Citation
Community educational processes

Material Information

Title:
Community educational processes group perceptions of energy issues
Creator:
Blalock, Carol Douglass, 1944-
Publication Date:
Copyright Date:
1980
Language:
English
Physical Description:
xi, 137 leaves : ; 28 cm.

Subjects

Subjects / Keywords:
Best available technology ( jstor )
Education politics ( jstor )
Educational resources ( jstor )
Energy technology ( jstor )
Environmental conservation ( jstor )
Environmental technology ( jstor )
Learning ( jstor )
Political attitudes ( jstor )
Political education ( jstor )
Political organizations ( jstor )
Curriculum and Instruction thesis Ph. D
Dissertations, Academic -- Curriculum and Instruction -- UF
Educational sociology ( lcsh )
Personal construct theory ( lcsh )
Power resources -- Public opinion ( lcsh )
Genre:
bibliography ( marcgt )
non-fiction ( marcgt )

Notes

Thesis:
Thesis--University of Florida.
Bibliography:
Bibliography: leaves 132-135.
General Note:
Typescript.
General Note:
Vita.
Statement of Responsibility:
by Carol Douglass Blalock.

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Copyright [name of dissertation author]. Permission granted to the University of Florida to digitize, archive and distribute this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Resource Identifier:
023374237 ( AlephBibNum )
06731880 ( OCLC )
AAL3812 ( NOTIS )

Downloads

This item has the following downloads:


Full Text













COMMUNITY EDUCATIONAL PROCESSES:
GROUP PERCEPTIONS OF ENERGY ISSUES








BY

CAROL DOUGLASS BLALOCK


A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL
OF THE UNIVERSITY OF FLORIDA
IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE
DEGREE OF DOCTOR OF PHILOSOPHY



UNIVERSITY OF FLORIDA

1980





















TO MY HUSBAND TONY

AND MY CHILDREN

JEANNE, PATRICIA, AND ELIZABETH

FOR THEIR LOVE AND ENCOURAGEMENT,

AND TO MY PARENTS

WHO STARTED IT ALL















ACKNOWLEDGMENTS


The present research could not have been accomplished

without the assistance and cooperation of many individuals.

The researcher wishes to express appreciation to her doc-

toral committee,Dr. Arthur J. Lewis, Chairman, Dr. Roderick

McDavis, and Dr. Lynn C. Oberlin,for the time and guidance

provided. In particular, she is indebted to Dr. Arthur J.

Lewis for his valuable advice at critical stages in the

design and writing of this dissertation.

The researcher would like to further recognize Dr.

James R. Kennedy for his skill in the development of the

software used for the repertory grid analysis and Dr. H.

Anthony Blalock for scientific consultation in the data

analysis of energy issues. In addition, special thanks are

extended to Dr. Eugene A. Todd, Dr. H. A. Ingley, III,

Mr. John Dykes, and Mr. Lem Lee for their assistance in

data collection.

The researcher also expresses gratitude to her husband,

Tony, and her daughters, Jeanne, Patricia, and Elizabeth, for

their love, encouragement, and patience over the years

required to complete her graduate work.

Finally, the researcher would like to acknowledge the

influence of the many theoreticians and researchers whose

work provides the conceptual base for this dissertation.
iii









They rightfully share any credit this dissertation reflects.

The researcher feels as Sir Isaac Newton did: "if I have

seen thus far, it is because I have stood on the shoulders

of giants."
















TABLE OF CONTENTS


CHAPTER PAGE

ACKNOWLEDGEMENTS . . . . . . 111

LIST OF TABLES . . . .... .. . . vii

ABSTRACT . . . . .... .. . .. ix

I INTRODUCTION . . . .... .. . . .. 1

The Problem and Its Context . . . 1
Personal Construct Theory . . .... ... 5
Design of the Study . . . . . .. 11
Organization of Dissertation . . .. 16

II REVIEW OF LITERATURE ...... .. . . 17

Analysis of Social Perceptions . . .. 17
Cultural Foundations of the Study . . 31
Summary . . . . . . ....... 44

III PROCEDURES . . . .... .. . . . 46

Selection and Identification of
Respondent Groups . . . ... 46
Design and Administration of Repertory
Grid Instrument . . . . . . 47
Data Reduction and Analysis ...... 54
Statistical Procedures and Tests of
Hypotheses . . .... .. . . . 57

IV RESULTS . . . .... .. . . . 63

Frequency of Element and Construct
Citation . . . . . . . . 64
Numbers of Constructs Elicited . . .. 67
Degree of Construct Association . . .. 68
Results of Factor Analyses of Grids . . 72

V CONCLUSIONS AND RECOMMENDATIONS .... 99

Group Perceptions of Energy Issues . . 102









CHAPTER


PAGE


Implications for Educational Planning
and Curriculum Development . . ... .105
Implications for Other Areas of
Educational Planning . . . ... .107


APPENDICES

A REPERTORY GRID INSTRUMENT . . . . .. 110
B DEMOGRAPHIC CHARACTERISTICS OF GROUPS . 113
C MASTER LIST OF ELEMENTS . . . ... 114
D MASTER LIST OF CONSTRUCTS . . . ... 116
E TYPICAL SPSS FACTOR ANALYSIS OUTPUT ... 119


REFERENCES . . . . . . . . ... .. . 132


BIOGRAPHICAL SKETCH . . . . . . . ... 136















LIST OF TABLES


TABLE PAGE

1 Demographic Characteristics of Groups . . 48

2 The Most Commonly Cited Constructs by
Groups . . . . . . . ... 65

3 The Most Commonly Cited Elements by
Groups . . . . . . . ... 66

4 ANOVA Table of Number of Construct ..... 69

5 ANOVA Table for Associations per Construct 71

6 Summary of Factor Structure for Engineering
Student Group . . . . . ... .73

7 Summary of Factor Structure for Education
Student Group . . . . . . . 75

8 Summary of Factor Structure for Community
Group . . . . . . . . ... 77

9 Unrotated Principal Components for
Engineering Students . . . . . 78

10 Unrotated Principal Components for
Education Students . . . . ... 80

11 Unrotated Principal Components for
Community Group . . . . . ... 82

12 Varimax Rotated Factor Matrix for
Engineering Student Group . . . ... 84

13 Varimax Rotated Factor Matrix for
Education Student Group . . . ... 86

14 Varimax Rotated Factor Matrix for
Community Group . . . . . ... .88

15 Major Factors for Engineering Student Group 90









TABLE PAGE

16 Major Factors for Education Student
Group . . . . . . . . 91

17 Major Factors for Community Group . . 92


viii















Abstract of Dissertation Presented to the Graduate
Council of the University of Florida
in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy



COMMUNITY EDUCATIONAL PROCESSES:
GROUP PERCEPTIONS OF ENERGY ISSUES



By


CAROL DOUGLASS BLALOCK



JUNE, 1980

Chairman: Arthur J. Lewis
Major Department: Curriculum and Instruction


The major challenge to the planning of community educa-

tional processes lies in the pluralistic nature of society.

Educational theorists agree as to the critical need for an

emerging role for education in providing linkage between the

various formal and functional groups in society. In accept-

ing this challenge educators must find tools for mapping and

quantifying similarities and differences for values and per-

ceptions across groups. The purpose of this study was to

demonstrate the utility of a method based on Kelly's per-

sonal construct theory for mapping and quantifying attitudes

about energy issues.









Kelly's personal construct theory views human thought

and learning as a process of construing. In this view indi-

viduals view ideas about the world which give meaning to

information and experiences. These constructs are bipolar

so that new information is seen as like one end of the polar-

ity and unlike the other. There are also constructs about

constructs at various levels. Thus, people develop systems

of constructs relating to any particular area of their lives.

One measure of a person's effectiveness is how congruent his/

her construct system is with reality and consequently how

well it permits prediction of events. In this sense, Kelly

saw all persons as scientists making and testing models of

reality.

A major purpose of the study was to demonstrate the

utility of repertory grid analysis in identifying and quan-

tifying perceptual differences among groups. Repertory grid

questionnaires were administered to three groups: a group

of senior education majors at the University of Florida, a

group of senior mechanical engineering majors at the Univer-

sity of Florida, and a group of adult learners enrolled in

energy awareness or radio/T.V. repair classes at Santa Fe

Community College. These groups were expected to have sub-

stantially different energy-related perceptions because of

their widely different educational experiences. Results of

this study demonstrated that all three groups shared similar

awareness of energy issues, but that they construed these

awarenesses differently. Differences were also noted with









regard to perceptual complexity and connectedness but not

centrality. Factor analysis of the pooled repertory grid

data for each group permitted the identification of common

dimensionalities of construing among the group. The use

of factor loading scores as potential indicators of values

is described and an example illustrated. Suggestions are

made for the incorporation of information derived from the

repertory grid studies into community educational planning.

Strategies are described for directing dialog among dimen-

sionalities identified as critical by factor analysis. The

extension of the method to a broad range of educational

activities is discussed.















CHAPTER I
INTRODUCTION


The Problem and Its Context

The purpose of this study was to examine the perceptions

of undergraduate students in engineering, education, and the

community about energy issues in order to map these percep-

tions and quantify similarities and differences among groups.

Data from this study provide groundwork for energy education

and the design of the study is a paradigm for studying group

perceptions about a variety of issues.

It has long been recognized that education is a life-

long process which takes place throughout the community

(Dewey, 1916; Cremin, 1976; Cross, 1979). Many educators

have identified the multiplicity of educative forces and

institutions in modern society. Illich (1971, p. 47) was

perhaps the first to propose the development of "learning

webs--voluntary networks that would permit any student at

any time to gain access to any resource that might help him

define and achieve his goals." Cremin (1976, p. 27) sug-

gests that such networks already exist to some extent and

expands the definition of education to include any "deliber-

ate systematic and sustained effort to transmit, evoke, or

acquire knowledge, attitudes, values, skills, or sensibili-

ties as well as any outcomes of that effort."









Society and communities in the United States are replete

with efforts to inform and persuade which are both deliber-

ate and sustained. However, the extent that these efforts

are broad and systematic in nature is more a reflection of

cultural homogeneity than the result of planning and coordin-

ation. What may be needed for both efficiency and greater

effectiveness is not more educational efforts, but coordin-

ation of educational efforts to be mutually supportive rather

than duplicative or divergent.

The anthropological literature describes an institution-

alized network for the dissemination of information which

provides most of the formal ("technical" in the sense of

Hall, 1959) communication in our society. Major components

of this network are public agencies and institutions, the

media, and specific public interest groups. This network

is characterized by information that is processed in a cen-

tral locus and disseminated through a multibranching series

of pathways. This type of network is designated as dendritic.

The nature of a dendritic network is such that it is possible

to coordinate the flow and control the quality of information.

It should be noted that information flow is bidirectional

while control and coordination goes downward.

People living in communities develop a series of net-

works of communication which are characterized by random

paths in which information flows out from a locus of percep-

tion. This type of network is designated as radiative. A

radiative network operates in the spreading of rumors and









the development of "grass-roots" movements. Because of the

dynamic and continuously changing nature of this network

information flows cannot be coordinated nor can information

quality be controlled. Still this radiative type of network

is very important in the life of a community and probably

operates strongly to condition values and perceptions. In

fact, it may be this type of network that drives the statis-

tical convergence in the Delphi technique (Mayer, 1964;

Myers, 1979).

It would seem reasonable that the values and perceptions

of an individual tend to be determined by those implicit in

the communication in the dendritic network unless a differ-

ent perception has been spread through the radiative network.

Consequently, each type of communication will affect a given

individual to a varying extent at different times on a single

issue and on different issues at the same time (Robinson,

1950).

Concepts of network types and influences can be brought

to bear on a variety of current issues. Many studies have

demonstrated that under certain conditions (which are poorly

understood) people behave in a way which is inconsistent with

their announced values. For example, studies sponsored by

the United States Department of Energy have shown that while

most people strongly support energy conservation they tend

not to support this behaviorally (Public Energy Education,

1979). Thus while we can make inferences about the values

and perceptions an individual holds from an analysis of that









person's ethnic and socioeconomic background, educational

status, and/or formal associations, these influences are

likely to be useful only in the absence of other strong

influences conditioning that individual's behavior. In

particular the influence of a radiative network on specific

perceptions would appear to be exceptionally strong (Mayer,

1964). This may account for the type of observed discrep-

ancies referred to above.

It is possible to determine functional groupings based

on clustering of like perceptions and values (Kluckholm and

Strodtbeck, 1961). Correlating these functional groups with

institutional associations can provide insight regarding the

value structure of a community (Kluckholm and Strodtbeck,

1961; Robinson, 1950). All managed networks by necessity are

dendritic; thus, it may be argued that managed networking will

be most effective where radiative networks are weak or absent.

Because they lack lateral communication capabilities, dendritic

networks cannot provide the value reinforcement which takes

place in radiative networks. This effect may be approximated

by the coordinated actions of several dendritic networks on an

individual belonging to all of these networks.

Agencies assuming responsibility for providing public

information and clarifying values on issues of social impor-

tance can improve their effectiveness by precisely adapting

their program content and delivery methods to the character-

istics of target groups (Torda, 1978; Rosenstein, 1978;

Cremin, 1976; Dewey, 1938). Unfortunately, there is at this






5


time no generally accepted methodology for mapping the values

and perceptions of such target groups. Were such a method-

ology available, its use could lead to the effective deline-

ation of target groups and their characteristics.

If educational activities relating to areas of social

concern are to be effective in facilitating participatory

decision-making, they must meet four criteria: relevance,

timeliness, accessibility, and compatibility with group

values and perceptions (Public Energy Education, 1979).

Much research has been done regarding the meeting of the

first three criteria; relatively little work,however, is

available in the area of mapping and analyzing perceptions.

It is the purpose of this dissertation to demonstrate the

use of personal construct theory as developed by George

Kelly (1955) as a tool for mapping and contrasting percep-

tual and evaluative activities of individuals and groups.

Personal Construct Theory

George Kelly's (1955) introduction and development of

personal construct theory is based on a unique set of assump-

tions regarding the nature of mankind and the human universe.

Three assertions are made in relation to the universe.

Humanity knows the universe to be real and not the result of

someone's imagination; humanity knows that the universe can

be understood only by repeated encounters over a period of

time; and humanity knows that all events are interrelated.










Kelly's assumption that the universe contains real

events and objects is complemented by his assumption that a

person's internal events are as real. Thoughts about exter-

nal objects or events also have an internal reality as

authentic as the happenings themselves. Kelly believed our

knowledge about the universe is determined by the extent to

which we can interpret it. For individuals, this interpre-

tation gradually approaches a true picture of events by

successive approximations. Unlike the purely subjective

existential or phenomenological theorists, Kelly holds to

the possibility of accurate knowledge of objective reality.

Mankind in Kelly's theory is trying out his interpretations

or constructions for their accuracy in predicting the world

he is beginning to understand.

Secondly, Kelly believes that some aspects of existence

can only be made comprehensible by the notion of time.

Behavior is explicable only in the context of the future as

well as the present and the past in Kelly's world view (1955).

The third aspect of Kelly's personal construct theory, the

notion of an integral universe, is consistent with the

development of a system of ideas to bring coherence to one's

world view. Kelly tried to describe how systems like this

operate. The concept of an integral universe is both a phil-

osophical and a psychological statement. People are seen as

continually striving to construe relationships where there

were none perceived before (Bannister, 1968).









Kelly visualizes humans as distinguished by their ability

to not only respond to life's events but able to represent

their environment. Personal construct theory is based on

people knowing the world by means of the constructions they

place upon it. According to Kelly humans are bound by events

to the degree that they have the ability to construe them.

All present perceptions are open to question and reconsidera-

tion. Kelly's philosophical position has been called construc-

tive alternativism (Bannister, 1968).

Personal construct theory (Kelly, 1955) can be applied

to the analysis of physical environments as well as personal

perceptions. Kelly postulated that knowledge precedes behavior

and that a person's knowledge is developed over time by accu-

mulating and categorizing information. People develop images

of their environment in this manner. Kelly calls these images

constructs. Construing means to Kelly placing an interpreta-

tion upon. Moreover, "in construing, the person notes features

in a series of elements which characterize some of the elements

and are particularly uncharacteristic of others" (Bannister,

1968, p. 7).

The technique used to visualize an individual's constructs

contains four steps, each of which will be described in detail

in a subsequent section of this work. The first task is the

listing of relevant elements, i.e., observations; the second

is the generation of bipolar constructs from these elements;

the third is the location of each element in total construct

space; and the fourth is the rotation of axes to identify









factors, combinations of constructs which divide element

groups. The first two tasks require only listing, the third

requires generation of a grid or matrix of element-construct

relationships, and the fourth a computerized analysis.

The analysis of a grid is to make clear the structure

in the grid; analytical techniques therefore should not

impose the structure of the experimenter. If each factor in

the grid represented were a separate unrelated opinion of

the subject, no structure would emerge. However, all grids

developed by people have inherent structure. In other words,

the relationships among individual elements which are sus-

ceptible to mathematical analysis are representative of

basic psychological processes. Grid analysis, then, should

be applicable in principle to any area of perception and

cognition.

Application of Personal Construct Theory to the Study of
Environmental Perceptions

Deutsch (1972) used Kelly's repertory grid test to ask

subjects to apply a group of their own constructs to a set

of known environments. Graduate students in architecture

were compared with graduate students in other fields. Sig-

nificant differences between groups were found in verbal

content and organizational structure. The results of this

study suggested further work could be done on the factor

analytic procedures used as well as on the reliability of

the coding methods used.









Stringer (1976) used personal construct theory (Kelly,

1955) to facilitate public participation in planning to re-

build Wistow Hill "Triangle," a London neighborhood, in 1970.

The general feeling from the public was that some form of

remodeling should be attempted. Stringer was seeking a

general understanding of environmental value systems. Using

personal construct theory to allow individual differences to

emerge was essential to obtaining a group of elements repre-

sentative of individual construct systems. Instead of using

simple preference orderings, Stringer believed the meanings

of elements are more significant when observed in relation

to each other. Fewer assumptions by the researcher then need

to be made. Stringer felt the repertory grid technique

"integrated perception and evaluation" at a time when politi-

cal rallying made participation a reality. Attitudes are

changing according to Stringer and "doing" has become more

important than having. Change has also become of paramount

importance in the process of environmental perception (Kelly,

1977).

Application of Construct Theory to Energy Education

The energy issue, like many other issues of our time,

presents both technological and philosophical problems. As

a case in point, the population of Sweden enjoys a comparable

standard of living to that of the United States at about 60

percent the energy cost using substantially the same tech-

nology (Schipper, 1976). As important to discovering addi-

tional oil and alternate energy resources is the development









of cooperative group processes to deal with the critical

social issues. The role of education is to facilitate the

decision making process by communicating information and

clarifying values. Personal construct theory can provide

specific information to deal with perceptions and values of

the population as a whole and certain groups specifically.

The methodology used in this study has yet to be demon-

strated in the context of energy perceptions or other criti-

cal social issues. The purpose of the study was to demon-

strate that personal construct theory in this form is useful

in delineating the energy perceptions of three groups which

can be expected a prior to differ in attitudes. A group

was chosen with obviously different background so that

there would be a maximal likelihood that each might look at

the energy problem differently. A group of senior under-

graduate engineers specializing in energy conversion was

selected to represent a group of individuals knowledgeable

in energy use and transfer. The second group was composed

of senior secondary education majors chosen to represent

lay knowledge on the energy problem as well as specialized

awareness of the participatory nature of our society with

certain social and political orientations. The last group

were representatives of the community at large. These

students were registered at Santa Fe Community College

receiving credit for either a radio/T.V. or an energy cor-

respondence course. It was believed they demonstrated a

wide spectrum of attitudes.









Design of the Study

The sample population included approximately 75 under-

graduate students of the University of Florida and Santa Fe

Community College. The first group of 28 were senior mechan-

ical engineering majors studying energy conversion. The

second group of 32 were University of Florida seniors also

majoring in secondary education. The third group of 14 com-

munity members were chosen from two classes at Santa Fe

Community College which were known to be heterogeneous in

terms of age and occupations. All of the subjects were asked

to complete a three part listing based on Kelly's repertory

grid method. In the first part the subject was to list 20

elements related to the energy problem. In the second part

the subject chose from the first part pairs of factors which

were the same and one opposite forming a triad. Subjects

were encouraged to state how factors were the same or oppo-

site (from constructs). The third part consisted of a grid

with elements along the X axis numbered one to twenty and

constructs located on the Y axis lettered A through Z.

Respondents were asked to correlate each construct with each

element.

The repertory grid was administered to the engineering,

teacher, and radio/T.V. classes allowing approximately an

hour for completion. If more time was required for comple-

tion, the researcher extended the time as needed. The reper-

tory grid was mailed along with the mid-term exam to the









community group studying energy. The researcher's phone

number was included to encourage questions.

Analysis of the data included a number of steps. First,

all element responses were tabulated to ascertain how many

different elements exist and how many times each element is

repeated. Secondly, constructs were tabulated to determine

how many constructs exist and their repetitions. An element

by construct grid was developed using factor analysis for

each subject. All grid data for a particular group were

combined to form a composite grid of elements, constructs,

and their relationships for each group. Finally, the three

groups were compared for similarities and differences

(Bannister, 1968).

The following hypotheses were tested:

1. The three groups have not chosen different ele-

ments which can be tested by analysis of the frequency

distribution.

2. The three groups have not developed different

numbers of constructs which can be tested by an analysis

of construct responses.

3. The three groups did not show different numbers

of positive relationships between elements and constructs

which can be tested by grid.

4. Analysis of group data does not lead to different

principal components which can be tested by factor analysis.









Definition of Terms

Fundamental postulate "A person's processes are psy-

chologically channelized by the ways in which he or she

anticipates events" (Kelly, 1955, p. 46).

Construction corollary "A person anticipates events

by construing their replications" (Kelly, 1955, p. 50).

Kelly expressed his assumption that all men act as scientists

in this world (Ryle, 1975). Leman (1970, p. 65) examined

this corollary from a linguistic philosophy and suggested

that 'the characteristically scientific activity is an oper-

ation with language . and that the scientist's most

important problems have to do with the relationship between

language and extra-linguistic reality . ." Leman empha-

sized the 'making-sense-of' aspect of men's constructing of

themselves and reality.

Individuality corollary "Persons differ from each

other in their constructions of events" (Kelly, 1955, p.

55).

Organization corollary "Each person characteristi-

cally evolves for his/her convenience in anticipating events

a construction system embracing ordinal relationships between

constructs" (Kelly, 1955, p. 56).

Dochotomy corollary "A person's construction system

is composed of a finite number of dichotomous constructs"

(Kelly, 1955, p. 59).









Choice corollary "A person chooses for herself or

himself that alternative in a dichotomized construct through

which he or she anticipates the greater possibility for the

elaboration of her system" (Kelly, 1955, p. 64).

Range corollary "A construct is convenient for the

anticipation of a finite range of events only" (Kelly, 1955,

p. 68).

Experience corollary "A persons's construction system

varies as he construes the replication of events" (Kelly,

1955, p. 72).

Elements "The things or events which are abstracted

by a person's use of a construct are called elements. In

some systems these are called object's (Bannister, 1968,

p. 219).

Core construct 'A core construct is one which governs

an individual's maintenance processes" (Bannister, 1968,

p. 221).

Factor analysis 'Factor analysis is a multivariable

method that has as its aim the explanation of relationships

among several difficult-to-interpret, correlated variables

in terms of a few conceptually meaningful, relatively inde-

pendent factors" (Kleinbaum, 1978, p. 276).

Assumptions

It is assumed that the diversity within the sample

groups and among the sample groups is a minimum estimator

of community diversity. It is also assumed that it will be

possible to functionally define groups in later work by









identifying clusters of individuals having between group

differences equal to or greater than those demonstrated by

the groups in this study.

Delimitations

The scope of the study included three well-defined

groups with clearly identifiable educational characteristics.

Thus, in no way could the groups be construed as either a

random or a polar representation of the community at large.

Rather, the value of the study lies in demonstrating a novel

and effective perception mapping technique and demonstrating

the application of this technique to the identification of

functional groupings.

Limitations

The selection of a repertory grid with subjects respond-

ing with negative or positive answers limits the sensitivity

of perceptions of relationships. Using an intensity scale

for each relationship, however, would have made administra-

tion of the repertory grid an extremely tedious process.

Administration of the repertory grid format to groups pre-

vented acquisition of contextual information by the researcher

which might help clarify the dimensionality of constructs.

However, the avoidance of the interview technique reduced

the possibility of experimenter contamination of results.

Interpretation of principal components is a qualitative

rather than a quantitative measure of group differences.

By this method it will be possible to identify how groups

differ in perceptions, but not possible to directly compare

two sets of between group differences.









Organization of Dissertation

In the remaining four chapters of this dissertation

the literature pertinent to this study is reviewed, the

experimental and statistical methodology are described, the

results reported, and the significant implications of those

results discussed. A review of literature pertinent to

social perception research, personal construct theory, fac-

tor analysis, and cultural foundations of this study is

presented in Chapter II. The rationale for the selection of

personal construct theory and repertory grid analysis as

the method of choice is described in Chapter III. A

description of the groups studied and a discussion of the

statistical methodology used to analyze and interpret the

results also are included in this chapter. The numerical

results of the study are presented in Chapter IV. Raw data

summaries, descriptive statistics, and analytical results

were tabulated and interpreted. A discussion of the find-

ings in light of the original experimental intent is pre-

sented in Chapter V. Additionally, a review of the impli-

cations of this study will suggest extensions into network

identification and applications to other areas of educa-

tional research.

For the sake of brevity and convenience, the groups of

engineering students, education students, and community

college students will be referred to from time to time in

the text and tables as "engineers," "teachers," and "commu-

nity groups," respectively.















CHAPTER II
REVIEW OF LITERATURE


In developing and demonstrating the methodology of this

study, it has been necessary to draw upon the fields of phil-

osophical foundations of education, cultural foundations of

education, social perception research, personal construct

theory, and matrix methods of statistical analysis. The pur-

pose of this study differs from the work previously done in

that it was designed to provide a vehicle for meaningful user

input to the development of educational programs. Of course,

the problems of relevancy and of the roles of education have

been of concern to many educators. In the following section,

a summary of the philosophical, educational, and methodolog-

ical bases of this work is presented.

Analysis of Social Perception

If culture is communication as Hall (1959) has said,

then education may be thought of as the summated social tech-

nology of communication. Examination of this assertion leads

us to ask three questions: (1) What are the activities and

institutions which communicate and educate in a modern

society? (2) Who decides what is to be communicated and

how? (3) How do the people in a community perceive and

influence the educational process? In answer to these ques-

tions a map of the network of educative forces and activities









occurring in society will be found. Identification will

also be made as to where that network breaks down: in iden-

tifying and incorporating the perceptions of the groups it

serves.

Educational Roles in Social Attitudes and Perceptions

Education has grown from providing only literacy and

vocational skills to modifying social values through the

broad development of all functional aspects of citizenship.

It is important to view education in its broadest possible

context: in its role as interpreter and modifier of social

attitudes. It will also be necessary to identify the role

of education in serving the needs of non-traditional learners.

In analysis of social perceptions a number of educa-

tional philosophers have interpreted and conditioned social

attitudes and values. Dewey (1916) realized that cultural

transmission was education in its broadest sense. He pointed

to the difference between the education most individuals get

simply from the process of living and the deliberate educa-

tion offered by the schools. For Dewey, the advancement of

civilization was a process of complexification. While the

young of savage groups could participate in society by mere

incidental learning, the young of complex cultures could

gain the same degree of participation only with intentional

learning. Dewey's theory of learning was essentially a

theory of the school as agent of society. Dewey worked on

reconciling the polarity between school and society. This

gap has remained. Even in the educational reform of the










1960's, there was ambivalence as to whether schools should

be improved or abolished altogether. In the 1970's, Ameri-

can opinion swung from an overreliance on the school as an

agent of socialization to a widespread disenchantment of

community support for schooling. Dewey anticipated this

phenomenon in the year 1933 when all public institutions

were suffering from a lack of public confidence.

Silberman (1970, p. 2) was aware of the same dilemma.

"If our concern is with education, we cannot restrict our

attention to the schools, for education is not synonymous

with schooling." He saw the need to emphasize the many edu-

cating aspects in American society other than schools.

Silberman's task was to make recommendations on the educa-

tion of educators, but he found this difficult without an

understanding of what education would be like in the years

ahead. Thus, Silberman supported by the Carnegie Corpora-

tion, undertook a four-year study of schools and other educa-

tional agencies. This study, the purpose of which was to

understand basic directional developments and synthesize

these into a coherent program, is today viewed by some as a

theoretical statement of educational progressivism.

Illich (1971) recognized, as had Dewey, that many insti-

tutions and situations educate and that the school is only

one such. He proposed the development of "convivial" educa-

tional institutions called learning webs which were networks

permitting free access to any resource that might help a

student achieve individual goals. He suggested four networks:









reference services in educational subjects, skill exchanges,

peer matching, and reference services to education at large.

Illich was thought to have achieved in design the Dewian

utopia of 1933 in which no schools existed at all. "The

most utopian thing about utopia is that there are no schools

at all" (Dewey, 1933, p. 236). All people would learn what

they needed to know from informal association with others.

Cremin (1976) suggested that Dewey's formulations of

democracy and education be revisited in order to attempt a

redefinition of education. He questioned education as a

result of which Dewey emphasized institutional origins rather

than functions. The theory of education thus created posits

major educative agencies performing a linking role with

respect to other agencies and society. Cremin (1976, p. 27)

has defined education . as the deliberate, systematic,

and sustained effort to transmit, evoke, or acquire knowledge,

attitudes, values, skills, or sensibilities, as well as any

outcomes of that effort." He saw education as a process more

limited than socialization (as perceived by sociologists) or

enculturation (as perceived by anthropologists). This def-

inition nonetheless has moved beyond focusing on the schools

and colleges to identifying all persons and institutions

that educate--siblings, churches, family, friends, libraries,

museums, and others. Finally, the definition allows that

education may produce outcomes which may be anticipated or

may be unintended, with the possibility that the latter could

be the more significant outcomes.









Cremin (1976) defined the role of education by his con-

cept of configurations of education. He envisioned a multi-

plicity of institutions relating to one another within the

larger society. Relationships between and among institutions

he saw as political, pedagogical, or personal. He predicted

a correlative relationship between configurations of educa-

tion and social change or stability. Through the encultura-

tion of the young, configurations of education have maintained

social continuity and stability (Cremin, 1976). Educational

institutions also have played their part in catalyzing social

change according to Cremin.

At the individual level, persons have related uniquely

to configurations of education (Cremin, 1976) according to

individual experience and perspective. Cremin (1976) held

that educational life history of an individual begins with

the efforts of critical others in creating appropriate atti-

tudes and behaviors and results in the individual becoming

a self-directed learner.

Fromm (1968) held a compatible view in that he saw each

person to be sacred and (ideally) to be united with his/her

world. He expressed concern that our educational system

lacked quality despite its great institutionalization and

despite the large numbers of college graduates. According

to Fromm our educational system must become alive and re-

sponsive to each participant. He believed that it is time

for man to assert himself and make the technological society

human.









Leichter (1974) visualized education as a lifelong

process which can take place in a variety of settings and

needs to be understood in each of the settings. She assumed

education took place on numerous levels often simultaneously

and both learning content and process need understanding.

Leichter (1974, p. 239) conceptualized educative style as a

"set of characteristic ways in which an individual engages

in, moves through, and combines educational experiences over

a lifetime." Here it is assumed that educationally signifi-

cant others taught in childhood such relevant educational

attitudes and that these are supported or modified by addi-

tional experience. Leichter's notion (1974) of educative

style focused on continual change linked with the continuity

of an individual moving from institution to institution

within configurations and among configurations. Movement

of this kind by each individual resulted in the development

of individualized networks of education for each person and

resulted in the individual's selection of a variety of

learning activities which reflected that individual's con-

structs pertaining to knowledge and growth.

Kohlberg (1966) realized the difficulty of understand-

ing the educational process without a thorough picture of

maturation, learning, and development. He believed like

Dewey (1933) and Piaget (1973) that understanding of intel-

lectual content and cognitive processes was essential to

the development of moral judgment. Kohlberg's (1966) formu-

lation of six developmental stages of moral judgment









described the potential for continuous growth in individuals

over a lifetime. Kohlberg's moral stages were redefined in

1975 with each of them examined in terms of operating "moral

motives." This categorization was supported by longitudinal

and cross-cultural studies. Kohlberg believed that moral

reasoning was clearly reasoning and advanced moral reasoning

depended upon advanced logical reasoning. A person's logical

stage created a ceiling on the attainment of moral stage. To

understand the learner and the learning process, we must

observe reasoning about choice.

This kind of reasoning defined the structure an individ-

ual chooses and was reflected in configurations of critically

relevant other persons or institutions. Thus, if educators

can understand what an individual finds valuable and why he/

she finds it valuable, they can begin to conceptualize not

only the content of his/her moral judgment but also how he/

she is motivated to learn. On this basis it may be possible

in theory to provide a more supportive climate in which

learning can occur.

Education has now recognized its obligation to meet

needs of the non-traditional learner and is beginning to

address these needs. Knowles (1977) described a new tech-

nology of androgogy or education of adults based on the

premise that adults are critically different than children

in experiencing learning. Cross (1979) emphasized the need

for equal education for 30 to 80 year olds. Researchers

estimated that between 80 and 90 percent of the adult









population carried out one self-directed learning experience

every year. Surveys show between 17 and 32 million adults

are now participating in classes, workshops, groups, or

other organized educational structure. By the year 2000

Cross (1979) forecasts the United States will become an

adult culture with 57 percent of the population over thirty

years of age.

Cross (1979) has shown that education for adults is

elitist with certain populations significantly underrepre-

sented in organized learning activities. She argued that

brokering services and education information centers repre-

sented the greatest hope for shaping the "learning society."

The development of education and information services is

perceived by Cross to involve three steps: (1) collecting

information about the educational resources available,

(2) reaching the intended audience, and (3) assisting clients

to identify and obtain the appropriate learning opportuni-

ties. A strengthening of these three linking functions

will thus contribute immeasurably to the success of lifelong

learning.

The recognition that the learner, especially the adult

learner, must be viewed contextually has led to the important

conclusions by Cremin (1976), Leichter (1974), and Illich

(1971). It is reasonable to expect this matrix of inter-

relationships in which individuals are embedded to influence

and condition values and perceptions. Clearly values and

perceptions affect not only the choices of educational









participation but also the motivation for participation.

These influences are not generally seen as falling within

the purview of education.

Sociocultural Aspects of Attitudes and Perceptions

If, however, one could define functional groupings

within a community based on values and perceptions with

regard to a particular area of interest, it would be pos-

sible to meet the needs of that group for education and

information. A variety of anthropological studies have

demonstrated methodologies for identifying and mapping

social influences as well as for mapping attitudes.

Mayer (1964) directed his efforts toward describing

the networks of relationships among individuals in urban

settings. He observed that people living in communities

developed a series of networks of communication which were

characterized by random paths in which information flowed

out from a locus of perception. He saw such a network

operating in the spreading of rumors and the development

of "grass-roots" movements. Mayer (1964) knew that be-

cause of the dynamic and continuously changing nature of

this network, information flow could not be coordinated

nor could information quality be controlled. Still this

type of network, according to Mayer (1964), was very impor-

tant in the life of a community and probably operated

strongly to condition values and perceptions. In fact,

it may be this type of network that moved the statistical

convergence in the Delphi technique (Myers, 1979).









While Mayer emphasized intra-group networks, Kluckholm

and Strodtbeck (1961) looked at the problem from the perspec-

tive of entire social groups in his study of five South-

western cultures. Using interview techniques informants from

five juxtaposed cultures were questioned about their percep-

tions of social and physical environments. Group mean scores

for the different cultures were compared by a one-way analysis

of variance for several predetermined dimensions. The results

obtained included an accurate map of perceptions along the

predetermined axes. This mapping elucidated the causes of

a great deal of conflict in values among the various groups.

The need for better tools in observing individual

values was expressed by Craik (1970), in his comprehensive

review of environmental psychology, who devoted a section to

the discussion of personality inventories. He pointed out

the neglect by psychologists of items or scales for assess-

ing environmental dispositions in spite of individuals' strong

orientations toward the physical environment, He suggested

a number of environmental dispositions, such as Pastoralism

scale, an Eccological Perspective scale, a Luddite scale,

an Urbanite scale, and measures of environmental sensitivity.

McKechnie (1970, p. 320) developed the Environmental

Response Inventory (ERI) for measuring environmental disposi-

tions, which he defines as "the configuration of attitudes,

beliefs, values, and sentiments .. ." of the individuals

being tested. The Environmental Response Inventory included

items relating to the areas of pastoralism, conservation,









science and technology, urban life, rural life, stimulus

preferences, cultural life, leisure activities, the outdoors,

geographic and architectural preferences, and environmental

memories and knowledge. The responses were factor analyzed

and a series of different factors for men and women surfaced.

The ERI scales did have an interesting correlation with

numerous traditional personality measures, and with environ-

ment related behaviors such as membership in conservation

and agricultural organizations. This result was consistent

with the results of the anthropological work of Kluckholm and

Strodtbeck (1961) and Mayer (1964).

Numerous studies of energy perceptions have been done

which demonstrated that under certain conditions (which are

poorly understood) people behave in ways inconsistent with

their announced values. For example, studies sponsored by

the United States Department of Energy (Public Energy Educa-

tion, 1979) have shown that while most people avowedly sup-

port energy conservation, they tend not to demonstrate this

support behaviorally. Thus while inferences can be made

about the values and perceptions an individual holds from an

analysis of that person's ethnic and socioeconomic background,

educational status, and/or formal associations, these in-

ferences are likely to be useful only in the absence of

information regarding other strong influences conditioning

that individual's behavior. In particular the influence of an

associational network on specific perceptions appeared to be










exceptionally strong to Mayer (1964). This may account for

the observed discrepancies referred to above.

Hall (1959) explored culture by means of communication

theory (information theory) of the electronics laboratory.

Hall (1959) likened communication theory to shorthand for

talking about communication events such as phonetics of

language, orthographics, and telephone signals. He noticed

that the process proceeded in one direction--toward symboli-

zation. He saw an individual's speech as an arbitrary vocal

symbol used to describe something that had taken place or

might have taken place with possibly no actual connection

between occurrences and symbols. Because of the workings of

culture, Hall (1959) saw talking as a highly selective

process. He believed that no culture had discovered a means

for talking without emphasizing some events at the expense

of others. He saw writing, then, as a symbol of a symbol.

Using communication theory (Hall, 1959) took this process

still further. He noticed entire messages of various dura-

tions with some less than a minute and others extending

over years. The study of culture thus could include events

of short duration whereas the study of individuals or govern-

ments involved communication over longer duration. He

developed a system based on tripartite theory which included

three kinds of time: formal time, informal time, and tech-

nical time. Hall (1959) discovered that man had three modes

of behavior and that at any point in time one of the three

orientations would dominate although all three would be









present. Formal activities were taught by a mistake being

made and a correction suggested. Informal learning empha-

sized a model for imitation in which thousands of details

could be passed through generations without specific under-

standing of the rules. Technical learning was shared in

explicit terms from teacher to student usually preceded by

a logical analysis. In summary, the formal mode was a two-

way process while informal learning involved identification

of a model to follow. The technical learning rested with

the teacher.

Numerous trends of social perceptions have been influ-

ential in the evolution of social communication. Joseph

(1979) forecasted a transformation in education through

technology to accommodate the tremendous mass of information

necessary for adaption to a highly complex society. The

information explosion has forced the emerging development

of a "technology for education which included imbedding

increasingly capable, but physically small, micro-processor

logic, digital storage/memory, sensors, communications cir-

cuits and links, and eventually voice actuated and reply

mechanisms for creating convivially smart machines--which

are more humanistic for students" (Joseph, 1979, p. 1).

Meadows (1974) emphasized the information explosion as

it related to the future course of human society or human

survival. He saw that human survival could depend on the

effectiveness with which the population related to the

world's problems or solutions. Meadows (1974) has plotted









the dimensions of time and space and demonstrated how every

human concern can be located at a point on a graph. He

felt life was a challenge for people, many of whom use total

effort to provide daily for their families. Other individuals

act on problems further out in time and space so the pressure

they feel is of a community nature.

Meadows (1974) observed that a person's use of time and

space dependson his/her cultural orientation, immediate prob-

lems, and past experience. Before an individual moved into

a larger space he/she must have solved the more immediate

problems. He further concluded that the more difficult prob-

lems involved longer time commitments made by smaller numbers

of people. The danger perceived by Meadows (1974) was that

by limiting their perspective too much, individuals lose the

ability to cope with problems at state, national, or global

levels.

McLuhan's (1967) global village concept accentuated the

current use of a community of information. He saw the entire

world knowing about other lifestyles through the perceptions

created by television, movies, radio, and all forms of tele-

communications. A hamlet in the Andes or a traffic jam in

Los Angeles has been made the immediate experience of each

individual. Drucker (1969) compared the closeness between

continents today with the less relative closeness between

mansions and slums during the eighteenth century. The apt-

ness of the concept of a global village in terms of communi-

cation can readily be observed in current world events.









Cultural Foundations of the Study

The work reviewed in the preceding sections has traced

the role of education in modern societies, identified the

need for better understanding the role of values and percep-

tions in the learning process, and delineated evolutionary

trends in the expanding mission of education to serve the

needs of all groups in society. In this section we present

a rationale for utilizing knowledge of learner attitudes and

perceptions in the design and development of innovative

learning networks.

Human Cognition and Perceptions in Social Decision Making

Personal construct theory (Kelly, 1966) has demonstrated

the usefulness of looking at human and scientific endeavors

as sharing relevant similarities in that in both, people

pinpoint issues, observe issues, become intimate with prob-

lems, form hypotheses, test hypotheses, relate results to

expectancies, control investments so they can understand

what leads to what outcomes, carefully generalize, and broaden

dogma in view of experience. Kelly (1955) considered man in

this context and delineated the unifying concepts brought

forth through explaining and charting strategies of human

experience, both over individual lifespans, and over centur-

ies. Kelly's (1977) unique contribution to psychology was

the introduction of a single language for explaining human

process making the psychological nature of scientific in-

quiry able to provide new insights into man's potential and

the nature of science. Kelly (1955) did not claim to reach









a total understanding of human process but he did feel the

idea of man-the-scientist was worth exploring.

Kelly (1955) assumed humans were real and not just

existing in fantasy, that the human universe could be under-

stood only through a time perspective, and that the universe

was integral such that given complete knowledge and a wide

enough viewpoint, all events would be seen as interrelated.

Kelly (1966) further accepted that not only was the universe

real but also that human internal events were also real.

Therefore, an individual could come to understand his/her

world only to the extent he/she could interpret it by moving

toward an accurate awareness of events through successive

approximations. This theory bypasses the groundlessness and

subjectivity of phenomenological or existential analysis and

views people as able to test out their own constructs for

completeness of world prediction. Kelly argued that human

life events could be understood only as people acted on them

in view of the present, past, and future. He saw man in a

continual effort of construing relationships where none were

observed before in an attempt to incorporate what was for-

merly diverse into a more integrated universe. This repre-

sentational model of the world allowed humans, according to

Kelly, to make sense from the world and choose behavior in

relation to it. Therefore, Kelly (1966) refused to accept

any once and for all construction of the universe because

he viewed all current interpretations as subject to revision

or replacement.









In a further exploration of the characteristics and

nature of constructs, Bannister (1968) demonstrated that a

construct is a way in which some things are seen as alike

and others as different. For each person the basis of dif-

ference can only be appreciated when the contrast is under-

stood. The range of usefulness of an applied construct is

also necessary for complete comprehension as it is possible

to use similar discrimination while using different ranges

of convenience. A construct is an interpretation imposed

on events, not part of events themselves and thus constructs

are useful inventions, not a part of nature. A construct

was for Kelly (1955) a tool of discrimination and structur-

ing of events in anticipation of future possibilities.

Kelly (1966) suggested that a person could be under-

stood by clearly seeing his/her construction system. Each

person's constructs represented a network of avenues along

which he/she can move. When movement was necessary, each

person was presented with a number of dichotomous choices

each channeled by a construct. Therefore, each construct

represented a pair of rival hypotheses within a system.

Consider, for example, the construct represented by the

triad: automobile, airplane, and railroad train. A hypo-

thetical respondent might suggest that airplanes are like

trains and unlike automobiles in that the former are

usually scheduled and the latter usually not. This construct

clearly dichotomizes travel decisions according to scheduled

and unscheduled arrangements. Note that the same triad









could have led to an entirely different construct; e.g.,

one in which ground transportation was constrasted with

more rapid air transportation. This illustrates the need

for careful interpretation of the like-unlike axis elicited

from the respondent.

Each construct system limits a person's perceptions

beyond which he/she cannot perceive and, therefore, controls

his/her behavior. In relation to sociological cultural con-

structs, decisions arise from individual behavior and per-

ceptions held by groups of people. Two individuals holding

similar construct systems in both discrimination and range

can be seen as having similar sets of constructs. A role

of education is to create values and facilitate participa-

tion of individuals in their life. Education has tradi-

tionally provided information and identified desirable be-

haviors while ignoring the critical link provided between

individuals of similar construction systems. Furthermore,

if education could change perceptions or concepts, our

society as a whole might be more effective in meeting the

challenge of survival. An example of the usefulness of per-

sonal construct theory is in accounting for the difficulties

we find in convincing other cultures to make good use of

transported commodities such as surplus food grain. The

range of convenience of their constructs regarding food may

not include the particular material provided. A perhaps

apocryphal example is the resistance of Central and South

American Indians to the use of high-lysine corn in tortillas.









Their perception of quality corn was that it was yellow; thus

they rejected the white imported corn even though it was

nutritionally superior.

Before education can deal with the development of use-

ful constructs, it must map the perceptions and values of the

target groups. Otherwise, it will be extremely difficult to

enlist the active support of members of the target group who

do not see relevance in the educational program goals. Many

learning theorists have developed hypotheses with related

models. Rogers (1970, p. 158) believed "significant learning

takes place when the subject matter is perceived by the stu-

dent as having relevance for his/her own purpose. Speed of

learning is also influenced by relevance. Probably one-third

to one-fifth of the present time allotment would be sufficient

if material were perceived by the learner as related to his

own purpose." Combs (1974) believed human beings had a

natural ability to learn. He saw motivation in people to

better understand themselves and their world. This fascina-

tion with learning, for experimentation, and extension of

cognitive limits could be released under ideal environmental

conditions according to Combs (1974).

Piaget (1973, p. 70) stated "the general culture which

education is to transmit to the student cannot be restricted

to abstract formation without roots in the structure and real

life of the society as a whole but must consolidate the dif-

ferent practical, technical, scientific, and artistic aspects

of social intercourse into a more organic whole." Piaget









(1973) supported Kelly's (1966) observation that the con-

structs of each individual are real for that person and

that the universe with its system of interrelationships is

also real.

Combs (1974, p. 126) stated "people do not behave in

response to stimuli but, rather, to the meanings these stim-

uli hold for them." Kelly (1966) further described stimuli

as constructs being individually formed and meaningfully used

as guides to further behavior. Kelly's (1955) personal con-

struct theory can be considered a metatheory or a theory

about theories in that it accounts not only for the behavior

of observed individuals, but also simultaneously it accounts

for the activities of the observer-theorist. Kelly's use of

language to structure human approaches to understand events

is unique.

Kelly (Bannister, 1968) developed techniques for elicit-

ing and measuring personal construct systems leading to his

repertory grid as the most sophisticated. The idea that con-

structs are individual bipolar abstractions with precise

ranges of convenience used for an individual's world struc-

ture is considered in the procedure for eliciting constructs.

The significance of exploring and understanding an individual's

system of constructs was recognized by Kelly (1955) in his

elicitation of numerous constructs and his design of statis-

tical techniques in which assessment of links among constructs

was possible.









The repertory grid technique of Kelly (1955) to be

described more fully in the next section lends itself not

only to the mapping of the construct fields of individuals

but also to that of groups. Although architectural and

urban planners are thus far the only non-psychologists to

make use of the method, the limited data available reveal

much promise for its application to a broad range of social

research fields.

Stringer (1976) discovered some important moral and

political implications as well as technical and theoretical

ones while working in the fields of planning and architec-

tural design. He was placed in charge of a study to examine

people's perceptions of alternative environmental aspects of

planning proposals to rebuild a decaying Victorian shopping

center in South London. In seeking general understanding

of environmental value systems, he felt his success would

depend upon not pre-empting the focus of inquiry too soon.

A comparable set of environmental elements would enable re-

spondents to give a clearer picture than one element or a

disparate set. Stringer (1976) saw a range of responses as

more richly definable by their observed relation to each

other. He further envisioned the repertory grid as preserv-

ing individual construct systems by integrating perceptions

and evaluations. Meaning was defined by the grid structure.

The planning context provided a future orientation to a

person's construing which conformed with personal construct

theory. Public participation has been a potential rallying









cry lately but it is in the planning field rather than edu-

cation or social services that most examples can be found

according to Stringer (1976). He examined 200 individual

grids to determine which redevelopment proposal drawn up

in different map formats constituted what respondents con-

sidered "adequate publicity" for local planning.

Leff and Deutsch(1972) did a pilot study using a modi-

fication of Kelly's (1955) repertory grid in which they re-

quested individuals to apply a set of their own constructs

to a set of environments known to them. Graduate students

in architecture, urban planning or studies, and graduate

students in other fields construed their physical environ-

ments. The two groups had significant organizational and

content differences between the environmental verbal con-

struct systems and between the results of this study and

semantic differential studies. The results of the study

suggested that there were two types of subjects in the pro-

fessional group, ones who construed environments in terms

of more qualities than lay persons and ones who construed

environments in terms of fewer qualities. "The finding

that professionals bracket lay persons suggests that there

might be a cognitive source of interactional problems in

that they are concerned with different numbers of environ-

mental aspects" (Leff and Deutsch,1972, p. 289). The find-

ing that some architects use fewer constructs than lay per-

sons and non-architects could suggest that when professionals

with backgrounds interact with lay persons most probably









conflict will occur because the lay persons most probably are

not concerned with as many environmental aspects. The find-

ing also suggested architects will have problems interacting

with environmental planners and designers. Alexander (1964)

saw that professionals are socialized by their educations to

think about environments very differently than non-professionals

and that such a cognitive gap could be a major hindrance to

cooperation. He saw cognitive differences translated readily

into interactional problems when groups think differently

about problems.

Hershberger (1969) carried out a study to compare the

environmental cognitions of students in architecture and

other fields using the semantic differential method. Even

though the study did show group differences the semantic

differential method required all students to use the same

set of scales provided by the experimenter and as a result

inter-individual differences were masked. Furthermore,

associations at the aggregate level did not necessarily

reflect relationships existing at the individual level

(Robinson, 1950).

Repertory Grid Methods

The work described in the previous section has demon-

strated the utility of personal construct theory and the

repertory grid method derived from it for describing the

perceptions of individuals and groups. In this section

will be described the methodology used in relevant work

and the appropriate statistical measures will be discussed.









Certain essential components of the repertory grid and

its conventions should be considered first (Ryle, 1975).

Among them are a list of things to be compared, the ele-

ments, and a list of terms used to compare and contrast

them, the constructs. The first list was compared systema-

tically against the second list to create a grid of figures

which is where the name "repertory grid" comes from. Con-

structs are bipolar according to Kelly (1966); therefore,

constructs should be elicited in the bipolar form. The sub-

ject has to have a list of elements which can be compared

to a list of constructs and vice versa. A researcher's

first responsibility in carrying out the repertory grid is

to elicit from the subject a list of elements and constructs.

The more subjects provide their own elements and constructs

the more they reveal about themselves according to Kelly

(Ryle, 1975).

The first stage in constructing a test is to put together

a set of elements. The next step is to obtain constructs.

The more freedom that is allowed the respondent the more

valid their constructs. The classical method has the tester

randomly choosing sets of three elements from the element

list and asking the subject to describe all the similarities

and differences in the triad. The tester then writes down

these descriptions. Triads are used until no new constructs

are formed. Three elements are considered rather than two

based on Kelly's (1955) stand on the bipolar nature of con-

structs and the fact that this is the minimum to allow for









obvious similarities and differences and to define both poles

of the construct. Constructs are recorded in each subject's

words unless constructs are supplied to describe often used

judgments. If the respondent did not feel comfortable with

the supplied construct, it could be modified.

The tester now has a list of elements and a list of con-

structs from the subject. To complete the test a respondent

is asked to rate each element with each construct by dichoto-

mising, by ranking, or by rating. The form which the test

thus takes is the form simply of rows and columns. The form

of the grid described can be called a "standard grid."

Other modifications can be made as required by the mathema-

tical analysis of the repertory grid.

Let us consider the following hypothetical case in

which a graduate student is asked to respond to the follow-

ing elements each of which is a potential topic for a doc-

toral dissertation:

1. Correlation of Myers-Briggs Extraversion Scales

with height.

2. Comparison of bilingual and monolingual children

on performance in a new language.

3. Correlation of Myers-Briggs thinking, feeling

category with computer literacy.

4. Use of computer-aided instruction to teach mathe-

matics in the middle school.

Our hypothetical respondent is asked to create triads

from the above elements. In each case, one pair of elements









will be alike and the third will be different in the same

way the pair is alike.

Alike Different How

A. 1 and 3 2 1 and 3 used Myers-Briggs

B. 3 and 4 2 3 and 4 involve computers
in education

C. 3 and 4 2 3 and 4 approved of by
dissertation chairman
and 2 is not

Having created the above constructs, the respondent is asked

to fill in the grid using a 2-point rating scale. A + will

be placed in a cell representing a construct and element

having a strong association. A will be placed in a cell

for the construct and element having a strong negative asso-

ciation. The responses given by our hypothetical graduate

student appear below.
Elements

1 2 3 4

A + --

Constructs B + +

C + +


This simple grid represents a mapping of the values and per-

ceptions about the various potential dissertation topics.

The reason for the use of Kelly's (1955) repertory

grid in specific is to facilitate the communication of con-

cepts relevant to understanding individuals and groups

(Ryle, 1975). Ultimately a grid of rows and columns is too

complex to be intelligible to inspection; therefore some









form of analysis and display is required before any conclu-

sions can be drawn. The purpose of mathematical analysis

of a grid is to make obvious the structure in the grid, not

to impose the experimenter's expected structure upon it.

If each point of the grid represents an independent judgment

of the respondent, there can be no structure to reveal. Ryle

(1975) stated that all grids created by humans are character-

ized by visible structure such that there are relationships

between individual ratings which are open to mathematical

analysis and which also relate to psychological process.

Complete analysis of essential properties of a grid (or the

relationships between elements and interaction of elements

and constructs) demands, according to that author, computer

analysis. Bannister and Mair (1968) described methods which

can easily be applied to grids based on dichotomization of

elements. Computer analysis also allows for examination of

element relationships and construct-element interactions

which are only fully explored by such an analysis. Problems

of centrality versus extremity of ratings and a skewed dis-

tribution of elements on constructs can also be addressed.

Two elements receiving precisely the same rating on

every construct must be perceived by the subject as the same

or indistinguishable in terms of constructs chosen. Two

elements rated at opposite ends of a construct must be per-

ceived as highly dissimilar. Similarity of any two elements

can be estimated by the measure of the distance between any









two elements and also the degree to which they are related

as similar in relation to the constructs used.

Ryle (1975) treated analysis of constructs as similar

or different in the same way as the elements. The associa-

tion between two constructs based on all the elements is

given with a value of -1 or +1 in the table of construct

correlations. Assuming enough elements have been rated

against constructs, the correlations between constructs

could understand his world. Knowledge of these could help

the researcher explain future behavior.

A graph can be created from an analysis of elements

to represent conceptual space in which the meaning is indi-

cated by the constructs and the location of each element

in relation to these constructs and to the other elements.

Summary

The literature review of this section has demonstrated

the need for educators to accept a role in the development

of society to serve non-traditional learners outside of

school settings. Very little is known about the values

and perceptions of these non-traditional learners with re-

gard to a variety of social issues which education will

need to address. In fact, very little is known about

methodology for acquiring information about values and

perceptions in a reliable way. In limited studies done

by architects and planners, the repertory grid method of

personal construct theory appears to have much promise for

mapping values and perceptions of individuals and groups.





45


This study demonstrates its utility for such a mapping with

regard to energy issues and suggests ways the results can

be used to plan curricula for energy education activities.















CHAPTER III
PROCEDURES


Kelly's theory of personal constructs and recent appli-

cations of its associated repertory grid method to map envi-

ronmental perception have provided a potentially powerful

tool for analyzing and interpreting group perceptual struc-

ture. Any technique for mapping complex perceptions is it-

self necessarily complex because of the many decisions

involved in its application.

The construction, administration, and analysis of the

repertory grid instrument are described in this chapter.

The makeup of respondent groups and the reduction of the

acquired data into forms suitable for hypothesis testing

are reported. Finally, the statistical procedures for

analyzing the data and testing the hypotheses are described.

Selection and Identification of Respondent Groups

The rationale for group selection was to find groups

of individuals who might be expected a prior to be differ-

ent in energy values and perceptions. Further, a group was

selected to represent a community sample with no prediction

made about member's homogeneity of values and perceptions.

A class of senior education majors at the University

of Florida was selected to represent a group with socio-

logical sophistication but limited technological expertise.









To represent a group with technological and energy exper-

tise, a class of senior mechanical engineering majors at

the University of Florida was selected. Community members

were sampled by soliciting responses from participants in

two classes at Santa Fe Community College which were known

to be heterogeneous in terms of ages and occupations One

class consisted of individuals participating in a "course

by newspaper" on energy; the second was made up of persons

studying radio-TV repair.

Valid responses were acquired from 28 engineering stu-

dents, 32 education majors, and 14 community students. The

average age of the groups was 24 for the engineering students,

25 for the students in education, and 31 for the community

group. Table 1 summarizes the demographic information avail-

able for each group. Appendix B lists majors and/or occupa-

tions for the groups.

Design and Administration of Repertory Grid Instrument

The instrument used to elicit repertory grid information

is found in Appendix A. Respondents were asked to list fac-

tors they believed were important in understanding the energy

problem. Several forces were implicit in the design of this

form. First among these was the decision not to provide

respondents with a list of elements, but rather to permit

free-form generation of element lists by each individual.

Providing a list of elements forces respondents to limit

their thinking to the researcher's supplied scope. The

eliciting of free-form lists, on the other hand, makes












Demographic


Average Range
Age Age


Engineering 24 21-36

Teacher 25 21-50

Community Group* 31 19-68



*Respondents from energy class = 4;


Table 1

SCharacteristics of Groups



Male Female Caucasian


28 0 27

7 25 32

12 2 12



from radio-TV class 10.


Black


0

0

1


Oriental Indian


1 0

0 0

0 1









direct comparison of individual grids impractical. This

latter difficulty can be dealt with, in both concept and

practice, by a procedure which will be demonstrated later

in this section.

The choice of requiring exactly 20 element responses

was made for the following reasons:

1. It was necessary to have uniformity of element

list size in order to meaningfully compare the resulting

construct list sizes.

2. Leff and Deutsch (1972) found 20 elements to be a

workable number of elements in their study.

3. Twenty three-digit element identification fields

leave a reasonable amount of space on an 80-column IBM

card for demographic and identification information.

Instructions for the first part of the response were

made as brief as possible. Subjects were asked to "List

20 factors you believe are important to understanding cur-

rent world energy problems." No attempt was made to define

"factor" nor was an energy related example provided. Indi-

vidual responses ranged from single words such as names of

oil companies to complex relational phrases such as economic

or political influences. In most cases individuals were

able to complete this listing in 20 to 30 minutes.

Below is a sample of a single individual's response.

Note that two numbers appear to the left of each response.

The first is the assigned element code from the categorical

list of Appendix C which is described in the next section









of this chapter. The second is just the number of the ele-

ment in the list. The category number was assigned by the

researcher on two criteria. The first was the meaning of

the listed element. The second was derived from examining

the grid column of construct associations to achieve con-

textual understanding of the element's meaning. For example,

note that element 8 "rising gas prices" could have been

recorded as a sociopolitical factor, a resource/distribution

factor, or an economic factor. Column 8 of the association

grid below, however, shows positive associations with con-

structs B, C, and E, all of which suggest economic dimension-

alities. Negative associations are recorded for constructs

A and D which suggest resource and technology dimension-

alities. Thus the element was assigned to category 80

"economics." Note that the underlying structure of the

grid would have been presented equally well had the category

assigned been 30 or 31, providing all other similar responses

by other individuals were treated consistently. This is true

because the purpose of the study was to analyze the response

structure of the groups rather than to interpret the precise

meaning of the responses.

Provision was made for the elicitation of from 0 to 26

constructs on the form. In the study reported by Deutsch

(1972) a maximum of 29 responses were found. It was antici-

pated that because the list of elements in this study was

not provided, fewer constructs would be produced. This was,

in fact, the case with 17 being the maximum number from a









single individual. Each construct was defined by informa-

tion of two types: the triad of elements which provided

the poles of the dichotomy, and the organizing concept

("How") which provided the dimensionality of the dichotomy.

Thus, the instrument could distinguish between constructs

using the same triads, but different dimensionalities. For

example, "car A is like car B and unlike car C." In one

case the dimensionality might be: "Because A and B are

sports cars while C is a station wagon"; in another "Because

A and B will run while C will not run." No responses were

considered valid for constructs if both triad and dimension-

ality was supplied.

Survey of Energy Problems

I. List 20 factors you believe are important to under-
standing current world energy problems.

73 1. Arab oil problems
72 2. Nuclear energy processing
74 3. Gasohol
80 4. Rising electric bills
68 5. Cleaning up coal use
06 6. Windmill generators
42 7. "Turn out the lights" syndrome
80 8. Rising gas prices
69 9. Use of natural gas
74 10. Use of geothermal energy
06 11. Use of solar energy
30 12. 55 m.p.h. speed limit
34 13. World politics
73 14. Third world's desire for energy
21 15. Rising rate of energy consumption/per cap
72 16. Difficulty of finding new oil
30 17. Emergency building temperature restrictions
52 18. Building insulation
51. 19. Car pooling and car pool lanes
80 20. Increased costs of all transportation









An example of the instructions and response to the con-

struct section of Part II using the same single individual

as in Part I follows:

II. Find pairs of factors which are alike. For each
pair find a third factor that is different in the
same way as the pair is alike. Continue to do
this until you can find no more contrasts. (Most
people will have more than 5 and fewer than 25
contrasts.) For your convenience, simply record
the numbers of the factors in each contrast in
the blanks provided below. In the space along-
side each triplet briefly describe how the factors
are similar or different.

Alike Different How

A. 10 11 5 018 Ways to find new energy

B. 1 3 15 031 Politics of the situation

C. 12 17 18 028 Federal mandate

D. 6 11 9 018 Still experimental

E. 4 20 18 009 Price-to-pay increase

Most respondents required 20 minutes or so to complete this

portion.

Constructs were examined for triad content. First,

the triad was recorded as in the element list. If a partic-

ular element code occurred twice other than in the alike

pair, the element coding was reexamined. The original ele-

ment triad was looked at, and the dimensionality axis ("How")

applied. The construct was then assigned to a number on the

construct category list. This list appears as Appendix D.

If an appropriate category did not appear on the list, one

was created and a new number was assigned.









For example, construct C above has the original ele-

ments of "55 m.p.h. speed limit" (12) and "emergency build-

ing temperature restrictions" (17) contrasted with "building

insulation" (18). The dimensionality ("How") is "federal

mandate." What this is interpreted to mean is "12 and 17

represent politically-motivated conservation requirements

while 18 is a conservation measure which is not required

politically but may be of value." Construct number 28 which

has the dimensionality "conservation versus politics" was

assigned to this response.

In filling out the association grid, the choice was

made to seek only a + or a response for each construct-

element association. This was done for reasons of expedi-

ency as well as for theoretical reasons. The use of a rat-

ing scale to estimate the degree of association (relevance)

between each element and each construct would be useful in

observing the evolution of attitudes of an individual over

time. In this case, however, it was felt the time required

for filling out the grid would be inordinate for an indi-

vidual who had developed possibly 20 constructs. Most

persons were able to complete the grid in 15 to 20 minutes.

An example grid for the above elements and constructs follows.

Interpretation of the simple dichotomous responses

elicited is relatively easy. "Construct X relates in some

way to element K" (+) or "Construct X does not relate in any

I think significant to element K" (-). Note that the latter

interpretation is difficult to distinguish from "I didn't









III. Look at the grid on the page. Each square in the
grid has a small letter and number on it. The
numbers refer to your list of factors from Part I
and the letters to your list of contrasts from
Part II. Start with box Al. Does contrast A
apply to factor 1? If so, place a + in box Al;
if not, place a in box Al. Repeat the process
for box A2 checking whether contrast A applies
to factor 2. Continue until you have checked
your contrasts against all 20 factors.

Factors

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

A - + + - + + - - -

B + + + + -- + + + + + + -

C + +- +- -+ + + + + -

D + ++ - + + + + -

E ++ + + + + + ++ + + + + + + + ++ + +


consider element K (construct X, or both) very important,

and therefore, didn't mention them." This point is concep-

tually central to the use of individual grids to study group

values and perceptions. It is on this basis that we can

merge non-overlapping element and construct responses into

group grids as discussed in the next section.

Data Reduction and Analysis

Inspection of the received responses revealed that

there was a great many repetitions of elements and, to a

lesser degree, of constructs. In many cases, generic ele-

ments were represented by specific examples; e.g., Texaco,

Exxon (oil companies) and Iran, Saudia Arabia (O.P.E.C.).

Both these phenomena were anticipated, and in fact, neces-

sary for meaningful analysis to occur.









Because of the redundancies, it was possible to com-

press the approximately 600 element responses and 250 con-

struct responses per group into an orderly and manageable

group grid. To accomplish this, a taxonomic listing of

constructs was developed and each individual's responses

were recorded according to that taxonomy. The taxonomic

listing is found in Appendices C and D. The more complex

constructs were less easily compressed by categorization;

nonetheless, the same approach was used to categorize these

responses. A listing of construct categories is found in

Appendix C.

Care was taken to retain the original meaning of the

responses. Thus, tests were designed to ensure minimum

information loss in the recoding of the responses. To

assure fidelity in the recoding of elements, recorded triads

were examined for cases in which a single element occurs

both as a member of the like pair and as the unlike element

as was described in the previous section. In such cases,

the coding was reexamined to ascertain if a better assign-

ment could be made. If not, a subdivision was created in

the taxonomy so as to preserve the degree of discrimination

in the original response. After recoding, less than 1 per-

cent of element responses contained the same element in

more than one occurrence.

Consistency of interpretation of constructs was assured

by inspection of the row vectors of associations. In cases









where a single individual used the same construct two or

more times, the grid scores of each occurrence were com-

pared. If more than two grid cells were scored differently,

the constructs were reexamined, and, if necessary, reassigned.

A further check was the inspection of combined scores on

group grids to assure non-randomness of scoring distributions

across members within a group for a single construct.

The scoring of individual grids was merged into group

grid scores. While individual grids were no larger than

20 x 17, the group grids included all constructs found in

the group and thus were grids of 20 x 81 for the engineer-

ing class, 20 x 77 for the education class, and 20 x 44

for the community group. It is from the group girds that

the intercorrelation matrix of construct correlations was

calculated to be placed in the SPSS factor analysis routine.

Ordinarily, one would provide raw group data to the

SPSS factor analysis package. This program, in turn, cal-

culates the correlation coefficients between all construct

pairs and constructs the correlation matrix (Nie et al.,

1970). In this study, however, the raw data were submitted

to several purely data processing steps to permit checking

of group response consistency prior to calculating correla-

tions coefficients. A second reason for computing correla-

tions coefficient matrices for input to the SPSS routine is

that this calculation, which needs to be done only once, is

the most expensive calculation in the routine. By inputing









the correlation matrices, a variety of alternative computer

runs could be achieved relatively inexpensively.

Statistical Procedures and Tests of Hypotheses

The data acquired from the respondent groups were

analyzed and statistics calculated to test the four hypotheses

stated in Chapter I on page 12. It was anticipated that dif-

ferent groups would focus on different aspects of energy

concerns. One indication of such a difference is the range

of element responses elicited from a single group. A hypoth-

esis which can be easily tested is that three groups would

each contribute the same categorical elements with the same

relative frequency. The element frequency distributions

were printed out by the computer program and appear in

Chapter IV. Note that this hypothesis can be tested by super-

position of normalized frequency distributions.

It is not reasonable to believe that reference to a

particular element would be a randomly distributed variable

in the population as a whole. On the contrary, Kelly's

theory suggests that the element list an individual creates

is very much the result of their way of construing reality.

This individual set of characteristics is in turn a product

of a person's experience and learning. Consequently, it is

expected that common citation of elements should occur in a

group with some commonalities of experience and training.

It would thus be inappropriate to attempt to compute the

usual parametric test statistics and confidence intervals.









The data reported in Chapter IV do, in fact, support the

above prediction as will be described in detail later.

The number of constructs generated by individuals in a

group is a measure of the dimensional complexity of thought

in that group. The more constructs reported, the more ways

the respondent has of organizing perceptions. Consequently

one would wish to examine group differences as to the num-

bers of constructs generated by their members.

A convenient testable hypothesis is that the group

means are equal to each other. The implication of this

hypothesis is that its acceptance suggests all three groups

are representatives of a single population. This hypothesis

was tested by a one-way analysis of variance and the results

discussed in Chapter IV.

The underlying assumption of repertory grid analysis

is that the pattern of element-construct associations re-

flects the structure of the respondent's perceptions. An

individual who sees the world as a series of relatively

isolated events would be expected to have few positive

association responses. Thus the grid of such an individual

would be expected to show large numbers of (-) responses

and few (+'s). Conversely an individual who perceived a

great deal of interconnectedness would be expected to

report many positive associations. Such a person's grid

response should reflect a preponderance of (+) answers.

As a hypothesis susceptible to testing, it was pro-

posed that the percent of (+) responses of individuals in









each group were equal. This hypothesis was tested by a

one-way analysis of variance and the results are presented

and discussed in Chapter IV.

If it is desirable to assess the degree of interre-

latedness of perceptions of individuals, it is difficult to

interpret the structure of highly associated perceptions.

Interpretation of the structure of group perceptions becomes

almost impossible from simple inspection of grids. Fortu-

nately, methodology exists which can combine groups of

variables into composite variables called factors which

permit simplification of structural complexity. This method-

ology is termed factor analysis.

Factor analysis, as used here, proceeds through three

stages: computation of correlation matrix, interactive

estimation of communalities, and extraction and rotation

factors. Each of these steps and their associated analyti-

cal options will be discussed in turn.

In the simplest form, factor analysis attempts to

clarify structure in a set of data by creating linear

combinations of the variables involved. The proportion

of the group variance accounted for by these combinations

can be computed and, if the linear combinations are uncor-

related orthogonall), their relative significance is esti-

mated. The SPSS factor analysis routine operates on a

correlation matrix to find the single linear combination of

the variables present which maximally accounts for the total

group variance. This is called the first factor. A second









combination is then sought which will meet two criteria:

it must be orthogonal to the first factor, and it must

account for the maximum amount of the remaining variance.

A third factor is then generated with the same criteria,

i.e., orthogonal to the first two factors and accounting

for the maximum amount of remaining variance. This process

continues until there are as many factors as there are

variables.

The process described above suffers with regard to

interpretability of the factors derived. The correlation

matrix is generated by calculating the Pearson product

moment intercorrelation between the pooled rows of construct

pair responses. This generates a square matrix of construct

by construct correlations with l's on the diagonal. If we

are looking for structure in the matrix, we can focus on

the significance of the correlations. The diagonal row of

unit correlations reflects the fact that each construct has

reported for it both variance common to the whole group of

responses, and unique variance due to its own nature

(Harris, 1975). It is the former variance in which we are

interested. This quality is called the communality of the

variable. If accurate estimates of communality could be

used to adjust the matrix, then the factor scores would be

directly interpretable as reflecting structure.

The program used in this study accomplishes this com-

munality estimation by an iterative process (Nie et al.,

1970). On the first pass, the l's are replaced by the









single highest correlation in the column and the values of

all other correlations adjusted. Then the diagonal values

are replaced with the highest correlation in each adjusted

column. This process is repeated until there is no change

in the communalities between two passes.

Factors are then generated by the program until they

no longer explain a significant amount of variance. In the

option selected for this study, only factors with eigenvalues

of one or greater were printed. An eigenvalue of one is

equivalent to the normalized variance of one of the original

variables. Thus each factor printed is at least as useful

in accounting for group variance as any one of the original

variables.

The interpretation of the factors can proceed at this

point, but visualization of the significance of the factor

loading scores is difficult. A rotation algorithm is pro-

vided by the program to produce factors with loadings as

close to 0 or 1 as possible for each variable. The program

performs rotations of pairs of factors, retaining orthogo-

nality. If the rotation results in one or both of the

factor axes becoming colinear with a structural dimension

of the data, the projection of the data points on the other

axis becomes 0 as does the factor loading score. This is

easily interpretable; the colinear axis has no influence on

the variance in the data.

The program performs rotations maintaining orthogo-

nality and maximizing the numbers of O's and l's in the






62


factor loadings, thus generating more easily interpretable

factors. Further, it prints the distribution of data

points along each pair of factor axes. This information

is discussed in detail in Chapter IV and program output is

found in Appendix E.















CHAPTER IV
RESULTS


This study was designed to utilize data from three

groups of college students to demonstrate the use of reper-

tory grid methods in mapping values and perceptions concern-

ing energy issues. Two of the groups--education majors and

mechanical engineering majors--were relatively homogeneous

with regard to age, sex, and educational background. The

third group was made up of individuals taking a course by

newspaper on energy offered by Santa Fe Community College

and of persons taking a course in radio-television repair.

This last group was quite heterogeneous in composition

(for more detail, see Chapter III).

Repertory grids were elicited, constructs and elements

categorized, and individual responses merged into group

data. Pearson product moment correlations were calculated

between constructs and the intercorrelation matrices sub-

jected to factor analysis by means of the SPSS subprogram

FACTOR. Average number of constructs, means and standard

deviations, and one-way analysis of variance were performed

for each group using the SPSS subprogram ONEWAY. This pro-

gram was also used to compare number of positive associations

per construct for each group. Frequency of citation was

calculated for each construct and element for each group.









The products of the computations described above were

utilized in testing the following hypotheses:

1. The same elements and constructs are cited fre-

quently by different groups.

2. Groups generate the same number of constructs per

response.

3. Groups generate the same number of positive associ-

ations between elements and constructs per response.

4. Groups generate the same factors.

Each of these hypotheses will be analyzed, tested, and

discussed in turn.

Frequency of Element and Construct Citation

As might be expected, not all elements and constructs

occurred with similar frequencies in grid responses. In

fact, a few elements and constructs were very frequently

cited by members of a particular group, others were cited

much less frequently, and most were cited rarely. Tables 2

and 3 show the ten most frequently occurring elements and

constructs for the three groups. It is clear from the

tables that there is substantial agreement among the groups

on the importance of certain elements. Five elements--numbers

34, 41, 70, 80, and 81--appear in all five groups' listings.

A sixth--number 30--occurs in the top ten for the engineer-

ing and education groups.

It is interesting to note that these elements vary

widely in content. Elements 30, 34, and 41 relate to social

and political dimensions of the energy problem, while elements









Table 2

The Most Commonly Cited Constructs by Groups


Number
Construct Cited Meaning
Engineers

18 59 Conventional Technology/Alternate
Technology
87 44 Resources/Population Growth
44 43 Conservation/Public Attitudes
74 42 Education/Resources
10 36 Economics Cost/Public Needs
106 35 Politics/Industry
9 34 Economics Cost/Conservation
45 30 Conservation/Consumption
60 29 Attitudes/Planning
74 28 Pressure for/against International
Welfare

Teachers

9 104 Economics Cost/Conservation
18 75 Conventional Technology/Alternate
Technology
31 72 Politics/Resources
67 64 Individual Behavior/Public Politics
7 49 Economics Cost/Available Resources
28 48 Conservation/Politics
75 47 Education/Resources
44 43 Conservation/Public Attitudes
11 34 Economics Cost/Policies
5 34 Economics Lifestyle/Resource Limit

Communi-ty Groups

60 28 Attitudes/Planning
48 23 Conservation/Regulation
18 23 Conventional Technology/Alternate
Technology
37 21 Technology-practical/unrealistic
115 17 Government/Population
45 16 Conservation/Consumption
81 16 Energy production/Consumption
65 14 Credibility of Information/Attitudes
80 13 Energy Needs/Luxuries
2 12 Economic Factors/Technological Forces









Table 3

The Most Commonly Cited Elements by Groups


Number
Element Cited Meaning
Engineers

80 83 Economics
34 72 Politics
81 63 Industry
100 62 Efficiency of Current
Technology
70 59 Resources
6 58 "Soft" Technology
43 55 Credibility of Information
30 54 Government Regulation
41 51 Social Attitudes
50 50 Conservation


Teachers

70 119 Resources
80 112 Economics
34 111 Politics
51 89 Transportation
69 78 Consumption
41 76 Social Attitudes
81 75 Industry
42 68 Lifestyle Expectancies
74 65 New Types of Fuel
30 65 Government Regulation

Community Groups

34 31 Politics
68 28 Pollution
80 26 Economics
70 26 Resources
75 24 Safety
69 22 Consumption
73 19 Import/Export
81 17 Industry
51 17 Social Attitudes
74 15 New Types fuel










70, 80, and 81 are associated with resources, economics,

and industry. Surprisingly, none of the groups was strongly

aware of technical factors such as energy efficiency, emerg-

ing technology such as fusion, or radical lifestyle change

options as determined by frequency of citation.

Construct citations can be interpreted as indicative

of group values. Since constructs are both dichotomous and

polar, there is a "preferred" pole implicit as well as an

"undesirable" emergent pole. Recall the example of Chapter

II "approved by the dissertation chairman." The preferred

pole is clearly the approved pole, while the emergent pole

is the not approved pole. The data in Table 2 reveal that

only one construct, number 18 (conventional vs. alternate

technology), appears among those cited most frequently by

all three groups. One interpretation of the combined fre-

quencies is that while members of all three groups perceive

similar factors as relevant to the energy problem, they

place distinctly different constructions on those factors.

The implication of this is that the groups hold different

values and priorities regarding this problem.

Numbers of Constructs Elicited

Since each respondent provided the same number of ele-

ments (20), it is possible to compare the numbers of con-

structs produced from the element sets and attach significance

to the result. The number of constructs produced is indica-

tive of the complexity which the respondent perceives in the

area under study. An individual showing few constructs thus









can be said to have a relatively simplistic view while an

individual who responds with many constructs either views

the area as highly complex or perceives a variety of unre-

lated bits and pieces of observations. This ambiguity will

be resolved below.

Table 4 presents the mean numbers of constructs for

the three groups along with ranges and standard deviations.

In this table are also found the analysis of variance data

for numbers of constructs. Note that the F value is signif-

icant at the 0.0000 level. Thus we can reject the null

hypothesis that all group means are equal with less than a

0.01 percent chance of error. To determine whether or not

all pairs of means are significantly different, a least

significant difference follow-up test was performed. This

particular follow-up was selected because of its exactness

for uneven cell sizes. Although it suffers from multiple

levels of criterion variables, this particular application

with only three levels of group membership is immune from

that weakness. The table reveals that all pairs of means

are unequal at the 95 percent level of significance. It

is thus possible to state with confidence that insofar as

number of constructs are concerned, the three groups

examined are significantly different.

Degree of Construct Association

It was noted above that a relatively large number of

constructs could be interpreted in two ways. One means of

resolving this ambiguity is to achieve a measure of the









Table 4

ANOVA Table for Number of Constructs


Source D.F. S.S. M.S. F F Prob.

Between Groups 2 180.6951 90.3475 11.462 0.0000

Within Groups 71 559.6424 7.8823

Total 73 740.3374


Group N Mean S.D. S.E. Min. Max. 95% C.I. for Mean

Education 32 8.5000 3.1315 0.5536 4.0 16.0 7.3710 to 9.6290

Engineering 28 6.3571 2.7650 0.5225 3.0 17.0 5.2850 to 7.4293

Community 14 4.3571 1.9457 0.5200 3.0 9.0 3.2337 to 5.4805

Total 74 6.9054 3.1846 0.3702 3.0 17.0 6.1676 to 7.6432


Multiple Range Test
(*) Denotes pairs of
Mean Group

4.3571 C

6.3571 E

8.5000 T


- LSD At the 0.050 Level Ranges: 2.82 2.82
groups significantly different at the 0.050 level
C E T


*

*


C = Community

E = Engineering majors

T = Education majors









"connectedness" of the respondents' perceptions. It will

be recalled that each construct was scored against all ele-

ments for association. For strict consistency, only three

associations are required as a minimum for each construct.

Positive associations in excess of this minimum number are

indicative of a perception of connectedness by the respon-

dent. It is this feature of the grid that permits calcula-

tion of the construct intercorrelations. An alternative

measure is to determine the average number of positive

associations per construct for each individual. To the

extent that persons report positive associations, they indi-

cate the degree of interrelationships they perceive. This

degree of interrelationship perceived is useful in designing

learning models. A more complete discussion will follow in

Chapter V.

Table 5 also reflects results of the one-way analysis

of variance for this parameter. The F ratio is indicative

of significance at a level of 0.0030. The null hypothesis

that the group means are equal can thus be rejected with

only a 0.3 percent probability of error. The LSD follow-up

tests show that all pairs of means are different at a con-

fidence level of 95 percent. From the information displayed

in Table 4, we can then conclude that (1) all groups show

a degree of interconnectedness in excess of the minimum

required for definition of constructs, and (2) the groups

are all significantly different with regard to the degree









Table 5

ANOVA Table for Associations per Construct


Source D.F. S.S. M.S. F F Prob.

Between Groups 2 326.6074 163.3037 6.318 0.8030

Within Groups 71 1836.1949 25.8462

Total 73 2162.8023


Group Count Mean S.D. S.E. Min. Max. 95% C.I. for Mean

Education 32 10.0470 5.3352 0.9431 1.7000 27.5002 8.1234 to 11.9705

Engineering 28 7.3679 5'.7138 1.0798 1.2000 30.7002 5.1524 to 9.5835

Community 14 4.4001 2.3406 0.6256 0.6000 6.9001 3.0487 to 5.7515

Total 74 7.9650 5.4417 0.6326 0.6000 30.7002 6.7042 to 9.2257


Multiple Range Test
(*) Denotes pairs of
Mean Group

4.4001 C
7.3679 E
10.0470 T


- LSD At the 0.050 Level Ranges: 2.82
groups significantly different at the 0.050 level
C E T


2.82


C = Community
E = Engineering majors
T = Education majors


* *









of interconnectedness of perceptions demonstrated. The impli-

cations of this observation for curriculum development are

discussed in Chapter V.

Results of Factor Analyses of Grids

Table 6 lists the unrotated principal components (fac-

tors) for the engineering student group, the eigenvalue

associated with each factor, the percent of the grid value

accounted for by that factor, and the cumulative variance

accounted for. Because the group grid is structured, only

41 factors are required to account for 100 percent of the

variance. Had there been no structure, all 81 potential

factors would have been required. Note that the first fac-

tor accounts for far more variance than any other. Tables 7

and 8 display the same information for the education student

and community groups, respectively. Again, perceptual struc-

ture is indicated by much of the variance being accounted

for by relatively few factors. The amount of variance

accounted for by the first factor can be interpreted as a

measure of centrality of thought, that is, the degree to

which a single dimensionality is perceived to structure the

perceptual field. Note that on this criterion engineering

students and community members score similarly while educa-

tion students are significantly different.

Tables 9, 10, and 11 display the factor structures for

the first ten factors for each group. Each row of these

tables can be thought of as the row of regression coefficients

for predicting the score of one of the variables (constructs)









TABLE 6
SUMMARY OF FACTOR STRUCTURE FUR ENGIHEERING STUDENT CROUP
C"JSTRJCT ESTIMATED FACTOR EIENVALUE PERCENT OF CUMULATIVE
CDMUNALITY URIANCE PERCENT
C002 0.61552 1 17.492 21.9 21.9
C004 0.64939 2 7.57143 9.5 31.3
C005 0.82320 3 6.65219 8.3 39.6
C006 0.50018 4 5.06291 6.3 46.0
C007 0.83722 5 4.34184 5.4 51.4
CO08 0.86618 6 3.65097 4.6 56.0
C009 0.47453 7 3.37728 4.2 60.2
C010 0.9331 8 3.23098 4.0 64.2
COil 0.67399 9 2.73076 3.4 67.6
C012 0.74915 10 2.62047 3.3 70.9
C013 0.69756 11 2.25849 2.8 73.7
C014 0.86618 12 1.99421 2.5 76.2
C015 0.49723 13 1.92815 2.4 78.7
C017 0.79582 14 1.71036 2.1 80.8
C018 0.64405 15 1.58909 2.0 82.8
C019 0.85540 16 1.47138 1.8 84.6
C020 0.48224 17 1.36750 1.7 86,3
C021 0.48244 18 1.26276 1.6 87.9
C022 0.49246 19 1.09909 1.4 89.3
C023 0.52421 20 0.91551 1.2 90.5
C024 0.46245 21 0.84537 1.1 11.5
C025 0.74742 22 0.80370 1.0 92.5
C026 0.37538 23 0.72572 0.9 93.5
C027 0.67399 24 0.64331 0.8 94.3
C029 0.79741 25 0.d0929 0.8 95.0
C030 0.79512 26 0.54703 0.7 95.7
C031 0.70119 27 0.52099 0.7 96.4
C035 0.80277 28 0.43228 0.5 96.9
C040 0.63522 29 0.364d9 0.5 97.3
C041 0.82320 30 0.31827 0.4 97.7
CO42 0.79823 31 0.27886 0.3 98.1
CO4 0.79741 32 0.25853 0.3 98.4
C044 0.72941 33 0.21631 0.3 98.7
C045 0.72843 34 0.19824 0.2 98.9
C046 0.88561 35 0.18460 0.2 99,2
C047 0.66023 36 0.15452 0.2 99.4
C08 0.81409 37 0.13924 0.2 99.5
COIf 0.48244 38 0.13128 0.2 99.7
C051 0.57327 39 0.11085 0.1 99.8
COdO 0.70498 40 0.0538 0.1 99.9
C063 0.59395 41 0.04677 0.1 100.0
C064 0.70680 42 0.01982 0.0 100.0
COd5 0.70371 43 0.01039 0.0 100.0
C066 0.72018 44 0.00002 0.0 100.0
CON4 0.65523 45 0.00001 0.0 100.0
C070 0.38378 46 0.00001 0.0 100.0
C071 0.94281 47 0.00001 0.0 100.0
C072 0.70680 48 0.00001 0.0 100.0
C073 0.85540 49 0.00001 0.0 100.0
C074 0.62088 50 0.00000 0.0 100.0
C075 0.72941 51 0.00000 0.0 100.0
C076 0.93934 52 0.00000 0.0 100.0
C077 0.42591 53 0.00000 0.0 100.0
CC80 0.38378 54 0.00000 0.0 100.0
C081 0.67198 55 0.00000 0.0 100.0
C082 0.74742 56 0.00000 0.0 100.0
C084 0.83722 57 0.00000 0.0 100.0
CC07 0.71039 58 0.00000 0.0 100.0
C091 0.63296 55 0.00000 0.0 100.0









TABLE 6
SUnMARY OF FACTOR STRUCTURE FOR ENGINEERING STUDENT GRUP


CONSTRUCT ESTIMATED
CDIOUNALITY

CO92 0.57793
C093 0.54763
C094 0.49024
C095 0.75037
CO0 0.58287
C0o8 0.4268
C101 0.75879
C102 0.72018
C103 0.80277
C104 0.62088
ClOd 0.75037
C107 0.54763
C108 0.48596
C109 0.42591
C111 0.69756
C112 0.72862
C113 0.81409
C11i 0.61942
C115 0.71039
Ci d 0.61489
Cl17 0.94281


FACTOR EIGENUALUE PERCENT OF
VARIANCE


0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
0.0000
0.09000
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
-.00001
-.00001
-.00001
-.00001


CUMULATIVE
PERCENT


0.0 100.0
0.0 100.0
0.0 100.0
0.0 100.0
0.0 100.0
0.0 100.0
0.0 100.0
0.0 100.0
0.0 100.0
0.0 100.0
0.0 100.0
0.0 100.0
0.0 100.0
0.0 100.0
0.0 100.0
0.0 100.0
0.0 100.0
0.0 100.0
0.0 100.0
0.0 100.0
0.0 100.0








TABLE 7
SUMMARY OF FACTOR STRUCTURE FOR EDUCATION STUDENT GROUP


CONSTRUCT ESTIMATED
CODMUNALITY
C001 0.71852
C002 0.61368
C004 0.80156
C005 0.82333
C007 0.83039
C008 0.71852
C009 0.83039
C01o 0.72850
C01 0.82158
C01l 0.48052
C014 0.82031
C015 0.54364
Co01 0.70771
C017 0.49397
C018 0.72105
C019 0.74675
C021 0.70895
C022 0.67306
C023 0.68494
C024 0.43252
C025 0.51929
C026 0.89301
C027 0.86017
C028 0.73976
C029 0.46857
C030 0.68052
C031 0.77447
CO32 0.72231
C033 0.70388
C034 0.53491
C035 0.85391
C036 0.68268
C038 0.76571
Co04 0.58779
C042 0.45196
C013 0.61020
C044 0.69601
C045 0.74289
C046 0.67976
CO49 0.67976
C050 0.73377
C051 0.70895
CO0O 0.65499
CC62 0.49799
CC06 0.71967
CC06 0.49156
CC67 0.77447
C069 0.67949
C073 0.74616
C074 0.85178
C075 0.69922
C076 0.72805
C077 0.49229
C078 0.74504
C079 0.85391
C080 0.66051
CC81 0.63038
CC82 0.73976
C083 0.86017


FACTOR EIGE1iVALUE PERCENT OF
UARIANCE

1 22.2013 28.8
2 7.25164 9.4
3 5.82969 7.6
4 4.60265 6.0
5 4.04365 5.3
6 3.74168 4.9
7 2.99490 3.9
8 2.92703 3.8
9 2.55866 3.3
10 2.28323 3.0
11 2.05944 2.7
12 1.87529 2.4
13 1.59244 2.1
14 1.44934 1.9
15 1.39991 1.8
16 1.28586 1.7
17 1.14581 1.5
18 1.04233 1.4
19 0.97265 1.3
20 0.71313 0.9
21 0.66515 0.9
22 0.58672 0.8
23 0.53504 0.7
24 0.48325 0.6
25 0.40610 0.5
26 0.38434 0.5
27 0.36624 0.5
28 0.29558 0.4
29 0.25154 0.3
30 0.22411 0.3
31 0.17917 0.2
32 0.16439 0.2
33 0.14679 0.2
34 0.11022 0.1
35 0.09746 0.1
36 0.08717 0.1
37 0.03317 0.0
38 0.01244 0.0
39 0.00005 0.0
40 0.00001 0.0
41 0.00001 0.0
42 0.00001 0.0
43 0.00000 0.0
44 0.00000 0.0
45 0.00000 0.0
46 0.00000 0.0
47 0.00000 0.0
48 0.00000 0.0
49 0.00000 0.0
50 0.00000 0.0
51 0.00000 0.0
52 0.00000 0.0
53 0.00000 0.0
54 0.00000 0.0
55 0.00000 0.0
56 0.00000 0.0
57 0.00000 0.0
58 0.00000 0.0
59 0.00000 0.0


CUMULATIVE
PERCENT
28.8
38.3
45.8
51.8
57.1
61.9
65.8
69.6
72.9
75.9
78.6
81.0
83.1
85.0
86.8
88.4
89.9
91.3
92.5
93.5
94.3
95.1
95.8
96.4
96.9
97.4
97.9
98.3
98.6
98.9
99.2
99.4
99.6
99.7
99.8
99.9
100.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0









TARLE 7
SUMMARY OF FACTOR STRUCTURE FOR EDUCATION SrUDEHT CROUP


CONSTIRCT ESTIMATED
COiMUNALITY

C084 0.82333
C087 0.49405
C088 0.41020
CO 0.49400
C091. 0.72281
COT9 0.80766
C094 0.82158
C094 0.82619
C097 0.65928
C098 0.82496
C099 0.76787
C101 0.dl0o5
C102 0.55699
C103 0.82619
C104 0.89301
C15 0.55554
C106 0.45196
C110 0.8650q


FACTOR E[GtEUALUE FERCEHT OF
UARIANCE


CUMULATIUE
PERCENT


40 0.00000 0.0 100.0
41 0.00000 0.0 100.0
62 0.00000 0.0 100.0
63 0.00000 0.0 100.0
64 0.00000 0.0 100.0
65 0.00000 0.0 100.0
6d 0.00000 0.0 100.0
67 0.0000 0.0 100.0
64 0.00000 0.0 100.0
69 0.00000 0.0 100.0
70 0.00000 0.0 100.0
71 0.00000 0.0 100.0
72 0.00000 0.0 100.0
73 -.00001 0.0 100.0
74 -.00001 0.0 100.0
75 -.00041 0.0 100.0
76 -.00001 0.0 100.0
77 -.00002 0.0 100.0










TARLE 8
SUMMARY OF FACTOR STRUCTURE FOR COMfUNITY GROUP
CONSTRUCT ESTIMATED FACTOR EICENUALUE PERCENT OF CUIULATIVE
CMUNALITY UARIANCE PERCENT
C002 1.00000 1 9.88800 22.5 22.5
C007 1.00000 2 4.62006 10.5 33.0
C009 1.00000 3 3.47582 7.9 40.9
C010 1.00000 4 3.25159 7.4 48.3
COll 1.00000 5 2.82219 6.4 54.7
C012 1.00000 6 2.63603 6.0 60.7
C013 1.00000 7 2.41758 5.5 66.2
C015 1.00000 8 1.92819 4.4 70.6
CO01 1.00000 9 1.65041 3.8 74.3
C017 1.00000 10 1.39879 3.2 77.5
C018 1.00000 11 1.27767 2.9 80.4
C021 1.00000 12 1.23360 2.8 83.2
C022 1.00000 13 1.07283 2.4 85.6
C023 1.00000 14 0.94358 2.1 87.8
C025 1.00000 15 0.89499 2.0 89.8
C027 1.00000 16 0.83737 1.9 91.7
C02q 1.00000 17 0.55595 1.3 93.0
C030 1.00000 18 0.52031 1.2 94.2
CO31 1.00000 14 0.46591 1.1 95.2
CO33 1.00000 20 0.38131 0.9 96.1
C03' 1.00000 21 0.31750 0.7 96.8
C035 1.00000 22 0.30508 0.7 97.5
C037 1.00000 23 0.25142 0.6 98.1
C040 1.00000 24 0.11222 0.4 98.5
C045 1.00000 25 0.17063 0.4 98.9
C048 1.00000 26 0.12654 0.3 99.2
C050 1.00000 27 0.10725 0.2 99.4
C040 1.00000 28 0.08092 0.2 99.6
CC65 1.00000 29 0.07204 0.2 99.8
Co66 1.00000 30 0.03859 0.1 99.9
C067 1.00000 31 0.03279 0.1 99.9
C075 1.00000 32 0.01545 0.0 100.0
C080 1.00000 33 0.01130 0.0 100.0
C081 1.00000 34 0.00311 0.0 100.0
C082 1.00000 35 0.00000 0.0 100.0
C083 1.00000 36 0.00000 0.0 100.0
C088 1.00000 37 0.00000 0.0 100.0
C092 1.00000 38 0.00000 0.0 100.0
C098 1.00000 39 0.00000 0.0 100.0
C099 1.00000 40 0.00000 0.0 100.0
C106 1.00000 41 0.00000 0.0 100.0
C115 1.00000 42 0.00000 0.0 100.0
C113 1.00000 43 0.00000 0.0 100.0
C119 1.00000 44 -.00001 0.0 100.0









TABLE 9
UNROTATED PRINCIF L COMHPNENTS FOR ENGINEERING STUDENTS
1 2 3 4 5 6 7 8 9 10


COSTRUCT
C002
C004
C005
COO

C008
009
C010
coil
C012
C013
C014
015
C017
Co018
C019
C020
C021
C022
C023
co24
C025
C026
C027
C029
C030
C031
C035
COIM
C040
C042
CO43
044
C045
C046
C047
C048
C049
C051
C060

Codi
COd5
Co46
C069
C070
C071
C072
C073
C074
COTS
C075
C076
C077
C081
COB1
C082
C087
C091
C092


0.475 0.496 -0.192 0.046 -0.071 0.156 -0.224
0.509 -0.007 -0.081 -0.460 0.288 0.083 0.085
0.425 0.293 -0.407 -0.327 0.262 -0.198 -0.158
0.325 0.097 -0.253 -0.035 0.121 0.187 0.039
0.578 0.370 0.026 -0.326 -0.030 -0.075 -0.008
0.560 0.259 0.536 -0.066 0.008 -0.281 -0.181
0.648 0.440 -0.119 -0.153 -0.304 -0.11 -0.045
0.715 -0.278 0.213 -0.462 -0.154 -0.001 -0.089
0.340 -0.216 -0.261 0.351 0.361 -0.151 -0.021
0.726 -0.306 -0.143 0.155 -0.197 0.297 -0.304
0.456 0.254 -0.243 0.340 0.290 0.089 -0.183
0.484 0.142 0.528 0.185 -0.088 -0.225 -0.174
0.241 0.461 0.092 -0.270 -0.213 -0.011 0.043
0.400 0.513 -0.165 0.139 -0.148 0.427 -0.315
0.622 0.186 0.330 0.367 0.084 0.207 -0.020
0.177 0.510 -4.024 0.341 -0.281 0.292 -0.229
0.014 -0.282 -0.O99 0.042 -0.080 0.102 -0.132
0.288 0.264 0.083 -0.334 -0.115 0.114 0.156
0.101 0.146 0.374 0.022 0.461 -0.030 0.105
0.444 -0.073 0.008 0.262 0.007 0.166 0.244
0.125 -0.119 0.296 0.379 0.313 -0.246 -0.096
0.285 0.371 0.442 0.206 -0.106 0.021 -0.012
-0.051 0.144 0.224 0.130 0.215 -0.091 0.050
0.245 -0.243 -0.193 0.372 0.260 -0.236 0.262
0.548 -0.139 -0.179 0.171 -0.355 -0.086 0.393
0.639 0.243 0.471 -0.099 -0.216 -0.224 -0.061
0.643 0.184 -0.173 -0.460 0.028 -0.086 0.134
0.368 0.121 -0.378 0.151 0.095 -0.390 -0.048
0.343 0.455 -0.337 0.323 -0.276 -0.123 -0.273
0.536 0.052 -0.327 -0.552 0.148 -0.098 -0.118
0.592 -0.371 0.442 0.176 -0.164 -0.154 -0.147
0.441 -0.171 -0.225 0.293 -0.205 -0.102 0.397
0.698 0.038 0.074 -0.043 -0.162 -0.163 0.161
0.678 0.025 -0.134 0.353 -0.120 -0.136 0.166
0.438 -0.610 -0.009 -0.174 -0.365 0.136 -0.175
0.324 -0.074 -0.312 0.243 -0.232 -0.229 0.284
0.450 -0.060 -0.336 0.345 0.009 -0.047 0.294
0.404 0.171 0.191 -0.113 -0.080 -0.020 0.246
0.104 0.042 0.646 0.225 0.408 0.021 0.042
0.506 0.160 0.592 -0.023 0.164 -0.124 -0.166
0.325 0.253 0.414 -0.286 -0.024 -0.251 0.140
0.062 -0.341 0.079 -0.187 0.015 0.494 0.110
0.393 -0.132 -0.310 -0.355 0.249 -0.188 -0.213
0.700 -0.277 0.072 0.222 0.022 0.067 0.011
0.415 -0.547 0.402 0.223 0.105 0.004 0.016
-0.041 -0.038 -0.049 -0.028 0.157 -0.088 0.058
0.484 -0.729 0.087 -0.017 -0.212 0.057 -0.274
0.145 -0.323 0.070 -0.374 -0.042 0.260 0.066
0.283 0.557 -0.187 0.254 -0.174 0.392 -0.175
0.630 0.057 0.359 -0.035 0.347 0.239 0.059
0.833 -0.159 -0.017 -0.024 -0.129 -0.042 -0.086
0.511 -0.705 0.028 0.008 -0.116 0.060 -0.243
0.154 -0.180 0.018 0.210 -0.433 -0.331 0.041
-0.115 -0.050 0.02q -0.047 0.391 -0.229 -0.107
0.493 0.078 0.359 -0.254 -0.261 -0.102 0.225
0.291 0.491 0.354 0.153 -0.162 0.158 -0.211
0.564 0.289 0.036 -0.277 -0.003 0.015 0.252
0.721 -0.121 -0.112 -0.095 0.056 -0.232 0.060
0.347 0.279 0.216 -0.402 0.146 -0.181 -0.074
0.454 0.292 0.416 -0.120 -0.030 0.237 -0.054


-0.176 -0.104 -0.024
0.113 0.101 0.116
0.150 0.041 -0.231
-0.242 -0.355 -0.272
-0.312 -0.068 -0.196
0.051 -0.217 -0.003
0.119 0.120 -0.090
0.083 0.114 0.053
-0.396 0.011 0.123
0.007 -0.045 0.022
0.205 -0.224 0.058
0.215 -0.108 0.029
-0.151 0.029 0.192
0.043 0.119 0.008
-0.073 0.033 -0.119
0.062 0.092 0.359
-0.019 -0.182 0.071
0.134 -0.012 0.210
-0.098 -0.080 0.148
-0.242 0.018 -0.238
-0.293 0.216 -0.089
-0.049 -0.017 -0.024
0.216 -0.007 0.034
-0.360 0.218 0.107
0.355 -0.124 0.037
0.195 -0.062 -0.107
-0.339 0.153 0.152
0.031 -0.132 -0.161
0.078 0.059 -0.140
0.263 0.252 -0.176
0.175 -0.080 0.107
0.218 -4.368 -0.116
0.109 -0.006 0.119
0.102 -0.216 -0.035
0.083 -4.250 0.051
0.347 0.240 0.119
-0.027 0.266 -0.144
0.101 0.064 0.392
0.028 0.046 -0.082
0.176 0.109 -0.120
0.167 -0.046 0.081
0.128 0.183 -0.185
0.212 0.270 -0.089
-0.286 0.138 0.377
0.134 0.131 -0.030
-0.120 0.112 0.371
-0.066 -0.080 0.082
0.208 0.355 -0.076
0.160 0.090 0.185
0.032 -0.047 0.154
-0.130 -0.062 0.141
-0.147 0.021 -0.040
0.225 -0.081 -0.194
0.077 -0.056 0.241
-0.065 0.330 0.034
-0.174 0.044 0.066
-0.328 -0.224 -0.187
-0.017 -0.145 -0.132
-0.227 -0.230 -0.048
-0.162 -0.063 -0.164









TABLE 9
UNROTATED FPRICIPAL CIMPXMENTS FUR ENGINEERING STUDENTS
CONSTRUCT 1 2 3 4 5 6 7 8 9 10


0.350 -0.190 0.327
0.262 -0.177 0.284
0.390 -0.328 -0.394
0.294 0.076 -0.473
0.298 0.105 -0.0O8
0.579 -0.523 0.039
0.621 -0.300 -0.24q
0.494 -0.074 -0.573
0.440 0.142 -0.207
0.548 -0.131 -0.267
0.16 -0.074 0.318
0.304 0.072 0.260
0.054 -0.169 0.10
0.334 0.273 -0.148
0.564 0.380 -0.157
0.647 0.177 -0.270
0.560 0.192 -0.252
0.707 -0.248 0.064
0.407 0.213 -0.320
0.543 -0.729 0.103


0.292 0.352 0.390 0.161
0.148 0.337 -0.006 0.155
-0.049 0.048 0.434 0.186
0.028 0.325 -0.138 -0.238
-0.218 0.334 -0.010 0.406
-0.269 0.135 -0.046 -0.213
0.154 0.228 -0.046 -0.294
0.055 0.166 -0.20 -0.108
0.075 0.358 0.211 -0.115
-0.134 0.311 0.316 0.198
0.015 0.224 0.334 -0.006
0.288 0.198 0.091 0.438
0.297 -0.144 -0.325 0.248
0.150 0.508 -0.204 -0.335
0.320 0.019 0.007 -0.073
0.293 -0.218 -0.115 0.095
0.067 -0.045 0.431 0.254
-0.222 -0.103 -0.033 0.097
-0.237 -0.061 0.055 0.364
-0.013 -0.202 0.029 -0.251


0.106 -0.093
0.020 0.261
0.099 -O.lb1
0.194. 0.138
0.134 -0.021
0.008 0.114
-0.344 0.037
0.075 -0.076
0.112 -0.321
0.322 -0.143
0.221 0.336
0.196 -0.270
-0.086 0.205
0.344 -0.114
-0.152 .3719
-0.027 0.400
-0.004 0.143
-0.351 -0.210
-0.364 -0.087
-0.043 -0.043


-0.154
-0.251
0.028
0.323
0.322
0.143
0.174
-0.048
-0.048
-0.137
-0.399
0.293
-0.176
-0.026
-0.094
-0.092
0.0998
-0.132
-0.084
0.081








TABLE 10

UNROTATED PRINCIPAL CifPONENTS FOR EDUCATION STUDENTS


CONSTRUCT
C001
C004
C005
C007
C008


C013
C001

C013
C014
C015
C014
C017
018
C019
C021
C022
C023
C024
C025
C026
C02?
C029
C029
C031
C032
C033
C034
C035
C036
C039
C040
C042
C043
C044
C045
C046
C044
C051
C060
C062
C06O
CO66

C073
C074
CO75
C076
C077
C078
C079
COSO
CC080
C082
C083
C034


2 3 4 5
-0.215 -0.304 0.333 -0.139
0.145 -0.340 0.139 -0.233
-0.389 0.154 -0.242 0.180
0.489 -0.299 0.246 0.227
-0.116 -0.147 -0.252 -0.063
0.040 -0.396 0.172 -0.209
0.007 -0.316 0.036 0.132
-0.365 -0.066 0.133 -0.117
0.095 -0.136 0.026 0.016
0.026 -0.262 0.181 0.109
0.572 0.347 -0.022 -0.349
-0.231 -0.420 -0.218 -0.045
0.310 -0.179 -0.080 -0.070
0.101 0.436 -0.01 0.296
0.235 0.031 -0.164 0.217
0.363 -0.270 0.100 0.436
-0.136 0.135 0.517 -0.014
0.357 0.484 0.222 0.357
0.314 0.268 0.024 0.604
0.042 -0.083 0.020 0.234
0.059 0.033 -0.075 0.434
0.488 0.544 0.035 -0.250
-0.366 0.229 -0.412 0.202
0.268 -0.1! -0.002 0.053
0.255 -0.503 -0.018 -0.104
0.253 -0.242 0.155 -0.042
-0.255 -0.032 -0.135 0.061
0.255 0.28, -0.083 0.781
-0.079 -0.069 -4.231 -0.037
-0.083 0.181 0.004 0.116
0.245 -0.011 -.13 -0.200
-0.159 0.108 -0.076 -0.278
-0.419 0.184 -0.133 -0.092
-0.004 -0.362 0.197 0.161
-0.155 -0.070 -0.290 0.168
-0.067 -0.134 0.115 -0.146
-0.257 0.291 0.652 -0.213
0.434 0.137 -0.095 -0.277
0.134 0.160 0.172 -0.276
0.427 -0.022 0.071 -0.154
0.343 -0.071 -0.193 -0.136
0.050 0.200 0.374 -0.060
-0.508 -0.021 0.442 -0.073
-0.213 -0.179 -0.313 -0.107
-0.340 -0.076 0.059 -0.017
0.225 0.144 0.404 0.119
-0.154 0.086 0.056 -0.107
-0.118 0.376 0.501 0.064
0.152 -0.094 -0.202 0.202
-0.474 0.312 -0.315 -0.004
-0.307 0.409 -0.079 -0.063
-0.137 -0.333 0.115 -0.119
-0.203 -0.263 0.056 0.112
-0.420 -0.1$9 0.227 -0.007
0.332 0.021 -0.355 -0.247
0.015 -0.334 0.342 0.402
-0.296 -0.318 -0.073 0.046
0.067 -0.551 -0.193 -0.079
-0.419 0.142 -0.588 0.025
0.312 -0.040 4.244 0.413


6 7 8 9 10
-0.080 -O.016 -0.025 0.29'7 0.0O0
0.315 0.203 -0.393 -0.133 -0.157
-0.087 0.026 -0.189 -0.009 -0.145
0.175 -0.006 -0.177 0.107 -0.043
0.159 0.172 -0.149 0.055 0.073
0.012 0.093 -0.105 0.037 -0.016
0.216 0.121 -0.062 -0.000 0.015
-0.178 0.081 -0.171 -0.115 -0.034
-0.097 0.012 0.112 -0.347 -0.023
-0.145 -0.229 -0.191 0.058 0.280
0.051 -0.036 -0.090 0.105 -0.112
-0.284 0.321 0.240 -0.155 0.026
-0.191 0.042 0.301 -0.090 -0.140
-0.300 0.171 -0.036 -0.227 0.257
-0.106 0.003 -0.070 0.100 -0.363
-0.046 0.186 0.193 0.046 0.014
-0.164 0.273 -0.263 -0.119 0.028
0.012 0.208 0.033 -0.107 -0.134
0.045 0.119 0.055 -0.174 0. 05
-0.093 -0.190 -0.003 0.234 -0.465
0.385 -0.358 -0.23 -0.192 0.120
-0.083 -0.019 -0.059 -".022 -0.330
-0.053 0.150 -0.115 0.196 0.031
-0.357 0.082 -0.012 0.004 0.040
0.020 0.117 -0.091 -0.166 0.010
0.088 -0.466 -0.161 0.087 0.314
-0.227 -0.103 0.292 0.059 0.008
-0.116 0.087 0.179 0.212 -0.Q25
0.010 -0.428 -0.074 -0.031 -0.000
-0.216 0.431 -0.253 v.030 0.11i;
0.518 0.241 0.140 O.Ii6 0.184
-0.210 0.053 0.108 0.023 0.042
-0.250 -0.133 0.019 -0.191 -0.013
-0.117 -0.171 0.254 -4.358 -0.207
0.421 -0.057 0.095 -0.259 0.190
0.456 -0.171 0.091 0.105 -0.177
0.111 0.271 0.00. -0.125 0.005
0.237 -0.269 -0.074 0.156 0.025
0.000 0.032 0.161 -0.105 0.299
-0.270 0.010 0.143 0.027 0.086
0.091 -0.198 -0.045 -0.M60 0.104
-0.084 0.037 -0.065 -n.228 0.254
0.013 -0.267 0.04 0..142 -0.105
0.279 0.523 0.159 0.040 0.080
-0.050 0.052 0.265 -).063 -0.091
-0.076 -0.178 0.061 0.210 0.231
0.087 0.073 0.249 -0.121 -0.185
0.184 -0.000 0.175 -;.131 0.222
0.236 0.240 -0.085 0.213 -8.024
-0.172 -0.121 -0.153 0.086 -0.017
0.041 0.077 0.202 -0.045 0.142
-0.095 -0.063 -0.091 .011t -0.198
0.583 0.038 0.134 -0.013 -0.131
-0.096 -0.130 O.C47 -0.171 -0.037
0.443 0.240 0.!62 0.033 0.131
0.084 -0.123 -0.330 0.081 -0.153
0.064 -0.320 0.161 -0.255 0.098
-0.256 0.188 0.248 0.166 0.059
0.011 -0.090 -0.059 0.121 -0.0il
0.095 0.228 -0.116 0.145 0.022








TABLE 10
URRUTrTED PRINCIPAL COPONENTS FOR EDJUCTION STUDENTS
CONSTRUCT 1 2 3 4 5 6 7 8 9 10


0.384 ..140 0.005 -0.021 -0.084
0.244 -0.114 0.224 0.420 -0.363
0.254 -.2389 0.035 -0.040 0.144
-0.106 -0.05C 0.438 -0.329 -0.152
0.359 -0.017 -0.291 -0.041 -0.018
-0.322 -0.250 -0.184 -0.155 0.060
0.230 0.205 0.011 0.142 -0.037
0.237 -0.078 -0.013 -0.152 -0.204
0.164 0.391 -0.076 0.114 0.137
0.230 -0.393 0.044 0.122 -0.163
0.121 -0.103 0.074 0.082 0.065
0.100 -0.281 -0.170 -0.226 -0.125
0.315 0.208 -0.031 0.171 0.14'1
0.524 0.111 -0.327 -0.123 -0.076
-0.194 0.093 -0.337 -0.041 -0.044
-0.084 -0.095 -0.109 -0.452 -0.062
0.410 -0.048 -0.165 0.028 -0.082


0.2`2 -o.145 0.226
0.31' .0093 0.054
-0.271 O.050 0.037
0.332 0.295 0.141
-0.026 0.061 -0.994
0.18 -0.226 -0.179
0.035 5.220 0.023
-0.193 0.329 -0.096
0.070 0.265 -0.194
-0.089 -0.287 -0.027
-0.615 -0.41* 0.049
-0.035 0.209 0.379
0.155 0.375 -0.013
-0.0CS -I.116 -0.150
-0.020 -..22 0.166
-0.149 .113 0.243
0.174 -.152 -0.349


0.154 0.029
0.422 -0.076
0.356 0.081
0.410 0.375
0.662 0.347
0.728 0.205
0.595 -0.512
0.563 -0.230
0.254 -0.585
0.513 -0.315
0.514 -0.083
0.526 0.088
0.407 -0.507
0.451 0.505
0.246 0.553
0.203 -0.216
0.351 0.625









TIfELE' II
UNROTATED PRINCTRFL COMtPOXNETS FOR COMMUNITY CROUP
CONSTRUCT 1 2 3 4 5 6 7 8 9 10
C002 0.572 -0.286 0.196 0.013 0.002 0.370 -0.221 -0.177 0.149 0.027
C007 0.442 -0.095 0.412 -0.410 0.205 -0.106 0.107 0.492 0.175 0.003
C004 -0.041 0.764 -0.124 -0.062 0.413 0.077 -0.131 0.016 0.162 -0.089
C010 -0.154 -0.071 0.079 -0.163 -0.047 -0.074 0.123 0.271 0.056 -0.292
COl0 0.633 -0.325 -0.136 0.308 -0.155 0.208 -0.266 0.217 0.155 0.008
C012 0.036 0.746 -0.121 -0.068 0.424 0.085 -0.160 0.068 0.178 -0.067
C013 0.672 0.339 -0.109 -0.244 -0.435 -0.167 -0.075 -0.205 0.131 -0.122
C015 0.342 -0.169 0.374 -0.263 -0.181 0.310 0.430 0.218 0.119 0.035
Col0 0.044 0.259 0.457 0.096 0.365 0.134 0.535 -0.011 -0.203 0.013
C017 0.194 0.268 0.310 0.218 -0.610 0.404 -0.137 -0.094 0.204 -0.186
C018 0.147 0.299 0.577 0.080 -0.041 0.399 -0.172 -0.337 0.010 0.333
C021 -0.193 -0.080 0.176 -0.252 -0.008 -0.187 -0.060 0.218 0.374 0.395
C022 0.672 0.339 -0.109 -0.244 -0.435 -0.167 -0.075 -0.205 0.131 -0.122
C023 0.273 0.217 0.676 0.344 -0.061 0.106 -0.049 0.160 -0.193 -0.202
C025 0.668 0.398 0.260 -0.164 -0.157 0.040 0.181 0.161 0.118 -0.026
C02? 0.479 -0.012 0.244 -0.208 -0.396 -0.116 -0.091 -0.227 -0.396 -0.178
C029 0.053 0.136 -0.079 -0.262 -0.280 -0.326 0.162 0.131 0.501 0.267
C030 0.387 0.006 0.483 -0.296 0.357 -0.034 0.039 0.387 0.053 0.100
C031 0.247 0.130 0.530 0.502 0.009 -0.047 -0.342 0.237 0.041 -0.296
C033 0.121 -0.193 -0.123 0.356 0.323 0.132 0.474 -0.337 0.266 -0.094
C034 0.680 -0.411 -0.135 -0.261 0.123 0.079 0.127 -0.025 -0.057 -0.146
C035 0.844 -0.077 0.004 0.058 0.211 -0.206 0.098 -0.003 -0.015 0.030
C037 0.365 0.604 -0.156 0.120 -0.032 -0.068 0.439 -0.094 -0.229 0.157
C040 0.095 -0.144 0.068 0.490 -0.044 -0.241 -0.371 0.080 -0.175 0.125
C045 0.500 -0.170 0.151 -0.088 0.252 -0.311 -0.630 -0.073 0.041 0.132
CO48 0.619 -0.010 -0.007 0.540 -0.008 -0.086 0.259 0.011 -0.185 0.300
CO'5 -0.145 -0.023 0.092 0.026 -0.056 0.047 -0.036 -0.283 -0.080 0.501
CC60 0.612 -0.320 0.1(3 -0.070 0.126 0.336 -0.036 -0.235 0.131 0.263
C065 0.620 -0.132 -0.340 0.486 -0.121 0.023 0.009 0.180 0.184 0.066
C066 -0.061 0.057 -0.516 -0.007 0.006 0.554 -0.075 0.424 -0.274 0.075
C047 0.428 0.348 -0.246 -0.079 -0.203 0.507 0.079 0.307 -0.247 0.120
C075 0.471 0.176 -0.288 -0.408 0.177 0.016 -0.271 0.033 -0.174 -0.100
C080 0.688 -0.378 0.114 -0.041 -0.163 0.282 -0.066 -0.029 0.001 -0.033
C081 -0.012 0.827 0.055 0.175 0.216 0.243 0.057 -0.167 0.173 -0.080
C082 0.731 -0.177 -0.362 0.186 0.162 -0.167 -0.051 0.069 0.084 -0.048
C083 0.677 0.058 0.156 0.203 0.261 -0.355 0.062 -0.028 -0.259 0.117
C088 0.603 0.184 -0.061 -0.467 -0.151 -0.304 0.207 -0.165 -0.243 -0.012
C092 0.599 0.531 -0.207 0.151 -0.240 -0.324 0.028 -0.028 0.123 0.089
C008 0.680 -0.411 -0.135 -0.261 0.123 0.079 0.127 -0.025 -0.051 -0.146
C099 0.197 0.328 -0.206 -0.359 0.403 0.291 -0.489 -0.119 -0.030 0.028
C10 0.532 -0.221 -0.053 -0.260 0.337 0.312 0.003 -0.256 -0.066 0.047
C115 0.551 0.109 -0.557 0.361 -0.071 0.174 0.070 0.287 0.117 0.097
C118 0.602 0.055 0.071 0.350 0.428 -0.347 0.005 -0.054 0.076 -0.149
C19 0.018 -0.330 -0.154 0.110 0.177 0.148 0.235 -0.309 0.383 -0.326









from the factor score. The square of each coefficient then

estimates the proportion of the construct score variance

accounted for by the factor score. Note that these tables

have been adjusted for communalities; i.e., the variance

accounted for is common to the entire grid structure, not

unique to a particular variable. Thus the factor structure

matrix displayed in the tables is a structural depiction of

the innercorrelated variance. Further, variables having a

relatively large factor score coefficient are relatively

important in determining the meaning of the factor.

To maximize the possibility of distinguishing important

factor dimensionalities, Varimax rotation was performed on

the factors. This transformation forces the coefficients

to approach zero or one, making it easier to distinguish

heavy from light loadings. The Varimax rotated factors are

displayed in Tables 12, 13, and 14. The most significant

variables were extracted by ranking the coefficients in

decreasing order of magnitude and retaining those which

accounted for the first 50 percent of the factor eigenvalue.

Note that for the first factors, ten or more variables were

found as identified by circled coefficient values in the

tables. For subsequent variables, only a few significant

variables are required to account for 50 percent of the

eigenvalue (thus 50 percent of the variance) of the factor.

This is consistent with the literature which reports that

first factors tend to be quite general, loading heavily on

many variables, while subsequent factors tend to be bipolar,









TABLE 12
UARTIMX RUTTED FACTOR MATRIX FUR ENGIHEERIRG STUDENT CROUP
CUCSTRUCT 1 2 3 4 5 6 7 8 9 10
C002 0.019 0.112 0.605 0.407 0.178 -0.010 0.008 0.184 -0.146 -0.048
C004 0.184 0.102 -0.010 0.298 0.469 -0.060 -0.054 0.089 0.247 0.425
C00l -0.093 0.042 0.118 0.332 0.769 0.060 -0.051 0.125 -0.004 -0.039
C006 0.072 -0.098 0.090 0.499 0.075 0.024 0.054 0.434 0.005 -0.144
C007 0.048 0.295 0.205 0.732 0.239 -0.008 0.016 -0.027 -0.030 0.005
C008 0.169 0.833 0.040 0.242 0.108 0.040 -0.047 0.049 -0.163 0.034
CO09 0.044 0.279 0.480 0.414 0.372 0.356 -0.149 -0.150 -0.015 0.051
C010 0.638 0.341 -0.005 0.317 0.296 0.094 -0.150 -0.191 0.257 0.236
Co01 0.240 -0.113 0.025 0.089 0.149 0.011 0.695 0.236 -0.184 0.067
C012 0.751 0.031 0.418 0.113 0.135 0.200 0.080 0.214 0.149 -0.051
C013 0.032 0.148 0.410 -0.037 0.313 0.135 0.140 0.570 -0.118 0.042
C014 0.198 0.775 0.165 -0.044 0.009 0.195 -0.004 0.032 -0.085 -0.001
C015 -0.092 0.197 0.306 0.392 0.011 -0.046 -0.174 -0.215 -0.149 0.253
C017 0.008 0.064 0.839 0.146 0.140 -0.035 -0.097 0.132 0.101 -0.042
C018 0.144 0.511 0.422 0.179 -0.092 0.122 0.319 0.248 0.237 -0.002
C019 -0.047 0.015 0.816 -0.126 -0.140 0.018 -0.074 0.010 -0.161 0.143
C020 0.315 -0.156 -0.038 -0.075 -0.070 -0.007 -0.052 0.131 -0.069 -0.062
C021 -0.001 0.169 0.171 0.255 0.079 0.050 -0.302 -0.051 0.062 0.393
C022 -0.144 0.396 -0.11U 0.045 -0.046 -0.259 0.230 0.211 0.007 0.273
C023 0.147 0.060 0.140 0.327 -0.137 0.256 0.371 0.185 0.253 -0.046
C024 0.041 0.323 -0.071 -0.045 -0.032 -0.106 0.644 -0.045 0.005 -0.141
C025 -0.073 0.549 0.317 0.161 -0.237 0.029 0.084 -0.014 -0.005 -0.005
C026 -0.222 0.284 -0.048 -0.206 0.014 -0.001 0.032 0.111 0.017 0.084
C027 0.092 -0.135 -0.081 0.054 0.040 0.192 0.75 0.014 -0.062 0.165
C029 0.256 0.047 0.094 0.819 0.042 0.807 -0.080 0.126 0.048 0.222
C030 0.180 0.770 0.171 .266 0.126 0.280 -0.154 -0.093 0.023 0.025
C031 0.187 0.050 0.163 0.685 0.370 -0.004 0.127 -0.175 -0.019 0.400
C035 -0.010 0.033 0.063 0.173 0.434 0.348 0.210 0.187 -0.280 -0.145
C040 -0.079 0.059 0.629 0.083 0.272 0.338 0.042 0.007 -0.235 -0.283
C041 0.134 0.033 0.040 0.303 0.839 0.065 -0.143 -0.057 0.265 0.119
C042 0.637 0.568 -0.003 -0.111 -0.032 0.286 0.085 0.012 0.030 0.055
C043 0.188 -0.007 -0.043 0.146 -0.043 0.751 0.025 0.370 -0.050 0.039
C044 0.289 0.350 0.158 0.274 0.178 0.420 0.034 -0.019 -0.004 0.292
C045 0.234 0.218 0.233 0.198 0.098 0.606 0.198 0.307 -0.106 0.058
C046 0.850 -0.028 -0.000 0.087 -0.011 0.207 -0.245 0.059 0.068 -0.025
C047 0.057 -0.092 0.148 -0.153 0.218 0.683 0.113 -0.117 0.012 0.217
C048 0.137 -0.059 0.267 0.213 0.241 0.574 0.500 0.093 0.216 0.090
C049 0.062 0.285 0.151 0.150 -0.013 0.194 -0.044 -0.083 0.003 0.571
C051 -4.107 0.616 -0.107 -0.136 -0.167 -0.183 0.288 0.169 0.253 0.046
C060 0.108 0.810 0.096 0.055 0.177 -0.022 0.065 -0.014 0.209 0.022
C0o3 -0.052 0.580 -0.071 0.216 0.106 0.103 -0.190 -0.129 -0.025 0.277
C064 0.227 -0.143 -0.064 0.002 -0.071 -0.060 -0.125 0.039 0.632 0.034
C065 0.212 -0.017 -0.043 0.070 0.760 0.018 0.028 -0.064 0.155 0.055
C066 0.605 0.154 0.257 0.140 -0.021 0.108 0.492 0.030 0.016 0.373
C069 0.512 0.374 -0.169 -0.208 -0.053 0.179 0.302 0.056 0.352 0.055
C070 -4.013 -0.108 -0.057 -0.101 0.018 -0.155 0.167 -0.094 -0.158 0.320
C071 0.937 0.052 -0.090 -0.011 0.004 0.101 0.073 -0.004 0.083 -0.053
C072 0.261 -0.064 -0.098 -0.003 0.157 -0.018 -0.173 -0.220 0.559 0.151
C073 -0.110 0.019 0.837 -0.013 0.045 0.074 -0.122 0.155 0.006 0.113
C074 0.237 0.498 0.157 0.196 0.077 -0.098 0.150 0.324 0.270 0.393
C075 0.620 0.240 0.234 0.353 0.198 0.241 0.148 0.064 -0.044 0.186
C076 0.884 0.020 -0.047 0.046 0.086 0.063 0.207 -0.008 0.159 -0.136
C077 0.181 0.141 -0.069 -0.085 -0.034 0.569 -0.074 -0.137 -0.112 -0.257
C080 -0.053 0.089 -0.229 -0.227 0.212 -0.223 0.101 0.103 -0.204 0.178
C081 0.169 0.401 0.103 0.337 0.004 0.192 0.022 -0.441 0.211 0.318
C082 -0.030 0.450 0.55 0.190 -0.185 -0.137 0.024 -0.057 -0.055 -0.015
C084 0.025 0.230 0.079 0.794 0.058 0.097 0.028 0.319 0.021 0.114
C087 0.351 0.229 -0.040 0.404 0.369 0.341 0.157 0.171 -0.016 0.057
CO1q 0.023 0.408 -0.0O4 0.528 0.194 -0.211 -0.076 0.031 -0.164 0.077
C102 0.090 0.493 0.275 0.434 -0.101 -0.126 -0.049 0.073 0.217 -0.010






85


TABLE 12
UARINTAX ROTATED FACTOR MATRIX FOR EXCINEERIN STUDENT GROUP
CONSTRUCT 1 2 3 4 5 4 7 8 9 10

C093 0.140 0.304 0.004 -0.035 -0.173 0.039 0.244 0.4q9 0.483 0.079
CO4 0.021 0.306 -0.138 -0.023 0.066 0.064 0.409 0.064 0.429 -0.007
C095 0.378 -0.362 0.054 0.147 0.125 0.181 -0.030 0.438 u.303 0.229
C097 0.089 -0.128 0.238 -0.164 0.617 0.005 0.155 0.178 -0.204 0.259
C098 -0.098 0.046 -0.088 0.166 0.229 0.090 0.040 0.194 0.055 0.435
C101 0.715 0.141 -0.130 0.074 0.376 -0.043 0.131 -0.033 0.164 0.198
C102 0.598 -0.033 0.189 0.163 0.317 -0.057 0.527 0.174 -0.127 0.066
C103 0.218 -0.154 0.087 0.144 0.601 0.313 0.213 0.251 -0.189 -0.020
C104 0.102 0.098 0.221 0.176 0.302 0.005 0.037 0.426 0.032 0.039
C106 0.206 -0.038 0.029 0.207 0.381 0.235 -0.059 0.540 0.413 0.231
C107 -0.015 0.271 0.048 -0,081 0.074 -0.069 0.070 0.055 0.708 -0.134
ClO8 -0.040 0.332 0.009 -0.042 -0.239 0.273 0.112 0.434 0.025 0.489
C109 -0.030 0.098 -0.160 -0.027 -0.135 0.393 0.358 -0.239 0.036 -0.117
C111 -0.102 0.322 0.175 -0.162 0.614 -0.013 0.113 0.437 -0.147 -0.031
C112 -0.050 0.161 0.619 0.219 0.259 0.194 0.442 -0.044 0.012 -0.022
C13 0.094 0.056 0.510 0.214 0.273 0.528 0.371 -0.169 0.102 0.024
Cll4 0.094 -0.116 0.513 0.311 0.067 0.241 0.094 0.194 0.302 0.345
C115 0.551 0.187 -0.081 0.645 0.048 0.165 0.140 0.077 0.049 0.046
C116 -0.031 -0.187 0.111 0.721 0.086 0.178 0.091 0.063 -0.005 0.190
C117 0.947 0.089 -0.087 0.009 0.029 0.132 0.114 -0.025 0.102 -0.028








TALE 13
UVRIMAX ROTATED FACTOR 0TRIX FOR EDUCATION STUDENT GROUP
CONSTRUCT 1 2 3 4 5 6 7 19 10
CoC1 0.219 0.445 -0.019 0.559 0.007 0.496 0.123 -0.063 0.14? -0.058
C002 -0.060 0.226 0.161 0.098 -0.099 0.156 0.317 0.037 0.672 -0.130
C004 0.787 0.217 0.049 0.273 0.220 -0.022 0.036 0.119 0.199 -0.066
C005 -0.100 0.315 0.263 0.071 0.473 0.516 0.220 0.118 0.375 -0.206
C007 0.559 0.373 0.105 0.172 0.105 0.233 0.471 0.103 0.283 0.014
C008 0.095 0.508 0.116 0.261 -0.030 0.359 0.230 -0.045 0.333 -0.070
C009 0.254 0.476 0.045 0.262 0.289 0.308 0.376 0.212 0.343 -0.065
C010 0.487 0.420 0.048 0.459 -0.004 0.131 -0.030 -0.034 0.334 0.101
Co11 0.272 0.670 0.297 0.178 0.236 0.153 0.051 0.237 0.234 0.187
C013 0.239 0.331 -0.024 0.169 0.188 0.635 -0.067 0.075 v.204 0.073
C014 0.043 -0.049 0.804 -0.048 0.043 0.207 0.171 -0.084 0.062 0.007
C015 0.238 0.699 -0.231 6.010 -0.023 -0.121 0.171 -0.134 -0.005 0.103
C014 0.047 0.608 0.352 -0.032 0.163 0.088 0.101 0.021 -0.079 -0.022
C017 0.237 -0.050 0.127 -0.059 0.459 -0.084 -0.186 -0.111 0.003 0.465
C018 0.418 0.369 0.420 0.052 0.446 0.116 0.083 0.058 0.146 -0.317
C019 -0.067 0.487 0.062 0.022 0.647 0.230 0.196 0.040 0.044 -0.063
C21 0.046 0.128 0.109 0.524 0.209 0.085 -0.128 -0.242 0.439 0.239
C022 0.079 0.020 0.510 0.229 0.716 -0.071 0.008 0.049 0.140 0.142
C023 0.068 0.038 0.13, -0.038 0.773 -0.050 -0.002 0.204 0.081 0.129
C024 0.020 0.050 0.052 0.033 0.168 -0.016 -0.195 0.020 -0.071 -0.549
C025 0.175 -0.183 -0.047 -0.102 0.306 0.252 -0.025 0.593 0.292 -0.011
C026 0.145 -0.015 0.916 0.074 0.164 -0.016 -0.036 -4.077 0.077 -0.011
C027 0.830 0.072 -0.028 0.169 0.263 -0.014 0.245 -0.002 0.000 -0.010
C028 0.145 0.703 0.104 -0.062 0.254 0.395 0.051 -0.164 0.165 -0.034
C029 -0.152 0.442 0.005 -0.193 -0.037 0.197 0.181 -0.002 0.317 -0.035
C030 0.046 0.180 0.141 0.030 0.028 0.782 0.007 0.268 0.134 0.024
C031 0.591 0.574 0.112 0.347 0.186 0.191 0.060 0.149 -0.192 0.035
C032 0.173 -0.032 0.053 -0.049 0.405 0.000 -0.011 0.041 -0.254 -0.097
C033 0.511 0.280 0.189 0.048 -0.040 0.358 -0.017 0.372 0.051 -0.084
C034 0.390 0.063 0.023 0.169 0.333 0.026 0.134 -0.3,6 0.240 0.211
C035 0.016 0.068 0.272 0.091 0.062 0.184 0.748 0.148 0.033 0.098
C036 0.496 0.403 0.294 0.319 -0.046 0.135 0.120 -0.075 -0.034 0.175
C038 0.608 0.272 0.086 0.270 -0.082 -0.041 -0.184 0.119 0.002 0.183
C040 -0.077 0.644 0.011 0.159 0.145 0.037 -0.221 0.324 0.100 -0.059
C042 0.219 0.089 -0.147 -0.049 0.066 -0.026 0.310 0.539 0.077 0.173
C043 0.088 0.197 0.213 0.410 -0.067 0.222 0.328 0.400 0.121 -0.233
C044 -0.044 0.037 0.206 0.807 0.040 -0.050 0.024 -0.07.5 0.284 0.320
C045 0.228 0.094 0.671 0.041 0.032 0.498 0.302 0.227 0.072 -0.045
C046 0.107 0.256 0.381 0.285 0.035 0.280 0.173 0.045 0.005 0.436
C041 0.030 0.416 0.452 0.002 0.179 0.328 0.047 -0.149 -0.057 0.140
C050 0.293 0.319 0.441 -0.100 0.084 0.441 0.223 0.243 0.172 0.057
C051 0.159 0.228 0.307 0.407 0.227 0.287 -0.054 0.043 0.261 0.451
C0o0 0.142 0.130 -0.090 0.741 -0.126 0.194 -0.193 0.152 -0.025 -0.118
C062 0.204 0.250 -0.164 0.080 -0.030 -0.217 0.661 -0.034 0.055 0.087
C064 0.241 0.401 -0.041 0.399 0.014 -0.074 0.053 0.117 -0.062 0.028
C066 -0.050 0.019 0.227 0.310 0.337 0.494 -0.082 -0.006 -0.092 0.125
C067 0.343 0.474 0.339 0.489 0.117 -0.036 0.200 0.234 0.0O6 0.061
C069 -0.079 -0.107 0.139 0.573 0.222 0.066 -0.034 0.202 0.002 0.405
C073 0.365 0.257 0.167 0.092 0.429 0.218 0.532 0.079 0.207 -0.152
C074 0.904 0.086 0.100 0.279 0.023 0.051 -0.024 0.072 0.001 0.031
C075 0.573 0.156 0.242 0.462 0.139 -0.021 0.202 0.16d -0.103 0.339
C076 0.306 0.566 0.117 0.343 -0.014 0.280 0.040 0.056 0.285 -0.200
C077 -0.025 0.162 -0.144 0.321 0.050 -0.014 0.419 0.461 0.168 -0.198
C078 0.339 0.555 -0.041 0.547 0.005 0.143 -0.115 0.240 0.167 0.029
C079 0.147 0.152 0.390 -0.115 0.035 0.096 0.768 0.162 -0.007 0.106
C080 0.034 0.196 -0.120 0.254 0.388 0.383 -0.093 0.152 0.447 -0.337
CC81 0.247 0.449 -0.179 0.140 -0.099 0.188 -0.039 0.477 0.023 0.066
C082 0.077 0.658 -0.130 -0.101 0.004 0.201 0.271 -0.215 -0.136 -0.105
CC83 0.839 0.087 -0.014 0.032 -0.043 -0.050 0.196 0.181 -0.086 -0.113
C084 0.030 0.180 0.152 0.203 0.683 0.298 0.225 -0.018 0.275 -0.059






87

TABLE 13
VARIMAX ROTATED FACTOR MATRIX FOR EDUCATION STUDENT 1ODUP
CONSTRUCT 1 2 3 4 5 6 7 8 9 10

C087 0.012 -0.032 0.220 0.192 0.137 -0.006 -0.087 0.160 -..234 0.412
C088 0.220 -0.012 0.161 0.237 0.219 0.095 0.184 0.672 -4.310 0.029
COC9 0.512 -0.060 0.221 -0.168 0.181 0.016 0.194 -0.085 0.147 0.058
C091 0.079 0.389 0.100 -0.129 0.588 0.378 -0.058 0.015 -0.175 -0.042
CO93 0.293 0.143 0.753 0.160 0.114 0.206 0.148 -0.038 0.061 0.080
C094 0.263 0.803 0.352 -0.066 0.032 0.087 0.197 0.107 0.119 -0.026
CC09 0.458 0.009 -0.004 0.719 0.077 0.133 0.119 0.134 -0.006 0.018
C097 0.632 0.001 0.202 0.314 0.075 0.266 -0.060 -0.047 -0.019 -0.183
C098 0.175 -0.055 -0.105 0.807 -0.055 -0.113 0.057 -0.076 0.003 -0.129
C099 0.667 0.098 0.117 0.056 -0.001 -0.108 0.042 0.429 0.133 0.131
C101 0.485 0.031 0.123 -.01Ol 0.129 0.042 -0.032 0.183 0.728 0.235
C102 0.517 0.138 0.218 -0.074 0.014 0.476 0.144 -0.114 -).123 0.219
C103 0.345 -0.111 -0.018 0.748 0.071 -0.018 0.243 -0.032 -.153 -0.014
C104 0.087 0.038 0.910 0.C87 0.104 0.091 -0.080 -0.050 0.055 0.181
C105 -0.195 0.171 0.390 -4.116 -0.064 0.568 0.214 -0.193 -0.027 -0.037
C106 0.397 0.190 -0.073 0.022 -0.122 0.179 -0.212 -0.290 0.013 -0.023
C110 -0.004 0.096 0.911 -0.112 0.202 -0.102 0.029 0.143 -0.058 -0.018










TALE 14
0URI2 X RCETATED FiCTOR MATRIX FOR CDOMnNITY GROUP


COHSTRUCT
C002
C007
C0019
C010
coil
C012
C013
C015
cold
C017
C018
C021
C022
C023
C025
C027
C029
C030
C031
CO3l
C033
C034
C035
C037
C040
C045
C048
C050
C060
C066
C067
C075
1080

C083
0c0a
0o92
1098

Clod
Co199
C115
C115
C118
c119


1
0.705
0.467
-0.198
-0.082
0.473
-0.134
0.215
0.440
0.004
0.003
0.209
-0.056
0.215
0.032
0.280
0.290
-0.156
0.435
-0.034
0.135
0.742
0.503
-0.178
-0.131
0.460
0.120
-0.041
0.775
0.197
0.068
0.220
0.417
0.715
-0.261
0.404
0.252
0.298
-0.121
0.742
0.383
0.793
0.125
0.204
0.217


2 3 4
0.063 0.069 -0.048
0.051 0.196 0.022
0.069 0.025 0.883
-0.047 -0.136 -0.090
0.048 0.260 -0.188
0.072 0.063 0.887
0.908 0.069 0.105
0.130 -0.235 -0.260
-0.123 0.129 0.151
0.358 -0.324 -0.026
0.049 -0.106 0.233
-0.132 -0.070 -0.0ll3
0.908 0.069 0.105
0.025 0.258 -0.008
0.570 0.120 0.185
0.582 0.080 -0.273
0.371 -0.070 -0.036
-0.073 0.265 0.150
-0.083 0.33 0.063
-0.136 0.169 -0.013
0.218 0.180 -0.169
0.337 0.622 -0.002
0.503 0.199 0.242
-0.149 0.477 -0.204
0.077 0.567 0.118
0.175 0.543 -0.220
-0.074 -0.063 -0.096
0.050 0.117 -0.065
0.167 0.409 -0.136
-0.269 -0.350 0.169
0.270 -0.231 0.222
0.319 0.101 0.375
0.203 0.082 -0.271
0.155 -0.052 0.758
0.232 0.560 0.001
0.252 0.766 -0.003
0.758 0.171 -0.020
0.726 0.390 0.217
0.218 0.180 -0.169
0.021 -0.067 0.717
0.077 0.098 0.138
0.185 0.229 0.103
0.135 0.805 0.184
-0.103 -0.042 -0.044


5 6 7 8
0.095 0.341 -0.117 -0.105
-0.087 0.096 0.309 0.253
0.020 0.003 0.086 0.007
-0.097 0.010 0.060 0.025
0.561 0.305 -0.299 -0.035
0.066 0.021 0.075 0.038
0.113 0.115 -0.131 0.028
0.086 0.212 0.488 -0.019
-0.165 0.128 0.807 -0.091
0.172 0.778 -0.050 -0.075
-0.174 0.566 0.206 0.052
-0.142 -0.069 -0.071 0.131
0.113 0.115 -0.131 0.028
-0.065 0.747 0.350 0.183
0.174 0.314 0.374 0.109
-0.190 0.191 0.036 0.292
0.058 -0.142 -0.055 -0.067
-0.137 0.139 0.373 0.295
-0.008 0.791 -0.041 0.101
0.096 -0.089 0.223 -0.764
0.134 -0.199 0.04' -0.091
0.186 -0.025 0.153 -0.061
0.304 -0.116 0.524 -0.012
0.116 0.231 -0.275 0.185
-0.198 0.082 -0.430 0.275
0.467 0.114 0.315 -0.085
-0.100 -0.070 0.031 0.073
0.092 0.123 0.057 -0.134
0.704 0.129 -0.150 -0.224
0.657 -0.191 -0.011 0.290
0.667 0.055 0.269 0.319
0.100 -0.245 -0.138 0.302
0.218 0.245 -0.032 -0.014
0.060 0.275 0.274 -0.188
0.425 -0.131 -0.174 -0.158
0.046 0.017 0.259 0.10q
-0.083 -0.280 0.223 0.197
0.354 0.075 0.008 -0.005
0.134 -0.199 0.044 -0.091
-0.013 -0.128 -0.221 0.244
0.030 -0.157 0.096 -0.090
0.876 -0.034 -0.055 -0.130
0.037 0.105 0.073 -0.233
-0.045 -0.059 -0.096 -0.706


9

-0.006
0.519
0.011
0.086
0.029
0.050
0.051
0.290
-0.00 2
-0.059
0.083
0.664
0.051
-0.151
0.229
-0.297
0.669
0.397
-0.061
-0.102
-0.079
0.044
-0.142
0.144
-i.083
0.077
0.104
0.034
-0.246
-0.151
-0.138
-0.047
-0.08?
-0.023
-0.073
-0.031
0.119
-0.079
-0.122
-0.137
-0.010
-0.042
-0.076


10
0.171
-0.419
-0.035
-0.435
-0.038
-0.050
-0.014
-0.151
0.033
0.109
0.599
0.073
-0.014
-0.075
-0.121
0.005
-0.017
-0.227
-0.187
0.084
-0.218
-0.051
0.195
O.161
0.104
0.288
0.572
0.322
0.001
-0.077
0.025
-0.150
0.005
0.158
-0.130
0.080
-0.058
0.090
-0.218
0.113
0.123
-0.028
-0.110
-0.138









loading heavily on only a few variables. These combinations

of variables are, of course, more easily interpreted than

the more general first factors (Nie et al., 1970).

The first ten factors for each group are listed in

Tables 15, 16, and 17 along the significant constructs deter-

mining their dimensionality. Note that similar dimension-

alities occur across groups with the appearance of a resource

dimension, a politico-technological dimension, and a politico-

economic dimension most prevalent. It is possible to locate

individual variables (constructs) with regard to these axes

of dimensionality by use of the factor loading coefficients

from Tables 12, 13, and 14. Figure 1 displays the location

of the most frequently cited construct across groups (con-

struct 18 "conventional vs. alternate technology") for each

of the three groups on the resource X politico-technological

plane. Figure 2 displays the location of this variable on

the resource X politico-economic plane, and Figure 3 locates

construct 18 on the politico-technological X politico-economic

plane for each group.

What is clear from the figures is that using similar

dimensionalities and constructs, the different groups per-

ceive the energy problem (at least in this regard) differently.

Strong loadings on a particular dimensionality can provide

important clues to the value placed on a particular construct.

The implications of this interpretation for curriculum devel-

opment are explored in Chapter V.




Full Text

PAGE 1

COMMUNITY EDUCATIONAL PROCESSES: GROUP PERCEPTIONS OF ENERGY ISSUES BY CAROL DOUGLASS BLALOCK A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 1980

PAGE 2

TO MY HUSBAND TONY AND MY CHILDREN JEANNE, PATRICIA, AND ELIZABETH FOR THEIR LOVE AND ENCOURAGEMENT, AND TO MY PARENTS WHO STARTED IT ALL

PAGE 3

ACKNOWLEDGMENTS The present research could not have been accomplished without the assistance and cooperation of many individuals. The researcher wishes to express appreciation to her doctoral committee, Dr. Arthur J. Lewis, Chairman, Dr. Roderick McDavis, and Dr. Lynn C. Oberlin, for the time and guidance provided. In particular, she is indebted to Dr. Arthur J. Lewis for his valuable advice at critical stages in the design and writing of this dissertation. The researcher would like to further recognize Dr. James R. Kennedy for his skill in the development of the software used for the repertory grid analysis and Dr. H. Anthony Blalock for scientific consultation in the data analysis of energy issues. In addition, special thanks are extended to Dr. Eugene A. Todd, Dr. H. A. Ingley, III, Mr. John Dykes, and Mr. Lem Lee for their assistance in data collection. The researcher also expresses gratitude to her husband, Tony, and her daughters, Jeanne, Patricia, and Elizabeth, for their love, encouragement, and patience over the years required to complete her graduate work. Finally, the researcher would like to acknowledge the influence of the many theoreticians and researchers whose work provides the conceptual base for this dissertation. iii

PAGE 4

They rightfully share any credit this dissertation reflects The researcher feels as Sir Isaac Newton did: "if I have seen thus far, it is because I have stood on the shoulders of giants." IV

PAGE 5

TABLE OF CONTENTS CHAPTER PAGE ACKNOWLEDGEMENTS iii LIST OF TABLES vii ABSTRACT ix I INTRODUCTION 1 ' The Problem and Its Context 1 Personal Construct Theory . . . 5 Design of the Study 11 Organization of Dissertation 16 II REVIEW OF LITERATURE 17 Analysis of Social Perceptions . . . . . . 17 Cultural Foundations of the Study .... 31 Summary 44 III PROCEDURES 46 Selection and Identification of Respondent Groups 46 Design and Administration of Repertory Grid Instrument 47 Data Reduction and Analysis 54 Statistical Procedures and Tests of Hypotheses 57 IV RESULTS 63 Frequency of Element and Construct Citation 64 Numbers of Constructs Elicited 67 Degree of Construct Association 68 Results of Factor Analyses of Grids .... 72 V CONCLUSIONS AND RECOMMENDATIONS . 99 Group Perceptions of Energy Issues .... 102

PAGE 6

CHAPTER PAGE Implications for Educational Planning and Curriculum Development 105 Implications for Other Areas of Educational Planning 107 APPENDICES A REPERTORY GRID INSTRUMENT 110 B DEMOGRAPHIC CHARACTERISTICS OF GROUPS ... 113 C MASTER LIST OF ELEMENTS 114 D MASTER LIST OF CONSTRUCTS 116 E TYPICAL SPSS FACTOR ANALYSIS OUTPUT 119 REFERENCES 132 BIOGRAPHICAL SKETCH 136

PAGE 7

LIST OF TABLES TABLE PAGE 1 Demographic Characteristics of Groups .... 48 2 The Most Commonly Cited Constructs byGroups 65 3 The Most Commonly Cited Elements by Groups 66 4 ANOVA Table of Number of Construct 69 5 ANOVA Table for Associations per Construct . 71 6 Summary of Factor Structure for Engineering Student Group 73 7 Summary of Factor Structure for Education Student Group 75 8 Summary of Factor Structure for Community Group 77 9 Unrotated Principal Components for Engineering Students 78 10 Unrotated Principal Components for Education Students 80 11 Unrotated Principal Components for Community Group 82 12 Varimax Rotated Factor Matrix for Engineering Student Group 84 13 Varimax Rotated Factor Matrix for Education Student Group 86 14 Varimax Rotated Factor Matrix for Community Group 88 15 Major Factors for Engineering Student Group . 90

PAGE 8

TABLE PAGE 16 Major Factors for Education Student Group 91 17 Major Factors for Community Group 9 2

PAGE 9

Abstract of Dissertation Presented to the Graduate Council of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy COMMUNITY EDUCATIONAL PROCESSES: GROUP PERCEPTIONS OF ENERGY ISSUES By CAROL DOUGLASS BLALOCK JUNE, 19 80 Chairman: Arthur J. Lewis Major Department: Curriculum and Instruction The major challenge to the planning of community educational processes lies in the pluralistic nature of society. Educational theorists agree as to the critical need for an emerging role for education in providing linkage between the various formal and functional groups in society. In accepting this challenge educators must find tools for mapping and quantifying similarities and differences for values and perceptions across groups. The purpose of this study was to demonstrate the utility of a method based on Kelly's personal construct theory for mapping and quantifying attitudes about energy issues. ix

PAGE 10

Kelly's personal construct theory views human thought and learning as a process of construing. In this view individuals view ideas about the world which give meaning to information and experiences. These constructs are bipolar so that new information is seen as like one end of the polarity and unlike the other. There are also constructs about constructs at various levels. Thus, people develop systems of constructs relating to any particular area of their lives. One measure of a person's effectiveness is how congruent his/ her construct system is with reality and consequently how well it permits prediction of events. In this sense, Kelly saw all persons as scientists making and testing models of reality. A major purpose of the study was to demonstrate the utility of repertory grid analysis in identifying and quantifying perceptual differences among groups. Repertory grid questionnaires were administered to three groups: a group of senior education majors at the University of Florida, a group of senior mechanical engineering majors at the University of Florida, and a group of adult learners enrolled in energy awareness or radio/T.V. repair classes at Santa Fe Community College. These groups were expected to have substantially different energy-related perceptions because of their widely different educational experiences. Results of this study demonstrated that all three groups shared similar awareness of energy issues, but that they construed these awarenesses differently. Differences were also noted with

PAGE 11

regard to perceptual complexity and connectedness but not centrality. Factor analysis of the pooled repertory grid data for each group permitted the identification of common dimensionalities of construing among the group. The use of factor loading scores as potential indicators of values is described and an example illustrated. Suggestions are made for the incorporation of information derived from the repertory grid studies into community educational planning. Strategies are described for directing dialog among dimensionalities identified as critical by factor analysis. The extension of the method to a broad range of educational activities is discussed.

PAGE 12

CHAPTER I INTRODUCTION The Problem and Its Context The purpose of this study was to examine the perceptions of undergraduate students in engineering, education, and the community about energy issues in order to map these perceptions and quantify similarities and differences among groups. Data from this study provide groundwork for energy education and the design of the study is a paradigm for studying group perceptions about a variety of issues. It has long been recognized that education is a lifelong process which takes place throughout the community (Dewey, 1916; Cremin, 1976; Cross, 1979). Many educators have identified the multiplicity of educative forces and institutions in modern society. Illich (1971, p. 47) was perhaps the first to propose the development of "learning webs-voluntary networks that would permit any student at any time to gain access to any resource that might help him define and achieve his goals." Cremin (1976, p. 27) suggests that such networks already exist to some extent and expands the definition of education to include any "deliberate systematic and sustained effort to transmit, evoke, or acquire knowledge, attitudes, values, skills, or sensibilities as well as any outcomes of that effort." 1

PAGE 13

Society and communities in the United States are replete with efforts to inform and persuade which are both deliberate and sustained. However, the extent that these efforts are broad and systematic in nature is more a reflection of cultural homogeneity than the result of planning and coordination. What may be needed for both efficiency and greater effectiveness is not more educational efforts, but coordination of educational efforts to be mutually supportive rather than duplicative or divergent. The anthropological literature describes an institutionalized network for the dissemination of information which provides most of the formal ("technical" in the sense of Hall, 1959) communication in our society. Major components of this network are public agencies and institutions, the media, and specific public interest groups. This network is characterized by information that is processed in a central locus and disseminated through a multibranching series of pathways. This type of network is designated as dendritic, The nature of a dendritic network is such that it is possible to coordinate the flow and control the quality of information. It should be noted that information flow is bidirectional while control and coordination goes downward. People living in communities develop a series of networks of communication which are characterized by random paths in which information flows out from a locus of perception. This type of network is designated as radiative. A radiative network operates in the spreading of rumors and

PAGE 14

the development of "grass-roots" movements. Because of the dynamic and continuously changing nature of this network information flows cannot be coordinated nor can information quality be controlled. Still this radiative type of network is very important in the life of a community and probably operates strongly to condition values and perceptions. In fact, it may be this type of network that drives the statistical convergence in the Delphi technique (Mayer, 1964; Myers, 1979). It would seem reasonable that the values and perceptions of an individual tend to be determined by those implicit in the communication in the dendritic network unless a different perception has been spread through the radiative network. Consequently, each type of communication will affect a given individual to a varying extent at different times on a single issue and on different issues at the same time (Robinson, 1950) . Concepts of network types and influences can be brought to bear on a variety of current issues. Many studies have demonstrated that under certain conditions (which are poorly understood) people behave in a way which is inconsistent with their announced values. For example, studies sponsored by the United States Department of Energy have shown that while most people strongly support energy conservation they tend not to support this behaviorally (Public Energy Education, 1979) . Thus while we can make inferences about the values and perceptions an individual holds from an analysis of that

PAGE 15

person's ethnic and socioeconomic background, educational status, and/or formal associations, these influences are likely to be useful only in the absence of other strong influences conditioning that individual's behavior. In particular the influence of a radiative network on specific perceptions would appear to be exceptionally strong (Mayer, 1964). This may account for the type of observed discrepancies referred to above. It is possible to determine functional groupings based on clustering of like perceptions and values (Kluckholm and Strodtbeck, 1961). Correlating these functional groups with institutional associations can provide insight regarding the value structure of a community (Kluckholm and Strodtbeck, 1961; Robinson, 1950). All managed networks by necessity are dendritic; thus, it may be argued that managed networking will be most effective where radiative networks are weak or absent. Because they lack lateral communication capabilities, dendritic networks cannot provide the value reinforcement which takes place in radiative networks. This effect may be approximated by the coordinated actions of several dendritic networks on an individual belonging to all of these networks. Agencies assuming responsibility for providing public information and clarifying values on issues of social importance can improve their effectiveness by precisely adapting their program content and delivery methods to the characteristics of target groups (Torda, 1978; Rosenstein, 1978; Cremin, 1976; Dewey, 1938). Unfortunately, there is at this

PAGE 16

time no generally accepted methodology for mapping the values and perceptions of such target groups. Were such a methodology available, its use could lead to the effective delineation of target groups and their characteristics. If educational activities relating to areas of social concern are to be effective in facilitating participatory decision-making, they must meet four criteria: relevance, timeliness, accessibility, and compatibility with group values and perceptions (Public Energy Education, 1979). Much research has been done regarding the meeting of the first three criteria; relatively little work, however , is available in the area of mapping and analyzing perceptions. It is the purpose of this dissertation to demonstrate the use of personal construct theory as developed by George Kelly (1955) as a tool for mapping and contrasting perceptual and evaluative activities of individuals and groups. Personal Construct Theory George Kelly's (1955) introduction and development of personal construct theory is based on a unique set of assumptions regarding the nature of mankind and the human universe. Three assertions are made in relation to the universe. Humanity knows the universe to be real and not the result of someone's imagination; humanity knows that the universe can be understood only by repeated encounters over a period of time; and humanity knows that all events are interrelated.

PAGE 17

Kelly's assumption that the universe contains real events and objects is complemented by his assumption that a person's internal events are as real. Thoughts about external objects or events also have an internal reality as authentic as the happenings themselves. Kelly believed our knowledge about the universe is determined by the extent to which we can interpret it. For individuals, this interpretation gradually approaches a true picture of events by successive approximations. Unlike the purely subjective existential or phenomenological theorists, Kelly holds to the possibility of accurate knowledge of objective reality. Mankind in Kelly's theory is trying out his interpretations or constructions for their accuracy in predicting the world he is beginning to understand. Secondly, Kelly believes that some aspects of existence can only be made comprehensible by the notion of time. Behavior is explicable only in the context of the future as well as the present and the past in Kelly's world view (1955) The third aspect of Kelly's personal construct theory, the notion of an integral universe, is consistent with the development of a system of ideas to bring coherence to one's world view. Kelly tried to describe how systems like this operate. The concept of an integral universe is both a philosophical and a psychological statement. People are seen as continually striving to construe relationships where there were none perceived before (Bannister, 1968).

PAGE 18

Kelly visualizes humans as distinguished by their ability to not only respond to life's events but able to represent their environment. Personal construct theory is based on people knowing the world by means of the constructions they place upon it. According to Kelly humans are bound by events to the degree that they have the ability to construe them. All present perceptions are open to question and reconsideration. Kelly's philosophical position has been called constructive alternativism (Bannister, 1968). Personal construct theory (Kelly, 1955) can be applied to the analysis of physical environments as well as personal perceptions. Kelly postulated that knowledge precedes behavior and that a person's knowledge is developed over time by accumulating and categorizing information. People develop images of their environment in this manner. Kelly calls these images constructs. Construing means to Kelly placing an interpretation upon. Moreover, "in construing, the person notes features in a series of elements which characterize some of the elements and are particularly uncharacteristic of others" (Bannister, 1968, p. 7) . The technique used to visualize an individual's constructs contains four steps, each of which will be described in detail in a subsequent section of this work. The first task is the listing of relevant elements, i.e., observations; the second is the generation of bipolar constructs from these elements; the third is the location of each element in total construct space; and the fourth is the rotation of axes to identify

PAGE 19

factors, combinations of constructs which divide element groups. The first two tasks require only listing, the third requires generation of a grid or matrix of element-construct relationships, and the fourth a computerized analysis. The analysis of a grid is to make clear the structure in the grid; analytical techniques therefore should not impose the structure of the experimenter. If each factor in the grid represented were a separate unrelated opinion of the subject, no structure would emerge. However, all grids developed by people have inherent structure. In other words, the relationships among individual elements which are susceptible to mathematical analysis are representative of basic psychological processes. Grid analysis, then, should be applicable in principle to any area of perception and cognition. Application of Personal Construct Theory to the Study of Environmental Perceptions Deutsch (1972) used Kelly's repertory grid test to ask subjects to apply a group of their own constructs to a set of known environments. Graduate students in architecture were compared with graduate students in other fields. Significant differences between groups were found in verbal content and organizational structure. The results of this study suggested further work could be done on the factor analytic procedures used as well as on the reliability of the coding methods used.

PAGE 20

Stringer (1976) used personal construct theory (Kelly, 1955) to facilitate public participation in planning to rebuild Wistow Hill "Triangle," a London neighborhood, in 1970. The general feeling from the public was that some form of remodeling should be attempted. Stringer was seeking a general understanding of environmental value systems. Using personal construct theory to allow individual differences to emerge was essential to obtaining a group of elements representative of individual construct systems. Instead of using simple preference orderings, Stringer believed the meanings of elements are more significant when observed in relation to each other. Fewer assumptions by the researcher then need to be made. Stringer felt the repertory grid technique "integrated perception and evaluation" at a time when political rallying made participation a reality. Attitudes are changing according to Stringer and "doing" has become more important than having. Change has also become of paramount importance in the process of environmental perception (Kelly, 1977) . Application of Construct Theory to Energy Education The energy issue, like many other issues of our time, presents both technological and philosophical problems. As a case in point, the population of Sweden enjoys a comparable standard of living to that of the United States at about 60 percent the energy cost using substantially the same technology (Schipper, 1976). As important to discovering additional oil and alternate energy resources is the development

PAGE 21

10 of cooperative group processes to deal with the critical social issues. The role of education is to facilitate the decision making process by communicating information and clarifying values. Personal construct theory can provide specific information to deal with perceptions and values of the population as a whole and certain groups specifically. The methodology used in this study has yet to be demonstrated in the context of energy perceptions or other critical social issues. The purpose of the study was to demonstrate that personal construct theory in this form is useful in delineating the energy perceptions of three groups which can be expected a priori to differ in attitudes. A group was chosen with obviously different background so that there would be a maximal likelihood that each might look at the energy problem differently. A group of senior undergraduate engineers specializing in energy conversion was selected to represent a group of individuals knowledgeable in energy use and transfer. The second group was composed of senior secondary education majors chosen to represent lay knowledge on the energy problem as well as specialized awareness of the participatory nature of our society with certain social and political orientations. The last group were representatives of the community at large. These students were registered at Santa Fe Community College receiving credit for either a radio/T.V. or an energy correspondence course. It was believed they demonstrated a wide spectrum of attitudes.

PAGE 22

11 Design of the Study The sample population included approximately 75 undergraduate students of the University of Florida and Santa Fe Community College. The first group of 28 were senior mechanical engineering majors studying energy conversion. The second group of 32 were University of Florida seniors also majoring in secondary education. The third group of 14 community members were chosen from two classes at Santa Fe Community College which were known to be heterogeneous in terms of age and occupations. All of the subjects were asked to complete a three part listing based on Kelly's repertory grid method. In the first part the subject was to list 20 elements related to the energy problem. In the second part the subject chose from the first part pairs of factors which were the same and one opposite forming a triad. Subjects were encouraged to state how factors were the same or opposite (from constructs) . The third part consisted of a grid with elements along the X axis numbered one to twenty and constructs located on the Y axis lettered A through Z. Respondents were asked to correlate each construct with each element. The repertory grid was administered to the engineering, teacher, and radio/T.V. classes allowing approximately an hour for completion. If more time was required for completion, the researcher extended the time as needed. The repertory grid was mailed along with the mid-term exam to the

PAGE 23

12 community group studying energy. The researcher's phone number was included to encourage questions. Analysis of the data included a number of steps. First, all element responses were tabulated to ascertain how many different elements exist and how many times each element is repeated. Secondly, constructs were tabulated to determine how many constructs exist and their repetitions. An element by construct grid was developed using factor analysis for each subject. All grid data for a particular group were combined to form a composite grid of elements, constructs, and their relationships for each group. Finally, the three groups were compared for similarities and differences (Bannister, 1968). The following hypotheses were tested: 1. The three groups have not chosen different elements which can be tested by analysis of the frequency distribution. 2. The three groups have not developed different numbers of constructs which can be tested by an analysis of construct responses. 3. The three groups did not show different numbers of positive relationships between elements and constructs which can be tested by grid. 4. Analysis of group data does not lead to different principal components which can be tested by factor analysis.

PAGE 24

13 Definition of Terms Fundamental postulate "A person's processes are psychologically channelized by the ways in which he or she anticipates events" (Kelly, 1955, p. 46). Construction corollary "A person anticipates events by construing their replications" (Kelly, 1955, p. 50). Kelly expressed his assumption that all men act as scientists in this world (Ryle, 1975). Leman (1970, p. 65) examined this corollary from a linguistic philosophy and suggested that 'the characteristically scientific activity is an operation with language . . . and that the scientist's most important problems have to do with the relationship between language and extralinguistic reality . . . ." Leman emphasized the 'making-sense-of ' aspect of men's constructing of themselves and reality. Individuality corollary "Persons differ from each other in their constructions of events" (Kelly, 1955, p. 55) . Organization corollary "Each person characteristically evolves for his/her convenience in anticipating events a construction system embracing ordinal relationships between constructs" (Kelly, 1955, p. 56). Dochotomy corollary "A person's construction system is composed of a finite number of dichotomous constructs" (Kelly, 1955, p. 59) .

PAGE 25

14 Choice corollary ''A person chooses for herself or himself that alternative in a dichotomized construct through which he or she anticipates the greater possibility for the elaboration of her system" (Kelly, 1955, p. 64). Range corollary "A construct is convenient for the anticipation of a finite range of events only" (Kelly, 1955, p. 68). Experience corollary "A persons' s construction system varies as he construes the replication of events" (Kelly, 1955, p. 72) . Elements "The things or events which are abstracted by a person's use of a construct are called elements. In some systems these are called objects" (Bannister, 1968, p. 219). Core construct "A core construct is one which governs an individual's maintenance processes" (Bannister, 1968, p. 221). Factor analysis '-'Factor analysis is a multivariable method that has as its aim the explanation of relationships among several dif f icultto-interpret , correlated variables in terms of a few conceptually meaningful, relatively independent factors" (Kleinbaum, 1978, p. 276). Assumptions It is assumed that the diversity within the sample groups and among the sample groups is a minimum estimator of community diversity. It is also assumed that it will be possible to functionally define groups in later work by

PAGE 26

15 identifying clusters of individuals having between group differences equal to or greater than those demonstrated by the groups in this study. Delimitations The scope of the study included three well-defined groups with clearly identifiable educational characteristics. Thus, in no way could the groups be construed as either a random or a polar representation of the community at large. Rather, the value of the study lies in demonstrating a novel and effective perception mapping technique and demonstrating the application of this technique to the identification of functional groupings. Limitations The selection of a repertory grid with subjects responding with negative or positive answers limits the sensitivity of perceptions of relationships. Using an intensity scale for each relationship, however, would have made administration of the repertory grid an extremely tedious process. Administration of the repertory grid format to groups prevented acquisition of contextual information by the researcher which might help clarify the dimensionality of constructs. However, the avoidance of the interview technique reduced the possibility of experimenter contamination of results. Interpretation of principal components is a qualitative rather than a quantitative measure of group differences. By this method it will be possible to identify how groups differ in perceptions, but not possible to directly compare two sets of between group differences.

PAGE 27

16 Organization of Dissertation In the remaining four chapters of this dissertation the literature pertinent to this study is reviewed, the experimental and statistical methodology are described, the results reported, and the significant implications of those results discussed. A review of literature pertinent to social perception research, personal construct theory, factor analysis, and cultural foundations of this study is presented in Chapter II. The rationale for the selection of personal construct theory and repertory grid analysis as the method of choice is described in Chapter III. A description of the groups studied and a discussion of the statistical methodology used to analyze and interpret the results also are included in this chapter. The numerical results of the study are presented in Chapter IV. Raw data summaries, descriptive statistics, and analytical results ere tabulated and interpreted. A discussion of the findings in light of the original experimental intent is presented in Chapter V. Additionally, a review of the implications of this study will suggest extensions into network identification and applications to other areas of educational research. For the sake of brevity and convenience, the groups of engineering students, education students, and community college students will be referred to from time to time in the text and tables as "engineers," "teachers," and "community groups," respectively. w

PAGE 28

CHAPTER II REVIEW OF LITERATURE In developing and demonstrating the methodology of this study, it has been necessary to draw upon the fields of philosophical foundations of education, cultural foundations of education, social perception research, personal construct theory, and matrix methods of statistical analysis. The purpose of this study differs from the work previously done in that it was designed to provide a vehicle for meaningful user input to the development of educational programs. Of course, the problems of relevancy and of the roles of education have been of concern to many educators. In the following section, a summary of the philosophical, educational, and methodological bases of this work is presented. Analysis of Social Perception If culture is communication as Hall (1959) has said, then education may be thought of as the summated social technology of communication. Examination of this assertion leads us to ask three questions: (1) What are the activities and institutions which communicate and educate in a modern society? (2) Who decides what is to be communicated and how? (3) How do the people in a community perceive and influence the educational process? In answer to these questions a map of the network of educative forces and activities 17

PAGE 29

18 occurring in society will be found. Identification will also be made as to where that network breaks down : in identifying and incorporating the perceptions of the groups it serves . Educational Roles in Social Attitudes and Perceptions Education has grown from providing only literacy and vocational skills to modifying social values through the broad development of all functional aspects of citizenship. It is important to view education in its broadest possible context: in its role as interpreter and modifier of social attitudes. It will also be necessary to identify the role of education in serving the needs of nontraditional learners, In analysis of social perceptions a number of educational philosophers have interpreted and conditioned social attitudes and values. Dewey (1916) realized that cultural transmission was education in its broadest sense. He pointed to the difference between the education most individuals get simply from the process of living and the deliberate education offered by the schools. For Dewey, the advancement of civilization was a process of complexification. While the young of savage groups could participate in society by mere incidental learning, the young of complex cultures could gain the same degree of participation only with intentional learning. Dewey's theory of learning was essentially a theory of the school as agent of society. Dewey worked on reconciling the polarity between school and society. This gap has remained. Even in the educational reform of the

PAGE 30

19 1960 's, there was ambivalence as to whether schools should be improved or abolished altogether. In the 1970's, American opinion swung from an overreliance on the school as an agent of socialization to a widespread disenchantment of community support for schooling. Dewey anticipated this phenomenon in the year 1933 when all public institutions were suffering from a lack of public confidence. Silberman (1970, p. 2) was aware of the same dilemma. "If our concern is with education, we cannot restrict our attention to the schools, for education is not synonymous with schooling." He saw the need to emphasize the many educating aspects in American society other than schools. Silberman 's task was to make recommendations on the education of educators, but he found this difficult without an understanding of what education would be like in the years ahead. Thus, Silberman supported by the Carnegie Corporation, undertook a four-year study of schools and other educational agencies. This study, the purpose of which was to understand basic directional developments and synthesize these into a coherent program, is today viewed by some as a theoretical statement of educational progressivism. Illich (1971) recognized, as had Dewey, that many institutions and situations educate and that the school is only one such. He proposed the development of "convivial" educational institutions called learning webs which were networks permitting free access to any resource that might help a student achieve individual goals. He suggested four networks

PAGE 31

20 reference services in educational subjects, skill exchanges, peer matching, and reference services to education at large. Illich was thought to have achieved in design the Dewian Utopia of 1933 in which no schools existed at all. "The most Utopian thing about Utopia is that there are no schools at all" (Dewey, 1933, p. 236). All people would learn what they needed to know from informal association with others. Cremin (1976) suggested that Dewey's formulations of democracy and education be revisited in order to attempt a redefinition of education. He questioned education as a result of which Dewey emphasized institutional origins rather than functions. The theory of education thus created posits major educative agencies performing a linking role with respect to other agencies and society. Cremin (1976, p. 27) has defined education "... as the deliberate, systematic, and sustained effort to transmit, evoke, or acquire knowledge, attitudes, values, skills, or sensibilities, as well as any outcomes of that effort." He saw education as a process more limited than socialization (as perceived by sociologists) or enculturation (as perceived by anthropologists) . This definition nonetheless has moved beyond focusing on the schools and colleges to identifying all persons and institutions that educate — siblings , churches, family, friends, libraries, museums, and others. Finally, the definition allows that education may produce outcomes which may be anticipated or may be unintended, with the possibility that the latter could be the more significant outcomes.

PAGE 32

21 Cremin (1976) defined the role of education by his concept of configurations of education. He envisioned a multiplicity of institutions relating to one another within the larger society. Relationships between and among institutions he saw as political, pedagogical, or personal. He predicted a correlative relationship between configurations of education and social change or stability. Through the enculturation of the young, configurations of education have maintained social continuity and stability (Cremin, 1976). Educational institutions also have played their part in catalyzing social change according to Cremin. At the individual level, persons have related uniquely to configurations of education (Cremin, 1976) according to individual experience and perspective. Cremin (1976) held that educational life history of an individual begins with the efforts of critical others in creating appropriate attitudes and behaviors and results in the individual becoming a self-directed learner. Fromm (1968) held a compatible view in that he saw each person to be sacred and (ideally) to be united with his/her world. He expressed concern that our educational system lacked quality despite its great institutionalization and despite the large numbers of college graduates. According to Fromm our educational system must become alive and responsive to each participant. He believed that it is time for man to assert himself and make the technological society human.

PAGE 33

22 Leichter (1974) visualized education as a lifelong process which can take place in a variety of settings and needs to be understood in each of the settings. She assumed education took place on numerous levels often simultaneously and both learning content and process need understanding. Leichter (1974, p. 239) conceptualized educative style as a "set of characteristic ways in which an individual engages in, moves through, and combines educational experiences over a lifetime." Here it is assumed that educationally significant others taught in childhood such relevant educational attitudes and that these are supported or modified by additional experience. Leichter's notion (1974) of educative style focused on continual change linked with the continuity of an individual moving from institution to institution within configurations and among configurations. Movement of this kind by each individual resulted in the development of individualized networks of education for each person and resulted in the individual's selection of a variety of learning activities which reflected that individual's constructs pertaining to knowledge and growth. Kohlberg (1966) realized the difficulty of understanding the educational process without a thorough picture of maturation, learning, and development. He believed like Dewey (1933) and Piaget (1973) that understanding of intellectual content and cognitive processes was essential to the development of moral judgment. Kohlberg's (1966) formulation of six developmental stages of moral judgment

PAGE 34

2 3 described the potential for continuous growth in individuals over a lifetime. Kohlberg's moral stages were redefined in 1975 with each of them examined in terms of operating "moral motives." This categorization was supported by longitudinal and cross-cultural studies. Kohlberg believed that moral reasoning was clearly reasoning and advanced moral reasoning depended upon advanced logical reasoning. A person's logical stage created a ceiling on the attainment of moral stage. To understand the learner and the learning process, we must observe reasoning about choice. This kind of reasoning defined the structure an individual chooses and was reflected in configurations of critically relevant other persons or institutions. Thus, if educators can understand what an individual finds valuable and why he/ she finds it valuable, they can begin to conceptualize not only the content of his/her moral judgment but also how he/ she is motivated to learn. On this basis it may be possible in theory to provide a more supportive climate in which learning can occur. Education has now recognized its obligation to meet needs of the non-traditional learner and is beginning to address these needs. Knowles (1977) described a new technology of androgogy or education of adults based on the premise that adults are critically different than children in experiencing learning. Cross (1979) emphasized the need for equal education for 30 to 80 year olds. Researchers estimated that between 80 and 90 percent of the adult

PAGE 35

24 population carried out one self-directed learning experience every year. Surveys show between 17 and 32 million adults are now participating in classes, workshops, groups, or other organized educational structure. By the year 2000 Cross (1979) forecasts the United States will become an adult culture with 57 percent of the population over thirty years of age. Cross (1979) has shown that education for adults is elitist with certain populations significantly underrepresented in organized learning activities. She argued that brokering services and education information centers represented the greatest hope for shaping the "learning society." The development of education and information services is perceived by Cross to involve three steps: (1) collecting information about the educational resources available, (2) reaching the intended audience, and (3) assisting clients to identify and obtain the appropriate learning opportunities. A strengthening of these three linking functions will thus contribute immeasurably to the success of lifelong learning. The recognition that the learner, especially the adult learner, must be viewed contextually has led to the important conclusions by Cremin (1976), Leichter (1974), and Illich (1971). It is reasonable to expect this matrix of interrelationships in which individuals are embedded to influence and condition values and perceptions. Clearly values and perceptions affect not only the choices of educational

PAGE 36

25 participation but also the motivation for participation. These influences are not generally seen as falling within the purview of education. Sociocultural Aspects of Attitudes and Perceptions If, however, one could define functional groupings within a community based on values and perceptions with regard to a particular area of interest, it would be possible to meet the needs of that group for education and information. A variety of anthropological studies have demonstrated methodologies for identifying and mapping social influences as well as for mapping attitudes. Mayer (1964) directed his efforts toward describing the networks of relationships among individuals in urban settings. He observed that people living in communities developed a series of networks of communication which were characterized by random paths in which information flowed out from a locus of perception. He saw such a network operating in the spreading of rumors and the development of "grass-roots" movements. Mayer (1964) knew that because of the dynamic and continuously changing nature of this network, information flow could not be coordinated nor could information quality be controlled. Still this type of network, according to Mayer (1964), was very important in the life of a community and probably operated strongly to condition values and perceptions. In fact, it may be this type of network that moved the statistical convergence in the Delphi technique (Myers, 1979).

PAGE 37

26 While Mayer emphasized intra-group networks, Kluckholm and Strodtbeck (1961) looked at the problem from the perspective of entire social groups in his study of five Southwestern cultures. Using interview techniques informants from five juxtaposed cultures were questioned about their perceptions of social and physical environments. Group mean scores for the different cultures were compared by a one-way analysis of variance for several predetermined dimensions. The results obtained included an accurate map of perceptions along the predetermined axes. This mapping elucidated the causes of a great deal of conflict in values among the various groups. The need for better tools in observing individual values was expressed by Craik (1970) , in his comprehensive review of environmental psychology, who devoted a section to the discussion of personality inventories. He pointed out the neglect by psychologists of items or scales for assessing environmental dispositions in spite of individuals' strong orientations toward the physical environment. He suggested a number of environmental dispositions, such as Pastoralism scale, an Eccological Perspective scale, a Luddite scale, an Urbanite scale, and measures of environmental sensitivity. McKechnie (1970, p. 320) developed the Environmental Response Inventory (ERI) for measuring environmental dispositions, which he defines as "the configuration of attitudes, beliefs, values, and sentiments . . ." of the individuals being tested. The Environmental Response Inventory included items relating to the areas of pastoralism, conservation,

PAGE 38

27 science and technology, urban life, rural life, stimulus preferences, cultural life, leisure activities, the outdoors, geographic and architectural preferences, and environmental memories and knowledge. The responses were factor analyzed and a series of different factors for men and women surfaced. The ERI scales did have an interesting correlation with numerous traditional personality measures, and with environment related behaviors such as membership in conservation and agricultural organizations. This result was consistent with the results of the anthropological work of Kluckholm and Strodtbeck (1961) and Mayer (1964). Numerous studies of energy perceptions have been done which demonstrated that under certain conditions (which are poorly understood) people behave in ways inconsistent with their announced values. For example, studies sponsored by the United States Department of Energy (Public Energy Education, 1979) have shown that while most people avowedly support energy conservation, they tend not to demonstrate this support behaviorally . Thus while inferences can be made about the values and perceptions an individual holds from an analysis of that person's ethnic and socioeconomic background, educational status, and/or formal associations, these inferences are likely to be useful only in the absence of information regarding other strong influences conditioning that individual's behavior. In particular the influence of an associational network on specific perceptions appeared to be

PAGE 39

28 exceptionally strong to Mayer (1964) . This may account for the observed discrepancies referred to above. Hall (1959) explored culture by means of communication theory (information theory) of the electronics laboratory. Hall (1959) likened communication theory to shorthand for talking about communication events such as phonetics of language, orthographies, and telephone signals. He noticed that the process proceeded in one direction-toward symbolization. He saw an individual's speech as an arbitrary vocal symbol used to describe something that had taken place or might have taken place with possibly no actual connection between occurrences and symbols. Because of the workings of culture, Hall (1959) saw talking as a highly selective process. He believed that no culture had discovered a means for talking without emphasizing some events at the expense of others. He saw writing, then, as a symbol of a symbol. Using communication theory (Hall, 1959) took this process still further. He noticed entire messages of various durations with some less than a minute and others extending over years. The study of culture thus could include events of short duration whereas the study of individuals or governments involved communication over longer duration. He developed a system based on tripartite theory which included three kinds of time: formal time, informal time, and technical time. Hall (1959) discovered that man had three modes of behavior and that at any point in time one of the three orientations would dominate although all three would be

PAGE 40

29 present. Formal activities were taught by a mistake being made and a correction suggested. Informal learning emphasized a model for imitation in which thousands of details could be passed through generations without specific understanding of the rules. Technical learning was shared in explicit terms from teacher to student usually preceded by a logical analysis. In summary, the formal mode was a twoway process while informal learning involved identification of a model to follow. The technical learning rested with the teacher. Numerous trends of social perceptions have been influential in the evolution of social communication. Joseph (1979) forecasted a transformation in education through technology to accommodate the tremendous mass of information necessary for adaption to a highly complex society. The information explosion has forced the emerging development of a "technology for education which included imbedding increasingly capable, but physically small, micro-processor logic, digital storage/memory, sensors, communications circuits and links, and eventually voice actuated and reply mechanisms for creating convivially smart machines-which are more humanistic for students" (Joseph, 1979, p. 1). Meadows (1974) emphasized the information explosion as it related to the future course of human society or human survival. He saw that human survival could depend on the effectiveness with which the population related to the world's problems or solutions. Meadows (1974) has plotted

PAGE 41

30 the dimensions of time and space and demonstrated how every human concern can be located at a point on a graph. He felt life was a challenge for people, many of whom use total effort to provide daily for their families. Other individuals act on problems further out in time and space so the pressure they feel is of a community nature. Meadows (1974) observed that a person's use of time and space depends on his/her cultural orientation, immediate problems, and past experience. Before an individual moved into a larger space he/she must have solved the more immediate problems. He further concluded that the more difficult problems involved longer time commitments made by smaller numbers of people. The danger perceived by Meadows (1974) was that by limiting their perspective too much, individuals lose the ability to cope with problems at state, national, or global levels . McLuhan's (1967) global village concept accentuated the current use of a community of information. He saw the entire world knowing about other lifestyles through the perceptions created by television, movies, radio, and all forms of telecommunications. A hamlet in the Andes or a traffic jam in Los Angeles has been made the immediate experience of each individual. Drucker (1969) compared the closeness between continents today with the less relative closeness between mansions and slums during the eighteenth century. The aptness of the concept of a global village in terms of communication can readily be observed in current world events.

PAGE 42

31 Cultural Foundations of the Study The work reviewed in the preceding sections has traced the role of education in modern societies, identified the need for better understanding the role of values and perceptions in the learning process, and delineated evolutionary trends in the expanding mission of education to serve the needs of all groups in society. In this section we present a rationale for utilizing knowledge of learner attitudes and perceptions in the design and development of innovative learning networks. Human Cognition and Perceptions in Social Decision Making Personal construct theory (Kelly, 1966) has demonstrated the usefulness of looking at human and scientific endeavors as sharing relevant similarities in that in both, people pinpoint issues, observe issues, become intimate with problems, form hypotheses, test hypotheses, relate results to expectancies, control investments so they can understand what leads to what outcomes, carefully generalize, and broaden dogma in view of experience. Kelly (1955) considered man in this context and delineated the unifying concepts brought forth through explaining and charting strategies of human experience, both over individual lifespans, and over centuries. Kelly's (1977) unique contribution to psychology was the introduction of a single language for explaining human process making the psychological nature of scientific inquiry able to provide new insights into man's potential and the nature of science. Kelly (1955) did not claim to reach

PAGE 43

32 a total understanding of human process but he did feel the idea of manthe-scientist was worth exploring. Kelly (1955) assumed humans were real and not just existing in fantasy, that the human universe could be understood only through a time perspective, and that the universe was integral such that given complete knowledge and a wide enough viewpoint, all events would be seen as interrelated. Kelly (1966) further accepted that not only was the universe real but also that human internal events were also real. Therefore, an individual could come to understand his/her world only to the extent he/she could interpret it by moving toward an accurate axvareness of events through successive approximations. This theory bypasses the groundlessness and subjectivity of phenomenological or existential analysis and views people as able to test out their own constructs for completeness of world prediction. Kelly argued that human life events could be understood only as people acted on them in view of the present, past, and future. He saw man in a continual effort of construing relationships where none were observed before in an attempt to incorporate what was formerly diverse into a more integrated universe. This representational model of the world allowed humans, according to Kelly, to make sense from the world and choose behavior in relation to it. Therefore, Kelly (1966) refused to accept any once and for all construction of the universe because he viewed all current interpretations as subject to revision or replacement.

PAGE 44

3 3 In a further exploration of the characteristics and nature of constructs, Bannister (1968) demonstrated that a construct is a way in which some things are seen as alike and others as different. For each person the basis of difference can only be appreciated when the contrast is understood. The range of usefulness of an applied construct is also necessary for complete comprehension as it is possible to use similar discrimination while using different ranges of convenience. A construct is an interpretation imposed on events, not part of events themselves and thus constructs are useful inventions, not a part of nature. A construct was for Kelly (1955) a tool of discrimination and structuring of events in anticipation of future possibilities. Kelly (1966) suggested that a person could be understood by clearly seeing his/her construction system. Each person's constructs represented a network of avenues along which he/she can move. When movement was necessary, each person was presented with a number of dichotomous choices each channeled by a construct. Therefore, each construct represented a pair of rival hypotheses within a system. Consider, for example, the construct represented by the triad: automobile, airplane, and railroad train. A hypothetical respondent might suggest that airplanes are like trains and unlike automobiles in that the former are usually scheduled and the latter usually not. This construct clearly dichotomizes travel decisions according to scheduled and unscheduled arrangements. Note that the same triad

PAGE 45

34 could have led to an entirely different construct; e.g., one in which ground transportation was constrasted with more rapid air transportation. This illustrates the need for careful interpretation of the like-unlike axis elicited from the respondent. Each construct system limits a person's perceptions beyond which he/she cannot perceive and, therefore, controls his/her behavior. In relation to sociological cultural constructs, decisions arise from individual behavior and perceptions held by groups of people. Two individuals holding similar construct systems in both discrimination and range can be seen as having similar sets of constructs. A role of education is to create values and facilitate participation of individuals in their life. Education has traditionally provided information and identified desirable behaviors while ignoring the critical link provided between individuals of similar construction systems. Furthermore, if education could change perceptions or concepts, our society as a whole might be more effective in meeting the challenge of survival. An example of the usefulness of personal construct theory is in accounting for the difficulties we find in convincing other cultures to make good use of transported commodities such as surplus food grain. The range of convenience of their constructs regarding food may not include the particular material provided. A perhaps apocryphal example is the resistance of Central and South American Indians to the use of high-lysine corn in tortillas

PAGE 46

35 Their perception of quality corn was that it was yellow; thus they rejected the white imported corn even though it was nutritionally superior. Before education can deal with the development of useful constructs, it must map the perceptions and values of the target groups. Otherwise, it will be extremely difficult to enlist the active support of members of the target group who do not see relevance in the educational program goals. Many learning theorists have developed hypotheses with related models. Rogers (1970, p. 158) believed "significant learning takes place when the subject matter is perceived by the student as having relevance for his/her own purpose. Speed of learning is also influenced by relevance. Probably onethird to one-fifth of the present time allotment would be sufficient if material were perceived by the learner as related to his own purpose." Combs (1974) believed human beings had a natural ability to learn. He saw motivation in people to better understand themselves and their world. This fascination with learning, for experimentation, and extension of cognitive limits could be released under ideal environmental conditions according to Combs (1974) . Piaget (1973, p. 70) stated "the general culture which education is to transmit to the student cannot be restricted to abstract formation without roots in the structure and real life of the society as a whole but must consolidate the different practical, technical, scientific, and artistic aspects of social intercourse into a more organic whole." Piaget

PAGE 47

36 (1973) supported Kelly's (1966) observation that the constructs of each individual are real for that person and that the universe with its system of interrelationships is also real. Combs (1974, p. 126) stated "people do not behave in response to stimuli but, rather, to the meanings these stimuli hold for them." Kelly (1966) further described stimuli as constructs being individually formed and meaningfully used as guides to further behavior. Kelly's (1955) personal construct theory can be considered a metatheory or a theory about theories in that it accounts not only for the behavior of observed individuals, but also simultaneously it accounts for the activities of the observertheorist . Kelly's use of language to structure human approaches to understand events is unique. Kelly (Bannister, 1968) developed techniques for eliciting and measuring personal construct systems leading to his repertory grid as the most sophisticated. The idea that constructs are individual bipolar abstractions with precise ranges of convenience used for an individual's world structure is considered in the procedure for eliciting constructs. The significance of exploring and understanding an individual's system of constructs was recognized by Kelly (1955) in his elicitation of numerous constructs and his design of statistical techniques in which assessment of links among constructs was possible.

PAGE 48

37 The repertory grid technique of Kelly (1955) to be described more fully in the next section lends itself not only to the mapping of the construct fields of individuals but also to that of groups. Although architectural and urban planners are thus far the only non-psychologists to make use of the method, the limited data available reveal much promise for its application to a broad range of social research fields. Stringer (1976) discovered some important moral and political implications as well as technical and theoretical ones while working in the fields of planning and architectural design. He was placed in charge of a study to examine people's perceptions of alternative environmental aspects of planning proposals to rebuild a decaying Victorian shopping center in South London. In seeking general understanding of environmental value systems, he felt his success would depend upon not pre-empting the focus of inquiry too soon. A comparable set of environmental elements would enable respondents to give a clearer picture than one element or a disparate set. Stringer (1976) saw a range of responses as more richly definable by their observed relation to each other. He further envisioned the repertory grid as preserving individual construct systems by integrating perceptions and evaluations. Meaning was defined by the grid structure. The planning context provided a future orientation to a person's construing which conformed with personal construct theory. Public participation has been a potential rallying

PAGE 49

38 cry lately but it is in the planning field rather than education or social services that most examples can be found according to Stringer (1976) . He examined 200 individual grids to determine which redevelopment proposal drawn up in different map formats constituted what respondents considered "adequate publicity" for local planning. Leff and Deutsch (1972) did a pilot study using a modification of Kelly's (1955) repertory grid in which they requested individuals to apply a set of their own constructs to a set of environments known to them. Graduate students in architecture, urban planning or studies, and graduate students in other fields construed their physical environments. The two groups had significant organizational and content differences between the environmental verbal construct systems and between the results of this study and semantic differential studies. The results of the study suggested that there were two types of subjects in the professional group, ones who construed environments in terms of more qualities than lay persons and ones who construed environments in terms of fewer qualities. "The finding that professionals bracket lay persons suggests that there might be a cognitive source of interactional problems in that they are concerned with different numbers of environmental aspects" (Leff and Deutsch, 1972 , p. 289). The finding that some architects use fewer constructs than lay persons and non-architects could suggest that when professionals with backgrounds interact with lay persons most probably

PAGE 50

39 conflict will occur because the lay persons most probably are not concerned with as many environmental aspects. The finding also suggested architects will have problems interacting with environmental planners and designers. Alexander (1964) saw that professionals are socialized by their educations to think about environments very differently than non-professionals and that such a cognitive gap could be a major hindrance to cooperation. He saw cognitive differences translated readily into interactional problems when groups think differently about problems. Hershberger (1969) carried out a study to compare the environmental cognitions of students in architecture and other fields using the semantic differential method. Even though the study did show group differences the semantic differential method required all students to use the same set of scales provided by the experimenter and as a result interindividual differences were masked. Furthermore, associations at the aggregate level did not necessarily reflect relationships existing at the individual level (Robinson, 1950). Repertory Grid Methods The work described in the previous section has demonstrated the utility of personal construct theory and the repertory grid method derived from it for describing the perceptions of individuals and groups. In this section will be described the methodology used in relevant work and the appropriate statistical measures will be discussed.

PAGE 51

40 Certain essential components of the repertory grid and its conventions should be considered first (Ryle, 1975). Among them are a list of tilings to be compared, the elements, and a list of terms used to compare and contrast them, the constructs. The first list was compared systematically against the second list to create a grid of figures which is where the name "repertory grid" comes from. Constructs are bipolar according to Kelly (1966); therefore, constructs should be elicited in the bipolar form. The subject has to have a list of elements which can be compared to a list of constructs and vice versa. A researcher's first responsibility in carrying out the repertory grid is to elicit from the subject a list of elements and constructs. The more subjects provide their own elements and constructs the more they reveal about themselves according to Kelly (Ryle, 1975). The first stage in constructing a test is to put together a set of elements. The next step is to obtain constructs. The more freedom that is allowed the respondent the more valid their constructs. The classical method has the tester randomly choosing sets of three elements from the element list and asking the subject to describe all the similarities and differences in the triad. The tester then writes down these descriptions. Triads are used until no new constructs are formed. Three elements are considered rather than two based on Kelly's (1955) stand on the bipolar nature of constructs and the fact that this is the minimum to allow for

PAGE 52

41 obvious similarities and differences and to define both poles of the construct. Constructs are recorded in each subject's words unless constructs are supplied to describe often used judgments. If the respondent did not feel comfortable with the supplied construct, it could be modified. The tester now has a list of elements and a list of constructs from the subject. To complete the test a respondent is asked to rate each element with each construct by dichotomising, by ranking, or by rating. The form which the test thus takes is the form simply of rows and columns. The form of the grid described can be called a "standard grid." Other modifications can be made as required by the mathematical analysis of the repertory grid. Let us consider the following hypothetical case in which a graduate student is asked to respond to the following elements each of which is a potential topic for a doctoral dissertation: 1. Correlation of Myers-Briggs Extraversion Scales with height. 2. Comparison of bilingual and monolingual children on performance in a new language. 3. Correlation of Myers-Briggs thinking, feeling category with computer literacy. 4. Use of computer-aided instruction to teach mathematics in the middle school. Our hypothetical respondent is asked to create triads from the above elements. In each case, one pair of elements

PAGE 54

43 form of analysis and display is required before any conclusions can be drawn. The purpose of mathematical analysis of a grid is to make obvious the structure in the grid, not to impose the experimenter's expected structure upon it. If each point of the grid represents an independent judgment of the respondent, there can be no structure to reveal. Ryle (1975) stated that all grids created by humans are characterized by visible structure such that there are relationships between individual ratings which are open to mathematical analysis and which also relate to psychological process. Complete analysis of essential properties of a grid (or the relationships between elements and interaction of elements and constructs) demands, according to that author, computer analysis. Bannister and Mair (1968) described methods which can easily be applied to grids based on dichotomization of elements. Computer analysis also allows for examination of element relationships and construct-element interactions which are only fully explored by such an analysis. Problems of centrality versus extremity of ratings and a skewed distribution of elements on constructs can also be addressed. Two elements receiving precisely the same rating on every construct must be perceived by the subject as the same or indistinguishable in terms of constructs chosen. Two elements rated at opposite ends of a construct must be perceived as highly dissimilar. Similarity of any two elements can be estimated by the measure of the distance between any

PAGE 55

44 two elements and also the degree to which they are related as similar in relation to the constructs used. Ryle (1975) treated analysis of constructs as similar or different in the same way as the elements. The association between two constructs based on all the elements is given with a value of -1 or +1 in the table of construct correlations. Assuming enough elements have been rated against constructs, the correlations between constructs could understand his world. Knowledge of these could help the researcher explain future behavior. A graph can be created from an analysis of elements to represent conceptual space in which the meaning is indicated by the constructs and the location of each element in relation to these constructs and to the other elements. Summary The literature review of this section has demonstrated the need for educators to accept a role in the development of society to serve nontraditional learners outside of school settings. Very little is known about the values and perceptions of these nontraditional learners with regard to a variety of social issues which education will need to address. In fact, very little is known about methodology for acquiring information about values and perceptions in a reliable way. In limited studies done by architects and planners, the repertory grid method of personal construct theory appears to have much promise for mapping values and perceptions of individuals and groups.

PAGE 56

45 This study demonstrates its utility for such a mapping with regard to energy issues and suggests ways the results can be used to plan curricula for energy education activities.

PAGE 57

CHAPTER III PROCEDURES Kelly's theory of personal constructs and recent applications of its associated repertory grid method to map environmental perception have provided a potentially powerful tool for analyzing and interpreting group perceptual structure. Any technique for mapping complex perceptions is itself necessarily complex because of the many decisions involved in its application. The construction, administration, and analysis of the repertory grid instrument are described in this chapter. The makeup of respondent groups and the reduction of the acquired data into forms suitable for hypothesis testing are reported. Finally, the statistical procedures for analyzing the data and testing the hypotheses are described Selection and Identification of Respondent Groups The rationale for group selection was to find groups of individuals who might be expected a priori to be different in energy values and perceptions. Further, a group was selected to represent a community sample with no prediction made about member's homogeneity of values and perceptions. A class of senior education majors at the University of Florida was selected to represent a group with sociological sophistication but limited technological expertise. 46

PAGE 58

47 To represent a group with technological and energy expertise, a class of senior mechanical engineering majors at the University of Florida was selected. Community members were sampled by soliciting responses from participants in two classes at Santa Fe Community College which were known to be heterogeneous in terms of ages and occupations One class consisted of individuals participating in a "course by newspaper" on energy; the second was made up of persons studying radio-TV repair. Valid responses were acquired from 28 engineering students, 3 2 education majors, and 14 community students. The average age of the groups was 24 for the engineering students, 25 for the students in education, and 31 for the community group. Table 1 summarizes the demographic information available for each group. Appendix B lists majors and/or occupations for the groups. Design and Administration of Repertory Grid Instrument The instrument used to elicit repertory grid information is found in Appendix A. Respondents were asked to list factors they believed were important in understanding the energy problem. Several forces were implicit in the design of this form. First among these was the decision not to provide respondents with a list of elements, but rather to permit free-form generation of element lists by each individual. Providing a list of elements forces respondents to limit their thinking to the researcher's supplied scope. The eliciting of free-form lists, on the other hand, makes

PAGE 59

48 in

PAGE 60

49 direct comparison of individual grids impractical. This latter difficulty can be dealt with, in both concept and practice, by a procedure which will be demonstrated later in this section. The choice of requiring exactly 20 element responses was made for the following reasons: 1. It was necessary to have uniformity of element list size in order to meaningfully compare the resulting construct list sizes. 2. Leff and Deutsch (1972) found 20 elements to be a workable number of elements in their study. 3. Twenty three-digit element identification fields leave a reasonable amount of space on an 80-column IBM card for demographic and identification information. Instructions for the first part of the response were made as brief as possible. Subjects were asked to "List 20 factors you believe are important to understanding current world energy problems." No attempt was made to define "factor" nor was an energy related example provided. Individual responses ranged from single words such as names of oil companies to complex relational phrases such as economic or political influences. In most cases individuals were able to complete this listing in 20 to 30 minutes. Below is a sample of a single individual's response. Note that two numbers appear to the left of each response. The first is the assigned element code from the categorical list of Appendix C which is described in the next section

PAGE 61

50 of this chapter. The second is just the number of the element in the list. The category number was assigned by the researcher on two criteria. The first was the meaning of the listed element. The second was derived from examining the grid column of construct associations to achieve contextual understanding of the element's meaning. For example, note that element 8 "rising gas prices" could have been recorded as a sociopolitical factor, a resource/distribution factor, or an economic factor. Column 8 of the association grid below, however, shows positive associations with constructs B, C, and E, all of which suggest economic dimensionalities. Negative associations are recorded for constructs A and D which suggest resource and technology dimensionalities. Thus the element was assigned to category 80 "economics." Note that the underlying structure of the grid would have been presented equally well had the category assigned been 30 or 31, providing all other similar responses by other individuals were treated consistently. This is true because the purpose of the study was to analyze the response structure of the groups rather than to interpret the precise meaning of the responses. Provision was made for the elicitation of from to 26 constructs on the form. In the study reported by Deutsch (1972) a maximum of 29 responses were found. It was anticipated that because the list of elements in this study was not provided, fewer constructs would be produced. This was, in fact, the case with 17 being the maximum number from a

PAGE 62

51 single individual. Each construct was defined by information of two types: the triad of elements which provided the poles of the dichotomy, and the organizing concept ("How") which provided the dimensionality of the dichotomy. Thus, the instrument could distinguish between constructs using the same triads, but different dimensionalities. For example, "car A is like car B and unlike car C." In one case the dimensionality might be: "Because A and B are sports cars while C is a station wagon"; in another "Because A and B will run while C will not run." No responses were considered valid for constructs if both triad and dimensionality was supplied. Survey of Energy Problems I. List 20 factors you believe are important to understanding current world energy problems. 73 1. Arab oil problems 72 2. Nuclear energy processing 74 3. Gasohol 80 4. Rising electric bills 68 5. Cleaning up coal use 06 6. Windmill generators 42 7. "Turn out the lights" syndrome 80 8. Rising gas prices 69 9. Use of natural gas 74 10. Use of geothermal energy 06 11. Use of solar energy 30 12. 55 m.p.h. speed limit 34 13. World politics 73 14. Third world's desire for energy 21 15. Rising rate of energy consumption/per cap 72 16. Difficulty of finding new oil 30 17. Emergency building temperature restrictions 52 18. Building insulation 51. 19. Car pooling and car pool lanes 80 20. Increased costs of all transportation

PAGE 63

52 An example of the instructions and response to the construct section of Part II using the same single individual as in Part I follows: II. Find pairs of factors which are alike. For each pair find a third factor that is different in the same way as the pair is alike. Continue to do this until you can find no more contrasts. (Most people will have more than 5 and fewer than 25 contrasts.) For your convenience, simply record the numbers of the factors in each contrast in the blanks provided below. In the space alongside each triplet briefly describe how the factors are similar or different. A. 10 B. _1 C. 12 D. _6 E. 4 Alike Different How 11 _5_ 018 Ways to find new energy 031 Politics of the situation _3 17 11 20 15 18 028 Federal mandate 18 018 Still experimental 009 Price-to-pay increase Most respondents required 20 minutes or so to complete this portion . Constructs were examined for triad content. First, the triad was recorded as in the element list. If a particular element code occurred twice other than in the alike pair, the element coding was reexamined. The original element triad was looked at, and the dimensionality axis ("How") applied. The construct was then assigned to a number on the construct category list. This list appears as Appendix D. If an appropriate category did not appear on the list, one was created and a new number was assigned.

PAGE 64

53 For example, construct C above has the original elements of "55 m.p.h. speed limit" (12) and "emergency building temperature restrictions" (17) contrasted with "building insulation" (18). The dimensionality ("How") is "federal mandate." What this is interpreted to mean is "12 and 17 represent politically-motivated conservation requirements while 18 is a conservation measure which is not required politically but may be of value." Construct number 28 which has the dimensionality "conservation versus politics" was assigned to this response. In filling out the association grid, the choice was made to seek only a + or a response for each constructelement association. This was done for reasons of expediency as well as for theoretical reasons. The use of a rating scale to estimate the degree of association (relevance) between each element and each construct would be useful in observing the evolution of attitudes of an individual over time. In this case, however, it was felt the time required for filling out the grid would be inordinate for an individual who had developed possibly 20 constructs. Most persons were able to complete the grid in 15 to 20 minutes. An example grid for the above elements and constructs follows Interpretation of the simple dichotomous responses elicited is relatively easy. "Construct X relates in some way to element K" (+) or "Construct X does not relate in any I think significant to element K" (-) . Note that the latter interpretation is difficult to distinguish from "I didn't

PAGE 65

54 III. Look at the grid on the page. Each square in the grid has a small letter and number on it. The numbers refer to your list of factors from Part I and the letters to your list of contrasts from Part II. Start with box Al . Does contrast A apply to factor 1? If so, place a + in box Al ; if not, place a in box Al. Repeat the process for box A2 checking whether contrast A applies to factor 2. Continue until you have checked your contrasts against all 20 factors. Factors 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 C _ + __ + __ + _-+ --+ + + D _ + + _ + + ___ + + _--+ -+ E + + + + + + + + + + + + + + + + + + + + consider element K (construct X, or both) very important, and therefore, didn't mention them." This point is conceptually central to the use of individual grids to study group values and perceptions. It is on this basis that we can merge non-overlapping element and construct responses into group grids as discussed in the next section. Data Reduction and Analysis Inspection of the received responses revealed that there was a great many repetitions of elements and, to a lesser degree, of constructs. In many cases, generic elements were represented by specific examples; e.g., Texaco, Exxon (oil companies) and Iran, Saudia Arabia (O.P.E.C). Both these phonomena were anticipated, and in fact, necessary for meaningful analysis to occur.

PAGE 66

55 Because of the redundancies, it was possible to compress the approximately 600 element responses and 250 construct responses per group into an orderly and manageable group grid. To accomplish this, a taxonomic listing of constructs was developed and each individual's responses were recoded according to that taxonomy. The taxonomic listing is found in Appendices C and D. The more complex constructs were less easily compressed by categorization; nonetheless, the same approach was used to categorize these responses. A listing of construct categories is found in Appendix C. Care was taken to retain the original meaning of the responses. Thus, tests were designed to ensure minimum information loss in the recoding of the responses. To assure fidelity in the recoding of elements, recoded triads were examined for cases in which a single element occurs both as a member of the like pair and as the unlike element as was described in the previous section. In such cases, the coding was reexamined to ascertain if a better assignment could be made. If not, a subdivision was created in the taxonomy so as to preserve the degree of discrimination in the original response. After recoding, less than 1 percent of element responses contained the same element in more than one occurrence. Consistency of interpretation of constructs was assured by inspection of the row vectors of associations. In cases

PAGE 67

56 where a single individual used the same construct two or more times, the grid scores of each occurrence were compared. If more than two grid cells were scored differently, the constructs were reexamined, and, if necessary, reassigned. A further check was the inspection of combined scores on group grids to assure non-randomness of scoring distributions across members within a group for a single construct. The scoring of individual grids was merged into group grid scores. While individual grids were no larger than 20 x 17, the group grids included all constructs found in the group and thus were grids of 20 x 81 for the engineering class, 20 x 77 for the education class, and 20 x 44 for the community group. It is from the group girds that the intercorrelation matrix of construct correlations was calculated to be placed in the SPSS factor analysis routine. Ordinarily, one would provide raw group data to the SPSS factor analysis package. This program, in turn, calculates the correlation coefficients between all construct pairs and constructs the correlation matrix (Nie et al., 1970) . In this study, however, the raw data were submitted to several purely data processing steps to permit checking of group response consistency prior to calculating correlations coefficients. A second reason for computing correlations coefficient matrices for input to the SPSS routine is that this calculation, which needs to be done only once, is the most expensive calculation in the routine. By inputing

PAGE 68

57 the correlation matrices, a variety of alternative computer runs could be achieved relatively inexpensively. Statistical Procedures and Tests of Hypotheses The data acquired from the respondent groups were analyzed and statistics calculated to test the four hypotheses stated in Chapter I on page 12. It was anticipated that different groups would focus on different aspects of energy concerns. One indication of such a difference is the range of element responses elicited from a single group. A hypothesis which can be easily tested is that three groups would each contribute the same categorical elements with the same relative frequency. The element frequency distributions were printed out by the computer program and appear in Chapter IV. Note that this hypothesis can be tested by superposition of normalized frequency distributions. It is not reasonable to believe that reference to a particular element would be a randomly distributed variable in the population as a whole. On the contrary, Kelly's theory suggests that the element list an individual creates is very much the result of their way of construing reality. This individual set of characteristics is in turn a product of a person's experience and learning. Consequently, it is expected that common citation of elements should occur in a group with some commonalities of experience and training. It would thus be inappropriate to attempt to compute the usual parametric test statistics and confidence intervals.

PAGE 69

The data reported in Chapter IV do, in fact, support the above prediction as will be described in detail later. The number of constructs generated by individuals in a group is a measure of the dimensional complexity of thought in that group. The more constructs reported, the more ways the respondent has of organizing perceptions. Consequently one would wish to examine group differences as to the numbers of constructs generated by their members. A convenient testable hypothesis is that the group means are equal to each other. The implication of this hypothesis is that its acceptance suggests all three groups are representatives of a single population. This hypothesis was tested by a one-way analysis of variance and the results discussed in Chapter IV. The underlying assumption of repertory grid analysis is that the pattern of element-construct associations reflects the structure of the respondent's perceptions. An individual who sees the world as a series of relatively isolated events would be expected to have few positive association responses. Thus the grid of such an individual would be expected to show large numbers of (-) responses and few (+'s). Conversely an individual who perceived a great deal of interconnectedness would be expected to report many positive associations. Such a person's grid response should reflect a preponderance of (+) answers. As a hypothesis susceptible to testing, it was proposed that the percent of (+) responses of individuals in

PAGE 70

59 each group were equal. This hypothesis was tested by a one-way analysis of variance and the results are presented and discussed in Chapter IV. If it is desirable to assess the degree of interrelatedness of perceptions of individuals, it is difficult to interpret the structure of highly associated perceptions. Interpretation of the structure of group perceptions becomes almost impossible from simple inspection of grids. Fortunately, methodology exists which can combine groups of variables into composite variables called factors which permit simplification of structural complexity. This methodology is termed factor analysis. Factor analysis, as used here, proceeds through three stages: computation of correlation matrix, interactive estimation of communalities , and extraction and rotation factors. Each of these steps and their associated analytical options will be discussed in turn. In the simplest form, factor analysis attempts to clarify structure in a set of data by creating linear combinations of the variables involved. The proportion of the group variance accounted for by these combinations can be computed and, if the linear combinations are uncorrected (orthogonal), their relative significance is estimated. The SPSS factor analysis routine operates on a correlation matrix to find the single linear combination of the variables present which maximally accounts for the total group variance. This is called the first factor. A second

PAGE 71

60 combination is then sought which will meet two criteria: it must be orthogonal to the first factor, and it must account for the maximum amount of the remaining variance. A third factor is then generated with the same criteria, i.e., orthogonal to the first two factors and accounting for the maximum amount of remaining variance. This process continues until there are as many factors as there are variables . The process described above suffers with regard to interpretability of the factors derived. The correlation matrix is generated by calculating the Pearson product moment intercorrelation between the pooled rows of construct pair responses. This generates a square matrix of construct by construct correlations with l's on the diagonal. If we are looking for structure in the matrix, we can focus on the significance of the correlations. The diagonal row of unit correlations reflects the fact that each construct has reported for it both variance common to the whole group of responses, and unique variance due to its own nature (Harris, 1975). It is the former variance in which we are interested. This quality is called the communality of the variable. If accurate estimates of communality could be used to adjust the matrix, then the factor scores would be directly interpretable as reflecting structure. The program used in this study accomplishes this communality estimation by an iterative process (Nie et al . , 1970). On the first pass, the l's are replaced by the

PAGE 72

61 single highest correlation in the column and the values of all other correlations adjusted. Then the diagonal values are replaced with the highest correlation in each adjusted column. This process is repeated until there is no change in the communalities between two passes. Factors are then generated by the program until they no longer explain a significant amount of variance. In the option selected for this study, only factors with eigenvalues of one or greater were printed. An eigenvalue of one is equivalent to the normalized variance of one of the original variables. Thus each factor printed is at least as useful in accounting for group variance as any one of the original variables . The interpretation of the factors can proceed at this point, but visualization of the significance of the factor loading scores is difficult. A rotation algorithm is provided by the program to produce factors with loadings as close to or 1 as possible for each variable. The program performs rotations of pairs of factors, retaining orthogonality. If the rotation results in one or both of the factor axes becoming colinear with a structural dimension of the data, the projection of the data points on the other axis becomes as does the factor loading score. This is easily interpretable ; the colinear axis has no influence on the variance in the data. The program performs rotations maintaining orthogonality and maximizing the numbers of O's and l's in the

PAGE 73

62 factor loadings, thus generating more easily interpretable factors. Further, it prints the distribution of data points along each pair of factor axes. This information is discussed in detail in Chapter IV and program output is found in Appendix E.

PAGE 74

CHAPTER IV RESULTS This study was designed to utilize data from three groups of college students to demonstrate the use of repertory grid methods in mapping values and perceptions concerning energy issues. Two of the groups-education majors and mechanical engineering majors--were relatively homogeneous with regard to age, sex, and educational background. The third group was made up of individuals taking a course by newspaper on energy offered by Santa Fe Community College and of persons taking a course in radiotelevision repair. This last group was quite heterogeneous in composition (for more detail, see Chapter III). Repertory grids were elicited, constructs and elements categorized, and individual responses merged into group data. Pearson product moment correlations were calculated between constructs and the intercorrelation matrices subjected to factor analysis by means of the SPSS subprogram FACTOR. Average number of constructs, means and standard deviations, and one-way analysis of variance were performed for each group using the SPSS subprogram ONEWAY. This program was also used to compare number of positive associations per construct for each group. Frequency of citation was calculated for each construct and element for each group. 63

PAGE 75

64 The products of the computations described above were utilized in testing the following hypotheses: 1. The same elements and constructs are cited frequently by different groups. 2. Groups generate the same number of constructs per response. 3. Groups generate the same number of positive associations between elements and constructs per response. 4. Groups generate the same factors. Each of these hypotheses will be analyzed, tested, and discussed in turn. Frequency of Element and Construct Citation As might be expected, not all elements and constructs occurred with similar frequencies in grid responses. In fact, a few elements and constructs were very frequently cited by members of a particular group, others were cited much less frequently, and most were cited rarely. Tables 2 and 3 show the ten most frequently occurring elements and constructs for the three groups. It is clear from the tables that there is substantial agreement among the groups on the importance of certain elements. Five elements—numbers 34, 41, 70, 80, and 81--appear in all five groups' listings. A sixth — number 30--occurs in the top ten for the engineering and education groups. It is interesting to note that these elements vary widely in content. Elements 30, 34, and 41 relate to social and political dimensions of the energy problem, while elements

PAGE 76

65 Table 2 The Most Commonly Cited Constructs by Groups Construct Cited Meaning Engineers 1. 18 59 Conventional Technology/Alternate Technology 2. 87 44 Resources/Population Growth 3. 44 43 Conservation/Public Attitudes 4. 74 42 Education/Resources 5. 10 36 Economics Cost/Public Needs 6. 106 35 Politics/Industry 7. 9 34 Economics Cost/Conservation 8. 45 30 Conservation/Consumption 9. 60 29 Attitudes/Planning 10. 74 28 Pressure for/against International Welfare Teachers 1. 9 104 Economics Cost/Conservation 2. 18 75 Conventional Technology/Alternate Technology 3. 31 72 Politics/Resources 4. 67 64 Individual Behavior/Public Politics 5. 7 49 Economics Cost/Available Resources 6. 28 48 Conservation/Politics 7. 75 47 Education/Resources 8. 44 43 Conservation/Public Attitudes 9. 11 34 Economics Cost/Policies 10. 5 34 Economics Lifestyle/Resource Limit ~ Community Groups Attitudes /Planning Conservation/ Regulation Conventional Techno logy /Alternate Technology Technologypractical /unreal is tic Government/ Population Conservation/ Consumption Energy production/Consumption Credibility of Information/Attitudes Energy Needs/Luxuries Economic Factors/Technological Forces 1.

PAGE 77

66 Table 3 The Most Commonly Cited Elements by Groups

PAGE 78

67 70, 80, and 81 are associated with resources, economics, and industry. Surprisingly, none of the groups was strongly aware of technical factors such as energy efficiency, emerging technology such as fusion, or radical lifestyle change options as determined by frequency of citation. Construct citations can be interpreted as indicative of group values. Since constructs are both dichotomous and polar, there is a "preferred" pole implicit as well as an "undesirable" emergent pole. Recall the example of Chapter II "approved by the dissertation chairman." The preferred pole is clearly the approved pole, while the emergent pole is the not approved pole. The data in Table 2 reveal that only one construct, number 18 (conventional vs. alternate technology), appears among those cited most frequently by all three groups. One interpretation of the combined frequencies is that while members of all three groups perceive similar factors as relevant to the energy problem, they place distinctly different constructions on those factors. The implication of this is that the groups hold different values and priorities regarding this problem. Numbers of Constructs Elicited Since each respondent provided the same number of elements (20), it is possible to compare the numbers of constructs produced from the element sets and attach significance to the result. The number of constructs produced is indicative of the complexity which the respondent perceives in the area under study. An individual showing few constructs thus

PAGE 79

68 can be said to have a relatively simplistic view while an individual who responds with many constructs either views the area as highly complex or perceives a variety of unrelated bits and pieces of observations. This ambiguity will be resolved below. Table 4 presents the mean numbers of constructs for the three groups along with ranges and standard deviations. In this table are also found the analysis of variance data for numbers of constructs. Note that the F value is significant at the 0.0000 level. Thus we can reject the null hypothesis that all group means are equal with less than a 0.01 percent chance of error. To determine whether or not all pairs of means are significantly different, a least significant difference follow-up test was performed. This particular follow-up was selected because of its exactness for uneven cell sizes. Although it suffers from multiple levels of criterion variables, this particular application with only three levels of group membership is immune from that weakness. The table reveals that all pairs of means are unequal at the 95 percent level of significance. It is thus possible to state with confidence that insofar as number of constructs are concerned, the three groups examined are significantly different. Degree of Construct Association It was noted above that a relatively large number of constructs could be interpreted in two ways. One means of resolving this ambiguity is to achieve a measure of the

PAGE 80

69 P! o u m o P O < &

PAGE 81

70 "connectedness" of the respondents' perceptions. It will be recalled that each construct was scored against all elements for association. For strict consistency, only three associations are required as a minimum for each construct. Positive associations in excess of this minimum number are indicative of a perception of connectedness by the respondent. It is this feature of the grid that permits calculation of the construct intercorrelations . An alternative measure is to determine the average number of positive associations per construct for each individual. To the extent that persons report positive associations, they indicate the degree of interrelationships they perceive. This degree of interrelationship perceived is useful in designing learning models. A more complete discussion will follow in Chapter V. Table 5 also reflects results of the one-way analysis of variance for this parameter. The F ratio is indicative of significance at a level of 0.0030. The null hypothesis that the group means are equal can thus be rejected with only a 0.3 percent probability of error. The LSD follow-up tests show that all pairs of means are different at a confidence level of 95 percent. From the information displayed in Table 4, we can then conclude that (1) all groups show a degree of interconnectedness in excess of the minimum required for definition of constructs, and (2) the groups are all significantly different with regard to the degree

PAGE 82

71 x H +-> u f-i 4-1 t/1 a o u Oh o 03 EO < &

PAGE 83

72 of interconnectedncss of perceptions demonstrated. The implications of this observation for curriculum development are discussed in Chapter V. Results of Factor Analyses of Grids Table 6 lists the unrotated principal components (factors) for the engineering student group, the eigenvalue associated with each factor, the percent of the grid value accounted for by that factor, and the cumulative variance accounted for. Because the group grid is structured, only 41 factors are required to account for 100 percent of the variance. Had there been no structure, all 81 potential factors would have been required. Note that the first factor accounts for far more variance than any other. Tables 7 and 8 display the same information for the education student and community groups, respectively. Again, perceptual structure is indicated by much of the variance being accounted for by relatively few factors. The amount of variance accounted for by the first factor can be interpreted as a measure of centrality of thought, that is, the degree to which a single dimensionality is perceived to structure the perceptual field. Note that on this criterion engineering students and community members score similarly while education students are significantly different. Tables 9, 10, and 11 display the factor structures for the first ten factors for each group. Each row of these tables can be thought of as the row of regression coefficients for predicting the score of one of the variables (constructs)

PAGE 84

7 3 TAP4.E 4 SUflllARY OF FACTOR STRUCTURE FDR ENGINEERING STUDENT GROUP CONSTRUCT

PAGE 85

74 TftELE 4 SUmtoRY OF FACTOR STRUCTURE FDR ENGINEERING STUDENT GROUP CONSTRUCT

PAGE 86

75 TABLE 7 SUWIftRY OF FACTOR STRUCTURE FOR EDUCATION STUDENT GROUP HSTRUCT

PAGE 87

76 TABLE 7 SUMARY DF FACTOR STRUCTURE FDR EDUCATION STUDENT GROUP CONSTRUCT ESTIMATED FACTOR EKEJWftLUE FERCENT OF CUflULATIUE CDtttlUNALITY VARIANCE PERCENT C084 0.82333 id 0.00 WO 0.0 100.0 C087 0.41405 41 0.00000 0.0 100. C088 0.41020 il 0.00000 0.0 100.0 COCT 0. 41400 61 0.00000 0.0 100.0 C011 0.72281 64 0.00000 0.0 100.0 COTJ O.SO?^ 6"5 0.00000 0.0 100.0 C014 0.82158 66 0.00000 0.0 100.0 Cm 0. 82611 67 0.00000 0.0 100.0 C017 0.6-5128 6S 0.00000 0.0 100.0 C018 0.824166"1 0.00000 0.0 100.0 C011 0.76"787 70 0.00000 0.0 100.0 C101 0.6106-5 71 0.00000 0.0 100.0 C102 0. 55,511 72 0.00000 0.0 100.0 C103 0.826-11 73 -.00001 0.0 100.0 C104 0.81301 74 -.00001 0.0 100.0 C105 0.55554 75 -.00001 0.0 100.0 C106" 0.4511676" -.00001 0.0 100.0 C110 0.86-501 77 -.00002 0.0 100.0

PAGE 88

TABLE 8 SUttflARY OF FACTOR STRUCTURE FDR CDHflUKITY GROUP 77 CONSTRUCT

PAGE 89

TABLE 1 UMROTATED PRIHCIFftL CDHPDHENTS FDR EHGIHEERIHG STUDENTS CONSTRUCT 123^5
PAGE 90

79 TABLE 1 UNRDTATED PRIHUIPftL CSflPOSEHTS FOR ENGINEERING STUDENTS CONSTRUCT 1 2 3 4 5 6 7 8 1 10 zm

PAGE 91

80 TABLE 10 UNROTATED PRINCIPAL COHPQHEHTS FDR EDUCATION STUDENTS CONSTRUCT

PAGE 92

TnBLE 10 IMITATED relHCIM. COrtP'JfOTS FCR EDUCiTIDN STUDENTS CONSTRUCT

PAGE 93

82 TittLE tl UHRQTATED PRINCIPAL CDfiPDXENTS FDR CDttflUNITY GROUP CONSTRUCT

PAGE 94

83 from the factor score. The square of each coefficient then estimates the proportion of the construct score variance accounted for by the factor score. Note that these tables have been adjusted for communalities ; i.e., the variance accounted for is common to the entire grid structure, not unique to a particular variable. Thus the factor structure matrix displayed in the tables is a structural depiction of the innercorrelated variance. Further, variables having a relatively large factor score coefficient are relatively important in determining the meaning of the factor. To maximize the possibility of distinguishing important factor dimensionalities, Varimax rotation was performed on the factors. This transformation forces the coefficients to approach zero or one, making it easier to distinguish heavy from light loadings. The Varimax rotated factors are displayed in Tables 12, 13, and 14. The most significant variables were extracted by ranking the coefficients in decreasing order of magnitude and retaining those which accounted for the first 50 percent of the factor eigenvalue. Note that for the first factors, ten or more variables were found as identified by circled coefficient values in the tables. For subsequent variables, only a few significant variables are required to account for 50 percent of the eigenvalue (thus 50 percent of the variance) of the factor. This is consistent with the literature which reports that first factors tend to be quite general, loading heavily on many variables, while subsequent factors tend to be bipolar,

PAGE 95

84 TABLE 12 VARIHAX ROTATED FACTOR tIATRTX FDR EKGIHEERIHG STUDEKT GROUP CDHSTRUCT

PAGE 96

TABLE 12 UfiRIflAX ROTATED FACTTJR HATRIX FOR ENGINEERING STUDENT GRTJUP CONSTRUCT 12345^78^10 C013 0.140 0.30M 0.004 -0.035 -0.173 0.031 0.244 0.411 0.483 0.071 COm 0.021 0.304 -O.oe -0.023 0.044 0.044 0.401 0.044 0.421 -0.007 C015 0.378 -0.342 0.054 0.147 0.125 0.181 -0.030 0.438 0.303 0.221 C017

PAGE 97

86 TABLE 13 VARUM ROTATED FriCTDR ilATRIX FDR EDUCATIDH STUDENT GSBUP CONSTRUCT

PAGE 98

87 TABLE 13 VARItlAX ROTATED FACTOR ilATRB FDR EDUCATION STUDENT QRQUP :tINSTRUCT 123L,5
PAGE 99

88 TABLE 1M VfiRIHfiX RGTATED FACTOR HATRIX FDR COrtliUHITY GROUP CONSTRUCT 123M5478R10 C002

PAGE 100

89 loading heavily on only a few variables. These combinations of variables are, of course, more easily interpreted than the more general first factors (Nie et al . , 1970). The first ten factors for each group are listed in Tables 15, 16, and 17 along the significant constructs determining their dimensionality. Note that similar dimensionalities occur across groups with the appearance of a resource dimension, a politicotechnological dimension, and a politicoeconomic dimension most prevalent. It is possible to locate individual variables (constructs) with regard to these axes of dimensionality by use of the factor loading coefficients from Tables 12, 13, and 14. Figure 1 displays the location of the most frequently cited construct across groups (construct 18 "conventional vs. alternate technology") for each of the three groups on the resource X politicotechnological plane. Figure 2 displays the location of this variable on the resource X politico-economic plane, and Figure 3 locates construct 18 on the politicotechnological X politico-economic plane for each group. What is clear from the figures is that using similar dimensionalities and constructs, the different groups perceive the energy problem (at least in this regard) differently. Strong loadings on a particular dimensionality can provide important clues to the value placed on a particular construct. The implications of this interpretation for curriculum development are explored in Chapter V.

PAGE 101

90 Table 15 Major Factors for Engineering Student Group 1. Politico-economic Population growth International relations Conservation vs. international relations Economic costs Politics vs. research Economics of public needs 2. Socio-political Economics of waste Attitudes toward planning Alternate technology, resources Consumption by transportation Attitudes vs. resources 5. Socio-economic Attitudes toward technology National vs. foreign policies Environmental vs. economic policies Energy conservation vs. waste 4. Resource-Consumption Resource policies Resource economics Politics vs. resources Energy consumption and waste 5. Expectations Historic consumption Economic expectations vs. limits Attitudes toward credibility 6. Conservation policies Conservation in transportation Conservation, public vs. private 7. Appropriate Technology Technology, individual vs. group Economic-costs of policies Alternate technology and resources 8. Politicotechnological Technological vs. social forces Resource distribution, global International relations and industry Energy productions consumption 9. Attitudes Attitudes and policies Social vs. special interests 10 Conservation Industry resource consumption Conservation and resources

PAGE 102

91 Table 16 Major Factors for Education Student Group 1. Multidimensional International welfare pressures Resource limits and consumption Individual vs. group technologyEconomics of industry ethics Urban vs. rural economics Education vs. resources Politics vs. resource consumption 3. Policy Planning Technology and long-term planning Alternate technology and resources Resource distribution, global Risks of technological development 5. Technology Technology and industry Soft vs. hard technology Risks of technological development 7. Resource-Policy Resources vs. transportation Attitudes and international relations Socio-economic Economics vs. technology Energy needs vs. luxuries Technology vs. conservation 2. Politico-economic Alternate energy and international relations Politics of conservation Technology and resources Economic cost of policies Efficiency of energy production 4. Conservation Public attitudes and conservation Attitudes and planning Politics vs. technology 6. Resource-Technology Transportation consumption Technological vs. social forces Expectations vs. limits 8. Socio-technological Renewable vs. fossil fuels Technology and consumption Energy conservation and alternate energy 10. Attitudes Public attitudes and technology Alternate technology vs. resources Conservation and international relations Attitudes, domestic vs. international Population growth and resources

PAGE 103

92 Table 17 Major Factors for Community Group 1. Politico-technological Politics and industry Attitudes and planning Industry and resources Politics and alternate energy Energy needs vs. luxuries Economic vs. technological forces 2 . Socio-economic Social vs. technological forces Risk of technological development Renewable vs. fossil fuels Pollution vs. resource consumption 3. Regulation Politics and kickbacks Consumption vs. limits Transportation and resources 4. Politico-economic Costs vs. conservation Economics of energy 5. Attitudes Government vs. the public Credibility of information 6. Resource-Policy Politics and consumption Soft vs. hard technology 7. Technology Alternate technology and conservation Practical vs. unrealistic technology 8. Industry Industry and waste Industry and lag time 9. Conservation Technology vs. conservation Transportation vs. conservation 10. Policy Conservation and alternate technology Housing vs. transportation

PAGE 104

Figure 1. Factor Loadings for Construct 18 -Conventional vs. Alternate Technology. Vertical axis -resource dimension Horizontal axis -politicotechnological dimension E = Engineering students; T = Education students; C = Community group members.

PAGE 105

94 Politico-technological

PAGE 106

Figure 2. Factor Loadings for Construct 18 -Conventional vs. Alternate Technology. Vertical axis -resource dimension Horizontal axis -politico-economic dimension E = Engineering students; T = Education students; C = Community group members.

PAGE 107

96 Eoli ti<9-e
PAGE 108

Figure 3. Factor Loadings for Construct 18 -Conventional vs. Alternate Technology. Vertical axis -politico-economic dimension Horizontal axis -politicotechnological dimension E = Engineering students; T = Education students; C = Community group members.

PAGE 109

98 Pol it icotech no logic a I

PAGE 110

CHAPTER V CONCLUSIONS AND RECOMMENDATIONS A major challenge to the planning and management of community educational processes lies in the pluralistic nature of our society. As has been reviewed in previous chapters, educational theorists agree as to the critical need for an emerging role for education in providing linkage between the various formal and functional groups in society. In accepting this challenge educators must find tools for mapping and quantifying similarities and differences for values and perceptions across groups. The purpose of this study was to demonstrate the utility of a method based on Kelly's personal construct theory for mapping and quantifying attitudes about energy issues. Kelly's personal construct theory views human thought and learning as a process of construing. In this view individuals possess ideas about the world which give meaning to information and experiences. These constructs are bipolar so that new information is seen as like one end of the polarity and unlike the other. There are also constructs about constructs at various levels. Thus, people develop systems of constructs relating to any particular area of their lives. One measure of a person's effectiveness is how congruent his/her construct system is with reality and consequently 99

PAGE 111

100 how well it permits prediction of events. In this sense, Kelly saw all persons as scientists making and testing models of reality. In order to test the effectiveness of an approach based on personal construct theory to mapping group perceptions and values, it was decided to examine three groups which could be expected to differ substantially in their views of energy issues. These three groups were a group of senior students in the College of Education at the University of Florida, a group of senior students in the College of Engineering at the University of Florida, and a group of adult learners at Santa Fe Community College. Each member of these groups completed a repertory grid of the type developed by Kelly. In this instrument each person listed 20 factors they considered relevant to current energy problems. These became the elements for repertory grid analysis. Triads of elements were combined by each respondent to generate constructs about energy issues. A grid was then completed indicating all associations between every construct and each element. Analyses were made of pooled group responses to identify commonalities and differences among groups. The groups demonstrated substantial amounts of overlap in the elements reported. While 74 respondents generating 20 elements each could have produced 1,480 distinct elements, in fact only 44 clearly different elements were found. When the frequency of citation of elements was examined, the same five elements occurred among the most frequently cited 10

PAGE 112

101 elements in all three groups. This similarity of awareness is taken as evidence of the operation of pervasive means of information transfer such as newspapers and television. The construct responses of the groups, however, did not follow the same pattern as that of the elements. With an overall mean of almost 7 constructs per respondent, 74 respondents could have generated over 500 district constructs. In fact, 119 clearly distinguishable constructs were found. This is a proportionally much less compact body of responses than the element responses. It will be recalled that construing involves predicting. A construct contains implicitly the position "this is the way the world is." The implication of that position is "this is the way it seems reasonable for the world to be given my past experiences and observations." In this sense constructs are valuational entities. In light of these observations it is not surprising that the group construct responses showed only one common construct among the most frequently cited 10 for each group. The data reported in the previous chapter show substantial differences in perceptions among the groups studied. For purposes of educational design, it is desirable to examine not only differences in perceptions but also reasons for these differences. In order to do this it is necessary to distinguish perceptions from the values associated with these differences. In the section below will be found a discussion of these two aspects of cognition as found in the groups studied, application of these results to a

PAGE 113

102 strategy for developing effective educational programs, and an examination of the paradigmatic virtues of the approach used in this study as it might apply to other educational development activities. Group Perceptions of Energy Issues In the previous chapter it was reported that by both statistical and empirical criteria, the groups studied demonstrated substantial differences in reported perceptions of facts and values about energy. Of equal interest, they demonstrated some surprising commonalities of awareness. In the discussion below, these findings are examined in the light of personal construct theory and of the specific features of the energy problem. Differences in Perceptions of Fact That there occurred significant overlap of awareness of fact is demonstrated by the obersvation that the same five elements occurred among the most frequently cited 10 elements for all three groups. While no precise probability can be assigned to the random occurrence of this effect (because there is no exhaustive enumeration of all possible elements), it is clearly small. A possible explanation of this observation is that persons in the groups studied probably acquired factual information from the same general pool of resources. It would be interesting to test this hypothesis by correlating the frequency of citation of these five elements with the frequency of occurrence of the same elements in the various media, such as newspaper articles,

PAGE 114

103 television documentaries, etc., as a check on the influence of information transfer. That the same dimensionality occurred in three of the first 10 factors for each of the three groups indicates that there is a moderate degree of congruence of organizing constructs. This effect, however, is not as strong and distinct as the common element effect. This may suggest that there is a difference between the nature and quality of experience that generates elements from that which generates constructs. One explanation of this observation is that the transfer of supraordinate constructs (hierarchical or organizing principles) requires higher level mental activity than the transfer of factual information. Thus it is the implicit aim of most learning models to maximize the internalization of hierarchical structures by the learner. Both advanced organizers and discovery learning models, for example, are aimed toward nurturing the development of a coherent hierarchical understanding on the part of the learner; in the case of the former by an explicit description of the teacher's perceived heirarchy, in the case of the latter by careful teacher participation in the emerging process. Differences in Valuation If the generation of constructs was simply the elicitation of random relational entities of elements, the degree of commonality in constructs among groups should parallel that of elements. However, the results of this study show that members of the different groups, given much the same

PAGE 115

104 pool of information, generated distinct sets of relational entities or constructs. While the influence of individual differences cannot be discounted, the data show commonalities in the high incidence of citation of certain constructs within groups. The failure of this high incidence to carry over across groups suggests that there are group specific influences in operation. A discussion in a later section examines this observation in the light of the network models discussed in the first chapter. A clue to the specific nature of valuational differences can be found in the relative location of constructs on common factor axes among groups. Examination of the factor loading coefficients for a given construct among groups reveals those dimensionalities on which the construct scores highest. Interpretation of the factor can suggest the value held by a group for the specific construct. For example, among education majors construct number 23, "soft" versus "hard" technology, loaded quite heavily on factor number 5, "technical development policy factor." The same construct loaded quite heavily on factor 6 for the community group. This factor had the dimensionality of political resource conservation policy making. Among engineers construct 23 loaded most heavily on factor 7, a factor having the dimensionality of the impact of economic policy on technological development. These loadings point to the aspect of the "soft" versus "hard" technology issue in the responses of each group.

PAGE 116

105 The results of this study suggest that this method provides only a clue, however, for two reasons. First, the constructs which occur across groups tend to be quite general, and, as is the case in information theory, the greater the generality, the lower the information content, i.e., the interpretability . Second, only one frequently cited construct was found common to all three groups in this study, alternate versus conventional technology. It would probably be a case of overinterpretation of data to attempt to draw conclusions about a topic as complex as energy issues from a single construct. A possible approach to the problem of developing several construct or group associations is to repeat the study using an instrument with a predetermined element set and including certain key constructs to be tested. This approach would of course suffer from imposing researcher biases and values, but should provide readily interpretable results on several constructs. It would be of interest to perform such a twostep study on subsets of a single group. Implications for Educational Planning and Curriculum Development In designing a program of energy education, it would be valuable to utilize the commonly perceived elements of the target groups as bridges. Where groups can agree on perceptions, they can be led to discuss, clarify, and develop valuations related to those perceptions. It is where facts

PAGE 117

106 and premises are questioned from the beginning that community processes and communication breakdown. For example, for the community and education groups Table 2 shows construct 18, "conventional technology versus alternate technology," occurring as the second and third most frequently cited construct, respectively. This is the only construct appearing on the list for both groups. However, inspection of the table reveals common themes of economics, public attitudes, and conservation. These themes are consistent with the structure of the factors in Tables 15, 16, and 17 and indicate awareness and concern on the part of members of both groups with regard to the issues they represent. A program of community education aimed at these groups might profitably focus on discussions of such issues as prevalent attitudes toward conservation, the question of finite resources, priorities for technological development, and the role of the political process in planning for an energy limited future. From discussions of these topics already known to be of interest to the target groups linkages can be developed to topics which are less familiar to the target populations . Although specific perceptions will be valued differently by the groups, the existence of common perceptions and of common dimensionalities of valuing should greatly facilitate communication. This approach to curriculum planning and educational strategy recognizes the difficulty of comprehensively

PAGE 118

10.7 informing through a broad based community education program and accepts the more limited goal of clarifying and consolidating the pre-existing common body of information and ideas. While such a goal is restrictive, it is more comprehensive than what appears to have been achieved thus far. Finally, if community educational processes are seen as ongoing and evolving, it can be assumed that the commonalities will themselves expand and evolve as better communications occur. Thus, over time, perhaps an extensive dialogue can take place. Implications for Other Areas of Educational Planning The data reported and discussed above clearly demonstrate the existence of differences of perception and valuation among the groups studied. In any community, there exist functional groupings of individuals which share values and perceptions regarding one or more areas of interest. The methodology described provides a way of identifying the commonalities and differences which contribute to the group identity. As an extension of the present work, disciminant analysis can be applied to the group grids and a function developed which might assign individuals to group membership based on valuations on the principal factor axes. It is consistent with the observation of various anthropological workers described in Chapter I to suggest that element perceptions on culturally pervasive issues such as energy arise from the dendritic network of agency, institutional and media communications, while the constructs are strongly conditioned by radiative networks of family,

PAGE 119

108 interest group, or occupational communications. Because of the strong non-verbal aspects of construing much informal, non-written communication of information can occur. An ethnographic study correlating valuational behavior with construct structure for an informal group might bear interesting fruit in terms of accounting for this type of behavior. It is probable that educational programs which ignore the valuational aspect of social issues will fail in achieving effective communication with membership of groups which differ significantly from the values assumed in the curriculum design. The method described in this work provides a paradigm for mapping such values with regard to any social issue. Educators have traditionally accepted responsibility for the transfer of information, the organization of that information into rational, heirarchical structures, and the use of knowledge as a base for ethical participation in the democratic process. Historically, these responsibilities have been restricted to the young. An emerging role of education is the extension of these functions to society as a whole. If curriculum is the plan by which education functions, then workers in this field must find methods to present and develop organized knowledge and thought in target populations. While a variety of learning theories are available which provide specific strategies for implementation of the educational process, there is a need for a unifying theory of perception and thought that transcends

PAGE 120

109 differences in strategy. Personal construct theory provides us with that unifying picture. It permits the identification of various levels of intellectual functioning for which the educator must take responsibility. Thus, curriculum planning can begin with the transfer of factual information, proceed through the development of organizing constructs, explicitly facilitate the testing of constructs against reality, and culminate in the evolution of high level supraordinate constructs which serve as guides for moral action. The techniques demonstrated in this study can be used as tools to visualize the status of an individual or group at any of these levels, thus providing baseline data on which planning can take place. One can readily create lists of areas which could profitably be examined by repertory grid analysis. For example, such issues as population control, community planning and zoning, and educational policy and priorities could all easily be dealt with. It has been said that a good scientific hypothesis always creates more questions than it answers. Perhaps this work in a limited way shares that desirable property.

PAGE 121

APPENDIX A Thank you for participating in this study. It will permit us to have a better picture of public feelings of the energy problem. Please fill in all blanks below but do not identify yourself by name or social security number. Age Sex Race Major or occupation SURVEY OF ENERGY PROBLEMS I. List 20 factors you believe are important to understanding current world energy problems. 1. 11. 2. 12. 3. 13. 4. 14. 5. 15. 6. 16. 7. 17. 8. 18. 9. 19. 10. 20. 110

PAGE 122

Ill II. Find pairs of factors ivhich are alike. For each pair find a third factor that is different in the same way as the pair is alike. Continue to do this until you can find no more contrasts. (Most people will have more than 5 and fewer than 25 contrasts.) For your convenience, simply record the numbers of the factors in each contrast in the blanks provided below. In the space alongside each triplet briefly describe how the factors are similar or different. Alike Different How C. D. E. F. G. H. I. J. K. L. M. N. 0. P. QR. S. T. U. V. w. X. Y. z.

PAGE 123

112 103 III. Look at the grid on the page. Each square in the grid has a small letter and number on it. The numbers refer to your list of factors from Part I and the letters to your list of contrasts from Part II. Start with box Al. Does contrast A apply to factor 1? If so, place a in box Al ; if not, place a in box Al . Repeat the process for box A2 checking whether contrast A applies to factor 2. Continue until you have checked your contrasts against all 20 factors. FACTORS 1 2 3 4 S 6 7 8 9 10 11 12 15 14 IS 16 17 IS 19 Al | A2 j A3 1 A4 IAS

PAGE 124

APPENDIX B LIST OF OCCUPATION/MAJOR BY GROUP Community Group Load Operator (1) Electrician (1) Receiving Clerk (1) Owner of Business (1) Sales (1) Nurse (1) Radio/T.V. Student (5) Part-time Student (2) Pre-engineering Student (1) Engineering Students Major Mechanical Engineering (2 8) Education Students Health Education (7) Curriculum and Instruction (2) Social Studies (3) Music Education (2) Biology (1) Art Education (3) English Education (1) Math Education (3) Physical Education (5) Business Education (2) Speech Pathology (1) Not Given (2) 113

PAGE 125

APPENDIX C ELEMENTS Technology 01 Physical Laws 100 Efficiency of Current Technology 03 Development of New Technology 04 Financial Support for Research and Development 05 "Hard" Technologies 06 "Soft" Technologies 07 Education *08 Education *09 Transportation 10 Safety of Transportation 11 New Auto Technology (small, electric) 12 Mass Transit Socio-Political *30 Government Regulation 31 Price Controls 32 Environmental Regulations 33., Taxes 34 Politics *40 Public Information and Attitudes 41 Social Attitudes 42 Lifestyle Expectancies 43 Credibility of Information *50 Conservation 51 Transportation (car pools, bikes, walking) 52 Waste Heat Recovery 54 Energy from Waste *60 Social and Governmental Organizations Geography /Economics *80 Economics 81 Industry *20 Population 21 Growth Rate 22 Distribution 23 Labor 114

PAGE 126

115 *70 Resources 71 Distribution 72 Development 73 Import/Export 74 New Types (alcohol, etc.) 75 Safety 76 Storage 69 Consumption 68 Pollution 67 History *77 Needs (climate, weather, etc.) 78 Food 'Major Organizational Categories

PAGE 127

APPENDIX D MASTER CONSTRUCT LIST Economics 001 Economic factors versus individual action 002 Economic factors versus technical forces 003 Economic cost versus regulatory agencies 004 Economic cost versus industry 005 Economic lifestyle expectations versus resource limitations . . 006 Economic control over technological decisions 007 Economic cost versus availability of resources 008 Economic cost versus energy waste 009 Economic cost versus conservation 010 Economic cost versus public needs 011 Economic cost versus policies 012 Economic cost versus known cost of energy 105 Economic cost versus alternate energy 118 Economic kickbacks versus politics 111 Economics versus education 114 International economics versus national economics Technology 013 Technological forces versus social forces 014 Alternate technology versus alternate energy resources 015 Technology versus resources 016 Alternate technology versus conservation 017 Technology versus public attitudes 018 Conventional technology versus alternate technology 019 Environmental cost of energy technology versus politics 020 Technology versus regulation 021 Technology versus conservation 022 Risk of technological development 023 Soft versus hard technology 024 Alternate technology versus resources 025 Technology versus consumption 026 Current technology versus long-range planning 027 Individual technology versus group technology 32 Technology versus industry 037 Technological-practical versus unrealistic 096 Politics versus technology 107 Public attitudes versus technology 109 Technology versus research 116

PAGE 128

117 Conservation 040 Conservation of energy versus energy waste 041 Conservation of energy versus historical trends 042 Conservation of energy versus alternate energy 043 Public conservation versus private conservation 044 Conservation versus public attitudes 045 Conservation versus consumption 046 Conservation versus international relations 047 Residential versus public conservation 048 Conservation versus regulation 049 Conservation versus resources 050 Conservation in transportation versus conservation in buildings 029 Conservation versus transportation 033 Conservation versus industry Attitudes 060 Attitudes and planning 061 Attitudes and regulation 062 Attitudes and international relations 063 Attitudes and population growth 064 Attitudes and policies 065 Credibility of information versus attitudes 066 Effective education of energy consumption versus solutions 067 Individual behavior versus public policy 068 Ethical industrial behavior versus unethical 069 International attitudes versus domestic attitudes 070 International relations versus population growth 071 Public growth and regulation 036 Politics versus health 051 Attitudes versus resources 077 Transportation versus government control of information 097 Education versus industry 075 Education versus resources 076 International relations versus standard of living 099 Education versus international relations 073 National policies versus foreign affairs 074 Pressure toward and against international warfare 095 International relations versus industry 101 Politics versus research 103 Politics versus waste 106 Politics versus industry 108 Education versus alternate energy 113 Education versus law 115 Government versus population 117 Politics versus transportation 051 Attitudes versus resources

PAGE 129

118 Energy Use 080 Energy needs versus luxuries 081 Energy production versus consumption 082 Energy production versus efficiency 083 Resource limits versus consumption 084 Resources versus policy 085 Limited resources versus growth economy 086 Energy consumption versus energy shortage 087 Resources versus population growth 088 Renewable resources versus fossil fuels 089 Worldwide distribution of oil versus no distribution 090 Alternate energy versus energy regulation 091 Energy waste versus consumption 092 Environmental resources versus pollution 093 International resources versus national resources 094 Alternate energy versus international relations 030 Transportation versus consumption 031 Politics versus resources consumption 039 Industry versus energy waste 034 Politics versus alternate energy 035 Transportation versus resources depletion 038 Urban industry versus rural farming 077 Transportation versus housing 079 Food versus shelter 098 Industry versus resources 102 Resources versus research 104 International relations versus resources 110 Research versus alternate energy 112 Housing versus resources 116 Regulation versus consumption 119 Industry versus time Organizations 100 Functioning versus nonfunctioning concerns

PAGE 130

— > 5 r: rL>-JL B« y -r 119 t it _, r. c.

PAGE 131

120 iooc-oc-.« — — „ — __— _r tr o c — rv <" c1") c ^z JUl'UULUUtl .'UUUUUt W U V

PAGE 132

121 NtDBC ; Cc C

PAGE 133

122 C <• rj S" •• •• DC NlB <\i O P 1 * O CC CO c s c c '/Mrt>c"Wntfif. 1 fkj'C i, c-N«ffir)XMccn1 \'i'e(fiCKrc •«m*MCC > c s c ac — w r ^ it t£ k* e c o — : ooe-^c-cc-cooc'^oooc

PAGE 134

123 cnccco:;cr t w. C s c !> r — )OOocoor.

PAGE 135

124 MiCCC "." C *tf 1 C ? — C 3 flri«1.\i£iO'i»1£( «f rc vn r — rv, -r c o coeoococ ~-ot • c c r c * ji fOf^tsr; c^I"-^^^-c:^ccv•!""* K ' , 'T^^cf. *» — r, *i-.? * »£ s *r f, * *. c — c fj — c .c _ r — — t? *n f» r rv cr— r* c rr* o <* r »* i~ c f ->-p:i; c •Of^rr T-'r'.-Ci-cr.r.-c-'-ir.-'f.-f c; cc " C* O c o r — C — r>j — ry L -i « * r. l" c — * — cM; , f.C':^f'<-![,f.*nL--cc#( -. r. — — c frcei"**cr<.rr. Nt^ct'C c r j l* c. c c c r; c r : -. „ » r c — mt: f c — £ n l' — *** ••o*e i* »«• c*'*) c r*. — a; •*". — * — * o i r — *• o — if o* ^c k o a — •«. — o «• r: < -«e it r. c c © — » •£ c it ** a. « c c *r it l*> *f ** eo c 0>o f• cr.-r-trc'.fi-^ffi'corcn-.-ceP'r .'^;^f-!:ro^' ^ -"oc-« l ;(«cn^».<•^-rci.'.^-^w 1 ^l'•Cflv'^l^f , ^ ' x * f. Jir> «c I"r. ^ ^ < fv « c — o r. cr v c r a'*C c in ^ r 1' t* c ^• * •* f\i f" rj ^\ r w r, c IT c ^ E C* e — rj r^ oocoooococcoooocooc;ocC'cocooooecccoacco^c'ocoO!r'Cr-oo c ooooocn(~(^cccoce'c ' u u o u w u u u ij tj ij u il.' o c u ^ : u u y u w ^ l. w u u u u t. 1 c (_•• u u u v-" c c w l c l; u ^ u u u u u u l, ^ uyuuucuuuuuui.'

PAGE 136

>rOocc-j( C *~ C . C Li IT v c r ; •*"c c f J rr — r w e c ^ ftc IT K if. C O v r c ei~ -* o l<"it ! 125

PAGE 137

126 « C * K Z, O • '• =: ; tc c o c # P or . -r c 0" •J — tv*". cu «r * — r 4 cC'jic > o ° c cecoccooe^"?ee-c^cccc orcooc » o — r\j n
PAGE 138

127 *? *c r o c c ir * «•.' »rj — ir> c K
PAGE 139

128 •_' p.ft *» *c k r — »•: l' cct Vjpco r hr ' ; a V— ~ ~ r* # k ?« i,-** o cr t • C n c # n c* joPcorooo--— « — r.rur " -f !•» CC O c a C — f . r> « l~ * r

PAGE 140

C ~> C t: O o O C . r. o c ' " 1 * X O L" 129 Ji>r-r-C
PAGE 141

ocorrooc 130 7 C C 7 <

PAGE 142

131

PAGE 143

REFERENCES Alexander, C. No tes on the Synthesis of Form . Cambridge: Harvard University Press, 1964 . Bannister, D. New Perspectives in Personal Construct Theory . New York: Academic Press, 1977 . Bannister, D. and Mair, J. M. The Evaluation of Personal Constructs . London and New York: Academic Press, T968. Combs, Arthur. Some Basic Concepts in Perceptual Psychology. Curriculum Planning: A New Approach . Hass, Glen; Wiles, Kimball ; and Bondi, Joseph (Eds . ) , Boston: Allyn § Bacon, 1974. Craik, Kenneth. Environmental Psychology. New Directions in Psychology . New York: Holt, Rinehart, and Winston, Inc., 1970, 4. Cremin, Lawrence A. Public Education . New York: Basic Books, Inc., 1976. Cross, K. Patricia. Growing Gaps and Missing Links. The Interstate Compact . 1979, 14-15, 29. Deutsch, Paul S. Meaning in the Environment as Construed by Environmental Designers and Layman. Submitted in Partial Fulfillment of the Requirements for the Degree of Bachelor of Science at Massachusetts Institute of Technology. May, 1972. Dewey, John. Democracy and Education . New York: Macmillan, 1916. Dewey, John. Dewey Outlines Utopian Schools. The New York Times . April 23, 1933. Dewey, John. Experience and Education . New York: Macmillan, 1938. Drucker, Peter F. The Age of Discontinuity . New York: Harper § Row, 1969. Fromm, Erich. The Revelation of Hope Toward a Humanized Technology" New York: Harper 5 Row Publishers, 1~968 . 132

PAGE 144

133 Hall, Edward T. The Silent Language . New York: Fawcett Publications, 1959. Harris, R. J. A Primer of Multivariate Statistics . New York: Academic Press, 1975. Hershberger, R. A Study of Meaning in Architecture. Environmental Design Research Association Conference . Pennsylvania: Halsted Press, 1969. Illich, Ivan. Deschooling Society . New York: Harper § Row, 1971. Joseph E. C. Long-Term Electronic Technology Trends: Forecasted Impact on Education. Testimony presented to the Congress of the United States, Subcommittee of Elementary, Secondary, and Vocational Education, April 25, 1979, 1. Kelly, George A. A Theory of Personality . New York: Norton § Company, Inc. , 1955. Kelly, G. A. A Brief Introduction to Personal Construct Theory . Unpublished manuscript. Brandeis University, 1966. Kelly, George. The Psychology of the Unknown. New Perspec tives in Personal Construct Theory . New York: Academic Press, 1977. Kleinbaum, David. Applied Regression Analysis and Other Multivariable Methods . Massachusetts: Duxbury Press, 1979. Kluckholm, Florence and Strodtbeck, Fred. Variations in Value Orientation . Evanston, Illinois: Row, Peterson, 1961. Knowles, I. Androgogy: Understanding the Adult as a Learner, Journal of Reading . 1977, 20 (5), 361-363. Kohlberg, Lawrence. Moral Education in the Schools: A Development. School Review . 1966, _74 (1), 1-30. Kohlberg, Lawrence. The Cognitive-Developmental Approach to Moral Education. Phi Delta Kappan . 1975, 56* (10) , 670-677. Leff, H. S. and Deutsch, P. S. Construing the Physical Environment: Differences Between Environmental Professionals and Lay Persons. Environmental Design Research Association Conference . Pennsylvania. Halsted Press, 1972.

PAGE 145

134 Leichter, Hope Jensen. The Concept of Educative Style. Teachers College Record . 1974, 75_, 239-250. Leichter, Hope Jensen. Some Perspectives on the Family as Educator. Teachers College Record . 1974, 7_6_, 213-215. Leman, G. Words and Worlds. Perspectives in Personal Construct Theory . London: Academic Press, 1970. Mayer, P. Labour Migrancy and the Social Network. Problems of Transition: Proceedings of the Social Sciences Research Conference . University of Natal, Durban, South Africa, 1964. McKechnie, George E. Measuring Environmental Dispositions with the Environmental Response Inventory. Proceedings of the First Annual Environmental Design Research Asso ciation Conference . Pennsylvania: Halsted Press, 1970. McLuhan, Herbert Marshall. The Medium is the Message . New York: Bantom Books, 19~5T. Meadows, Donella H. The Limits to Growth . New York: Universe Books, 1974. Myers, D. G. How Groups Intensify Opinions. Human Nature . 1979, 3, 34-38. Nie, N. H. ; Hull, C. H. ; Jenkins, J. G. ; Steinbrenner , K. ; Bent, D. H. Statistical Package for the Social Sciences New York: McGraw-Hill Book Company, 1970. Piaget, Jean. To Understand is to Invent: The Future of Education . New York: Grossman Publishers, 1973. Public Energy Education . United States Department of Labor Conference. Washington, D.C., 1979. Robinson, W. S. Ecological Correlations and the Behavior of Individuals. Journal of the American Statistical Association . 1950, 30, 517-536. Rogers, Carl. Learning to be Free. Readings in Curriculum . Hass, Glen; Wiles, Kimball, and Bondi, Joseph (Eds . ) , Boston: Allyn $ Bacon, 1970. Rosenstein, Allen B. Education for the Professions and the Future Quality National Life. Technological Horizons in Education . 1978, 5_ (5), 45-50. Ryle, Anthony. Frames and Cages . London: Sussex University Press, 1975.

PAGE 146

135 Schipper, Lee. Efficient Energy Use and Well-Being: The Swedish Example. Science . 1976, 194, 1001-1013. Silberman, Charles E. Crisis in the Classroom: The Remak ing of American Education . New York: Random House, 1970. Stringer, Peter. Repertory Grids in the Study of Environmental Perception. Explorations of Intrapersonal Space . London: John Wiley, 1976. Torda, Paul. Need for Change in Professional Education. Technological Horizons in Education . 1978, _5_ (5), 54-56. —

PAGE 147

BIOGRAPHICAL SKETCH Carol D. Blalock was born December 25, 1944, in Buffalo, New York. In June, 1964, she graduated from Lyons Township High School in LaGrange, Illinois. In June, 1968, she received a Bachelor of Science degree in Education from the University of South Dakota. In August, 1976, she received a Master of Arts degree in Education and a Specialists degree in Education with a major in counselor education from the University of Florida. Since that time she has pursued work toward the degree Doctor of Philosophy in the College of Education Department of Curriculum and Instruction. Ms. Blalock has held a number of positions in education, having been employed as an elementary and high school teacher, counselor of senior citizens, and coordinator of the environmental education project at Santa Fe Community College in Gainesville, Florida. For her graduate training she worked with the Department of Education on a team to explore possible futures of education in Florida. Carol D. Blalock is a member of Phi Delta Kappa, Kappa Delta Pi, Florida Association of Supervision and Curriculum Development, Florida Foundation for Future Scientists, and the World Futures Society. She received the Bingham Environmental Education Award in 19 78 from the University of Florida. 136

PAGE 148

137 Carol D. Blalock is married to Dr. H. Anthony Blalock and has three daughters, Jeanne, Patricia, and Elizabeth.

PAGE 149

I certify" that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy.
PAGE 150

UNIVERSITY OF FLORIDA 3 1262 08553 0664