“ LIBRA R Y
MiChIgan State
University

      

g Illlll ll l lllll lll ll ll ll‘llllcll

3 1293 10154 287

This is to certify that the

thesis entitled

EFFECTS OF A VISUAL ADVANCE ORGANIZER ON THE
VISUAL-LEARNING DISCRIMINATION PERFORMANCE
OF PERSONS OF VARYING CEREBRAL
HEMISPHERIC FUNCTIONING

presented by

John L. Phelps

has been accepted towards fulfillment

of the requirements for
Instructional Development
Ph.D. degree in and Technology/Education

 

 

 

r7 . * ,
xké/Zr/Zlaééz‘g“

Major professor

Dan: November 18, 1977

 

0-7 639

 

EFFECTS OF A VISUAL ADVANCE ORGANIZER ON THE
VISUAL-LEARNING DISCRIMINATION PERFORMANCE
OF PERSONS OF VARYING CEREBRAL
HEMISPHERIC FUNCTIONING

By.

John L. Phelps

A DISSERTATION

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

DOCTOR OF PHILOSOPHY
Area of Instructional Development and Technology

1977

ABSTRACT

EFFECTS OF A VISUAL ADVANCE ORGANIZER ON THE VISUAL-
LEARNING DISCRIMINATION PERFORMANCE OF PERSONS OF
VARYING CEREBRAL HEMISPHERIC FUNCTIONING

By
John L. Phelps

The purpose of this study was to investigate (l) the efficacy
of a visual advance organizer compared to a verbal advance organizer,
(2) the differential learning performance of persons who varied in
vividness of imagery, and (3) the interaction of type of advance
organizer and vividness of imagery on learning performance. The
independent variables were type of organizer (visual, verbal, control)
and level of imagery (high and low). The dependent variable was dis-
crimination learning as measured by a learning stimulus posttest.

One hundred fifteen undergraduate and graduate students in
various disciplines at Michigan State University participated in the
study. Seventy-two subjects with complete sets of data were cate-
gorized as high and low imagers using Boutwell's Noun Test and

Sheehan's QMI. Of these, eleven low-imagery and seventeen high imagery

. subjects were identified; forty-four subjects could not be categorized

and were deleted.

Subjects were randomly assigned to one of three treatments:
Visual advance organizer, verbal advance organizer or control group.
The visual advance organizer consisted of a series of six photographs

showing Art Nouveau objects in relation to other art objects. The

John L. Phelps

verbal advance organizer described the relationship of these objects.
The control treatment was written material unrelated to Art Nouveau.

A learning stimulus using projected slides and controlled
presentation rate was used. A learning stimulus posttest was designed
that required subjects to visually identify examples of Art Nouveau.

The hypotheses tested were that scores on the posttest would
(1) be greatest for subjects in the visual advance organizer treatment,
(2) be greatest for high-imagery subjects, and (3) indicate an inter-
action between type of advance organizer and imagery level.

The procedure consisted of (l) administering Boutwell's Noun
Test and Sheehan's QMI, (2) randomly assigning subjects to three treat-
ment groups, administering (3) the treatments, (4) the learning
stimulus slides, (5) the posttest, and (6) the Sherman-Kulhavy
Laterality Test.

A 3 x 2 factorial design was used. The primary statistic was
a two-way analysis of variance. Additional statistics included corre-
lation and factor analysis. The statistical significance level was
(set at .05. Analysis of the data yielded these findings.

There were no significant differences in the posttest scores
associated with (l) the three treatments: visual advance organizer,
verbal advance organizer, and control group; (2) imagery level;

(3) the interaction of treatment and imagery level.

In addition, correlation and factor analysis indicated that
the three imagery instruments were nonoverlapping; the correlation
between instruments was low and each instrument yielded strong and
separate factors.

The hypotheses, therefore, were not supported.

John L. Phelps

Since the imagery scores correlated only to a small extent
and resulted in separate and strong factors, the imagery measure had
low construct validity in this study. Also, the instructional treat—
ment, art-style slides, represented a sample of one; other treatments
should be tested. Further, for the two-way analysis of variance the
small cell size resulted in a situation where treatment differences of
great magnitude would have been needed to make the advance organizers
treatment variable statistically significant. Finally, the small
differences in the dependent variable may have been due to the rela—
tively small amount of learning that took place, or perhaps were a
result of the short experimental treatment and learning stimulus.

Other studies, if warranted, should consider including such
variables as instructional treatments, amount of learning, and length
of treatment. Imagery measures should continue to be studied with the

aim of increasing construct validity.

ACKNOWLEDGMENTS

First, I am indebted to my family, Patricia and Michael, for
their patience during my many absences, their tolerance of my many
demands, and their very real support and assistance during all the
projects that make up a dissertation study.

Second, I gratefully acknowledge the guidance of Dr. Kent
Gustafson, who chaired the doctoral committee. His steady hand and
effective leadership contributed greatly to this learning experience.

Finally, I want to thank the other members of the committee
for their time, support and critical review: Dr. Norman Bell,

Dr. John E. Hunter, and Dr. James Nord.

ii

TABLE OF CONTENTS

LIST OF TABLES . . . . . . .

LIST OF FIGURES.

Chapter
I.

II.

III.

IV.

IDENTIFICATION OF THE PROBLEM.

Introduction. . . . . . .. .
Purpose . . . . . . . .
Importance . . . . . . .
Generalizability . . . . .
Hypotheses . . . . .

REVIEW OF LITERATURE.

Stimulus Aspects . .

Summary of Preinstructional and Advance Organizer

Literature. . . . .
Advance Organizers as Stimuli .

Summary of Stimulus Aspects of Advance Organizers

and Pictures . . . .
Learner Characteristics .

Summary of Learner Characteristics Related to

Advance Organizers and Pictures.
Situation Characteristics . . .
Summary of Literature Review . .

DESIGN OF THE STUDY . . . .

Overview of the Chapter .

Research Questions. .

Research Hypotheses for This Study
Definitions of Important Terms.
Design of the Study

Procedures for the Study.

Chapter Summary. . .

ANALYSIS OF THE DATA.

Findings of the Study.
Chapter Summary. . . . .

iii

Page

vi

—J

0301th

Chapter
V.

CONCLUSIONS AND IMPLICATIONS FOR FURTHER RESEARCH.

Sumnary . . . . .
Conclusions. . . . .
Discussion . .

Implications for Future Research.

APPENDICES
Appendix
A. Learning Stimulus Posttest Audio-Taped Instructions .
Description of Test Slides . .
B. Learning Stimulus Audio- -Taped Instructions . . .
Description of Learning Stimulus Slides and Slide
Sequence. . r
C. Visual Advance Organizer Treatment.
D. Verbal Advance Organizer Treatment.
E. Control Group Treatment . . . . . . . . .
F. Fifty Noun Test words with Means, Standard Deviations

and High Words/Excluded Words/Low Words Grouping

REFERENCES . . . . . . . . . . . .

iv

86
87

88
89
91
93
94

95
97

Table

LIST OF TABLES

Page
The Two Modes of Consciousness: A Tentative Dichotomy.' . . 32

Cell Sizes, Means and Sex Distribution--Two-way
Analysis of Variance . . . . . . . . . . . . . 7l

Two-Nay Analysis of Variance Table for Learning
Stimulus Posttest Scores. . . . . . . . . . . . 72

Variables Used for Each Subject in Correlation
Procedure. . . . . . . . . . . . . . . . . 73

Correlation of Variables . . . . . . . . . . . . 74

Principal Factors and Highest Factor Loadings Related
to Variables. . . . . . . . . . . . . . . . 75

LIST OF FIGURES

Figure

01 U1 3 (A) N
o o o o 0

Two Dimensional Stimulus Matrix . .
Three Dimensional Stimulus Matrix

Visual Fields Related to Hemispheres

Scale Used in Learning Stimulus Posttest .

Art Style Picture Combinations
Timing of Projected Slides.
Variable Matrix . . . . .

Flowchart of Categorization Procedure .

vi

Page
T7
22
29
53
53
55
6l
68

CHAPTER I

IDENTIFICATION OF THE PROBLEM

Introduction

 

It is generally accepted by teachers and others in instruc-
tional settings that presenting material that is entirely new to
students does not facilitate learning. Overviews, summaries, outlines,
general remarks, and other means are often used to "bridge the gap'I in
introducing new material to students.

The research literature on preinstructional Strategies extends
back almost thirty years. As part of the Pennsylvania State College
Instructional Film Research Program, LathrOp (1949) studied the effects
of film introductions on learning from motion pictures. During the
same program, Stein (1952) studied the effects of pre-film tests on
learning from films. May and Lumsdaine (1958) found that the use of
pre-film test questions increased certain aspects of learning.

Interest in preinstructional strategies has continued in both films

and other strategies, from these early studies to the present time.

In a recent review of preinstructional strategies, Hartley and Davies

(1976) concluded that the meaningful-reception (i.e., other than

rote and discovery) learning situation is more likely to benefit

from these strategies than other types of situations (p. 258).
Advance organizers have been a major component of preinstruc-

tional strategies since the early 19605 when Ausubel first defined and

studied the concept (Ausubel, 1960). Lawton and Nanska (1977, p. 234)
describe advance organizers as consisting of ". . . concepts and
propositions that are maximally stable and discriminable from the
schematic arrangement of concepts and propositions existing in the
learner's cognitive structure, and are presented in advance of new
learning materials." Advance organizers are viewed as functioning

to ". . . induce, through a particular form of learning, organizing
and explanatory concepts, propositions, and principles" (Lawton and
Hanska, l977, p. 235). The organizer itself becomes a cognitive
structure during the initial learning process.

Research in the use of advance organizers has concentrated
almost completely on organizers in the yerbal_mode, either written or
spoken. In their review, Barnes and Clawson (1975) cite only two
studies using visual advance organizers. Considering the extensive use
of visuals in the instructional process and the different functions
that visuals serve, it would seem useful to attempt to extend knowledge
of how advance organizers Operate in visual instructional procedures.

Concurrently, a renewed interest has developed in mentalistic
concepts that include imaging, verbal processes and cerebral hemis-
pheric specialization. Recent work in this area indicates that the
familiar notion of the transverse relationship between the hemispheres
and the body (e.g., the left hemisphere controls the right side of the
body) extends to perceptual processes as well, each hemisphere
."specializing" in the processing of certain kinds of external stimuli.
This work is especially valuable as researchers begin to assess the

role of the right hemisphere in stimulus perception and processing.

Ornstein (1972) describes the basic nature of cerebral speciali-
zation:

Both the structure and the function of these two "half-brains"
in some part underlie the two modes of consciousness which
simultaneously coexist within each one of us. Although each
hemiSphere shares the potential for many functions, and both
sides participate in both activities, in the normal person
the two hemispheres tend to specialize. The left hemisphere
(connected to the right side of the body) is predominantly
involved with analytic, logical thinking, eSpecially in
verbal and mathematical functions. Its mode of operation is
primarily linear. This hemisphere seems to process informa-
tion sequentially. This [linear] mode of operation of neces-
sity must underlie logical thought, since logic depends on
sequence and order. Language and mathematics, both left-
hemiSpheric activities, also depend predominantly on linear
time.

If the left hemisphere is specialized for analysis, the right
hemisphere, again, remember connected to the left side of the
body seems specialized for holistic mentation. Its language
ability is quite limited. This hemisphere is primarily
responsible for our orientation in space, artistic endeavor,
crafts, body image, recognition of faces. It processes
information more diffusely than does the left hemisphere,

and its reSponsibilities demand a ready integration of many
inputs at once. If the left hemisphere can be termed pre-
dominately analytic and sequential in its operation, then

the right hemiSphere is more holistic and relational, and
more simultaneous in its mode of operation (p. 51).

Studies of cerebral specialization have yielded results con-
sistent with Ornstein's description. Kimura (1973) used a stereo tape
system to play different material to subjects' right and left ears
(each ear is connected predominantly to the opposite hemisphere).

With simultaneous melodies, Kimura found a left ear/right hemisphere
superiority in picking out tunes; she found a complementary right ear/
left hemiSphere superiority in processing certain human speech sounds.
Considering the visual field, one-half of each eye's visual field is
connected to the right hemisphere; the other half is connected to the

left hemisphere. Using visual tachistoscopic tests, Kimura (1973)

found a left hemisphere dominance for the visual perception of letters
and words. In a follow-up study to one of Kimura's earlier studies

. using dichotic listening materials (Kimura, 1964), Broadbent and
Gregory (1964) used the same techniques but different stimulus
materials and obtained the same results.

Considering the psychological constructs of visual and verbal
processing, Paivio (1971) views each as having a separate type of
coding system. Paivio associates visual imagery with parallel proces-
sing and verbal imagery with sequential processing (p. 9). Rather than
the global concept of hemispheric dominance, the notion of imagery
seems more amenable to measurement, and Paivio notes that the measure-
ment of imaginal versus verbal abilities as individual differences

constitutes a useful independent variable (p. 9).

Purpose

The present study was based on the two elements noted in the
Introduction: advance organizers in a visual mode, and individual
differences in hemispheric specialization as measured by imagery
abilities.

Specifically, this study had two purposes. First, the effects
of a verbal or visual advance organizer as a component of a learning
task that is primarily visual were examined. Second, the extent that
any differences in learning associated with each type of organizer are

functions of subjects' degree of imagery capability was examined.

Importance

 

This study is important for several reasons. First, consider-

ing how many visuals are used in instruction, little is known about

their functional properties, especially when used as advance organizers.
Second, it is important to extend the concept of advance organizers
to modes other than verbal learning since few studies to date have been
in other than verbal modes. Third, it is important to attempt to link
what is known about cerebral hemisphere differences to classroom tasks.
Rennels (1976) decries an educational system that emphasizes the linear
processing associated with the left brain to the neglect of the
capacity to imagine and visualize. He notes that schools have system-
atically eliminated these experiences, but they ". . . should become
concerned with developing equal qualities of cerebral functioning
in children" (p. 471).

Finally, it is important to attempt to consider individual
differences in crucial learning variables, not necessarily in the
sense of individualized instruction as an outcome of Specific trait-
treatment interactions, but as indicants of stimulus considerations

in the instructional design process.

Generalizability

 

The results of this study are limited. First, the under-
graduate and graduate subjects were drawn from available classes, not
randomly sampled from the university population..

Second, although the literature reviewed does not limit the
elements of visual imagery and cerebral specialization to projected
slides only, the single medium used represents a restricted sample of

one o

Hypotheses

 

The study is based on three hypotheses. First, it was hypothe-
sized that test scores associated with a visual advance organizer would
exceed those associated with a verbal advance organizer and a control
group.

Second, it was hypothesized that the scores of high imaging
subjects would exceed those of low imaging subjects.

Finally, a type of organizer by learner degree of imagery

interaction was hypothesized.

CHAPTER II

REVIEW OF LITERATURE

Using a suggested classification by Hovland et a1. (1949)
regarding independent variables, this review will be organized along
three dimensions. First, stimulus aspects of presentation materials
will be considered, including categories of preinstructional materials
generally, stimulus aspects of advance organizers, and general stimulus
characteristics of visuals or pictures. Second, learner character-
istics will be examined and related to advance organizers and some
aspects of cerebral dominance will be discussed. Third, situation
variables will be explored, such as the degree the learning content is

structured and length of treatment.

Stimulus ASpects

 

Introduction

 

The general concept of a "before instruction," or preinstruc-
tional treatment has been examined for a long time. Even the current
interest in visuals in preinstructional treatments can be traced back
to media research done in the armed forces through the period of World
War II. In a study of the effect of introduction to instructional
films on learners, Lathrop (1949) found that, with respect to

performance on a multiple-choice posttest of facts, high school

students' performance was not significantly different for two films,
while for a third, the introduction had a highly significant adverse
effect. 9

In a study from the same series (done at Pennsylvania State
College), Northrop (1952) varied the prominence of the organizational
outline within three kinds of instructional films sequences: discrete
item, logical, and chronological (dramatic). He found that a highly
prominent outline was most effective for the discrete item film.

Stein (1952) used a test of film learning in two conditions.
First as a pre-film test, where learners received total knowledge of
results. Then after the film, the same test was administered. Stein
found that this procedure was more effective than showing the film only
once, or showing it twice.

Although May (May and Lumsdaine, 1958, p. 91) obtained ambiva-
lent results when he experimented with attention directing techniques
using prefilm test questions, he did find indications that specific but
open-ended short-answer questions coupled with verbal directions
brelated to the content might be effective. May also concluded that
the relative importance of material in the film might be a key variable.
If all the material is of about equal importance, an overall pretest
is indicated, but if certain elements are more important, a shorter
pretest that focuses on these elements may highlight them.

Nulff and Kraeling (1961) compared two types of familiarization
techniques as part of a motion picture learning experience dealing with
the assembly of an automobile distributor. One group of subjects was
familiarized prior to the film showing. The subjects were shown the

various distributor parts, one at a time, in an irrelevant order. An

expert pointed out features of each part and handed each part to a
subject who named the important features. This group was then shown
the training film. A second group was exposed to familiarization
techniques gprjpg_the film. Filmed close-up inserts were prepared that
included visual and aural pointers. These inserts dealt with aspects
of the assembly process and were inserted just before the actual
assembly of the part. A third group watched the film with no treat-
ment. When the thirty-three subjects (eleven per group, aged 16-26
with little prior experience and mathced on the Minnesota Paper Form
Board scores) performed the task after the film, time and error
measurements were made. It was found that both familiarization proce-
dures improved performance and that the within-film procedure was
superior.

In addition to pretests, familiarization techniques and
organizational emphasis, other preinstructional procedures have been
studied, including behavioral objectives. overviews, and advance
organizers. In a review of preinstructional strategy studies, Hartley
(and Davies (1976) define and differentiate these procedures. The
authors, after examining instances of use for each strategy, concluded
that ". . . pretests alert, behavioral objectives jpjpgm, overviews
prepare, and advance organizers clarify" (1976, p. 246). Further, they
categorized the strategies in relation to the instructional experience.
Pretests and behavioral objectives were felt to have a list structure
and were given in advance of the instruction. Overviews and advance
organizers were described as more abstract prose formats, to be given

immediately prior to instruction.

10

Hartley and Davies (1976) describe advance organizers as being
more complex than overviews and having a unique function: to clarify
the task ahead. Advance organizers provide a broad framework, a con-
text, as differentiated from other preinstructional strategies.

Two types of advance organizers are described by Hartley and

Davies. First, expositogy_organizers are those generated and used when

 

new material is completely unfamiliar. When this is the case, material
familiar to the student is included in the organizer and it can be used
to make the learning task more understandable, because relationships
to existing knowledge is made clear.

When the material is familiar to some extent, the authors note

that comparative organizers are appropriate. These organizers operate

 

to differentiate the known and unknown information. This heightens
the contrast between the two and the resulting clarity reduces confusion
and some forgetting.

Hartley and Davies (1976) describe the functions of advance
organizers as:

(a) to provide ideational scaffolding for the more differ-
entiated learning task, or

(b) to increase the discriminability of the latter [learning

task] from related ideas in existing cognitive structuring,

or

(c) to effect integrative reconciliation at a level of

abstraction, generality and inclusiveness which is much

higher than that of the learning material itself

(p. 246).

The authors also suggest a reason for much of the difficulty

with research using advance organizers: since there are no procedures
for writing or creating advance organizers, it is very difficult to

detect the presence of an organizer.

11

The research evidence regarding advance organizers is almost
entirely related to the written or aural mode. These studies will be
examined after which a few non-verbal studies will be reported. Atten-
tion will then shift to stimulus characteristics of visuals or pictures,
and how these might relate to advance organizers.

Christie and Schumacher (1975) hypothesized that (1) subjects
not provided an advance organizer would try to generate their own;

(2) of these, older subjects would be more effective in creating
organizers and so, would be able to recall more; and (3) organizers
might have an effect on relevant versus irrelevant sentences. Pre-
senting a one-sentence organizer and a learning passage to six- and
nine-year olds aurally and analyzing audio-taped recall, they found
that (1) age was a significant factor, younger subjects not provided
an organizer apparently did not generate one; (2) subjects who did not
receive an organizer recalled more relevant than irrelevant information,
but subjects who did receive one recalled ppth_relevant apg_irrelevant
information; (3) older subjects did in fact generate organizers.
_Christie and Schumacher conclude that, while an advance organizer may
have no overall effect, its use may change the mixture of relevant

and irrelevant information learned.

Kennedy (1975) measured performance of learning and retention
of metric concepts. His treatments were a written historical intro-
duction that included various measurement concepts, a written advance
organizer, and a no-treatment control group. Kennedy examined per-
formance in terms of subject ability (grade-point average) and cogni-
tive structure (the number of science and math courses completed).

In addition to finding that high-ability students performed better and

12

that the cognitive strategy variable was not significant, Kennedy found
that after a thirty-day delayed posttest, students with the advance
organizer treatment performed significantly better.

In an examination of the advance organizer research over the
past fifteen years, Hartley and Davies (1976) describe the results as
confused. They do note a trend indicating that as studies become more
recent and have been conducted without some of the earlier methodo-
logical limitations, the results in favor of advance organizers has
diminished somewhat. They summarized the results as:

l. The majority of studies still indicate that advance

organizers facilitate both learning and retention, although
almost all the studies in this category were conducted with

undergraduate students as subjects;

2. Other studies have been unable to conclude that
advance organizers facilitate learning or retention;

3. The effects of organizers appear to be specific,
not general;

4. The effectiveness of advance organizers varies
with subject characteristics such as age and general
ability, but these effects are not stable;

5. Expository organizers may be most useful with
children who have low verbal and analytical ability;

6. Indications exist that formats other than con-
tinuous prose should be considered; and

7. Post organizers may sometimes have a greater
effect than advance organizers.

In another review of research, Barnes and Clawson (1975)
analyzed thirty-two advance organizer studies in terms of whether they
facilitate learning. When these studies were examined in terms of
frequency of significance, a pattern emerged that is similar to that
noted by Hartley and Davies: the relative frequency of nonsignificant

studies has increased in recent years.

13

As a final component of this section of the literature review
dealing with advance organizers, two studies will be reported which
used advance organizers in other than written form. Weisberg (1970)
studied the effects of two visual advance organizers on eighth grade
students as compared to a verbal passage (presented as an advance
organizer) and a no-treatment group. One visual advance organizer
consisted of a physiographic diagram of the North Atlantic Ocean floor.
The second visual advance organizer was a topographic profile of the
ocean floor. The verbal advance organizer was a five-hundred word
passage that dealt with the same material as the visual organizers.
The learning task was a 1,100-word passage dealing with continental
drift. Weisberg concluded that (l) the visual organizer groups did
better on the multiple-choice posttest than the written organizer or
control groups; (2) the visual advance organizers were most effective
with subjects who were in the middle level of prior knowledge (versus
high or low).

In a study where the advance organizer was a fabricated model
of a computer, Mayer (1976) examined the differences in performance
requiring concept transfer. The model was a three-by-two-foot board
with a small chalkboard, data cards, note pads, and index cards. In a
model versus no-model comparison, the subjects exposed to the model
also received a lSO-word written explanation; the no—model subjects
received the learning task (a 26-frame programmed text) directly.
Mayer found that the subjects exposed to the model and text were more
effective overall on a posttest, and also more effective in problems
requiring transfer. In a second study, Mayer found that the model used

as an advance organizer was more effective than as a post organizer.

14

Summary of Preinstructional and
Advance Organizer Literature

 

 

The strength inherent in this area due to the many studies and
period of time over which they were conducted is diffused when one
considers the lack of consistency and recent trends in results. One
reason may be the lack of an effort to Operationalize the advance
organizer especially with respect to stimulus characteristics. For
example, in the Weisberg and Mayer studies, both of which used visual
advance organizers, there was no attempt to specify the nature of the
visualizations, thus no meaningful comparisons are possible. In both
studies, other methodological problems notwithstanding, we are left
with indications of positive effects of visual advance organizers, but
also with no explanation of why this is so.

Concentrating on the stimulus aspects of organizers may provide
a high "pay off." The stimulus characteristics of media such as
pictures, motion or still, offer a degree of control and replication
and as such, greatly facilitate successful research on instruction
(Lumsdaine, 1963, p. 586). It should be noted that the term stimulus,
(as used here, embraces both characteristics of the medium itself, such
as the color, amount of detail, degree of motion or implied motion,
etc. agg_characteristics of the content of the medium, such as com-
plexity, level of abstraction, shape, size, and so on.

While not paralyzing research in this area, the lack of a con-
sistent Operational definition of advance organizers is a hindrance to
systematic study of their effects. Attending to stimulus attributes may

be partially effective but will not overcome the problem of definition.

15

Finally, the studies of advance organizers that use non-written
organizers (e.g., the Weisberg and Mayer studies) are noteworthy even
though not rigorous or consistent. They indicate that the basic notion
of advance organizers is not Specifically restricted to verbal learning

modes.

‘Advance Organizers as Stimuli

 

Turning more directly to the stimulus aspects of advance
organizers, literature will be examined regarding the nature and
organization of the stimulus content of media. Studies related to
sensory modality and realism versus abstraction will be discussed.
Finally, material concerned with the relative concreteness of stimuli
will be noted.

Several dimensions of the content of media are noted. Levie
and Dicke (1973) discuss sign types--specifically iggfli£_versus digital
sign types. Digital sign types imply an arbitrariness, such as words
or numbers. To understand a digital sign, one must "know the code,"
such as English letters, numbers, and so on. The elements of digital
signs have a temporal sequence; they are linear. Iconic signs resemble
the things they represent, to a greater or lesser degree, such as
photographs, maps, sculptures, and so on; in each case, "critical"
aspects of the real object are present. Iconic signs "picture" a
complex object.

Levie and Dicke (1973) made the following contrast in the two
sign types. Iconic signs can easily picture a complex object but are
difficult to use to communicate an abstract concept, like "mammal."

Iconic signs communicate knowledge pf_the world; digital signs

l6

communicate knowledge apppt_the world. Digital signs can be used to
make statements; iconic signs can only say "it exists." The authors
note Gropper's comment that pictures are poor for communicating
generalizations and qualifications; they cannot communicate the con-
cepts "all, many or most." Iconic signs, on the other hand, are related
to divergent thinking, hypothesis formation, and creative activity.

In a chapter on human information, McCormick (1964) defines
"loading" and "speed" as two elements of a visual display. Loading
refers to the variety, in terms of type and number, of stimuli to which
responses must be made. He defines spggg_as the number of stimuli per
unit time or the time available per stimulus.

Gagné (1973) describes the components of the stimulus situation

in learning. First he defines integral stimuli as task (the printing

 

of an equation in a test) or context (the organization of a graph) as
stimuli that are inherent to the medium used. Second, he describes
activating stimuli as those that activate previously established pro-
cesses or states. Examples of activating stimuli include (1) giving
directions ("Look at the picture on the left"), (2) activating recall
of previously learned capabilities, (3) activating a performance set,
and (4) arousing a motivational state. A third component of Gagné's
stimulus situation is stimuli from recall such as recalling cognitive
strategies that may provide directions for the learning event. Gagné's
structure may be useful in specifying or designing aspects of the
stimuli used in advance organizers. He does associate one type of
activating stimuli (activating recall) with advance organizers.

With regard to advance organizers, the stimulus characteristics

specified by Levie and Dicke, McCormick, and Gagné can be organized

17

into a matrix that might differentiate verbal from visual advance
organizers and differentiate advance organizers from non-advance

organizers. Such a matrix is given in Figure l.

 

 

 

 

 

 

 

 

Task
Integral ‘—
U, stimuli Context
3 a:
m Lu 0 I O
35. E.’ Giv1ng d1 rect1ons
> <2:
33 33 Activating recall X1 X2
5 o Activating
<2: 8 stimuli Activating
3 E performance set
0 Z
23' ‘1: Arousing
<2: 5 motivational set
> z
D
‘F Stimuli
from recall
Visual/ Verbal/
Iconic Digital

 

 

TYPE OF ADVANCE
ORGANIZER

Figure 1. Two Dimensional Stimulus Matrix.

In the matrix, the X's indicate the cells where advance
organizers are specified. X1 is a visual advance organizer and X2 is a
verbal advance organizer. Above or below these two cells, non-advance
organizers are specified. Each entry in the matrix could be associated
with a load factor and a speed factor

This matrix by no means specifies the stimulus stiuation
completely; but it does, in two dimensions and two factor specifica-
tions, provide a potential method for creating or sorting organizers.

An additional dimension will be added to the matrix below.

 

18

In addition to the stimulus characteristics just discussed, the

sensory modality of advance organizers is an important characteristic.

 

Sensory modality here refers to whether the stimulus is received by the
eye or the ear, i.e., visual or aural. In addition to the issue of
which, if either, mode is superior (and if so, under what conditions),
the issue of simultaneous stimulus reception must be considered.

There is no clear trend in either case. Lumsdaine (1953)
reviewed appropriate research and concluded that the situation is com-
plex and the results inconsistent. Levie and Dicke (1973) report a
study by Mowbray, who found that subjects could not pay attention to
two sense modalities at the same time. Travers (1967) developed a
model for information transmission by means of audiovisual materials,
and used it to imply that multiple sensory inputs would be of value only
when the rate of information input is very slow. This supports the
notion, for example, of alternating print and pictures in an instruc-
tional sequence. May and Lumsdaine (1958, p. 167) also commented that
combinations of modalities would be most effective for verbal learning.

The mixtures of sensory modalities best suited for groups,
individuals or Specific content sequences are not known, however. The
study of such mixtures would be quite complex. For that reason, the
advance organizers generated for the present study should each be
restricted to a single sense modality.

In the visual mode, the degree of realism versus abstraction

 

must be considered. Realism in pictorial media can be enhanced by the
use of color, greater detail, use of motion, and so on. These vari-

ables will be discussed below.

19

Any visual advance organizer that is designed would include
color or black-and-white stimuli, or a mixture of the two modes. Color
could be an important design factor for visual advance organizers. The
use of color in instructional media has been studied for decades. In a
review of the literature in this area, Lumsdaine (1963, p. 635) noted
that ". . . no really definitive studies have been made on specific
ways in which color may contribute to learning from instructional
media." Lumsdaine summarized the general status of the color versus
black-and-white issue as one where the role of color has been perhaps
overrated; the actual utility for color may be in cases where color
cues are essential for learning. Levie and Dicke's conclusions were
consistent; they reported "no significant difference" generally in
studies of color versus black-and-white.

Winn used a multivariate approach to examine the relationships
between the stimulus presentation mode (color or black-and-white) and
cognitive structure, in this case the structure of resulting multiple
free associations (Winn, 1976). Among the results, Winn found that
color emerged as a crucial factor only when central to the concept in
- question. This supports Lumsdaine's earlier conclusion.

Katzman and Nyenhuis (1972) also tried to study something other
than straightforward meaures of learning. After exposing subjects to
color and black-and-white slides of posters of various content, when
recall of materials in the posters was tested, the use of color was
associated with an increased recall of peripheral, as opposed to
central material. This increase in recall appeared to be limited to

pictorial, not written, material.

20

It appears that color would be used as part of a visual
advance organizer if (1) color were a key aspect of the conceptual
content, or (2) if it were desired to increase the likelihood of
responding to all parts of a visual organizer, instead of just the
"central" parts.

The amount of detail or "quality" of material in pictures can
easily be varied. In examining studies of pictorial quality, Lumsdaine
reported that in several instances noted, rough, crude sketches of
instructional materials were apparently as effective as "polished"
versions using motion-picture photography of "real" people and objects.
The studies examined included some methodological problems, but in any
event, no clear case is made that "high quality high detail" materials
are distinctly superior.

Levie and Dicke (1973), surveying studies in this area, found
essentially the same results. They did note a study by Dwyer who found
that low-detail materials were better for fixed-pace instruction but
high-detail materials were more effective when the pace was learner
controlled.

A final variable in this class of picture stimulus related
considerations is motion versus still. In this area, Lumsdaine (1963)
commented that the evidence does not demonstrate a clear general
superiority for one mode over the other, but does indicate that the
advantage of motion related to the nature of the task involved.

In summary, there appears to be no crucial set of types and
modes of stimulus combinations in pictures that is consistently

superior with learning tasks. This is consistent with Travers'

21

implication, with regard to his model for audiovisual information trans-
mission, that the quest for realism is not useful.

In addition to the stimulus aspects of pictures that have
just been discussed, the semantic content or meaning of advance

organizers must be Specified to the extent possible. In Imagery and

 

Verbal Processes, Paivio (1971) developed an approach to images and
verbal processes that conceptualizes alternative coding systems or
modes of symbolic representation; these modes are linked to experi-
ences with concrete objects and events, and language. This approach
will be treated more fully in the second major division of this litera-
ture review, but it is pertinent that Paivio does define stimulus
characteristics of the semantic content as major independent
variables.

A major dimension of the content of pictures is one of
abstractness versus concreteness. Paivio builds the relationships of
concrete semantic meaning to visual imagery, and abstract semantic
meaning to verbal imagery; "the more concrete or 'thinglike' the
stimulus or task situation, the more likely it is to evoke memory
images . . ." (1971, p. 9)-

This abstract-concrete dimension, as incomplete conceptually
as it is, may provide a third dimension for the matrix presented
earlier. The expanded matrix is given in Figure 2.

As in the first version of the matrix, X] and X2 represent the
layer of advance organizers; all other layers are non-advance
organizers. Each cell, then specifies a stimulus event or media

presentation along three major dimensions:

22

.chumz mapaswpm

Pocowmcmswo mmczh

.N mesmmd

 

 

 

 

 

 

 

 

 

 

 

 

mmNHz<wmo
uuz<>o< no mm>h
Pmuwmwo uwcoum
9% ya? 4/ >222, :23;
$9
firmwm wouoovv :38 :5ch
99%? 2:53
cquv «mm pmcowum>wpos
.///mvw mcwmsoc<
pmm mucmscomcma F
.353.
\\\\\\\\\\\\\\\\\\\\\\ mcwum>_uu< mcwpm>muu<
Nx _x F_momc mcwum>_pu<
mcowpumcwu mcw>wo
wpsewum
\\\\\\\\\\\ uxmucou _mcmmucH
\\\\\\\\\\\ xmmp

 

 

UHZINVDUO BDNVAOV'NON
SOSUEA UBZINVDUO BDNVAOV

23

1. advance organizer vs. non-advance organizer;
2. type of advance organizers; and
3. type of semantic content.
In addition each stimulus event or media presentation is specified

in such stimulus characteristics as:

1. speed;
2. load;
3. color;

4. quality/amount of detail; and

5. motion.

Summary of Stimulus Aspects of Advance
Organizers and Pictures

 

 

The first major division of the literature review has been
aimed at relating current research and thinking regarding stimulus
characteristics of media, especially pictures, to stimulus aspects of
advance organizers. The stimulus aspects of the physical media presen-
tations have been differentiated from the semantic content differences
of stimulus presentations. A matrix relating all these factors has
been created which it is felt can be used to define stimulus charac-

teristics of advance organizers.

Learner Characteristics

 

The second major division of this literature review focuses on
learner characteristics. Studies will be reviewed that deal with
learner characteristics and media and learner characteristics related

to the use of advance organizers.

24

One of the most basic learner characteristics is intelligence
or general ability. Hovland, Lumsdaine and Sheffield (1949), in a
study of the effects of films on servicemen of different intellectual
ability, found that, with regard to "factual knowledge,” higher-ability
men learned more from films, and that the degree of difficulty of the
material did not seem to be a significant factor.

In a study cited earlier, Northrup (1952) varied the prominence
of the organizational outline of three instructional films, and found
that making the outline more pronounced had the greatest facilitating
effect on groups of lower intelligence.

Countering these indications that general ability might be a
factor in preinstructional strategies, Stein (1952) found that the
improvement in scores on a post-film test due to the use of a pre-film
test was not a function of intelligence. And in the same vein, Barnes
and Clawson (1975), in a review of eighteen studies using advance
organizers spanning 1966 to 1973, found five significant with respect
to ability level, and thirteen nonsignificant. They conclude that no
trends were identifiable. 0n the other hand, Hartley and Davies (1976)
note that both university students in early studies and above-average
School children have benefited from advanced organizers, while educable,
mentally retarded adolescents and less able adolescent school children
have not. They conclude that it is ". . . probably best . . . [if the
learners involved are of] above-average ability, maturity and sophis-
tication" (p. 260).

The present author's conclusion with respect to learner ability
is that it would be best to involve persons with average or above-

average general abilities.

25

A second learner characteristic to examine is age. The
reviews cited earlier by Barnes and Clawson and Hartley and Davies also
examined advance organizers in terms of differences with respect to
age. In the Barnes and Clawson review, half (sixteen) of the studies
used subjects who were at the post secondary level. Of these, eight
reported significant effects in favor of advance organizers; this
pattern did not hold at the lower levels, so no conclusions were made.
In their review, Hartley and Davies note that the more recent studies
of advance organizers, where the age range was greater, tended to be
nonsignificant. They did not, however, factor out age per se as a
factor.

The extent of prior knowledge is a relevant learner character-
istic. May and Lumsdaine (1958) completed studies that indicate a
greater relative increase in learning for persons who had had instruc-
tion prior to instructional films. They comment, ". . . to him that
hath shall be given" (p. 119).

In the Weisberg study comparing visual and verbal advance
organizers, subjects with a middle level of prior knowledge made the
greatest gains in both treatments.

These two studies do not constitute trend. However, they do
indicate that prior knowledge, if not a major variable in a study,
should be accounted for.

The sex variable was included only in Weisberg's study. He
found that, for the eighth graders in the study, the sex variable was
nonsignificant. I

The final learner characteristic to be considered is cerebral

or lateral specialization. It has been hypothesized for some time

26

that the two halves of the brain are specialized to a considerable
extent. Ornstein (1972), in a chapter titled "Two Sides of the Brain,"
describes the left hemisphere (which controls the right side of

the body) as analytical, logical, verbal, linear and oriented to
language, mathematics, and 1nformation processing. He describes the
right hemisphere (which controls the left side of the body) as
holistic, artistic, relational, simultaneous and oriented to crafts,
body image, recognition of faces and integrating multiple inputs
simultaneously. These characteristics, Ornstein notes, are "purest"
for right-handed persons; left-handers are less consistent, reversed,
or mixed in these characteristics.

Regarding the relationship of handedness to cerebral func-
tioning, Harris (1975) describes the general population as being 90 to
98 percent right-handed. Of the right-handed people, about 99 percent
have left hemisphere laterality for language, i.e., their language
functions are primarily in the left hemisphere. In contrast he reports
that only 53 to 65 percent of left-handed people have the same left
hemisphere laterality, and concludes that left-handed persons are less
well-lateralized, i.e., their lateralization is weaker than right-
handed persons. Harris counsels knowing the handedness of subjects in
experiments.

White, in a review of laterality differences (1969), notes
that the criterion of handedness as a sole criterion of cerebral
dominance may be too simplistic, and concludes that ". . . there is no
clear-cut relationship between handedness . . . and cerebral dominance

(which. again, is not always a matter of singular lateralization)"

27

(p. 399). White does, however, accept the idea that handedness is in
some way related to lateral dominance.

Crinella et a1. (1971) lament the notion that handedness is an
indicant of some aspects of cerebral dominance, "Despite an overwhelming
amount of evidence indicating that there is no absolute correspondence
between these peripheral [such as handedness] measures and cerebral
dominance, weak or inconsistent lateral preference is often said to be
an indication of similarly inconsistent hemispheric organization"

(p. 2035). They cite evidence that lateral preference is not geneti-
cally transmitted, but is randomly assigned and culturally modified in
some left-handed individuals.

Calder notes that right handedness as the human norm ". . . is
probably connected with the occurrence of the speech area in the left
of the brain" (1970, p. 248). He describes the situation as complex
and little understood.

Kimura, in a series of experiments with "normal" (as distinct
from brain-damaged) persons (1973), investigated the extent that the
aural and visual senses "cross," i.e., the extent that the left hemi-
sphere responds to stimuli from the "opposite" side. Given that such
crossing existed, she was especially interested in its nature. For
the auditory system, she found evidence that the aural system is not
completely crossed. She developed a dichotic listening task in which
a headset played back two melodies Simultaneously, one to each ear.
When "normal" subjects were asked to pick out the two melodies from
four presented to both ears, they were able to pick out the melody
associated with the left ear better, indicating a dominant connection

with the right hemisphere.

28

The crossing of the visual system is more complex. For each
eye, each half of the visual field is connected to the opposite hemi-
sphere. In the left eye, for example, the right visual field is
connected to the right hemisphere. Thus, given visual activity and
eye movements, each hemisphere receives stimuli from the general visual
field of the person who is looking.

If a person fixates on a particular point, however, the visual
input to each hemisphere can be stabilized (see Figure 3).*

When this fixation procedure was set up, Kimura found that
words and letters (when presented tachistoscopicly) were reported more
accurately from the right visual (left hemisphere) than from the left
field (right hemisphere). She concluded that ". . . recognition of
visual verbal material is more accurate when such material initially
stimulates the left hemisphere" (p. 4). Using Levie and Dicke's terms,
the left hemisphere may be more reSponsive to digital stimuli.

Does then, the right hemisphere respond more effectively to
iconic stimuli? Kimura, working with Dunsford, created a device that:
(l) fixated a subject's gaze; (2) provided for stereo-optical images;
and (3) provided for tachistosc0pic exposure of the three-dimensional
material. They found that the right hemisphere was clearly superior
in analyzing information about where objects are located in space. In
related studies, Kimura found: (1) a small but consistent superiority
for the right hemisphere in slope identification; and (2) absence of
right hemisphere superiority in the perception of form. She concludes

that portions of the right hemisphere are involved with direct analysis

 

*Adapted from Kimura, p. 5.

29

__Left Field
ll
1\ l.
‘ \ I 1

Right Field

 

 

 

1
l
1
l
1
1
1 /\
1
I
l
1
l
1

 

s; I
‘\

\.
Left ' ‘

\ .
Hemisphere , Right

Hemisphere

Figure 3. Visual Fields Related to Hemispheres.

30

of information about the external environment, spatial relations, and
processing nonSpatial stimuli such as melodic patterns and nonsense
designs.

Kimura describes the interaction of the hemispheres, via the
corpus callosum, as complex, with much to be learned about how transfer
processing takes place. The effect of transfer can be explained; right
hemisphere superiority in spatial perception can be demonstrated even
though the subject gives a verbal response that is controlled by his
left hemisphere. It may be that primary analysis takes place! in the
right hemisphere and the verbal response is secondary.

This hemisphere specialization has been studied by Gazzaniga
(1967) using subjects who had "split brains," cerebral hemispheres in
which the corpus callosum has been separated. Using a fixed-point
tachistoscopic setup similar to Kimura's, Gazzaniga presented pictures
of objects, a spoon for example. When these pictures were presented
to the right visual field, the subjects, via their left hemisphere,
had no trouble identifying them. When presented to the left visual
field, however, the pictures failed to evoke appropriate responses;
subjects either guessed or said or wrote nothing. Yet when the pic-
tures were shown to the left visual field, a cigarette for example,
patients gpplg feel around with the left hand among objects screened
from sight and select an ashtray, on being asked to choose the article
most closely related to the picture.

Calder, cited above, also illustrated the functions of split-
brained persons. He described a situation where such a person was
presented a picture of a face: a woman's half—face was on the left

and a girl's half-face on the right. When the person was shown

31

several other faces (including the woman's and the girl's) and was
asked to §3y_what was seen, will say "girl." But if asked to pgipt_
(i.e., implying recognition), the person pointed to the woman's face.
The left hemisphere controls speech, and so the person "says" what is
seen on the right-~the girl's face. But since the right hemisphere
controls recognizing and guiding eyes and hands, the person points to
what is "seen," the woman's half-face. Normally the c00peration of the
hemispheres ma§k§_their separateness.

The emotional components of the individual hemispheres was
studied by Dimond et a1. (1976). They prepared two sets of special
contact lenses for normal subjects. One set limited vision to the
right hemifield, the other to the left; Thus, when subjects watched
three motion picture films (each about three minutes long), some per-
ceived images only in the right hemisphere (via the left hemifield) or
the left hemisphere (via the right hemifield). The three films were:
(1) a Tom and Jerry cartoon; (2) a medical operation; and (3) a
travelogue. After each film, the twenty-one subjects (18-24 years old)
rated it on four nine-point scales: (1) unknown, (2) unpleasant,

(3) horrific, (4) pleasant. The results indicated that (l) the left
hemisphere ratings were the same as a control group without contact
lenses, (2) the right hemisphere ratings were significantly more
unpleasant for all three films and more horrific for all films except
the travelogue. The authors suggest that ". . . the two hemispheres
of the brain of man possess an essentially different emotional vision
of the world" (Dimond et al., p. 691). The right hemisphere adds an

emotional dimension that perceives a thing as more unpleasant and

32

horrible. Each hemisphere makes a contribution to the whole, but the
authors assume the left hemisphere predominates.

Ornstein has generalized the specializations of the hemispheres
as two modes of consciousness. While avoiding the idea of binary
classifications, he gave a table of right and left hemispheric charac-

teristics (Table 1).

Table l.--The Two Modes of Consciousness: A Tentative Dichotomy.*

 

 

Left Hemisphere Right Hemisphere
Intellectual Sensuous
Active Receptive
Explicit Tacit
Analytic Gestalt
Lineal Nonlineal
Sequential Simultaneous
Focal 1 Diffuse

Time Space

Verbal Spatial
Intellectual Intuitive
Argument Experience

*Adapted from Ornstein, p. 67.

Ornstein also indicates that some persons habitually prefer, or
at least operate, in one mode or the other. This raises the issue of
concern of cerebral dominance, or preference and the general emphasis

on one hemisphere at the expense of the other.

33

Grady, in an article that describes the school system as giving
the right hemispheres of students minimal attention compared to the
left, discusses using instructional strategies to educate both hemi-
spheres (1976). He describes the first phase of instructional
technology as media ". . . done to students . . .“ (p. 48) and notes
the move to the current, second phase, as one where the student has a
more active, creative role, and so increases right-hemisphere use.

Samples, writing about science education, describes attempts
to create classroom environments ". . . in which the intuitive-
metaphoric functions of the right cerebral hemisphere were as legiti-
mate as the left hemispheric function" (Samples, 1975, p. 23).

Rennels, cited above, has done laboratory research using
electroencephalographic equipment and has concluded that differences
do exist in cerebral dominance among adults. He notes that graduate
students in art, for example, have lower GRE scores in the linear
areas of studies, and infers that they have maintained a highly
developed right-hemisphere dominance in spite of educational and socio-
logical pressures. He concludes that education should concentrate on
neurological symmetry.

Crinella, cited above, notes that ". . . there seems to be
little doubt that our educational system is characterized by a heavily
'left-brained' (or 'right-winged') philosophy, valuing the acquisition
and manipulation of language and symbols foremost, at the expense of
self-awareness, spatial schemata, music and the fine arts“ (p. 2050).
Current social problems are associated with the demise of the culture

that ". . . 'educated the hemisphere that wields the sword'" (p. 2050).

34

The right/left hemispheric specializations can be linked to
constructs of internal processing. Paivio, cited earlier, describes
visual imagery as parallel processing, storing and manipulating infor-
mation about spatially organized objects and events; it is dynamic and
can perform swift symbolic transformations. Verbal imagery is sequen-
tial, more static, and related to serial memory (p. 9). Both imaginal
abd verbal processes interact with stimulus conditions to determine
perceptual phenomena. Also, recognition is a function of percpetion
interacting with long-term memory; thus there is an information pro-
cessing component of recognition.

Although the visual and verbal processes Paivio describes are
essentially learner characteristics, he does Specify elements of
stimulus meaning. These are: (l) familiarity; (2) verbal codability;
and (3) word concreteness and meaningfulness. It appears that these
three elements can be related to the three-dimensional stimulus matrix
defined earlier.

Paivio also cites work by Brenda Milner and her associates at
the Montreal Neurological Institute as providing some “. . . of the
clearest evidence for the verbal-nonverbal functional distinction . . .“
(Paivio, p. 522). In studies of patients with lesions in either right
or left hemispheres, left hemisphere lesions were associated with
behavioral deficits that were conceptualized as a loss in verbal
symbolic abilities, and right hemisphere lesions produced selective
deficits in patients' abilities to process concrete (nonverbal) infor-
mation from memory. Paivio concludes that ". . . brain lesions

apparently can produce individuals whose abilities correspond roughly

35

to the verbal and nonverbal imagery types of the post-Galton era"
(p. 523).

Levie and Levie performed a series of experiments to obtain
evidence regarding the dual-coding hypothesis (such as Paivio's) or
the single-system hypothesis of human memory (1974). They conclude
that: (1) there is indeed a separate visual iconic memory system, but
the separateness may be preserved only when no verbal transactions
are necessary; (2) retention of pictures will be good if the measure is
recognition; and (3) loss of picture information will be high if
tested by verbal report.

Paivio notes than Sheehan (1966) found accuracy of visual
memory to be related to individual differences in imagery. Sheehan
described poor imagers as being more dependent on symbolic coding
(perhaps verbally) in order to remember them. Sheehan concludes that
vivid imagers perceive literally, while poor imagers use coding devices
(e.g., habit or ability) to organize their perceptions. Paivio also
speculated that high imagers would perceive and retain incidental
visual cues because there is no verbal coding, but with intentional
cues, there may be no difference because low imagers code and store the
cues verbally.

Other work links imagery and components of instructional proce-
dures. Di-Vesta (1971) conducted a series of seven studies investi-
gating interactions between imagery-ability and treatments related to
instructional procedures. He tried to relate imagery-ability as an
individual difference to stimulus aspects and to learner information
processing components. Additionally, Di-Vesta compared subjective and

objective reports of imagery ability. General findings included:

36

l. Imaginal and verbal processing are distinct. Imaginal pro-
cessing was found to be more effective for concrete words and verbal
processing was more effective for abstract words.

2. Imagery-ability was found to be a general indicant of
ability to process information by verbal or imaginal strategies.

3. The acquisition by all subjects was facilitated more by
pictorial than verbal presentations.

4. Picture-word order was more effective for learning,
transfer and recall than word-picture order.

5. Verbal ability and imaginal-ability factors are orthogonal
rather than bi-polar. Subjective reports of imagery contained a strong
element of social desirability.

6. Paivio's two-stage model of associative learning (imaginal/
verbal) was supported.

The implications of Di Vesta's studies for the present study
are first, that with regard to verbal and visual organizers, imagery
appears to be an adequate and relevant measure. Di Vesta's stimulus
materials were isolated units--words and pictures as individual
stimuli, in various combinations, sometimes as paired-associate
sequences. In the present study, the stimuli are more complex; para-
graphs and sequences of pictures would be used.

Second, Di Vesta's finding that the pictorial mode was
generally more effective implies the hypothesis for the present study
that the visual organizer should be significant, with the visual-
organizer means being higher.

Third, the verbal organizer treatment could represent a word-

picture treatment, which was one of Di Vesta's combinations, and so

37

should represent an ineffective combination. This should heighten the
main effect noted above. Thus the conclusion is made for the present
study that imagery is centrally related to the functioning of the right
hemisphere, although the nature of the relationship is not completely
understood.

Instruments to assess imagery for the present study include one
based on Paivio's work and a second developed by Sheehan. Additionally
an instrument designed by Sherman and Kulhavy was used.

Boutwell's test of visual imagery is based on one developed by
Paivio et a1. (1968). Paivio generated a list of 925 nouns that sub-
jects scaled on (1) abstractness-concreteness, (2) imagery, and
(3) meaningfulness, in terms of each word's capacity to arouse non-
verbal images. The assumption was made that because of their consis-
tent association with specific objects and events, concrete nouns were
particularly effective stimuli for the arousal of sensory images.
Paivio et a1. hypothesized that stimulus-evoked images ". . . could
function as mediators of response recall in paired-associate (PA)
learning situations, much as they have long been assumed to do in
mnemonic techniques that involve the use of imagery . . ." (Paivio
et al., p. 2).

Paivio et a1. further comment that "this heuristic theory has

been supported by the findings that concrete nouns exceed abstract

38

nouns in imagery according to rating scales . . ., reaction time,

. and physiological evidence . . ." (p. 2). The method of cate-
gorizing persons as high or low imagers in his work was the mean of
subjects' ratings, on a seven-point scale, of the ease with which a
word aroused sensory images.

There were thirty subjects in Paivio et al.'s studies all of
whome were university students ranging in age from 17 to 46 years
(median 19-20 yrs.). The resultant overall mean was 4.97 with a
standard deviation of 1.93. The frequency distribution of the scaled
attributes indicated that imagery was negatively skewed, with a high
proportion of nouns falling within the range 6-7 (high imagery). The
intergroup reliability was calculated to be 0.94. The imagery values
also correlated 0.87 with scale values from Paivio's previous research.
The Pearsonian correlation, indicating within-group stability, was
0.98.

Boutwell (1974) extracted fifty-four nouns from Paivio's
original set of 925 for use in a study in differentiated imagery.

Means of the 54 nouns correlated 0.98 to Paivio's list. The method
Boutwell used to categorize subjects as High or Low imagers is
described in a later section.

In a study that used both the noun test and Sheehan's QMI
(described next), Boutwell administered the instruments to ninety-eight
undergraduate students. The correlation of scores on the two tests led
Boutwell to conclude that construct validity existed.

The second mental imagery instrument used in the present study
was developed by Sheehan (1967a). Sheehan created a short form of

Betts' questionnaire on mental imagery. In 1909 Betts developed a

39

150-item questionnaire to measure mental imagery that remains the most
comprehensive test of imagery available. The questionnaire predicts

a subject's capacity to image in a variety of sensory modalities,
visual, auditory, cutaneous, etc. Sheehan used factor analysis and
determined that a Single factor (or component) was largely responsible
for the variance of scores within each modality in the Betts' instru-
ment. Seven modalities were included: visual, auditory, cutaneous,
kinaesthetic, gustatory, olfactory, and organic. Sheehan selected
five items from each modality, based on: (1) a high loading on the
main component (around 0.69 and 0.70); (2) similar means and standard
deviations, and (3) no appreciable differences in correlation for the
two sexes.

Sheehan then administered this 35-item instrument to a Sample
of sixty college student subjects (32 female). The resulting data
were factor analyzed and again, he found a Single component to be
largely responsible for the variance. He also obtained a correlation
of 0.92 between the total scores based on the complete Betts scale and
the shortened 35-item form. Sheehan concluded that the shortened
instrument predicted subjects‘ overall imagery score about as well as
Betts' complete questionnaire. He later replicated the study with a
sample of teacher's college students and obtained a correlation of
0.98.

Finally, since the Shortened QMI was developed in Australia,
Sheehan reported a study to establish the reliability of the test and
its suitability for testing with American college students (1967b).

In a test-retest procedure spaced over a seven—month period, results

were (1) individual differences were reliably shown, (2) the

4o

test-retest of both the total and visual scales as measured by the
Pearson correlation method, was 0.78, (3) there were no significant
differences in the vividness ratings over the two samples (except for
the organic modality: American reported higher imagery), and (4) there
was no significant difference between the American and Australian
samples in the average ratings.

The third and final mental imagery instrument was the Laterality
Assessment Inventory develOped by Sherman and Kulhavy (1976). This
instrument was designed to determine "cerebral laterality," i.e., to
discriminate between right- and left-dominant groups. The instrument
assesses both fine and gross motor activities of the hands, arms, legs,
and feet. Subjects respond on a five-point scale for each of the forty-
five items. For each subject, the score generated is a continuous
variable, extending from pure right- to pure left-handedness.

Sherman and Kulhavy began development of the instrument by
reviewing four general techniques for determining right- or left-
sidedness: (1) those focusing on morphological and morphofunctional ‘
characteristics, (2) batteries of manual tests of sidedness, (3) special
techniques, and (4) questionnaires about sidedness. They concluded
that the most practical and reliable methods involve questionnaires.

Then, currently-used questionnaires were criticized for:

(1) not getting at the underlying structure of the laterality dimen-
sion, (2) Splitting the laterality dimension into discrete left- and
right-sections. The present instrument viewed laterality as a con-

tinuous variable, extending from pure right- to pure left-handedness.

Sherman and Kulhavy generated forty-five action descriptions,

each requiring the use of hands. A five-point scale used varied from

41

"Right Always" to "Left Always." This instrument was given to one thou-
sand undergraduate students (367 males, 633 females). For each subject,
two scores were computed. First, a total "Right" score was obtained by
summing the "Right Always" and "Right Mostly" responses. The same was
done to obtain a "Left" score. The Igtip_of these two Scores provided
an estimate of the relative dominance of one side over the other.

For the second score, each response category was weighted
from "Right Always" = l, to "Left Always" = 5. The total number of
responses for each category was multiplied by that category's weight.
The five products were then summed, forming a composite laterality
coefficient that ranged from 45 (all "Right Always" responses to 225
(all I'Left Always" reSponses).

Analysis of the subject data yielded several findings. First,
using the laterality coefficients, Sherman and Kulhavy found that the
pr0portion of handedness across sex was not significant. The distri-
bution of laterality coefficients revealed that most of the group was
right-dominant. Only about 10 percent of the subjects were clearly
left-dominant.

Second, Sherman and Kulhavy computed Cronbach's coefficient
alpha for the composite instrument and obtained a value of 0.97. The
split-half reliability was 0.98.

Third, Sherman and Kulhavy performed a factor analysis by
intercorrelating the responses using Pearson product-moment correlation
coefficients. Kaiser's Varimax rotation procedure was also used with
normalization. Sherman and Kulhavy inferred from the structure a
general laterality factor that accounted for 81.8 percent of the total

variance.

42

Sherman and Kulhavy also obtained data from a separate group
of 200 undergraduates to assess the overall predictability of a dis-
criminant function, previously computed. Discriminant scores were
computed for these individuals, based on the weights derived from the
analysis of the original sample. Predictions of group membership
based on this discriminant function were found to be 99 percent
accurate, with only 2 cases classified incorrectly. Thus, this func-
tion appears to distinguish accurately between groups of right- and
left (mixed)-dominant persons.

Finally, in an effort to obtain additional validity data,
Sherman and Kulhavy, in a follow-up study, asked thirty subjects
(fifteen right-handed, fifteen left-handed) to return two months
after filling out the instrument. These subjects were then asked to
perform a series of eighteen activities. These performances related
to the three items loading heaviest on each of the six factors
originally identified. Performances were scored 1 for right, 3 for
both, and 5 for left responses. Pearson product-moment correlations
were computed between original verbal report and actual manual per-
formance. Except for one item, these correlations were all highly
significant. Also, the overall correlation coefficient between verbal
report and manual performance was .98, indicating that the instrument

has a high degree of predictive validity.

43
Summary_of Learner Characteristics Related
tp_Advance‘Organizers and'Pictures

The second major division of the literature review has been
devoted to examining relevant learner characteristics related to
advance organizers, especially in a visual mode. Learner character-
istics were general ability, age, prior knowledge, sex and cerebral
functioning. It appears that this set of variables can be controlled
by population selection, or measured and accounted for in an experi-
mental procedure. For cerebral functioning, a linkage has been made
between: (1) hemispheric specialization and dominance, and (2) visual
and verbal imagery. Finally, instruments to assess imagery have been

described.

Situation Characteristics

 

The third and final major division of this literature review
is related to situation characteristics of an instructional situation,
especially when advance organizers are used.

At the conclusion of the Hartley and Davies review of pre-
instructional strategies, cited earlier, they discuss when advance
organizers should be used, and say that ". . . situations requiring
some sort of conceptual framework that students can subsequently use
to help clarify the task ahead" (p. 259) may be best. They add that
the content should already possess a dominant structure that relates
well to students' existing knowledge and that the learning tasks
should be relatively short in length.

However, Barnes and Clawson, in their review, state that the

length of treatment is not a critical variable. As a result of

44

several studies, they further conclude that there is no clear relation-
ship between the effect of advance organizers and the content area of
the instruction.

Regarding conditions then, the Hartley and Davies and Barnes
and Clawson reviews indicate that (1) there should exist a "need" for
a clarifying structure or framework for the learning task, and (2) the
content of the learning task should have a strong, visible structure
that is not completely novel to students. Finally, the visual learning
task Should not be predominantly analytic, i.e., a consideration of

parts, but holistic, based on considerations of whole patterns.

Summary of Literature Review

 

The literature has been examined in terms of stimulus, learner
and situation characteristics of an instructional task where advance
organizers, especially visual advance organizers, might be used. The
primary conclusions of the literature review can be stated as tentative
assertions. First, the Opportunities to specify key components of
advance organizers may be enhanced by using a stimulus matrix such as
has been described. For the present study, advance organizers may be
generated that are characterized as activating recall, i.e., sensi-
tizing leaners to earlier, relevant cognitive strategies. Additionally,
advance organizers for the present study could be categorized as having
concrete semantic content, having shape, texture and form as dominant
stimulus aspects. Finally, such advance organizers could be either

visual or verbal in nature.

45

Second, left and right hemispheric functioning, especially as
expressed in individual differences, may be an important learner vari-
able with respect tO advance organizers.

Finally, the literature review indicates that both learner

characteristics and situation variables can be Specified.

CHAPTER III

DESIGN OF THE STUDY

Overview of the Chaptep_

 

This chapter contains description Of: the research questions
and hypotheses; definitions of important terms; the research design
including descriptions Of the instruments used; the learning stimulus
and experimental treatments; and subjects used. Internal and external
validity factors are discussed and the design over variables is given.
Finally the statistical hypotheses are listed and the procedures used

in the study are described.

Research Questions

 

The following research questions are drawn from the purpose
of the study and review Of the literature.

1. Given a visual discrimination learning task, is there a
difference in the effectiveness of visual and verbal advance organizers?

2. Given a visual discrimination learning task, is there a
difference in the performance of students with different cerebral
hemispheric functioning?

3. Is differential hemispheric functioning a factor in per-
formance differences associated with visual and verbal advance

organizers?

46

47

From these research questions, the following specific research

hypotheses are formulated.

Research Hypotheses for This Study

 

l. The mean score associated with the yiggal_advance organizer
will exceed those of the yg:pal_organizer and a control group.

2. The mean scores associated with high_imaging subjects will
exceed those of'lp! imaging subjects.

3. There will be a type of advance organizer by subject
imagery-level interaction.

Definitions of Important Terms

1. Verbal Advance 0rganizer--A written paragraph or two given
to a learner prior to an instructional task. ". . . organizers provide
relevant ideational scaffolding, enhancing the discriminability of the
new learning material from previously learned related ideas, and other-
wise effect integrative reconciliation at a level Of abstraction,
generality, and inclusiveness which is much higher than that of the
learning material itself" (Ausubel, 1963, p. 214).

2. Visual Advance 0rganizer--A series of photographs given to
a learner prior to an instructional task. The nature of a visual
advance organizer springs from Ausubel's writing, as does that of
verbal organizers. Ausubel's organizers are generated for "didactic
exposition," (Ausubel, 1963, p. 81), i.e., expository teaching.
Organizers also are based on the principle of integrative reconcili-
ation, i.e., by showing how previously learned, related ideas are
similar or different from the new ideas to be presented (Ausubel, 1963,

p. 82). The goal of an organizer is to Specify clearly, precisely

48

and explicitly the important Similarities and differences between the
new ideas and related extant knowledge (Ausubel, 1963, p. 83).

A visual advance organizer is not a visual argument, but is
expository, meaning ". . . to Open up, to place out, to lay open, to
unfold" (Friedman and McLaughlin, 1969, p. 54). Considering exposi-
tion, a visual advance organizer could include elements of illustration,
comparison and contrast, and classification. Visual classification
could include subsumption, i.e., the presentation of one class Of
objects as being part of a larger class. AS exposition, the visual
advance organizer would not include attempts of definition, analysis,
or presenting chronological order.

Further, the visual would not be objective in the sense that
it would analyze ". . . the Object into its significant parts and
discuss each in turn, following an order that reflects a principle of
organization inherent in the object itself" (Kane and Peters, 1969,
p. 155). Instead it would be subjective: the Observer would be a
visual participant.

The major goal of the visual advance organizer is relational.

 

Both stimulus differentiation and generalization Of the concepts would
be used. Thus the organizer would have both parallel and sequential
stimulus presentations. As part of the learning task, it would present
simple wholes, starting the presentation process toward complex wholes.
3. Cerebral hemispheric functioningr-The association of

different modes of information processing with each of the cerebral
hemispheres. The general association is one Of verbal/logical pro-
cesses related to the left hemisphere and pictorial/holistic processes

related to the right hemisphere.

49

4. Imagery-~The extent to which individuals have competence
in or preference for clear and strong images in the various sensory
modes: visual. aural, olfactory, etc. Measurement Of the extent of
imagery is used to infer the extent of right hemisphere specialization
competence or preference.

5. Learners--College students who are subjects in the experi-
ment. Learners are divided into two groups:

a. High Imagers--Those learners who have an extensive
competence in or preference for imagery as measured by imagery tests.
Right hemispheric competence or preference is associated with high
images.

b. Low Imagers-—Those learners who have a reduced compe-
tence in or preference for imagery as measured by imagery tests. A
reduced right hemisphere competence or preference is associated with
low imagers.

6. _jsual discrimination learning tagkr-An audiovisual
presentation, such as a slide/tape program, where the primary instruc-
tional goal is to teach the visual discrimination of things or events ‘

in a holistic or patterned way.

Design of the Study
The overall design had several main elements.
1. Observations consisted of three instruments to assign
subjects to a High or Low imagery group, and one instrument to assess

the hypothesized differences in learning--the dependent variable.

50

2. The learning task consisted of a series of projected
slides designed to teach the difference between art Objects from the
Art Nouveau period from objects of other periods.

3. Three treatments were administered, each to a subject
group. First, a verbal advance organizer, in written form was gen-
erated. Second, a visual advance organizer, in pictorial form was
generated. Finally, a control treatment was generated, in written
form, that occupied a time equal to each Of the other treatments.

4. Graduate and undergraduate students at MSU were recruited
as subjects. Data from seventy-two subjects were analyzed.

5. The primary statistic used in the study was an analysis
of variance. In addition, correlation and factor analyses were under-
taken.

The overall procedure had the following sequence:

1. Administration of two of the three instruments to assess
imagery.

2. Random assignment of treatments to subjects and adminis-
tration of the treatments.

3. Administration of the learning task.

4. Administration of the learning task posttest.

5. Administration of the third instrument to assess imagery.

6. Limited debriefing.

The study took the form of random assignment of subjects to
one of three treatments, and Observations to assess degree of imagery

and treatment effects. The design is illustrated below:

Treatment I 01A 01B RX1 01C 02

1A 018 sz 01c 02

1A 013 Rxs 01c 02

Treatment II 0

Treatment III 0

0 013, and 01C are instruments designed to assess vividness

1A’
of imagery and handedness.
02 is a learning performance instrument.
X1 used a visual (pictorial) advance organizer.
X2 used a verbal (written) advance organizer.

X3 (control group) used verbal (written) irrelevant material.

Instrumentation

 

Measures. Two measures were used in the study. First, each
subject was assessed in terms of his/her mental imagery. This measure-
ment was made in order to categorize each subject as either a High or
Low imager. The rationale for the measure was that the degree of mental
imagery could be used to infer the extent to which the subject's right
\ hemisphere was dominant. The measure as utilized in this study had
three individual-difference components: (1) the degree of imagery
associated with each word in a set Of nouns, (2) the degree Of imagery
vividness conjured up by short descriptions of various situations,
and (3) the extent of left- or right-handedness. The three instruments

that comprise this measure are described in the Instruments section

 

below.
The second measure was related to the learning task in that it
assessed the degree of learning that took place as a result of the

treatments. This measure was constructed by the researcher to provide

52

data on the dependent variable. The rationale for the measure was that
differences due to both the treatment and imagery category would be
Observed in the data associated with this measure. That is, effects
due to different advance organizers could be seen as higher or lower
scores, and effects due to imagery categories (high or low) could also

be seen as higher or lower scores.

 

Instruments. As noted above, 01 consisted of three instruments
while 02 consisted of a Single instrument.

The three instruments used to assess mental imagery were
(1) Boutwell's test of visual imagery, (2) Sheehan's short form of
Betts' questionnaire on mental imagery, and (3) the Sherman-Kulhavy
Laterality Assessment Inventory.

Boutwell's test consisted of fifty nouns, presented to subjects
in the format of ten words per page. A computer system was used to
randomly order and print the words, so that separate test orders were
available to subjects. Each subject rated each of the fifty words on
a seven-point scale from "High" to "Low" imagery, based on the extent
that each noun aroused mental images for a subject.

Sheehan's QMI test consisted of thirty-five items. Subjects
were asked to rate each descriptive situation on a seven-point scale,
from "perfectly clear" to "no image present at all."

The Sherman-Kulhavy Laterality Assessment instrument consisted
of fOrty-five items. For each item, subjects were asked to indicate
whether they used their right or left hands, and their consistency Of
using that hand. A five-point scale was used, ranging from "Right

always" to "Left always."

53

The final instrument was one designed by the researcher to
measure the learning that took place as a result of the treatments.
The instrument consisted of twelve items. Each item was made up of
three projected slides. The first two slides were photographic copies
of art objects. The third slide presented the scale to be used in

responding and is illustrated in Figure 4.

l 2 3 4
First Second Both Neither
slide slide slides slide

Figure 4. Scale Used in Learning Stimulus Posttest.

For each of the twelve pairs of art-style slides, subjects
were to decide which slide or slides were examples of Art Nouveau and
mark their responses on answer sheets. The slides showing Art Nouveau

objects are shown in Figure 5.

1 First 5. Both 9. Second
2. Second 6. Both 10. Second
3. First 7. First 11. Neither
4 Neither 8.

Second 12. Neither

Figure 5. Art Style Picture Combinations.

After the frequency of each response was determined, the
sequence of items was determined by use of a table of random numbers.

A c0py of the audio-taped instructions and a short description
of each of the test slides is included in Appendix A. i i

To test the instrument, it was given to four people who had not
seen the learning stimuli and had no particular specialization in art.

The mean score correct was 3.5; the maximum score correct was 4; and

54

the minimum score correct was 3. Additionally, the instrument was
administered to a faculty member in the MSU Art Department. This
faculty, who was familiar with the art-styles involved, scored all

twelve items correctly.

Learning_Stimulus

The basic task of the learning stimulus was tO teach the
differentiation of Art Nouveau from two other, somewhat temporally
related, art-styles (Baroque and Art Deco). Thus, the learning task
was basically visual discrimination. The stimulus materials were
photographic copies of various art objects in art picture books. Slide-
film was used and the resulting slides sequenced. The sequence had two
components. The first component included examples and non-examples of
Art Nouveau arranged sequentially; each example was keyed with a small
hole punched in the upper right corner of the slide (occupying 1.0
percent of the picture area). The second component differed in that
each example was first presented unkeyed (without the punched hole) and
then followed by a second, keyed repetition. Thus, subjects had the
Opportunity to judge whether the slide was an example of Art Nouveau
or not, and then to receive confirmation. Nonexample slides were
followed by a black slide (blank screen). 1

Timing of the slide presentations for each of the two com-
ponents is shown in Figure 6.

It can be seen that during the second sequence, where unkeyed
examples were followed by keyed examples, the total exposure of the
examples was twice that of the nonexamples. Each slide however,

whether example (keyed or not), nonexample, or black slide, was

55

.mau__m nauuanoca co memsac .m acsm_a

.mmuwpm xumpn

an uszFPoa mmFQmemco: use mmPamem nmamx an nozopFom umxmxcz mo mucmacmm go; mavewh

 

 

 

we?»
vmmampm
mm P am mm mp m— m N a
_ _ _ m r _ _ p
. . 4. _ u. z 41. _ u A _
umm F mumm N umm _ mumm 5 6mm _ mOOm n umm _ mumm m
wmcmzu muwpm mmcmgu mFQmem mmcmcu mpqemxm mmcmgu mpasmxm
gumpm -coz umxmx vmxuxcs
r111 L
pcmucoo mamm
.mmPQmemcoc use mmngmxm umxmx we mucozcmm cot nave?»
wave
vmmampm
am mm mF mF m N o
r_ F _ Pl P p
. . 41. A, _ a _ _ A
.mumm n 6mm P momm 5 6mm F momm N

m_a5mxm mucosa OPQmemcoz «mango mFamem

56

presented for seven seconds. Slide-change pulses were recorded on
magnetic tape to effect these changes, and to increase the timing
precision for each group of subjects.

The transition from the first sequence of slides to the
second sequence was made by use of audio-taped instructions and an
explanation; the screen was black during this explanation. Each slide
in the instructional sequence was presented for seven seconds. The
text of the audio comments is included in Appendix B, along with a

short description of each slide and an indication of the sequence.

Experimental Treatments

 

Three experimental treatments were used, each followed by the
learning stimulus, described above. The treatments were distributed
randomly among subjects.

1. Visual Advance Organizers. The first treatment consisted
of Six photographs presented in sequence to each subject. An attempt
was made to illustrate the relationships between Art Nouveau and
other art styles, and between art styles and their general visual
antecedents. The art objects pictured were photographically copied
from art books. Descriptions of the six photographs making up the
visual advance organizer are included in Appendix C.

2. Verbal Advance Organizer. The second treatment consisted
of a verbal advance organizer, in written form. An attempt was made
to describe the learning task at a high level of generality, and to
link it with other, general knowledge. This treatment is included in

Appendix D.

57

3. Control group. The control group treatment consisted of
a paragraph of material about art masterpieces not related to the
learning task or the posttest. This paragraph is included in

Appendix E.

Subjects
Subjects were 115 graduate and undergraduate students at

MSU. Two processes for selecting students were used. First, a set
of instructors was solicited to permit the experimental procedure
to occur during classtime, with all class members participating.'
Second, a procedure was used whereby individual student volunteers
were solicited from various classes. These volunteers then met at
a common time for the experimental procedure. Students who partici-

pated in this second procedure received extra course credit.

Subject Imagery Assignment Procedure

 

Considering imagery level, a procedure suggested by Boutwell*
using scores Obtained from Boutwell's Noun Test and Sheehan's QMI was
modified and used. First, the Noun Test data were analyzed in such a
way that the mean and standard deviation of responses for each word
were computed. The words were then rank-ordered, from low mean to
high mean. The researcher then examined the middle set of words, where
the mean was near the middle of the seven-point scale and the standard
deviations were greatest. Nine words were discarded leaving Sixteen
"Low" words and twenty-five "High" words. The words, with means and
standard deviations, are presented in Appendix F. The mean of the

"High" words and the mean of the "Low" words were then computed.

 

*Telephone conversation including Boutwell and Phelps.

58

The data were then analyzed subject by subject. For each
subject, two means were computed, one for the "High" words and one
for the "Low" words. Each subject's score was then computed using
responses from Sheehan's QMI test. The overall QMI mean was also
computed.

Categorization of each subject as a high- or low-imager was
determined using three means: (1) the Noun Test AHigh" words, (2) the
Noun Test "Low" words, and (3) the Sheehan QMI mean. A subject was
categorized as a HIGH imager if: (1) his/her mean for the "High“ words
was higher than the overall "High" words mean; agg_(2) his/her mean for
the "Low" words was higher than the overall "Low" words mean, agg.

(3) his/her QMI mean was higher than the overall QMI mean. A subject
was categorized as a LOW imager if: (1) his/her mean for the "High"
words was lower than the overall "High" words mean, agg_(2) his/her
mean for the "Low" words was lower than the overall "Low" words mean,
gpg_(3) his/her QMI mean was lower than the overall QMI mean. Subjects
who did not fall into these categories were eliminated from further

analysis.

Validity700ncerns

As in all research, the internal and external validity Of the
study was affected by the design and events occurring during the course
of the study. Validity will be discussed in terms of Campbell and
Stanley's (1963) discussion.

Internal Validity. 1. Of 115 subjects initially tested,

 

thirty-one completed the entire experimental process in one setting.

The remaining eighty-four completed the process in two sessions,

59

Spaced over a period of two to four days. For these subjects,
Boutwell's Noun Test and Sheehan's QMI were administered during the
first session, but the learning stimulus, posttest, and the Sherman-
Kulhavy Test were given during the second session. Thus, the threat
of effects due to history were minimized due to having a single setting
for over one-fourth of the subjects, and keeping the delay for the
remaining subjects to four days or less. Since the two instruments
administered prior to the treatments and learning stimuli (Boutwell's
Noun Test and Sheehan's QMI) were not related to the content of the
learning stimuli (Art Nouveau and other art-styles), it is unlikely
that any relevant learning took place due to these two instruments.

2. Statistical regression may have been a threat to internal
validity because subjects were assigned to high- and low-imager groups
on the basis of extreme scores (high and low) on the Noun Test and QMI
instruments.

3. Internal validity was generally strengthened because Of
random assignment to treatments.

4. Experimental mortality did occur in the present study.
Data from forty-three subjects had to be discarded because it was
incomplete. All but three subjects in this category were ones in
classes where the experimental procedure was in two sessions, as
described earlier. Approximately half of these subjects did not appear
for the second experimental session; in the remaining cases, control
numbers on some of the response sheets were missing or incorrect.
Although these subjects were evenly distributed among the classes that
had two sessions, the experimental mortality does represent a threat

to the internal validity Of the study.

60

External Validity. Interaction effects of selection biases

 

and the treatments represented a threat. Sampling from the population
was not random, but consisted of blocks of volunteers (individual
classes) and individual subject volunteers. This limits the generali-
zations that can be made from the study.

Further, the administration of the Noun Test and Sheehan's QMI
may have increased the subjects' sensitivity to the treatments in some
undetermined way other than "learning" as measured by the posttest
instrument.

Finally, the subjects were aware they were part of an experi-

ment.

Statistical Model of Analysi§_

 

The independent variables were (1) imagery level (High or Low)
and (2) treatment group (visual advance organizer, verbal advance
organizer, and control).

Since subjects were randomly assigned to one of three treat-
ments (visual advance organizer, verbal advance organizer, or control
group), there were three levels of the independent variable.

The dependent variable was the learning-stimulus posttest score
for each of the subjects. The data were analyzed in terms of the
number correct out of the total, twelve.

The variable matrix, showing the relationship of independent
variables, is given below in Figure 7.

The hypotheses were tested via a two-way factorial analysis of
variance at the a = 0.05 level. Considering the variables defined

above, it was assumed that the population sampled was normally

61

 

Treatment Group

 

l 2 3
Imagery Visual Advance Verbal Advance
Level Organizer Organizer Control

 

 

High

 

Low

 

 

 

 

 

Figure 7. Variable Matrix

distributed. No occurrences or events were observed by the researcher
that tended to counter this assumption. Further, it was believed that
the homogeneity of variances in the groups was the same due to random
assignment. Finally, the dependent measures, learning stimulus post-
test scores, were continuous data, and had equal intervals on the
posttest scale. It was concluded by the researcher that the assump-

tions for the use of analysis of variance were met.

Statistical Hypotheses. Using the learning stimulus posttest

 

scores as the dependent variable the following hypotheses were tested:
1. There will be no difference in the means of the three
treatment groups.
2. There will be no difference in the means of the two
imagery level groups.

3. There will be no treatment by imagery level interaction.

62

Procedures for the Study

 

Pilot Test

The experimental materials were pilot tested on a group of
thirteen students. After the learning stimulus posttest, subjects
were interviewed. This data and the posttest data revealed that the
learning stimulus was inadequate. Subjects had expressed confusion
regarding the task, and the data indicated that overall mean of correct
responses was about 25 percent. Changes made based on these data
included: (1) changing the pattern of the slides; (2) introducing the
technique of delayed visual feedback, wherein subjects were asked to
decided whether or not a slide was an example before a confirming slide
was projected, and (3) revising the audio taped instructions. These
changes resulted in substantially higher scores from subjects in a
second pilot test. In this test, the mean was 6.46 with a standard

deviation of 1.34 and an N of thirteen.

Experimental Procedures

Two procedures were followed; they differed in only one aspect.
As noted earlier, 25 percent of the subjects received the Noun Test
and Sheehan's QMI four days prior to the rest of the procedure. The
description that follows will detail the procedures used for the other
75 percent.

The procedure was as follows:

1. The Noun Test and Sheehan's QMI were given to each subject,
along with a treatment package (including the posttest), and the

instrument on handedness. Subjects were told to read the instructions

63

for the Noun Test, complete it, and then read the instructions for
Sheehan's QMI, and complete it.

i 2. After the subjects completed the two instruments, these
data were collected. Subjects were then asked to listen to audio-taped
instructions; these instructions were played. Subjects were asked to
examine the treatment materials for a minute and a half.

3. After the ninety-second pause (the audio tape provided
timing), taped instructions asked that the subjects stop their activity
and direct their attention to a screen. The method of keying the con-
cept to be learned was explained, and the learning stimulus slides
sequence was shown, synchronized by the audio tape.

4. After the slide sequence, the taped instructions introduced
the posttest and the response scale to be used. The tape was then
stopped and the subjects asked if there were any questions (there were
only two or three, with no pattern apparent). The tape was then
restarted and the test slide sequence was shown, again synchronized
by the tape.

5. At the end of the test sequence, the instrument on handed-
ness was distributed. Subjects were asked to read the instructions
and complete it.

6. When the handedness instrument was completed, the nature
of the procedure was explained and subjects' questions were answered.

The procedure was repeated with subject groups ranging in size
from five to forty. In two instances, the slide synchronization system '
partially failed and some of the slides during the learning stimulus
slides were advanced by hand, using the sweep-second hand of a wrist

watch for timing. Timing errors appeared not to exceed one second.

64

Chapter Summary
After presenting the research questions of the present study
and listing definitions of terms, this chapter presented the design of
the study. Finally, procedures of the study were described. Chapter IV
will present the results of these procedures and will report the out-

comes of hypotheses tests.

CHAPTER IV

ANALYSIS OF THE DATA

The data and their analyses are presented in this chapter.
These results are then used to test the hypotheses of the study.
Interpretation of these findings and conclusions will be presented in
Chapter V.

The present study had two purposes. First, to examine the
effects of verbal and visual advance organizers as components of a
learning task. Second, to examine the effect of subjects' degree of
imagery capability on differences in learning associated with verbal
and visual advance organizers.

The hypotheses drawn from these purposes for testing after a
review of relevant literature were:

1. Test scores associated with a visual advance organizer
will exceed those associated with a verbal advance organizer and a
control group.

2. Test scores of subjects categorized as high imagers will
exceed those of subjects categorized as low imagers.

3. A type of organizer by learner degree of imagery inter-
action will be Significant.
The primary statistic to test the hypotheses was a two-way factorial

analysis of variance. Additional analyses included a correlation

65

66

procedure and factor analysis. For tests of significance of
the findings, the alpha level was set at the .05 level on a two-tailed

distribution.

Findings of the Study

One hundred fifteen subjects participated in the experimental
procedure, in seven groups, ranging in size from six to forty. Data
from forty-three of these subjects was discarded due to incompleteness.
Thus, data from seventy-two subjects were entered into the high- and
low-imagery categorization procedure.

Categorization of subjects as high- or low-imagers followed a
specific sequence:

1. A mean scale rating for each of the fifty words in the
Boutwell Noun Test was obtained. The fifty words were first placed in
alphabetical order. Then the subjects' (N = 72) ratings of the first
word were summed and the mean calculated. The same was done for the
second word, and so on. The fifty means ranged from a low of 1.69 to
a high of 6.61. Thus the means of the fifty words covered a range of
4.92 on the seven-point scale.

2. The means of the words were then used to rank the words
themselves, from low-imagery words to high-imagery words. The words
were thus distributed on an imagery continuum. "City” had the
highest imagery rating; "impropriety" had the lowest.

3. The researcher then visually scanned the means of the
words and eliminated a group of words in the center, thus creating a
group of high-imagery words and a group of low imagery words. Nine

words in the center were eliminated. The means of the high-imagery

67

words ranged from 6.61 to 4.60; the means of the low-imagery words
ranged from 3.01 to 1.69. Twenty-five words were thus determined to
be high-imagery words and sixteen determined to be low-imagery words.

4. The mean of the twenty-five high-imagery words was then
calculated (5.602). In the same manner the mean of the sixteen low-
imagery words was calculated (2.383).

5. These two means were then used as the first part of the
high-imagery, low-imagery categorization process described below.

6. The mean score for all the subjects' ratings of Sheehan's
QMI Test was then calculated. This mean was found to be 5.732.* This
mean was used as the second part of the imagery categorization process.

7. The imagery categorization procedure itself had two steps
and each subject was categorized individually. The complete categori-
zation procedure is presented as a flowchart in Figure 8.

The first step was to compare each subject's high words mean
and low words mean from the Boutwell Noun Test to the overall high
words and low words means, previously calculated. Three outcomes were
possible: (1) categorization of the subject (tentatively) as a high
imager if both subject scores were higher than the overall means,

(2) categorization as a low imager if both subject scores were lower
than the overall means, and (3) discarding the data if the subject's
high score was higgg§_than the overall high mean but his/her low score

was lower than the overall low mean, or if the subject's high score

 

*In Boutwell's Noun Test, a scale-value of 7 indicated high-
imagery, while the same scale value in Sheehan's QMI Test indicated
low-imagery. For consistency in reporting results, the QMI scores
are transformed by the equation 8-R = S, where R = subject rating and
S = scale value used in reporting.

68

.mcaumuocm.comumNPLommumu mo penguzopm .m mesmwm

 

 

 

 

 

 

 

 

 

 

 

 

~_uz - __uz
cmmmEH vauz cmmmEH
Imam a me name Homfiaam god a ma
:OWHMchommpmu yo ccmOmwo cowuwchommumg
maoouzo
:8 :5 1 M M.
cmzmmcm _
”N amum zome m>on< Zofimm

N N

: :

I. 9
........................................................ ,m-.--------.111----------------------A1----------------------1-
“map caoz

_Pm3pzom

c_ amam

 

 

69

was lgggr.than the overall high mean but his/her low score was higggr
than the overall low mean. Second, if the subject was tentatively
categorized as a high- or low-imager, his/her score on Sheehan's QMI
Test was compared with the overall QMI Test mean. Once again, three
outcomes were possible: (1) final categorization as a high-imager if
(s)he was "high" on the Noun Test and his/her QMI Test score was above
the overall mean, (2) final categorization as a low-imager if (s)he
was "low" on the Noun Test and his/her QMI Test score was below the
overall mean, and (3) discarding the data for that subject if his/her
categorization from the QMI Test score was inconsistent with that of
the Noun Test. If the Noun Test category was high and the QMI Test
category was low, or if the Noun Test category was low and the QMI
Test was high, the subject's data were discarded.

As can be seen by referring to Figure 8, a subject was cate-
gorized as a high-imager if his high words mean exceeded 5.602, his/her
low words mean exceeded 2.383, and his QMI Test score exceeded 5.732.
In a similar manner, a subject was categorized as a low-imager if
his/her three scores fell below the three means listed. Any other
combination of scores caused the subject data to be discarded.

Figure 8 further reveals that seventeen subjects were cate-
gorized as high-imagers and eleven subjects as low-imagers. These
twentyheight subjects comprise only about 39 percent of the original
group of seventy-two subjects. This outcome resulted in reduced cell

sizes for the analyses of variance, to be described below.

70

Test Hypotheses

 

For purposes of statistical testing, the hypotheses of the

present study will be expressed in null and alternate form.

Hypothesis 1.

 

Null: There will be no difference in the means of the visual
advance organizer (Tl), verbal advance organizer (T2), and control
group (T3). H01 = pT] = “T2 = “T3

Alternate: The means of the visual advance organizer (T1) will
exceed those of the verbal advance organizer (T2), and the means of

the verbal advance organizer will exceed those of the control group

(T3)° H11 = 1'11 > “12 > “13

Hypotheses 2.

 

Null: There will be no difference in the means of the high-
imagery subject group (11) and the low-imagery subject group (12).
Ho2 = “11 = 1'12
Alternate: The means of the high-imagery subject group (11)

will exceed those of the low-imagery subject group (12). H12 =
“11 > “12

Hypothesis 3.

 

Null: There will be no type of advance organizer (T1, T2, T3)
by subject imagery-level (11, I2) interaction.
Ho3 = “11 = “12 = “1311‘

“T1 = “12 = “13'12

71

Alternate: There will be a type of advance organizer (T1, T2,
T3) by subject imagery-level (11, 12) interaction.

H13 ‘ “11 i “12 I “1311‘
“11 i “12 i “13'12

Two-Way_Ana1ysis of Variance

 

A two-way analysis of variance was performed on data from
twenty-eight subjects. The dependent variable was the mean score on
the learning stimulus posttest. The independent variables, cell means

and sizes and sex distribution are presented in Table 2.

Table 2.--Cell Sizes, Means and Sex Distribution-~TwO-Way Analysis of

 

 

 

 

 

Variance.
Treatment Group
Imagery
Level Visual Advance Verbal Advance Control Totals
Organizer Organizer
Size N=7 N=3 N=7 N=l7
Hi h Sex M=2 F=5 M=O F=3 M=2 F=5 M=4 F=l3
9 Mean 4.57 4.00 5.00
8.0. 2.15 1.00 1.83
Size N=3 N=4 N=4 N=ll
Low Sex M=1 F=2 M=2 F=2 M=1 F=3 M=4 F=7
Mean 5.33 5.75 4.50
5.0. 0.58 1.26 1.73
N=10 N=7 N=ll N=28
T°ta‘S M=3 F=7 M=2 F=5 M=3 F=8 M=8 F=20

 

 

 

 

 

 

 

The two-way analysis of variance results are presented in

Table 3.

72

Table 3.--Two-Way Analysis of Variance Table for Learning Stimulus
Posttest Scores.

 

 

Source of Variation df Mean Square F p
Treatment 2 0.09610 .03297 .96762
Imagery Level 1 1.74075 .59716 .44789
Interaction 2 2.68233 .92017 .41325
Error 22 2.91504

 

Table 3 reveals that neither the treatment nor the imagery
level main effects were significant at the a = .05 level. In addi-
tion, the interaction term was not Significant at the a = .05 level.

Thus, null hypotheses H01, H02 and H03 were not rejected.

Correlation Procedure

 

In order to investigate the relationship of variables in the
present study, a correlation procedure was used. For this procedure,
data from seventy-one subjects were included (one subject was deleted
because of a lack of data on the sex variable). Again, the larger
group of subjects was used in the analysis in order to increase the
size of the sample. Further, the procedure and results of the cate—
gorization procedure could be deleted because it was desired to
examine the data as continuous scores rather than as discrete factors.

The eleven variables used for each subject in the correlation

procedure are presented in Table 4 .

73

Table 4.--Variables Used for Each Subject in Correlation Procedure.

 

 

Variable Variable Name
Number
1 Sex
2 Treatment Group--Visual Advance Organizer, Verbal Advance
Organizer, Control
3 Boutwell Noun Test--Mean of all fifty words
4 Boutwell Noun Test--High words mean
Boutwell Noun Test--Low words mean
6 Sheehan's QMI Test--Mean of all items
7 Sheehan's QMI Test--Mean of Visual items (items 1-5)
8 Sherman-Kulhavy Test--Handedness Ratio
9 Sherman-Kulhavy Test--Latera1ity Coefficient
10 Learning stimulus posttest mean correct responses
11 The experimental group (classroom) each subject occupied.

The resultant correlation matrix is presented in Table 5.

The interpretation of these correlation coefficients is based on the
heuristic that any important relationship will be indicated by a
coefficient Of .35 or greater, indicating that at least 12 percent
of the variance in the two variables being correlated is shared.

The correlation data reveal several patterns. First, when the
treatments variable (variable 2) is compared with the dependent vari-
able, the posttest mean correct responses (variable 10), the resulting
correlation coefficient of -.04 indicates that there was essentially
no relationship between the experimental treatments and the posttest

SCOY‘ES .

74

Table 5.--Correlation of Variables.

 

 

1 2 3 4 5 6 7 8 9 10 11

1 100 0 17 26 -4 14 30 -17 -14 10 5

2 0 100 O -1 1 13 15 7 7 -4 -7

3 17 0 100 85 76 14 11 2 -1 2 -11

4 26 -1 85 100 35 5 2 7 3 5 3

5 -4 1 76 35 100 12 11 -7 -5 -1 -19

6 14 13 14 5 12 100 61 -8 -6 5 O

7 30 15 11 2 11 61 100 13 15 -4 -11

8 -17 7 2 7 -7 -8 13 100 88 -5 3

9 -14 7 -1 3 . -5 -6 15 88 100 6 O

10 10 -4 2 5 -1 5 -4 -5 6 100 -14
11 5 -7 -11 3 -19 0 11 3 O -14 100

 

Second, when the sex variable (variable 1) is compared with
the other ten variables, it appears that sex did not relate importantly
to any of these.

Third, when the experimental group (classroom, variable 11)
the subject occupied for the procedures is compared with the other ten
variables, it appears that it, too, was not an important factor.

Fourth, none of the variables associated with (1) Boutwell's
Noun Test (variables 3, 4, 5) or (2) Sheehan's QMI Test (variables 6, 7)
correlated at important levels. The coefficients were all positive,
but at low levels.

Finally, none of the variables associated with (1) Boutwell's
Noun Test and Sheehan's QMI Test (variables 3, 4, 5, 6, 7) or (2) the
Sherman-Kulhavy Test (variables 8, 9) correlated above the pre-

established important level of .35.

75

Factor Analysis

 

The same set of eleven variables used in the correlation pro-
cedure was entered into a factor analysis procedure. Three principal
components resulted from the analysis. These components are listed in
Table 6. The factors on which each variable had its highest loading

are marked by an asterisk.

Table 6.--Principa1 Factors and Highest Factor LOadings Related to

 

 

 

Variables.
Factors
Variables

l 2 3
1 24 -14 -25*
2 4 9 -18*
3 94* 2 20
4 79* 6 26
5 72* -5 13
6 -29* 3 70*
7 -28 -17 74*
8 -2 94* 6
9 -3 93* 2
10 4* -l -1
ll -12* 0 5

‘_.

*Note--Entries marked with an asterisk (*) were highest in
loadings;

The first factor consisted of variables three, four, and five,
the scores associated with Boutwell's Noun Test.

The second factor conSisted of the scores associated with the
Sherman-Kulhavy Test, variables eight and nine.

The final factor consisted of the scores associated with

Sheehan's QMI, variables six and seven.

76

The results of the correlation and factor analyses are consis-
tent with the results of the two-way analyses of variance. No
relationships or factors were found that indicated an effect due to the

three treatments or to imagery.

Chapter Summary

 

In this chapter data were analyzed, the results were presented
and hypotheses tested. Additional analysis of the data was also pre-
sented. The purpose of the study was to investigate (l) the effects
of a verbal and visual advance organizer on performance associated
with a learning task, and (2) the effect of subjects' degree of imagery
capability on learning differences associated with the two advance
organizers.

One hundred fifteen subjects participated in the experimental
procedure. Of these, data from forty-three was discarded due to
incompleteness; data from seventy-two subjects were entered into the

imagery categorization procedure.

Each subject was categorized as a high- or low-imager by
comparing hl'S/her scores on (1) a list of high-imagery words on Bout“
well's Noun Test, (2) a list of low-imagery words on Boutwell‘s “0““
Test, and (3) Sheehan's QMI Test with overall group means in the$3
three categories. Subjects who scored above the mean of all three
groups were categorized as high-imagers; those who scored below the
means of all three groups were categorized as low-imagers. Of the

seventy-two subjects, seventeen were categorized as high-imagem and

eleven as low-imagers.

77

Null forms of the hypotheses were tested by use of a two-way
factorial analysis of variance. The independent variables were
(1) level of imagery (high or low), and (2) treatment (visual advance
organizer, verbal advance organizer, control group). The dependent
variable was the score on the learning stimulus posttest (mean correct).

The results of the two-way analysis of variance indicated that
none of the null hypotheses could be rejected. Specifically, (1) no
difference due to the three treatment groups (visual advance organizer,
verbal advance organizer, control group) were found; (2) no differences
due to imagery level (high, low) were found; and (3) no significant
interaction of treatments and imagery levels were found.

A correlation procedure was used to investigate the relation-
ship of variables in the study. The treatments variable was not found
to relate with the dependent variable. The sex and experimental group
variables were found not to relate with other variables. Finally, the
variables associated with Boutwell's Noun Test and Sheehan's QMI Test
were found not to relate with each other or to the Sherman-Kulhavy
Test.

Finally, a factor analysis procedure, using the same variables
as used in the correlation procedure, revealed a three-factor structure.
The first factor consisted of the Boutwell Noun Test variables. The
second factor consisted of the Sherman-Kulhavy Test variables, and
the third factor consisted of the Sheehan QMI Test variables. No
other factors were found.

These findings along with conclusions are discussed in

Chapter V.

CHAPTER v
CONCLUSIONS AND IMPLICATIONS FOR FURTHER RESEARCH

Summar

The purpose of this study was to investigate the effectiveness
of a visual advance organizer compared to a verbal advance organizer,
and to examine differences in learning resulting from each type of
organizer that could be attributed to imagery level as an individual
difference.

The importance of the study is related to extending knowledge
about the functional properties of visuals, especially when used as
advance organizers. Further, it is important to extend the concept of
advance organizers to non-verbal modes. Finally, linking student
differences in cerebral hemisphere functioning to classroom tasks is
important, as educators attempt to improve the effectiveness of
instruction.

A review of the literature indicated that the verbal advance
organizer concept might be extended to the visual mode. The review
also revealed that individual differences in right and left cerebral
hemispheric Specialization, as measured by degree of visual imagery,
might be an important variable with respect to advance Organizers.

The basic hypothesis of the present study was that the type of
advance organizer and category of imagery would result in differential

learning performance, and that these two factors would interact.

78

79

Subjects in this study were one hundred fifteen graduate and
undergraduate students in various disciplines at Michigan State Univer-
sity. Of these, seventy-two subjects with complete data were cate-
gorized as high- or low-imagers, or discarded if data were inconsistent.
Eleven low-imagery and seventeen high-imagery subjects were identified,
using Boutwell's Noun Test and Sheehan's QMI Test.

Three experimental treatments were generated, a visual advance
organizer, a verbal advance organizer and a control treatment. In
addition, a learning stimulus and posttest were generated by the
researcher. .

The experimental procedure consisted of administering Boutwell's
Noun Test and Sheehan's QMI Test. These two tests were designed to
measure relative mental imagery. Second, subjects were exposed to the
learning stimulus, a series of projected slides containing material
designed to teach discrimination of Art Nouveau from other art styles.
Third, the learning stimulus posttest was administered that recorded
subject responses to visual discrimination items.

Following the posttest, the Sherman-Kulhavy Test, purporting to
assess cerebral laterality, was administered. The experimental proce-
dure was administered to seven groups of six to forty subjects.

Three hypotheses were tested to determine whether type of
advance organizer and degree of imagery would affect learning perfor-
mance, and whether these factors would interact. The data were tested
with a two-way factorial analysis of variance with the alpha level set
at .05. Other statistical procedures included correlation and factor

analyses.

80

The findings of the study did not support the three hypotheses.
No significant differences were found for factors in any of the analyses
of variance. Further, correlation and factor analyses revealed that
the two instruments used to measure imagery had a low correlation with
each other and that each was a strong, distinct and separate factor

in the factor analysis.

Cpnclusions

 

The alternate hypotheses will be considered in turn.

H11: The means of the visual advance organizer (Tl) will
exceed those of the verbal advance organizer (T2), and the means of
the verbal advance organizer will exceed those of the control group
(T3).

This hypothesis was not supported by the two-way analysis of
variance. As noted, however, these results are equivocal because of
the restricted sample size resulting from the imagery categorization
procedure. Thus, the conclusion is made that there is insufficient
data in the present study to conclude one way or the other regarding
the advance organizer treatments.

The results of the correlation and factor analyses were consis-
tent with the two-way analysis results. That these additional analyses
did not detect any effects due to the advance organizer treatments
indicate that further study is needed in the design of advance
organizers, both verbal and nonverbal.

H12: The means of the high-imagery subject group (11) will

exceed those of the low-imagery subject group (12).

81

This hypothesis was not supported by the two-way analysis of
variance. These results, however, are equivocal for two reasons.
First, this hypothesis test suffers from the sample size restriction
that made a conclusion based on H11 difficult. Further, the correla-
tion and factor analyses indicate that a composite imagery factor was
not, in fact, measured. Thus, the conclusion is made that imagery was
not appropriately measured, and no conclusion can be made regarding
H12. Additional study is needed to increase the knowledge of what
constructs mental imagery instruments measure. Further, the relation-
ship and consistency of existing mental imagery instruments warrants
continued investigation.

H13: There will be a type of advance organizer (T1, T2, T3)
by subject imagery level (11, 12) interaction.

This hypothesis was not supported by the two-way analysis of
variance. Again, both the factors of small cell size and apparent
multiple components in the imagery measure make this result equivocal.
The conclusion is made that the interaction was not adequately tested.

Additional correlation and factor analyses yielded results
that were consistent with the earlier findings, and thus support these
conclusions.

Two factors partially vitiated the findings of the present
study. First, the relatively high incidence of subject mortality
due to data incompleteness weakens internal validity of the study.
Second, although sex was not found to be an important variable in the
literature review of advance organizers and mental imagery, the sex

distribution of the present study (eight males, twenty females)

82

represents an imbalance that may have biased the results in an uncon-

trolled manner.

Discussion

 

The results of testing the three hypotheses do not support the
assertions that type of advance organizer or imagery level are signifi-
cant factors in learning performance. Several factors common to all
three hypotheses preclude outright rejection of the assertions, how-
ever. First, an effect of the imagery categorization procedure using
Boutwell's Noun Test and Sheehan's QMI Test was to eliminate forty-
four of seventy-two subjects, a reduction of 61 percent. The remaining
twenty-eight subjects constituted a quite small sample for a two-way
factorial analysis of variance. Thus, the negative results of the two-
way analysis are equivocal.

Second, the correlation analysis, performed with seventy-one
subjects (one less than the total sample available) revealed that the
three scores of the Boutwell Noun Test (high words mean, low words
mean and overall mean) and the mean of items on Sheehan's QMI Test
had very small correlation coefficients. This indicates that the
assumption that these instruments were measuring the same thing,
imagery, is questionable. This concern is heightened by the results
of the factor analysis. That analysis revealed that the two instru-
ments measured strong and separate factors for the sample of subjects
in this study.

Third, the art-style slides as the instructional treatment
represents a sample of one. Thus the variables examined in this study

should be tested using several different instruments.

83

The negative statistical results thus may not be surprising,
given the three factors described: small cell size, multiple factors
in the imagery measure, and an instructional treatment sample of one.
For the sample size used, treatment differences of great magnitude
would have been required to make the treatments variable statistically
significant. The power of the treatments used, compared to the
learning stimulus and posttest is not known. Either the power was not
sufficient to overcome the limited cell Size, or the treatments in fact
did not affect learning in this study.

In light of the multiple-factors aspect of the imagery measure,
it would have been difficult to interpret differences in learning
performance due to "imagery." As it was, the lack of imagery as a
significant factor may be due to a problem of: cell Size; multiple
factors in the imagery measure; or to the fact that no imagery differ-
ences existed that affected learning performance.

In the case of the interaction term, treatments by imagery
level, the small cell Size and multiple-factors aspect of the imagery
measure may well be responsible for the negative result. Another
possibility is that the interaction effect was not strong enough to be.
significant.

Finally. the specific instructional treatment used may have
been inappropriate in content, organization or presentation.

Additional possible causes of no significant differences in the
variables include the relatively small amount of learning that took
place. Treatment differences may exist only within some optimal range

of learning. Further, both the treatments and learning stimulus were

'84

short, and the dependent measure was administered immediately after the
learning stimulus. The particular values of these variables may have

reduced treatment differences.

Implications for Future Research

 

Future research should be continued to the extent that this
study is inconclusive. Several sets of variables lend themselves to
further exploration. The basic relationships between advance organi-
zers, instructional treatments, and tests of learning could be
explored. Appropriate variables would include the length of the
advance organizers, the length of the instructional treatment, the
type of content, type of learning test and length of learning test
delay. Pilot studies of materials in each of these areas would be
useful.

Other studies could vary the type of instructional stimulus
used. If the findings of this study were confirmed using different
instructional content, organization and presentation, broader conclu-
sions could be made.

An additional direction of investigation regarding the imagery
instruments is one of conStruct validity. The results of the present
study indicate that Boutwell's Noun Test and Sheehan's QMI Test
measure quite different things. Research is needed to learn more
about the constructs related to each instrument and especially how the
constructs of each instrument relate to each other. The development
of common theoretical constructs that relate across instruments is
seen as a central need for the use of multiple-instrument measures of

imagery such as the measure used in the present study. Studies that

85

lead to effective combinations of mental imagery instruments are
needed. Replication of such studies would be useful.

The present study was limited by the small number of subjects
actually involved in the statistical tests; the imagery categorization
procedure reduced the available subject group by over half. Future
research Should examine the efficacy of this categorization procedure
in terms of the number of subjects that are deleted. A fruitful
approach might be to both modify the procedure and attempt to acquire
larger numbers of subjects for studies such as these. Studies leading
to an effective combination of mental imagery instruments that use an
appropriate but minimum number of subjects would be useful.

Finally, the need for additional studies of both advance
organizers and mental imagery implies that studies such as the present
one, where the interaction effects of these two classes of variables
are examined, will not be fruitful until progress is made in each area
alone. Studies of interactions are dependent on the stability of the
constructs in each independent variable classification. Studies of
advance organizer and imagery interaction may have to await a greater

stability of constructs in each Of these areas.

APPENDIX A

PART I: LEARNING STIMULUS POSTTEST
AUDIO-TAPED INSTRUCTIONS

PART II: DESCRIPTION OF EACH TEST SLIDE

LEARNING STIMULUS POSTTEST
AUDIO-TAPED INSTRUCTIONS

"The art style to be learned is called Art Nouveau. To see
whether you can identify examples of Art Nouveau among other art
styles, you will see twelve test items. Each test item will consist
of two slides of art styles and a third slide of the scale to use for
your response. Look at the scale of the printed test instruction.

For each test item please decide whether the art style to be learned,
Art Nouveau, is the first slide, second slide, both Slides or neither
slide. Then mark yOur choice on the answer sheet. For each test item

you will have ten seconds to decide after you see both art styles."

(Twelve test items are shown.)

86

Test Item*
1
2
3

10
11
12

87

DESCRIPTION OF TEST SLIDES

First Slide

 

Glass vase*
Bead necklace

Porcelain statuette
of girl*

Painting of woman

Ceramic vase*
Painting of woman*
Necklace*

Framed mirror
Stuffed chair
Glass sculpture
Glass vase

Building

Second Slide

 

Glass vase

Upright phonograph*
Jewelry pin
Painting of man,
woman, dog

Metal vase*
Painting of woman*
Jewelry pin

Table*

Lounge chair*
Glass sculpture*
Statue

Chair

*Note: The slides marked with (*) are examples of Art Nouveau.

APPENDIX B

PART I: LEARNING STIMULUS AUDIO-TAPED INSTRUCTIONS

PART II: DESCRIPTION OF LEARNING STIMULUS SLIDES
AND SLIDE SEQUENCE

LEARNING STIMULUS AUDIO-TAPED INSTRUCTIONS

"The purpose of this learning task is to teach you to identify
one particular art style among several art styles. First, you will
examine some materials in the package you have to help you in the
learning task. Then you will view some slides that include examples
of the art style to be learned. Finally, you will be tested to see if
you can identify the particular art style to be learned, among others.
After the learning task a questionnaire will be distributed. Following
its completion, any questions about the experience will be answered.
Please turn the page and examine these materials for the next minute
and a half," (Ninety-second pause.)

"Now, please direct your attention to the screen and we'll start
the slides. Remember that each slide showing the art style to be
learned will have a bright spot in the upper right corner. Like this.
Only slides of the art style to be learned will have this bright Spot.
Now view these slides. No sound will accompany them!’ (Twenty-four
slides are shown.)

"Now, for the next series of slides, pictures of the art style
to be learned will first be shown without the bright spot, then
repeated with the Spot. Please try to decide whether each of the
following slides is the art style to be learned before the second,
confirming slide. A dark screen following a slide will confirm that

slide as a non-example of the art style."

88

Slide Number

mem-DOON-d

NNN-fl—J-fi—l—l—od—l—l—l
N—JOKOCDVOSU‘l-wa—IO‘D

three.

89

DESCRIPTION OF LEARNING STIMULUS SLIDES

AND SLIDE-SEQUENCE

Content

Black

Non-art example

Black

Statuette

Ceramic vase

Three statuette/figures
Ceramic vase
Illustrated plate
Illustrated plate
Three statues

Statue

Two statues

Table

Chair

Chair

Table and chairs
Cabinet

Cabinet

Cathedral interior dome
Group of statues
Porcelain statuette of girl
Cathedral

Art Nouveau example marked by (E), nonexample by (N).
slides marked by (*).

Keyed

This slide was also used in the posttest sequence, in item

23
24
25
26
27
28
29
3O
31
32
33

35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52

 

3
played here.

90

Statue

Skyscraper

Mezzanine

Statue

Building

Black

Lounge chair

Lounge chair (same)
Ceramic forms

Black

Illustrated plate
Black

Vanity

Vanity (same)

Pewter pitcher
Pewter pitcher (same)
Cathedral

Black

Building Interior
Black

Clock

Black

Painting of port and city
Black

Entrance to building
Entrance to building (same)
Ceramic forms

Black

Statue

Black

Audio comment regarding delayed confirmation of examples

APPENDIX C

VISUAL ADVANCE ORGANIZER TREATMENT

VISUAL ADVANCE ORGANIZER TREATMENT

The visual advance organizer consisted of six photographs;
each photograph included several Objects. In the diagrams below, the

mark (*) indicates an object related to Art Nouveau.

 

 

 

 

 

 

 

 

 

flowers* pliers
l.
.3
U)
m d)
P 7; s
01 flowers* pliers +4 .3
Q.
m 3,.
C 3
.,_ U
3 in
2.
k
n O
, g 3-p1ece sculpture m
0.) O in
in S. rd
in .D it >
> in
m (D
'3 (U l_' S... _v.
i5 23 f * ' .3-3 8
o m lowers plTers a, Q ,_
4: u c E 3 U
m 0'.- U a,
U) 2 m in
ft! (6
> >
3.

 

 

 

91

 

92 -

 

Photo 4 has the same objects
as photo 3, but a large black
"X" crosses the objects on
the right

 

 

 

 

 

 

 

.k
_c
U
0 i1!
4: o m
(D S- U1
U1 .Q-k to
1U x I—
> 8 °' flowers*
I— (D
U C
a E
O)
m
M
>
5.
'k
g 3 flowers*
.Pw
30--
U)

 

 

 

APPENDIX D

VERBAL ADVANCE ORGANIZER TREATMENT

VERBAL ADVANCE ORGANIZER TREATMENT

Consider how an art-style might develOp and thus be identified.
One style might develop out of an awareness of nature, where many forms
are smooth, flowing, and exhibit a resolution or completeness. A
second style might develop from an awareness of the world of mechanical
forms and shapes, where shapes are joined based on function, utility
and design. Each style, as a dynamic concept, grows and develops in
its given direction, never rejecting its beginnings. Over time, many
styles are created; they grow and develOp in many directions. .

We are concerned here with the first art-style noted, where the
basis is a flowing, smooth form based on natural lines. AS this style
deveTOped, all art-forms were affected: sculpture and glass work,
painting, furniture and architecture. In each of these art-forms,
flowing flowery designs were executed. Smooth, circular shapes were
etched with detailed patterns found in nature. These patterns were

well integrated with the Shapes of the art-forms.

93

CONTROL GROUP TREATMENT

Masterpieces are created by tradition as well as by individuals.

The original meaning of masterpiece was Specialized. It designated an

 

object which qualified a young worker in the arts for the title of
master in medieval guilds. The guild itself, like its successor and
rival the humanistic art academy, was a powerful instrument of tradi-
tion. As a continuing institution, it 'handed on' from one generation
to the next a body of principles and attitudes toward the making of
art. It could provide a foundation, a basis of information and
training--a platform--from which a man of genius might rise to create
the consummate examples of craft and imagination which later periods
come to call masterpieces in rather looser interpretation of the
original meaning of the word.

A masterpiece may mean simply an unusually fine example of a
style or period, rather than an absolute summit of achievement. Even
so, the older relationship with tradition persists. For a masterpiece

is not recognized until it is seen against a large background.

94

APPENDIX F

FIFTY NOUN TEST WORDS WITH MEANS, STANDARD
DEVIATIONS AND HIGH WORDS/EXCLUDED
WORDS/LOW WORDS GROUPING

Word

Impropriety
Interim
Surtax
Instance
Forethought
Criterion
Distinction
Misconception
Elaboration
Concept
Outcome
Chance

Duty

Satire
Sobriety
Predicament
Management
Amount
Position
Effort

Idea
Hardship
Safety
Underworld
Discipline

FIFTY NOUN TEST WORDS WITH MEANS, STANDARD

DEVIATIONS AND HIGH WORDS/EXCLUDED
WORDS/LOW WORDS GROUPING

3
(D
9.1
:3

#hbwwwwwwwNNNNNNNNNNN-fld—l-fl

NOOGJVNNUT-DOKDQVO‘U'IU'IU‘I-bwfvo\DKONOS
«hm-HOGU'IN-kkodeO‘mO‘OONN-DmkoO-ho

Standard
Deviation

95

.31
.37
.60
.44
.45
.61
.61
.73
.65
.73
.74
.61
.73
.80
.99
.68
.oo
.09
.79
.80
.04
.96
.87
.05
.94

Low Imagery Words

l

Low Imagery Words
Excluded Words

1

1

Excluded Words

Word

Amazement
Length
Captive
Islander
Committee
Infection
Crime
Mathematics
Candidate
Victory
Breeze
Festivity
Destruction
Industry
Angle
Banner
Cell

Jury
Thief
Tomb

Monk

Moss
Accordian
Prison
City

3
(D
m
3

 

mmmmmmmmmmmmmmmmmmmmmbbh4:-

mbNNN—o—aoooooocnmh-c-Aww—ioooo\10101
#010501dmdhomwwowmmmN-‘NODO‘NO

96

Standard
Deviation

ddddddddd—J—Jddddddd—‘ddmddd

.81
.93
.93
.OO
.80
.88
.80
.98
.69
.72
.73
.56
.64
.68
.45
.54
.61
.52
.27
.58
.33
.56
.59
.09
.58

High Imagery Words

1

High Imagery Words

REFERENCES

REFERENCES

Ausubel, David P. "The Use of Advance Organizers in the Learning and
Retention of Meaningful Verbal Material." Journal of
Educational Psychology 57 (1960):261-72.

 

. The Psychology of Meaningful Verbal Learning. New York and
London: Grune and Stratton,’l963.

Barnes. Buckley R., and Clawson, Elmer W. "00 Advance Organizers
Facilitate Learning? Recommendations for Further Research
Based on an Analysis of 32 Studies." Review of Educational
Research 45 (Fall 1975):637-59.

 

Boutwell, Richard C. "Differential Imagery Ratings Using Nouns, QMI,
Prose and Inducing Encoding Strategy." Florida State
University, 1974. E0 110479.

Broadbent, D. E., and Gregory, Margaret. "Accuracy of Recognition for
Speech Presented to the Right and Left Ears.“ Quarterly
Journal of Experimental Psychology 16 (l964):359-60.

Calder, Nigel. The Mind of Man. New York: Viking Press, 1970.
Campbell, Donald T., and Stanley, Julian C. Experimental and Quasi-

Ex erimental Desi ns for Research. Chicago: Rand McNally
PuEIisfiing Co., 1963.

Christie, David 0., and Schumacher, Gary M. "Advanced Organizers,
Age, and the Recall Of Relevant Versus Irrelevant Thematic
Information." Paper presented at the annual meeting of the
American Educational Research Association, Washington, D.C.,
April 1975. ED 109619.

 

Crinella, Francis M.; Beck, Francis W.; and Robinson, James W.
"Unilateral Dominance Is Not Related to Neuropsychological
Integrity." Child Develgpment 42 (l97l):2033-54.

Dimond, Stuart J.; Farrington, Linda; and Johnson, Peter. "Differing

Emotional Responses from Right and Left Hemispheres."
Nature 261 (June 24, 1976):690-92.

97

98

Di Vesta, Francis J. and others. "Project Ikon: Studies of Imagery.
Section 1, Institutional Strategies: Multivariate Studies of
Psychological Processes Related to Instruction." Pennsylvania
State University, 000., July 1971. ED 055447.

Friedman, Norman, and McLaughlin, Charles A. "Two Forms: Exposition
and Argument." In Forms of Rhetoric: 0rdering_Experience.
Edited by Tom E. Kakonis andTJames C. Wilcox. New York:
McGraw-Hill, 1969.

 

Gagne, Robert M. "Learning and Instructional Sequence." In Review
of Research in Education. Edited by Fred N. Kerlinger
Itasca, IllihOis: AERA, Peacock Publishers, Inc., 1973.

 

Gazzaniga, Michael S. "The Split Brain in Man.“ Scientific American
Off rints, no. 508. San Francisco: W. H. Freeman and 00.,
I967.

 

Grady, Michael P. "Students Need Media for a Balanced Brain."
Audiovisual Instruction 21 (November l976):46-48.

Harris, Lauren Jay. “Functional Specialization of the Cerebral
Hemisphere in Infants and Children: New Experimental and
Clinical Evidence." Paper presented at the Biennial Meeting
of the Society for Research in Child Development, Denver,
Colorado, April 1975. ED 118271.

Hartley, James, and Davies, Ivor K. "Preinstructional Strategies:
The Role of Pretests, Behavioral Objectives, Overviews and
Advance Organizers." Review of Educational Research 46
(Spring l976):239-65.

Hovland, Carl 1.; Lumsdaine, Arthur A.; and Sheffield, Fred 0.
Experiments on Mass Communication. Vol. 111. Princeton,
New Jersey: Princeton UnTVersityTPress, 1949.

Kane, Thomas S., and Peters, Leonard J. “Objective and Subjective
Description." In_£prms of Rhetoric: Ordering Experience.
Edited by Tom E. Kakonis and’James CTTWilcox. New York:
McGraw-Hill, 1969.

Katzman, Nathan, and Nyenhuis, James. "Color vs. Black-and-White
Effects on Learning, Opinion, and Attention." AVCR 20
(Spring 1972):l6-28.

Kennedy, Keith. "The Effectiveness of a Comparative Advance Organizer
in the Learning and Retention of Metric System Concepts."
Paper presented at the annual meeting of the National Associ-
ation for Research in Science Teaching, Los Angeles, Cali-
fornia, March 1975. ED 107487.

99

Kimura, Doreen. "The Asymmetry of the Human Brain." Scientific
American Offprints, no. 554. San Francisco: W. HT‘Freeman
and’Co.,11973.

 

 

Lathrop, C. W., Jr. "Contributions of Film Introductions and Film
Summaries to Learning from Instructional Films." Technical
Report SDC 269-7-8, The Pennsylvania State College Instruc-
tional Film Research Program, November 1949.

Lawton, J. T., and Wanska, S. K. "Advance Organizers as a Teaching
Strategy: A Reply to Barnes and Clawson." Review of Educa-
tional Research 47 (1977):233-44.

 

 

Levie, W. Howard, and Dicke, Kenneth E. "The Analysis and Application
of Media." In Second Handbook of Research on Teaching,
Edited by Robert M. W. Travers,*Project Of the—American
Education Research Association. Chicago: Rand McNally and
Co., 1973.

Levie, W. Howard, and Levie, Diane 0. "Is There a Separate Visual
Iconic Memory System? Final Report." AVCR 23 (Spring 1974):
81-97. ED 095519.

Lumsdaine, Arthur A. "Instruments and Media of Instruction." In
Handbook of Research on Teaching, pp. 583-682. Edited by
NJTL. Gage. Chicago:TRand McNale and Co., 1963.

 

May, Mark A., and Lumsdaine, Arthur A. Learning_from Films.
New Haven: Yale University Press, 1958.

 

Mayer, Richard E. "Differences in Breadth of Transfer Due to Advance
Organizers and Subject Control of Frame Sequencing." Paper
presented at the annual meeting of the American Educational
Research Association, April 1976. E0 121284.

Northrop, Dean S. "Effects on Learning of the Prominence of Organi-
zational Outlines in Instructional Films." Documents
0210.2 SCD 269-7-33, Pennsylvania State College, Instruc-
tional Film Research Program, October 1952.

Ornstein, Robert E. The Ps cholo of Consciousness. San Francisco:
W. H. Freeman and Co., 197%.

 

Paivio, Allan. Imagery and Verbal Processes. New York: Holt, Rine-
hart and Winston, Inc., 1971.

Paivio, Allan; Yuille, John 0.; and Madigan, Stephen A. "Concrete-
ness, Imagery, and Meaningfulness Values for 925 Nouns."
Journal of Experimental Psychology Monograph Supplement 76
(January 1968), 25 pp.

Rennels, Max R. "Cerebral Symmetry: An Urgent Concern for Education."
Phi Delta Kappan 57 (March 1976):471-72.

100

Samples, Bob. "Educating for Both Sides of the Human Mind." Science
Teacher 42 (January 1975):21-23.

Sheehan, Peter W. "Functional Similarity of Imaging to Perceiving:
Individual Differences in Vividness of Imagery." Perceptual
and Motor Skills 23 (1966):lOll-33.

 

 

. "A Shortened Form of Betts' Questionnaire upon Mental
Imagery." Journal of Clinical P§ychology_23 (1967a):386-89.

 

. "Reliability of a Short Test of Imagery." Percgptual and
Motor Skills 25 (l967b):744.

 

 

Sherman, Jay L., and Kulhavy, Ray W. "The Assessment of Cerebral
Laterality: the Sherman-Kulhavy Laterality Assessment
Inventory." Technical Report No. 4, April 1976. E0 123861.

Stein, J. J. "The Effect of a Pre-film Test on Learning from an
Educational Sound Motion Picture." The Pennsylvania State
College, Instructional Film Research Program, State College,
Pennsylvania, November 1952.

Thorndike, E. L., and Lorge, I. The Teacher's Word Book of 30:000
Words. New York: Bureau of PfibliCatiOns, TeaChers COllege,

19447'

Travers, Robert M. W. Research and Theory Related to Audiovisual
Information Transmission. U.S. Department of Health,
EdUcation, andTWelfare, 1967.

Weisberg, Joseph S. "The Use of Visual Advance Organizers for
Learning Earth Science Concepts." Paper presented at annual
meeting of the National Association for Research in Science
Teaching (43rd), Minneapolis, Minnesota, March 5-8, 1970.

White, Murray J. "Laterality Differences in Perception: A Review."
Psychological Bulletin 72 (1969):387-405.

 

Winn, William. "The Structure of Multiple Free Associations to
Words, Black-and-White Pictures, and Color Pictures."
AVCR 24 (Fall 1976):273-93.

Wulff, J. Jepson, and Kraeling, Doris. "Familiarization Procedures
Used as Adjuncts to Assembly-Task Training with a Demonstra-
tion Film." In Student Response in Programmed Instruction,
p. 41. Edited by A. A. Lumsdaine. ‘Publication 9431
Washington, 0.0.: National Academy of Sciences, National
Research Council, 1961.

 

"Il1lllllllllllllllllllls