Close
Home
Collections
Login
USC Login
Register
0
Selected
Invert selection
Deselect all
Deselect all
Click here to refresh results
Click here to refresh results
USC
/
Digital Library
/
University of Southern California Dissertations and Theses
/
The relationship between human values and moral reasoning as components of moral behavior
(USC Thesis Other)
The relationship between human values and moral reasoning as components of moral behavior
PDF
Download
Share
Open document
Flip pages
Contact Us
Contact Us
Copy asset link
Request this asset
Transcript (if available)
Content
INFORMATION TO USERS
This manuscript has been reproduced from the microfilm master. UMI
films the text directly from the original or copy submitted. Thus, some
thesis and dissertation copies are in typewriter face, while others may
be from any type of computer printer.
The quality of this reproduction is dependent upon the quality of the
copy submitted. Broken or indistinct print, colored or poor quality
illustrations and photographs, print bleedthrough, substandard margins,
and improper alignment can adversely affect reproduction.
In the unlikely event that the author did not send UMI a complete
manuscript and there are missing pages, these will be noted. Also, if
unauthorized copyright material had to be removed, a note will indicate
the deletion.
Oversize materials (e.g., maps, drawings, charts) are reproduced by
sectioning the original, beginning at the upper left-hand corner and
continuing from left to right in equal sections with small overlaps. Each
original is also photographed in one exposure and is included in
reduced form at the back of the book.
Photographs inchided in the original manuscript have been reproduced
xerographically in this copy. Higher quality 6" x 9" black and white
photographic prints are available for any photographs or illustrations
appearing in this copy for an additional charge. Contact UMI directly
to order.
A Beil & Howell information Company
300 North Zeeb Road. Ann Arbor. M l 48106-1346 USA
313/761-4700 800/521-0600
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
THE RELATIONSHIP BETWEEN HUMAN VALUES AND MORAL
REASONING AS COMPONENTS OF MORAL BEHAVIOR
By
Lynn Weaver Frost
A Dissertation Presented to the
FACULTY OF THE GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
in Partial Fulfillment of
Requirements for the Degree
DOCTOR OF PHILOSOPHY
May 1997
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
UMI Number: 9734862
UMI Microform 9734862
Copyright 1997, by UMI Company. All rights reserved.
This microform edition is protected against unauthorized
copying under Title 17, United States Code.
UMI
300 North Zeeb Road
Ann Arbor, MI 48103
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
UNIVERSITY OF SOUTHERN CALIFORNIA
THE GRADUATE SCHOOL
UNIVERSITY PARK
LOS ANGELES. CALIFORNIA 90007
This dissertation, written by
Lynn W eaver F r o s t
under the direction of hJrF. Dissertation
Committee, and approved b y all its members,
has been presented to and accepted by The
Graduate School, in partial fulfillm ent of re
quirements for the degree of
DOCTOR OF PH ILOSOPH Y
Dean of Graduate Studies
Date
DISSERTATION COMMITTEE
S > .
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
DEDICATION
This work is dedicated to the following people who are very dear to me:
To Peggy and Frank, who sponsored my undergraduate education,
To Jack and Yvonne, who provided for the first half of my doctoral
program,
And to Mark, who helped me run the final mile.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
ACKNOWLEDGMENTS
I would like to express my appreciation to those who helped make this
dissertation a reality:
My deepest gratitude goes to my committee chairperson, Dr. William B.
Michael. His kindness, dignity, extraordinary availability, encouragement, and
support made the writing of this study an exciting and enjoyable process.
Special thanks go to Dr. Donald Polkinghome for his help in selecting and
administering the psychometric instrumentation. His interest and support were
crucial to getting the project started.
I also appreciate the interest and support of the members of my dissertation
committee, Dr. Dennis Hocevar and Dr. Dallas Willard.
Finally, my appreciation is extended to Dr. Tony Guerino for his statistical
expertise, to Ms. Susan Putnam for her administrative assistance, and to the
subjects who gave so generously of both their time and their opinions.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
TABLE OF CONTENTS
List of T ables.....................................................................................
Chapter
I. THE PROBLEM: THEORETICAL BACKGROUND
AND PERSPECTIVES..................................................
Area of Concern
Theoretical Perspectives
Theoretical Perspectives Extended: A Selective Review
of Existing Literature as Related to the Proposed
Contributions of the Current Study
Purpose of the Study
Research Question
Hypothesis
Delimitations
Conceptual Assumptions
Organization of the Remainder of the Study
H. METHOD AND PROCEDURES...................................
Research Sample
Variables and Their Measures
Procedures in Data Collection
Data Analysis
Methodological Assumptions
Limitations
m . ANALYSIS, INTERPRETATION, AND DISCUSSION
OF RESULTS............................................................
Analysis of Findings
Interpretation of the Data in Relation to the Tenability
of the Research Hypothesis
Discussion
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
TABLE OF CONTENTS (Continued)
Chapter Page
IV. SUMMARY, CONCLUSIONS, AND RECOMMENDATIONS . . 60
Summary
Conclusions
Recommendations
REFERENCES..................................................................................................... 68
APPENDIX A........................................................................................................ 72
APPENDIX B ........................................................................................................ 74
v
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
LIST OF TABLES
Table Page
1. Findings from Studies Regarding Relationships of
Expressed Levels of Preference for (Importance of)
Selected Terminal Value Statements to Indicators of
Standing on Measure of Moral Reasoning..................................... 19
2. Findings from Prior Research Studies Regarding
Relationships of Expressed Levels of Preference for
(Importance of) Selected Instrumental Value Statements
to Indicators of Standing on Measures of Moral Reasoning .... 20
3. Indicators in Various Demographic Subgroups of Only
Those Relationships That Were Statistically Significant
Between Scores on the DIT (Measure of Moral Reasoning)
and Rank Orders of Importance Assigned to Terminal Value
Statements from the R V S ............................................................... 47
4. Indicators in Various Demographic Subgroups of Only
Those Relationships That Were Statistically Significant
Between Scores on the DIT (Measure of Moral Reasoning)
and Rank Orders of Importance Assigned to Instrumental
Value Statements from the RVS..................................................... 50
5. A Comparison of Statistically Significant Correlational Results
Between Scores on a Measure of Moral Reasoning and Level
of Importance Assigned to Terminal Value Statements Among
Investigations................................................................................. 55
6. A Comparison of Statistically Significant Correlational Results
Between Scores on a Measure of Moral Reasoning and Level
of Importance Assigned to Instrumental Value Statements Among
Investigations................................................................................. 56
vi
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER I
THE PROBLEM: THEORETICAL BACKGROUND
AND PERSPECTIVES
Area of Concern
The Importance of Studying Moral Reasoning and Human Values
Both human values and moral reasoning have been researched in an attempt to
understand human behavior; however, neither domain by itself provides a sufficient
explanation of behavioral influences. In pursuit of this goal, researchers have
attempted to find relationships between human values and moral reasoning (Weber,
1993).
Feather (1988) suggested two reasons why a relationship between values and
stages of moral reasoning might be expected: (a) theorists such as Kohlberg (1984)
associate different values with different developmental stages of moral reasoning and
(b) solutions to moral problems would be expected to be influenced by learning in past
social situations, because both moral reasoning and preferences for values have been
related to socialization experiences. Unfortunately, although consistent relationships
between values and moral reasoning might be predicted, research has failed to produce
consistent findings. Indeed, the results have been better distinguished by a nearly total
inconsistency of findings. A handful of investigations has been concerned with this
relationship. Although all but one have produced significant results, the findings have
1
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
been, almost without exception, dissimilar (Feather, 1988; Glover, 1991; Lockley,
1976; McLellan, 1970; Parish, Rosenblatt, & Kappes, 1979-1980; Weber, 1993).
The Moral Domain
Colby and Kohlberg (1987) distinguished between moral judgment and other
types of judgment:
Moral judgments, then, have certain properties that make them moral
judgments. They are, first, judgments of value, not of fact. This
distinguishes them from cognitive reasoning and judgments studied by
Piaget. Second, they are social judgments, judgments involving people.
Third, they are prescriptive or normative judgment, judgments of ought,
of rights and responsibilities, rather than. . . judgments of liking and
preference, (p. 10)
In Kohlberg’s conception, decisions of moral "oughts" are primarily interpreted as
increasingly differentiated conceptions of justice (i.e., fairness) constituting three levels,
which are further subdivided into six stages of moral reasoning development (Colby &
Kohlberg, 1987).
Moral Reasoning and Moral Behavior
Moral reasoning has been found to be statistically related to moral action, albeit
inconsistently (Blasi, 1980). Varying degrees of empirical support for the correlation
have existed for different areas of behavior. Higher stages of moral reasoning have
been associated with the ability to resist pressure to conform in judgment, but not
behavior. Only modest support has been found for a relationship between moral
reasoning and altruism or honesty.
2
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
From a theoretical point of view, the inconsistency of relationship between
moral behavior and moral cognition should not be surprising. In developing his method
of assessment, Kohlberg (Colby & Kohlberg, 1987) moved progressively away from
content and toward thought structures as the basis for his developmental stages. In his
theory, each stage is defined not by an action choice, but by the reasons given in
support of the action. Opposite moral actions are allowed to coexist within the same
developmentally defined stage (Colby & Kohlberg, 1987). Theoretically, any stage of
development would be compatible with any action choice possible within Kohlberg’s
moral dilemmas. Thus, it would seem that the ability to predict action from stage of
reasoning would be expected to be low, given the stages as defined.
Human Values and Behavior
In contrast to these findings, human values have been shown to be correlated
with numerous behaviors (e.g., Ball-Rokeach, 1984; Rokeach, 1973, 1979). Rokeach
(1973) defined a value as “an enduring belief that a specific mode of conduct or end-
state of existence is personally or socially preferable to an opposite or converse mode
or conduct or end state of existence” (p. 5). He further described values as “the main
genotypes that underlie or determine social behavior” (p. 122).
Moral Reasoning and Values as Components
of Moral Behavior
In response to the irregularity of findings concerning moral reasoning and
behavior, Blasi (1980) suggested the need for greater emphasis on the “psychological
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
nature of integrity or of personal consistency” (p. 40). He criticized the study of
moral/cognitive structures in absence of their psychological environment. The current
investigation represents an attempt to respond to this critique and to bring an existing
theoretical position to bear on the lack of consistency which has characterized research
in the area of moral judgment, human values, and moral behavior. Through a
theoretical integration, perhaps a new understanding of old findings may emerge.
Rest (Narvaez & Rest, 1995) has proposed a theory of moral behavior which
may serve to explain and to integrate the discrepant findings previously indicated.
According to Rest’s theory, moral behavior is determined by the following four
processes or components which are hypothesized to interact with one another: moral
sensitivity, moral judgment, moral motivation, and implementation (Narvaez & Rest,
1995). (See Appendix A for complete descriptions of components.)
In this view, moral judgment and human values are actually two separate
processes which interact with personality dimensions in producing behavior. Whereas
moral judgment (Component II) involves deciding between right and wrong acts,
human values (Component 1H) constitute the motivation (or lack thereof) to carry out
the chosen act. This model would predict that, as separate processes, values and moral
reasoning each would exhibit a stronger relationship to moral behavior than to one
another. This interpretation is consistent with past research findings.
This current study is an attempt to replicate past research on the relationship
between values and moral reasoning. It is anticipated that further inconsistent findings
4
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
may result. Yet, this outcome in itself would serve to advance the body of knowledge
in this area. It is hypothesized that any significant findings which emerge (or which
have been observed in prior studies) may be explained in terms of sample
characteristics (e.g., demographic factors) rather than stage of moral reasoning. The
twofold contribution to be made in this instance would be (a) the advancement of the
four component theory of moral behavior and (b) a further clarification of the proposed
functionally independent (or at least externally moderated) relationship between moral
reasoning and human values.
Theoretical Perspectives
The Cognitive Developmental Legacy:
Piaget. Kohlberg. and Rest
Jean Piaget: The theoretical beginnings. The cognitive-developmental
paradigm of moral theory began with the work of Jean Piaget (1965). Prior to his
work, morality was considered to be an unsuitable topic for scientific investigation.
This view stemmed partially from the principles of logical positivism which
predominated the philosophy of science governing psychological inquiry. These
principles demanded that science should be value free or neutral. Facts were to be
clearly distinguished from values, which were considered to be subjective and therefore
inappropriate for empirical investigation. As morality was centered around values
which determine the good and the right, the study of morality clearly was outside the
providence of science as it was conceived at the time. However, advances in thinking
5
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
within the philosophy of science brought these assumptions into question. Specifically,
after it became more widely accepted that the human sciences in particular could not be
value free, the emphasis then moved to questions of how values influence research,
rather than i f they do. These developments opened the way for the empirical study of
moral development (Kurtines & Gewirtz, 1995).
The theory of moral development formulated by Piaget (1965) was an extension
of his work concerning the cognitive development of children. He proposed that
through the processes of assimilation (acquiring new information through experience)
and accommodation (adapting to the newly acquired knowledge), individuals progress
through developmental stages. Piaget identified four stages of cognitive development
present during the human lifespan, the exposition of which has been adapted from the
text in Papalia and Olds (1995, p. 29) as follows:
-Sensorimotor stage (ages birth to 2 years): The early infant exhibits mostly
reflex responses. He learns through sensory and motor activities to organize his
behaviors in relation to his environment.
- Preoperational stage (ages 2 to 7 years): The child leams the use of
symbols to represent things as he constructs a representational system of
himself and his environment.
- Concrete Operations stage (ages 7 to 12 years): Child develops
ability logically to solve problems focused in the here and now.
- Formal Operations Stage (ages 12 years through adulthood): Abstract
and hypothetical thinking becomes possible.
In applying his understanding of cognitive development to the moral realm,
Piaget (1965) posited two stages of moral development which loosely parallel two of
his stages of cognitive development. All individuals were said to proceed through the
stages in a specified order, but with the timing of progress to vary among persons.
6
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
The first stage, morality o f constraint, or heteronomous morality, corresponds
approximately with the preopenational stage of cognitive development. At this level,
children possess an egocentric view of things. They are incapable of adopting another's
point of view. This egocentrism leads to a rigid, rule-based style of moral judgments
which is accompanied by severe punishment for any infractions. One is constrained by
moral rules which are externally imposed.
Piaget's second stage, the morality o f cooperation or autonomous morality,
roughly coincides with his concrete operations stage of cognitive development.
Increased social experience has expanded the child's ability to assume points of view of
individuals other than herself. The child also has been exposed to environments
offering a richer, and sometimes conflicting, account of what is good and acceptable.
He or she begins to regard rules as being made by, and therefore as being modifiable by
human beings. This flexibility, combined with her or his new understanding of
intention, yields a view of punishment which is in proportion to the transgression
(Papalia & Olds, 1995).
Kohlberg's Dissertation:. An expansion of Piaget's stage theory. In his
dissertation, Kohlberg (1958) initially intended to validate Piaget's (1965) model of
moral development. However, he concluded that “Piaget had defined a series of
cognitive-developmental dimensions of moral development. . . (which) did not define
qualitative levels and did not cluster together to define a ‘structured whole’ as the stage
concept would imply” (Kohlberg, 1984, p. xxix). He proposed an alternative model to
7
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Piaget’s (1965). He cautiously called the stages “ideal types" because he was
concerned that a single study did not support the rigorous conditions of developmental
stages including invariant sequence, structured wholeness, and hierarchical integration
(to be discussed subsequently) (Kohlberg, 1984). Future research ultimately provided
enough cross-cultural, cross-gender, and longitudinal evidence to give Kohlberg the
confidence to call his types "structural stages" (Colby & Kohlberg, 1987).
After proposing a six-stage developmental theory of morality as an alternative
to Piaget’s (1965) view, Kohlberg spent the balance of his career refining, expanding
upon, and defending his position. His work, in turn, has been continued and modified
by his former student, James Rest.
Theoretical Similarities Between Kohlberg
and Piaget: Characteristics and
Philosophical Assumptions
Much as did Piaget (1965), Kohlberg (1984) distinguished the form or structure
of moral thinking from its content. By abstracting the organizing principles of moral
thought from situationally imbedded content, Kohlberg maintained that he had found
generalizable forms of reasoning (Colby & Kohlberg, 1987).
In a vein similar to that of Piaget (1965), Kohlberg accepted traditional criteria
for his stage model:
1. Stages imply a qualitative difference in structures (modes of thinking)
that still serve the same basic function. . . at various points in
development.
8
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
2. These different structures form an invariant sequence . . . in individual
development. Cultural factors . . . do not change its sequence.
3. Each of these different and sequential modes of thought forms a
'structural whole.' A given stage response. . . represents an underlying
thought organization. The implication is that various aspects of stage
structures should appear as a consistent cluster of responses in
development.
4. Stages are hierarchical integrations. . . stages form an order of
increasingly differentiated and integrated structures for fulfilling a
common function. Accordingly, higher stages displace (or, rather,
integrate) the structures found at lower stages (Colby & Kohlberg,
1987).
Finally, both theories are based on evolving conceptions of justice (Colby & Kohlberg,
1987).
However, as striking as the similarities are between the two theories, their
differences are perhaps more profound. Piaget (1965) developed his view from direct
observations of children, whereas Kohlberg (1984) based only his lower stages (1-4) on
more scientifically controlled empirical observations. In contrast, Kohlberg’ s principled
stages (5-6) are grounded heavily in the Kantian and Rawlsean tradition of Western
philosophy (Snarey, 1985).
Kohlberg's Three Levels and Six Stages of Moral Development.
Kohlberg (1984) expressed his theory in several formulations across the more
than thirty years of his career. Alterations were made in response to methodological
and theoretical critiques and to empirical evidence of flaws in his conceptualization (see
Kohlberg, 1984, or Colby and Kohlberg, 1987, for an extended discussion). The final
incarnation is the most thoroughly conceived and empirically discriminating of the
variations.
9
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Kohlberg (1958,1984) has defined moral reasoning as being classifiable within
three levels which are further divided into six stages of development as follows:
Level 1: Pre-Conventional Morality.
Stage 1: Heteronomous Morality.
Stage 2: Individualistic, Instrumental Morality.
Level 2: Conventional Morality.
Stage 3: Interpersonally Normative Morality.
Stage 4: Social System Morality.
Level 3: Postconventional Morality.
Stage 5: Human Rights and Social Welfare Morality.
Stage 6: Morality of Universalizable, Reversible, and Prescriptive
General Ethical Principle(s): (Kohlberg, 1984) (See Appendix
B for complete definitions of the stages.)
All persons are said to progress through this sequence of stages in the same order,
although the rate and end-point of development may vary. That is, some individuals
may move through the stages more rapidly than others, and not every person proceeds
through all stages. In fact, as the upper two stages are quite rare, Kohlberg eventually
allowed the possibility of conceiving the sixth stage as a theoretical entity not
evidenced empirically, which was necessary to inform the teleology of the preceding
five stages (Colby & Kohlberg, 1987). However, it is largely accepted that individuals
do move through the stages in the order specified without skipping stages or
10
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
regressing. Evidence for the hypothesis of sequentiality was presented by Walker
(1982). Subjects presented with reasoning two steps above their current stage showed
advances to the stage one step higher than their own, but not to the second higher stage
which was modeled (Walker, 1982).
Primary Research Instruments Associated
with Kohlberg's Theory
Kohlberg (Kohlberg, Colby, Gibbs, Speicher-Dubin, & Power, 1977) developed
his interview and scoring techniques over the more than thirty years of his career. He
did so in response to inconsistent findings, cross-cultural limitations, and theoretical
critiques. The emphasis was on clarifying definitions of the unit of analysis and on the
differentiation of moral content from its structure. The end products of his work have
been the Standard Issue Moral Judgment Interview (MJI) and the Standard Issue
Scoring procedures (Kohlberg et al., 1977). The Interview is in three parallel forms,
each of which contains three hypothetical dilemmas and each of which is followed by
9 to 12 standardized questions used to probe the subject's reasoning process. Each
dilemma represents a conflict between two important underlying goals, such as a
conflict between preserving life versus upholding the law.
A former student of Kohlberg's, James Rest (1993) developed a multiple-choice
questionnaire to assess moral stages which are comparable to Kohlberg's stages. Both
methods share moral dilemmas as stimulus materials, with Rest having adapted
Kohlberg's stories for use on his multiple choice questionnaire. Although the task is
11
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
generative in Kohlberg’ s methodology, it is one of recognition of and of preference for
particular influences on decision making in Rest's methodology. Subjects must
spontaneously produce their responses to the Moral Judgment Interview (MJI)
(Kohlberg et al., 1977). However, on Rest's (1993) Defining Issues Test (DIT),
subjects choose between possible courses of action and then assign degrees of
importance to statements that may have influenced their decision.
In addition to differences in methodology, the data produced by the two
procedures also differ. Kohlberg (1984) posited a simple-stage model (subjects may be
located only at a single stage, or at most, two adjacent stages). The level of reasoning
of an individual is inferred from his or her interview responses. However, the multiple
influences on action choices offered on the DIT are representative of many levels of
moral reasoning. Therefore, through their ranking of possible influences subjects will
exhibit the logic of several stages at once. In Rest’s complex stage model a subject is
not “at” a particular stage. Instead, the subject simultaneously exhibits greater and
lesser percentages of high importance being assigned to the various reasoning
characteristics associated with several developmental stages (Rest, 1979).
The DIT and the MJI have produced empirical results which have been
moderately correlated with one another: .70 for heterogeneous samples and .43 with
homogeneous groups (Rest, 1979). (It should be noted that homogeneous samples
produce scores of more limited range~a circumstance which results in the attenuated
correlation.) Rest (1979) maintained that these levels were adequate for making
12
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
comparisons between the two methodologies because scores on the Standard Issue MJI
were more highly correlated with scores on the DIT than they were with scores on
some of the earlier versions of Kohlberg's Moral Judgment Interview (Kohlberg, 1958,
1978).
Rolceach’s rnnceptiialization of Human Values
Rokeach (1973) listed five assumptions which guided his work with human
values:
1. The total number of values that a person possesses is
relatively small
2. All men everywhere possess the same values to different
degrees
3. Values are organized into value systems
4. The antecedents of human values can be traced to culture,
society and its institutions, and personality
5. The consequences of human values will be manifested in
virtually all phenomena that social scientists might consider
worth investigating and understanding (p. 3).
Rokeach's work focused on the values people hold, rather than on the values objects
are said to have, as human values may function as criteria for evaluations.
Rokeach attributed the enduring quality of values to their being taught and
learned in an all-or-nothing manner, as well as in isolation from other values. (For
instance, children initially learn that it is always right to be honest.) With maturation
and experience, they learn to integrate the separate values into a hierarchical system
of values. It is this hierarchical arrangement of values relative to one another that
accounts for the possibility that value systems may change (Rokeach, 1973).
13
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Rokeach further conceived of values as beliefs. He distinguished between
existential/descriptive beliefs (which can be true or false), evaluative beliefs (the
quality of whether a thing is bad or good), and prescriptive/proscriptive beliefs
(which evaluate some means or end of action to be desirable or undesirable). Thus,
when activated, values prescribe particular ways of behaving or end states of
existence for the individual who holds them.
Rokeach (1973) divided values into two types: instrumental (which are
concerned with modes of conduct) and terminal (which are concerned with end-
states of existence). His types represent two distinct but interconnected systems.
Generally, instrumental values are employed in the service of attaining terminal
values.
These two types of values were further subdivided into four subtypes.
Terminal values may be either personal or social. Personal terminal values are
concerned with intrapersonal end-states of existence; social terminal values are
concerned with interpersonal end-states of existence. Instrumental values were
subdivided into moral and competence values. Moral values have an interpersonal
focus. When moral values are violated, the individual experiences guilt feelings.
McLellan (1970) cited Rokeach as identifying the following instrumental values as
being moral values: “ clean, forgiving, helpful, honest, loving, obedient, polite,
responsible, and self-controlled" (p. 101). Competence values, which are
intrapersonal in focus, are concerned with self-actualization, and are accompanied by
14
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
feelings of shame and feelings of personal inadequacy when violated (Rokeach,
1973).
Rokeach distinguished values from attitudes: values are single beliefs
concerned with desirable ways of being or end-states of existence, whereas attitudes
are several beliefs which are organized around a particular situation or object. In
being more central to the personality, values may guide attitudes. Furthermore,
values may be contrasted with personality traits. Rokeach (1973) proposed that an
individual’s character, viewed externally as a cluster of traits, may be reformulated
from an internal view as a system of values. However, although traits are considered
to be fixed, value systems allow greater flexibility for change.
Finally, Rokeach (1973) regarded values as standards which guide social,
political, and religious behavior. He indicated that values influence ways in which
individuals present themselves, evaluate themselves, compare themselves and others,
and rationalize their behavior to themselves when preserving their self-esteem.
Various forms of the Rokeach Value Survey (RVS) (Rokeach, 1983) have
been employed during some thirty years of values research. The RVS is the primary
instrument associated with Rokeach’s (1973) theory of human values.
A Brief Critique of Kohlberg'sandRokeadi's^Theories
Kohlberg's (1984) theory has been criticized primarily on the issue of
universality. Gilhgan (1982/1993) maintained that Kohlberg's (1984) theory, which
was derived from research with an all male sample, contained a bias in favor of male
15
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
subjects. Her critique has not been supported by empirical findings (Moon, 1986;
Walker, 1984). However, some evidence does exist for an alternative "care"
morality which does not split along gender lines (Cortese, 1989). Similarly, cross
cultural studies have shown that Kohlberg’ s (1984) conception of morality is not
exhaustive of all possible conceptions (Snarey, 1985).
Rokeach has been criticized primarily on methodological issues. His rank
ordering procedure does not allow researchers to know the true distance between
values in their hierarchical arrangement. It also does not allow two values to be tied
in preference. Finally, the relations among values on the RVS are ipsative: if one
value is given a high ranking, this assignment necessarily results in other values
receiving a lower ranking (Mueller, 1984).
Feather’s Value. Valance, and Expectancy
Theory of Motivation
Rokeach declared that values can be motivational by functioning as goals of
behavior. Adopting Rokeach’s definition of values, Feather (1975,1982) expanded
on this view by integrating values into an expectancy-valence theory of motivation.
Feather (1982) identified values more specifically as motives which have a
“normative or oughtness quality about them” (p. 220). He argued that values as
motives “influence a person’s subjective definition of the situation so that certain
objects, activities, and states of affairs within the immediate environment acquire
positive valance (become attractive) or negative valence (become aversive)” (p. 220).
16
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Values may induce valences, or tendencies to approach or to avoid objects or
ways of being, by selectively sensitizing individuals to certain aspects of their
environment. Feather (1982) compared this ability to the way in which hunger can
sensitize a person to aspects of the environment related to acquiring food. Similarly,
if an individual prized being imaginative, that person may attend more easily to novel
solutions than one who has given that value a lower degree of priority (Feather,
1975). Or an individual who has assigned high priority to “a comfortable life” on the
RVS (Rokeach, 1983) may experience a strong positive valence toward money.
Feather (1975) noted that actions are associated within the individual with a
degree of perceived likelihood of occurrence. This concept of expectancy occurs
along with surrounding actions which the individual is actually capable of
performing. Both the present perceived reality and past learning achieved by
individuals influence their perception that they are capable of a particular response,
as well as their belief that the response will be effective in achieving the desired
consequences. Feather (1982) further proposed that the relationship between
valences and expectations is multiplicative in its determination of motivational force:
one who has zero valence toward an object or zero expectation that one's action will
be effective will then experience zero motivation toward that action. This
relationship is posed in contrast to a mere additive relation.
17
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Theoretical Perspectives Extended: A Selective Review
of Existing Literature as Related to the Proposed
Contributions of the Current Study
Studies of the Relationship Psto een Values
and Moral Reasoning
Seven studies of the relationship between human values and moral reasoning
have been identified. The findings across the studies have shown some vague
tendencies toward patterns in their results; however, the results may be just as easily
described by their inconsistencies. The outcomes of five of the studies are
summarized in Table 1 and Table 2. Two of the studies were not included in the
tables for reasons specified in the discussion of their findings. [In the following
discussion of research findings, the directions of associations found by some
researchers were reversed. Some researchers had coded low value rankings with
high numbers, and others had coded low value rankings with low numbers. In the
effort to make sense of the data, all findings have been standardized such that a high
priority ranking for a value associated with a high level of moral reasoning
development is discussed as a positive correlation.]
McLellan (1970) investigated males in grades 7, 9, and 11 by using the RVS
and the MJI. He found the value statements offreedom and obedient to discriminate
among stages of moral reasoning, with high levels of expressed preference for
freedom associated with Stages 2 and 5 and with high levels of indicated preferrence
18
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Table 1
Findings from Studies Regarding Relationships of Expressed Levels of Preference for
(Importance of) Selected Terminal Value Statements to Indicators of
Standing on Measure of Moral Reasoning
Terminal
Value Statement
Related Investigations
McLellan
1970
Lockley Standring Parish etal. Feather
1976 1976 1979-80 1988
A Comfortable Life [Pri]
Equality Pri Pri
An Exciting Life Pre Pri
Family Security [Pri] Pri/m
Freedom S2, S5 [Pre] Pri
Inner Harmony Pri Pri Pri
Mature Love [Pri/m] Pri
National Security [Pri] [Pri/f]
Pleasure Pri
Salvation [Pri] [Pri]
Social Recognition Pri/m
A World at Peace Pri
Note: Pri = Principled reasoning [ ] = negative correlation
Pre = Preconventional reasoning S2 = Kohlberg's Stage 2 of moral reasoning
m = male subjects only S5 = Kohlberg's Stage 5 of moral reasoning
f = female subjects only
(Cheerful, courteous, and happiness have been omitted from this analysis, as
they were not included in both forms D and G of the RVS)
19
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Table 2
Findings from Prior Research Studies Regarding Relationships of Expressed Levels of
Preference for (Importance of) Selected Instrumental Value Statements to
Indicators of Standing on Measures of Moral Reasoning
Instrumental
Value Statement
Related Investieations
McLellan
1970
Lockley Stanching Parish et al. Feather
1976 1976 1979-80 1988
Ambitious Pre [Pri]
Broadminded Pri Pri
Clean [Pri] [Pri]
Courageous [Pre]
Forgiving Pri/f
Honest [Pri/m]
Logical Pri Pri
Obedient SI, S4 [Pri] [Pri]
Polite Pre, [Pri]
Note: Pri = Principled reasoning [ ] = negative correlation
Pre = Preconventional reasoning S1 = Kohlberg's Stage 1 or moral reasoning
m = male subjects only S4 = Kohlberg's Stage 4 of moral reasoning
f = female subjects only
(Cheerful, courteous, and happiness have been omitted from this analysis, as
they were not included in both forms D and G of the RVS)
20
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
for obedient associated with Stages 1 and 4. He concluded that the RVS was better
suited for use with adult subjects than with school-aged subjects. Because he stated
that the value instrumentation was ill-suited to his sample, his findings are not central
to this analysis.
hi another study of older subjects, Lockley (1976) employed the DIT and the
RVS with her sample of 190 male undergraduates. She found a predominance of
principled moral reasoning to be negatively associated with level of preference for
value statements designated as a comfortable life, salvation, ambitious, clean,
obedient, and polite. A predominance of conventional reasoning was positively
associated with preference level for value statements of inner harmony,
broadminded, courageous, imaginative, and logical. Placement at the level of
preconventional reasoning was positively associated with rank order of preference
for the value statements of an exciting life, happiness, ambitious, and polite, but
negatively associated with stated preference order for the value statements of
freedom and courageous. However, when she analyzed the distributions by chi
square, only three values statements were ranked differently by individuals classified
as principled or conventional moral reasoners. She pointed out that Rokeach (1973)
had cautioned that of the 36 values, two should turn out to be shown significant in
correlational studies at the .05 level by chance alone. For the sake of discussion,
only her correlational data are included in the present analysis of research, and her
chi square data have been omitted in response to her cautionary statement. This
21
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
decision was also supported by the desire to compare results across studies, all of
which employed correlational analyses.
Standring’s 1976 research was reported by Rest (1979) and Parish et al.
(1979-1980). Using a sample that included 33 college students, Standring (1976)
employed the DIT and the portion of the RVS concerned with terminal values.
Standring found positive associations between placement on the meassure of moral
judgment and level of preference for value statements of equality, an exciting life,
freedom, inner harmony, pleasure, and a world at peace. A negative relationship
existed between standing on the moral judgment measure and level of preference for
the value terms fam ily security and national security.
In another study with 75 undergraduate students, Parish et al. (1979-1980)
also used the DIT and RVS. They found that scores on the measure of moral
judgment were positively related to the rank order assigned to value statements of
equality and pleasure for both males and females. Additionally, among males, level
of moral judgment as measured was negativley associated with rankings given to
value statements of both mature love and being honest. Moral judgment scores were
positively correlated with rank orders of value statements offamily security and
social recognition among males and with those of value terms of happiness,
forgiving, and cheerful among females. (It should be noted that the value statements
of happiness and cheerful have been omitted from later forms of the RVS.)
22
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Using the DIT and RVS, Feather (1988) investigated 133 Australian high
school students. He found level of moral judgment as measured to be positively
related to the rank order given to value statements of inner harmony, ambitious, and
being logical, but negatively associated with rankings assigned to the value terms of
clean and obedient.
Glover (1991) investigated a sample which included male and female students
in grades 6,9, and 12. Among her sample, 87.5% was Caucasian, and 73.1%
belonged to the Church of Jesus Christ of Latter-day Saints. She found no
practically significant differences in value statement rankings across stages of
measured moral reasoning. The lack of significant findings within this extremely
homogeneous group is to be pivotal to the current interpretation of existing literature
concerning the relationship between scores on a measure of moral reasoning and
rank orders assigned to preference levels for statements of various human values.
Her findings are not included in Table 1 or Table 2 because of the lack of significant
relationships between rank orders indicating level of preference for value statements
and scores on the measure of moral reasoning.
In a study with 111 students in a professional graduate school of business,
Weber (1993) obtained data on subjects, most of whom were White (89%) and/or
male (69%). Weber’s investigation was not directly comparable to the previous
studies for three reasons: (a) he used a 7-point Likert scale instead of rank-ordering
the values, (b) he did not provide correlations between stage of development and
23
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
level of value preference, and (c) he analyzed his data according to “value
orientations" instead of by individual values. For these reasons, his findings also are
not included in Table I or Table 2.
In his study, Weber assigned each value statement a score on a 7-point Likert
scale. The score for each value item was then multiplied by a weighting score (1 to
5) based on how consistently in past research that particular value statement had
been grouped into (personal, social, moral, and competence) subcategories. Each
weighted value was summed and then divided by the weighted coefficient to arrive at
a weighted group mean for each of the subcategories. If the subject’s mean score for
the value term was greater than that for a subgroup mean, then the subject was
presumed to have that value orientation.
The findings of Weber’s (1993) study in which the resulting value
orientations were related to a corresponding stage of moral reasoning were as
follows:
Stage 3 reasoning personal-competence orientation
Stage 4 reasoning social-competence orientation
and personal-moral orientation
Stage 5 reasoning social-moral orientation
Weber’s work represented a possible alternative way of viewing the interaction
between moral reasoning and human values. However, the extreme differences in
methodology in the study by Weber and in that of other reviewed investigations
24
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
make direct comparisons impossible. For this reason, his work has not been included
in the current analysis.
Proposed Theoretical Explanation of Past Findings
As was stated before, only the research findings of Lockley (1976), Standring
(1976), Parish et al. (1978-80), Feather (1988), and Glover (1991) have been
analyzed. Some of the previously cited findings exhibited consistency between two,
but never more than two, studies. Others showed contradictory findings between
two studies. Even when two studies yielded similar compatible findings, the results
would constitute a pattern which existed in only 40% of the occasions (two of five
studies). Certainly, strong interpretations cannot be made on the basis of such weak
evidence.
The lack of significant findings in the Glover (1991) study may be interpreted
in a way which illuminates the apparent disorder which characterizes the data.
Glover found the value terms of mature love, salvation, and fam ily security to be the
most frequently preferred across grade in school and gender. She noted that “ while
the intent of this project was not to examine the relation of religious affiliation to
value system, it seems apparent that the overwhelming homogeneity of values across
grade, stage of moral reasoning, and sex of this sample is most probably a product of
the influence of the Church of Jesus Christ of Latter-day Saints (LDS) in the lives of
this population” (p. 936). Thus, she attributed the lack of relationship of value
25
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
statement preferences and scores on a measure of moral reasoning to demographic
characertistics, rather than to level of moral reasoning.
Other Investigations . Unkmg Value Preference
tQ-Hemogaphic. Variables
Variations in value preference have been linked to a number of individual
differences: age (Rokeach, 1973), anti-nuclear activist behavior (Matyon &
Fumham, 1994), attitude towards taxation (Fumham, 1983), cultural identity
(Feather, 1986; Fumham, 1984; Katz, Juni, & Shope, 1993; Reynolds, 1984),
ethnicity (Matyon & Sangster, 1992; Rokeach, 1973), gender (Femald & Solomon,
1987; Lau, 1988, Rim, 1984), income and education (Rokeach, 1973), intelligence
(Rim, 1984), personality dimensions (Rim, 1984), and religious affiliation (Sohlberg,
1986). Thus, although significant findings appeared in several of the studies linking
moral development with value preference, it may be that these findings were
coincidental: the findings may have been primarily a result of the influence of
demographic characteristics within the sample, rather than because of any important
relationship between moral reasoning and human values. The lack of consistent
findings across studies, as well as in the extreme case of Glover’s (1991)
extraordinarily homogeneous sample producing consistent value ratings, may be
interpreted as evidence of this alternative explanation of the potential influence of
demographic variables.
26
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
A Rival Hypothesis: Values and Other
Demographic Variables
One possible alternative explanation for the findings is that expressed
preference levels for statements of human values are more strongly correlated with
demographic characteristics than with scores on a measure of moral reasoning. This
explanation would advance the possibility that the relationships found between
operationalized indicators of moral reasoning and of human values were only sample-
specific, and thus not generalizable. For instance, several values statements which
differentiated subjects by income level in Rokeach’s (1973) work also catagorized
subjects by stages of moral reasoning in the present studies reviewed.
Rokeach (1973) found that clean was the value term which most clearly
differentiated subjects by income level, with low income subjects ranking the
statement clean higher than high income subjects. Salvation and obedient value
terms similarly separated subjects by income level. In each of these cases the levels
of preference for the statements of values mentioned were negatively correlated with
scores on a measure of principled reasoning in two of the four studies of moral
reasoning and human values. Furthermore, rank orders of preference for inner
harmony and logical value statements were found by Rokeach (1973) to be related
to higher levels of income. Preference levels for inner harmony as a value term
were correlated positively with scores on the measure of principled reasoning in three
of the four studies. The degree of preference for the value term of logical was
27
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
correlated positively with scores on the measure of principled reasoning in two of the
four studies.
Unfortunately, this relationship also was not infallible. Although rank orders
assigned to the value statement offamily security were correlated negatively with
scores on the measure of principled reasoning in two of four studies, the negative
relationship was more characteristic of affluent than of less advantaged subjects in
Rokeach’s (1973) report.
Another demographic variable included in Rokeach’s (1973) discussion was
level of education. In this area, subjects low in level of education also valued the
term clean more than subjects in high level of education. Among two of the four
studies on values and moral reasoning, it was reported that level of preference for the
value term clean reflected an inverse relationship with standing of a measure of
principled reasoning. Similarly, the term logical was valued more by subjects high in
education level than by respondents low in education level in Rokeach’s (1973)
review. The level of preference for this value statement showed a direct relationship
with scores on a measure of principled reasoning in two of the present studies.
The hypothesis that values statements when rank ordered for preference level
are more highly correlated with demographic variables than with scores on a measure
of moral reasoning was further supported by the Glover (1991) study. Her sample
was extremely homogeneous across religious and racial variables. She achieved
highly consistent findings among the value terms preferred by individuals in her
28
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
sample, although preference levels for these values statements were totally unrelated
to scores on a measure revealing level of moral reasoning.
This information may also be compared with the finding that moral reasoning
(as measured by the DIT) has been shown to be moderately correlated with high
levels of education but weakly as well as inconsistently with social economic status
(SES) (Rest, 1979). (Rest noted that SES itself is often confounded with education
and that the data reviewed concerning SES were imprecise. Futhermore, SES often
includes measures of variables other than income level.)
This line of thought is not intended as a formal re-interpretation of past
findings. There is no basis for assuming comparability among the samples of subjects
for whom data have been reported or among many groups included in Rokeach’s
(1973) review. Moreover, even if the groups were acceptable for purposes of
comparison, demographic data were largely missing for the samples in the studies
reviewed. Instead, the current analysis of past studies is presented merely to
illustrate a possible rival hypothesis which is to be investigated in the current
research project.
Purpose of the Study
The twofold purpose of this study was to (a) determine for a total sample of
66 undergraduate university students the relationship of scores on a measure of
moral reasoning to the perceived relative levels of importance of statements of value
that Rokeach (1973) conceptualized as being instrumental (states of being) or
29
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
terminal (end states of existence) and (b) ascertain for various subgroups classified in
terms of certain demographic characteristics whether the pattern of relationships
between the scores on the measure of moral reasoning and the perceived relative
levels of importance assigned to the value statements were comparable to those for
the total sample.
It is possible that the apparent relationships between measures of human
values and those of moral reasoning which were proposed in earlier research
investigations were in fact false, a function of spurious demographic variables instead
of an indication of true relationships. If so, this circumstance would point toward the
possibility that measures of moral judgments and of human values provide data
suggesting their orthogonality, rather than a systematic relationship, as had been
previously hypothesized. If the patterns of relationship between preference for value
statements and level of moral reasoning as measured vary across demographically
defined subgroups, then this circumstance may serve as evidence of the possibility
that human values and moral reasoning are not related in any universal or necessary
way. In this case, an argument may be made for the theoretical integration of values
(and valence/expectancy theory) into the motivational domain of Rest’s four
component theory of moral behavior (Narvaez & Rest, 1995).
Research Question
In the present study, moral reasoning has been operationalized as
performance on the Defining Issues Test (Rest, 1993). Human values has been
30
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
operationalized as performance on the Rokeach Value Survey - Form G (Rokeach,
1983). Consistent with the purpose of this investigation, one central research
question was posed as follows:
For each of several subgroups from a total sample of 66
undergraduate university students which had been formed in terms of
a specific demographic classification, what was the degree of
relationship of scores on a test of moral reasoning developed by Rest
(1993) to the rank orders of importance that members of each of
these subgroups assigned to each of (a) 18 terminal value statements
and (b) 18 instrumental value statements proposed by Rokeach
(1983)?
Hypothesis
Consistent with the research question just stated, the following hypothesis is
advanced pertaining to the present research study.
Statement of the Hypothesis
Across various demographically defined subgroups, there would be no
consistent patterns of relationship between scores on the moral reasoning measure
and the rank orders of importance that members of each of these subgroups assigned
to each of (a) the 18 instrumental value statements and (b) the 18 terminal value
statements. It was further anticipated that any patterns which might exist within the
31
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
subgroups would be inconsistent with any pattern which may emerge for the larger
group as a whole, as well as inconsistent with patterns found in prior research.
Rationale for the Hypothesis
As was revealed in Table 1 and Table 2, prior research studies have produced
inconsistent, and sometimes contradictory, findings regarding the hypothesized
relationship between level of moral reasoning and priority rankings for value
statements. It is possible that any relationship which appears to exist between level
of moral reasoning as measured and preference rank for value statements is sample-
specific— a relationship that is an artifact of the influence of demographic variables on
a value preference.
Some evidence within past investigations would indicate that demographic
variables may have asserted a powerful influence on value preference. For instance,
Parish et al. (1979-1980) discovered gender differences within the observed
relationships between level of moral reasoning and preference for value statements
(i.e., in some instances, different value statements were related to comparable levels
of moral reasoning for males and females). Furthermore, Glover (1991) concluded
that one demographic variable (religious affiliation) influenced value statement
preference strongly enough to negate the possibility for any significant and consistent
relationship to exist between value statement preferences and level of moral
reasoning.
32
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
The current hypothesis extends the scope of past research by suggesting that
inconsistent patterns of correlation may exist between the level of moral reasoning
and strength of preference for value terms across several demographically defined
subgroups and furthermore that these patterns would be inconsistent with any pattern
which may emerge for the larger group as a whole.
Delimitations
The following delimitations were applicable to this investigation:
1. The present research did not represent an attempt to investigate moral
behavior. Instead, an effort was made to exmine the relationship between rank
orders of importance assigned to human value statements and level of moral
reasoning as measured within each of several demographically defined subgroups, as
a step toward integrating the two constructs as possible components of the actual
manifestation of moral behavior.
2. Only one undergraduate class of university students participated.
3. Only one measure of moral reasoning and only one scale of value
preferences were employed.
Conceptual Assumptions
The following conceptual assumptions were important to the research study
being conducted:
1. Psychological constructs such as human value systems and moral
reasoning exist.
33
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
2. Rokeach's (1973) theoretical formulation of human values and Rest's
(1979) theoretical formulation of moral reasoning offer a conceptual framework
within which the results of this study might be interpreted.
Organization of the Remainder of the Study
Chapter II contains a description of (a) the sample employed in this research,
(b) the variables and instruments used to measure them, (c) procedures in data
collection, (d) methods of data analyses, (e) methodological assumptions, and
(f) limitations. Chapter in comprises three parts: (a) an analysis of the findings
within the context of the research question, (b) an interpretation of the data in
relation to the tenability of the research hypothesis, and (c) a discussion of the
results. Chapter IV includes a summary of the entire research endeavor as well as
pertinent conclusions and recommendations.
34
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER n
METHOD AND PROCEDURES
This chapter affords information regarding the method and procedures used in
the study. The six major areas of this chapter include a description of the:
(a) research sample, (b) variables studied and their measures, (c) procedures employed
in data collection (d) the methods used in data analysis, (e) methodological
assumptions, and (f) limitations present.
Research Sample
The research sample originally consisted of 93 students enrolled in an elective
undergraduate university class in ethics reflecting aspects of personal development. A
check for the internal consistency of responses on the previously identified DIT
measure for moral reasoning indicated that the scores of 23 subjects were likely to
indicate a lack of understanding of the items and thus to reflect questionable validity.
After these cases had been discarded, the number of remaining subjects for whom
dependable responses could be analyzed was reduced to 66. A qualitative analysis of
the rejected cases did not reveal a proportionately heavier concentration of the
eliminated individuals for any particular demographic group. (Henceforth, the total
sample for the research project was considered to include only the remaining 66
subjects.)
35
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Variables and Their Measures
Among the variables in the study, eight were classifactory and two were
criterion measures.
Classifactory Variables
Students provided information concerning classificatory (demographic)
variables listed on a sheet with the heading "Demographic Information Page."
Information pertained to such characteristics as (a) religious upbringing, (b) ethnicity,
(c) chronological age, (d) gender, (e) amount of financial aid for which the respondents
were eligible, (f) level of education of father, and (g) level of education of mother.
Religious upbringing. Subjects described their religious upbringing as falling
within seven categories: Buddhist, Catholic, Hindu, Judaism, Muslim, Protestant, or
other. Because of the small numbers of subjects indicating religious affiliation within
many of the groups, separate statistical analyses according to religious group were
calculated only for the subjects with Catholic (n = 21) or Protestant (n = 24)
upbringings. Of the 45 participants, 46.7% and 53.3%, respectively, were Catholic or
Protestant in religious identification.
Ethnicity or race. The students defined their racial identity as being primarily
African-American/Black, Asian-American/Asian, Hispanic/Latino/Chicano, Native-
American, Pacific Islander, White, or Other. Racial groups large enough to warrant the
calculation of separate analyses were the following: African/American/Black (n = 8),
Asian-American/Asian (n = 12), Hispanic/Latino/Chicano (n = 10), and White (n = 33)
36
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
subjects. For the total number of 63 individuals associated with these four subgroups,
the respective percentages were 12.7, 19.0, 15.9, and 52.4.
Gender. Females comprised 56.1% (n = 37), and males represented 43.9%
(n = 29) of the total sample of 66 subjects.
Age. The chronological age of the subjects ranged from 17 to 34 years with a
mean of 20.2 years. For the purposes of analysis, subjects were dichotomized into two
groups comprising those 22 years of age or younger (n = 55) or 24 years of age or
older (n = 11). Corresponding percentages of representation were 83.3 and 16.7.
Amount of financial aid eligibility. Subjects indicated their level of eligibility for
need-based financial aid in increments of $2,000, ranging from no eligibility to
eligibility for $20,000 or more in aid. For the purposes of statistical analysis, this
information was categorized with reference to two groups: (a) subjects eligible for
$8,000 or less of need-based financial aid or (b) subjects eligible for $12,001 or more
of need-based financial aid. With the elimination of 11 cases (or the 16.7% of the
subjects who did not disclose their status of need for financial aid), 51.5% of the
remaining subjects (n = 34) was in the group of lower financial aid eligibility, and
31.8% of the remaining subjects (n = 21) was eligible for higher levels of need-based
aid.
Level of education of father. The individuals participating in the study provided
information concerning the level of formal education achieved by each of their parents.
In both cases, subjects chose between the categories of grade school, junior high, high
37
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
school, some college, college graduate, or graduate degree. Similarly, in both cases the
information was dichotomized into groups of fathers and mothers who received a
baccalaureate degree or an even higher degree or those parents who were not
graduated from college. Of the 66 respondents 56.1% (n = 29) indicated that their
fathers had attained at least a baccalaureate level of education, whereas 33.9% (n = 37)
reported that their fathers had not been graduated from college.
Level of education of mother. Within the sample of 66 students, 40.9% of the
subjects (n = 27) indicated that their mothers had been graduated from college, and
56.1% of the participants (n = 37) described their mothers as not having been
graduated from college. Two of the subjects, or 3.0%, neglected to report on their
mothers' level of educational achievement.
Dependent Variables
Moral reasoning. Level of moral reasoning was measured by the Defining
Issues Test (DIT) (Rest, 1993). As was indicated in Chapter I, the DIT is a multiple-
choice test of moral reasoning. An internal reliability check for possible ambiguities in
meaning or consistent intrepretability of responses was built into the DIT. This
checking device was intended to identify those subjects whose answers revealed a lack
of understanding of the items.
The P% (percentage) score is the one that has been primarily reported in
research with the DIT. It represents the percentages of responses associated with
principled reasoning (i.e., Stages 5 and 6 in moral reasoning) to which subjects
38
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
assigned a high priority, when compared with the relative importance they assigned to
responses representing lower stages of moral reasoning.
Human values. Preferences among statements of human values were measured
by the Rokeach Value Survey-Form G (RVS) (Rokeach, 1983). The RVS involves
placing in rank order two sets of 18 statements of value preference (18 terminal values
--desirable ends or states of existence— and 18 instrumental values— desirable ways of
being). Subjects assigned rank orders of preference from 1 (highest in preference) to
18 (lowest in preference) for the two sets of 18 instrumental and 18 terminal statements
of values. As indicated previously, the rank orders were reversed in the scoring
process so that 18 was the highest possible score and 1 was the lowest possible score
corresponding to the most preferred and least preferred statements.
Procedures in Data Collection
On the first day of class (in September, 1995), the students were invited to take
part in a research study concerning college students' beliefs. The professor of the class
urged student participation and introduced the researcher to the class. The researcher
informed the students that this activity was voluntary and that it would have no bearing
on their grade in the class. All students were assured that their responses would be
kept in complete confidentiality. The teaching assistants then handed out packets
including the (a) Informed Consent Page, (b) Demographic Information Page,
(c) Defining Issues Test (DIT), and (d) Rokeach Value Survey (RVS).
39
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
The researcher read through the printed instructions for each of the instruments
contained in the research packets. The students were asked whether they had any
questions regarding the project. They were requested to self-administer the
instruments at home and were told to return the completed forms at the next class
meeting. They were asked not to discuss their answers with anyone else and were
given the researcher's home telephone number in case they had questions or concerns
which needed to be discussed.
Data Analysis
The relationships between level of moral reasoning as revealed by scores on the
DIT and rank orders indicative of relative degree of preference for each of the value
statements on the RVS were analyzed for the entire group and for each of several
demographically defined subgroups within the total sample of 66 undergraduate
students. Primary attention was given to data relevant to the various demographic
subgroups as posed in the research question in Chapter I.
For the total sample, Pearson product-moment correlation coefficients were
calculated: (a) one between the P% scores earned on the DIT (moral reasoning
measure) and the rank order of preference assigned to each of the 18 terminal value
statements of the RVS (the measure of human values) and (b) another between the P%
scores earned on the DIT and the rank order assigned to each of the 18 instrumental
statements of the RVS. For the total sample, as well as for each of the 16 subsamples
or subgroups associated with a demographic classification (with frequencies high
40
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
enough to permit meaningful statistical analyses in the judgment of the researcher),
both chi square analyses and phi coefficients were calculated to ascertain the degree of
relationship between performance on the DIT and the RVS. Scores on the DIT were
categorized as being above or below the group mean. Rank orders of preference for
both the terminal and instrumental sets of value statements on the RVS were classified
into three categories of high (13 to 18), medium (7 to 12), or Iqjy (1 to 6).
In the instance of rank orders assigned to value statements in the RVS,
placement in the high, medium, or low groups was derived in terms of the rank orders
generated by each individual, not in the terms of mean values assigned to rank orders
either by members of the total sample or by individuals in the various subgroups.
Whenever subgroups exceeded 20 in number the previously mentioned chi square
analyses were carried out as described. In the instance of groups smaller than 20 the
classification of rank orders for the expressed importance of value statements was
dichotomized into high (10 or above) or low (9 or below). In addition, the Fisher
Exact Probability Test was applied to the resulting four-fold contingency tables. Phi
coefficients from the 3 x 2 contingency tables were obtained after reducing the tables
into a 2 x 2 format (four-fold contingency table) by collapsing into one category any
two adjacent categories associated with the three groupings of rank orders.
In the previous section of this chapter concerned with the description of the
various subgroups arising from demographic classifications, statements were made
concerning those subgroups of subjects for whom statistical analyses were carried out.
41
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Admittedly, several of the subgroups were quite small— a fact which might have limited
to some extent the meaningfulness of the results arising from the statistical analyses.
The alpha level of .05 was chosen as the indicator of statistical significance for the
correlation coefficients.
Methodological Assumptions
The following methodological assumptions were made:
1. The research design, data collection procedure, and methods of statistical
analysis were appropriate to the intent of the investigation.
2. The measurement instruments selected for use in the study were sufficiently
reliable and valid indicators of the constructs they were intended to represent.
3. The tests were administered properly and scored correctly.
4. The data were recorded and analyzed accurately.
5. The condition of data gathering did not affect the subjects.
6. The subjects responded candidly and accurately in their responses to items
on the research questionnaires (unless indicated otherwise by the built-in indicator of
internal consistency on the DIT).
Limitations
The following limitations might exist for this study:
1. The findings might not be generalizable to all socio-economic and age
groups.
42
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
2. Approximately 30% of the sample was lost to checks of intemal-consistency
on the DIT, as a result of identification of what appeared to be invalid responses.
3. The ipsative nature of the RVS made difficult the determination of an
intemal-consistency estimate of reliability of scores.
4. The ipsative characteristic of the RVS could lead to a relatively low ranking
of the importance of a given vahie statement even though at an absolute level the
respondent would consider that value to be of at least moderate to high importance.
5. In many instances, the subgroups were too small for reliable and valid
generalizations of the findings.
43
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER IE
ANALYSIS, INTERPRETATION, AND DISCUSSION
OF RESULTS
This chapter is organized in three major sections: (a) an analysis of the
statistical outcomes, (b) an interpretation of the findings within the framework of the
major hypothesis, and (c) a brief discussion of the results in relation to those obtained
by other investigators. In the third and closing portion of this chapter, the results of
this current investigation are related to only those other relevant inquiries completed
since 1976.
Analysis of Findings
This section of the chapter is divided into two parts. The first part is concerned
with the correlational results obtained for the total sample of 66 subjects. The second
part, which is relevant to the research question posed as well as to the major hypothesis
of the study, pertains to the statistical outcomes associated with each of the 16
subgroups corresponding to a specific demographic classification.
Correlational Results Obtained for the Total
Sample of 66 Subjects
When for the total sample a Pearson product-moment coefficient was calculated
between the P% scores earned on the DIT (moral reasoning measure) and the rank
order of preference assigned to each of the 18 terminal value statements of the RVS
(the measure of human values), the coefficients varied between -.178 and .239, none of
44
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
which was statistically significant at the .05 level. Ten of the coefficients were negative
and eight were positive. When the variables were categorized as described in the
previous chapter only two phi coefficients of .355 and .375 were statistically
significant— namely those for the value terms of a comfortable life and freedom,
respectively. This first coefficient reflected a tendency for those individuals who placed
at a low level in the measure o f moral reasoning to express a medium degree of
preference for the value statement of a comfortable life. The second coefficient
indicated a tendency for those individuals who placed high in the measure of moral
reasoning to reveal a high degree of preference for the value statement offreedom.
Correspondingly when Pearson product-moment correlation coefficients were
calculated between scores on the DIT and rank orders assigned for preferences of
instrumental value statements, the range of the correlational indexes fell between -.144
and .105, none of which was statistically significant. Nine of the coefficients were
positive, and nine were negative. Only one phi coefficient of 0.314 was statistically
significant— namely that between scores on the DIT and the rank orders given to the
instrumental value statement of capable. This coefficient was interpreted to suggest
that those individuals low in their level of moral reasoning tended to have a medium or
moderate preference for the value statement of capable. (This investigator would place
a higher level of confidence in the information provided by the Pearson i coefficients
involving continuous or ordered data than in the phi coefficients reflecting the extent of
45
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
association between roughly categorized variables in which a substantial loss of
information has occurred.)
Statistical Outcomes Corresponding to 16
Demographically Identified Subgroups
Relationship between scores on the PIT (measure of moral reasoning’ ) and rank
order of preference for terminal value statements on the RVS. In Table 3 only those
results which yielded statistically significant relationships are reported. The phi
coefficients in this table reflected the amount of association between scores categorized
as being at or above the mean or below the mean in the DIT (the moral reasoning
measure) and placement in one of three categories of being high, medium, or low in the
rank order of preference assigned to terminal value statements of the RVS (the measure
of human values) are cited. Of the 288 possible phi coefficients associated with the 16
demographically identified subgroups, only 17 yielded statistically significant phi
coefficients. This frequency indicates that only 5.9% of the correlational indexes was
statistically significant at the .05 level. This feet suggests that by chance alone one
would have expected 14.4 coefficients to have obtained statistical significance relative
to the .05 level. The reader may examine Table 3 to ascertain for her or his satisfaction
which correlation coefficients were indeed statistically significant.
Mention might be made of the fact that relative to a more stringent significance
level of .01, five coefficients achieved statistical significance at the .01 level. One
would expect that by chance alone about 3 of the coefficients would be significant at
46
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission.
Table 3
Indicators in Various Demographic Subgroups of Only Those Relationships That Were Statistically Significant
Between Scores on the DIT (Measure of Moral Reasoning) and Rank Orders of Importance Assigned to
Terminal Value Statements from the RVS
Demographic Group
and Value Statement
N
X2
Phi
P
Fisher's Exact
Probability Test
Dominant Cell in
Contingency Table
Catholic
A Comfortable Life 21 7.431 0.595 0.024
-
B, E
Freedom 21 6.496 0.556 0.039
-
E
True Friendship 21 7.289 0.589 0.026
- B
Protestant
A Comfortable Life 24 6.436 0.518 0.040 - E, C
White
Pleasure 33 6.240 0.435 0.044
-
C
True Friendship 33 9.809 0.545 0.007 - D
Female
A Comfortable Life 37 7.680 0.455 0.021
Inner Harmony 37 6.115 0.407 0.047
E
D
Age 22 and younger
A Comfortable Life 55 12.663 0.480 0.002
Freedom 55 9.585 0.417 0.008
E
D
^4
Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission.
Table 3 (Continued)
Demographic Group
and Value Statement
N
x2
Phi
P
Fisher's Exact
Probability Test
Dominant Cell in
Contingency Table
$12,000 and above financial aid
Freedom 34 6.734 0.445 0.034
.
D
$8,000 and below financial aid
A Comfortable Life 21 6.077 0.538 0.048
-
E
Father not college graduate
Freedom 37 10.274 0.527 0.006
-
D
Mother bacalaureate degree or more
A Comfortable Life 27 7.200 0.516 0.027 E
Pleasure 27 8.168 0.550 0.017
-
C
A World at Peace 27 6.435 0.488 0.040 - A
Mother not college graduate
Freedom 37 9.847 0.516 0.007 - E
Note: A = low level moral reasoning/low value statement preference
B = low level moral reasoning/high value statement preference
C = high level moral reasoning/low value statement preference
D = high level moral reasoning/high value statement preference
E = low level moral reasoning/medium value statement preference
F = high level moral reasoning/medium value statement preference
OO
the .01 level. Specifically, the five phi coefficients achieving such a level of statistical
significance occurred between categorized scores on the DIT and one of three levels of
expressed preference (high, medium, or low) for value statements on the RVS as
follows: (a) true friendship for Whites (those high in level of moral reasoning tending
to be high in the value term), (b) a comfortable life for those 22 years or younger in
age (those being low in level of moral reasoning likely to assign a medium value to the
statement), (c) freedom for those 22 years of age of younger (those placing high in
level of moral reasoning tending also to register a high level to the term), (d) freedom
for those fathers who had not been graduated from college (those placing high in level
of moral reasoning also putting a high value on the statement), (e) freedom for mothers
who had not been graduated from college (those placing low in level of moral
reasoning likely to express a medium value for the statement).
Relationship between scores on the PIT (measure of moral reasoning) and rank
order of preference for instrumental value statements on the RVS. In Table 4, only
those results which yielded statistically significant relationships between scores
categorized as being at or above the mean in the DIT (the moral reasoning measure)
and placement in one of three categories of being high, medium, or low in the rank
order of preference assigned to instrumental value statements of the RVS (the measure
of human values) are cited. One exception to this statement resulted with reference to
the sample of only 8 African-American/Black subjects for whom placement in the RVS
was in only two categories of high versus low (for this sample, the Fisher Exact
49
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission.
Table 4
Indicators in Various Demographic Subgroups of Only Those Relationships That Were Statistically Significant
Between Scores on the DIT (Measure of Moral Reasoning) and Rank Orders of Importance Assigned to
Instrumental Value Statements from the RVS
Demographic Group
and Value Statement
N
X2
Phi
P
Fisher's Exact
Probability Test
Dominant Cell in
Contingency Table
Catholic
Capable 21 7.471 0.596 0.024 - B, C
African-American/Black
Loving 8 8.000 - 1.000 0.005 0.036 B
Female
Loyal 37 7.695 0.456 0.021 - F
Age 22 and younger
Clean 55 6.040 0.331 0.049
- A, C
$8,000 and below financial
Capable
aid
21 7.469 0.596 0.024 C
Helpful 21 7.870 0.612 0.020 - D
Independent 21 7.891 0.613 0.010
- B
O
Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission.
Table 4 (Continued)
Demographic Group
and Value Statement
N
x2
Phi
P
Fisher's Exact
Probability Test
Dominant Cell in
Contingency Table
Father bacalaureate degree or more
Capable 29 6.032 0.456 0.049 E
Mother bacalaureate degree or more
Helpful 27 6.193 0.479 0.045 - E
Note: A = low level moral reasoning / low value statement preference
B = low level moral reasoning / high value statement preference
C = high level moral reasoning / low value statement preference
D = high level moral reasoning / high value statement preference
E = low level moral reasoning / medium value statement preference
F = high level moral reasoning / medium value statement preference
U \
Probability Test of significance was also calculated). Of the 288 possible coefficients
associated with the 16 demographically identified subgroups, only 9 yielded statistically
significant phi coefficients. This frequency indicates that only 3.1 % of the correlational
indexes achieved statistical significance when by chance alone one would anticipate
about 14 or 15 would have afforded statistically significant results. Again, the reader is
directed to consult Table 4 for details.
Of possible interest is the fact that only two coefficients obtained statistical
significance at the .01 leveL In the instance of the demographic subgroup of African-
American/Black the value term loving afforded a statistically significant degree of
association with placing high or low on the measure of moral reasoning. There was a
marked tendency for those exhibiting a low level in moral reasoning to assign a high
value to this value statement. It should be pointed out, however, that only eight
subjects were involved and that the occurrence of zeros in two cells of the four-fold
contingency table led to the result of a phi coefficient of -1.0, which probably would
not have been obtained in a much larger sample. Application of the Fisher Exact
Probability Test provided a significance level of 0.036. The other phi coefficient of
0.613 that was statistically significant at the .01 level occurred for the demographic
subgroup of 21 subjects represented by those students whose need for financial aid was
$8,000 or less. This coefficient was associated with the value term of independent.
There was a tendency for those placing low in their level of moral reasoning to assign a
high value to this statement.
52
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Interpretation of the Data in Relation to the Tenability
of the Research Hypothesis
It was hypothesized in Chapter I that across various demographicaUy defined
subgroups, no consistent patterns of relationship would be found to exist between
scores on the moral reasoning measure (DIT) and the rank orders of importance that
members of each of these subgroups would assign to each of the (a) 18 terminal value
statements of the RVS and (b) 18 instrumental value statements of the RVS.
It was further anticipated that any patterns which existed within the subgroups
would be inconsistent with any pattern which emerged for the larger group as a whole,
as well as with any patterns found in prior research. As was clearly illustrated by the
statistical outcomes for the total group and by those realized for the demographicaUy
identified subgroups (for whom data are reported in Tables 3 and 4), no patterns of
results were found to exist consistently across the demographicaUy defined subgroups,
and not one of the patterns for any subgroup matched the pattern of relationships
provided by the total group.
Similarly, as can be seen in Tables 5 and 6, the findings of the present study
were inconsistent with those of past researchers. (Explanation of the entries of these
last two tables are given in the foUowing section entitled Discussion.) Therefore, the
conclusion would be that the findings could be interpreted as being in support of the
research hypothesis which in itself constituted a nuU statement.
53
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Discussion
Table 5 furnishes a comparison of the correlational results obtained between
scores on a measure of moral reasoning (not necessarily the DIT) and preference for
terminal value statements of the RVS as reported by five other researchers as well as by
this investigator. Table 6 provides comparable information in the instance of
instrumental value statements. Some additional explanation may be needed for the
symbols in Tables 5 and 6. The symbol Pri refers to the occurrence of principled
reasoning as being indicative of a higher level of moral reasoning, whereas Pre stands
for preconventional reasoning, which portrays a relatively lower level of moral
reasoning. Brackets around either one of these symbols point to the occurrence of a
negative relationship. Symbols of m and f designate, respectively, an outcome
associated with male or female subjects only. Finally, the expressions of S1, S2, S4,
and S5 reflect placement at the first, second, fourth, and fifth stages of moral reasoning
as postulated by Kohlberg.
In light of the statistical findings of the current investigation, it would appear
that the constructs of moral reasoning, as measured by the DIT, and human values, as
measured by the RVS, were not related in any consistent or necessary way. This
finding may be considered important to the study of human morality because it
represents a first step toward integrating two of the four components which have been
hypothesized to interact in producing a moral act (Narvaez & Rest, 1995). Rest
(Narvaez & Rest, 1995) suggested that both moral reasoning and moral motivation
54
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Table 5
A Comparison of Statistically Significant Correlational Results Between Scores on a
Measure of Moral Reasoning and Level of Importance Assigned to Terminal
Value Statements Across Investigations
Terminal
Value Statement
Related Investigations
McLellan Lockley Standring Parish etaL Feather
1970 1976 1976 1978-80 1988
Frost
1997
A Comfortable Life [Pri] None
Equality Pri Pri for
An Exciting Life Pre Pri any
Family Security [Pri] Pri/m Pearson
Freedom S2, S5 [Pre] Pri r
Inner Harmony Pri Pri Pri
Mature Love [Pri/m] Pri
National Security
[Pri]
[Pri/f]
Pleasure Pri
Salvation [Pri] [Pri]
Social Recognition Pri/m
A World at Peace Pri
Note: Pri = Principled reasoning [ ] = negative correlation
Pre = Preconventional reasoning S2 = Kohlberg's Stage 2 of moral reasoning
m = male subjects only S5 = Kohlberg's Stage 5 of moral reasoning
f = female subjects only
{Cheerful, courteous, and happiness have been omitted from this analysis, as they were
not included in both forms D and G of the RVS)
55
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Table 6
A Comparison of Statistically Significant Correlational Results Between Scores on a
Measure of Moral Reasoning and Level of Importance Assigned to Instrumental
Value Statements Across Investigations
Instrumental
Value Statement
Related Investigations
McLellan
197Q
Lockley Stanching Parish etal.
1976 1976 1978-80
Feather
1988
Frost
1997
Ambitious Pre [Pri] None
Broadminded Pri Pri for
Capable any
Clean [Pri] [Pri] Pearson
Courageous [Pre] r
Forgiving Pri/f
Honest [Pri/m]
Logical Pri Pri
Obedient SI, S4 [Pri] [Pri]
Polite Pre, [Pri]
Note: Pri = Principled reasoning [ ] = negative correlation
Pre = Preconventional reasoning SI = Kohlberg's Stage 1 of moral reasoning
m = male subjects only S4 = Kohlberg's Stage 4 of moral reasoning
f = female subjects only
{Cheerful, courteous, and happiness have been omitted from this analysis, as they were
not included in both forms D and G of the RVS)
56
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
(conceived as involving human values), moral sensitivity, and implementation are four
separate components which together result in a moral action. Having produced
evidence of a relationship of independence between moral reasoning and human values,
the current study may be interpreted as advancing Rest's four component theory of
moral behavior.
If moral reasoning and human values are truly independent constructs, the
theoretical and practical implications are enormous. Theoretically, this conclusion is a
preliminary step toward satisfying Blasi's (1980) request for greater insights into the
relationships between moral cognition and moral action: if values serve a motivational
function in behavior which is separate from moral judgment, then the relative priorities
which an individual holds for particular values may help explain that individual's
disposition to carry out or to abstain from carrying out a moral act which has been
judged to be correct.
According to Feather’ s (1982) expectancy-valence theory of motivation, the
values held by an individual may serve to guide that individual's actions. Values are
said to motivate actions by inducing valences, or tendencies to approach or to avoid
objects or ways of being, and by selectively sensitizing individuals to certain aspects
within their environment. Thus, if an individual expects that he or she is capable of
carrying out a particular moral act and if he or she holds a hierarchically organized
value system which is compatible with the chosen moral act, then that individual would
likely be motivated to carry out that act. In supporting the hypothesized relationship of
57
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
independence between moral reasoning and human values, the current study suggests
that further theoretical work along this vein might prove fruitful.
The current study also has practical implications within the realm of predicting
and controlling moral acts. Whereas past studies have represented an endeavor to
explain and to predict moral behavior through an exploration of relationships between
human values and moral action (Weber, 1993), perhaps even more behavior might be
explained if the constructs of values and moral reasoning are not consistently related.
If the two constructs represent separate influences on moral behavior, and if one
obtains information on both constructs for an individual, then by a process of
triangulation a greater percentage of the individual's behavior becomes explicable. In
contrast, if moral reasoning and human values are significantly overlapping in content,
then these two concepts would represent only a single perspective on a given moral act
and thus would explain less of its nature.
The current study has even further practical implications for the field of moral
education. If both moral reasoning and human values may be implicated in the
production of a moral act, then education in either domain alone might be insufficient
for successfully modifying moral behavior. An individual who has received education
in the area of moral reasoning might be highly capable of making sophisticated moral
judgments, but lack the moral motivation to implement the chosen act. Similarly, an
individual who has been educated in the content of human values might be motivated to
act morally, but might lack the skills of moral cognition needed to arrive at a moral
58
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
decision. Therefore, if future research upholds the current findings, then moral
education might benefit from containing content applicable to both moral reasoning and
human values.
59
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER IV
SUMMARY, CONCLUSIONS, AND RECOMMENDATIONS
Summary
Background
A handful of studies has been concerned with the possible relationships that
exist between moral reasoning and human values (Feather, 1988; Glover, 1991;
Lockley, 1976; McLellan, 1970; Parish, Rosenblatt, & Kappes, 1979-80; Weber,
1993). All but one have revealed significant results; however, the findings have been
inconsistent with one another. The rationale behind this investigation was that the two
constructs of human values and moral reasoning actually represent two separate
component processes involved in the production of moral behavior— a position
consistent with that implied by the four component theory of moral behavior (Narvaez
& Rest, 1995). As such, no particular relationships necessarily would be expected to
exist between these two constructs.
In conceptual rather than operational terms, the central hypothesis in this study
was that there would be no consistent patterns of relationship between the two
constructs of moral reasoning and human values across several demographicaUy
defined subgroups that were formed from a sample of undergraduate university
students. Moreover, it was anticipated that this lack of a pattern of a consistent
relationship between the two constructs would hold not only in the total sample but
also in the samples studied by other researchers. In other words, it was expected that
60
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
any patterns of findings in the current investigation would not replicate in a consistent
way those found in other studies.
Purpose of the Study
The twofold purpose of this study was to (a) determine for a total sample of 66
undergraduate university students the relationship of scores on a measure of moral
reasoning to the perceived relative levels of importance of statements of value that
Rokeach (1973) conceptualized as being instrumental (states of being) or terminal (end
states of existence) and (b) ascertain for various subgroups classified in terms of certain
demographic characteristics whether the pattern of relationships between the scores on
the measure of moral reasoning and the perceived relative levels of importance assigned
to the value statements were comparable to those for the total sample.
It is possible that the apparent relationships between measures of human values
and those of moral reasoning which were proposed in earlier research investigations
were in fact false, a function of spurious demographic variables instead of an indication
of true relationships. If so, this circumstance would point toward the possibility that
measures of moral judgments and of human values provide data suggesting their
orthogonality, rather than a systematic relationship, as had been previously
hypothesized. If variability exists across demographically defined subgroups in the
patterns of relationship between preference for value statements and level of moral
reasoning as measured, then this circumstance may serve as evidence of the possibility
that human values and moral reasoning are not related in any universal or necessary
61
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
way. In this case, an argument may be made for the theoretical integration of values
(and valence/expectancy theory) into the motivational domain of Rest’s four
component theory of moral behavior (Narvaez & Rest, 1995).
Research Question
In the present study, moral reasoning has been operationalized as performance
on the Defining Issues Test (Rest, 1993). Human values has been operationalized as
performance on the Rokeach Value Survey - Form G (Rokeach, 1983). Consistent
with the purpose of this investigation, one central research question was posed as
follows:
For each of several subgroups from a total sample of 66 undergraduate
university students which had been formed in terms of a specific
demographic classification, what was the degree of relationship of
scores on a test of moral reasoning developed by Rest (1993) to the
rank orders of importance that members of each of these subgroups
assigned to each of (a) 18 terminal value statements and (b) 18
instrumental value statements proposed by Rokeach (1983)?
Methods and Procedures
Total sample and subgroups. The total sampie for this research study
consisted of 66 undergraduate students in an ethics class with an emphasis upon
personal development. From this total sample of students 16 different demographic
subgroups were formed in terms of (a) religious upbringing (two subgroups of 21
62
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
and 24 being associated, respectively, with the Catholic and Protestant faiths),
(b) ethnicity or race (four subgroups comprising 8 African-American/Black, 12
Asian-American/Asian, 10 Hispanic/Latino/Chicano, and 33 White students),
(c) gender (37 females and 29 males), (d) age (two subgroups of 55 students 22
years of age or younger and 11 students 24 years or older), (e) amount of financial
aid for which eligibility existed (two subgroups consisting of 34 students eligible for
$8,000 or less of need-based financial aid and 24 students eligible for $12,001 or
more of need-based financial aid), (f) level of education of father (two subgroups of
students of whom 29 indicated that their fathers had achieved at least a
baccaulaureate level of education and 37 students of whom reported that their fathers
had not been graduated from college), and (g) level of education of mother (two
subgroups of students of whom 27 indicated that their mothers had achieved at least
a baccaulaureate level of education and of whom 37 reported that their mothers had
not been graduated from college).
Measures. Moral reasoning was measured by the Defining Issues Test (DIT)
(Rest, 1993), a multiple-choice measure of stages of moral reasoning comparable to
those proposed by Kohlberg. Preference for human values was measured by the
Rokeach Value Survey - Form G (RVS) (Rokeach, 1983). Two sets of values (18
terminal values, or desirable end-states of existence, and 18 instrumental values, or
desirable ways of being) were rank ordered according to preference by the subjects.
63
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Data analysis. The relationships between level of moral reasoning as revealed
by scores on the DIT and rank orders indicative of relative degree of preference for
each of the value statements on the RVS were analyzed for the entire group and for
each of several demographicaUy defined subgroups within the total sample of 66
undergraduate students. Primary attention was given to data relevant to the various
demographic subgroups just described.
For the total sample, Pearson product-moment correlation coefficients were
calculated: (a) one between the P% scores earned on the DIT (moral reasoning
measure) and the rank order of preference assigned to each of the 18 terminal value
statements of the RVS (the measure of human values) and (b) another between the
P% scores earned on the DIT and the rank order assigned to each of the 18
instrumental statements of the RVS. For the total sample, as well as for each of the
16 subsamples or subgroups associated with a demographic classification (with
frequencies high enough to permit meaningful statistical analyses in the judgment of
the researcher), both chi square analyses and phi coefficients were calculated to
ascertain the degree of relationship between performance on the DIT and the RVS.
Scores on the DU were categorized as being above or below the group mean. Rank
orders of preference for both the terminal and instrumental sets of value statements
on the RVS were classified into three categories of high (13 to 18), medium (7 to
12), o rio a (1 to 6).
64
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
In the instance of rank orders assigned to value statements in the RVS,
placement in the high, medium, or low groups was derived in terms of the rank
orders generated by each individual, not in the terms of mean values assigned to rank
orders either by members of the total sample or by individuals in the various
subgroups. Whenever subgroups exceeded 20 in number the previously mentioned
chi square analyses were carried out as described. In the instance of groups smaller
than 20 the classification of rank orders for the expressed importance of value
statements was dichotomized into high (10 or above) or low (9 or below). In
addition, the Fisher Exact Probability Test was applied to the resulting four-fold
contingency tables. Phi coefficients from the 3 x 2 contingency tables were obtained
after having collapsed two adjacent categories associated with the three groupings of
rank orders into one category. (In all statistical analyses, the alpha level was set at
0.05.)
Sstegted Findings
The following major statistical findings resulted from the analyses of the
research data:
1. In the instance of the total sample, the Pearson product moment
coefficients between the P% scores earned on the DIT (moral reasoning measure)
and the rank order of preference (or importance) assigned to each of the 18 terminal
value statements of the RVS (the measure of human values) varied between -. 178
65
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
and .239 (all p > .05); correspondingly, for the 18 instrumental value statements,
between -.144 and . 105 (all p > .05).
2. For the 16 subgroups arising from the various demographic classifications,
the frequency of statistically significant indexes of association between categorized
scores on the DIT and categorized rank orders of preference (or importance)
assigned to the (a) 18 RVS terminal value statements and (b) the 18 instrumental
value statements closely approximated what one would expect by chance alone. No
consistent patterns of relationship were noted across the total sample and the various
demographically determined subgroups of that total sample.
Conclusions
This study of the relationship between the constructs of human values and
moral reasoning as components of moral behavior suggested the following
conclusions:
1. In the total sample of respondents, the level of moral reasoning as
measured by the DIT exhibited virtually no relationship to expressed degree of
preference for (i.e., importance of) any Rokeach value statement, terminal or
instrumental.
2. In general, little, if any, relationship was evident across the 16
demographically defined subgroups between one's level of moral reasoning and the
degree of importance one tended to place on terminal or instrumental value
statements.
66
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
3. The overall pattern of findings in the current investigation revealed not
only support for the central hypothesis of the study but also a lack of congruence
with the patterns of results reported by other investigators.
Recommendations
In line with this research experience, the following recommendations for
future research are offered:
1. Any further research on the hypothesized relationship between moral
reasoning scores and level of preference for human value statements should include a
detailed analysis of differences in the patterns of relationships across demographically
defined subgroups. This recommendation is motivated by the possibility that any
apparent relationships which may be observed are in reality artifacts of demographic
influences on value preference.
2. Confidence in the current research findings would be strengthened by a
similar study which would include a larger number of subjects, both within the total
group and within each of the various demographically defined subgroups.
3. Ultimately, the area of study would benefit from an investigation of moral
behavior which would be organized in a design centered around the four component
theory (Narvaez & Rest, 1995) and which would be constructed to test more directly
any interactions likely to occur among the four components in the production of a
moral act.
67
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
REFERENCES
Ball-Rokeach, S., Rokeach, M., & Gurbe, J. (1984). The great American
value test. New York: Free Press.
Blasi, A. (1980). Bridging moral cognition and moral action: A critical review
of the literature. Psychological Bulletin. 88(1). 1-45.
Colby, A. & Kohlberg, L. (1987). The measurement of moral judgment
(Vol. 1). New York: Cambridge University Press.
Cortese, A. J. (1989). The interpersonal approach to morality: A gender and
cultural analysis. The Journal of Social Psychology, 129(41. 429-441.
Feather, N. T. (1975). Values and education in society. New York: Free
Press.
Feather, N. T. (1982). Values, expectations, and the prediction of social
action: An expectancy-valence analysis. Motivation and Emotion. 6(3). 217-244.
Feather, N. T. (1986). Value systems across cultures: Australia and China.
International Journal of Psychology. 21. 697-715.
Feather, N T. (1988). Moral judgement and human values. British Journal of
Social Psychology, 27. 239-46.
Femald, L. W. & Solomon, G. T. (1987). Value profiles of male and female
entrepreneurs. Journal of Creative Behavior. 21(3). 234-247.
Furnham, A. (1983). The Protestant work ethic: Human values and attitudes
towards taxation. Journal of Economic Psvchologv.3 . 113-128.
Furnham, A. (1984). Value systems and anomie in three cultures.
International Journal of Psychology. 19. 565-579.
Gilligan, C. (1982/1993). In a different voice: Psychological theory and
women's development. Cambridge, Massachusetts: Harvard University Press.
Glover, R. J. (1991). Value selection in relation to grade in school and stage
of moral reasoning. Psychological Reports. 68. 931-937.
68
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Katz, B., Juni, S. & Shope, C. (1993). The values of Chinese students: At
home and abroad. International Journal of Psychology. 28(6). 761-773.
Kohlberg, L. (1958). The development of modes of moral thinking and choice
in the years 10 to 16. Unpublished doctoral dissertation, University of Chicago.
Kohlberg, L. (1984). The psychology of moral development. San Francisco:
Harper and Row.
Kohlberg, L., Colby, A., Gibbs, J., Speicher-Dubin, B. & Power, C. (1977).
Standard issue scoring manual. Cambridge, MA.: Center for Moral Development and
Education, Harvard University.
Kurtines, M. K., & Gewirtz, J. L. (1995). Moral development: An
introduction. Needham Heights, MA- Allyn and Bacon.
Lau, S. (1988). The value orientations of Chinese university students in Hong
Kong. International Journal of Psychology. 23. 583-596.
Lockley, O. E. (1976). Moral reasoning and choice of values among students
at Rutgers University. Unpublished doctoral dissertation, State University of New
Jersey, Rutgers.
Mayton, D. M. & Furnham, A. (1994). Value underpinnings of antinuciear
political activism: A cross-national study. Journal of Social Issues, 50(41 117-128.
Mayton, D. M. & Sangster, R. L. (1992). Cross-cultural comparison of values
and nuclear war attitudes. Journal of Cross-Cultural Psychology. 23(31. 340-352.
McLellan, D. D. (1970). Values, value systems, and the developmental
structure of moral judgment. Unpublished master’s thesis, Michigan State University,
Lansing.
Moon,Y. (1986). A review of cross-cultural studies on moral judgment
development using the defining issues test. Behavior Science Research, 20(1-41. 147-
177.
Mueller, D. J. (1984). Rokeach value survey. In D. J. Keyser & R. C.
Sweetland (Eds.). Test critiques (Vol. I). Kansas City, Missouri: Test Corporation of
America.
69
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Narvaez, D. & Rest, J. (1995). The four components of acting morally. In W.
M. Kurtines & J. L. Gewirtz (Eds.). Moral development: An introduction, (pp. 385-
400).
Papalia, D. E. & Olds, S. W. (1995). Human development. New York:
McGraw-Hill.
Parish, T. S., Rosenblatt, R. R., & Kappes, B. M. (1979-80). The relationship
between human values and moral judgment. Psychology: A Quarterly Journal of
Human Behavior. 16(41.1-5.
Piaget, J. (1965). The moral development of the child. New York: The Free
Press.
Rest, J. R. (1979). Development in judging moral issues. Minneapolis:
University of Minnesota Press.
Rest, J. R. (1993). The defining issues test. Minneapolis University of
Minnesota (206-A Burton Hall, 17 Pillsbury Drive S. E., Minneapolis): Center for the
Study of Ethical Development.
Reynolds, B. K. (1984). A cross-cultural study of values of Germans and
Americans. International Journal of Intercultural Relations, 8. 269-278.
Rim, Y. (1984). Importance of values according to personality, intelligence
and sex. Personality and Individual Differences. 5(2). 245-246.
Rokeach, M. (1973). The nature of human values. New York: Free Press.
Rokeach, M. (Ed.). (1979). Understanding human values. New York: Free
Press.
Rokeach, M. (1983). The Rokeach value survey. Palo Alto, CA Consulting
Psychologists Press.
Snarey, J. R. (1985). Cross-cultural universality of social-moral development:
A critical review of Kohlbergian research. Psychological Bulletin. 97(21. 202-232.
Sohlberg, S. C. (1986). Similarity and dissimilarity in value patterns of Israeli
kibbutz and city adolescents. International Journal of Psychology. 21. 189-202.
70
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Standring, D. E. (1976). Values, value change, and attitudes to occupations as
a function of moral judgment level. Unpublished master’ s thesis, Victoria University of
Wellington, New Zeland.
Walker, L. J. (1982). The sequentiality of Kohlberg’s stages of moral
development. Child Development (53). 1330-1336.
Walker, L. J. (1984). Sex differences in the development of moral reasoning:
A critical review. Child Development. 55. 677-691.
Weber, J. (1993). Exploring the relationship between personal values and
moral reasoning. Human Relations. 46(41. 435-463.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
APPENDIX A
Rest’s Four Component Theory of Moral Behavior
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Rest’s Four Component Theory of Moral Behavior
I. Moral Sensitivity:
... involves receptivity of the sensory perceptual system to social situations and
the interpretation of the situation in terms of what actions are possible, who and
what would be affected by each of the possible actions, and how the involved
parties might react to possible outcomes,
n. Moral Judgment:
... involves deciding which of the possible actions is the most moraL The
individual weighs the possible choices and determines what a person ought to
do in such a situation.
HI. Moral Motivation:
... implies that the person gives priority to the moral value above all other values
and intends to fulfill it.
IV. Implementation:
... combines ego strength with the social and psychological skills necessary to
carry out the chosen action, (from Narvaez & Rest, 1995, p. 386).
73
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
APPENDIX B
Kohlberg’s Theory of Moral Reasoning
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Kohlberg’s Theory of Moral Reasoning
(Paraphrased by the Investigator)
Level I: Pre-Conventional Morality.
Stage 1: Heteronomous Morality. Perspective is naive moral realism. Right
and wrong are determined by authority. Moral significance of an
action is an inherent and unchanging quality of an action. Punishment
is associated with bad actions, and there is a lack of mediating
concepts (e.g., intentionality) and a lack of attention to particulars of
circumstance. Differences in individual perspectives are not
appreciated.
Stage 2: Individualistic, Instrumental Morality. Individual develops an
awareness that persons may have differing, but equally valid,
perspectives of a given situation. The right is particular to the
situation and the individual perspective. The assumption is that all
persons are working to satisfy as many needs and desires as possible,
while avoiding negative personal consequences. Judgments are
pragmatic; mutual benefit is the goal.
Level II: Conventional Morality
Stage 3: Interpersonally Normative Morality. Social norms are generalized
across individuals and situations. These are superior to Stage 1 rules
in that they include numerous individual perspectives which have been
75
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
coordinated into social agreements. Goal is maintaining interpersonal
trust and social approval
Stage 4: Social System Morality: Individual interests are subordinated to the
maintenance of a sociomoral system. A social structure consisting of
institutions and social roles protects the common good. Conflicts may
occur between (even good) occupants, and the system is needed for
resolving conflicts. Perspective is either that of a legal/societal system
with formalized rules and laws, or that of higher religious or moral law
within the individual's conscience. Judgments rely on institutional
code.
Level III: Postconventional Morality
Stage 5: Human Rights and Social Welfare Morality: Here the individual is
aware of universal values and rights which are prior to and supersede
institutional law. Rules are good to the extent that they uphold higher-
order rights and values. Based on rule-utilitarian goal of long-term
consequences for individual members of society.
Stage 6: Morality o f Universalizable, Reversible, and Prescriptive General
Ethical Principle(s): Each individual is considered with equal
importance in decisions. An example of this view is philosopher John
Rawls' "original position" of justice in which a choice is made without
knowledge of which effected party one is within the group, thus
76
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
insuring the best and fairest choice from each point of view. All
human beings share intrinsic worth, dignity, and equality (Kohlberg,
1984).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Abstract (if available)
Linked assets
University of Southern California Dissertations and Theses
Conceptually similar
PDF
The relationship between human values and moral reasoning as components of moral behavior
PDF
The relationship between high school anti-smoking policies and adolescent cigarette use
PDF
Therapists' perceptions of marital counseling for older couples and the relation of perceptions to selected therapist demographics
PDF
The relationship between identity style and self-regulated learning
PDF
The effects of family environment variables on young children's school behavior and peer interactions
PDF
Developmental task achievement among student leaders at a small private liberal arts college
PDF
Maternal risk factors and child behavior problems
PDF
The relationship between disordered eating behaviors and coping
PDF
The impact of allied health education on moral reasoning and attitudes toward classroom and clinical misconduct
PDF
Preadmission predictor variables and their relation to academic and clinical performance in the University of Southern California primary care physician assistant program
PDF
Intrinsic and extrinsic factors which influence academic, especially science, major choice in high school students participating in the National Educational Longitudinal Study of 1988 (NELS '88)
PDF
The significance of naming speed in reading development
PDF
Cognitive and non-cognitive factors as predictors of retention among academically at-risk college students: A structural equation modelling approach
PDF
Frequency of trainees' nondisclosures to their supervisors as affected by relationship quality and trainee reactance
PDF
The relationship between success in medical school and both selected academic and nonacademic prediction factors
PDF
Distant Cousins: The Relationship Between Jews From The Former-Soviet Union and American Jewish Voluntary Associations In Los Angeles.
PDF
The effects of state and trait worry, self-efficacy, and effort on statistics achievement of Malay and Chinese undergraduates in Malaysia: A causal modeling approach
PDF
Maternal health and psychopathology: Impact on child behavior and development.
PDF
Maternal and infant factors related to the security of the infant-mother attachment relationship when considering employed and stay-at-home mothers
PDF
The effect of parental education, prior achievement, self-efficacy, goal orientation, and effort on undergraduate science performance of Latinos and Caucasians
Asset Metadata
Creator
Weaver Frost, Lynn
(author)
Core Title
The relationship between human values and moral reasoning as components of moral behavior
Degree
Doctor of Philosophy
Degree Program
Psychology
Publisher
University of Southern California
(original),
University of Southern California. Libraries
(digital)
Tag
education, educational psychology,Education, higher,OAI-PMH Harvest,philosophy,psychology, developmental
Language
English
Contributor
Digitized by ProQuest
(provenance)
Permanent Link (DOI)
https://doi.org/10.25549/usctheses-c17-284448
Unique identifier
UC11349977
Identifier
9734862.pdf (filename),usctheses-c17-284448 (legacy record id)
Legacy Identifier
9734862.pdf
Dmrecord
284448
Document Type
Dissertation
Rights
Weaver Frost, Lynn
Type
texts
Source
University of Southern California
(contributing entity),
University of Southern California Dissertations and Theses
(collection)
Access Conditions
The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the au...
Repository Name
University of Southern California Digital Library
Repository Location
USC Digital Library, University of Southern California, University Park Campus, Los Angeles, California 90089, USA
Tags
education, educational psychology
psychology, developmental