[‘12)fo .77.“ - J.‘
Jr 1 H144. -w. ..I u. ' ,2v 2““. it" L.‘°:'.'r-{'lh
".'l."""(“" l"

“‘r'
. . w. W “H

“II I I" V
‘59“ ' ”W 'H'
“1': WNW“. #:‘I

’14}

'fi'l‘ijr

I “"P'
M‘
- L

I,
‘ c
I

‘5‘ 1
‘cs 3'3“».
.flifi 3‘5 NJ.

‘1 ' ‘15

.3256 391' 5:4 32.:

shuv‘ .151”; ‘ “‘v
4
”L‘ll'jlll' u;

'm
. ‘ ' I. .L’
nfpjtbdifii IE ‘ 1%”
I ML. . “A“ at E"
D' I .‘ .A I n J
“ta-"r yt-j‘ J:$:rl. _‘|‘:—' ' .-
n J ' I
'v‘. .'-J' I ‘ Y, i
‘ ‘ 3“.” -. [=34

‘ ‘D~WT.‘

‘1‘“

h uV:_: 111» J': \‘fgenr
.4 '- ' -. A

‘1 ”.7. u".

3hr

:1. J
I 14 "
‘3‘“. ' n ‘ ' ',
“fl“:lfl'ifijffllhfi ', N

"J'hw'fl. ‘ .‘ 4.“_‘ .I ‘

-4::I‘I“t,-Talfi- ’, ~‘ - i-L ~. ,

.,r" - I > >
V _ e ’55 "
n ‘ - “9.; ’ “f . '-
_t. I’..‘I"< .1‘ I" 'Jé;' .
- . ”at?“

«. ;, I
7 .1".
A ‘ 3:va .h’:
.l-I )3.th ‘ "" J."
,JW

 

f,“ :1};fo 36a”

m

 

; JWillIzlfliflflllllllflliwfllfllfiml 7 __

00702 6390 1' a,
LIBRARY
Michigan State
I University 1

K I

——
fl

 

 

This is to certify that the

dissertation entitled
THE RELATIONSHIP OF SORTING ACTIVITIES
TO THE TORRANCE TESTS OF CREATIVE THINKING:
PICTURE COMPLETION TASK

presented by

John Bruce Shattuck

has been accepted towards fulfillment
of the requirements for

 

 

Ph.D. degree in Education

lb
Major of r
William Joyce

Date April 11, 1985

 

MS U is an Affirmative Action/Equal Opportunity Institution 0-12771

 

 

MSU

 

 

 

RETURNING MATERIALS:
Place in book drop to

 

 

LIBRARIES remove this checkout from
w your record. FINES Will
be charged if book is
returned after the date
stamped below.
, _.51§z
AU
5 21 2002

 

 

 

 

THE RELATIONSHIP OF SORTING ACTIVITIES
TO THE TORRANCE TESTS OF CREATIVE THINKING:
PICTURE COMPLETION TASK

John Bruce Shattuck

A Dissertation
Submitted in Partial Fulfillment of the
Requirements for the Degree of
Doctor of Philosophy
at
The Michigan State University
East Lansing, Michigan

1985

ABSTRACT
THE RELATIONSHIP OF SORTINC ACTIVITIES
TO THE TORRANCE TESTS OF CREATIVE THINKING:
PICTURE COMPLETION TASK
By

John Bruce Shattuck

The majority of creativity research has attempted to
investigate broad relationships between such areas as
creativity and personality traits, intelligence, socio—
economic levels, and school learning environments. The
classroom teacher, while frequently the subject of role
associated functions as they relate to creativity, is in
need of empirically sound approaches that can be integrated
into daily classroom activities and curricular models.

The need for this investigation, then, was promulgated
by (I) the apparent minimal attention to the selection of
researchable curriculum materials and activities, and (2)
the dearth of classroom oriented experimental investiga-
tions that utilize the classroom as a frame of reference to
buttress theoretical constructs.

The sample for this study included ninety—two randomly
selected fifth grade students in a suburban public school.

A Solomon four—group was employed as the research design.

-ii—

John Bruce Shattuck

The intervention consisted of two exposures to sorting
activities. The sorting activity was basically defined as
the ability to classify or group a collection of objects
based upon either a specified rule or pre-established
criteria selected by the student. The Torrance Test of
Creative Thinking, Picture Completion Task Forms A and B
were utilized as the independent and dependent variables,
respectively.

Major findings included: (1) subjects who were
administered the pretest received higher posttest means
regardless of treatment, and (2) subjects who received the
intervention received higher posttest means than those not
exposed to the intervention.

This study demonstrated the need to identify
appropriate classroom activities that appear to facilitate
creative thinking. Sorting activities do appear to have
this potential, however, the data only partially supported

this notion.

—iii-

DEDICATION

To Kathy
Without whom this document would

still be an unfinished thought.

_iv—

ACKNOWLEDGEMENTS

I would like to express my appreciation to the follow-

ing people for their contributions to this effort.

To William Joyce, for his long continued support and
guidance that went well beyond this manuscript;

To William Schmidt, for his research expertise,
teaching and review of statistical interpretations;

To Bob Cipriano and Rocco Orlando for their confidence
so generously given;

To Harvey Clarizio, for his assistance in reviewing
this document.

-v-

TABLE OF CONTENTS

PAGE

ABSTRACT ........................................... ii
DEDICATION ...... ...... ...... ...... ................. iv
ACKNOWLEDGEMENTS .................... ... ....... ..... v
TABLE OF CONTENTS .................................. vi
LIST OF TABLES ....... ..... .......... . ..... ......... viii
LIST OF FIGURES ........................... .... ..... ix
CHAPTER I ..... . .................................... 1
THE PROBLEM ...... . .......................... 1
Statement of the Problem .................... l

The Hypotheses .............................. l

The Delimitations ......................... .. 2

The Definition of Terms ...... . .............. 2
Assumptions . .................... . ........... 3
Importance of the Study ...... . .............. 3

Need for the Study ................... . ...... 4
CHAPTER II ......................................... 12
THE REVIEW OF THE RELATED LITERATURE ........ 12
Introduction .... ................... . ........ 12
Environmental Conditions .................... 13
Teaching Methodologies and Curricula ....... . 22
Summary ................. ...... ........... ... 29

.vi;’

TABLE OF CONTENTS Page

CHAPTER III ........................................ 31
METHODOLOGY AND PROCEDURES .................. 31
The Setting ....... .................... ...... 31
The Population ........... . .................. 31
The Instrument ......... ......... .... ........ 34
Procedures . .................. ..... .......... 38
Sorting .. .................... .. ............. 40
CHAPTER IV ..... ...... .............................. 45
PRESENTATION OF THE DATA ..... . .............. 45
The Design ......... ’ ........... . ........ ..... 45
Homogeneity of Variance .. ................... 45
Pre and Posttest Data ....................... 46
One—Way Analysis of Variance ................ 52
Analysis of Covariance ...................... 56
Two-Way Analysis of Variance ................ 65
CHAPTER V .......................................... 74
SUMMARY, CONCLUSIONS, AND RECOMMENDATIONS ... 74
Summary ..................................... 74
Conclusions ................................. 77
Recommendations ............................. 84
REFERENCES ......................................... 88
APPENDIX A ......................................... 93

-vii-

LIST OF

Analysis Groups .....

TABLES

Summary Statistics, Fluency ....

Summary Statistics, Flexibility

Summary Statistics, Originality

Summary Statistics, Elaboration ..... .... ..
ANOVA, Fluency .... .............. . ...... .....
ANOVA, Flexibility .... ......... . ............
ANOVA, Oirginality ......................... .
ANOVA, Elaboration . ..... . .................

T-Test Values and Probabilities Matrix:

Fluency .............

T-Test Values and Probabilities Matrix:

Flexibility . ........

Originality .........

T-Test Values and Probabilities Matrix:

T-Test Values and Probabilities Matrix:

Elaboration .................................
ANCOVA, Fluency .............................
ANCOVA, Flexibility .........................
ANCOVA, Originality .........................
ANCOVA, Elaboration .........................
Two-Way ANOVA, Fluency ......................
Two—Way ANOVA, Flexibility ..................
Two—Way ANOVA, Originality ........... . ......
Two-Way ANOVA, Elaboration ..................

-viii-

Page
45
47
48
49
50
51
51
52

52

53

54

55

56

59
61
63
66
68

7O

LIST OF FIGURES

FIGURE Page
1.1 Adjusted and Unadjusted Fluency Means ....... 58
1.2 Adjusted and Unadjusted Flexibility Means ... 60
1.3 Adjusted and Unadjusted Originality Means ... 62
1.4 Adjusted and Unadjusted Elaboration Means ... 64
2.1 Two-Way ANOVA Model .............. ........... 65
2.2 Two-Way ANOVA, Flu. Adjusted Means .......... 67
2.3 Two-Way ANOVA, Flex. Adjusted Means ......... 69
2.4 Two-Way ANOVA, Orig. Adjusted Means ......... 71
2.5 Two-Way ANOVA, Elab. Adjusted Means ......... 73

nix...

CHAPTER I

THE PROBLEM

Statement of the Problem

 

This research proposes to investigate if a relationship
exists between sorting activities and the Torrance Tests of

Creative Thinking: Picture Completion Task.

The Hypotheses

 

The first hypothesis is that subjects exposed to the
sorting activity will exhibit higher overall posttest scores
than subjects not exposed to the sorting activity as measured
by the Torrance Tests of Creative Thinking: Picture
Completion Task, Form B.

The second hypothesis is that subjects exposed to the
sorting activity will score highest on measurements of
originality on the posttest measure of the Torrance Tests of
Creative Thinking: Picture Completion Task, Form B.

The third hypothesis is that following originality,
subjects exposed to the sorting activity will then exhibit
highest posttest scores for measures of flexibility followed
by fluency and elaboration as measured by the Torrance Tests

of Creative Thinking: Picture Completion Task, Form B.

The Delimitations

 

The study will be limited to the fifth grade population
attending the Beecher School North, Woodbridge Public
Schools, Woodbridge, Connecticut.

The study will not attempt to identify creativity per se
in elementary school children or to compare subjects in the
study with other school populations.

The study will not attempt to examine the merits of the
school learning environment of this investigation as it
relates to the identification, extension, or facilitation of
creativity.

The study will not investigate possible relationships

between intelligence and creative thinking.

The Definition of Terms

 

Creativity. The ability to generate new responses to

 

problems presented by the available information or body of
knowledge.

Sorting. The process of classifying or grouping
objects, materials or elements according to a predetermined
criteria or rule.

Torrance Test of Creative Thinking: Picture Completion

 

Task. Part of a battery of tests designed to measure
creative thinking. The Picture Completion Task figuratively

measures frequency of response (originality), imagination

and exposition of detail (elaboration), quantity of figures
completed within apportioned time (fluency), and the number

of different categories into which the responses fall

(flexibility).

Assumptions

 

The first assumption. The first assumption is that
creative thinking abilities can be defined and measured
among elementary school populations.

The second assumption. ”The second assumption is that
activities can be developed to provide a facilitative
learning environment for creative growth among elementary
school populations.

The third assumption. The third assumption is that the

 

inclusion of empirically evaluated activities will be an

educational asset to the elementary school program.

Importance of the Study

 

The majority of creativity research seeks to establish
broad relationships between creativity and personality,
intelligence, socio-economic levels, environmental factors,
and personality traits. Much less has been accomplished to
illuminate approaches to the kinds of activities that appear

to be productive to creative thinking. The classroom

teacher, while frequently the subject of role function as it
relates to creativity, is in need of empirically sound
approaches that can be integrated into daily classroom

activities and curricular models.

Need for the Study

 

The process of education predates research in education
by a considerable margin. Historically, inquiry into
professional areas that included direct Observation or
experimental manipulation of variables constituted intel—

lectual heresy. As Traxler1 noted:

"...if the application of research to the under—

standing and control of the physical world is of
comparatively recent origin...the use of research
in the study of man and his development is indeed
a modern process. There is very little which can
properly be called research in any of the social
sciences that can be traced back as far as the
middle of the nineteenth century."

While educational research, particularly of an experi-
mental nature is a comparatively recent endeavor, creativity
research is in the initial stages of development. Research
devoted to virtually any dimension of creativity prior to

1950 is considered non-existent. The chronology is brief and

 

1Traxler, Arthur E., "Some Comments on Educational
Research at Midcentury," Journal of Educational Research,
XLVII, (1954), p. 359—66.

this fact must temper judgements relative to tendencies,
trends, and missing elements.

Contemporarily, creativity research has nonetheless
provided us with what can generally be referred to as
empirically defensible notions. These notions are largely
contained in broad descriptions of children's creative
characteristics and environmental factors related to those
characteristics. For example, creative children tend to be
more confident about themselves and their abilities to
conceptualize problems.2 They also seek to develop
independence and to rely on their own judgements. Creative
children appear to possess a spirit of inquiry--an
investigative prowess that seeks to differentiate relevant
from irrelevant information as a component of the problem
solving process.3

Environmental parameters generally agreed upon include:
(1) various social motivational and situational dimensions as
Aprofoundly effecting creative expression4, (2) a
considerable part of creative behaviors are learned and the

behaviors that underlie the creative process can be

 

2DeVito, Alfred, "Survival Through Creative Education,"
Journal of Creative Behavior, 10, No. 1, (1976), p. 45-61.

 

3E. Paul Torrance, Guiding Creative Talent, (Englewood
Cliffs: Prentice Hall, Inc., 1963), p. 16—42.

4Gowan, J.C. and Olson, M., "The Soceity Which
Maximizes Creativity," Journal of Creative Behavior, 13, NO.
2, (1979), p. 194—210.

 

facilitated by appropriate instructions, (3) educational
theory appears to recognize the need for attention to be
rooted in a dependence upon the transmission of infor-
mation6, and the creative process is both a basic part of
learning and a primary component by which the problem solving
process is accomplished7.

Once again it should be cautioned that inasmuch as the
chronology of creativity research is brief but rapidly ex-
panding, judgments relative to trends, tendencies and missing
elements should be tempered and placed in its proper longi-
tudinal perspective. The necessity for improvements in
educational practices per se are seldom dissimilar to those
addressing the proliferation of creative opportunity. Both
are largely dependent upon the nature of curriculum materials
and activities that ultimately surface in the teaching class-
room.

The need for this investigation is promulgated by: (1)

The minimal attention given to the selection of curriculum

 

5Crutchfield, R.S., "Nurturing the Cognitive Skills of
Productive Thinking," in Rubin, L.J., Life Skills in School
and Society, (Washington,'D.C.: Association for Supervision
and Curriculum Development, N.E.A., 1969).

 

 

6Cole, H.P., and Bernstein, 8., "Summary of Process-
Promoting Units Resulting from Curricular Analysis Activi-
ties," (Paper delivered at the Eastern Regional Insitute for
Education, Syracuse, New York, 1969).

7J. P. Guilford, The Nature of Intelligence, (New York:
McGraw~Hill, 1967), p. 67.

 

materials and activities as indicated by the literature, and
(2) The dearth of classroom oriented investigations that
utilize the classroom as a frame of reference to buttress
theoretical constructs.

Appropriate materials and activities-~the Objects of
instruction rather than the information of instruction--help
shift the focus of interaction from the teacher as a source
of information to the teacher as a facilitator of a problem
solving process. Teachers in this role are apt to be less
judgmental and more accepting of divergent thinking critical
to the creative process. Instruction utilizing objects are
less relevant upon "correctness" and instead promote a spirit
of inquiry to approach solutions to problems.

The literature indicates certain interpersonal/psycho-
logical dimensions as having a relationship to basic com-
ponents of the creative process. Brainstorming was described
by Osborn as a "creative conference for the purpose of
producing ideas that lead to the solution of a problem."8
During brainstorming, judgment is deferred, ideational
quantity is encouraged, and combinations and improvements are
sought. Since brainstorming offers an atmosphere with
limited evaluation in which free expression is encouraged, it

may be helpful in developing creative fluencies in children.

 

8A. F. Osborn, Applied Imagination, (New York:
Scribner's and Sons, 1963), p. 39.

 

Studies by Parnes and Meadow9 using brainstorming have
beenconducted at various grade levels and with students of
different levels of ability and appears to substantiate its
value as a method of increasing subjects' creative fluencies.
Sharp10 applied brainstorming techniques to educationally
handicapped children with similar results. As a corollary to
focusing attention on brainstorming per se, sorting activi-
ties appear to be suitable for using this technique as a
vehicle for instruction during the activity. Empirical data
is needed to lend credence to this notion.

Sorting activities also require involvement and active
participation by the learner. Educational research has
indicated the apparent positive relationship between higher
levels of involvement and motivation during periods of in-
struction. These components encourage intrinsically oriented
reward patterns that are a part of creative discovery and

exploration.11 Minimal investigations conducted during

 

9Parnes, S.J., and Meadow, A., Effects of Brainstorming
Instructions on Creative Problem Solving by Trained and
Untrained Subjects," Journal of Educational Pyscholo v, SO,
1963.

 

10Sharp, Lawrence Wesley, "The Effects of a Creative
Thinking Skills Program on Intermediate Grade Educationally

Handicapped Children," Journal of Creative Behavior, 7, No.
5, (1973) p. 37.

 

11Mihaly Csikszentmihaly, Beyond Boredom and Anxiety,
(San Francisco: Jossey—Bass, Inc., 1975), p. 32.

 

high learner involvement are recorded and more are needed to
expose the function of involvement with the creative process.
Similarly, will higher levels of involxement characteristic
of sorting activities capitalize on pupil interest and
thereby reduce boredom? Intuitively, it would appear that
levels of boredom would be inversely proportional to creative
performance. The relationship of boredom as a dysfunctional
variable to creative performance has been suggested by
Schubertlz.

Scattered in the literature are reports indicative Of the
importance for selection of activities that are facilitative

13. However, most of these

of creative learning atmospheres
are completed prior to elaboration on the specific kinds of
activities and materials that can be assimilated into the
classroomla.

The second area of need for this investigation relates to

the nature of the setting for creativity research itself.

The teaching environment in which learning strategies unfold

 

12Schubert, Daniel S.P., "Boredom As An Antagonist of

Creativity," The Journal of Creative Behavior, 11, No. 4,
1974.

 

13Yawkey, Thomas, D., "Facilitating Creative Thought
Through Object Play in Young Children," The Journal of
Creative Behavior, 17, No. 3, 1982.

 

 

14Penick, John E., "What Research Says: Encouraging
Creativity," Science and Children, 20, No. 5, (1983), p.
32-330

 

10.

represent an appropriate milieu for impacting upon the
conduct of education. The apparent hiatus between the
research sophistication available and the lack of pragmatic
applications tends to create a practice versus research
dichotomy. The school classroom can provide the needed
orienting frame of reference for investigations. Torrance's
work with practitioners in the field poignantly addressed the
impact of studies within the everyday classroomls.
Investigations are also needed to obtain greater levels
of generalizability to the data. It would appear helpful to
pose questions to the frame of reference that initially
prompted the hypothesis. This is a direct vehicle to extend
the meaning of the study beyond the querry itself. Most of
creativity research does make use Of what is generally
accepted or at least defensible, but makes minimal impact on
the modification and extension of curriculum efforts due to
the lack of generated hypotheses that focus upon classroom

experiences as the frame of reference.

 

15Torrance, E. Paul, "Predicting the Creativity of
Elementary School Children--and the Teachers Who Made a
Difference," The Gifted Child Quarterly, 25, No. 2, (1981),
p. 55-61.

 

11.

Sorting activities are unique in that they lend them-
selves to experimental manipulation. Given proper design
considerations, it is less arduous to measure experimental
interaction effects with sorting activities than broader,
multi-dimensional approaches encompassing school attitudes,
school environments, and educational learning styles.
Knowledge of the dynamics of the aforementioned are necessary
and provide important data, but are experimentally cumber—
some.

The burden of responsibility relative to creative
behavior ultimately rests with the classroom practitioner.
The need to assist the classroom practitioner towards this
objective is an important responsibility of the educational
researcher. Conducted within the appropriate frame of
reference, classroom research can leave behind a legacy of

applicability.

12.

CHAPTER II

THE REVIEW OF THE RELATED LITERATURE

Introduction

The Creative Education Foundation listed some 4,176 items
pertaining to the nurture of creative behavior prior to 1967.
Since that date, this number has nearly quadrupled and is ex—
pected to reach 20,000 items in 1988. Seven basic types of
studies are found in the literature: (1) Differences between
high and low creative individuals regarding personality
structures, cognitive functioning, and biographical data; (2)
Research comparing individual creativity with that of groups;
(3) Analysis of relationships among creativity, intelligence,
and some measurement index of achievement; (4) Studies of
factors that inhibit creative thinking; (5) Studies evaluat-
ing programatic approaches to foster creative behavior; (6)
Studies of environmental variables affecting creativity; and
(7) Studies designed to identify the role of educators,
psychologists, and others in a helping relationship to better
understand the dynamics of creativity. Frequently discus-
sions of the related literature fail to make the nexus be-
tween discussions of the literature others have produced and
the researchable problem at hand. By establishing an appro-
priate criteria for inclusion, selectivity of entries can

both display relevancy and a sense of overall connectedness.

13.

In the discussions that follow, inclusion of
bibliographical data will meet the following criteria: (1)
Studies will directly relate to school populations or possess
intrinsic transfer value when considering school populations;

(2) Studies will focus upon specific variables that buttress
an understanding for creativity experimentation in the
schools; (3) Studies will expose a theoretical framework
significant to the research problem of this investigation;
and (4) Studies will be representative of other parallel

studies and will be included for illustrative purposes.

Environmental Conditions

 

Investigations relative to modes of classroom instruction
(i.e. open vs. traditional, process vs. content, inductive
vs. deductive) occupy a large segment of the literature.

Many educators believe that the open classroom facilitates
creative growth through its emphasis on self direction, cur—
riculum integration and decision making strategies. Assuming
that open classrooms necessarily encompass these strategies,
however, is researchably as unsound as assuming these

elements are lacking in traditional classrooms.

14.

Golub and Hahn1 compared the effects of creative
thinking programs on randomly assigned third-grade students
at a suburban public elementary school. During four thirty—
minute sessions per week for eight weeks subjects in the
experimental groups were given a variety of creativity
facilitating exercises and techniques. Both the pre and post
measures consisted of Parallel Forms A and B of the Torrance
Tests of Creative Thinking.2 Responses were scored for
fluency, flexibility, originality and elaboration.

No significant differences were found between the
traditional and open classroom control groups. Significant
differences between the control and experimental groups were
revealed on all measures following exposure to training. In
addition, gains in creativity were significantly greater for
experimental subjects in the open classroom, supporting the
view that with training the open classroom facilitates

creative expression.3

 

lGolub, Sharon and Hahn, Karen Sorci, "Training
Creative Thinking in the Open and Traditional Classroom,"
Journal of Creative Behavior, 17, No. 3 (1976), 217.

 

2E. Paul Torrance, Torrance Tests of Creative Thinking,
(Bensenville, Illinois: Scholastic Testing Service, 1974).

3Golub and Hahn, op. cit.

15.

In an investigation into the relationship of high
creative children and common characteristics of educational
backgrounds, Hahn4 examined student preferences for
classroom environments, learning styles, teaching methods and
teacher characteristics. Sixteen hypotheses were developed
for the study. They stated that higher creatives would have,
as example, attended nursery school; been early readers;
preferred independent learning; preferred a democratic
classroom structure and an open-style form of
instruction.5 Although high creatives in the study did
have preferences for open classrooms and independent learning
experiences, only these two of the sixteen hypotheses were
statistically supported.6

Despite the apparent lack of verified support, the Wright
study does have merit, however. The search for commonalities
that appear among a target population could lend support and
provide valuable data for longitudinally oriented
investigations. The weakness of the Wright study was not one

of poorly developed hypotheses that were empirically unsound,

 

4Hahn, Beverly A., "Common Characteristics in the
Educational Backgrounds of High Creative Children and Their
Preferences Regarding Classroom Behavior," Journal of
Creative Behavior, 15, No. 4 (1981), 283-285.

 

 

51bid, p. 284.

61bid, p. 285.

16.

but was due to a basic design flaw that would have com-
promised the validity of any results. The population con-
sisted of all 343 fifth grade students in the Auburn, Maine
Public Schools. The forty-five students who attained the
highest combined scores (figural and verbal) on the Torrance
Tests of Creative Thinking were identified as high creatives.
Thus, in excess of 13 percent of the total population were
labeled "high creatives"——a statistical near impossibility.
The r.ndomized use of forty—five scores should have been
deleted in favor of national norms for high creativity in the
fifth grade, or less than five percent.

External validity of the investigation was further mini—
mized by such factors as demographic mobility (some students
may have attended several different schools); quality of edu—
cational opportunity (relative age at when reading abilities
were professionally diagnosed if at all); preferences by
students relative to learning atmospheres (presumes students
have basis for comparison).

To reinterate, the intent of the Hahn study has merit.
The methodological problem was contained in researchers only
identifying what can be measured and measuring only what can
be quantified. In the Hahn investigation, what was label.ed

as "high creative' was that which could be quantified to fit

the study.

17.

Much of the research findings appear inconsistent when
examining the relationship of classroom environments and
creative thinking. Haddon and Lytton7 reported superior
creativity scores for children in progressive schools than
traditional ones. Research conducted by Wilson, Stuckey and
Langevin8 found open-class children were less creative than
those in a traditional class. While results appeared to
focus on either the positive or negative attributes of kinds
of classroom environments, other alternative findings added

9

to overall disparities. Bennett's extensive work with
elementary school children found little or no differences in
creativity as compared to school environments. Wright's10
findings provided collaboration to this theory.

It would appear more educationally sound to address the

classroom environment question from other than traditional

vs. progressive perspective. That is, perhaps certain kinds

 

7Haddon, F.A. and Lytton, H., "Teaching Approach and
the Development of Divergent Thinking Abilities in Primary
Schools," British Journal of Educational Pyschology, 38,
(1968), 171—180.

 

8Wilson, R.S.; Stuckey, T.; and Langevin, R., Are
Pupils in the Open Plan School Different?", Journal of
Educational Research, 66, (1972), 115-118.

 

 

9N. Bennett, Teaching Styles and Pupil Progress,
(Cambridge, MA: Harvard University Press, 1976).

 

10Wright, R. J., "The Affective and Cognitive
Consequences of an Open Education Elementary School,"
American Education Research Journal, 12, (1975), 449-468.

 

18.

of measurable creativity would appear to be more
appropriately compared to certain kinds of classroom
environments. Consistent with this notion, Ramey and
Piper11 examined a broader range of classroom continuums
and found a strong relationship of open-classroom figural
creativity and traditional classroom superiority in verbal
creativity.

Some consideration for the heterogeneity of results
should be discussed. The basic problem may be contained
within the labels attached to school types. One can assume a
reasonably close chronological population within a given
grade level, or a reasonably accurate description of
content/skill development exposure contained within a school
year. What is needed is a more behaviorally oriented
definition of school types. The terms "open" or "free" are
not precise ”nd when applied to a research endeavor can
represent confounding variables not accounted for. Terms
employed by educators may frequently be more impressionistic
than behaviorally reliable. While "open" and "free" may
indicate what the school is not--precisely what the school is
may represent undetected variables to the investigator.

A study by Thomas and Beck12 explored the effects of

 

11Ramey, C.T. and Piper, V., "Creativity in Open and
Traditional Classrooms," Child Development, 45, (1974),
557-560.

 

19.

varying school environments on the development of creativity
over the course of a school year but improved upon prior
research in the following ways: (a) By systematically
measuring variations in schooling along a continuum. The
study focused upon an intermediate level of school formality

in which there existed "

...a dual emphasis on both fact
acquisition and leeway for self-expression that provided the
optimal environment for growth in creative ability;"13 (b)
Since children's capacity for creativity had been predicted
by other studies to be based upon school environments, Thomas
and Beck investigated school learning interactions with
teachers' views of student behaviors under those learning
interactions; and (c) In order to appropriately evaluate the
equivalence of school samples, parental characteristics
related to creativity were added. These variables included
mothers' and fathers' own creativity and their attitudes
about child characteristics' associated with creativity.14

This added design factor towards a greater assessment of

subject equivalence had not been observed in other studies.

 

12Thomas, Nancy C. and Berk, Baura E., "Effects of
School Environments on the Development of Young Children's
Creativity," Child Development, 52, (1981), 1153-1162.

 

13Ibid., p. 1154.

141b1d., p. 1155.

20.

All children, grades one and two in six schools partici-
pated. There were 126 boys and ninety-nine girls. The over-
all population was middle class. An interview system
gathered data on the school descriptions; these descriptions
were then rated along a continuum from extremely informal to
extremely formal. The raters were specialists in elementary
education and/or child development. Ratings were based upon
ten dimensions. These included fact acquisition, subject
matter areas, academic achievement, evaluation, verbal and
artistic expression, self-awareness, peer relationships,
authority, range of individual behavior exhibited, and group
behavior exhibited. Parents were asked to respond to two
tasks: (1) to assess their own creativity, the Torrance
Thinking Creativity With Pictures, Form A was administered,
and (2) to investigate parental values of personality
characteristics associated with childrens' creativity, The
Torrance Ideal Child Checklist was administered.15

The results of the study indicated the relationship
between school environments and creativity development to be
more complex than previously observed. It depended upon (a)
type of school; (b) sex of the child; and (c) the particular

d.15

component of creativity assesse The findings supported

the hypothesis that children in the intermediate environment

 

15Ibid.

16Ibid, p. 1160.

 

   

21.

tended to perform better than children in the formal environ-
ment. Girls showed the greatest gain in originality in the
informal schools, modest gains for intermediate schools, and
a decrease in the formal schools. Boys gained in originality
in all three school environments. On the flexibility
measure, both boys and girls gained most in the informal
schools. For the elaboration measure, girls showed greatest
gains in the intermediate school and boys showed greatest
gains in the formal environments and lower gains in the
informal settings.17

In considering school environments the Thomas and Berk
investigation recognized the multiplicity of interrelated
elements and addressed the need to avoid oversimplification
of formal vs. informal educational settings.

Ogilvie18 added credence to the Thomas and Berk study
with her statistically supported hypothesis that a curvi—
linear relationship exists between the degree of school
formality and children's creativity. Ogilvie suggested that
the concepts of formal and informal be thought of as extremes
on a continuum. Her rationale was that highly formal
environments allow little room for nonconformity and highly

informal environments offer insufficient information

 

17Ibid., p. 1161

18Ogilvie, E., "Creativity and Curriculum Structure",
Educational Research, 16, (1974), 126-132.

22.

reservoirs on which creative associations can be based.19
Ogilvie states "The environment most favorable to creativity
will be that which provides for both freedom of expression

and good quality association reservoirs."20

Teaching Methodologies and Curricula

Investigations concerning school environments and their
relationship to creativity development were usually conducted
after the establishment of the individual school programs.
That is, traditional, open and alternative settings were not
established to reflect their relative impact upon the crea—
tive process, but later became the mileau in which hypothesis
could be tested. What then of specific curricular
approaches, teaching styles, training efforts, classroom
activities and "climate" variables and their relationships to
create behavior?

Possien,21 using map studies as a vehicle, evaluated
specific teaching methodologies in the development of problem
solving skills among sixth grade children. Included were the

inductive method, with a focus upon the processes of

 

191b1d, p. 127.
201b1d, p. 129.

21Possien, Wilma Martens, "A Comparison of the Effects
of Three Teaching Methodologies on the Development of the
Problem Solving Skills of Sixth Grade Children," Journal of
Creative Behavior, 1, No. 2, (1967).

 

 

23.

searching and self discovery; the deductive method, comprised
of information giving by the teacher and memorization by the
students; and a third method deductive in nature but
"detailed explanations of the causal relationships underlying
the concepts were added to lesson plans."22 Data for com-
parison of methodological effects were derived from pre and
post scores on the map reading section of the Iowa Test of
Basic Skills: Form I and on mental age by the California Test
of Mental Maturity. Hypotheses for the investigation were
twofold: (1) Differences in teaching method would not effect
differences in achievement; and (2) Variations in methodology
would not effect differences in problem solving behavior
exhibited by the pupils.23

Results for the hypothesis that differences in teaching
method would not effect differences in achievement were
supported and this hypothesis was accepted. For the second
hypothesis, that variations in methodology would not effect
differences in problem solving were found to be significant
and this hypothesis was rejected. Possien concluded that
students trained in the use of inductive procedures exhibit

some characteristics of problem solving more frequently than

 

22mm, p. 87.

23Ibid, p. 89.

24.

pupils taught by the deductive method.24

Intuitively it would appear that inductive approaches
offer a more stimulating learning atmosphere and therefore a
greater opportunity for individual growth. However the
Possien investigation cannot be considered upon intuition but
on the merits of its research methodology per se. Validity
of the Possien study was reduced in two primary areas.

First, the treatment indicated each of the classes were
instructed for thirty minutes each day over a three week
period. All instruction involving each of three teaching
approaches were taught by the investigator. Given the
latitude of potential instructor/student interaction during
the instructional periods, experimentor bias could be
significant. Further, the experimentor could have preferred
one method over the other two, had prior training in a
particular teaching style, or unknowingly possessed greater
communicative expertise in one mode of instruction. The pro‘
cedures of the study suggest an alternative hypothesis:
Students exposed to the Possien method of inductive teaching
will exhibit higher gains in problem solving than other
methods employed by the same instructor.

Second, pre and post instruments were identical. While

the appropriateness of the instruments can be left to

 

24mm, p. 90.

25.

conjecture, the utilization of parallel forms would have
reduced the possibility of pre and post interaction as a
confounding variable. The consideration of test-retest
effect was not addressed in the investigation.

It is nearly two decades since E. Paul Torrance reported
the fourth~grade slumps in creative teaching.25 Torrance's
hypothesis was that certain periods of stress in children's
lives contribute to behaviors that cause discontinuities in
creative growth. More specifically, it is at about the
fourth grade level that for the first time in pupils' lives
they are expected to be regimented into certain academic
molds imposed by teacher, peer, and parental pressures for

26 The work of Torrance in this area

school success.
pointed out the relationship between self worth and creative
growth as a major variable in determining fourth grade losses
in creative thinking.

An investigation by Williams27 began as a primary

prevention mental health project for schools, parents and the

community at large. The project was an effort to reduce

 

25E. Paul Torrance, Understanding the Fourth Grade
Slump in Creative Thinking, (Washington, D.C.: U.S. Office of
Education Final Report, December, 1967).

 

 

26F. E. Williams, Classroom Ideas for Encouraging
Thinking and Feeling, (Buffalo, New York: D.O.K. Publishers,
1970), p. 93.

 

 

27Williams, F. E., "Rediscovering the Fourth Grade

Slump in a Study of Children's Self-Concept," Journal of
Creative Behavior, 10, No. 1 (1976), p. 15-28.

 

 

26.

excessive stress by building coping mechanisms that would
assist people to better handle daily situations prior to the
development of mental problems requiring treatment.28
Although not a new concept in the field of mental health, it
was innovative to education. During this program, teachers
and counselors were expected to use self—concept scales on a
diagnostic basis by prescribing individual exercises in the
classroom. These exercises were designed to develop
children's feelings about themselves, others and school. The
salient question of the project was whether the technique,
when applied to children in a school setting, would eliminate
or reduce the fourth grade slump in creative thinking.29

The program was undertaken in two inner-city schools
having a population of fifty—five teachers and in excess of
one thousand students in grades one through six. All staff
participated in the training but pre and posttests were
administered only to grades two, three and four. The
instrument utilized for both pre and post measures was the
Self Concept and Motivation Inventory (SCAMIN). Two scores
are tabulated from this instrument: school self concept and
motivation. Student self concept scores were derived from

measured pupil role expectations and feelings of self

 

281b1d, p. 17.

291bid, p. 23.

27.

adequacy. School motivation scores were derived from
achievement investment--or the degree of desire to avoid
school failure.3O All pretests were conducted during the
first full week of the school year.

Treatment for the experimental classrooms included weekly
teacher inservice training sessions on strategies and
procedures for improving school self concept. Classroom
activities were intended to capitalize on student successes.
Positive reinforcement throughout the year focused upon
student trust, confidence and security in school. Activities
were integrated into all aspects of the school curricula in
an attempt to encourage learning while dealing with
feelings. Class and small group discussions on attitudes and
values clarification were further tied to content areas.31

Posttest scores revealed no fourth grade slump in school
self-concept or motivation among experimental groups after
treatment. T-tests indicated a significant (.05 signifi-
cance) growth of fourth grade post over pretest scores.
Treatment appeared to not only deter the fourth grade slump
but improved pupils' feelings about school and

themselves.32

 

3OIbid.
31Ib1d, p. 34.

321bid.

28.

In discussion, work previously cited by Torrance appears
to have been at least a partial catylist for the Williams'
investigation into school self—concept and motivation.

Indeed Williams had reinforced the need and relative value of
providing the necessary training for educators in order to
equip them with the mechanisms necessary to dually address
emotional and academic appetities of the elementary school
child. Eliminating or at least reducing the fourth grade
slump in school self-concept and motivation was statistically
supported. The relationship of school self-concept and moti-
vation upon creative thinking was implied but not empirically
tested. Follow-up investigations parallel to the Williams'
study could contribute to these dynamics by the inclusion of:
(1) Use of creativity instruments to identify specific rela-
tionships to figural and verbal forms of creative thinking;
(2) Base line data that considers performances on cognitive
parameters during programs specifically designed to address
affective factors; and (3) Utilization of populations with
broader generalizability to other school groups. While there
is a humanitarian need for support data to improve the
learning conditions of inner city children, their needs in
the areas of self concept and school motivation may be

different both in nature and degree.

29.

Summary

The development of creative thinking in elementary school
children is a complex issue. Research has been partially
successful in finding some of the ingredients but much more
is needed to improve the blend. Research to search for the
optimal school environment, program and teacher characteris—
tics have been met with a wide continuum of success.

Teacher workshops,33 undergraduate courses and
seminars,34 specialized degree programs,35 and research
fellowships36 all have made a contribution but appear to
lack grass-roots endurance. Nonetheless, the evidence does
suggest that creative abilities can be identified, measured,
and developed.

The most complicated arena of all is the classroom. In—
vestigations conducted in the classroom under day to day

learning conditions may be the most beneficial to both

 

33Feldhasen, John F. and Treffinger, Donald J., "Design
and Evaluation of a Workshop on Creativity and Problem-Solv-

ing for Teachers," Journal of Creative Behavior, 10, No. 1,
(1976), p. 12-14.

 

34Linder, Toni W., "Organizing a Course on Creativity:
Theory and Practice," Journal for the Education of the
Gifted, 4, No. 3, (1981), p. 211-224.

 

3SHolman, E. Riley, "A Master of Education Degree in
Elementary Education With An Emphasis on Creative Teaching-
Learning," Creative Child and Adult Quarterly, 11, No. 2,
(1977), p. 98—100.

36Harmin, Merrill, "N.E.X.T.E.P.: New Exploratory
Teacher Education Program," Southern Illinois University,
Edwardsville, Illinois, (1969).

30.

learners—-the teacher and the student. Simply stated, how
can teacher-student—material interactions best be structured
to allow the creative process to occur? Regardless of how we
measure the final product, the final product is a reflection

of the creative process.

31.

CHAPTER III

METHODOLOGY AND PROCEDURES

The Setting

 

Woodbridge, Connecticut is a suburban community of New
Haven with a population of 9,800. Essentially representative
of a middle to upper middle class community, Woodbridge has a
large proportion of professional citizenry associated with
New Haven and its three Universities.

The school district has an elementary grade enrollment of
775. Secondary and junior high schools are provided on a
regional basis. Woodbridge ranks in the top one-fourth of
169 school districts statewide in per pupil expenditures.
There are two elementary schools, with students attending
kindergarten through grade three in the primary unit and
grade four through six in the intermediate unit. Regular
classroom instruction is complemented in the areas of
mathematics, art, music and physical education. French and
literature instruction is also provided in the intermediate
school. Psychological services and special education are

both provided within the district.

The Population

 

There are ninety—one students enrolled in the fifth

grades in the intermediate school. With certain exceptions,

32.

all students are randomly assigned to each of the four fifth
grades at the beginning of each academic year. Exceptions
include siblings who are generally not placed in the same

class group and students considered to be behavior problems
who are not placed with other students of the same conduct.
Methodology of randomization by the school district involved
an overall ability ranking of each student prior to entering
the next school year. Teachers assigned numerical weights of
one through five to each student, with one as the highest.
Students were then selected to each classroom with equal
numbers of one, two, three, etc., assigned to each of the
four fifth grade classes. For purposes of this investiga-
tion, the four fifth grade classes were then randomly
selected to experimental and control groups. True randomiza—
tion of subjects from the initial phases of this investiga—
tion would have been more favorable and enhanced the external
validity of the study, however daily scheduling of the
classes for specialized instruction presented problems and
could not be performed. Results of the randomization process
in terms of the homogeneity of the groups is presented in
Chapter 4. .

Organization for instruction is largely based upon
content areas of the curriculum, with each of the four fifth
grade teachers responsible for one or possibly two content

areas. During a typical day, students move from class to

33.

class and are exposed to the content and methodologies of
each teacher. Support areas of physical education, French,
literature, art and music are each taught by specialists to
all students. All classes can be considered heterogeneously
grouped and all students receive instruction from the same
teaching staff.

The organization for instruction component was an
important design consideration for the study. In most
elementary schools classrooms are basically self contained
with added specialists in music, art and physical education
where budgetary restraints permit. Therefore, the bulk of
instruction emanates from a single source. In this investi-
gation no attempt has been made to consider the effectiveness
of teaching styles relative to creative stimulation. In
classrooms where a single teacher is responsible for instruc-
tion, measurements could be influenced by the manner of
instruction and not the experimental treatment while inter-
pretation of results would be based upon the treatment.
Classroom studies seldom can be totally isolated from teacher
induced influences. However, when subjects are exposed to
the same groups of teachers the potential for any unidirec-
tional influence is reduced. The assumption basic to this
premise is that all of the teachers responsible for the sub—

jects' instruction do not have the same approach to teaching.

34.

The Instrument

The Journal of Creative Behavior describes in excess of
one-hundred instruments for studying creative behavior.1
This proliferation, together with the focus upon development,
has created a spectrum of reactions among educational
researchers.2 Uses and abuses of existing creativity tests
were also identified by Torrance.3 Additionally, how
researchers perceive creativity tests and how the creativity
testing development movement has resulted in a significant
decrease in educators seeking to find new ways to encourage
creativity in a general sense appears to suggest a loss of
perspective.4

The importance of the study addressed the need for

classroom applied research involving classificatory oriented

activities. Replication would offer greater generalizability

 

1The reader is referred to the following previous
issues of The Journal of Creative Behavior: "Instruments
Useful in Studying Creative Behavior and Creative Talent,
Part I, Commercially Available Instruments," 5, No. 2, 1971;
"Instruments Useful in Studying Creative Behavior and
Creative Talent, Part II, Noncommercially Available
Instruments," 5, No. 3, 1971; "Additional Instruments Useful
in Studying Creative Behavior and Creative Talent, Part III,
Noncommercially Available Instruments," 6, No. 4, 1972.

 

2Crockenberg, Susan 8., "Creativity Tests: A Boon or

Boondoggle for Education?", Review of Educational Research,
2, No. 1, (1972), p. 27-45.

 

3Torrance, E. Paul, "Creativity Testing in Education,"
Creative Child and Adult Quarterly, 1, No. 3, (1976), p. 36.

4Crockenberg, op. cit., p. 30.

35.

if results appeared consistently significant. Since replica—
tion suggests the involvement of classroom practitioners,
choice of instrumentation presented a unique challenge if
replication objectives were to be met. The instrument would
have to possess acceptable levels of interscorer reliability
using either trained or untrained scorers. To reduce possi—
ble effects of testing, parallel forms would be necessary.
The instrument itself had to be accessible to classroom
personnel. Accessibility to classroom teachers does not
suggest the use of teachers to conduct their own investiga-
tions with their own classrooms, but the availability and
subsequent familiarity with use of the instrument should
encourage classroom personnel to be part of the research
process with other populations.

The Torrance Tests of Creative Thinking5 were ten years
in development. The tests are available in parallel forms
appropriately labeled "A" and "B". Test booklets, directions
manual and scoring guide, and a norms-technical manual are
available for purchase.6

Evidence concerning interscorer reliability among trained

 

5E. Paul Torrance, Torrance Tests of Creative Thinkipg,
(Bensenville, Illinois: Scholastic Testing Service, Inc.)
1974.

61bid.

36.

scorers were consistently in the acceptable range.7 Data
obtained from untrained scorers who used only the scoring
guides were of particular importance. Evidence from
experiments in which untrained classroom scorers evaluated
samples of test booklets completed by children at the grade
level at which they were teaching yielded favorable results
when compared with trained scorers who examined the same
booklets. Mean reliability coefficients for the Incomplete
Figures Activity ranged from .88 for originality to .96 for
fluency. Measurements for flexibility and elaboration
revealed reliability coefficients between these extremes.8

Statements relative to overall validity are at best

permeated by disagreement. Torrance has stated that an
overall validity coefficient for tests of creative thinking
is inappropriate,9 preferring to think in terms of kinds of
criteria for creative behavior. I

10

Reviews by Taylor and Yamamoto11 described the

 

7E. Paul Torrance, Torrance Tests of Creative Thinking:
Norms-Technical Manual, (Bensenville, Illinois: Scholastic
Testing Service, Inc., 1974). p. 17

 

 

81bid., p. 18.

91bid., p. 21.

10C.W. Taylor, (Ed.), Creativity: Progress and
Potential, (New York: McGraw-Hill Book Co., 1964), p. 81-92.

 

 

11Yamomoto, K., "Validation of Tests of Creative

Thinking" A Review of Some Studies," Exceptional Children,
31, (1965), 281-290.

 

37.

intricacies of validity assessment and lacked specific find—
ings on overall measures of validity per se. Treffinger12
cautioned that in the quest for developing creativity tests
that are managable from a statistical perspective, the tasks
contained within the tests may have minimal application to
real life creative accomplishments in the future. Similar
precautions relative to predictive validity were further
buttressed by Yaird,13 declaring that the Torrance Tests of
Creative Thinking need to be more firmly linked to reality by
showing predictive value in terms of socially valuable
behavior.

Numerous studies in support of the overall validity of
the Torrance battery were presented by Torrance in the Norms-
Technical Manual previously cited. The eficacy of these
findings should be left to the judgement of the reader. Iflg_
Torrance Test of Creative Thinking, Incomplete Figures

Activityj Form A was selected as the pretest for this investi—

 

gation. Parallel Form B of the Incomplete Figures Activity

was selected for administration as the posttest measurement.

 

12Treffinger, D.J., J.S. Renzulli, and J. F. Feldhusen,
"Problems in the Assessment of Creative Thinking," Journal of

 

Creative Behavior, 5, (1971), 104-112.

 

13Yaird, Leonard E., "The Torrance Tests of Creative
Thinking," in the Seventh Menta‘ Measurements Yearbook, Oscar
Krisen Buros, Ed., (Highland Park, New Jersey: 1972), 1, 837.

 

38.

Procedures

 

The four fifth grade classroom teachers at the intermedi-
ate school were provided with a brief description of the
study and were invited to participate. As stated earlier,
fifth grade classes are randomly formed at the beginning of
the school year and are heterogeneous in nature. Further
randomization solely for the investigation represented a
scheduling imposition for the school and did not occur. The
four fifth grade classes, however, were randomly assigned as
Groups I-IV.

Necessary test administration and activities were con-
ducted by individual class basis over a three week period.
The initial week was devoted to pretesting, the second for
the sorting activity, and the third for collection of post-
test data.

Groups I and III were administered the Torrance Incom-
plete Figures Activity, Form A. Prior to the start of each
pretesting session, subjects were read the following instruc—
tions.

"By adding lines to the incomplete figures on this

and the next page, you can sketch some interesting

objects or pictures. Try to think of some picture

or object that no one else will think of. Try to

make it tell as complete and as interesting a

story as you can by adding to and building up your

first idea. Make up an interesting title for each

of your drawings and write it at the bottom of

each block next to the number of the figure. All
right, go ahead. You will have ten minutes."

 

14E. Paul Torrance, Torrance Tests of Creative
Thinking: Directional Manual and Scoring Guide, (Bensenville,
Illinois: Scholastic Testing Service, Inc., 1974), p. 8.

 

39.

As per direction guidelines, the aforementioned repre-
sented the total verbal assistance to the subjects during
test administration. Apparently, even slight variations in
verbal instructions can alter performance on many of the
tasks typically included on tests designed to measure
divergent thinking.15 Research also indicates most
measures of creative thinking are susceptible to a myriad of
variables related to testing conditions and procedures.16
Significantly, exactly what the precise effects of time
imposed constraints, low pressure psychological atmospheres
and verbal variations during instruction are not known due to
the presence of conflicting results in parallel studies.17
Uses and abuses of testing conditions were also addressed by
Torrance18 as a potential confounding variable influencing

the data.

 

15McCormack, Alan J., "Nonverbal Administration
Protocols for Figural Tasks of the Torrance Tests of Creative
Thinking," The Journal of Creative Behavior, 9, No. 2,
(1975), p. 89

 

16Manske, M.E. and Davis, G.A., "Effects of Simple
Instructional Biases Upon Performance in the Unusual Uses
Test", Journal of General Psychology, 79, (1968), p. 25.

 

17Van Mondfrans, A., et al., "The Effect of Instruc-

tions and Response Time on Divergent Thinking Test Scores,"
Psychology in the Schools, 8, (1971), p. 65-69.

 

18Torrance, E. Paul, "Creativity Testing in Education,"
Creative Child and Adult Quarterly, 1, No. 3, (1976), p. 136—
148.

40.

Sorting19

Sorting activities were not initially composed to
investigate their possible relationship to creative thinking.
The objective of sorting was to permit lower and middle
elementary school children to practice classificatory skills
using a collection of unrelated objects for which the student
established classes or groups based upon whatever criterion
he chose to select. As the activity was trial taught in the
classroom, it became clear that while a majority of students
would classify on the basis of more obvious attributes such
as color and shape, some would create more original
categories related to function (items one could build or
construct with) or form (items that need to be processed to
be used versus items available as they naturally occur).

Additionally, while some students encountered difficulty
in classifying all of their items, others were able by
creating new or additional criteria, to classify all of their
materials. By amending their initial criterion, some
students succeeded in expanding or changing the actual number
of groups in a given array of materials. It was following
several observations of students engaged in sorting that the

question of possible relationships to some of the parameters

 

19See Appendix A

41.

of creativity measurement in the Torrance Battery were
considered. Specifically, would sorting experiences
facilitate performance in the areas of fluency (number of
figures completed), flexibility (number of different
categories), originality (unusual or infrequent categories),
and elaboration (exposition and use of detail) as presented
in the Torrance Incomplete Figures Activity?

Drawing relationships among people, or things and events,
internalizing ideas and images, and forming patterns are all
essential components of thinking. Such basic mental
processes as observing, making inferences, problem solving,
and classificatory skills are all necessary elements in need
of coordination to facilitate thinking. Of these, a major
vehicle for the development of thinking in young children
involves concrete experiences with classification
activities.20

With roots in the sensori—motor period of development,
classification is a process that continues to be refined
through the period of formal operations.21 It is both

sequential and hierarchial. As more sophisticated

 

20Isenberg, Joan P. and Jacobs, Judith E., "Classifica—
tion: Something to Think About," Childhood Education, May/
June 1981, p. 284.

 

21Jean Piaget and B. Inhelder, The Psychology of the
Child, (New York: Basic Books, 1969), p. 103.

 

42.

classification tasks are presented to children, the more they
become involved in higher level thinking.22
Sorting, then is basically the ability to classify or

group a collection of chosen objects according to either a

specified rule or pre-established criteria.23 Discounting
valuables and large unwieldly items, virtually any objects
can be utilized for the sorting array. Metal, wood, and
plastic materials; items chosen for attributes of texture;
plain or multicolored materials; household objects and
workshop scraps have all been used. For specific purposes,
the assortment can be narrowed to suit desired learning
objectives. For instance, primary children often confuse
shapes and sizes. By providing an array of shapes, but with
two or three sizes, the sorting experience can assist with
skill development in this area. Sorting experiences designed
to improve classificatory skills include a variety of
materials as previously indicated.24 Materials that are
essentially similar (such as an assortment of leaves or
shells) should be avoided as they tend to restrict imagina—

tive responses.

 

22Ibid.

H

23Shattuck, J. Bruce, "Sorting Activities, Science and

Children, 6, No. 6, (1969), p. 14—16.

 

 

2“Ibid, p. 14.

43.

Important procedural elements needed to be observed
during sorting activities. Initially the students were asked
to empty the contents (See Appendix A) of their sorting boxes
on their desks, tables, or on the floor. At this point
students were given several minutes to freely explore the
array and become thoroughly acquainted with the objects.
After their initial exploration, guided instructions were
administered to begin the process of sorting itself.25

The process began when students were verbally encouraged
to arrange objects in combinations or groups. The students
were encouraged to create any number or type of combinations
provided each item in a group had something in common with
every other object. In situations where objects had poorly
defined or unknown attributes, students were asked to start a
pile of "not sure" articles. Decisions relative to this
group could be delayed until later in the sorting process,
with all objects ultimately sorted. At designated times
during the activity, students were asked to find ways to
increase or decrease the number of groups, or establish new
ones by altering the criterion for inclusion of an object
into a group.26

InstruEtor assistance during the sorting activities were

 

251bid, p. 15.

261bid.

44.

designed to be clarifying27

and to avoid judgmental or
evaluative traits. The purpose of any dialogue between the
teacher and student was to raise questions in the mind of the

student relative to the rationale for the criterion employed

for grouping objects. While the key ingredient of the experi—
mental intervention was the sorting activity itself, care was
taken to avoid placing adult values upon students' selections.
The "correctness" of choice was a valid approximation of the
student's perceptions of an object, therefore the relative
importance of correct or incorrect grouping was displaced by
the selection of a criterion to be employed. The use of clari-
fiers as a method of encouraging divergent thinking has not
been statistically verified, however classificatory oriented
activities appear well suited to its application.28

Groups I and III received the sorting activity during two
thirty minute sessions. All sorting activities were adminis—
tered during week two of the investigation.

During the third and final week of the investigation, all

subjects were administered parallel Form B of the Torrance Test

 

of Creative Thinking, Incomplete Figyres Activity. Methodology

 

of administration was identical to those employed during the

pretest utilizing Form A of the same instrument.

 

27Merrill Harmin, et. al., Values and Teaching,
(Columbus, Ohio: Charles Merrill Pub. Co., 1966), p. 51-82.

 

28Shattuck, J. Bruce, "Values Clarification from

Science Instruction," Science and Children, 8, No. 4, (1971),
p. 16-180

 

45.

CHAPTER IV
PRESENTATION OF THE DATA

The Design

 

To investigate if a relationship existed between sorting
activities and the Torrance Incomplete Figures Task, 3 Solomon
four-group design was employed. In this four—group form, there
are included two control (Groups II and IV) and two experimental
(Groups I and III) groups. Only one of each of the two groups
is pretested. All groups are posttested at the conclusion of
the study. Inclusion of the groups that are not pretested en-
ables one to determine both the main effects of pretesting and
the interaction of pretesting with the experimental variable.
The solomon four—group design makes a strong effort to disen-
tangle the possible effects of experimental treatment, the
pretest contemporaneous events, and interactions of pretest and
treatment. The presentation of the groups for analysis is
illustrated in Table 1.

Table 1. Analysis Groups

G__q_2 1 11 111 1.!
Pretest X X

Treatment X X

Postest X X X X

Homogeneity of Variance

 

Initial analysis of the data was to observe if randomiza—
tion of subjects resulted in an observed homogeneity of vari-

ances at the pretest level for Groups I and II.

46.

F max values were consistently low (fluency 1.05; flexi-
bility 1.11; originality 1.12; elaboration 1.25), suggesting
that homogeneity of variances existed. Probabilities were
consistently greater than .05 for each of the four measure-
ments. This allowed the variances to be pooled and utilize a
pooled standard error of estimate in an independent t-test of
significant difference. One can observe from Tables 2.1 -
2.4 that the revealed t values at the pretest for Groups I
and II were not statistically significant. This analysis
supported the assumption that the random assignment to Groups
I and II resulted in no statistically significant
differences. Accordingly, it was additionally assumed that
Groups III and IV were equivalent although no pretest data

existed.

Pre and Posttest Data

 

Referring to Tables 2.1 — 2.4, one can observe pre and
posttest means for both Groups I and II. Related T scores
indicated a statistically significant (p<.05) correlation for
Group I on measures of fluency, flexibility and originality.
For elaboration, related T-scores revealed no statistical
significance. On pre and post mean scores for Group 11,
there was a statistically significant correlation (p<.05) on

all four measures.

47.

Table 2.1 Summary Statistics, Fluency

 

 

 

 

Fluency Pretest
n x SD.
Group I 23 6.13 2.49
Group II 23 6.87 2.44
t-test* 1.32 (p>.05)
n x S.D
Group III 23
Group IV 23
t—test*

* independent t—test

** related t-test

M
x S.D. t-test
9.78 0.67 6.86 (p<.05)
9.22 1.17 3.63 (p<.05)
—2.01 (p<0.5)
x S.D. t—test**

8.43 1.99

7.04 2.25

2.22 (p<.05)

48.

Table 2.2 Summary Statistics, Flexibility

 

 

 

 

 

Flexibility Pretest Posttest
n x SD. x S.D. t-test
Group I 23 4.87 1.91 7.65 1.07 6.14 (p<.05)
Group II 23 5.26 1.81 8.35 1.26 7.09 (p<.05)
t—test* 0.71 (p>.05) -2.01 (p<.05)
n x S.D. x S.D. t—test**
Group III 23 7.34 1.72
Group IV 23 5.96 2.03
t-test* 2.50 (p<.05)

* independent t—test

** related t-test

Originality

 

Group I
Group II

t-test*

Group III

Group IV

t-test*

* independent t-test

49.

Table 2.3 Summary Statistics, Originality

 

7.96
6.96

392112.51

S.D. t-test

2.65 3.47 (p<.05)

2.53 2.54 (p<.05)

1.31 (p>.05)

 

 

Pretest

n x SD.

23 5.43 2.42

23 5.00 2.30

.62 (p>.05)

n x S.D
23
23

“* related t—test

x S.D. t-test**
6.83 2.53

5.30 2.38

2.10 (p<.05)

50.

Table 2.4 Summary Statistics, Elaboration

 

 

 

Posttest
x S.D. t-test
16.30 4.41 1.68 (p>.05)
21.30 7.29 2.71 (p<.05)

-2.81 (p<.05)

 

 

Elaboration Pretest
n x SD.
Group I 23 13.17 6.65
Group II 23 16.56 5.94
t-test* -1.82 (p>.05)
n x S.D
Group III
Group IV
t—test*

* independent t—test

** related t-test

x S.D. t-test**
15.74 7.14
12.43 4.60

1.87 (p>.05)

51.

One-Way Analysis of Variance

One-way analysis of variance was then employed to estab—

lish if a significant
test means for Groups

calculations for each

Source

 

Between Gps.

Within Gps.

Total

Source

 

Between Gps.

Within Gps.

Total

88

91

88

91

difference (p<.05) existed on the post-
I-IV. Tables 3.1 - 3.4 reveal ANOVA

of the four posttest measurements.

Table 3.1 ANOVA

Fluency
Sum 89. X Sq. F Ratio F. Prob.
97.2498 32.4166 11.964 0.0000
238.4344 2.7095
335.6841

Table 3.2 ANOVA

Flexibility

Sum 89. X Sq. F Ratio F. Prob.
69.6090 23.2030 9.427 0.0000
216.6083 2.4615

285.2173

52.

Table 3.3 ANOVA

Originality

 

 

Source ELF; Sum 89. X 59. F Ratio F. Prob.
Between Gps. 3 82.6520 27.5507 4.313 0.0069
Within Gps. 88 562.0865 6.3873

Total 91 644.7385

Table 3.4 ANOVA

Elaboration

 

Source D.F. Sum Sg. X 89. F Ratio F. Prob.
Between Gps. 3 924.8981 308.2993 8.511 0.0001
Within Gps. 88 3187.8206 36.2252

Total 91 4112.7148

As indicated by Tables 3.1 — 3.4, F max probability values

were consistently statistically significant (p<.05) at the
posttest for each of the four measures. While the one-way
analysis of variance revealed a significant difference between
the means did exist, it did not reveal which means were
significant and for what level of measurement. Matrices
utilizing separate T values and their probabilities for each
measurement and by groups were then profiled to determine

which means were significant. See Tables 4.1 - 4.4.

53.

Table 4.1

T-Test Values and Probabilities (p<.05) Matrix

 

Fluency
_l;__ _ll__ _lll’ _£1_ IJIII/II,IV
I 2.01 3.07 5.60
11 (.05 -1.62 4.12
III >.05 (.05 2.22
IV (.05 (.05 (.05
I,III/II,IV 2.85

As one may observe from Table 4.1, the following posttest
means were statistically significant (p<.05) for fluency:
Groups I and II; I and III; I and IV; II and IV; III and IV;
I/III and II/IV. Contrasts between Groups II and III were not

statistically significant.

54.

Table 4.2

T-Test Values and Probabilities (p<.05) Matrix

 

Flexibility
_l__ _II_ _III _ll_ I,III/II,IV
I -2.01 .72 3.54
II <.05 -2.24 4.79
III >.05 <.05 2.50
IV <.05 <.05 <.05
I,III/II,IV 1.06

>.05

As one may observe from Table 4.2, the following posttest
means were statistically significant (p<.05) for flexibility:
Groups I and II; I and IV; II and III; II and IV; 111 and IV.
Contrasts between Groups I and III and I/III and II/IV were

not sitnificant.

55.

Table 4.3

T-Test Values and Probabilities (p<.05) Matrix

 

Originality
._l__ _ll_ _lll_ _IX_D IJIII/IIJIV
I 1.31 1.48 3.57
11 >.05 —O.l7 2.28
III >.05 >.05 2.10
IV <.05 <.05 <.05
I,III/II,IV 2.40

<.05

Observing Table 4.3, one finds the following posttest
mears statistically significant (p<.05) for originality:
Groups I and IV; II and IV; III and IV; and I/III and Il/IV.
Contrasts between Groups I and II, I and III and II and III

were not statistically significant.

56.

Table 4.4
T-Test Values and Probabilities (p<.05) Matrix

Elaboration

 

_l__ _;;__ _;;; I_;y_ I,III/II,IV
I -2.81 0.32 2.91
11 <.05 -2.61 4.93
III >.05 <.05 1.86
1v <.05 <.05 >.05
I,III/II,IV —0.68
>.05

Observing Table 4.4, one can observe the following post-
test means as statistically significant (p<.05) for elabora-
tion: Groups I and II; I and IV; II and III; II and IV.
Contrasts with I and II; III and IV; I/III and II/IV were not

statistically significant.

Analysis of Covariance

 

The one-way analysis of variance compared all four groups
for significance at the posttest. Use of the T-test values
and probabilities matrix presented which means were
significant and for what groups. At this juncture the
relative impact of group assignment will be investigated.

Pretest influence has not been taken into consideration as a

57.

covariate upon the posttest means. The analysis of covariance
permits the factoring out of the covariate to observe if once
the covariate is eliminated any significance remains between
the posttest means for Groups I and II. Table 5.1 summarizes
the results of an ANCOVA of fluency.

Table 5.1

ANCOVA - Fluency

 

 

Source of Sum of Mean Signif.
Variation Sguares _D§_ Sguare F of F.
Preflu 2.970 1 2.970 3.385 0.073
Group 2.805 1 2.805 3.198 0.081
Explained 5.775 2 2.888 3.291 0.047
Residual 37.725 43 0.877

Total 43.500 45 0.967

One can observe from Table 5.1 that 13.2 percent of the
total variance in the posttest scores is explained by the pre—
test scores and the sorting activity.* The proportion of
explained variance is statistically significant at the .05
level of significance. Slightly more than half of the
explained variance (51.4%) is accounted for by the

covariate.** The proportion of variance in the total variance

 

Explained - Total

**Pref1u - Explained

58.

explained by either the pretest or the sorting activity is not
statistically significant at the .05 level of significance.
Their respective F values were 3.39 and 3.20. Figure 1.1

shows the adjusted and unadjusted fluency means for Groups I

 

 

 

and II.
Figure 1.1
Adjusted and Unadjusted
Fluency Means
Groups I and II
10_ ’ Experimental
9_ Control
8..
7—
6—
5—
l 1
Pretest Posttest
(X) = Adjusted
X = Unadjusted

59.

Table 5.2 summarizes the results of an ANCOVA of flexi-

 

 

bility.
Table 5.2
ANCOVA - Flexibility

Source of Sum of Mean Signif.
Variation Squares _2£_ Square F of F.
Preflex 0.646 1 0.646 0.462 0.500
Group 5.228 1 5.228 3.739 0.060
Explained 5.874 2 2.937 2.100 0.135
Residual 60.126 43 1.398

Total 66.000 45 1.467

One can observe from the Table that 8.9 percent of the
total variance in the posttest scores are explained by the
pretest scores and the sorting activity. The proportion of
explained variance is not statistically significant at the .05
level of significance. The covariate accounted for 10.9 per-
cent of the explained variance. The proportion of variance in
the total variance explained by either the pretest or the
sorting activity is not statistically significant at the .05
level of significance. Their respective F values were 0.46
and 3.74. Figure 1.2 shows the adjusted and unadjusted

flexibility means for Groups I and II.

 

60.

Figure 1.2
Adjusted and Unadjusted
Flexibility Means

Groups I and 11

Control

Experimental

 

//;}
//////////
Y/
y'
1 l
Pretest Posttest
(X) = Adjusted
X = Unadjusted

61.

Table 5.3 summarizes the results of an ANCOVA of original—

 

 

ity.
Table 5.3
ANCOVA - Originality

Source of Sum of Mean
Variation Squares _DE_ Square F
Preorig. 0.048 1 0.048 0.007
Group 11.462 1 11.461 1.666
Explained 11.510 2 5.755 0.836
Residual 295.902 43 ‘ 6.881

Total 307.412 45 6.831

One can observe from the Table that 3.7 percent of

total variance in the posttest scores are explained by

Signif.

_0_f__E_-_

0.934

0.204

0.440

the

the

pretest scores and the sorting activity. The proportion of

explained variance is not statistically significant at

the .05

level of significance. Less than one percent of the explained

variance is accounted for by the covariate. The proportion of

variance in the total variance explained by either the
or the sorting is not statistically significant at the
level of significance. Their respective F values were
1.67. Figure 1.3 shows the adjusted and unadjusted

originality means for Groups I and II.

pretest

.05

.01 and

 

62.

Figure 1.3
Adjusted and Unadjusted
Originality Means

Groups I and II

Experimental
0

Control

/®

 

I |

Pretest Posttest
(X) Adjusted

><
ll

Unadjusted

63.

Table 5.4 summarizes the results of an ANCOVA of

elaboration.

Table 5.4

ANCOVA - Elaboration

 

 

Source of Sum of Mean Signif.
Variation Squares '_QE_ Square F of F.
Preelab. 24.761 1 24.761 0.666 0.419
Group 262.987 1 262.987 7.070 0.011
Explained 287.748 2 143.874 3.868 0.029
Residual 1599.488 43‘ 37.197

Total 1887.236 45 41.939

One can observe from the Table that 15.2 percent of the
total variance in the posttest scores are explained by the
pretest scores and the sorting activity. The proportion of
explained variance is statistically significant at the .05
level of significance. Of the explained variance, 8.6 percent
is accounted for by the covariate. The proportion of variance
in the total variance explained by either the pretest or the
sorting activity is statistically significant at the .05 level
of significance. Their respective F values were .67 and 7.07.
Figure 1.4 shows the adjusted and unadjusted elaboration means

for Groups I and II.

22

21

20_

19_

184

17

164

14

13_

 

64.

Figure 1.4
Adjusted and Unadjusted
Elaboration Means
Groups I and 11

Control

Experimental

 

//
K’
L l
Pretest Posttest
(X) = Adjusted
X = Unadjusted

65.

Two-WayyAnalysis of Variance

 

To statistically evaluate the significance of group assign—
ment and the covariate with the independent variable, a two-
way analysis of variance was employed. As an extension of
simple analysis of variances, the data was organized into a
double entry table, with one of the independent variables
placed into columns and the other into rows. Figure 2.1

illustrates the procedure.

Figure 2.1

Two-Way ANOVA

Group
1 11 II 11:
Pre
Yes N=23 N=23
No N=23 N=23
I I
I I
I I
Exp A Exp B I !
I I
Pre I I
I
Yes N=23 N=23 I '
| I
|
I

66.

Table 6.1 summarizes the results of a two-way analysis of

variance of fluency.

Source of Sum of
Variation Squares
Main Effects 93.326
N Group 22.011
N Preflu 71.315
Two-Way Inter 3.924
Explained 97.250
Residual 238.433
Total 335.683

Table 6.1

Two-Way ANOVA

Fluency

DF

88

91

Mean

M

46.663

22.

71.

32.

One can observe from the Table that

total explained variance was explained.

011

315

.924

417

.709

.689

 

17.222
8.124
26.321
1.448

11.964

28.9 percent

Signif.

of F.

(L000
(L005

(L000

(L232

(L000

of the

Of the explained

variance, 95.9 percent was accounted for by the main effect.

Of the main effect, 14.4 percent was accounted for by the

group assignment and 76.4 percent by the pretest.

All of the

aforementioned ratios were statistically significant at the

67.

.05 level of significance. F values for the group and pretest
were 8.12 and 26.32 respectively. Interaction, or the con—
comitant effect of the two independent variables was observed
at 4.03 percent and was not statistically at the 0.5 level of
significance. Adjusted fluency means are entered in the

appropriate cells in Figure 2.2.

Figure 2.2
Two—Way ANOVA

Fluency Adjusted Means

Treatment

Yes .No

u
\o
'6.
o
I

Yes x
Pretest

 

No

><

ll

\1

L,

;.\
I-

 

x = 9.11 x
I |

8.13

Observations for fluency means revealed subjects who
received the pretest scored higher on posttest means regard—
less of treatment (9.50 to 7.74). Subjects who were
administered the treatment scored higher on posttest means
than those not administered the treatment (9.11 to 8.13).

Table 6.2 summarizes the results of a two-way analysis of

variance of flexibility.

68.

Table 6.2

Two-Way ANOVA

 

Flexibility
Source of Sum of Mean Signif.
Variation Squares _I§;_ Squggg F of F.
Main Effects 44.565 2 22.283 9.053 0.000
N Group 2.783 1 2.783 1.130 0.291
N Preflu 41.783 1 41.783 16.975 0.000
Two-Way Inter 25.043 1 25.043 10.174 0.002
Explained 69.609 3 23.203 9.427 0.000
Residual 216.607 88 2.461
Total 286.216 91 3.145

One can see from the Table that 24.0 percent of the total
variance was explained. 0f the explained variance, 64.3
percent was accounted for by the main effect. Both of the
aforementioned ratios were statistically significant at the
.05 level of significance. Of the main effect, 6.2 percent
was accounted for by the group and was not statistically
significant. The pretest, however, accounted for 93.5 percent
of the main effect and was statistically significant at the

.05 level of significance. F values for group and pretest

69.

were 1.13 and 16.97 respectively. Interaction was observed as
accounting for a statistically significant 35.9 percent of the
explained variance. Adjusted flexibility means are entered in

the appropriate cells in Figure 2.3.

Figure 2.3
Two-Way ANOVA

Flexibility Adjusted Means

Treatment
Yes No

Yes x - 7.90 -—
Pretest

 

No
x - 6.66 -

 

x a 7.50 x a 7.16
l I

Observations for flexibility means revealed subjects who
were administered the pretest scored higher in posttest means
regardless of treatment (7.90 to 6.66). Subjects administered
the treatment scored higher on posttest measures than those
not receiving the treatment (7.50 to 7.16).

Table 6.3 summarizes the results of a two-way analysis of

variance of originality.

70.

Table 6.3

Two-Way ANOVA

Originality
Source of I Sum of Mean Signif.
Variation Squares _£E;_ .§922£2. ____F__ of F.
Main Effects 81.087 2 40.543 6.348 0.003
N Group 36.565 1 36.565 5.725 0.019
N Preorig 44.522 1 44.522 6.970 0.010
Two—Way Inter 1.565 '1 1.565 0.245 0.622
Explained 82.652 3 27.551 4.313 0.007
Residual 562.079 88 6.387
Total 644.732 91 7.085

One can observe from the Table that 12.8 percent of the
total variance was explained. This ratio was statistically
significant at the .05 level of significance. 0f the
explained variance, 98.1 percent was accounted for by the main
effect. 0f the main effect, 45.1 and 54.9 percent were
accounted for by the groups and the pretest, respectively.
Both of these ratios were statistically significant at the .05
level of significance. F values were observed at 5.72 and

6.97 respectively for group and pretest. There was no

71.

significant interaction observed. Adjusted means are entered

in the appropriate cells in Figure 2.4

Figure 2.4
Two~Way ANOVA

Originality Adjusted Means

Treatment

Yes No

.-<
(D
(I)
X
II

7.46 —1
Pretest §

No i
6.06 —J

><
ll

 

x = 7.39 x = 6.13
I I

 

Observations for originality were similar to fluency and
flexibility. Subjects administered the pretest scored higher
on posttest means regardless of treatment (7.46 to 6.06).
Subjects administered the treatment scored higher on posttest
measures than those not receiving the treatment (7.39 to
6.13).

Table 6.4 summarizes the results of a two—way analysis of

elaboration.

72.

Table 6.4

Two-Way ANOVA

 

Elaboration
Source of Sum of Mean Signif.
Variation Squares _l§:_ .§322£2. F of F.
Main Effects 528.369 2 264.185 7.293 0.001
N Group 16.533 1 16.533 0.456 0.501
N Preelab 511.837 1 511.837 14.129 0.000
Two-way Inter 396.533 1 396.533 10.946 0.001
Explained 924.902 3_ 308.301 8.511 0.000
Residual 3187.817 88 , 36.225
Total 4112.719 91 45.195

One can observe from the Table that 22.5 percent of the
total variance was explained and was statistically significant
at the .05 level of significance. The main effect accounted
for 57.1 percent of the explained variance and was also
statistically significant. Of the main effect, 1.7 percent
was accounted for by the group and 55.3 percent by the pre-
test. The former was not statistically significant; however
the pretest was statistically significant at the .05 level of

significance. P values were observed at 0.456 and 14.13,

73.

respectively. There was a significant interaction of 42.8

percent. Adjusted means are entered in the appropriate cells

in Figure 2.5.

Figure 2.5
Two—Way ANOVA

Elaboration Adjusted Means

Treatment
Yes No

Yes I ' x a 18.81 ——-
Pretest

 

No
x = 14.09 -—

 

x = 16.03 x a 16.87
I |
Observations for elaboration means indicated subjects who
received the pretest again scored higher in posttest means
regardless of treatment (18.81 to 14.09). For elaboration,
subjects who were not administered the treatment scored
slightly higher (16.87 to 16.03) than subjects who were

administered the treatment.

74.

CHAPTER V

SUMMARY, CONCLUSIONS AND RECOMMENDATIONS

The purpose of this chapter is to present a restatement
of the problem and present a synopsis of the research
procedures and findings. In addition, the conclusions

regarding the hypotheses are summarily discussed.

Summary

The purpose of the study was to investigate if a
relationship existed between sorting activities and creative
thinking. Specifically, would participation in the sorting
task impact upon the dependent variable, the Torrance Test of
Creative Thinking: Incomplete Figures Activity Form B.

By 1988, the Creative Education Foundation projects some
20,000 items listed as creativity research. Six basic
categories of investigation have been included:

1. Differences between high and low creative
individuals regarding personality structures, cognitive
functioning, and biographical data.

2. Research comparing individual creativity with that
of selective groups.

3. Investigations designed to identify the role of
educators, psychologists, and others in a helping

relationship to better understand the dynamics of creativity.

75.

4. Studies of factors that appear to inhibit creative
thinking.

5. Analysis of relationships among creativity,
intelligences, and some measurement index of achievement.

6. Studies of environmental variables affecting
creativity, such as learning and home environmental factors,
educational background, and sociological patterns.

This investigation directly involved a classroom teaching
device and its relationship to creative thinking. The
majority of the aforementioned categories, while of value to
the field of creativity research in general, were not directly
related to the researchable problem at hand. Appropriate
criteria were established for inclusion of related
investigations in order to gain greater relevancy and a sense
of overall connectedness to the study. Inclusion of
bibliographical data needed to meet the following criteria:

1. Studies that were directly related to school popula-
tions or possessed intrinsic transfer value when considering
school populations.

2. Studies that focused upon specific variables that
buttressed an understanding for creativity experimentation in
the schools.

3. Investigations that expressed a theoretical framework
significant to the research problem of this investigation.

4. Research representative of other parallel studies

useful for illustrative purposes.

76.

Examination of literature that met the above criteria
revealed a dearth of appropriate investigations. While
sufficient numbers of classroom oriented investigations were
observed, these generally were not of an experimental nature
or did not sufficiently isolate parameters that could
facilitate empirical evaluation. Classificatory modes have
long been considered a contributory attribute of science
education, however their role as a vehicle for creative
stimulus appears to rest in a neophytic milieu and were
conspicuously absent.

The largest component of classroom investigations
involved comparisons of classroom structure, such as "open"
versus "traditional" approaches. Findings in this area
appeared to be more related to structures of the classroom
than structures of the activities designed to develop and
extend creative thinking.

It would appear more educationally sound from a research
perspective to address the classroom environment question from
other than a "traditional" versus "progressive" perspective.
Perhaps certain kinds of measurable creativity would appear to
be more appropriate compared to certain kinds of classroom
environments. This notion was buttressed by Ramez and Piper
(1974) in observing a strong relationship of open-classroom
figural activity and traditional classroom superiority in

verbal creativity. The work of Thomas and Buck (1981)

77.

indicated school environments to be an intensely complicated
area of investigation. Dimensions such as type of school
were the beginnings of other factors--these factors included
sex of the child, age, and the component of creativity
assessed.

The cumbersome approach to linking creative thinking
with environmental factors may account for the range of
findings encountered in the literature. Haddon and Lytton
(1968) found a positive relationship of creativity sources
for children in progressive schools. Research conducted by
Wilson, Stuckey and Langivin (1972) found open-class children
were less creative than those in a traditional class.
Bennett (1976) and Wright (1975) found little or no

correlation between creativity and school environments.

Conclusions

This study investigated if there is a relationship
between sorting activities and the Torrance Test of Creative
Thinking: Incomplete Figure Activity. This study was
conducted in the classroom utilizing ninety—two randomly
selected fifth grade students.

Since the limits of interpretation are established upon
choice of design, it was imperative to select a strong and
methodologically appropriate blueprint. The precision
afforded by pretesting was desirable but the potential for

interaction between pretest and treatment needed to be

78.

addressed. An after-only design would be free of the

interaction problem but interpretation of results would have
been limited. If for example groups receiving the treatment
received higher mean scores on the dependent variables all one
could infer is just that. The literature is spotted with
findings in which the investigators assume the reason for
differences on posttest results has been caused by
experimental manipulations.

The Solomon four-group design makes the strongest effort
to disentangle the possible effects of experimental treatment,
the pretest, contemporaneous events, and interactions of
pretest and treatment. The Solomon four-group design was
selected for this investigation after explicit consideration
was given to its regard for factors relative to external
validity.

Investigations that are classroom based should be con-
ducted in order to consider policy making decisions within the
school framework as regards curriculum. Use of the Solomon
four-group design appears infrequently in the literature,
however, its value in pinpointing the effects of the factors
under consideration merit its widespread consideration and
broader use. The methodological value of this investigation,
prior to any statistical analysis or significance of findings,
appear to be: (1) the study fulfills a need for more

classroom oriented research in the area of instructional

79.

strategies and (2) use of the Solomon four-square design with
the population selected will help to facilitate generaliz-
ability of findings to other school populations.

Hypotheses for this investigation were established as

follows:

H01: Subjects exposed to the sorting activity would
exhibit higher posttest means than subjects not
exposed to the sorting activity.

H02: Subjects exposed to the sorting activity would
score highest on measurements of originality on
the posttest instrument than subjects not exposed
to the sorting activity.

H03: Following originality, subjects exposed to the

sorting activity would then exhibit highest
posttest scores for measures of flexibility
followed by fluency and elaboration.

Initial analysis of the data revealed homogeneity of
variance for Groups I and II existed at the pretest level.
For each of the four measures, T-values were consistently
greater than the .05 level of significance. In the Solomon
four-group design, it is assumed that if the remaining
experimental and control groups had received the pretest
measurement, their scores would have been similar to the first
experimental and control groups, since all four groups had

been originally constituted by random assignment. Homogeneity

80.

of variance was therefore assumed for all four groups.

Observations of summary statistics for Groups I and II
yielded the following:

1. For Group I pre and posttest contrasts exhibited an
increase in posttest means for each of the four measures.
These changes were statistically significant for fluency,
flexibility, and originality. No statistical significance was
observed for elaboration.

2. For Group II pre and posttest contrasts also
exhibited an increase in posttest means for each of the four
measures. These changes were statistically significant for
all four measures.

The increase in posttest mean for Group I was desirable.
However, the accompanying increase in posttest mean for the
control group minimized identification of possible causal
relationships. Broadly, there was a change and this change
was largely significant within and between the two groups.

Observations of summary statistics for Groups III and IV
yielded the following:

1. Group III, as the second experimental group, attained
higher posttest scores for each of the four measurements than
did Group IV, the second control group.

2. Exclusive of the measurement for elaboration which
was not statistically significant, the measurements for
fluency, flexibility, and originality were statistically

significant.

81.

The first hypothesis, that subjects exposed to the
sorting activity would attain higher posttest means than
subjects not exposed to sorting was partially supported for
Groups I and II. The data supported the first hypothesis with
reference to Groups III and IV.

The second hypothesis, that subjects receiving the treat—
ment would score highest on the measurement for originality,
was not supported by the data for any of the four groups. It
should be noted, however, that while attained scores did not
reflect support for the second hypothesis, subjects receiving
the sorting activity did score higher on the posttest measure
of originality than subjects not receiving the intervention.
For posttest results of Groups III and IV, this was
statistically significant.

The third hypothesis, that following originality,
subjects exposed to the sorting activity would then exhibit
highest posttest means for measures of flexibility followed by
fluency and elaboration, was not supported by the data. It
should be noted that the third hypothesis was largely
dependent upon relative support of the data for the second
hypothesis. It therefore lacked any degree of interpretive
integrity on its own merit. It should also be noted that the
intended future replication of this investigation will not
include attempts to hypothesize hierarchial results involving

four levels of measurement at the posttest.

82.

Analysis of covariance calculations investigated the
relative impact of group assignment. Posttest means for
Groups I and II had been previously observed as statistically
significant both within and between groups. ANCOVA of
fluency, flexibility, originality, and elaboration revealed no
significance once the covariate was factored out. Fluency
calculations revealed the strongest effort, however, with
slightly more than half of the explained variance accounted
for by the covariate. The reader is reminded that the measure-
ment for fluency involves the total number of pictures that
are completed by the subject at both the pre and post levels.

Perhaps the most revealing observations occurred as a
result of the two-way analysis of variance. This function was
performed in order to statistically evaluate the significance
of group assignment and the covariate with the independent
variable. As no pretest data was available for Groups III and
IV as per the Solomon four—square design, the data was organ-
ized into a double entry table, with one of the independent
variables placed into columns and the other into rows. In
this fashion, Groups I and III actually became New Group I and
Groups II and IV New Group II. Main effects of group assign-
ment and pretest were statistically significant for fluency,
with the pretest accounting for a large proportion of (76.4
percent) of the main effect. For flexibility, in excess of

ninety percent of the main effect was accounted for by the

83.

pretest. For originality, 98.1 percent of the variance was
explained by the main effect, with over half of this ratio
accounted for by the pretest. For elaboration, observations
for main effect yielded in excess of 55.3 percent accOunted
for by the pretest.

Adjusted means for the two—way ANOVA fluency, flexibility
and originality yielded two significant findings:

1. Subjects who received the pretest scored higher in
posttest means regardless of the presence or absence of the
treatment.

2. Subjects who were administered the treatment scored
higher on posttest means than those subjects not administered
the treatment.

Adjusted means for the two-way ANOVA of elaboration
yielded the following findings:

1. As with fluency, flexibility and originality,
subjects who received the pretest again scored higher on
posttest means regardless of treatment.

2. Subjects who were not administered the treatment
scored slightly higher than subjects who were administered the
treatment.

The prevalent impact upon the dependent variable appears
to have been the result of the synthesizing of subjects by the
pretest. As interpreted by the finding, this impact was

significant. Sorting activities appear to have had a

84.

secondary influence although its impact was an identifiable
one. Posttest scores from Groups III and IV lend credence to
the role of sorting in creative thinking. This notion was
further supported through observations of the two-way ANCOVA.
Sorting encourages active participation to problem solve
and categorize with dependence upon the students criteria for
grouping. This in itself is basic to original thinking and
continuous appraisal of the available alternatives. Correct
answers are not the students goal, but a continual refinement
of thinking leading to intrinsically motivated rewards. Dur-
ing sorting, the students are encouraged to shift criteria in
order to increase or decrease the number of groups that can be
formed by their sorting array, which is related to the flexi-
bility and fluency component of the instrument. Elaboration
appears to be the least related to the sorting activity on the
basis of the findings of this investigation. The cause for
this may well be that elaboration scores reflect the quantity
of detail added to the original idea. This process may be
more dependent upon artistic fluency and not be as strongly

related to original thinking.

Recommendations
The following represent suggested directions and design
approaches that may serve as a coagulant to research

approaches in creative thinking. They are most appropriate

85.

when applied to experimental studies investigating the
relationship of specific activities and creative testing.
Recommendations include the following:

1. More focus should be given to future classroom
studies in terms of what is done with the data once col-
lected. Treatment of the data needs to do more than indicate
change in the presence of what appear to be causal
relationships. Possein (1967) concluded that students trained
in the use of inductive procedures exhibited some
characteristics of problem solving more frequently than pupils
taught by the deductive method. However pre and post
instruments were identical and the question of test-retest
effects were not addressed.

2. Too many parameters are frequently investigated
simultaneously and create an anatomically cumbersome
investigation. Future contributions may be facilitated by the
identification of the most appropriate activities that are
found through statistical analysis to be intrinsically
valuable relative to creative development. By using the
activities in a systematic manner, their subsequent value can
be networked into the myriad of labels that have been bestowed.
upon the classroom. Investigations most often appear to
emanate from another direction, such as parental influences
(Thomas and Berk, 1981), attendance at a pre-school program
(Hahn, 1981), and preferences in teaching styles (Selkirk,

1979). While these approaches and choices of investigative

86.

channels to contribute to a profile of information, their
limitations in terms of classroom applications remain.

3. Subjects involved in this investigation were exposed
to two thirty minute sessions with sorting. The possible
combinations of the activity are endless. Students may well
be hesitant to attempt original responses for fear of
"incorrect" responses. Although the activity is non—judg-
mental in terms of evaluation, this is not made clear to the
students during the instructions. This leaves a single thirty
minute session in which the students become aware of the
opportunity to freely explore grouping alternatives. A mini-
unit could be developed in which students have the opportunity
to practice their decision making skills during several
sessions over a period of weeks. Randomly selected fifth
grade students should be involved to broaden the data base.
Selection of other age/grade populations would erode a desired
focus of research. A second motivation for utilizing fifth
grade populations involves the already statistically supported
notion of creativity slumps that periodically occur (Torrance,
1967; Williams, 1976) but have not been reported with fifth
grade students.

4. The results of this investigation suggest a signifi—
cant impact of the pretest upon posttest scores. Each time
subjects were administered the pretest they scored higher on

posttest means than subjects not administered the pretest

87.

regardless of treatment. The absence of a pretest eliminates
test-retest effect but sharply reduces the validity of the
study by disregarding homogeneity of the populations strength
of the randomization, and removal of crucial comparative
data. What may be present in creativity experimentation is
the tendency of creativity instruments in general to
disproportionately synthesize subjects via the pretest. If
inferences are based upon‘analysis of the data collected by
the instruments, then what may be needed is closer
examination of the hypothesis that it is the testing and not
the treatment that encompasses the main effect. Although
appropriate statistical analysis should normally address the
question, experimental designs need to be created that have
as their primary function the effects of the instrument
rather than the treatment. Certain types of research,
particularly classroom research, have a distinct tendency to
produce unequal variances. It could be possible that
educators have used classroom research in the same manner the
village inebriate utilizes the lamppost for support instead

of illumination.

88.

REFERENCES

Anderson, Harold H., ed., Creativity and Its Cultivation.
New York: Harper and Brothers, 1959.

Bennett, N. TeachingStyles and Pupil Progress. Cambridge:
Harvard University Press, Inc., 1976.

Campbell, Donald T. and Julian C. Stanley. Experimental and
Quasi-Experimental Designs for Research. Chicago: Rand
McNally College Publishing, 1963.

 

Champion, Dean J. Basic Statistics for Social Research.
New York: Macmillan Publishers, 1981..

Cole, H.P. and S. Bernstein. "Summary of Process-Promoting
Units Resulting from Curricular Analysis Activities."
Paper delivered at the Eastern Regional Institute for
Education, Syracuse, New York, 1969.

Crockenberg, Susan B. "Creativity Tests: A Boon or
Boondoggle for Education?" Review of Educational
Research, 2, No. l, (1972), 27-45.

 

Crutchfield, R.S. "Nuturing the Cognitive Skills of
Productive Thinking." in Rubin, L.J., Life Skills in
School and§ociety. (Washington, D.C.: Association for
Supervision and Curriculum Development, N.E.A.), 1969.

Csikszentmihaly, Mihaly. Beyond Boredom and Anxiety. San
Francisco: Jossey-Bass, 1975.

DeVito, Alfred. "Survival Through Creative Education."
Journal of Creative Behavior, 10, No. 1, (1976), 45-61.

Feldhasen, John F. and Donald J. Treffinger. "Design and
Evaluation of a Workshop on Creativity and Problem-
Solving for Teachers." Journal of Creative Behavior,
10, No. 1, (1976), 12-14.

Golub, Sharon and Karen Hahn. "Training Creative Thinking
in the Open and Traditional Classroom." Journal of
Creative Behavior, 17, No. 3 (1976), 217.

Gowan, J.C. and M. Olson. "The Society Which Maximizes
Creativity." Journal of Creative Behavior, 13, No. 2,

(1979), 194-210.

89.

Guilford, J.P. The Nature of Intelliggnce. New York:
McGraw Hill, I967.

 

Haddon, F.A. and H. Lytton. "Teaching Approach and the
Development of Divergent Thinking Abilities in Primary
Schools." Bgitish Journal of Educational Psychology,
38, (1968), 171-80.

Hahn, Beverly A. "Common Characteristics in the Educational
Backgrounds of High Creative Children and Their
Performances Regarding Classroom Behavior." Journal of
Creative Behavior, 15, No. 4, (1981), 283-85.

Harmin, Merrill. "N.E.X.T.E.P.: New Exploratory Teacher
Education Program." Edwardsville, Illinois: Southern
Illinois University. 1969.

. Values and Teaching. Columbus, Ohio: Merrill
Publishing Company, 1966.

 

Holman, E. Riley. "A Master of Education Degree in
Elementary Education With An Emphasis on Creative
Teaching-Learning." Creative Child and Adult Quarterly,
11, No. 2, (1977), 98-100.

Huck, Schuyler W., et al. Reading Statistics and Research.
New York: Harper and Row, 1974.

Isenberg, Joan P. and Judith E. Jacobs. "Classification:
Something to Think About." Childhood Education. May/
June, (1981), 284. .

Kerlinger, Fred N. Foundations of Behavioral Research. New
York: Holt, Rinehart and Winston, 1973.

Leedy, Paul D. Practical Research: Planningypnd Design,
Second Edition. New York: Macmillian Publishers, 1980.

Linder, Toni W. "Organizing a Course on Creativity: Theory
and Practice." Journal for the Education of the Gifted,
4, No. 3, (1981), 2II424.

Manske, M.E. and G.A. Davis. "Effects of Simple Instruc-
tional Biases Upon Performance in the Unusual Uses
Test." Journal of General Psychology, 79, (1968), 25.

McCormack, Alan J. "Nonverbal Administration Protocols for
Figural Tasks of the Torrance Tests of Creative
Thinking." The Journal of Creative Behavior, 9, No. 2,
(1975), 89.

90.

Miel, Alice, ed. Creativity in Teaching. Belmont,
California: Wadsworth Publishing, 1961.

Ogilvie, E. "Creativity and Curriculum Structure."
Educational Research, 16, (1974), 126-32.

Osborn, A.F. Applied Imagination. New York: Scribner's
and Sons, 1963.

 

 

Parnes, S.J. and A. Meadow. "Effects of Brainstorming
Instructions on Creative Problem Solving by Trained and
Untrained Subjects." Journal of Educational Psychology,
50, (1963), 87-91.

Penick, John E. "What Research Says: Encouraging
Creagivity." Science and Children, 20, No. 5, (1983),
32—3 .

 

Piaget, Jean and B. Inhelder. The Psychology of the Child.
New York: Basic Books, Inc., 1969.

 

Possien, Wilma Martens. "A Comparison of the Effects of
Three Teaching Methodologies on the Development of the
Problem Solving Skills of Sixth Grade Children."
Journal of Creative Behavior, 1, No. 2, (1967), 85-90.

Ramey, C.T. and V. Piper. "Creativity in Open and
Traditional Classrooms." Child Development, 45, (1974),
557-60.

 

Schubert, Daniel S.P. "Boredom As An Antagonist of

Creativity." The Journal of Creative Behavior, 11, No.
4, (1974), 56.

 

Selltiz, Claire, et. a1. Research Methods in Social
Relations. New York: Holt, Rinehart and Winston, 1976.

 

 

Sharp, Lawrence W. "The Effects of a Creative Thinking
Skills Program on Intermediate Grade Educationally

Handicapped Children." Journal of Creative Behavior, 7,
No. 5, (1973), 37.

 

Shattuck, J. Bruce. "Sorting Activities." Science and
Children, 6, No. 6, (1969), 14-16.

 

"Values Clarification From Science

 

Instruction." Science and Children, 8, No. 4, (1971),
16-18.

91.

Taylor, Calvin W. Creativity: Progress and Potential. New
York: McGraw-Hill, 1964.

Thomas, Nancy C. and Laura E. Berk. "Effects of School
Environments on the Development of Young Children's
Creativity." Child Development, 52, (1981), 1153-62.

 

Torrance, E. Paul. "Creativity Testing in Education."
Creative Child and Adult Quarterly, 1, No. 3, (1976),
136-38.

. Guiding Creative Talent. Englewood Cliffs:
Prentice Hall, 1962.

 

 

"Predicting the Creativity of Elementary School
Children - and the Teachers Who Made a Difference." The

 

 

Gifted Child Quarterly, 25, No. 2, (1981), 55-61.

 

. Torrance Tests of Creative Thinking,
Bensenville, Illinois, 1974.

 

 

. Torrance Tests of Creative Thinking: Directions
Manual and ScoringyGuide. Bensenville, Illinois:
Scholastic Testing Service, 1974.

 

 

 

. Torrance Tests of Creative Thinking: Norms-
Technical Manual. Bensenville, Illinois: Scholastic
Testing Service, 1974.

 

 

 

Understanding the Fourth Grade Slump in
Creative Thinkipg. U.S., U.S. Office of Education Final

 

 

 

Report, No. . Washington, D.C.: U.S.0.E., December,
1967.

Traxler, Arthur E. "Some Comments on Educational Research
at Midcentury." Journal of Educational Research, XLVII,

 

(1954), 359-66.

Treffinger, D.J., J.S. Renzulli, and J. F. Feldhusen.
"Problems in the Assessment of Creative Thinking."
Journal of Creative Behavior, 5, (1971), 104-112.

 

Triola, Mario F. Elementary Statistics. Menlo Park,
California: Benjamin Cummings Publishers, 1980.

 

Van Mondfrans, A., et. al. "The Effect of Instructions and
Response Time on Divergent Thinking Test Scores."
Psychology in the Schools, 8, (1971), 65-69.

92.

Williams, F.E. Classroom Ideas for EncouragingyThinking‘and
Feeling. Buffalo: D.0.K. Publishers, 1970.

"Rediscovering the Fourth Grade Slump in a Study
of Children's Self-Concept." Journal of Creative
Behavior, 10, No. 1, (1976), 15-34.

 

Wilson, R.S., et. a1. "Are Pupils in the Open Plan School
Different?" Journal of Educational Research, 66,
(1972), 115-18.

Wright, R.J. "The Affective and Cognitive Consequences of
an Open Education Elementary School." American
Education Research Journal, 12, (1975), 449-68.

 

 

Yaird, Leonard L. "The Torrance Tests of Creative
Thinking." In the Seventh Mental Measurements Yearbook,
Oscar Krisen Buros, Ed. Highland Park, New Jersey:
Gryphon Press, 1972. p. 837.

Yamamoto, K. "Validation of Tests of Creative Thinking: A
Review of Some Studies." Exceptional Children, 31,
(1965), 281-290.

 

Yawkey, Thomas D. "Facilitating Creative Thought Through
Object Play in Young Children." The Journal of Creative

 

Behavior, 17, No. 3, (1982), 61-64.

nxcnronn sran UNIV. LIBRARIES
IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII
31293007026390