two Educational Levels
Seniors (n=30) Graduates (n=14)
Variables Mean SD Mean SD t
Verbal Ability 58.6 22.55 77.93 26.89 -2.33**
Prior Achievement (GPA) 3.39 .33 3.68 .24 -3.34***
Perceived Benefit from Lecture 6.33 .66 6.07 .73 NS
Perceived Benefit from Group 5.97 .81 5.93 .99 NS
Individual Learning Style
Active/Reflective (AE-RO) 3.03 6.73 6.43 4.55 -1.96**
Learning Style Type:
Active/Accommodator and Converger
Reflective/Assimilator and Diverger
n=15
n=15
n=
n=
10
4
Treatment: Lecture n=14 n= 8
Groups n=16 n= 6
Note. *p < .10
**p < .05
***p < .01
— t-1
Ui
Table 22
Means and Standard Deviations of Outcome Variables,
Achievement and Problem-Solving for Two Educational Levels
Variables
Seniors
Mean
(n=30)
SD
Graduates (n=14)
Mean SD t
Achievement 28.36 3.63 29.78 3.38 NS
Problem-Solving
first case
on PMP:
Proficiency 49.76 21.98 51.93 17.66 NS
Efficiency 66.23 12.33 69.07 11.91 NS
Overall Competence 42.40 21.13 44.71 17.35 NS
second case Proficiency 45.97 28.52 55.07 24.53 NS
Efficiency 61.17 19.58 60.78 22.39 NS
Overall Competence 39.17 27.15 46.43 23.01 NS
Overall Course Score
(combined z-score of achievement
and overall competence on both PMP's) -.25 2.09 .54 1.71 NS
H
i t *
Table 23
Means and Standard Deviations of Outcome Variables,
Study Time and Attitude Toward Instruction, for Two Educational Levels
Variables
Seniors
Mean
(n=30)
SD
Graduates (n=14)
Mean SD t
Amount of Study Time Outside Class:
For every class session 3.16 1.12 2.71 .47 1.89*
For exams 2.73 1.14 2.07 .92 2.06**
Attitude Toward Instruction:
Benefit 3.87 .82 4.28 1.71 NS
Enjoyment 3.57 1.19 3.93 1.14 NS
Confidence 3.37 1.03 3.57 1.09 NS
Overall Attitude 3.60 .85 3.93 .98 NS
Note, *p < .10
**p < .05
! Ln
146
Table 24
2-Way Analysis of Variance, Treatment by
Educational Level on Study Time and Attitude Toward Instruction
Educational Level
Seniors Graduates
(n=14) (n=8)
Study Time:
for every class
for exams
2.63(-)
2.38
2.79
2.79(+)
Attitude:
Benefit 4.00 4.07
Enjoyment 3.75
Confidence 3.50 3.57
Overall Attitude 3.86 3.75
(n=16) (n=6)
Study Time:
3.50(+)
2.69
for every class
for exams
2.83
1.67(-)
Attitude:
Benefit 3.69(-) 4.67(+)
Enjoyment 3.25(-) 4.17 (+)
3.19(-) 3.67(+) Confidence
4.17(+) Overall Attitude 3.37(-)
Note. Numbers in cells are means for the respective outcome;
variables.
(+) = highest score
(-) = lowest score
Study Time for every class— Treatment main effect F= 4.381, p=.04
for exams— Education main effect F=3.856, p=.05
Attitude Toward Instruction: Benefit-Interaction effect F=3.745,
____________________________ P=,06___________ : _________-
147
had the highest attitude; the lowest is reported by seniors
in groups (the only significant result is benefit, F =
3.745, p = .06). It seems that the educational level did
not differentiate within the lecture method.
The pattern indicated in Table 25 for cognitive
outcomes shows the graduates to achieve higher on all mea
sures, fluctuating between the methods on various measures.
The last two tables (2 6 and 27) present data from
a 3-way ANOVA, treatment by education by learning style
type. The sample size in this breakdown became very small
and therefore the interpretation of these trends should be
viewed very cautiously? nevertheless, some interesting
directions seem to appear.
Table 26 presents the data for the cognitive out
comes. It is divided into two sub-tables, one for reflec
tive type and one for active type. It is helpful to
inspect the table also in relation to Tables 19 and 25
which present the 2-way analysis of the same variables.
It seems that graduates in lecture whether reflective or
active have the highest scores on all measures besides the
first PMP. Overall it seems that reflective-graduates-in-
lecture did better than any other sub-group, while active-
seniors-in-lecture have the lowest scores (course score:
1.84 vs. -.89).
Table 27 is constructed in the same manner as
Table 25 and presents data for the study time outcomes and
:148
Table 25
2-Way Analysis of Variance, Treatment by
Educational Level on Achievement and Problem-Solving
Educational Level
Seniors Graduates
(n=14) (n=8)
Achievement 28.50 30.00(+)
1
PMP I P 49.43 50.25
+J
o
a
PMP I E 65.86(-) 67.88
PMP I OC 42.43 42.88
PMP II P 49.43(-) 60.88(+)
PMP II E 61.14 68.50(+)
PMP II OC 43.07 51.88.(+)
-p
s
Course Score (z) -.06 .72(+)
i
rd
CD
(n=16) (n=6)
&
Achievement 28.25(-) 29.50
PMP I P 49.63(-) 54.17(+)
PMP I E 66.56 70.67(+)
PMP I OC 42.38(-) 47.17(+)
PMP II P 42.94 47.33
Ui
PMP II E 61.19 50.50(-)
&
g
PMP II OC 35.75(-) 39.17
M
0
Course Score (z) -.42(-) .30
Note. Number in cells are means for the respective outcome variables.
(+)-.= highest score
(-) = lowest score
Table 26
3-Way Analysis of Variance, Treatment x Educational Level x Learning Style Type
on Achievement and Problem Solving
Learning Style Type-—Reflective Learning Style Type— Active
Educational Level Educational Level
Seniors Graduates Seniors Graduates
(n=7) (n=2) (n=7) (n=6)
( U
Achievement 28.29 32.00(+) Achievement 28.71 29.33(+)
B
PMP I P 60.57 61.00 PMP I P 39.29(-) 46.67
■P
U
PMP I E 71.29 71.00 PMP I E 60.43(-) 66.83
PMP I OC 52.86 52.50 PMP I OC 32.00(-) 39.67
PMP II P 60.29 63.50(+) PMP II P 38.57(-) 60.00(+)
PMP II E 62.43 67.50(+) PMP II E 59.86 68.83(+)
4J
PMP II OC 52.43 54.00(+) PMP II OC 33.71 51.17(+)
Course Score (z) .76 1.84(+) Course Score (z) -.89(-) .34 (+)
f d
£
(n=8)
(n=2)
(n=8) (n=4)
E h
Achievement 28.63 30.50 Achievement 27.88(-) 29.00
PMP I P 45.13(—) 68.50(+) PMP I P 54.13(+) 47.00
PMP I E 64.63(—) 75.50(+) PMP I E 68.50(+) 68.25
PMP I OC 37.50(-) 60.00(+) PMP I OC 47.25(+) 40.75
PMP II P 43.63 39.00(-) PMP II P 42.25 51.50
W
Q
PMP II E 66.38 41.50(-) PMP II E 56.00 55.00(-)
B
PMP II OC 38.00 28.00(-) PMP II OC 33.50 44.75
n
Course Score (z) -.47(-) .80 Course Score (z) -.36 .06
hrote. Numbers in cells are means for the respective outcome variables.
(+) = highest score
(-) = lowest score ^
Table 27
3-Way Analysis of Variance, Treatment x Educational Level x Learning Style Type
on Study Time and Attitude Toward Instruction
Learning Style Type-—Reflective Learning Style Type— Active
Educational Level Educational Level ,
Seniors Graduates . Seniors Graduate
(n=7) (n=2)
(n=7) (n=6)
Study Time: Study Time:
S
for every class 2.71 2.50(-) for every class 2.86 2.67(-)
for exams 2.43 2.50 for exams 3.14 2.33
s
Attitude: Attitude:
Benefit 4.00 4.50 Benefit 4.14 3.83
Enjoyment 3.57 4.50 Enjoyment 4.29(+) 3.50
4J
Confidence 3.86 4.00(+) Confidence 3.29 3.33
G
Overall Attitude 3.81 4.33 Overall Attitude 3.90 3.56
( D
(n=8) (n=2) (n=8) (n=4)
H
Study Time: Study Time:
for every class 4.13 (+) 3.00 for every class 2.88(+) 2.75
for exams 2.88(+) 2.00(-) for exams 2.50 1.50(-)
Attitude: Attitude:
w
Benefit 3.75(-) 5.00(+) Benefit 3.63(-) 4.50(+)
£
Enjoyment 3.25(-) 5.00(+) Enjoyment 3.25(-) 3.75
Confidence 3.13(-) 4.00(+) Confidence 3.25(-) 3.50(+)
O
Overall Attitude 3.37(-) 4.67(+) Overall Attitude 3.37(-) 3.92(+)
t -
Ln
C
Note* Numbers in cells are means for the respective outcome variables. (+)^highest score, (-) -lowest score
151
attitude, (Tables 20 and 24 are related). It indicates
that reflective seniors in groups report the largest
amount of study time for both variables among their type,
while reflective graduates report least time for every
class in lecture, and for exams in groups. Among the
active type, seniors in lecture report more time for exams,
and in groups for every class. The least time is reported
by active graduates in lecture for every class and in
groups for exams. Overall, the largest amount of time for
every class is reported by reflective-seniors^in-groups
and the least for reflective-graduates-in-lecture. In
relation to study time for exams the largest overall
amount is reported by active-seniors-in-lecture and the
least by active-graduates-in-groups.
The attitude variables show a similar trend over
both learning types. Graduate students have higher atti
tude in groups, and seniors in groups have the lowest atti
tude. Only in regard to enjoyment do active seniors report
higher values in the lecture method. This is interesting
in contrast to their performance which is the lowest over
all (Table 26).
For the following variables, significant results
were found in Table 27: study time for every class--
treatment effect (F = 4.832, p = * .03), interaction-
treatment by type (F = 4.096, p = .05); study time for
exams— education effect (F = 3.684, F = .06); attitude
152
benefit— interaction-treatment by education (F 3. 248
.08); attitude enjoyment--interaction-education by
type (F 3.341
Tables of Pearson r correlations among all vari
ables for each treatment are provided in Appendix E. Dis
cussion of the results and elaboration on the various
findings is provided in Chapter V.
153
CHAPTER V
SUMMARY, CONCLUSION, DISCUSSION,
AND RECOMMENDATIONS
Summary of the Study
The overall general purpose of this study was the
investigation, (via a controlled experiment) of instruc
tional effectiveness in occupational therapy education,
conceived as an Aptitude—Treatment Interaction (ATI)
study. More specifically, the interaction between indi
vidual learning style and teaching method in affecting
various learning outcomes, was the main focus of this
study. Another main aspect of the study was the investi
gation of occupational therapy students1 problem^-solving
ability as a learning outcome, besides achievement and
attitude toward instruction. Construction of Patient
Management Problems (PMP) in occupational therapy and
their use in measuring problem-solving was an additional
important objective of this study.
The main hypothesis referred to the interaction
between learning style and teaching method and was
stated as follows: Students whose learning style "matches"
the teaching method will score higher on measures of
problem-solving and achievement than students whose
154
learning style and teaching method are "mismatched."
In addition main effects for the group-discussion method,
and the active learning style for problem-solving scores
were hypothesized, and the "match" between them was pre
dicted to yield the highest scores on the PMPs.
The additional variables obtained in the study
were entered for exploratory reasons, i.e., seeking to
determine what relationships "amount of study time,"
"attitude toward instruction," and "educational level"
have to the main independent variables.
The literature review introduced first the con
ceptualization of the teaching/learning process as a
three component model including learner characteristics,
instructional variables, and learning outcomes (Bloom,
1976; Bruner, 1966; Gagne, 1977). Following, the ATI
approach is presented from its theoretical—conceptual
premise which is congruent with the teaching/learning
model outlined above, and in addition its research method
ology is briefly described (Cronbach, 1975; Cronbach &
Snow, 1977; Snow, 1976b, 1980). Studies utilizing the
ATI approach are numerous and generalizations from them
are very difficult to draw, due to the use of many dif
ferent combinations; of variables investigated. But, as
Cronbach & Snow (.19 77) suggest, investigating interaction
effects provides better understanding of subgroup per
formance than main effects, only.
155
The additional sections in the review of the
literature deal with the three components of the teaching/
learning process chosen to be investigated in this study:
(1) teaching methods were defined as lecture and group-
discussion, and the differences between them were classi
fied as: instructor-centered vs. student-centered dis
cussion; large class vs. small class size? limited feedback
and involvement in class vs. high degree of feedback from
peers and instructor, and active involvement in class;
reading assignments only vs. assignments including case
studies; active experimentation'with materials and pre
paration of issues to be discussed in class; and limited
responsibility for class function vs. high responsibility
for class function. (2) individual learning style as
measured with Kolb1 s LSI (.1976) (and specifically the
active/reflective dimension) was investigated as the main
individual difference variable. But, as suggested by the
ATI; methodology additional aptitude variables were obtained
which were found in previous studies to correlate with
learning outcomes, such as, verbal ability/prior
achievement, perceived benefit from learning situations
and educational level (Snow, 19 76). (31 problem-solving
skill was considered as the main outcome of interest, and
was investigated through the use of the PMP technique.
The three other learning outcomes included were; achieve
ment (defined as knowledge of the subject area), amount
156
of study time for class and exams (student self-report),
and attitude toward■instruction (including benefit,
enjoyment and confidence variables and measured with a
student self-report questionnaire).
By integrating the three components on the basis
of previous theories and research findings (provided in
Chapter I!) the preferential model (Salomon, 1971) or
"matching" concept led to the study's interaction hypothe
sis predicting that a "match" between lecture and a
reflective learning style, as well as a "match" between
group-discussion and an active learning style, will result
in higher student performance, while the contrary will
hold for the •"mismatched" subgroups (Randolph & Posner,
19791. In addition, main effects for group discussion and
active style were predicted on the basis of findings indi
cating that learning in group method is more beneficial
when problem-solving is considered (..McKeachie & Kulik, 1975).
Thus, the Research Method and the study's design
were conceived as a 2x2 factorial design— a posttest-only-
control-group-design (Campbell & Stanley, 1963; Cronbach
& Snow, 1977; Kerlinger, 1973). The design can be
described also as a randomized-block design as subjects
were blocked first on the Learning.style variable into
active and reflective subgroups, and then randomly
assigned to teaching methods within these subgroups.
The study's sample consisted of 44 occupational
therapy students, seniors and graduates f studying at________
157
USC in the academic year 19 80-1981. Following the random
assignment procedure the teaching methods were comprised
of 22 students in one lecture class and two group-,’ ;
discussion classes, each with 11 students. Students*
scores on the LSI resulted in 25 active style and 19
reflective style which were randomly distributed as
follows: lecture— 9 reflective and 13 active, group-
discussion— 10 reflective and 12 active. All subjects
in the study agreed to participate and signed an informed
consent form and with, one exception, all were females.
As Indicated by the design and the methodology
followed in this study, data gathering of aptitude vari
ables was obtained before instruction started (and before
subjects knew to what method they were assigned), and
at the end of the instructional period, students were
tested with the outcome instruments. The instruments
employed in this study included: (1) Learning Style
Inventory (Kolb, 1976) , (2) Students * rating of situa
tions: that facilitate their learning (adopted partly from
Kolb, 19 76). , (3) Concept Mastery Test (Terman, 1956) .
The next three Instruments listed below were used to
assess the outcome variables.; (4) Multiple—choice-question
exam, (5) two Patient Management Problems (PMP) following
McGuire et al., 1976 construction guide book), (6) Student
evaluation questionnaire.
158
The data analysis followed the ATI methodology
and regression analysis (multiple and simple) was pre—
dominently used. In addition, analysis of variance, t—test
and Pearson r correlation were used, as well as descrip
tive statistics: means, standard deviations, and fre
quencies .
Major Results
A summary of the major results presented in
Chapter IV is provided in this section.
1. Descriptive statistics and t—.test between
scores for the two teaching methods show no significant
difference tor any of the variables except for study
time for every class (t=-2.08 p=.05).
2. Results of the multiple 'regression analysis
for the problem-solving outcome show significant main
effects of GPA with all scores at .05 level, and for
perceived benefit from group with the second PMP at .10
level. Significant interaction was found with learning
style type for the first PMP at .05 level, and a similar
trend for the second PMP.
Results for the overall course score show signi
ficant main effect with. GPA, and significant interaction
with perceived benefit from lecture, both at the .01
level. The same is true for achievement outcome.
159
Findings for the study time outcome indicate sig^
nificant main effect of verbal ability (p <.01), and
treatment (p < .10) for both time variables. In addition,
perceived benefit from lecture showed significant main
effect (p <.01), and interaction (p< .10) with study time
for every class. For both variables significant inter
action was found with learning style or type (p < .05).
Results; for the attitude variables show signifi
cant main effects of verbal ability and perceived benefit
from lecture (p < .05), with the exception of confidence
Cp < .10). Significant interaction was found with per
ceived benefit from lecture--benefit (p < .10) , enjoyment
(p < .01) and overall attitude (p < .05)— with the exception
of confidence.
3. In a 2-way analysis of variance, treatment by
learning style type, the following trends in cell means
are indicated;
Cognitive outcomes the highest scores occur in cell ■ I
(Achievement and
-Problem—solving) for reflective-type—in-lecture, while
the other cells show inconsistent
trends.
Study time for
every class - highest scores; occur in cell III
reflective-type-in-groups (Scheffe*s
test, F=12.31, p < .05) , and lowest
in cell I reflective-in-lecture.
Study time for exam - lowest scores in cell IV.
Attitude toward instruction - all except for enjoyment are
highest in cell I reflective-
in-lecture, but the means
are actually very similar.
4. Using correlations and simple regression
between the cognitive outcomes and attitude (as predictor),
significant positive correlation was found for achievement
in the lecture method (r=.34, p < .05), and for problem
solving in the group-discussion method (overall competence ,
PMPIOC r—.27, PMPIIOC r=.36, p< .10). Looking at the
separate attitude scores students in lecture are higher
on bene fit (r == .36, p< .05) and students in group-discussion
are higher in conf idence ( r= .40, £ < .05).
5. Using educational leye1 as an independent
variable, significant t-tests were found for the following
variables; verbal ability (t=-2.33, p^c .05), prior
achievement (GPA) (t=3.34, p < .01), learning style (t=l.96,
p < .05) , both study time variables (t=l.89, t=2.06, £ < .10).
In a 2-way analysis of variance, treatment by
educational level, the following trends in cell means are
indicated:
Cognitive variables - graduates main effect over both
methods: most scores are highest
for graduates in lecture, but the
PMP first case is higher in groups.
161
Study time for eyery class - seniors in groups report more
study time than any other
s uhg roup ( . S che f f e ' s test,
F=6. 79 , p <.10) , and graduates
in lecture report least time.
Study time for exams - seniors in lecture report more time
and graduates in groups report least
time.
Attitude toward instruction — highest attitude show grad- >
uates in groups and lowest
seniors in groups.
6. A 3-way analysis of yariance, treatment by
education by learning style, shows the following trends
in cell means;
Cognitive Variables - graduates in lecture, regardless of
type, perform higher on all measures
except for the first PMP, which shows
inconsistent pattern. The overall
: course score over both types shows
the. highest score for refiective-
graduates-in-lecture/ and the lowest
for actiye-sehiors^in-lecture.
Study time for every class '-over both types f reflective-
seniors-in-groups report the largest
amount of time and reflective-'/* 1
graduates-in-lecture report the least.
162
Study time for exams - over both types active-seniors-in
lecture report the largest amount
of time and the least, active-'./'
graduates-in-groups.
Attitude toward
Instruction - for both types the pattern seems
similar? graduates in groups have
higher attitude and seniors in
groups lowest, except for enjoyment
which shows active-seniors-in-
lecture to have the highest atti
tude .
Hypotheses
This section restates the four hypotheses formu
lated originally for problem-solving and achievement and
indicates whether they were or were not supported.
1. Students whose learning style "matches." the
teaching method will score higher on jaeasures of problem-
splylng and achievement than students; whose learning
style and teaching method are "mismatched."
This hypothesis was supported for the problem-
solving outcome? the match between reflective learning
style and the lecture method showed the highest scores.
The hypothesis, was not supported for the achievement
outcome which showed no difference. The combined course
163
score of both outcomes supports the hypothesis, especially
within the lecture method, as reflectives scored the
highest, and actives, the lowest.
2. Students whose learning style "matches" the
group-discussion method will score the highest on measures
of problem-solving.
This hypothesis was not supported; as indicated
for the first hypothesis, the match, with the lecture show
ed the highest problem-solving scores.
3. Students participating in group-discussion
will score higher on measures of problem-solving than
lecture students. No significant difference will be found
in achievement (as, measured on a multiple—choice-quesfion
exam for knowledge and comprehension).
The first part of the hypothesis regarding the
main effect of group-discussion was not supported. The
second part was supported as no difference was found
between methods for achievement.
4. Students whose learning style emphasizes the
active dimension over the ref lective (AE-RO)., will score
higher on measures of problem-solving. No significant
difference will be found in achievement (knowledge and
comprehension).
The first part of this hypo thesis; was; not support
ed because active students did not score higher on
164
problem-solving. The second part was supported and no
significant difference in achievement was found.
Conclusions and Discussion
The focus of the discussion in this section is
on the interaction between teaching method and individual
learning style (in relation to each of the outcome vari
ables). , which was the study's main purpose. In addition,
main effects and interactions with: other aptitudes found
in the analysis are discussed, and supplemented with a
discussion of educational level as independent variable.
Before discussing interactions it is important
to indicate that no significant main effect differences
were found between the treatments for any of the variables,
except for "study time for every class" a difference which
was expected due to the treatment manipulation procedure.
This finding is in accordance with many conventional com
parative studies (Dubin & Taveggia, 196 8) regarding the
achievement outcome, but, is in conflict with, most findings
indicating main effects for group methods in regard to
problem-solving outcome (Berliner & Gage, 19 76) McKeachie,
1978? McKeachie & Kulik, 1975? Zimmerman & King, 19 63)
although in a recent study a similar result was found by
Shumway and Donahue (.29 80) in medical education. The
lack of treatment main effect while significant interac
tion exists supports the ATI approach, (and others) that
16 5
no one superior teaching method exists and learners inter
act differently with various methods (Abrahamson, 19 76;
Cronbach, 1975; Cronbach & Snow, 1977; Snow, 1976b). As
stated by Shaw and Bunt C19 79) "ATI designs enable
researchers to detect effects on subpopulations which would
be missed in a standard main—effect analysis" (p. 44).
In their study which followed a "matching" hypothesis
was also conducted with college students and the same
pattern was found: no treatment main effect, but there
was an interaction between low course structure and high
preference for structure, resulting in the lowest mean
achievement.
In the following paragraphs every outcome variable
is discussed separately starting with problem-solving
and following with achievement, study time and attitude
toward ins t rue tion.
Problem-Solying Skil1
The interaction of learning style type and teach
ing method was significant in regard to the first PMP,
while a partly similar trend can be seen also in the
second PMP. This interaction accounts for approximately
5 percent of the variance in the first PMP and is signi
ficant at the .05 level, which when considered after all
main effects were accounted for seems to explain a sub
stantial portion of the total variance. By inspecting
166
Figures 7 and 10, and Table 19 in Chapter IV, it can be
seen that the subgroup reflective-type-in-lecture performed
the highest on all PMP scores except for PMPIIE (which
is actually almost equal to active in lecture). In the
remaining three subgroups active-type-in-groups performed
higher on the first PMP and active-type-in-lecture per
formed better on the second PMP, while the lowest perfor
mance seems to be either for reflective in group or
active in lecture over the two PMPs.
Referring back to the study's hypotheses and its
design, the two match subgroups— reflective-type-in-lecture
and actiye-type—in—group— seem to do better, while the
mismatched sub—groups succeed less well. This finding
supports the first main hypothesis but, the relationship
seems much, stronger for the reflective—type—in—lecture
method, thus opposing the second hypothesis which pre^
dieted highest scores for active-type-in—groups.
Only one additional interaction existed between
PMP I IE and verbal ability, which is significant at the ,01
level, and accounts for 8 percent of the variance. As
seen in Figure 8 it is a disordinal interaction indicating
that high—ability students do better in lecture while v
low—ability students: perform better in groups Con this
particular efficiency score). In regard to main effect
the only consistend significant result over all PMP
scores is for prior achievement (GPA) which accounts for
167
approximately 10 percent of the variance. In addition,
perceived benefit from group had also significant main
effect for the second PMP.
When educational level was considered as an inde
pendent variable, it seems that graduate students do better
on all PMP scores but differ between the two PMPs (grad
uates in groups do better on the first PMP but those in
lecture do better on the second) (see Table 25). Taking
in account treatment, education and type, the picture
becomes less consistent (Table 26), and because of the
very small sample size in each cell of this breakdown,
can hardly be interpreted.
One additional finding regarding problem-solving
was found when attitude was used as a predictor. A
significant positive correlation existed between these
variables, indicating that the higher the attitude toward
instruction the higher the PMP scores in the group method,
while no relationship was found in lecture. Looking at
both together, students: with high attitude do almost the
same in both methods, but students with low attitude
perfotm much lower in the group method. On the separate
attitude measures, confidence seems to have the highest
correlation with problem-solving in the group method,
while benefit correlated highest in the lecture method.
16 8
In summary, it can be concluded that although
the multiple regression as a whole accounted only for
approximately 30 percent of the total variance in problem
solving, the interaction with learning type beyond prior
achievement main effect, is significant. In reference
to the hypotheses the results indicate that it may be
beneficial to match students" learning style and the method
in which, they study, which is in accordance with Kolb's
theory (.19 74, 1979), with the Randolph arid Posher (.1979)
application of it to learning situations, and also with
a preferential model as presented by Salomon (.1971) . The
lower performance of active-type students in the group
could have been caused by the short period of the instruc
tion, as suggested also by Cronbach and Snow (19 77); a
longer period of habituation is needed for a method to
be effective, especially for a new method. Reflective-
type students in the lecture method, on the other hand,
comprised the right fit, and were already familiar with,
the method, thus no extended habituation time was needed,
allowing them to perform better. A mismatch between
learning style and method seems to be detrimental when
an outcome like problem-solving is regarded.
Another consistend finding which is important to
note is in regard to educational level, namely, that
graduate students performed higher on the problem-solving
measures. This result may be explained by two factors;
169
(1) graduates were found to have significantly higher
GPA scores, as well as higher verbal ability scores, and
as indicated before, GPA had significant main effect with
the PMP scores; and (2) graduate students were signifi-
cantly more active-learning-type which probably enabled
them to do better on an instrument requiring more involye-
ment, regardless of the method they were placed in. (This
interpretation is, of course, very tentative and much
more data are needed to confirm it.
The relationship between attitude and problem
solving has some similarity to Strom and Hocevar (19 8.1)
data showing that students with low attitudes do lower
in a moderate'structure treatment (could be equated to
group in this study), and higher in high-structure (lec
ture in this study). It seems that a method which requires
more student involvement, responsibility and preparation
is more influenced by students * .low attitude. In regard
to the "constructive motivation hypothesis" (developed
by Cronbach & Snow, .19 77; and Snow, 2976a; end tested in
Strom & Hoceyar's 2982 study), the findings of this;
study provide support for it. Students who have higher
attitude toward the group-discussion method show higher
problem—solving ability. It should be noted that this
relationship was not found in regard to achievement in
groups, which may suggest that positive attitude is even
170
more important when an outcome like problem-solving is
considered.
Achievement
The findings in regard to achievement support
hypotheses (3&4) of no difference in methods of learning
style. Hypothesis (1) which predicted higher scores for
the matched subgroups was not supported. The cell means
for achievement are almost equal; thus, the match between
learning style and method, which seems to be important
in regard to problem-solving, has no influence on a
traditional measure of knowledge of a subject matter. The
s.ame conclusion is of course indicated by 'the lack of
interaction between treatment and learning style in the
multiple regression analysis. A potential explanation
of this result is provided by Hilgard in Dubin and
Taveggia (1968) where he states that ". , . most studies
have relied very heavily on a common textbook . , . and
we measure what the student learned from his textbook , . ,
which may be so powerful as to override differences in
teaching" (p. 47). It seems not to be the case for
measures of problem-solving which are novel and require
additional skill not provided directly in the text.
The full model of the multiple regression accounts
for 5 8 percent of the total variance in achievement,
almost half of it; 26.8 percent is accounted for by a
171
GPA main effect indicating that higher GPA is associated
with higher achievement regardless of method. An addi
tional main effect was found for perceived benefit from
group (R %=10.5) which is difficult to understand because
it means that higher perception of benefit from group
led to higher achievement in both methods. One interaction
was found regarding perceived benefit from lecture which
accounts for 12 percent of the total variance and is
significant at the .01 level. Inspecting Figure 6, the
interaction indicates that students who had low perception
of benefit from lecture and participated in groups per
formed much higher than those in the lecture. In other
words, a student who perceives low benefit from a method
he/she is, in actually benefits: less from it, while low
perception from the opposite method may, by inference,
indicate high perception of benefit from his/her method,
thus, higher performance.
Relating this result to the findings when attitude
was used as. predictor, the same relationship seems to
exist in the lecture method where higher attitude is
correlated with higher achievement, especially in relation
to the separate benefit score (r=. 3 6 , p < ,05},
Educational level as an independent variable shows
a slight difference in favor of graduate students but
the actual means are very similar, and no real difference
172
seems to exist (Table 25). In the 3-way breakdown (Table
27) /which needs to be considered very cautiously,
reflective-graduates-in-lecture seem overall to achieve
the highest, and active-seniors-in-r^groups seem to have
the lowest scores.
For exploratory reasons a combined overall course
score was calculated, a procedure which is done in most
educational practices when a grade has to be calculated
(not used in this study as a grade) and also in many
research studies. The investigator1s unstated hypothesis
was that this procedure would result in eliminating
differences between the two cognitive measures and that
it is not appropriate. From the findings regarding the
overall score it seems that the: above prediction was- •
partly supported.
The regression analysis provides almost identical
results with the achievement outcome. The GPA main
effect is even larger (R %=33.4) due to its significance
for all cognitive scores, and no other main effect exists.
The only interaction is with perceived benefit from lec
ture, which is similar to its effect on achievement,
but even more pronounced (see Figure 5), The interactions
with, learning style occuring for problem-solving were
eliminated probably because they were not large enough,
but their possible explanation power is also lost if no
173
other analyses are conducted. in order to examine this
relationship, simple regression was drawn (Figure 9) and
Table 19 was inspected. In both, it seems that the
learning type differentiates within the lecture method
because the reflective type have the highest score and the
active type the lowest, while all group members have
exactly the same mean. In general this finding highlights
the conclusion about the match between the reflective type
and the lecture method, while almost no difference can
be seen among the other three subgroups, which is only
a slight indication that the mismatch of active type in
lecture is the most detrimental. Considering the addi
tional information provided when educational level is used,
the main effect of graduates is seen once more, but show
ing also graduates in lecture to haye the highest scores
and seniors; in group the lowest (Table 25). Looking at
the 3-way breakdown (Table 26), it seems that, overall,
reflective-graduates-in-lecture have the highest scores
( . 2 5 —+1.84) f and active-seniors-in-lecture have the lowest
(,z=-. 89) . According to this trend in the analysis the
"match" hypothesis could be extended over the three inde
pendent variables, and over 8 cells, to suggest that the
best match occurs for reflective-type-graduate-in-lecture,
and the greatest mismatch occurs for actrye-type-seniors-
in-lecture. This relationship could become an interesting
hypothesis for further research, and probably an important
174
one for practical implementation of instruction in
similar situations.
In general/ comparison between the two cognitive
outcomes, achievement and problem-solving, and the rela
tive contribution of the independent variables to their
explanation can be summarized as follows: //'
2. Prior achievement (GPA) predicts performance
gn all measures, but is much more significant for the
achievement outcome. In addition, perceived benefit from
small group shows main effect for achievement and the
second PMP, but the direction for both is different (posi
tive with achievement and negative with the PMP}/ iThis
last relationship is. unclear because it does not show an
expected interaction differentiating between its influence
in relation to the treatment. The relationship is
understandable for students in the group who do better
on achievement, but is not as understandable for students
who do better when they are in lecture. The negative
correlation with the second PMP is unclear in regard to
the group unless students perceived the group method to
be different than it actually was; for the lecture, it
is more understandable. More research is needed in order
to understand better the above mentioned relationship.
2. The main difference between the two outcomes
relates to the interaction terms. Perceived benefit from
lecture showed significant interaction for achievement.
175
while learning type or educational level made no differ
ence. It seems that the best predictors for a conventional
measure of achievement using a multiple-choice-exam were
prior achievement (GPA) , which was probably measured in
the same manner, and in addition, students' perception of
benefit from the method indicating a positive relationship,
thus; supporting the notion that student perceptions
influence achievement (Salomon, 19 8.1; Shaw & Bunt, 1979) .
In contrast, individual learning type showed a
significant interaction for problem-solving outcome, and ■
in addition, educational level seem to show main effect
indicating higher performance for graudate students. These
findings may suggest that problem-solving requires more
from the students than can be compensated by individual
learning from textbook as stated by Hilgard (in Dubin
& Taveggia, 1968). According to Gagne (19771, problem
solving is placed at the top of the intellectual-skills
capabilities (learning outcomes), thus requiring many
previously learned rules (which may be integrated in an
individual learning style). Gagne also emphasizes the
conditions of learning to include conditions within the
learner, and conditions in the learning situation, which
preferably should accommodate each other. These theoretical
postulates seem to indicate the importance of the "match"
between the learner's style and the method, referred to
176
in this study, especially when a higher—order skill such
as problem-solving is considered. In a recent review
by Vu (19,80) about teaching and problem-solving in medical
education, problem-solving is defined as a process includ
ing 1 1 „ . . three main skills: collecting, interpreting,
and integrating data (cues)" (p. 440). These skills seem
to differ among individuals and seem to be related also
to ". , .the interaction between the problem and the
problem solver's ability" (Vu, 19 80, p, 443). Extended
a step further, it seems that the interaction among the
problem, the individual style, and the teaching method
may explain even more of the problem-solving skill.
Higher performance of graduate students noted
earlier may also refer to the individual learner's ability
part of the teaching/learning process triangle, in that
their GPA and verbal ability were significantly higher
and their learning style was significantly more active.
Problem-solving ability in itself is not seen as a
general stable trait (Neufeld, 1977), but it probably
depends, to a greater extent on individual intellectual,
perceptual and personality factors which may be incorpor
ated in the measure of individual learning style as
conceptualized by Kolb (19 74, 19 79). Although individual
differences are acknowledged in medical education dealing
with predicting and teaching of problemsolving, no ATI
177
study was conducted in this area and no effort seemed to
be made to try to teach problem—-solving with different
methods to different individual learners.
There is one additional fact worth noticing which
caused concerns to many who used the PMP technique:
examinees' inconsistent performance over various PMPs,
(Elstein, 1978, Vu, .1980) which is the case also in this
study. According to Vu's review of various studies this
inconsistency ". . . appears to be case-related . . .
while information gathering ability may generalize from
case to case, decision-making ability (diagnostic and
management) appears related to the content of the case’ 1
Cp. 443) . In the present study, the two PMPs differ in
the role the examinee has to play and in the kind of
decisions he/she has to make, thus perhaps explaining the
difference in performance. It still raises a question
about the reliability of the PMP technique in general and
in this study as well, even though their validity seems
to be high.
Amount of Study Time Outside Class
There were no stated hypotheses for this outcome,
and there is not much literature about it; therefore it
was entered in the analysis for exploratory reasons. Thus
the results and their discussion in this section should
be viewed as hypothesis-generation for further research.
178'
The first thing to .mention is the treatment main
effect which should he considered as a result of the
treatment manipulation pointed out before. Looking at
the other main effects, verbal ability shows a significant
effect at the point .01 level, which explains approximately
one third of the total variance (15% out of 5.1.65%, and
13.9% our of 39.93%) . It seems that in contrast to GPA
which accounted for a great part of the variance in
achievement, verbal ability has greater relationship to
the amount of time a student has to spend in preparing for
class and going over reading materials by himself f and
thus may be better predictor for it. This relationship
is. further strengthened when considering that, in the
analysis with educational level, graduate students had
significantly higher verbal ability and reported signifi
cantly less study time (Table 4 , Figure 15) .
One other variable, "perceived benefit from lec
ture" shows significant main effect and interaction with
study time for every class. This relationship is depicted
in Figure 11 showing an ordinal interaction, where it
seems clear that this variable makes no difference for
students, in groups as expected; but in lecture a positive
correlation exists suggesting that when students have
higher perception of benefit they also study more. This
relationship may be, viewed as Salomon ' (1981)' .postulates
179
in terms of the association between "perceived demand
characteristics (PDC)" and "amount of invested mental
effort (AIME)": " . . . the way one perceives the stimulus
demand characteristics determines how much and what kind
of mental effort will be invested in processing it" (draft,
ch. 5, no page number) (emphasis added). Although Salomon
does not mean that only amount of time is vital to AIME;
it certainly is part of it, and the relationship he sug
gests is supported in this study. Salomon further suggests
that AIME in turn influences achievement, a condition also
partly supported in this, study. When a regression of study
time as predictor on achievement was performed, a signi
ficant positive correlation was found for the lecture
method (Ff=2.92, jd < *05; F= 7,56, p < .01) respectively for
both time variables. But, in the group method the rela
tionship was not significant and was even slightly nega
tive, a fact which is harder to explain, unless educational
level is considered, where graduate students who report
the least amount of time still achieve higher.
The main interaction found for both time variables
is with individual learning style or type, as shown in
Figures 12 and 13 and also in Table 20. Study time for
every class shows an ordinal interaction which s u g g e s t s
that active type report approximately the same amount
of time regardless of method, but reflective type report
much more study time when placed in group than in lecture.
■. .180
The largest amount of time which was reported by reflec
tive in group is significantly different from all other
cell means (Scheffe's test), while reflective in lecture
seem to report the least amount. Extending the hypothesis
of "matching" to the study time outcome, it can be seen
that in regard to the reflective type in lecture, the
match hypothesized resulted in the least amount of study
time reported, while the mismatch of reflective with the
group method resulted in the largest amount of time.
Taking into account also the educational level, it seems
that the largest amount of time is reported by reflective-
seniors-in-groups which is also considered as one of the
mismatches, in contrast to reflective-graduates-in-lecture
who report the least amount and comprise probably the
best match, in this study.
The disordinal interaction for study time for
exams seems to indicate that active-type-in-lecture ,
studies more than active-in-group, while the opposite
is true for reflective type, who reports to study in group
the greatest amount before exam. Overall, the least
amount is reported by active-in-group.
Comparing both time variables, the reflective—
type-in-group, or the reflective-senior-in-group reported
to study more for every class and before exams. This
subgroup is considered as a mismatch, and is also among
181
the two lowest subgroups on the cognitive scores. On
the other hand, the least amount of time for every class
is reported by reflective-in-lecture or reflective-
graduate-in-lecture which comprised the best match in the
study in terms of their performance. The amount of time
a student reports to study seems to be a function of the
interaction between his learning style and the method.
This conslusion seems to be strengthened also when con
sidering that the least study time before exams was
reported by active-in-group or active-graduate-in-group.
Another potential explanation for this last relationship
can be that these active students in the group method
were involved and studied more all along, thus requiring
less time before the exams, while the reflective-in-
lecture studied less for each class and compensated for
it before the exam.
In summary, it seems that amount of study time
is influenced by verbal ability, by perceived benefit
from the method and by the interaction between the indi
vidual's learning style and the method he/she is placed
in, to which educational level adds power.
Attitude Toward Instruction
The attitude outcome was the second exploratory
variable entered in the analysis. The aptitudes which
show main effects for it are similar to the ones for
182
study time, verbal ability and perceived benefit from
lecture. In addition, GPA shows significant main effect
for benefit and overall score. All of the above show
positive correlations indicating that higher aptitude is
associated with higher attitude regardless of method.
The "perceived benefit from lecture" variable can actually
be considered as pretest to the attitude-benefit posttest
outcome ( the correlation between them is approximately
.30). Thus, when considering the significant interaction
found (Figure 14), this fact has to be taken in account.
What is unclear is the positive correlation between per
ceived benefit from lecture and attitude of students in
groups (this question was raised earlier also in regard
to main effect of perceived benefit from group, and
achievement for students in lecture).
Considering the relationships involved in learning
style, it seems that reflective-type-in-lecture have the
highest attitude and active-type-in-groups the lowest,
but the cell means are actually so close that it does
not seem to give much information (Table 20). Educational
level seems to be more related, and graduate students in
groups exhibit the highest attitude (Table 24); the. bene
fit variable shows an interaction indicating lowest score
for seniors in groups (Figure 16). What is also indicated
in Figure 16 is the main effect of lecture method for the
benefit score, as presented also by the high positive
183
correlation of benefit within the lecture method. In the
further breakdown (3-way ANOVA) of type, education, and
treatment (Table 27), graduate students in groups (regard
less of type) have higher attitude scores. White seniors-
in-groups have the lowest scores, except for enjoyment
which was higher for active-senior-in-lecture. This last
relationship is in opposition to this subgroupfs lowest
performance on the cognitive measures (Table 26), and
may lend some support to Clark's (1981) hypothesis of
negative correlation between achievement and enjoyment.
A similar trend is also indicated within the graduate
group itself, where their performance is higher in the
lecture method on almost all scores but their attitude is
higher in the group method.
In summary, it seems that after ability and a
pretest like perception measure are accounted for, atti
tude toward instruction is more related to educational
level, where graduates seem to have higher attitude
toward group-discussion and seniors value it the least.
This higher attitude of graduates is negatively correlated
to their performance, while for the senior group it is
more in accordance with their lower performance in group
method.
184
General Conclusions
Tables 28 and 29 provide summaries of the relation
ships found in the study. They organize the findings so
that some consistent patterns between the predictors and
the outcome variables may be viewed. Prior achievement
(GPA) accounts for most variance in the cognitive outcomes,
especially for the conventional measure of achievement.
But, in addition to it, achievement is explained by the
interaction with perceived benefit from lecture, while
problem-solving is explained more by the interaction with
learning type. Achievement as defined here seems to be
a known and familiar learning outcome, to which students
are accustomed and for which students probably know how
to study, and what to expect from the method. GPA, which
is comprised mainly of similar kinds of measures, accounts
for most of the variance and is therefore a good predictor.
Students are probably able to compensate by themselves
for any mismatch between their learning style and method.
On the other hand, problem-solving which requires other
skills like data gathering, hypothesis generations,
decision-making, more application of content, as well as
analysis of materials in some kind of sequential order
(Andres., 1972; Elstein et al. , 1978 ; McGuire et al. , 1976;
Vu, i980; and Ways, 1973), is explained more by a variable
Table 28
Summary of Variables Which Show Significant Contributions
in Explaining Variance for All Outcomes
..... ...... . - - —
OUTCOME VARIABLES
Predictors
Problem-Solving Achievement Study Time Attitude
PMP I PMP II class exams AB AE AC OA
Aptitudes
Main effects:
Prior Achievement (GPA)
Verbal Ability
Perceived Benefit from
Lecture
Perceived Benefit from
Group
— — —
Aptitude x Treatment
Interaction effects:
Perceived Benefit from
Lecture
Learning Type
Partial
Trend
—
______________ ____
Note. Exact values are shown in Tables 10-16. Attitude abbreviations: AB=benefit, AE=enjoyment,
AC-confidence, 0A=overall attitude.
H
.00
Vi
Seniors
Learning Style
Reflective Type Active Type
Education
Highest
Course score
PMPs
Attitude
Least study time
for every class
Highest study time
for every class
Lowest
Course score
Least study time
for exams
Lowest:
Attitude
Graduates
Highest Course Score
Achievement
PMP II
Least study time
for class
Highest
PMP I
Attitude
Least study time
for exams
Highest study
time for exams
Highest study
time for every
class
Lowest attitude
and course score
Lecture
Treatment
Groups
Figure 17.
Summary of Trends for All Outcomes as a Function of Learning Type, Educational Level, and Treatment
Note. Exact means are shown in Tables 19, 20, 24-27. Highest and lowest scores for achievement, and
lowest scores for PMPs not indicated, show no difference or inconsistent trend.
---- i_j
oo'
CT\
Table 29
Trends Combining the Three Dimensions: Learning Type, Educational Level,
and Treatment for All Outcomes
Variables Highest Scores Lowest Scores
Course Score Reflective-Graduate-in-Lecture Active-Senior-in-Lecture
Achievement Reflective-Graduate-in-Lecture Active-Senior-in-Group
PMP I Reflective-Graduate-in-Lecture Active-Senior-in-Lecture
PMP II Reflective-Graduate-in-Lecture Reflective-Graduate-in-Lecture
Study Time
for class Reflective-Senior-in-Group Reflective-Graduate-in-Lecture
for exams Active-Senior-in-Lecture Active-Graduate-in-Group
Attitude Reflective-Graduate-in-Group Active and Reflective-Seniors-
in-Group
H
oo
-u
188
as learning style which is assumed to incorporate intel
lectual and personality factors (Kolb, 1976, 1979).
It seems that the match between reflective learning
style and the lecture method resulted in the highest per
formance on the PMPs, while the other cells showed no
consistent pattern. Learning style showed no significant
main effect, and as emphasized by Kolb (1981) is probably
situational, thus its influence may be most significant
when allowed to interact with an appropriate situation
(in this case, teaching method). It is interesting
to point out that although, the other cells in the analysis
did not show consistent difference, when the overall
course score was calculated, the greatest mismatch seemed
to occur between active type and the lecture method. The
active/reflective dimension of the learning style seemed
to interact especially with the lecture method in the
hypothesized direction.
Additional information is provided by educational
level, which indicates that graduates are the best match
with reflectiye and lecture, while an active senior in
lecture is the least match. The importance of the learning
type may also be inferred when the analysis by education
alone is considered. The lowest course score in this
case is present for seniors in groups regardless of type;
however, when type is also taken in account a more hidden
relationship may be encountered.
189
When examining the other two outcomes it seems
that verbal ability accounts for variance in both, while
GPA contributes very little, in contrast to the cognitive
outcomes. The relationship indicated suggests that
study time outside class and attitude are explained more
by a student*s verbal ability than by his/her previous
achievement. It seems that less able students need more
time to study in order to compensate for their deficiency,
and therefore may also have lower attitude (indicated
also in the difference between educational level).
Another variable which seemed to contribute to the
variance in study time and attitude is perceived benefit
from lecture as main effect and interaction. In both
instances it seems that the significant relationship is
in the lecture method and indicates that students who
perceived to benefit from a method study for it more,
and more involved with it (Salomon, 1981), and they
also continue to have a higher attitude toward it. The
question which is. asked in many studies is whether stu
dents* perception accounts for their performance, or as
Snow (1980) formulated it, "can learner control of instruc
tion accommodate individual differences1 1 (p. 151) ?
Cronbach and Snow (19 77) in their review and analysis came
to the conclusion that students' preferences do not
necessarily predict their achievement. And as stated
by Snow (1980) learner control does not compensate for
190
the effects of individual differences. Support for these
conclusions may be indicated also in this study, where
perception interacted in the achievement outcome which was
in general not influenced by individual difference beyond
the effect of prior achievement. But, for the problem
solving outcome which seems to be more dependent on funda
mental individual differences and their interaction with
instruction, like the learning style, no interaction with
perception was found.
Besides problem-solving, amount of study time
shows significant interaction with learning style. This
relationship as discussed before seems to be in accor
dance also with a matching hypothesis, indicating that
students whose learning style and teaching method match,
perform better and report less time stpet in study outside
class. Amount of study time a student has to spend can
be viewed as the amount of compensation he/she needs to
do to overcome a mismatch between his/her individual
learning style and the teaching method.
Between the two time variables there seemed to be
a clear pattern in regard to the methods. Least amount
of study time for every class is reported in lecture,
whether the analysis is by type or by educational level,
and the opposite holds for study time before exams, where
the least amount is reported in groups. This relation
ship seems to be a pretty accurate representation of the
191
instructional requirements in terms of the study time for
every class session, but what it also indicates is that
as a result of the group learning process and the larger
amount of time spent for it, at the end before exams
students needed less time to study. The question, which
can't be answered in this study, is what in the long-run
is more effective in terms of knowledte aquisition or
problem-solving ability. This question is an important :
one to investigate because ultimately learning outcomes
should have long-lasting effects.
Overall conclusions or generalizations are hard
to draw from a study which included so many variables
and was based on a small sample size. The findings and
their interpretations should be viewed tentatively as
hypotheses to be further investigated. The list of
recommendations presented in the next section is comprised
of the important relationships found, which are recommend
ed for further investigation.
Re c ommendation s
Suggestions for further research which arose
from the implementation of this study are as follows.
1. The ATI approach followed in this study proved
itself in studying instructional effectiveness. Relation
ships. which otherwise would not be detected were observed
in spite of many methodological limitations. Thus, it
192 *
is recommended to employ the ATI approach, whenever
research concerning learning outcomes, as a function of
learner characteristics and instructional coihponents,
is conducted.
2. Individual Learning Style as a learner charac
teristic seems to be a significant construct which needs
to be further investigated. The interaction between
learning type and instructional components could be
investigated along other dimensions of the LSI and the
teaching methods. Further investigation of Kolb’s
Learning Style Inventory (LSI)'s measurement properties
(reliability and validity) is important, so that future
users c a n be confident in their decisions based on it.
3. The group-discussion method employed in this
study had many advantages in terms of more students' •
involvement and responsibility for their learning, com
bined with, less in-class—time required from the instructor,
But, its effectiveness needs further research with a
longer experimental period of time to allow for possible
effects to show.
4. The "matching" hypothesis between learning
style and teaching method seems to account for a signifi
cant amount of variance (especially for problem-solving
and study time), and has considerable explanatory power.
This relationship needs further investigation with the
same variables., as well as others.
193
5. Educational level showed significant differ
ences as a main effect and also in its interaction with
the teaching method and learning type. To understand
these differences and interactions, more research is *
needed which takes this variable in account, especially
in populations where two or more such groups learn togeth
er .
6. Problem-solving skill as an outcome variable
was mainly explained by the interaction with learning
style. This relationship can have important implications
for teaching toward this goal, and thus needs more
research. The Patient Management Problem (PMP) technique
also seems to have great potential in determining problem
solving ability in occupational therapy, but many more
PMPs have to be constructed and administered in order to
assess their, value as measures of occupational therapists*
problem-solving skills. Studies which investigate their
power to predict actual clinical performance are essential.
7. Amount of study time outside class seems to
be an important variable to investigate in relation to
learning style, method, and other outcome variables. Its
potential as a moderator variable between learning type
and the teaching method would be important to investigate.
194
8. Attitude toward instruction and some of its
negative correlations with performance need more investi
gation.
9. Research in the area of occupational therapy
education is needed and various components of this
process should be investigated. Replication of this
study with other occupational therapy students, in other
settings, and with other content areas is suggested in
order to better understand the relationships found in
this study.
195
REFERENCES
196 1
REFERENCES
Abrahamson, S. Helping individual students to learn.
Journal of Medical Education, 1976, _51, 1025.
Anastasi, A. Psychological testing. New York: Macmillan
Publishing Co., 1976.
Andrew, B. J. An approach to the construction of simulated
exercises in clinical problem solving. Journal of
Medical Education, 1972, 47, 952-958.
Atkinson, R. C. Ingredients for a theory of instruction.
American Psychologist, 1972, 27, 921-931.
Bane, C. L. The lecture vs. the class discussion method
of college teaching. School and Society, 1925, 21,
532; 300-302.
Barro, A. R. Survey and evaluation of approaches to
physician performance measurement. Journal of Medical
Education, 1973, 48, 1048-1093.
Barrows, H. S., & Tamblyn, R. N. An evaluation of problem
based learning in small groups utilizing an simulated
patient. Journal of Medical Education, 1976, 51_, 52-54.
Berliner, D. C., & Cahen, L. S. Trait-treatment inter
action and learning. In F. N. Kerlinger (Ed.), Review
of Educational Research. Itasca, Illinois: F. E.
Peacock, 1973, pp. 58-94.
Berliner, D. C., & Gage, N. L. In N. L. Gage (Ed.), The
psychology of teaching methods. University of Chicago
Press, National Society for the Study of Education,
1976.
Berner, E. S. Teasing apart the problem-solving process.
Conference of Medical Education Research, 1976, pp.
155-160.
Berner, E. S., Hamilton, L. A., & Best, W. R. A new
approach to evaluating problem-solving in medical stu
dents, Journal of Medical Education, 1974, 49, 666-672.
w
Bloom, B. S. Taxonomy of educational objectives. New
York: David McKay Co., 1956.
Bloom, B. S. Human characteristics and school learning.
New York: McGraw-Hill Co., 1976.
Briggs, A. K., Duncombe, L. W., Howe, M. C., & Schwarts-
berg, S. L. Case simulations in psychological occupa
tional therapy"! Philadephia: W~. A~. Davis Co. , 1979 .
Briggs, L. J. (Ed.) . Instructional design. ' Englewood
Cliffs, NJ: Educational Technology Publications, 19 77.
Bruner, J. S. Toward a theory of instruction. New York:
Norton & Co., 1966.
Buros, 0. K. (Ed.). Concept Mastery Test (CMT). In The
Fifth Mental Measurement Yearbook. New Jersey: Gryphon
Press, 1959, pp. 321; 449-450.
Campbell, D. T., & Stanley, J. C. Experimental and quasi-
experimental design for research. Chicago: Rand
McNally College, 1963.
Canfield, A. A., & Lafferty, J. C. Learning Styles Inven
tory Manual. Plymouth, Mich.: Experimental Learning
Methods, 19 74.
Certo, S. C. Stages of the Kolb-Rubin-Mclntyre Experimen-
tial Learning Model and perceived trainee learning: A
preliminary investigation. Academy of Management Pro
ceedings, 1977, pp. 21-24.
Clark, R. E. Is achievement enjoyable? Paper presented at
the American Educational Research Association National
Convention, Los Angeles, 1981.
Coop, R. H., & Brown, L. D. Effects of cognitive style and
teaching method on categories of achievement. Journal
of Educational Psychology, 1970, 6JL(5), 400-405.
Corno, L., Mitman, A., & Hedges, L. The influence of
direct instruction on student self-appraisals: A hier
archical analysis of treatment and aptitude-treatment
interaction effects. American Educational Research
Journal, 1981, 18, 39-61.
19 8
Cronbach, L. J., & Snow, R. E. Aptitude and instructional
methods. New York: Irvington Publications, 19 77.
Cronbach, L. J., & Webb, N. Between-class and within-class
effects in a reported Aptitude Treatment Interaction:
Reanalysis of a study by G. L. Anderson. Journal of
Educational Psychology, 1975, £7(6), 717-724.
DeTornyay, R. Measuring problem-solving skills by means of
the simulated clinical nursing problem test. The
Journal of Nursing Education, August 1968, £4(5), 3-8.
Dewey, J. Logic: The theory of inquiry. New York:
Henry Holt, 19 38.
Dietrick, D. C. Review of research. In R. J. Hill, (Ed.),
A comparative study of lecture and discussion methods.
Mew York: McGraw-Hill, 1960.
Domino, G. Interactive effects of achievement orientation
and teaching style on academic achievement. Journal of
Educational Psychology, 1971, £3(5), 427-431.
Domino, G. Aptitude by treatment interaction effects in
college instruction. Paper presented to American
Psychological Association, 1974.
Donnelly, M. B. Measuring performance on patient manage
ment problems. Conference of Medical Education Research,
1976, pp. 161-166.
Dowaliby, F. J., & Schumer, H. Teacher-centered vs.
student-centered mode of college classroom instruction
as related to manifest anxiety. Journal of Educational
Psychology, 1973, £4(2), 125-132.
Dubin, R., & Taveggia, T. C. The teaching learning paradox.
Eugene, Oregon: Center for the Advanced Study of
Educational Administration, 1968.
Elstein, A. S., Shulman, L. S., & Sprafka, S. A. Medical
problem solving. Harvard University Press, 1978.
Feightner, J. W., & Norman, G. R. Concurrent validity of
patient management problems by comparison with the
clinical encounter. Conference of Medical Education
Research, 1976, pp. 149-154.
199
Freedman, R. D., Stumpl, S. A. What can one learn from
the learning style inventory? Academy of Management
Journal, 1978, 21^, 275-282.
Gagne, R. M. (Ed.). Learning and individual differences.
Columbus, Ohio: Charles E. Merrill, 1967.
Gaghe, R. M. The conditions of learning (3rd ed.). New
York: Holt, Rinehart and Winston, 1977.
Gagne, R. M., & Briggs, L. J. Principles of instructional
design (2nd ed.). New York: Holt, Rinehart and Win
ston, 1979 .
Gall, M. D., & Gall, J. P. The discussion method. In
N. L. Gage (Ed.), The psychology of teaching methods.
University of Chicago Press, National Society for the
Study of Education, 19 76, pp. 166-216.
Gehlbach, R. D. Individual differences: Implications for
instructional theory, research and innovations. Educa
tional Researcher, 1979, 8_ , ' 8-14 .
Geller, L. M. Reliability of the Learning Style Inventory.
Psychological Reports, 1979, 4_4, 555-561.
Glaser, R. Adaptive education: Individual diversity and
learning. New York: Holt, Rinehart and Winston, 1977.
Glaser, R. Relationships between aptitude, learning and
instruction. In R. E. Snow, et al., Aptitude, learning
and instruction (Vol. 2). Hillsdale, NJ: Lawrence
Erlbaum Associates, 1980.
Goodenough, D. R. The role of individual differences in
field dependence as a factor in learning and memory.
Psychological Bulletin, 1976 , £3^(4), 675-694.
Goran, M. J., Williamson, J. W., & Gonnella, J. S. The
validity of patient management problems. Journal of
Medical Education, 1973, 48_, 171-177.
Gover, V. F. The NPSI: A nursing performance simulation
instrument. In E. Jacobi & L. E. Notter, (Eds.), Ameri
can nurses association eighth nursing research confer
ence , 1971.
200
Green, J. C. Individual and teacher/class effects in
aptitude treatment studies. American Educational
Research Journal, 1980, 17^(3), 291-302.
Grieve, T. D., & Davis, J. K. The relationship of cogni
tive style and method of instruction to performance in
ninth grade geography. Journal of Educational Research,
1971, 65 (3) , 137-141.
Grochow, J. Cognitive style as a factor in the design of
decision-support systems. (Unpublished doctoral disser
tation, MIT Sloan School of Management, Cambridge,
Mass., 1974.
Harasym, P. H. The underlying structure of clinical
problem-solving: Process or content? Proceeding 18th
Annual Conference on Research in Medical Education,
1979, pp. 67-83.
Heifer, R. E., & Slater, C. H. Measuring the process of
solving clinical diagnostic problems. British Journal
of Medical Education, 1971, 48-52.
Hill, W. M. F. Learning thru discussion. Beverly Hills,
CA: Sage, 1960.
Holcomb, J. D., & Garner, A. E. (Eds.). Improving teaching
in medical schools. Springfield, Illinois: Charles
C. Thomas, 19 73.
Hopkins, H. L., & Smith, H. D. (Eds.). Willard and Spack-
man's occulational therapy. Philadelphia: J. B.
Lippincott, 1978.
Hopkins, H. L., & Tiffany, E. G. In H. L. Hopkins & H. D.
Smith, Willard and Spackman1s occupational therapy,
Philadelphia: J. B. Lippincott, 1978.
Horne, D. J. Large 'small1 groups: Their place in higher
education. Medical Education, 1979, 1_3, 242-246.
Hubbard, J. P. Measuring medical education. Philadephia:
Lea and Feliger, 1978.
Huckabay, L. M., Cooper, P. G., & Neal, M. C. Effects of
specific teaching techniques. Nursing Research, 1977,
26(5), 380-385. (1979, 28(4), 228-233).
201
Hunt, D. E. Person environment interaction: A challenge
found wanting before it was tried. Review of Educa
tional Research, 1975, -45_(2) , 209-230.
Irby, D. M. Teaching and learning style preferences of
family medicine preceptors and residents. The Journal
of Family Practice, 1979 , M5), 1065-1067.
Janicki, T. C., & Peterson, P. L. Aptitude-treatment
interaction effects of variations in direct instruction.
American Educational Research Journal, 19 81, 118, 6 3-8 2.
Joyce, B. R. Selecting learning experiences. The Associa
tion for Supervision and Curriculum Development, 19 78.
Joyce, B. R., & Weil, M. Models of teaching. Englewood
Cliffs, NJ: Prentice-Hall, 1972.
Kerlinger, F. N. Foundations of behavioral research.
New York: Holt, Rinehart & Winston, 19 73.
Kerlinger, F. N. , & Pedhazur, E. J. Multiple regression
in behavioral research. New York: Holt, Rinehart &
Winston, 1973.
Kolb, D. A. On management and the learning process. In
D. A. Kolb, et al., (Eds.), Organizational psychology.
Englewood Cliffs, NJ: Prentice-Hall, 1974.
Kolb, D. A. Learning style inventory manual. Boston:
MA: McBer & Co., 19 76.
Kolb, D. A. Experiential learning theory and the learning
style inventory: A reply to Freedman and Stumpf.
Academy of Management Review, 1981, 6_ , 289-296 .
Kolb, D. A., Rubin, I. M., & McIntyre, J. M. Organiza
tional psychology: An experiential approach (3rd ed.).
Englewood Cliffs, NJ: Prentice-Hall, 1979.
Lamb, S. W., & Certo, S. C. The Learning Styles Inventory
(LSI) and instrumental bias. Academy of Management
Proceedings, 1978, pp. 28-32.
Llorens, L. A., & Adams, S. P. Entering behavior-student
learning style. In Ford and Morgan (Eds.), Teaching in
the health professions. St. Louis: Mosby Co., 1976.
202
Llorens, L. A., & Adams, S. P. Learning style preferences
of occupational therapy students. American Journal of
Occupational Therapy, 1978, _32(3) , 161-164.
Matarazzo, J. D. Wechsler's measurement and appraisal of
adult intelligence (5th ed.). Baltimore: Williams &
Wilkins,197 2.
Mayou, R. Psychiatric decision making by medical students.
British Journal of Psychiatry, 1978, 1_3(2) , 191-194.
McCarthy, W. H. Improving classroom instruction: A pro
grammed teaching method. Journal of Medical Education,
1971, 46_, 605-610.
McGaghie, W. C. Medical problem solving: A reanalysis.
Journal of Medical Education, 1980, 55_, 912-921.
McGuire, C. H., & Babbott, D. Simulation technique in the
measurement of problem-solving skills. Journal of
Educational Measurement, 1967, _ 4 (1) , 1-10.
McGuire, C. H., Solomon, L. M., & Bashook, P. G. Construc
tion and use of written simulations. New York: The
Psychological Co., 1976.
McKeachie, W. J. Student centered vs. instructor centered
instruction. The Journal of Educational Psychology,
1954, 45, 143-150.
McKeachie, W. J. Research on teaching at the college and
university level. In N. L. Gage (Ed.), Handbook of
research on teaching. Chicago: Rand McNally Co., 196 3.
McKeachie, W. J. Research on college teaching: A review.
Report 6, 1970. (ERIC Document Reproduction Service
No. ED 043-789)
McKeachie, W. J. Psychology in America's bicentennial
year. American Psychologist, December 1976, pp. 819-833.
McKeachie, W. J. Teaching tips: A guideline for the
beginning college teacher (7th ed.). Lexington, MA:
D. C. Heath Co., 1978.
McKeachie, W. J., & Kulik, J. A. Effective college
teaching. In F. N. Kerlinger (Ed.), Review of educa-
tional research. Itasca, Illinois: F. E. Peacock,
1975, pp. 165-209.
203
McLaughlin, F. E. Nurses and physicians performance on
clinical simulation test: Hypertension. Research in
Nursing and Health, 1979, 2^, 61-72.
McLaughlin, F. E. Selected psychometric properties of
two clinical simulation tests. Journal of Medical
Education, 1980, 55_, 375-376 .
McLeish, J. The lecture method. In N. L. Gage (Ed.),
The psychology of teaching methods. University of
Chicago Press, National Society for the Study of Educa
tion , 19 76.
McIntyre, H. M., MeDona1, F. J., Bailey, J. T., & Claus,
K. K. A simulated clinical nursing test. Nursing
Research, 1972, 2JL(5) , 429-435.
McMullan, W. E., & Cahoon, A. Integrating abstract concep
tualization with experiential learning. Academy of
Management Review, 1979, £(3), 453-458.
Miller, G. E. (Ed.). Teaching and learning in medical
education. Harvard University Press, 1961.
Myers-Briggs, I. Introduction to type. Swarthmore, PA:
1970.
Myers-Briggs Type Indicator. Princeton, NJ: Educational
Testing Service, 1962.
Netsky, M. G., Banghart, F. W. , & Hain, J. D. Seminar vs.
lecture, and prediction of performance by medical
students. Journal of Medical Education, 1964, 39_(41) ,
2-9.
Neufeld, V. R., & Barrows, H. S. The McMaster philosophy:
An approach to medical education. Journal of Medical
Education, 1974, 49_, 1040-1050.
Neufeld, V. R. Clinical reasoning of medical students--A
longitudinal study. Proceedings of the 16th Annual
Conference in Medical Education, Washington, D.C.,
1977, 351-355.
Nie, N. H. Statistical Package for the Social Sciences
(SPSS). New York: McGraw-Hill, 1975.
Olmstead, J. A. Small group instruction, theory and prac
tice . Human Resource Research Organization (HUNRRO),
Alexandria, Virginia, 1974.
204
Page, G. G. Written simulation nursing. Journal of
Nursing Education, 1978, 17(4), 28-32.
Page, G. G., & Fielding, D. W. Performance on PMPs and
performance in practice: Are they related? Journal
of Medical Education, 1980, 55, 529-537.
Palva, I. P. Measuring clinical problem-solving. British
Journal of Medical Education, 1974, £, 52-56 .
Peterson, P. L. Interactive effects of student anxiety,
achievement orientation, and teacher behavior on stu
dents' achievement and attitudes. Journal of
Educational Psychology, 19 77, 69_(6), 779-792.
Peterson, P. L. Aptitude treatment interaction effects of
teacher structuring and student participation in college
instruction. Journal of Educational Psychology, 1979,
71 (4) , 521-533.
Peterson, P. L., & Janicki, T. C. Individual characteris
tics and children's learning in large-group and small-
group approaches. Journal of Educational Psychology,
1979, 71(5), 677-687.
Peterson, P. L., Janicki, T. C., & Swing, S. R. Aptitude-
treatment interaction effects of three social studies
teaching approaches. American Educational Research
Journal, 1980 , 3/7 (3), 339-360 .
Plovnick, M. S. Primary care! career choices and medical
student learning styles. Journal of Medical Education,
1975, 5£ (9) , 849-855.
Porteus, A. W. Teacher-centered vs. student-centered
instruction: Interactions with cognitive and motiva
tional aptitudes. (Doctoral dissertation, Stanford
University, 1976). Dissertation Abstracts International,
1976, University Microfilms No. 77-7148)
Randolph, W. A., & Posner, B. Z. Designing meaningful
learning situations in management: A contingency,
decision tree approach. Academy of Management Review,
1979, £(3), 459-467.
Rezler, A. G., & French, R. M. Personality types and
learning preferences of students in six allied health
professions. Journal of Allied Health, 1975, £, 2CH26.
20 5
Rogers, J. C., & Hill, D. J. Learning style preferences
of bachelor's and master;s students. Paper presented
at the 59th annual conference of American Occupational
Therapy Association, 1979.
Sadler, G. R,, Plovnick, M. , & Snope, F. C. Learning
styles and teaching implications. Journal of Medical
Education, 1978, 5_3^ 847-849 .
Salomon, G. Heuristic models for the generation of apti
tude treatment interaction hypothesis. Review of Edu
cational Research, 1971, 42^(3), 327-343.
Salomon, G. Communication and education: An interactional
approach. A book to be published by Sage Publications,
Beverly Hills, 1981.
Sedlacek, W. E., & Natress, L. W. A technique for deter
mining the validity of patient management problems.
Journal of Medical Education, 1972, 47, 263-266.
Sharan, S., & Sharan, Y. Small group teaching. Englewood
Cliffs, NJ: Educational Technology Publications, 1976.
Shaw, R., & Bunt, M. M. Behavior of college students when
matched or mismatched for course structure. The
Journal of Educational Research, 1979, 13_, 41-45.
Shuell, T. J. Learning theory, instructional theory and
adaptation. In R. E. Snow, et al. Aptitude, learning
and instruction (Vol. 2). Hillsdale, NJ: Lawrence
Erlbaum Associates, 1980.
Shumway, J. M., & Donahue, J. F. Examination of the
. effects of structured small group format on medical
students' problem-solving performance. Conference of
Medical Education Research, 1980, pp. 215-220.'
Skakun, E. N. A preliminary investigation of computerized
patient management in relation to other examinations.
Educational and Psychological Measurement, 1979, 39,
303-310.
Smith, D. G. College classroom interactions and critical
thinking. Journal of Educational Psychology, 1977,
69(2), 180-190.
'206
Snow, R. E. Aptitude treatment interactions and individual
ized alternatives in higher education. In S. Messick
(Ed.), Individuality in Learning. San Francisco:
Jossey-Bass, 1976. (a)
Snow, R. E. Research on aptitude for learning: A progress
report. In Review of Research in Education. Itaska,
Illinois: F. E. Peacock, 1976. (b)
Snow, R. E. Individual differences and instructional
theories. Paper presented at the American Educational
Research Association, New York, 1977.
Snow, R. E. Aptitude, learner control, and adaptive
instruction. Educational Psychologist, 1980, 15,
151-158.
Snow, R. E., Federico, P. A., & Montague, W. E. (Eds.).
Cognitive process analyses of aptitude. Aptitude,
learning and instruction (Vol. 1). Hillsdale, NJ:
Lawrence Erlbaum Associates, 1980).
Stanford, G., & Roark, A. E. Human interactions in educa
tion . Boston: Allyn & Bacon, 1974.
Stovall, T. F. Lecture vs. discussion. Phi Delta Kappan,
1958, pp. 255-258.
Strom, B., College, K., & Hocevar, D. Interactive effects
of student satisfaction and course structure on student
achievement. Paper presented at the Conference of the
American Educational Research Association, Los Angeles,
1981.
Strom, B., & Hocevar, D. Influence of course structure
on student affect: The structural affect hypothesis.
Paper presented at the Conference of the American
Educational Research Association, Boston, MA, 1980.
Strumpf, S. A., & Freedman, R. D. The learning style
inventory: Still less than meets the eye. Academy
of Management Review, 1981, 6_ , . 297-299.
Terman, L. M. Concept Mastery Test (CMT) - Form T. Manual
1956 with a supplement 19 73, New York: The Psycho
logical Corporation.
20 7
Tobias, S. Achievement treatment interactions. Review of
Educational Research, 1976, 46_, 61-74.
Vu, N. V. A discriminating, teaching and predicting
medical problem-solving: A review. Evaluation and the
Health Professions, 1980, _3, 435-459.
Wallen, N. E., & Campbell, M. L. A. Vocabulary and non
verbal reasoning components of verbal analogies tests
(Miller Analogies Test and Concept Mastery Test). The
Journal of Educational Research, 1967, 6_1(2) , 87-89.
Ways, P. O., Loftus, G., & Jones, J. M. Focal problem
teaching in medical education. Journal of Medical
Education, 1973, 48^, 565-571.
Witkin, H. A. Cognitive style in academic performance and
in teacher student relations. In S. Messick (Ed.),
Individuality in learning, San Francisco: Jossey-Bass,
1976.
Wolf, D. M., & Kolb, D. A. Career development, personal
growth and experiential learning. In D. A. Kolb et al.
Organizational psychology (3rd ed.). Englewood Cliffs,
NJ: Prentice-Hall, 1979.
Wunderlich, R., & Gjerde, C. L. Another look at learning
style inventory and medical career choice. Journal of
Medical Education, 1978, 5_3, 45-54.
Yalow, E. Aptitude-treatment interactions with verbal and
figural instructional supplements. Conference of the
American Educational Research Association, Los Angeles,
1981.
Yerxa, E. J. The occupational therapist as consultant and
researcher. In Hopkins and Smith (Eds.), Occupational
Therapy. Philadelphia: J. B. Lippincott, 1978.
Zimmerman, J. M., & King, T. C. Motivation and learning
in medical schools III: Evaluation of the student
centered group. Surgery, 1963, 54, 152-156.
• 208
APPENDIXES
*20.9
APPENDIX A
INFORMED CONSENT FORM
210
COPY
The study in which you are asked to participate deals with
"instructional effectiveness in occupational therapy education."
The principal investigator will be Noomi Katz, MA, OTR. The study
is part of the investigator's PhD dissertation for the school of
education at USC.
Participants will be asked to take a few short group tests
during approximately two hours to permit the investigator to
obtain their O.T. GPA; and for one section of the course in
psychiatric O.T. (484) which will last for three weeks during the
spring semester 1981, participants will be randomly assigned to
one of three groups, and at the end will take an exam on the subject
matter taught.
Participation in this study will not change significantly the
usual learning requirements or schedule, and it will not result in
any discomfort or risk. Participants are free to withdraw their
consent and discontinue their participation in this study at any
time without affecting their education, but will be still required
to finish the course and take the exam for their grade in this
section.
All information collected will be confidential and no one
will be identified by names. Social security numbers and a file
number will be used for identification. Names of participants will
not be used in case of publication of any report of the data.
On the other hand, by participation in the study you may
benefit from the first hand experience in a research project. The
findings will be presented to you after the data analysis, and
elaboration of the research objectives and methodology will be
provided. Thus, participation may be beneficial to you in your
future studies and research efforts in occupational therapy.
The investigator will be very thankful for your participation.
Signed: Participant
Date
Witness
2ii
APPENDIX B
STUDENTS RATINGS OF SITUATIONS THAT
FACILITATE THEIR LEARNING
C O P Y
212
Social ’ .Security No.
INSTRUCTION
Each of the 16 learning situations listed on the next
page may he considered by you as facilitating your learning,
and therefore beneficial to you. You may be indifferent
to them. Or, on the contrary they may seem detrimental
to your learning.
Using the scale on the next pate, rate each item
according to the degree to which you feel it would promote
learning for you. CIRCLE the appropriate number beside
each item.
Turn now to the next page.
213
1--Would be Extremely Bene fic i a1
2 - Would be Generally Bene fie ia1
3--Would be Slightly Beneficial
4--Would make No Difference at all
5--Would be Slightly Detrimenta1
6-Would be Generally Detrimental
7 - Would be Extremely Detrimenta 1
Lectures .1 2 3 4 5 6 7
Seminars 1 2 3 4 5 6 7
Case Study 1 2 3 4 5 6 7
Small Group
Discussions, 1 2 3 4 5 6 7
Readings on
Theory 1 2 3 4 5 6 7
Projects 1 2 3 4 5 6 7
Term Papers 2. 2 3 4 5 6 7
Exercises &
Simulations 1 2 3 4 5 6 7
Faculty Feedback 1 2 3 4 5 6 7
Student Feedback 1 2 3 4 5 6 7
Summaries 1 2 3 4 5 6 7
Thinking Alone .1 2 3 4 5 6 7
Exams
1 2 3 4 5 6 7
Ind. Fclty Conf. 1 2 3 4 5 6 7
Talks by Experts 1 2 3 4 5 6 7
Homework 1 2 3 4 5 6 7
214
Correlations Between Learning Style Inventory Scores and
Student Ratings of Situations that Facilitate
Their Learning
CE RO AC AE • AC-CE AE-RO
Lectures .04 . 18** -.05 -.25** -.04 -.24**
Seminars .01 -.07
o
i
-.01 -.08 -.04
Case Studies -.03 -.12 # 22**
.09 .09 .11
Small Group Disc. -.05 .07 -.05 .16* .01 .14*
Readings on Theory -.21** .00 .34** -.13 .31** -.06
Projects .08 -.09 .03 .18 -.02 .15*
Term Papers .13 -.09 -.03 -.13 -.07 -.10
Exer. & Simula. .13 -.05 -.15 .10 -.12 .06
Faculty Feedback .00 .00 -.06 .09 -.02 .02
Student Feedback .13
00
o
i
-.11 . 20** -.14* .13
Summaries .03 .07 -.06 -.03 -.06 -.05
Thinking Alone .00 . 04 .17* -.07 .08
00
0
1
Examinations .05 .10 -.11 .02 -.10 -.07
Ind. Fclty Confs. .00 .05 -.11 -.03 -.06 -.07
Talks by Experts .06 .03 -.15* .02 -.15 -.01
Homework .10 .01 -.08 .19* -.12 .11
Note. From ’’ Learning Style Inventory, Technical Manual” by Kolb, D.A.
Boston; McBer & Company, 1976.
*p< *05
**p ^ .01
***p <:.001 2 tailed test
215
APPENDIX C
PATIENT MANAGEMENT PROBLEM
216
Diagram of First PMP
The role of the examinee is an occupational
therapist on the ward
Opening Scene
Section A
initial data
gathering
(options 1-5)
Bridging (opeions 6, 7)
Section C
Expert Advice
(.option 18-20)
Section B
continue data gathering
interview
(options 8-14)
Section E
^ initial placement U--
in treatment group
Copt ions 21-26) .
Section D
Expert Advice
(option 15-17)
/
/
/
JL
Section F
objectives and observations
in initial treatment group
(options 27—32)
/ Section G
next step in treatment
Response #39
right conduct
END OF PROBLEM
Response #38
wrong conduct
END OF PROBLEM
Note. solid arrows identify the optimal route.
217
Diagram of the Second PMP
The role of the examinee is an occupational
therapist leading a basic craft group.
r
±
Operi^rig Scene N
'.and craft description
Section C
correction of
preparation (8,9)
e
x
Section E
conducting with
wrong preparation
(13,14)
j --------
Section D
direction
group (.10-
to
-12)
Section A
preparation of
material for the
group (options 1-3)
V
•
v
Section B (1&2)
on—
4-7)
conducting the sessi
direction (options
\
/
Section I
conducting— second step
(options 38, 39)
Section H
observation and
reporting of cognitive
level (options 33-37)
END OF PROBLEM
Section J
observation and appli
cation of cognitive
levels Coptions 40-43)
Section F
observation and reporting
of cognitive level
Copt ions 18-32)
v
Section G
recommendation of treatmenl
for a new patient (options
15-17) END OF PROBLEM
Note. solid arrows identify the optimal route.
213
C O P Y
Social Security Number
COGNITIVE DISABILITY
Direction for the "Patient Management Problem,1 1
a Simulated Case Study ____ ' • : '
In each section of the problem there is a list of possible
items to select. Be sure to follow the instruction at the
beginning of each section, regarding the number of choices
allowed. Read each item in the section and then select
the appropriate one(s). After you have selected an item
"develop" the latent-image response with the special pen
by rubbing gently across the number you chose. The response
will give you the information you requested, present you
with the results of your actions, or direct you to the
appropriate next section of the exercise. Be sure to
develop the entire response; the end is indicated by an A.
When you receive a response that says END OF PROBLEM you
have finished the exercise.
219-
STMULATED CASE STUDY
You are an occupational therapy student at USC-LAC ;Medical
Center, Which is a public, in-patient psychiatric hospital.
It is the third week of your field work experience and
your supervising therapist has just asked you to assume
responsibility for a newly admitted woman, Mary Gonsales.
Neither you nor the supervising OTR know anything about
the patient beyond her age which is thirty-five years
old.
TURN TO SECTION A
v 2 20
Section A — Choose as many as you wish
1. Ask a nursing attendant to point
the patient out to you. 1.
2. Ask the patient*s MD to describe
the patient. 2.
3. Ask the OTR for information
about the patient. 3.
4. Read the medical chart. 4.
5. Ask the head nurst about the
patient's behavior on the ward. 5.
You would choose now only one.
6. Conduct an interview with the
patient. 6.
7. Attend staff rounds tomorrow in
order to gather more information. 7.
SECTION A — Data Gathering — Choose as many as you wish
1. Ask a nursing attendant to point
the patient out to you. 1 . A Spanish—Amer.
woman in isolated
in a corner of the
day room looking
at a magazine.
Clothing is slight
ly disheveled,A
2, Ask the patient * s M.D to
describe the patient 2. M.D. is in a
hurry and says the
patient will be
discussed in staff
. rounds tomorrow.A
3, Ask the OTR for information
about the patient.
3. OTR has already
told you she does
not know anything
and is annoyed. A
221
4. Read the chart
5. Ask the head nurse about
patient's behavior on the
ward.
You would choose now only one.
6. Conduct an interview with
the patiend.
7. Attend staff rounds
tomorrow in order to
gather more information.
4. Patient brought to
hospital by husband who
says she hasn't been
eating or sleeping well.
In addition she has not
been attending to her
housework or her chil
dren. Initial admitting
diagnosis is Primary
Affective Disorder; Uni
polar Type, rule out
substance abuse,
schizophrenic reaction .A
5, Mary is a quiet
woman who has posed no
management problems. A
6 . Turn to Sec tion B .A
7. The staff wonders
why you don't know more
about the patient and
instructs you to con
duct an interview as
soon as possible.
Turn to Section B. A
r
222
APPENDIX D
STUDENT EVALUATION QUESTIONNAIRE
223
COPY
STUDENT EVALUATION
1. How many hours per class session did you use to
prepare for class?
0
1-2
j3—4
_4-5
5 or more
2. How many hours did you use to prepare for the exam?
__l-3
4-6
_7-9
10 or more
3. Evaluate the teaching approach which you participated
in. I found the teaching approach:
Extremely Beneficial
Generally Beneficial
Neut 2:8,1
Generally Detrimental
Extremely Detrimental
224
STUDENT EVALUATION cont.
4. Evaluate your personal preference for the teaching
approach you were in.
personal pleasure and satisfaction
neutral
personal discomfort and dissatisfaction
5. After each session were you confident that you under
stood the material?
_____ very confident
_reasonable confident
neutral
slightly uncertain
very uncertain
6. Did your degree of confidence change between the first
class session and the exam?
confidence increased a lot
confidence increased a little
no change
confidence decreased a little
confidence decreased a lot
225
STUDENT EVALUATION cont.
7. Class assignments contributed to a comprehension of
the material:
jgreatly
_____some
neutral
_____ a little
_____ not much
8. Rank order the degree of fit between the teaching
approach and the class content (Rate 1 through 5)
Definitions of Cognition Disability
Cognitive Levels and Task Analysis
Assessment and Management of Change
Evaluation Instruments and Research
Conducting Task Groups and Reporting
226
APPENDIX E
CORRELATION MATRIX
227
LIST OF ABBREVIATIONS (USED IN THE MATRIX)
VA Verbal Ability
GPA Prior Achievement in O.T.
PBL Perceived Benefit from Group
ILS Individual Learning Style (Reflective/Active)
ACH Achievement
PMPIP Proficiency on the first PMP
PMP I E Efficiency on the first PMP
PMPICC Overall competence on the first PMP
PMP IIP Proficiency on the second PMP
PMPIIE Efficiency on the second PMP
PMP H OC Overra ll c ompe tenc e on the s econd PMP
ACHZ Overall course score, (a combined score)
STS Study Time for every class session
STE Study Time for Exams
AB Attitude toward ins.truction/Benefit
AE Attitude toward instruction/Enjoyment
AC Attitude toward instruction/confidence
OA Overall Attitude (a combined score)
PEARSON r CORRELATIONS BETWEEN ALL VARIABLES
FOR THE GROUP DISCUSSION METHOD (n=22)
VA GPA PBL PBG ILS ACH PMPIP PMPIE PMPIOC PMPIIP PMPIIE PMPIIOC
VA
1.00 .47 -.48 -.01 . .52 .16 ,14 .10 .53 -.16 -.'43 -.19
GPA 1.00 -.30 -.31 .29 .54 .39 .33 ,41 .26 .10 .26
PBL 1.00 .05 -.51 -.40 -.24 -.02 -.20 -.04 .08 . -.03
PBG 1.00 -.2.1 .18 ,13 .17 ,14 -.46 -.21 . ^.47
ILS 1.00 .02 ,17 .17 ,20 -.07 -.26 -.07
ACH 1.00 .41 .34 ,41 .20 .28 .23
PMPIP 1.00 .87 .99 .00 -.02 -.02
PMP IE 1.00 .91 ■ ' .05 -.02 .02
PMPIOC 1.00 ,00 -.02 -.02
PMPIIP 1.00 .78 .99
PMPIIE 1.00 .80
PMPIIOC 1.00
ACHZ
.06 .59 -.31. -.06 .07 .80 .67 .62 .67 .57 .51 .58
STS
-.56 -.47 .64 .25 -.76 -.28 -.25 -.25 -. 26 -.09 .25 -.07
STE
-.56 -.19 .30 .30 -.29 -.12 -.23 -.28 -.22 -.38 .02 -.36
AB.
.32 .23 .27 -.06 .00 -.01 .16 .28 -.26 .27 -.01 -.07
AE
.18 .11 .33 .03 -.07 -.02 .13 .26 .19 .25 .09 .26
AC
.03 .11 .22 -.12 .05 .01 .36 .39 .35 .49 .16 .47
OA
.20 .16 • .31 ,00 : . 24 .35 .36 -.04 -.01 .27 .38 .09
ACHZ STS STE • AB • AE.... AC OA •
.......
. . . . . . .
ACHZ 1.00 -.30 -.34 .21 .18 .40 ,30
STS
1.00 .52 -.03 -.01 -.10 -.05
STE
1.00 -.56 -.42 -.47 -.54
AB
1.00 .87 .64 .94
AE
AC
OA
1.00 .52
1.00
.91
.80
NJ
ro
1.00
00
Note. coeff ients. greater than .30 are significant
PEARSON r CORRELATIONS BETWEEN ALL VARIABLES
FOR THE LECTURE METHOD (n=22)
VA GPA ::EBL PBG ILS ACH PMPIP PMPIE PMPIOC PMPIIP PMPIIE PMPIIOC
VA 1.00 .09
.21 -.49. -.21 -.19 ,02 -.10 .00 .15 .19 . .13
GPA 1.00 -.16 -.34 -.14 .45 ,27 ,33 ,27 ,30 .35 .30
PBL 1.00 -.40 -.15 .07 .08 ,09 ,09 .13 .09 .15
PBG 1.00 ,39 .13 -,23 -.12 -.21 -.22 -.30 -.21
ILS 1.00 .28 -.24 -.14 -,24 -.12 -.20 -.12
ACH 1.00 .02 .07 ,01 .07 .12 .12
PMPIP 1.00 ,84 ,99 .14 .01 .11
PMPIE 1,00 ,89 ,19 .11 .16
PMPIOC 1.00 .15 .01 .12
PMPIIP 1.00 .84 .99
PMPIIE 1,00 .85
PMPIIOC 1.00
ACHZ -*.02 .55 .16 -.15
o
i
O'
o
.60 .60 .60 .65 .53 .66
STS
*-.35 -.14 .22 .03 .07 .35
CM
O
1
-.09 -.05 -.13 -.20 -.10
STE
-.50 -.03 .15 .20 .39 .52 .16 .17 .15
oo
1 —1
1
-.26 -.17
AB
.29 .25 .34 -.16 -.08 .40 .15 .13 .17 .05 .25 .10
AE
.47 .25 .27 -.28 .14 .20
CM
O
1
o
T— H
1
-.03 -.16 -.04 -.14
AC
.49 .27 .25 -.37 -.31 .31 .01 -.12 .02 .13 .16 .17
OA
.50 .30 .33 -.33 -.09 .34 .05 -.05 .04 .00 .13 .04
ACHZ
ACHZ STS STE AB AE AC OA
1 ; 00 .11 .27 .36 .01 ,27 .23
STS
1.00 .47 -.04 -.28 ,0.1 -.13
STE
1.00 -.09 .00 -.16 -.10
AB.
1.00 .60 .64 ,85
AE
1.00 .53 .85
AC
1.00 .85
to
to
OA VO
1.00
Note, coefficients greater than..30 are significant
230
APPENDIX F
LEARNING STYLE INVENTORY
2 31
T his in v en to ry is designed to assess your m ethod of learning. As you take the
in v en to ry , give a high rank to tho se w ords w hich best describe the w ay you learn an d a low
ran k to the w o rd s w hich least describe y o u r learning style.
You m ay find it h ard to choose the w ords th at best characterize your learning style.
N ev erth eless, keep in m ind th at th ere are no right or w ro n g answ ers —all the choices are
equally acceptable. The aim of the inventor}' is to describe how you learn, n o t to evaluate
y o u r learn in g ability.
INSTRUCTIONS
T here are n in e sets, of four w ord s listed below . Rank ord er each set of four w ords
assig n in g a 4 to the w ord w hich best characterizes your learning style, ? 3 to the w ord
w hich next b est characterizes your learning style, a 2 to the next m ost.characteristic w ords,
a n d a 1 to the w ord w hich is least characteristic of you as a learner. Be sure to assign a different
rank number to each of the four words in each set; do not make ties.
! ___ discrim inating j ___ tentative ! ___ involved
1 1
___ practical j
j ___ receptive ; ___ relevant ! ___analytical i __impartial j
I___ feeling i ___ watching ' ___ thinking ;
! ■ !
-----ooing j
i ___ acceoiihe j
j . V.) •
___ risk-taker j ______ evaluative j
j 1
! ___ aware f
.... - — .
1___ intuitive 1
... !
___ productive | ----- logical | ___ questioning ;
; - — abstract j
i 1
___ observing 1 ___ concrete j
1
___ active j
j ___ present-oriented j ___ reflecting
i
i ___ future-oriented 1
i !
___ pragmatic j
! - ;
I___ experience j ___ observation j ___ conceptualization ' ___ experimentation j
! !
i ______ intense ; ______ reserved i ______ rational ;
i J
______ responsible j
Copyright5 David A. Kolb, 1976
Published by McBer and Company
2 3 2
I n s t r u c t i o n s
T here are nine sets of four w ords listed below . Ranh o rd er the w ords in each set by
assigning a 4 to th e w ord w hich best characterizes your learning stele, a 3 to the w ord
w hich next best characterizes y o u r learning style; a d to the next m ost characteristic w ord,
an d a 1 to the w ord w hich is least characteristic of you as a learner.
You mav find it h ard to choose the w ords that best characterize vour learning stole.
N evertheless, keep in m ind th at there are no rig h t or w rong answ ers —ail the choices are
equally acceptable. The aim of the inventory is to describe how you learn, not to evaluate
y o u r learning ability.
Be sure to assign a diffei
make ties.
a it rank number to each of the four words in each set; do not
1 . _ _ discriminating — tentative ___involved _ _ practical
2. _ _ receptive _ _ relevant ___analytical ___impartial
3. _ _ feeling _ _ watching ___thinking ___doing
4. _ _ accepting __risk-taker _ _ evaluative . __.aware
___intuitive ----productive ___logical ___questioning
6. _ _ abstract _ _ observing ___concrete _ _ active-
7. __.present-oriented _ _ reflecting ___future-oriented ___pragmatic
S. _ _ experience _ _ observation ___conceptualization___experimentation
9. _ _ intense _ _ reserved ___rational _ _ responsible
S c o r i n g
The four colum ns of w ord s above correspond to the four learning style scales:
CE, RO, AC, an d AE. To com pute vour scale scores, w rite yo u r rank num bers in the
boxes below only for th e desig n ated item s. For exam ple, in the third colum n (AC), you
w o uld fill in the rank n u m b ers you have assigned to item s 2, 3, 4, 5, 8, and 9. C om pute
y o u r scale scores by ad d in g the rank num bers for each set of boxes.
Score items:
2 3 4 5 7 8 1
Score items: Score items: , Score items:
3 6 7 8 9 2 3 4 5 8 9 1 3 6 7 8 9
r i i i i 11 r i .1 . i 1 1 1 S i ! 1 I ; 1 M !• 1 1 : !
CE = RO = ...... AC = ' AF «... _
To co m p u te the tw o com bination scores, subtract CE from AC and subtract RO from
AE. P reserve negative signs if they appear.
AC
r —!
AC-CE:
CE AE RO
f i F l i i
-i J= AE-RO:|_J-L>_ _
233
LEARNING STYLE TYPE GRID
P e r c e n t i l e s
0 - t -12
1
I
-id
s
-7
1 -6
1
-4
1
-3
2 0 4
i
_ 1
i
J
i
Accom modator
-1
30 - r
S
0
1
►
1
40 j
•Diverger
0 - 1 - - 2 - 3 - 4 1 1
. f 7 1 5 1 3 1 2 1 1 1 0 9 8 7 6
Assimiiator
Converger
10 o
Percend'.es
Copyright® David A. Kolb, 1976
Published by McBer and Com pany
234
LEARNING STYLE PROFILE
N orm s for the L earning Style In v en to ry
' Concrete
Experience
60%
70
/Active j
Experimentation.
1 6 j 1 7 1 9 2 2 j
10Q%
A b stract
Concern tu alizatio n
Copyright5 ' ' David A. Kolb, 1976
Published by McBer and Company

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

An occupational therapy study of cognition in adult inpatients with depression

PDF

The effects of analogical instruction on the transfer of procedural knowledge and its interaction with learner ability

PDF

The interactive effects of arousal level, cognitive style, and color on affective learning

PDF

The effects of aptitudes and structurally different methods of teaching mathematics on achievement and satisfaction

PDF

Student nurse learning styles based on experiential learning theory from three types of nursing programs

PDF

Effects of sex, professional status, and attitudes toward women on female student nurses' reactions to clinical professionals' success and failure

PDF

The interaction of response modes and learner aptitudes on Newtonian physics achievement: A computer-based interactive videodisc exploratory study

PDF

The curriculum classification of junior college students

PDF

Interaction effects of introversion/extraversion and goal structure on performance and attitude

PDF

The effects of differential instruction treatments on the problem identification performance of field dependent-independent nursing students

PDF

The effects of some master learning techniques on achievement scores and mastery rate of junior college students

PDF

The effectiveness of problem-based learning in nurse practitioner education

PDF

A study of the effects of presenting literature to first grade students by means of five visual-verbal presentation modes

PDF

The influence of nursing curriculum and social factors on nursing students' choice of rural practice in Thailand

PDF

Assessment of selected learning outcomes of an experiential education program

PDF

The effectiveness of the meaning method in learning arithmetic

PDF

An experimental study of the effect of a listening skills training program in kindergarten on the development of pre-reading skills

PDF

An investigation of aptitude-treatment interaction between student-preferred learning modality and presentation method

PDF

The interactive effects of cognitive style and selected instructional strategies on a complex psychomotor skill

PDF

An investigation of the use of calculators with low achieving 4th grade students in mathematics achievement and attitude

Asset Metadata

Creator Katz, Noomi (author)

Core Title The interactive effects of occupational therapy students' learning style with teaching methods (lecture vs. group-discussion) on their problem-solving skills, achievement, study time and attitude...

Contributor Digitized by ProQuest (provenance)

Degree Doctor of Philosophy

Degree Program Education

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag education, curriculum and instruction,Health Sciences, Occupational Therapy,OAI-PMH Harvest

Language English

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c26-490098

Unique identifier UC11246025

Identifier usctheses-c26-490098 (legacy record id)

Legacy Identifier DP24781.pdf

Dmrecord 490098

Document Type Dissertation

Rights Katz, Noomi

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the au...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus, Los Angeles, California 90089, USA