Title Page
 Table of Contents
 List of Tables
 The threat index
 The Collett-Lester fear of death...
 The personal data questionnair...
 Participant informed consent...
 Obituary/biography experience
 Farewell letter experience
 Stress management experience
 Biographical sketch

Title: Reduction in death threat as a basis for optimal functioning
Full Citation
Permanent Link: http://ufdc.ufl.edu/UF00097436/00001
 Material Information
Title: Reduction in death threat as a basis for optimal functioning the test of a central existential hypothesis
Physical Description: viii, 135 leaves : ; 28 cm.
Language: English
Creator: Rigdon, Michael A., 1944-
Publication Date: 1982
Copyright Date: 1982
Subject: Death -- Psychological aspects   ( lcsh )
Psychology thesis Ph. D   ( lcsh )
Dissertations, Academic -- Psychology -- UF   ( lcsh )
Genre: bibliography   ( marcgt )
non-fiction   ( marcgt )
Thesis: Thesis (Ph. D.)--University of Florida, 1982.
Bibliography: Bibliography: leaves 124-134.
Additional Physical Form: Also available on World Wide Web
General Note: Typescript.
General Note: Vita.
Statement of Responsibility: by Michael A. Rigdon.
 Record Information
Bibliographic ID: UF00097436
Volume ID: VID00001
Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
Resource Identifier: alephbibnum - 000334621
oclc - 09460253
notis - ABW4264


This item has the following downloads:

PDF ( 5 MBs ) ( PDF )

Table of Contents
    Title Page
        Page i
        Page ii
    Table of Contents
        Page iii
        Page iv
    List of Tables
        Page v
        Page vi
        Page vii
        Page viii
        Page 1
        Page 2
        Page 3
        Page 4
        Page 5
        Page 6
        Page 7
        Page 8
        Page 9
        Page 10
        Page 11
        Page 12
        Page 13
        Page 14
        Page 15
        Page 16
        Page 17
        Page 18
        Page 19
        Page 20
        Page 21
        Page 22
        Page 23
        Page 24
        Page 25
        Page 26
        Page 27
        Page 28
        Page 29
        Page 30
        Page 31
        Page 32
        Page 33
        Page 34
        Page 35
        Page 36
        Page 37
        Page 38
        Page 39
        Page 40
        Page 41
        Page 42
        Page 43
        Page 44
        Page 45
        Page 46
        Page 47
        Page 48
        Page 49
        Page 50
        Page 51
        Page 52
        Page 53
        Page 54
        Page 55
        Page 56
        Page 57
        Page 58
        Page 59
        Page 60
        Page 61
        Page 62
        Page 63
        Page 64
        Page 65
        Page 66
        Page 67
        Page 68
        Page 69
        Page 70
        Page 71
        Page 72
        Page 73
        Page 74
        Page 75
        Page 76
        Page 77
        Page 78
        Page 79
        Page 80
        Page 81
        Page 82
        Page 83
        Page 84
        Page 85
        Page 86
        Page 87
        Page 88
        Page 89
        Page 90
        Page 91
        Page 92
        Page 93
        Page 94
        Page 95
        Page 96
        Page 97
        Page 98
        Page 99
        Page 100
        Page 101
        Page 102
        Page 103
        Page 104
        Page 105
        Page 106
        Page 107
        Page 108
        Page 109
        Page 110
        Page 111
        Page 112
    The threat index
        Page 113
        Page 114
    The Collett-Lester fear of death scale
        Page 115
        Page 116
    The personal data questionnaire
        Page 117
        Page 118
    Participant informed consent form
        Page 119
    Obituary/biography experience
        Page 120
        Page 121
    Farewell letter experience
        Page 122
    Stress management experience
        Page 123
        Page 124
        Page 125
        Page 126
        Page 127
        Page 128
        Page 129
        Page 130
        Page 131
        Page 132
        Page 133
        Page 134
    Biographical sketch
        Page 135
        Page 136
        Page 137
        Page 138
Full Text








From the beginning of this project to its completion,

many people supported me. I could not have done it without

their help.

I am grateful to Franz Epting for being my mentor during

the past six years. I am grateful to Bill Froming, Scott

Miller, Dave Suchman, and Hannelore Wass, each of whom made

a unique contribution to my development as a psychological

researcher/practitioner. I am grateful to Wayne Senfeld,

Lisa Spear, and Mart Thurmond, who persistently and enthusi-

astically helped collect and code nearly one thousand data

points for each participant. I am graceful to Imogene and

Mary Lisa, my wife and daughter, who gave me the time and

energy to continue and who withstood my disagreeable moods.






Death Orientation . . . . . . .

Optimal Functioning . . . . . . .
Maslow's Model of the Self-Actualizing Person
Landsman's Beautiful and Noble Person . .
Frankl's Model of a Meaningful Life . . .
Coan's Model of the Optimal Personality . .
Kelly's Model of Optimal Functioning. . .

Empirical Studies on the Correlation of Death
Attitudes and Psychological Health. . . .
Tendencies toward Neurosis or Depression. .
General Anxiety and Other Indicators of
Maladjustment . . . . . . .

An Existential Hypothesis. . . . . .

A Personal Construct View. . . . . .

Rationale and Hypotheses . . . . . .
The Complexity of Death Orientation and
Optimal Functioning. . . . . . .
Two Experimental Hypotheses . . . .

CHAPTER II: METHOD . . . . . . . .

Subjects . . . . . . . . . .

Instruments . . . . . . . . .
Threat Index (TI) . . . . . . .
The Collett-Lester Fear of Death Scale (FDS).
The Purpose in Life Test (PIL) . . .
The Personal Data Questionnaire (PDQ) . .
Procedure . . . . . . . . .
First Session . . . . . . . .
Second Session . . . . . . .



. . . . . . ii

. . . . . . v
. . . . . .

. '. 34

. . 37

. . 41

. . 42
. . 45

. . 49

. . 49

. . 49
. . 49

CHAPTER III: RESULTS . . . . . . . .

Instrument Reliability . . . . . . .

First Hypothesis . . . . . . . .
Canonical Correlation Analyses . . . .
Multivariate Analyses of Variance . . .
Multiple Regression Analyses . . . .

Second Hypothesis . . . . . . .

Third Hypothesis . . . . . . . .

Additional Analyses . . . . . . .
Religious Beliefs and Practices . . . .
Previous Experience with Death and Dying. .
Alternative Scores for the Threat Index . .


Death Orientation and Optimal Functioning
Relationship . . . . . . . .

Effectiveness of the Death-related Experiences .

Strengths and Limitations of the Present Study

Implications for Psychotherapy and for
Future Research . . . . . . . .








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

Bi.OGRAPHICAL SKETCH . . . . . . . .


. . 59

. . 59

. . 62
. .. 63
* . 69
* . 71

. . 75

. . 84

. . 87
. . 89
. . 91
* . 93

* . 97

. . 97

. . 101

. . 103

. . 108

. . 113

S. 115

. . 117

. . 119

* . 120

* . 122

* . 123

* . 124

* . 135


Table Page

I Intercorrelations of Pretest Scores on
Four FDS Subscales for All Participants. 61

2 Post Hoc Tests of Differences between
Pretest and Posttest Mean Scores on
Differentiation and Integration Indices . . 61

3 Pretest Intercorrelations between Indices
of Differentiation and Integration . . . 64

4 Canonical Correlation Analyses of Pretest
Death Orientation and Optimal Functioning
Scores for All Participants . . . . . 66

5 Manova Summary Table for Pretest Death
Orientation and Optimal Functioning Scores .67

6 Analysis of Variance of Death Orientation
and Optimal Functioning Scores for Sex
Differences . . . . . . . . . 68

7 Canonical Correlation Analyses of Pretest
Death Orientation and Ootimal Functioning
Scores for Females and for Males . . . . 70

8 Manova Summary Table for Effects of High
Versus Low FIC and Chi-square Scores on
Death Orientation Scores . . . . . . 72

9 Manova Summary Table for Effects of High
Versus Low FTC and Ordination Scores on
Death Orientation Scores . . . . . . 73

10 Multiple Regressions of Death Threat Scores
on Selected Opt-im~al Functioning Variables. ..74

11 Anovas for Treatment Differences in Pretest
Death Orientation and Optimal Functioning
Scores . . . . . . . . . . 77

Table Page

12 Mean Scores for Death Threat, Ordination,
and Chi-square by Treatment Conditions
for Pretest and Posttest . . . . . .. .78

13 Manova Summary Table for Posttest Death
Orientation Scores . . . . . . ... 79

14 Anovas for Sex Differences in Posttest
Death Orientation Scores . .... . .. .81

15 Anovas for Treatment Differences in
Posttest Death Orientation Scores. . . .. .82

16 Multivariate Analysis of Covariance of
Posttest Death Orientation Scores with Pre-
test Death Threat Scores as Covariate. . . 83

17 Nanova Summary Table for Posttest Optimal
Functioning Scores . . . . . . ... 85

18 Anovas for Treatment Effects on Posttest
Optimal Functioning Scores . . . . . 86

19 Multivariate Analysis of Covariance of
Posttest Optimal Functioning Scores with
Pretest Ordination and Chi-square Scores
as Covariates. . . . . . . . .. .88

20 Canonical Correlation Analysis between
Pretest Death Orientation Scores and
Religious Beliefs/Practices. . . . . .90

21 Multiple Regression of Pretest Death Threat
Scores on Religious Beliefs/Practices
Variables . . . . . . . . . 92

22 Canonical Correlation Analyses for Pretest
Death Orientation and Optimal Functioning
Scores with WTI Scores . . . . . .. 95

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



Michael A. Rigdon

August 1982

Chairman: Franz R. Epting
Major Department: Psychology

Based on a review of previous research, a significant

relationship was predicted between death orientation and opti-

mal functioning, both conceptualized in multidimensional terms.

Hypotheses are drawn from existential philosophy and from per-

sonal construct psychology to suggest that resolving the issue

of personal mortality enables a person to live a more intense,

meaningful life. In operational terms, optimal functioning

depends on one's positive orientation toward death. Selected

components of death education courses, therefore, were predicted

to result in more positive death orientation and more optimal


Participants in the study were 96 undergraduate psychology

students. In the first experimental session, they completed

four questionnaires and participated in the experimental con-

dition to which each group was randomly assigned: completing


an obituary/biography, writing a farewell letter, or, for the

control group, participating in a stress management experience,

The questionnaires included a personal data questionnaire, a

slightly modified form of the Threat Index (TI), the Purpose ii

Life Test (PIL), and the Collett-Lester Fear of Death Scale (FI

The TI was modified for scoring two differentiation and two

integration indices. Using 13-point scales, participants rate

self, my own death, and 10 death-related elements on 30 dimen-

sions related to death. For the second session, 95 participan'

returned to complete again the final three questionnaires list<


The results indicated a significant relationship between

positive death orientation and the optimal functioning variable

with some differences in the pattern of this relationship for

males and for females. Multivariate analyses of variance fail

to support the impact of the death-related experiences on deat:

orientation or optimal functioning, perhaps because the select,

experiences were not strong enough to affect either complex,

stable attitudes toward death or a person's overall optimal

functioning level.

Results of post hoc analyses revealed significantly more

negati'/e death orientation for females, as well as the signifi

cant association of strong religious beliefs and practices wit

positive death orientation. Discussion focused on strengths

and limitations of the investigation and presented implications

for psychotherapy and for research in the areas of death orien

station and optimal functioning.



We Americans are an optimistic sort. We have long

believed that any problem can be solved with a lot of hard

work, the right technology, and a bit of luck. Our technology

seems helpless in the face of death, however, which may be

one reason our society has tended to deny the reality of

death (Weisman, 1972). American psychologists similarly

have focused primarily on positive experiences in their in-

vestigations of optimal functioning--the behavior and exper-

ience of those who are prime examples of the actualization

of human potential. Maslow (1968), for instance, acknowl-

edged that he "made no effort to explore . what may be

called the 'nadir experience,' e.g., the (to some) painful

and crushing insights into the inevitability of aging and

death ." (p. 84).

Lester (1971a), on the other hand, presented evidence

that suggested the death denial of earlier times has given

way to more frequent thoughts of death. In 1970, students

completed the identical death attitude questionnaire com-

pleted by students at the same college in 1935. Compared

to the 1935 students, contemporary students reported being

more preoccupied with and concerned about death. This wide-

spread concern about the imminence of death may be responsible

for the movement presently spreading in Europe and the United

States to freeze the number of nuclear weapons. Scholars

also began to explore psychological and sociological aspects

of the phenomena of death and dying, as is evident from

Pearson's (1969) 100-page annotated bibliography, Santora's

(1980, 1981) 17-page bibliography of dissertations centered

on death and dying topics in the 1970's, and two volumes of

annotated death education references by Wass and her colleague

(Wass, Corr, Pacholski, & Sanders, 1980; Wass, Corr, & Pachol:

with Forfar, in press). Although the issue of the relations

between death attitudes and optimal functioning has been in-

vestigated occasionally, the study presented here was design

to build a comprehensive framework for understanding death

attitudes, optimal functioning, and their interrelationship

in order to provide some answer to Murphy's (1959) questions

"Under what conditions can the facing of death be constructed

in mental health terms?" (p. 319).

The aim of this introduction is first to elaborate the

conceptualization and operationalization of death orientation

second to review five models of optimal functioning, third to

review the currently available data regarding the relationship

between death attitudes and aspects of healthy functioning,

fourth to present an existential approach to the relationship

between healthy functioning and one's attitude about death,

fifth to present a personal construct view of the same issue,

and finally to state the rationale for the present study and

the hypotheses to be investigated.

orientation, so much so that two successful journals are now

in existence: Death Education and Omega. In studies reported

in these journals, death orientation has been conceptualized

in a variety of ways and assessed by means of a variety of

instruments. This section will describe these conceptualiza-

tions and instruments used in research on death orientation.

Research on death orientation is concerned generally with

a person's beliefs, feelings, attitudes, and behaviors regard-

ing the concept of personal death or the death of others.

Investigators initially focused on several negative aspects

of death orientation which were loosely termed death anxiety,

fear of death, or concern about death. With the exception

of an occasional use of physiological measures, such as

galvanic skin response and projective instruments such as

the Thematic Apperception Test to assess the level of death

anxiety or fear (Kurlychek, 1978a), research in this area

involved the development of self-report measures, such as

Templer's (1970) Death Anxiety Scale (DAS), Boyar's (1964)

Fear of DeatLh Scale, Lester's (1967a) Fear of Death Scale

(FDS), and Dickstein's (1972) Death Concern Scale (DCS).

More recently, Ray and Najman (1974) developed a self-report

measure designated to assess death acceptance, conceptualiz-

ing death orientation in positive rather than negative terms.

The five self-report measures just described represent

attempts to operationalize death orientation in unidimensiona

terms. For some time, however, thanatologists have recognize

that death orientation is a complex phenomenon, requiring

measures able to tap several aspects of a person's approach

to death. The complexity and multidimensionality of death

orientation have been explored in terms of one's orientation

toward death at several levels of awareness and in terms of

one's orientation toward several aspects of death itself.

Let us examine these two ways of viewing the multidimension-

ality of death orientation.

Feifel and his colleagues (Feifel & Branscomb, 1973;

Feifel,Freilich, & Hermann, 1973; Feifel & Hermann, 1973;

Feifel & Jones, 1968) were the first to focus attention on

the levels of awareness in relation to death orientation.

They developed a set of measures designed to assess fear of

death not only at the conscious level of awareness, but also

at a level below awareness (unconscious level) and at a mid-

level of awareness (fantasy level). Analysis of the results

of this set of measures with a variety of populations has

convinced Feifel of a consistent pattern of conscious denial

of death fear and high fear of death at the below-awareness

level, with frequently ambiguous responses at the fantasy

level. This pattern would seem to support Feifel's insistence

on assessing death fear at several levels. A person might

deny any awareness of being afraid of death, but a compara-

tively high level of unconscious death fear might be essentic

to an accurate assessment of the same person's death orien-

tation. Furthermore, the death orientation of two individ-

uals may be deceptively similar if one simply examines their

relatively high unconscious fear of death. Their death

orientation may differ in complex ways, however, if one of

the two individuals is consciously aware of high death fear

and is, perhaps, in the process of coming to terms with this

fear, while the second individual denies any awareness of

death fear as a problematic issue in life. The ability of

Feifel's set of measures accurately to assess conscious,

mid-level, and unconscious fear of death would open up such

questions for systematic study. The little evidence avail-

able to date suggests that Feifel's measures do not, in

fact, assess three levels of death fear. A factor analysis

(Rigdon, in press) of the scores of 60 high school and

college students on Feifel's measures indicated that the

conscious and fantasy measures tap what might be called

most accurately "conscious fear of death," while scores on

the other five measures split into two or three factors.

These five measures do not seem to tap a unitary unconscious

fear of death. And, until research indicates the correlates

of the various factors, it is unclear precisely what aspects

of death orientation are being assessed by these seemingly

heterogeneous measures. Feifel was a pioneer in devising

a method for assessing the multidimensional meaning of

death, but his work has not yet resulted in a satisfactory

set of death orientation measures.

Cardner Murphy (1959) suggested an alternative line of

assessing the complexity of death orientation in relation to

a variety of aspects of death. He noted that a person could

develop attitudes toward at least seven aspects of death.

Death could be feared not only as the natural end and as the

loss of consciousness and mastery, but also because of the

associated fears of loneliness, of personal failure, and of

what might happen to one's dependents. Vernon (1970) and

Kastenbaum and Aisenberg (1972) similarly emphasized the

importance of examining one's response to a variety of fea-

tures of death. In addition to fear associated with aspects

of the death and dying of others, they explored the basic

fear of personal death as extinction, as the end of personal

experience, and as separation from loved ones, as well as

the fear of pain and indignity during the process of dying,

and the fear of the possible consequences of death such as

personal punishment and what might happen to the people and

projects one is responsible for.

Researchers have used three approaches to tackle the

issue of the multidimensionality of death orientation. The

first approach to be examined is the determination of rele-

vant aspects of death by experts during the process of test

construction. At least two attempts have been made to

develop measures that include subscales for more than one

aspect of the concept and experience of death. Collett and

Lescer (1969) developed a reliable Fear of Death Scale (FDS)

with four subscales to assess fear of one's own death or


dying and fear of the death or dying of others. Hooper and

Spilka (1970) constructed nine 20-item scales for assessing

the seven aspects of death fear indicated by Murphy, as well

as the positive attitudes of death as an afterlife of reward

and death as an expression of courage. The researchers who

developed these more complex measures of death orientation

agreed with Murphy, Vernon, and Kastenbaum and Aisenberg that

one simple score could not estimate adequately a person's

beliefs, feelings, and behaviors with respect to such a

complex event as death.

Rather than developing a questionnaire to assess prede-

termined aspects of death orientation, several investigators

used a factor analytic approach to determine empirically what

aspects of death orientation emerge from a set of self-report

items. This represents a second approach to the multidimen-

sionality of death orientation. A factor analysis of

Dickstein's DCS (Klug & Boss, 1976), for instance, revealed

not the expected single death concern factor, but two factors

for the conscious contemplation of death and the negative

evaluation of death. Using factor analysis, Nelson and Nelson

(1975) developed scales they identified as death fear, death

avoidance, death denial, and reluctance to interact with the

dying. Victor (1981) constructed a questionnaire with factors

for existential death concern and helplessness of dying concern.

Durlak and Kass (1981) used principal components analysis to

determine what aspects of death orientation actually are

frequently used in death orientation research. Their analy-

sis determined the overlap between eight existing scales.

Four of the instruments had single scores: the Threat Index

(Rigdon, Epting, i.eimeyer, & Krieger, 1979), Templer's DAS,

Ray and Najman's Death Acceptance Scale, and Lester's FDS.

Three of the instruments had multiple scores, as previously

described: the Collett-Lester FDS, Nelson and Nelson's death

anxiety measure, and Dickstein's DCS. The final scales in-

volved the rating of the concepts of death and being in pain

on semantic differential dimensions.

Two of the five factors in Durlak and Kass' analysis

could be described as a general negative response to personal

death and preoccupation with thoughts about death. A re-

analysis of Durlak and Kass' data (Rigdon & Epting, 1981)

indicated that two of the other factors identified as reluc-

tance to interact with the dying and reaction to the reminder

of death may be summarized better as a third factor of avoid-

ance of the dying and of death reminders, while Durlak and

Kass' fifth factor, negative reaction to pain, should be

ignored as trivial. Although the analysis by Durlak and

Kass and that by Rigdon and Epting produced results that

differed in some important respects, both analyses indicated

the importance of selecting more than one measure to assess

various aspects of a person's attitudes and concepts about


The Threat Index (TI) (Krieger, Epting, & Leitner, 1974)

is an instrument that represents a third way of assessing

death orientation with respect to several aspects of death.

Unlike the Collett-Lester FDS, however, the relevant aspects

of death are not predetermined when the test is constructed.

Unlike Durlak and Kass' approach to the multidimensionality

of death orientation, the relevant components of death atti-

tudes are not determined by a process of comparing responses

in a large sample of people. The TI, derived from personal

construct theory (Kelly, 1955), begins with a person gener-

ating a set of dichotomous meaning dimensions (personal

constructs) that are important to the person's view of death

(e.g., violent-peaceful or meaningful-useless). The person

next places the elements of self and my own death on each

construct. The more dimensions on which the self and death

are "split" (i.e., placed on opposite poles), the more the

person is taken to be threatened by death, in that more

change would be necessary for self and death to be compatible.

Thus, the TI assesses death orientation with respect to those

features of death that are most salient for each person.

Yet the TI provides in the number of splits a single score

which may be taken as a measure of one's level of threat or,

judging by the TI's high loading on Durlak and Kass' first

component, one's overall negative response to death.

The versatility of the TI is apparent in that one can

use the number of splits as a simple index of threat and, at

the same time, capitalize on the idiosyncratic multidimension-

ality of the TI by analyzing into principal components a

person's ratings of self, death, and other death-related

events on the death-related personal constructs. Warren

(1982) has indicated the usefulness of the principal com-

ponents analysis in conjunction with the TI. The split

score can provide an estimate of the individual's level of

threat, while the principal components analysis can provide

an indication of the possible directions in which the person

might move in an attempt to reduce the threat of death and

make self and death more compatible.

In conclusion, the three approaches to the multidimen-

sional conceptualization and assessment of death orientation

indicate the usefulness of assessing predetermined aspects

of death attitudes (e.g., fear of one's own death and fear

of the death of others) and the usefulness of having several

measures to tap the relatively independent components of

death as determined by Durlak and Kass' analysis. The flex-

ibility of the TI suggests that one might, at the same time,

examine the idiosyncratic multidimensionality of an individ-

ual's stance with respect to death. A combination of the

three approaches to multidimensionality would seem to be

most advantageous for capturing death orientation with

some accuracy.

Optimal Functioning

Upon meeting a psychologist, people often remark, "I

had better be careful what I say." As if the psychologist

would catch them in a "Freudian slip" or some indication of

theoretical understanding of mental illness. Within'this

framework, mental health would be simply the opposite of

mental illness and would he thought of as the absence of any

noticeable indications of psychological disorders. Jahoda

(1958) extended the concept of mental health to include var-

ious positive features of behavior and experience, such as

self-expressiveness, open-mindedness, the ability to deal

effectively with daily stress, realistic self-awareness

about one's strengths and weaknesses, and genuine concern

for others and the ability to establish meaningful relation-


Other investigators developed alternative models to

circumvent the illness-health dichotomy and to present healthy

functioning in a positive framework on its own terms. Just

as there would seem to be important physiological differences

between a normally healthy (i.e., "unsick") person and a

champion athlete, so there may be important differences be-

twenen a mentally healthy person who has escaped serious psycho-

logical disorders and a person who is an exceptional exa-mple

of the possibilities of human growth and development. Jung

(19163) focused an a life-long process he termed individuation.

Allpcrt (1961) develoIped the concepts of prjopriate functioning

and mature personality. Jourard (1968) emphasized self-

dis-closure. Rogers (1961) developed his notion of the fully

functioning person. These approaches share a focus on opti-

mal functioning, a term which can be used to describe the

level of growth attained by exceptional people or the exper-

ience and performance level of most people at those excep-

tional moments when we are "at our best." Investigators in

this area of psychology are interested in what a person can

become when human potentials are realized. The remainder

of this section will examine briefly five models of optimal

functioning and several methods for assessing this phenomenon

Maslow's Model of the Self-Actualizing Person

Maslow (1968), in his study of exemplars of what he

termed full humanness, noticed that, in the process of growth

certain basic needs were fulfilled first to maintain minimal

functioning. These are physiological needs, as well as

security needs and a person's need to be loved and respected.

Maslow called these deficiency needs because a person pursues

objectives that will fill the void which comprises each need.

A second set of needs relates to a person's full development.

These are aesthetic needs and the need to self-actualize--to

realize one's potential more completely. Unlike the deficien

needs, a person engages in the self-expressiveness of self-

actualization or in the pursuit of truth and beauty for the

sheer joy of the pursuit. The experience itself is the rewar

not the reduction of some deficit. Maslow referred to this a

meta-motivation, and he used the term peak experiences for

those joyful moments in which one surpasses deficit motivation

to actualize self or to attain truth and beauty. Among other

characteristics, Maslow described the exceptional people he

studied as having 1. clear, efficient perceptions of re-

ality, 2. openness to experience, 3. spontaneity and self-

expressiveness, 4. creativity, 5. self-acceptance, and 6. a

sense of autonomy and independence.

In order to assess the level to which people have de-

veloped the characteristics presented by Hlaslow and other

investigators of optimal functioning, Shostrom (1962) de-

veloped the Personality Orientation Inventory (POI). This

150-item self-report questionnaire has two major scales for

time competence and inner support, as well as 10 overlapping

minor scales for characteristics such as spontaneity, self-

acceptance, and capacity for intimacy. Some of the hundreds

of POI studies have provided moderate validity support in

its ability to discriminate clinical populations from more

normal, mentally healthy groups and in the predicted corre-

lations of some POI subscales with other characteristics

related to self-actualization. On the other hand, the POI

has been criticized on several grounds (Amerikaner, 1978;

Oakland, Freed, Lovekin, Davis, & Camilleri, 1978). First

of all, the test construction has been questioned because

some items are not clearly based on Maslow's theory or any

other. Other items are vaguely worded. Many items are

scored on one of the major scales and two of the minor scales.

Secondly, test-retest reliability coefficients for the sub-

scales are lower than one might wish, with a median of .59

in one study (Ilardi & May, 1968). Third, the POI profiles

of a large number of people fall within the self-actualized

range, although Maslow believed less than 1 percent of the

population could be called self-actualized. Finally, there

is evidence (Braun & LaFaro, 1969) that students can arti-

ficially raise their POI scores if they are taught self-

actualization concepts and terminology. Their ability to

give the impression of being self-actualized raises the

question of whether psychotherapy may sometimes result in

higher POI scores simply by giving clients information about

self-actualization rather than by producing it. The POI is

an interesting first attempt to operationalize the complex

notion of self-actualization, but some basic re-construction

needs to correct the problems just outlined to produce a

psychometrically adequate instrument.

Landsman's Beautiful and Noble Person

Rather than studying exceptional individuals, Landsman

(1974; Jourard & Landsnan, 1980) studied exceptional experi-

ences to develop his model of optimal functioning. His in-

vestigation of positive personal and interpersonal experience

remembered by a variety of people and of tragic experiences

which were later viewed as positive led him to propose three

general stages in the development of the basically normal

person into a beautiful and noble person. These stages are

1. the self-loving, self-expressive passionate self, 2. the

environment-loving self who appreciates the beauty of the

psionate behavior develops, especially in children.

Fakl's Model of a Meaningful Life

Following his concentration camp experiences during

Wrd War II, Frankl (1963) presented a model of optimal

fntioning which emphasized a person's ability to create a

maingful, purposeful life. Experiences such as the success-

flachievement of a goal, the sense of being one with nature

some other valued object, and interpersonal intimacy pro-

moethe discovery of meaning and purpose. But Frankl focused

osuffering as a special opportunity for the person to create

maing. Even under extremely desperate circumstances, each

pron remains free to determine the attitude with which to

apoach the situation. According to Frankl, the discovery

this freedom forms the basis of a meaningful life.

Crumbaugh and Maholick (19G9) designed a 20-item, Purpose

in ife Test (PIL) to assess the extent to which a person has

effctively created a meaningful, purposeful life. In-itial

reut:; supported the reliability of the PIL and indicated,

example, that it does distinguish between depressed indi-

v tdals and those who report a sense of well-being. High PIL

soes are also associated with positive self-concept and

sefesteem (Rekler, 1977). Several studies have investigated

the relaLionship between the PIL and variables with which

one could predict an association in light of Frankl's theory.

Sharpe and Viney (1973) analyzed college students' written

descriptions of their world view, finding that the world

views of the students with low PIL scores were judged nega-

tive, lacking in purpose, and lacking in transcendent goals.

Students with positive, purposeful world views and transcen-

ent goals had high PIL scores. Similarly, persons with high

Purpose in Life scores endorsed values such as responsibility

and salvation, but de-valued pleasure (Crandall & Rasmussen,

1975; Simmons, 1980). They seemed to have learned that, as

Frankly maintained, happiness or pleasure are achieved in-

directly by striving for personal values which directly make

life meaningful. The results of these studies support the

contention that the PIL is a good measure of meaning and

purpose in life, as conceptualized by Franki, although a

significant relationship with social desirability (r = .59)

indicates that PIL scores may be influenced by the desire to

give a good impression. A group of subjects did not, however,

raise their PIL scores after being told their first scores

were unacceptably low (Snavely, cited in Crumbaugh & Maholick,

1969). Furthermore, Durlak (1972, 1978) has failed to find

a significant correlation between PIL scores and social de-

sirability in three samples (r = .00, .01, -.10, respectively).

college students' responses to a variety of questionnaires

designed to assess characteristics of the optimal personality.

Rather than indicating one general trait of optimal function-

ing, Coan's analysis resulted in 19 factors, which he encap-

sulated in five basic features of efficiency, creativity

and openness to new experiences, harmony with self, the

ability to relate to others, and the ability to transcend

self in unity with a larger whole. Amerikaner (1978) criti-

cized the lack of validity support for Coan's questionnaires.

He also observed that a general optimal functioning factor

might be found in a sample of optimally functioning individ-

uals. Coan's failure to find a general trait may reflect on

the college sample rather than on the nature of optimal func-

tioning. On the other hand, his approach could represent

eventually the most comprehensive, empirical model for the

investigation and understanding of optimal functioning.

Kelly's Model of Optimal Functioning

Like the other investigators briefly reviewed, Kelly

(1955, 1980) developed his own unique approach to optimal

functioning, an approach elaborated by Epting and Amerikaner

(1980; see also Epting & Suchman, in press). Kelly was con-

cerned with the process according to which people change

their personally constructed meaning units (personal construts

a;dic Lhus their overall construct system in order more effec-

tively to interpret a complex, ever-changing world. Accord-

ing to Kelly, growthful change occurs through the five-stage

experience cycle. Most of us, unfortunately, tend to get

stuck at one stage or another, or we try to skip one of them.

Optimal functioning is characteristic of people who generally

complete the experience cycle. And most of us function opti-

mally at those exceptional moments when our experience cycle

is complete. The first step is active anticipation. When

an acquaintance asks to borrow your car, do you anticipate

that she is trustworthy and dependable or not? The second

stage is involvement: If you expect your acquaintance is

dependable, what are the implications of this expectation?

Will she also drive carefully? Will she clean the car before

she returns it? How important is it that you view her as

dependable? In the next step, encounter, you act as if your

acquaintance is dependable: You loan her your car and are

vulnerable and open to the results. The results of the

encounter will either confirm and validate your anticipation

or not (step 4). The invalidation of having your car returned

with a large dent in the left fender would require some activ-

it' at stage 5, constructive revision, where you might re-

construe your acquaintance as being accident-prone or a

problem drinker. Anticipation is essential to this entire

process, but growth cannot occur without taking some action

in terms of what one anticipates.

respectively. The creativity cycle refers to the process by

which a person alternately loosens and tightens the boundaries

of constructs in order to invent new construct dimensions to

interpret events. The activity cycle (or C-P-C cycle) is

similar to the experience cycle, but reflects the special

experience of making decisions. A full activity cycle has

three stages: 1. a circumspect consideration of a variety

of constructs which might describe the issues in this situa-

tion, 2. a preemptive selection of a single issue as most

salient and crucial (e.g., exercise is good for my health;

sitting to watch another football game is not), and 3. the

choice of one of the alternatives to carry into action. No

method has been developed, unfortunately, to assess directly

and quantitatively the extent to which the three cycles just

described are fully and efficiently operating in a person's


An alter-native approach to optimal. functioning within

the personal construct framework relates to the extent to

which the person is engaged in the process of developing a

more complex construct system for effectively anticipating

life events- Follow~ing Werner's (1957) orthogenetic prin-

ciple of cognitive development, personal construct theorists

(Adams-Webber, 1979; Crockett, 1965; Crum, 1978) have viewed

the development of complexity as a two-fold process of

differentiation and hierarchical integration. Differentiation

is the process of creating new construct dimensions with

which to interpret one's experience. This implies the pro-

gressive elaboration of construct subsystems with which to

make sense of a variety of events. Interpersonal experience,

career events, and the realm of death might each have an asso-

ciated construct subsystem. Integration, on the other hand,

refers to the process of drawing out the implications between

constructs so that the various constructs and subsystems fit

together in a unified framework.

There is a relationship between the two-fold process of

complexity and the creativity and activity cycles described

above. The level of differentiation would be an index of

the efficiency of the creativity cycle for generating con-

structs in response to new experiences, while the level of

integration would indirectly indicate the efficiency with

which a person could sort through a variety of constructs to

preemptively select one that best represents the issue on

which to make a choice for action. This preemptive selection

would be more efficient in an integrated network with its

clear lines of implication.

Fransella and Bannister (1977) and Adams-Webber (1979)

described aid evaluated about 20 indices to use reptest ratings

of elements in terms of a set of personal constructs for as-

sessing complexity or one of its aspects. The results provide

some support for differentiation and integration indices.

Smith and Leach (1972), for instance, found a relationship


between scores on their hierarchical organization measure

and abstractness as assessed by Harvey's (1966) This I Be-

lieve Test, designed to assess cognitive development from

a child's focus on the concrete to the mature adult's ability

to deal with the world in a more abstract way. More inte-

grated subjects were also higher in abstractness. Using

Adams-Webber's (1969) differentiation index, Olson and

Partington (1977) found that highly differentiated subjects,

as a result of having a variety of their own personal per-

spectives, were more able to take the perspective of others

when re-telling a story from the viewpoint of different char-

acters. Relative to optimal functioning, Thomas and Seeman

(1-971) reported that individuals identified by other group

members as matching Jahoda's (1958) characteristics of mental

health were more differentiated in their interpersonal con-

structs, in that they generated a greater number of constructs

to describe people. Wexler (1974) used the POI to identify

optimally functioning members in a group, whose descriptions

of emotional experience were judged more differentiated and

integrated than the descriptions of less optimally function-

ing group members. Besides providing some support for the

validity of the differentiation and integration indices, the

research to date also supports the optimal functioning of

more differentiated and integrated subjects.

Landfield (1971) and his colleagues (Landfield & Barr,

1975; Leitner, Lan~dfield, & Barr, 1976; Landfield & Schimittdiel,

differentiation, and two to operationalize integration. The

differentiation index is called FIC, for Functionally Inde-

pendent Constructions (the name is a definition of differen-

tiation), and represents the degree to which the person

applies constructs in a similar way to the elements. If

everyone who is friendly is also warm, the two constructs

are functionally equivalent. If there is little overlap, so

that knowing a woman is friendly does not help the person

decide whether or not she is also emotionally stable, for

instance, then the two constructs are functionally independent.

A person who uses constructs in a nonoverlapping, independent

manner is considered more differentiated than a person with

more overlapping constructions.

Landfield operationalized integration in two ways. Both

ordination and the chi-square statistic measure the extent to

which a person uses the same rating point to describe various

people on a construct dimension or, on the other hand, spreads

the descriptions across all the points on the scale. The

rationale for these two indices is that the process of mak-

ing fine discrimination along a construct dimension would

seem to require the use of superordinate constructs. A per-

son with a poorly integrated interpersonal system, for example,

might. tend to view Barb, Bill, and Wayne as equally friendly,

whereas a more integrated person might view Barb as friendlier

than Bill because she is also affectionate, but less friendly

than Wayne, who is not only affectionate, but also trustworthy.

The fine discrimination along the friendly-unfriendly dimension

are made possible because of the implications of the super-

ordinate constructs affectionate-not affectionate and


Crum (1978) demonstrated the usefulness of conceptual-

izing complexity in terms of the two processes of differen-

tiation and integration, especially as operationalized by

FJC and ordination. Although ordination was the best pre-

dictor of performance in his study, FIC scores clarified

the relationship between ordination and such variables as

intuitive reasoning and performance on an embedded figures

task. Wilkins (1978) found that subjects who were high in

both FIC and ordination were less likely than other subjects

to change their first impression about a person after being

presented conflicting information. As predicted, these sub-

jects were better able to see at the same time more than one

aspect of other people. Using his FIC, ordination, and chi-

square indices, Landfield (1979; Landfield & Schmittdiel,

1981) found that highly integrated group members (according

to their ordination and chi-square scores) could more easily

predict the way that other group members use their constructs.

Not only are they able to understand others, but others in

the group are also able to understand those who are more

integrated. The group had a hard time predicting the more

differentiated members, especially if they were also low in

integration. Highly differentiated, poorly integrated people

are confusing to others and hard to understand, and they do not

understand others very effectively.

While these results support the validity of FIC, ordina-

tion, and chi-square as indices of differentiation and inte-

gration, Amerikaner's (1979) investigation failed to find a

relationship between either differentiation (assessed by FIC)

or integration (assessed by ordination) and optimal function-

ing. Although optimally functioning persons have been more

differentiated and integrated in several studies described

earlier, this association has not resulted from research

using FIC and ordination. Amerikaner indicated that his

failure to find support for this relationship may be due

either to his determination of the differentiation and inte-

gration scores from provided constructs rather than the more

usual elicited constructs or to the lack of truly extreme


Maslow, Landsman, Frankl, Coan, and Kelly developed

models of optimal functioning that differ in many ways in

the conceptualization and assessment of this phenomenon.

Amerikaner (197S) suggested a comparison of optimal function-

ing models in terms of three aspects: process, structure,

and content. Let us explore how these three aspects can

help summarize models of optimal functioning, particularly

the five approaches described here.

The process aspects relate to how the person changes and

progresses toward the goal of optimal functioning. Optimal

functioning theories generally focus attention on this aspect.

Jung, for example, described the life-long process by which

the person becomes more individuated, and Rogers emphasized

Self-actualization is an end-state toward which we progress

without ever fully attaining it. Frankl's approach points

to the key process of creating personal meaning. The per-

sonal construct model describes the process by which a person

completes the cycles of experience, creativity, and activity.

The complementary processes of differentiation and integra-

tion are also important for the development of a richly

varied personality which operates as a unified whole.

The structural aspects of the optimally functioning per-

son refer less to the person's flexibility and more to the

arrangement of the personality components in a hierarchical

order, with some more important than others. M1,aslow, for

example, spoke of a need hierarchy, -with the basic deficiency

needs eventually subordinated to the meta-needs of self-

actualization and aesthetic experience. Vine compassionate-

self in Landsman's model is the highest stage in the hier-

archy of becoming the beautiful and noble person. Yelly

frequently referred to a person's core constructs--those

uuon hihone estdblishos an-d maintains self-identity.

Integration refers to the extent to which the implicat-ions

of these superordinate core constructs for the more incidental

subordinate constructs have been elaborated- The connections

between the core and incidental constructs ensure the system's

smooth, coordinated operation.

In Kelly's model, the constructs could be called the

content--the components which comprise the person's system.

The content aspect is important in several optimal function-

ing models. Maslow, Coan, and Landsman, for instance, in-

vestigated traits they considered constitutive of the opti-

mal functioning person. The POI was designed to measure such

optimal functioning qualities. And the PIL, designed to tap

into the person's process of creating a meaningful life,

directly measures the content of a person's beliefs about

the relative meaningfulness of a number of aspects of life.

A comprehensive approach to optimal functioning might profit-

ably attempt assessment in all three aspects of process,

structure, and content.

Empirical Studies on the Correlation of
Death Attitudes and Psychological Health

Proceeding from a variety of theoretical perspectives,

a considerable number of researchers have searched for a re-

lationship between attitudes toward death and a variety of

personality characteristics. Reasoning that a healthy

approach to life goes hand-in-hand with a positive, accept-

ing orientation toward death, these researchers have attempted

to associate positive death orientation (low death anxiety or

fear, for instance) with a positive life orientation or, at

least, with the absence of negative life orientation. After

reviewing the evidence regarding death orientation and both

specific and general indicators of healthy functioning, this

section will conclude with an examination of the data regard-

ing the association between death orientation and aspects

of inadequate, maladaptive behavior.

First of all, the expectation generally has been that

those who accept death will also be living a more meaningful

life, with higher self-esteem and a greater sense of compe-

tence than those who are anxious, fearful, or threatened by

death. Individuals who have a positive death orientation

have been predicted also to be lower in certain indicators

of inadequate functioning, such as general anxiety and scores

on MMPI scales, while those with a negative death orientation

were expected to be functioning less adequately, i.e., to

be more anxious and to score higher on the .MMPI scales.

Contrary to this general line of thought, Diggory (1966)

and his colleague (Diggory & Rothman, 1961) developed a ration-

ale for a rival hypothesis. They reasoned that people who

value and esteem themselves will have "more to lose" at death

and, therefore, be more fearful of death than people with low

self-esteem. Accordingly, Diggory predicted that persons who

pride themselves on their accomplishments and view life as

a time for achieving goals will have a special difficulty

facing death, which represents, after all, the end of goal-

directed behavior and which "will permanently frustrate us

by removing the possibility of any purposive activity" (p.

415). In support of their hypothesis, Diggory and Rothman

presented the results of a survey indicating that people

generally rank "the end of the opportunity to achieve goals"

high on their list of why death is a frightening event. This

result, while interesting, does not bear directly on their

hypothesis. Three studies that have more directly tested

Diggory and Rothman's hypothesis produced surprising results.

Nogas, Schweitzer, and Grumet (1974) failed to find a posi-

tive correlation between need for achievement and death

anxiety scores in their sample of 80 female college students

(r = -.13, p > .05). Their evidence did not support Diggory

and Rothman's hypothesis. Secondly, Alcorn (1977) found a

significant negative association between DAS scores and self-

esteem. People who value self are not more anxious about

death because of what they have to lose. The evidence indi-

cates that they have a more relaxed, accepting attitude than

those who have a less adequate level of self-esteem. Alcorn's

test of Diggory and Rothman's hypothesis resulted in support

for the competing notion of the positive relationship between

healthy life and death orientations. A third study (Aronow,

Rauchway, Peller, & DeVito, 1980) similarly reported signif-

icant negative correlations between death anxiety and sense

of '.'ell-being (-.28) and the similarity between self and

ideal-self ratings (-.19). Individuals who value self tend

to have lower death anxiety, contrary to Diggory and Rothman's


Further research has produced rather consistent evidence

for the association between positive death orientation and

specific aspects of healthy functioning, such as meaning in

reported consistent evidence that individuals who are living

a meaningful, purposeful life as assessed by the Purpose of

Life Test (PIL) tend to be less afraid and less anxious about

death. It should be noted, however, that these correlations

might be inflated by the presence on the PIL of two items

that relate specifically to one's orientation toward death.

Furthermore, Duke (1978) was unable to find support for a

relationship between the PIL scores of dying patients in a

private, religious-affiliated hospital and the staff's rating

of patients' acceptance of death.

People who have a good sense of their own competence are

less fearful and anxious about death, whether sense of com-

petence is assessed by a structured interview (Farley, 1971)

or by several subscales of the California. Personality Inven-

tory (Nogas (-et al., 1974). In other words, people who feel

capable of coping with their own death also tend to feel

capable of coping with life. Moses (1973) likewise found a

significant relationship between high self-esteem and low

fear of death, using Boyar's (1964) Fear of Death Scale,

a result not replicated, however, by Casciani (1979), who

used several scales toassess death orientation. Le-stex and

Collett (1970) and Neimeyef aind Chapman (1980) both found

support- for a similar notion th-at individuals who are rather

satisfied with the progress of their life project (little

discrepancy between their views of self and ideal self) are

able to face death with a more accepting, less fearful orien-

tation than are those who are less satisfied with who they

are and see the need for many changes (large discrepancy

between their views of self and ideal self).

Several studies have examined the relationship between

death orientation and a general index of healthy functioning

--Shostrom's (1962) Personality Orientation Inventory (POI)

to assess self-actualization. On the one hand, Wesch (1971)

and Wexler (1978) reported that individuals with less death

fear, anxiety, and threat had significantly higher scores

on several of the POI subscales. In Pollak's (1979) sample

of graduate students, low DAS scores were associated with

high scores on the POI scale for time competence (r = -.25).

The correlations with the other major scale and the 10 minor

subscales, however, were not significant.

In summary, no evidence has supported Diggory and Roth-

man's hypothesis that individuals with high self-esteem and

high need for achievement have more to lose by death and will

be more apprehensive about death. The data indicate, instead,

a consistent, moderate association of positive death orienta-

tion with specific features of optimal functioning, such as

meaning in life, sense of competence, and high self-esteem.

And, although positive death orientation has been related to

one or another aspect of self-actualization as assessed by

the POI, the pattern of this relationship has not been con-

sistent. The overall evidence supports a moderate, positive

relationship between positive deat orientation and various

aspects of optimal functioning.

The relationship between deat orientation and the pres-

ence or absence of certain aspect of inadequate function-

ing, including neurotic tendencies an general anxiety, has

also been explored. Let us now reiw the evidence from

these investigations. Almost 50 yas ago, Schilder (1936)

interviewed 31 imprisoned murderer to determine their

attitudes about death. Psychopati and psychotic killers

reported high fear of death and procupation with thoughts

of death as punishment, whereas kilrs with no particular

psychopathic trends were less afri of death and viewed

death as remote. Lester (1967b) adPollak (1979) reviewed

more recent studies which attempted to extend Schilder's

finding to other indicators of inaeuate functioning,

Including neurotic tendencies and hih general anxiety.

The evidence from these studies wilbe presented next.

Tendencies toward Neurosis or Deprsion

Several early studies on the reationship between death

attitudes and unhealthy, inadequate functioning involved aged

participants. According to Rhuck and Dibner (1961), high

death concerns among the aged are rated to so-called neurotic

trends (high scores on the MYPI scls for hypochondriasis,

depression, and hysteria) and to ig scores on the Cornell

Medical Index- (CMI) of physical an psychiatric symptoms.

significant relationship of death anxiety to .IL'PI depression

scores (r = .28) and to the psychiatric section of the CMI

(r = .54). Jeffers, Nichols, and Eisdorfer (1961) found

that elderly people with higher death fear more frequently

reported feelings of depression and rejection. Maladjusted

elderly people, especially those who are depressed, generally

are more fearful and anxious regarding death.

In young adults, Templer (1970) found that death anxiety

is related to high MMPI depression scores and also to the

MIPI scales for schizophrenia and psychasthenia. Like the

elderly, college students who are more anxious and fearful

concerning death also have more neurotic tendencies, with

elevated scores on the neurosis scales of the Eysenck Per-

sonality Inventory (Templer, 1972) (r = .36) and of the

Maudsley Personality Inventory (Moses, 1973). Templer and

Ruff (1971) reported that mean scores on the Death Anxiety

Scale (DAS) are generally higher for adult psychiatric patients

than for normal adults. Feifel and Hermann (1973), on the

other hand, found similar patterns in the death attitudes of

neurotic and normal subjects.

A study by Lester and Schumacher (1969) failed to find

the predicted elevation in the level of death fear of schizo-

phrenic patients compared to nonschizophrenics, perhaps be-

cause the nonschizophrenic comparison group was comprised

of many depressed patients (Fast & Pawl, 1970). It would

have been surprising if these depressed control subjects had

reduced death fear scores in light of the previously cited

indications that depression is linked with high death fear.

To summarize, in most empirical investigations, negative

death attitudes such as high death fear or high death anxiety,

have been associated, both in the aged and in younger adults,

with negative indicators of physical and psychological health,

particularly with the presence of depressed feelings and

other neurotic behaviors.

General Anxiety and Other Indicators of Maladjustment

Research consistently indicates a positive association

between death anxiety (DAS scores) or death fear (scores on

the Collett-Lester FDS) and general anxiety, as measured by

a variety of standard scales (Dickstein, 1972; Farley, 1971;

Handal, 1969; Handal & Rychlak, 1971; Lucas, 1974; Nogas et

al., 1974; Templer, 1970, 1971, 1972). Researchers have also

expected a relationship between death fear/anxiety and several

other negative indicators of psychological health. The

Repression-Sensitization Scale, for instance, identifies the

extent to which a person is aware of and actively responds

to arousing stimuli. The sensitized individuals are thought

of as tending toward anxiety, and prove to have more negative

attitudes toward death, whether these attitudes are assessed

by the DAS (landal & Rychlak, 1971; Tolor & Reznikoff, 1967),

DCS (Dickstein, 1972), TI (Tobacyk & Eckstein, 1980), or a

combination of items from a number of instruments (Farley,

1971). When presented with death-related material to read,

however, sensitizers did not report more anxiety than repres-

sors (Paris & Goodstein, 1966).

Feldman and Hersen (1967) reported a significant rela-

tionship between conscious death fear and frequency of night-

mares, a result not replicated by Lester (1968). Lester

(1967a) did find that those individuals who had attempted

suicide in a sample of adolescents were less afraid of death

than their nonsuicidal counterparts, but found no relation-

ship in a normal, adult population between one's attitudes

regarding suicide and those regarding death (Lester, 1971b).

In summary, investigators have generally reported that

people with higher death fear or death anxiety score higher,

as well, on a variety of indices that may be thought of

broadly as indicators of maladjustment. Let us now turn to

several theories which try to account for the relationship

between positive death orientation and positive life orien-


An Existential Hypothesis

Personal death is an important topic in the writing of

the existentialists. And, although there is variation in the

amount of attention they give to death and in the precise

manner in which death fits into their overall scheme, there

is a fundamental similarity in the way in which they deal with

the issue (Carse, 1980; Nagley, 1978; Olson, 1971). As a

group, Kierkegaard, Heidegger, Jaspers, and Sartre focused

their energies on understanding and resolving the general

malaise they observed in modern societies. Science and


technology have contributed to a higher standard of living

in which basic needs are satisfied. Rather than producing

happiness, this situation has resulted in the continualcrv

ing for more that is characteristic of the materialistic

consumer-oriented Western countries. Underneath the smiln

surface, the existentialists perceived a yawning gulf o

emptiness and boredom. What went wrong? After examining

the existentialists' analysis of the root of this emptiness

we will explore their notion of the awareness of personal

death and how this awareness is supposed to result in a mr

intense, meaningful life.

The existentialist diagnosis is that people have los

track of their individuality and have been swallowed up i

the attitudes and behavior of the herd. This happens ina

variety of ways. For one thing, we easily fall into the

trap of focusing on what we do rather than trying to dis-

cover who we are. Our conversation frequently betrays the

importance we give to the roles we play- "I am a counselr

what do you do? I-hat is your major? What do you teach?

Are you married? What does your husband do?" Becaus o

the complexity of modern life, people have come tobeiv

that truly worth-vwhile goals are accom-,pli shed by team ly

with everybody working together and contributing tothwol

through their own small role. The problem- is, according

thLe existentialists, we cannot follow group guidelinewihu

losing the self-consciousness which makes us uniquelyhmn

course," I am basing my behavior on something other than my

own experience and my own intentionality. An alternative way

of losing ourselves is to develop the belief that "I will be

happy and alive as soon as. . No matter how we finish

that sentence, we betray an attitude of de-valuing our present

experience. There seems to be little reason, therefore, to

be intent on the present. And if we find a new "as soon as

tomorrow, intentionality and human self-consciousness

may never begin (Slote, 1978). Small wonder, thought the

existentialists, that people experience an emptiness. People

have lost their individuality and ceased to live in a specif-

ically human mode by fading into the crowd or in some other

way failing to be intent and self-aware in the present.

The cause of the emptiness perceived by the existential-

ists seems to run to the very core of what it means to be

human. What sort of solution can they offer? Each person,

they believed, can re-discover the important aspects of human

existence through an encounter with death, not death in gen-

eral, but personal death, "my death." This encounter is an

awareness of my essential, absolute limitation: my personal

mortality, an awareness with intellectual, emotional, and

behavioral implications (Choron, 1963). According to exis-

tentialist thinking, bringing my future death into the

present reveals the present for what it is: all that a

person has. A person could turn away from this awareness.

People do, after all, fade back into the crowd. They pre-

tend to be making a contribution to some significant cause,

or they continue to live "as soon as ."--any method to

avoid the awareness of who they really are. On the other

hand, a person who resolutely faces personal death learns

that, while the escape routes preclude anxiety, they also

preclude genuine human existence. For the person who accepts

personal death and its implications for living a truly human

existence, the present acquires an urgency and intensity be-

cause it is all the person has (Koestenbaum, 1971). The

individual who is aware of personal mortality will focus on

the most important and meaningful possibilities of the moment

and will be aware of personal freedom rather than being pres-

sured by the opinions or behaviors of the group.

To summarize, the existentialists believed that without

the awareness of personal death, life remains artificial. A

uniquely human mode of living is possible only for the indi-

vidual who learns to face the necessity of personal mortality.

Personal death, in a certain sense, does not threaten my

existence. Instead, it is one of the essential features of

my existence (Gray, 1967). An intense, free, vigorously

human life is characteristic of the person with this view

of personal death. To live life well, a person must face


A Personal Construct View

Like other humanistic psychologists, Kelly focuses his

attention on the living, active person-in-process. At first

glance, a personal construct view of optimal functioning

seems to have less to do with death and more to do with com-

pleting the cycle of experience and with the process of de-

veloping a differentiated and integrated personal construct

system with which to approach life. Kelly (1965) did, it

is true, extend his theory to the phenomenon of suicidal

behavior, proposing the intriguing notion that suicide may

be thought of as a way of preventing further invalidation

and the total collapse of a person's system for interpreting

life. Rather than face a future of total chaos without a

system which can accurately anticipate the on-rushing events

of life (except the certainty of still further invalidation!),

the person may opt to preserve the remnants of his/her sys-

tem of meaning by suicide. Landfield (1976) and Lester

(1968, 1969, 1971b) pursued this hypothesis about suicidal

behavior with generally positive results. This application

of personal construct psychology speaks less to the meaning

of death, however, than to the lack of meaning in one's life.

Death as a possibility enters the picture only when life is

identified with chaos and a meaningful life seems impossible.

Three aspects of personal construct psychology reflect

more directly the relationship predicted by the existentialists

between a person's approach to death and optimal functioning.

First of all, Kelly frequently referred to death as an example

of an event which generally threatens the person. As Kelly

(1955) stated:

oThreatlikes the hawarenes tof theminent comlire-yt
heniveaot rsi changes in oneir core costructurssn.re

fcordn the theaty tohe significantigtieicanspectivechng

changeicalushane asubtatial persoh is for coexample.cph

Deangth iers wihhreathenining awa to motedulecVe describe
iasthreatening tour the because the y pumna erceive iton'
both asmp ike ito heappen oe a to thinn surielyn _onog
brings of bgouth drati change inteiifeontucs

Acordng tho Kelly thaed prcest ofvigifian person rslea ichmange

isilthwartwedndpersonratening conditones, which manino? Minht

superficialnchanger aslosrc the prorusforndoverntol mesapen the

hithraeningfeveto tieu wilhu eeec odlath tohend offcove ltheat

Comng thto terms with meaningis aortwaey, tomprtiule death'

theathnThe nature by psacholgicefndmnal meaning inprof cone'

lifeucompsatholeg with dath seodapiwilas onewa to enueanoging

Ona ta psothrhad must evfonerth personl resolver an icompa-

snout a peson, aftral coneblestrudcvalues fundlifenthal manino

his/heribe withoueth reeencea to dellyhs thenr discoe late

death? Theaatreo psychological meaning in perntson dalecticnl

in otheLr words, a person comes -to know a se-ries of events in

terms of the ways in which some events are si-milar to each

other and different from other events. To understand what

a person means by "playful," for example, we need to know

the opposite. One person may use the contrast "boring."

Playfulness means something quite different to another per-

son who thinks of people as "playful vs. hard-working." The

former person will relate to a playful person as a spontaneous,

joyful, exciting person; the latter might consider the playful

person to be lazy, childish, and irresponsible. The same is

true of a person's understanding of life. An exploration of

a person's understanding of death will provide information

essential to understanding what the person considers impor-

tant in life. Operating within a personal construct perspec-

tive, Rowe (1932) reported on her attempts to explore her

clients' "metaphor" about death in order to determine the

questions their life was meant to answer. Rowe believes that

one of the most significant questions a therapist can ask

clients is: "What comes after death?" Another is: "Do

you want to be buried or cremated?" Rowe's clinical evidence

supports her notion that clients who find an optimistic answer

to their questions about death will be able to live a fuller,

more meaningful life than are those clients who arrive at a

negative answer. Because of the dialectical nature of psycho-

logical meaning, a person cannot construe adequately the

meaning of life without construing the meaning of its


A third aspect of Kelly's theory also suggests that a

positive death orientation will result in more optimal func-

tioning. According to Kelly, the resolution of a conflict

r-equres the development of more abstract, superordinate

constructs, resulting in a more integrated construct system,

which may be considered a feature of optimal functioning.

A person who views life and death as falling on opposite

poles of a variety of personal constructs (high death threat)

will avoid behaviors which would imply that death is a real

possibility, compatible with his/her definition of core

self. In the process of resolving the incompatibility be-

tween self-definition and death, the person will actively

engage in death-related behaviors, which will require and

support the development of a more structured, integrated

construct system through the creation of superordinate,

linkageconstructs. In several ways, therefore, construct

theory supports the notion that coming to terms with death

by reducing death threat is a basis for optimal functioning.

Rationale and Hypotheses

Previous research has indicated a moderate relationship

between positive orientation toward death and aspects of

optimal functioning. People with less anxiety or fear

about personal death tend to have higher scores on measures

that tap self-esteem, for instance, and lower scores on

general anxiety and depression indices. Two weaknesses of

these previous investigations will be examined in this sec-

tion. After each is examined, a rationale for the present

study will be presented, together with the hypotheses to

be tested.

The Complexity of Death Orientation and
Optimal Functioning

The first limitation noted in previous research is that

death orientation and optimal functioning have been conceptu-

alized frequently in unidimensional terms, with single scale

questionnaires to assess death anxiety or death fear and

single scale questionnaires to assess self-esteem, purpose

in life, or general anxiety. Studies which have operation-

alized death orientation or optimal functioning in a more

complex manner have, nonetheless, examined individual cor-

relations between subscales rather than the overall pattern

of the relationships. Results which support a relationship

between low death fear, for instance, and high scores on

two or three Personality Orientation Inventory (POI) sub-

scales are more suggestive than conclusive. One might wonder

whether death orientation is related to optimal functioning

overall, or merely to the specific aspects measured by the

two or three POI subscales. The present study represents

an attempt to conceptualize multidimensionally both death

orientation and optimal functioning.

The multidimensional nature of death orientation has

been reflected through the use of the Collett-Lester Fear of

Death Scale (FDS) and the Threat Index (TI), each of which

represents a multidimensional approach to death orientation.

The FDC3 .was constructed to assess the four predetermined

aspects of death previously described, and the TI reflects

the idiosyncratic dimensions each person constructs to


understand and anticipate the phenomenon of death. Further-

more, the TI and the FDS represent a combination of the two

empirically derived (Rigdon & Epting, 1981) factors of nega-

tive evaluation of death (TI scores and scores on th& FDS

subscales for fear of death and dying of self and death of

others) and avoidance of the dying and of death reminders

(scores on the FDS subscales for fear of death and dying of

others). From several perspectives, therefore, the FDS an

the TI represent a multidimensional approach to death orien

station. The third empirically derived factor (the frequency

of thoughts about death) was not assessed in the present

study for two reasons. First of all, previous research mor

frequently has explored the relationship of aspects of opti

mal functioning with death attitude measures that load highly

on the first two factors of negative evaluation and avoidane

Secondly, there seems to be no compelling reason to believe

that optimally functioning persons think about death more

frequently or less frequently than other people.

An additional advantage of using the TI arises fromth

attempt to conceptualize death orientation in theoretical

terms. Whereas. other conceptualizations are in general tem

of fear, anxiety, concern, or acceptance, personal construc

theory, on whicn the 71 is based, provides a conceptualizato
of death orientation in terms of threat, i.e., the aarenes

at the prospect of imminent death, of comprehensive cange

in the constructs used to clarify and maintain one's dentty
Futeroe cntuc hor ousso teay n uget

how to reduce threat so that significant personality change

and growth can continue.

A multidimensional conceptualization of optimal function-

ing is reflected in this study through the use of differen-

tiation, integration, and purpose in life. Differentiation

represents an important process in optimal functioning--the

development of independent perspectives regarding a set of

events such as personal death and dying. Differentiation

was operationalized in two ways. The first was Landfield's

procedure for calculating the number of Functionally Inde-

pendent Constructions (FIC) each person uses for interpreting

the phenomenon of death. The second was the number of times

a person indicated that neither pole of a construct applied

or a lack of certainty about which pole applied. This index

of the applicability of the constructs is referred to as the

range of convenience (ROC). Crum (1978) reported a signifi-

cant correlation between FIC and ROC (r = .80), and he found

that ROC scores,like FIC scores, predicted performance on

several variables. The FIC effect, however, was more fre-

quently significant, whereas the effect of ROC was often

oni, marginally significant. Both process and structural

aspects of optimal functioning are reflected in integration

--the development of hierarchically arranged constructs which

permit the system to function as a coordinated network. Land-

field's ordination and chi-squared indices served as operations

of integration in the person's death-related construct system.

Finally, the content aspect of optimal functioning is reflected

in a person's set of beliefs relative to life's purpose and

meaning, assessed by the Purpose in Life Test (PIL).

The first hypothesis for this investigation arose from

the multidimensional conceptualization of death orientation

and optimal functioning and from previous research support-

ing an inverse relationship between death anxiety or fear

of death and aspects of optimal functioning. A significant

inverse relationship is predicted between death orientation

and optimal functioning. Specifically, low fear of death

and dying of self, low fear of the death and dying of others,

and low death threat are related to high purpose in life and

highly differentiated (as assessed by FIC and ROC) and highly

integrated (as assessed by ordination and chi-squared) con-

struct systems for anticipating death.

Two Experimental Hypotheses

A second weakness of previous studies, in addition to

the frequently unidimensional conceptualization of death

orientation and optimal functioning, has been a lack of an

experimental test for two rival hypotheses that may explain

the correlation between positive death orientation and

optimal functioning. It may be that a person's ability to

cope with stressful. life events determines the efficiency of

one's strategies for coping with death. On the other hand,

a person's ability to come to terms with death in a positive,

accepting way may determine one's ability to live a healthy,

productive life.

Investigators may hesitate to test these competing hy-

potheses because of the difficulty in changing death-related

attitudes. A variety of studies have investigated ways to

improve death attitudes. Preexisting groups such as Zen

meditators and Buddhists (Garfield, 1975), meditators (Curtis,

1980), and creative artists (Goodman, 1975) have been found

to have lower death anxiety than control subjects. Other

studies have used a quasi-experimental design, comparing

the death orientation of groups who completed death educa-

tion experiences to the orientation of control group subjects.

Some studies have not found the expected changes in death

attitudes. Bell (1975), for instance, reported no differ-

ences in fear of death, although participants in the death

education course thought about death more frequently than

control subjects and were more interested in discussing

death. And Rainey and Epting (1977) found lower death

threat in death education students not only after the course,

but also before it began, indicating that positive death

orientation may be a selection factor for death education

courses. In a similar study by Tobacyk and Eckstein (1980),

however, death education students, compared to another group

of students, had lower TT scores, when pretreatment differ-

ences in threat were statistically controlled. Several

other studies have also supported the effectiveness of

death education experiences. Murray (1974) and Whelan and

Warren (1930) found reduction in death anxiety when the post-

test was delayed for four and eight weeks, respectively, after

the death-related experience had ended, suggesting that it

may take some time for changes in death attitudes to be con-

solidated. Durlak (1978) found that an eight-hour didactic,

information-oriented workshop on death reduced death anxiety

(DAS scores) more than an experiential, feeling-oriented

workshop, while the experiential workshop resulted in lower

death fear (scores on Lester's FDS). In a similar experiment

(Kurlychek, 1978b), participants randomly assigned to an

experiential death education course had lower fear of the

death of others than participants assigned to an information-

discussion course on death and a waiting list control group.

But the information-discussion course produced lower

fear of death of self, compared to the waiting list controls.

Interestingly, Durlak and Kurlychek also investigated the

effect of the death-related experiences on purpose in life,

one aspect of optimal functioning as conceptualized in the

present study. Durlak reported no impact of either workshop

or. PIL scores, while Kurlychek found that the experiential

death education course, compared to the information course

and the waiting list group, produced higher PIL scores.

In light of the perspective previously presented from

existential philosophy and personal construct psychology,

the present study was an experimental test of the impact of

positive death orientation on optimal functioning. In addi-

tion to the previous correlational hypothesis, two experimen-

tal hypotheses were generated. First, it was hypothesized

that the treatment groups, compared to the control group,

will have more positive death orientation, i.e., lower fear

of death and dying of self and of others and lower death

threat. This is a check on the effectiveness of the treat-

ments. Second, the treatment groups, compared to the control

group, will be functioning at a more optimal level at post-

test. In other words, they will have more purpose in life

and will be more differentiated and integrated in their

constructed anticipations of death.

The importance of resolving this issue can be seen in

the implications for therapy, counseling for personal growth,

and death education. Death should be a central issue in

therapy and counseling, as Rowe (1982) believed, if a change

to a more positive death orientation results in healthier

functioning. Death education courses would acquire an im-

portance not only for their potential to change attitudes

about death and dying, but also for their potential to pro-

duce a higher level of functioning in life. Should evidence

support the impact of general life coping strategies on one's

approach to death and dying, then counseling for personal

growth might profitably focus on life coping skills, and

death education would lose some of its importance as a

personal growth experience.



Participants in the study were undergraduate students

in general psychology, who received course credit for their

participation. Of the 96 who participated, 79 were fresh-

men or sophomores. Their ages ranged from 17 to 32, with a

median age of 18.8. All but ten were in the 18 to 20 age

range. An equal number of 48 females and 48 males partici-

pated. Four of the subjects were black; the others were



The four questionnaires for this experiment were the

Threat Index (TI), the Collett-Lester Fear of Death Scale

(FDS), the Purpose in Life Test (PIL), and a Personal Data

Questionnaire (PDQ). Appendices A, B, and C contain a copy

of the questionnaires, except for the copyrighted PIL.

Threat Inde:r (TI)

Because the process of eliciting death-related constructs

from each person is time-consuming, the present investigation

used a provided form of the TI, consisting of 30 of the most

frequently elicited death-related constructs. Krieger,

Epting, and Hays (1979) initially developed the 40-construct

provided form, the TIp4O. The number of constructs was re-

duced later to 30 (Neimeyer, Dingemans, & Epting, 1977).

For the present study, a new TIp30 was introduced. Like the

former TIp30, it used 30 constructs appearing on the TIp40.

An examination of five of those constructs suggested that

they were elicited not by death as an event, but by the con-

crete stimulus "death" which was an element subjects used to

generate the initial constructs in the elicited version of

the TI. These five constructs were: relating to others-not

relating to others, specific-general, concrete-abstract,

personal-impersonal, and subjective-objective. Compared to

the other death elements, this vague "death" element could

reasonably be described by the right pole in each of the five

constructs. Accordingly, these five constructs were replaced

with five others'from the TIp40 which were more clearly re-

lated to death as a personal event: pleasure-pain, helping

o'her.3-beipg selfish, insecure-secure, useful-useless, and

sta:n.-it ion-growth.

The scalar form of the TI was used. According to this

form, the elements are rated along 13-point scales, with the

construct poles as the end points. Furthermore, in order to

score the TI [or the differentiation and integration indices,

one slight modification was initiated. In addition to rating

the standard elements self and my own death, each person rated

changes, from previous TI research (Krieger et al., 1974;

Rainey & Epting, 1977). To control for an order effect,

these 10 elements (listed in Appendix A) were presented to

each participant in random. order after the self and personal

death elements. The 30 constructs also were presented to

each participant in three different random orderings.

The TI was scored for the number of splits between the

way self and my own death were rated. A person, for instance,

who rated self as useful and death as useless received a scare

of one split. The total score was the number of splits for

the 30 constructs. Because the ratings allowed for a neutral

zero rating, each participant was requested to make a second

choice for those constructs on which the zero point was usecd

to rate self or personal death.

Various forms of the TI have Droduced acceptable esti-

mates of reliability, with a median test-retest coefficient

of .86 for a 4-week period and a median split-half coefficient

of .93 (Rigdon et al., 1979). Research has also supported

the validity of the TI as a measure of death orientation

(Rigdon et al., 1979). Threat index scores correlate moder-

ately with scores on a variety of other death orientation

instruments, for instance, and those who had planned their

funeral or provided for body disposal at death had, as ex-

pected, lower T1 scores than nonplanners.

In addition to the number of splits, the Threat Index

self and personal death were rated (WTI), an index used by

Warren (1982) in place of the split score. A person, for

example, who rated self as extremely useful (+6) and death

as slightly useful (+2) would receive a score of 4 on that

construct. The WTI score was the total for the 30 constructs.

The number of zeros the participant initially used to rate

personal death was also calculated because meaning depolari-

zation (i.e., the tendency to rate elements toward the

neutral midpoint) is considered in construct theory as an

indicator of possible anxiety (Rigdon et al., 1979)--an event

the person is unable to make any sense of. Although the

hypotheses for the experiment were not phrased in terms of

the difference score (WTI) or the number of zeros, post hoc

analysis would include these potentially useful variables.

The ratings of the 10 death-related elements were scored

for range of convenience (ROC) simply by counting the number

of zero ratings. The scoring for the Landfield FIC, ordina-

tion, and chi-square indices was more complex (Landfield &

Schmittdiel, 1981; Leitner et al., 1976). For FIC, each

construct dimension was matched with each other construct

dimension, and it was noted when there was agreement in the

application of the two constructs to the death-related events.

The fewer the instances of overlap, the more functional inde-

pendence of the two constructs; the more overlap, the more

functional similarity. Using the standard 30% overlap

criterion, two constructs were considered functionally

identical when the participant used the same construct pole

Lo describe eight of the ten death-related elements. The

same procedure was applied to the 10 death-related elements.

Two elements were considered functionally identical if the

participant used the same pole of 24 of the 30 constructs

to describe the two events. The number of functionally

independent constructs and the number of functionally inde-

pendent elements were surnmmed for a total FIC score.

To calculate the ordination score for a construct, the

10 element ratings on that construct were examined. If a

participant used only three different rating points (e.g.,

2, 5, and 6), a score of 3 was multiplied by the difference

between the highest and lowest ratings (6 2 = 4) for a

score of 12. This calculation w.as made for all 30 of the

30 scores. Similarly, ordination scores were calculated for

each of the elements. The number of different rating points

used in describing an element with the 30 constructs was

multiplied by the difference between the high and low ratings.

The ordination score for the elements was the average of the

10 element scores, which was then combined with the average

ordination score for constructs for a total ordination score.

To calculate the chi-square index for a construct, it

was necessary to count the number of times that each rating

point was used to describe the 10 death-related elements.

From these frequencies, the chi-square statistic was calcu-

lated as a measure of the difference from a rectangular

distribution in which all ratings are used with equal fre-

quency. After the chi-square statistic was calculated for

the 30 constructs, an average was taken as the chi-square

score for constructs. An average chi-square score for

elements was also calculated and added to the average chi-

square score for constructs to create a total chi-square


The Landfield indices for differentiation and integration

(including ROC) were calculated from the TI ratings by com-

puter (Landfield, Page, Barrows, Willis, Lavelle, & Schmittdiel,

1981). To ensure that threat scores be mathematically inde-

pendent of the differentiation and integration indices, no

ratings were scored for both threat and the latter indices.

Differentiation and integration scores were calculated on the

ratings of the 10 death-related elements only, while the rat-

ings of self and personal death were used to derive death

threat scores.

For clients in therapy, Landfield (1971) reported test-

retest reliability coefficients of FIC scores of approximately

.50. These low correlations are not so surprising, in that

therapy should produce change in the clients' construct

systems. In this population, one would expect change in

FIC scores rather than stability. The stability of FIC

scores needs to be examined with participants for whom no

intervention occurs from pretest to posttest. Crum (1978)

reported a test-retest reliability coefficient of .71 for

ROC scores and a low .48 for ordination scores. These co-

efficients may have been artificially reduced by the fatigue

of the participants, who completed the retest after a two-hour

experimental session. Crum reported that many subjects

seemed bored and completed the retest in a hurried manner.

Reliability estimates for chi-squared have not been reported.

The Collett-Lester Fear of Death Scale (FDS)

This questionnaire consist of 36 items to which a per-

son responds on a 6-point scale to indicate slight, moderate,

or extreme agreement or slight, moderate, or extreme dis-

agreement with each statement. The FDS is scored for four

aspects of death orientation: fear of the death or dying of

self and fear of the death or dying of others. As summarized

by Lester (1974) and by Wass (in press), the evidence supports

the test-retest reliability and internal consistency of the

FDS. The validity also has been supported by moderate corre-

lations between FDS subscale scores and scores on several

other death attitude scales.

The Purpose in Life Test (PIL)

Participants respond to the 20 PIL items on a 7-point

scale to indicate little meaning in life (score of 1) or

significant meaning in life (score of 7). Investigations of

the reliability of the PIL have resulted in a median split-

half reliability coefficient of .91 (Crumbaugh & Maholick,

1969; Reker & Cousins, 1979). Test-retest reliability ranged

from .83 for one week to .68 for 12 weeks (Reker, 1977). Evi-

dence supporting the validity of the PIL as a measure of mean-

ing in life was presented earlier.

The Personal Data Questionnaire (PDQ)

Items were constructed to assess demographic variables,

participants' religious beliefs and practices, and their pre-

vious experience of near-death or the death of significant

others. Analyses involving the variables assessed by the

PDQ are regarded as post hoc, in that no a priori hypotheses

were generated.


To carry out the experimental design required two ses-

sions. The first session involved the completion of the

pretest questionnaires and participation in a death-related

experience for the treatment groups or an alternative exper-

ience for the control group. In the second session, partic-

ipants completed the posttest questionnaires.

First Session

Participation in the experiment was in groups, ranging

in size from 9 to 12 members. The session began with a

description of the purpose of the study and of the procedures

as stated on the informed consent form (Appendix D). After

an opportunity to ask questions, the participants signed the

informed consent form. Each participant next received a

packet of four questionnaires in random order: the TI, the

FDS, the PIL, and the PDQ. When every member of the group

had completed the questionnaires (45 minutes to an hour),

the group members participated in the experience to which

they were randomly assigned: the death-related experience

of writing an obituary/biography (Appendix E) or of writing

a farewell letter (Appendix F) or the stress management ex-

perience (Appendix G) not related to death (control group).

The two death-related experiences were selected as com-

monly used components of death education courses that are

likely to reduce threat, in that both involve the playful,

imaginative realization of one's personal mortality. Writing

the farewell letter invites the person to behave "as if"

he/she were dying, and writing an obituary/biography invites

the person to act "as if" he/she were dead. Just as the

playful enactment of a new role in fixed role therapy (Kelly,

1955) provides for the development of new constructs to sup-

port the new behavior as "my own," so the enactment of oneself

as dying or dead would provide the opportunity to resolve the

split between self and death through the development of super-

ordinate construct dimensions which resolve the split. The

use of these two experiences to reduce death threat permits

the determination of whether these components of death educa-

tion courses are the ones that are generally responsible for

the change in death attitudes reported as the result of death

education courses, which are comprised of a multitude of

components, both informational and experiential.

The stress management experience (Apgar & Callahan, 1980)

was selected as a treatment which was not related to death,

but one that would be interesting and useful to the partici-

pants in the study.

Each experience began with a brief introduction and a

period of time for the participants to complete their written

response, followed by an opportunity to share with others

what they had written and to discuss the meaning of the ex-

perience and their reaction to it. The discussion phase of

the stress management experience was more structured, involv-

ing the presentation of information about stress and its

control, as well as an opportunity to practice a relaxation

exercise. Each experience lasted for about an hour, after

which participants selected a time to return for the second

session seven weeks later.

Second Session

A reminder of the second appointment was mailed to all

participants. When they arrived for the second session, they

completed the TI, the FDS, and the PIL, once again in groups

of 9 to 12. This session ended with an opportunity to re-

quest feedback about the results of the investigation.


This chapter will present the results of the analyses

of data provided by the participants in this study. After

data regarding instrument reliability are presented, the re-

sults pertaining to each hypothesis will be presented in order,

followed by a presentation of analyses for variables about

which no a priori. hypotheses were generated.

Instrument Reliability

The provided form of the T1 used in this study involved

minor changes from previous TI research. Evidence regarding

the reliability of this TI form was, therefore, important.

The data supported both the internal consistency of the TT

(split-half r = .90, corrected by the Spear-man-Brown formula)

and the test-retest stability for the 31 control participants

over seven weeks (r = .73). These reliability estimates are

comparable to those previously reportedI for other TI forms

(Rigdon et al., 1979).

While the stability of the- overall FDS scores (summing

the four subscale scores) was acceptable (r_ = .76), only the

DyS (fear of dying of self) subscale scores matched that level

of relationship (r_ ! .73). Pre-post correlations for the DS

(fear of death of self; .59), DO (fear of death of others;

.42), and DyO (fear of dying of others; .46) subscales,

though significant, are lower than one might wish, perhaps

due to the small number of items for some of the subscales.

Furthermore, the FDS subscale intercorrelations (Table 1)

were surprisingly high, some apparently higher than the

estimates of individual subscale stability. One would expect

greater independence between supposedly orthogonal aspects

of death orientation.

PIL scores, on the other hand, were quite stable (r =

.85). The other optimal functioning indices, unfortunately,

did not seem to have the same level of stability as PIL scores.

Pretest-posttest correlations for FIC (functionally independ-

ent constructions; .52), ROC (range of convenience; .60), and

chi-square (.65) scores indicated significant, though mod-

erate stability. There was no relationship, however, between

ordination scores derived from the two testing sessions (r =

.18, n.s.). Furthermore, there was post hoc evidence that

the scores of the 31 control group members significantly

decreased on the three most stable reptest indices, although

not on ordination scores (Table 2).

In summary, the results supported the reliability of the

TI, the PIL, and the FDS subscale for fear of the dying of

self (DyS). The moderate support for the stability of three

FDS subscales (DS, DO, and DyO) and of FIC, ROC, and chi-

square scores was less than one might expect. The results

simply failed to support the reliability of ordination scores.


DS --.63 .63 .28

Dys .61 .55

DO --.34

DyO -

Note. For all correlations, p < .003. DS=
Fear of the death of self. DyS = Fear of the dy-
ing of self. Do = Fear of the death of others.
DyO = Fear of the dying of others.

Table 2
Post IHoc Tests of Differences between
Pretest and Posttest Mean Scores on
Differentiation and Integration Indices

Index Pretest Posttestt(0
Mean Mean P_

ROC 43.6 20.5 4.42 <.001
(36.2) (20.7)
FIC 14.1 9.6 4.22 <.001
(6.5) (5.3)
CHI 55.1 48.9 2.34 <.026
(28.7) (20.8)
ORD 37.7 39.1 -.85 n.s.
(7.0) (7.5)

Note. Standard deviations are in parentheses.
ROC Range of convenience. FIC = Functionally inde-
pendent constructions. CHI = Chi-square. ORD

First Hypothesis

The first hypothesis predicted a significant inverse

relationship between death orientation and optimal function-

ing. More specifically, low fear of death and dying of self,

low fear of the death and dying of others, and low death

threat were predicted to be related to high purpose in life

and to highly differentiated and highly integrated construct

systems for anticipating death.

This hypothesis was tested using the pretest scores of

the 96 participants on the FDS, the PIL, and the TI (includ-

ing scores for the differentiation and integration indices).

As stated before, two scores were derived from the PIL: a

total score on all 20 items and a second score excluding

items 10 and 15, which appear to assess a person's attitudes

about personal death. To prevent participants' responses on

these two items from artificially inflating the relationship

between purpose in life and the death orientation variables,

the plan was to perform the analyses with and without these

two items. For the whole sample, the two scores were prac-

tically identical (r = .99). In order to have results that

are comparable to others using the PIL therefore, the analy-

ses which involved purpose in life used the participants'

total PIL scores. Pretest death orientation and optimal

functioning scores were analyzed by three multivariate pro-

cedures: canonical correlation, analysis of variance, and

multiple regression.

the pretest scores to estimate the amount of overlap between

the death orientation variables, on the one hand, and the

optimal functioning variables, on the other. An examination

of the correlation between the differentiation and integra-

tion indices, however, revealed strong relationships between

several of the optimal functioning variables (Table 3).

Entering all four variables in the same canonical analysis

would contradict the assumption of the relative independence

of the variables in each set in a canonical correlation.

One solution for this problem of multicallinearity (Green,

1973) is to drop from the analysis one highly overlapping

variable, retaining those that have the most theoretical

importance. For that reason, analysis of the pretest scores

initially used the FIC index because, as a differentiation

index, it is rooted more directly in construct theory than

ROC scores. As an integration index, thie first analysis

used chi-square scores which were develo-ped as an improve-

ment over ordination. Analyses were al-so performed using

heless highly related FIC and ordination scores as differ-

entiation and integration indices and, finally, the ROC and

ordination scores. Because of the potential multicollinearity,

scores which had a correlation greater than /.70/were not in-

clude-d toge-ther in any analysis: FIC and ROC (r = .74), ROC

Table 3
Pretest Intercorrelations between
Indices of Differentiation and Integration

Chi-square Ordination ROC

FIC .35* -.26 .74*

Cni-square -- -.82* .76*

Ordination -- -.51*


Note. Scores were based on responses of all 96
*p <.001.

and chi-square (r = .76), and ordination and chi-square (r =

-.82). The analyses (Table 4) supported a consistent rela-

tionship between positive death orientation and the optimal

functioning variables. Using a standard .30 value to gauge

the importance of each variable's contribution to the over-

all relationship indicates that low death threat, low fear

of one's own dying, and of the death and dying of others is

related to high purpose in life, high FIC and ROC scores

(highly differentiated views of death), and low ordination

and high chi-square scores (low integration in viewing death).

This relationship is evident in similar analyses that used

FIC, ordination, and chi-square scores for constructs only

(Table 4), instead of the total scores described earlier,

which were calculated for both constructs and elements. The

relationship between death orientation and optimal function-

ing was as predicted, except that positive death orientation

unexpectedly was related to low integration in viewing death

(low ordination and high chi-square scores).

A multivariate analysis of participants' scores on all

variables revealed evidence of initial sex differences (Table

5), due to differences in scores on four of the five death

orientation scales (Table 6). Female participants scored

significantly higher than males on fear of the death of

self (10.5 and 1.3), fear of dying of self (5.8 and 0.9),

fear of the death of others (10.6 and 3.7), and death threat

(17.7 and 13.5).

i I




Ln ,o
mn m


4-J -4J

0 u
. -i




0 -4

r 4-4
r .U


r u





Coi Ln
Ln a

o0 00
an un


C n %D Ln W in *'

N ,c r-N fM CD qo

N oo n r cL oo oo
e e a ui <

r-I r-4 C

r-I r-

OC, 0 U3 00 0

W M %D 0 M Ma r-

(U I

C-: t 0 -1
-0 -14
rO -4l

,4C --- C -l
E-x Q)
3 0 o

n a _, u)
I1 >, 43 -

.rl C: (C E
,c mE
0 0 o r -.i *,
-0 0 u

0 0, 4-H 40 )
U -) 2 4Q

k C *

0 H 0 rv

Oa) r r-

0 -0 -H

4 .im c
ro 4-H
,U ,C- Q 0
SC m 0

-, 0 (-
.0 Wu -j W 0
( m mC0

r L0 4-4.
0 0 Q0
C o
SW) I CD 0

>, >, 4-4

*H -I 0 CW

U-o raQ )
(u ( U 3 me
S in c
(D (1 4J 410 M .-,

X: .0 ( 0 iC

(C W I WO Ic -
-*H u ru -H U

4J CZ 0n)
4 Z CC 0
--1 0 ( E rv

*C E c E r, *4-
(O O 4 v v v
0 Ca 0 Ca c
O -a CO 0. Ol i 1
C; r- U .0J

a -4 r- -,1 5

0 0 H cO 0

U-I -

(C 0

O 4J

IH 4-4

C; M


0 -H
a, a

an Ln

(N O m




. 00

rj >




Table 5
Manova Summary Table for Pretest
Death Orientation and Optimal Functioning Scores

Source of variation df F

Sex of participant 10,81 2.81*

Treatment condition
Contrast between control and
average of two treatments 10,81 1.00
Contrast between two treatments 10,81 <1

Treatment by sex, roots 1 and 2 20,162 <1

Treatment by sex, root 2 9,81.5 <1

Note. The effect of treatment condition was tested
by the two a priori orthogonal comparisons. Tests of
significance used the Wilks Lambda criterion.
*E <.005.

Table 6
Analysis of Variance of Death Orientation
and Optimal Functioning Scores
for Sex Differences

Dependent variable MS F(1,90) p

Fear of death of self (DS) 1932.89 16.96 <.001
Fear of dying of self (DyS) 532.02 16.21 <.001
Fear of death of others (DO) 1066.17 .14.16 <.001
Fear of dying of others (DyO) 17.35 <1
Death threat (TI) 422.14 7.80 <.006
Purpose in life (PIL) 26.30 <1
Functionally independent
constructions (FIC) 101.62 2.41 <.124
Range of convenience (ROC) 1050.90 <1
Chi-square 188.02 <1
Ordination 4.38 <1


Because of the initial sex differences in death oren

station, canonical analyses were performed separately on

scores of the female and the male participants (Table7_

For the females, the results were similar to those ofth

whole group, with low death threat and low fear of the dyin

of self and of the death of others significantly associte

with high purpose in life, high differentiation (FIC), n

low integration (high chi-square and low ordination).Fo

the males, lowq fear of the dying of self and others wasre

lated significantly to high purpose in life, with theote

variables not making an important contribution to therea

tionship between death orientation and optimal functionng

Multivariate Analyses of Variance

In addition to exploring the linear relationship bewen

optimal functioning and death orientation variables, a secnd

method was used to explore the additive effects of some var-

able sets. Previous investigations (Crum, 1978; Landfed

1979; Landfield & Schmittdiel, 1981; Wilkins, 1978) hav

explored, for instance, the interaction between FIC an ori-

nation indices by analysis of variance, using a median pi

on both indices. In order to examine the Landfield indce

in a manner comparable to previous studies, similar nlic

that categorized participants by high versus low,!soes,,r

performed to explore potential interactive effectsatog

no a prior~i hypotheses w,.ere generated for the presetsuy



in in Lf




E l

I- -co

+ -- -
O 0 ,-I

c- rj CN

r mcc



In Ia I

M re) n qr CM 1r-
LALAn N 4N (4

V. o mO C co inLA
a r- o co r-

HAN NY 000
rN co Coo

n o o oI mI
ZT N m oa Ln Ln

" in r- co CN %.D
cMNo oC r-q r-



.r: -H 4.4

41 00

l l0 0


0 0

no uc

o, n U)-

a-lU) a,
U r-

S) .- 0

C) 4-1 Q) a,
40 V 0
4C) i rf U

i"O -H
O r

O -rO

$-J .0
m EO
0-) 1 0,
4 ) U)

0 r -
ra a Q

0 0 ....
(0 V0

.0 0- o
(14 r-i ) i4 v
10 C 0

Q) x Eu

0 *,--

a a4 C
p E 0 )

a r

4 ra r E
o u raJa
r1 gC
*H 10

r-1 *H

Q Or-I
4. rd-


o F- -H
OE -<

4J) 4) 0

0O E
aI 0

U 3

.0 -H
(t0 J
O-H 4
}- 4J)
i c
> 0)

0 0

0 4-
-r-1 IT,

rd -j

4 4J
.-1 -H
0 3

inter-ctlon between FIC and chi-square (Table 8) or between

FIC an.d ordination (Table 9). Nor was there evidence from

these analyses of any significant main effects on death orien-

tation scores for FIC, chi-square, or ordination.

Mult iple Regression Analyses

A final method for studying the relationship between

death orientation and optimal functioning was to examine the

ability of the optimal functioning variables to predict scores

on the TI as a single criterion, because the TI can be consid-

ered by itself to be a multidimensional measure of death atti-

tudes. Due to the multicollinearity problem mentioned before,

the multiple regression analyses once again used the differen-

tiation and integration indices in pairs that were not highly

correlated. Although the overall regression equations approached

significance (Table 10), there was no evidence that the optimal

functioning variables accurately predicted TI scores.

In summary, the results of the canonical correlation

analyses support the hypothesized association between positive

death orientation and the optimal functioning variables. The

relationship was as predicted, except that participants with

oositiLve death orientation tended to have lower, rather than

higher integration in their views of death. Although females

generally had higher scores on the death orientation scales,

the relationship between positive death orientation and opti-

mal functioning held for both females and males. There was a

difference, however, in the pattern of this relationship.

Table 8
Manova Summary Table for Effects of
High Versus Low FIC and Chi-square Scores
on Death Orientation Scores

Source of variation df F p

FIC 5,88 1.12 n.s.

Chi-square 5,88 1.87 <.11

FIC by chi-square 5,88 <1


Tests of significance used the Wilks Lambda

Table 9
Nanova Summary Table for Effects of
High Versus Low FIC and Ordination Scores
on Death Orientation Scores

Source of variation df F p

FIC 5,88 1.13 n.s.

Ordination 5,88 1.64 <.16

FIC by ordination 5,88 <1

Note. Tests of significance used the Wilks Lambda

Table 10
Multiple Regressions of Death Threat Scores
on Selected Optinal Functioning Variables

Predictor variables

PIL, FIC, and ordination

PIL, FIC, and chi-square

PIL, ROC, and ordination













For Lhe females, the pattern was similar to that of the whole

group, in that low threat and low fear ofthe dying ofself

and of the death of others were associated with high purpose

in life, high differentiation, and low integration. For the

males, low fear of the dying of self and of others was asso-

ciated with high purpose in life. For the males, neither the

differentiation and integration scores nor the other death

orientation scores contributed to the relationship. There

was no evidence from the multivariate analyses of variance

of an interactive effect of differentiation and integration

indices on death orientation scores. Ior did the optimal

functioning variables, according to the multiple regression

analyses, accurately predict TI scores taken as a single

estimate of death orientation.

-Second Hypothesis

The second hypothesis represents a check on the effective-

ness of the two treatments which were predicted to increase

the positive death orientation of the participants. Compared

to the control group participants, those who participated in

the death-related experiences were expected to have, at post-

test, lower fear of the death and dying of self and others

and lower death threat. To ensure an initial similarity in

death orientation and optimal functioning between the treat-

ment groups, a multivariate analysis of the pretest scores

on the five death orientation and the five optimal functioning

variables was performed, with sex and treatment level as the

two independent factors (Table 5). There was no evidence for

any interaction between treatment condition and participant

sex. As previously noted, the analysis did indicate higher

scores for female participants on four of the five death

orientation scales. There was no evidence, how'.'ever, of any

overall difference between the two treatment groups, or be-

tw:een the treatment groups and the control group on the death

orientation and optimal functioning variables. Even a post

hoc examination of the univariate tests associated with each

dependent variable (Table 11) indicated little evidence for

an, differences, with only the apparent differences in TI

scores, ordination scores, and chi-square scores approaching

significance (Table 12). The results indicate relative simi-

larity between participants in the three treatment conditions

in their initial level of death orientation and optimal


A multivariate analysis of variance of the death orien-

tation scores of the 95 participants who returned for posttest

(Table 13) produced results similar to the results from the

prete.t scores. Once again, there was no evidence for a sig-

nificant interaction between participant sex and treatment

condition. The sex differences remained stable, with the

females continuing to exhibit a more negative death orienta-

tion as at pretest. Specifically, they had higher mean scores

for the same death orientation scales: TI (19.1 for females

and 1C.3 for males), DS (6.7 and 1.9), DyS (7.1 and 2.1), and

Table 11
Anovas for Treatment Differences in Pretest
Death Orientation and Optimal Functioning Scores

Variable MS F(1,90) P MS F(1,90) p

DS 25.45 <1 152.25 1.34 <.25
DyS 1.22 <1 23.84 <1
DO 99.68 1.32 <.25 86.73 1.15 <.29
DyO 3.23 <1 3.90 <1
TI 132.31 2.45 <.12 39.46 <1
PIL 20.25 <1 65.57 <1
FIC .38 <1 97.35 2.31 <.13
ROC 4058.65 1.88 <.17 3330.37 1.54 <.22
CHI 8728.50 3.56 <.063 137.54 <1
ORD 332.66 3.52 <.064 9.45 <1

Table 12
Mean Scores for Death Threat, Ordination, and
Chi-square by Treatment Conditions for
Pretest and Posttest

Treatment Variable
TI Ordination Chi-square

Obituary/biography 14.0 34.4 76.1
Farewell letter 15.5 33.2 74.6
Stress management control 17.3 37.8 54.7

Obituary/biography 16.3 36.4 62.5
Farewell letter 17.5 33.5 64.3
Stress management control 19.2 39.0 45.2

Note. For each group, n = 31, except for the farewell
letter treatment group at pretest (n = 34) and at posttest
(n = 33).

Table 13
Manova Summary Table for Posttest
Death Orientation Scores

Source of variation df F p

Sex of participant 5,85 3.25 <.01

Treatment condition
Contrast between control and
average of two treatments 5,85 <1
Contrast between two treatments 5,85 1.06 <.39

Treatment by sex
Roots 1 and 2 10,170 1.30 <.235
Root 2 4,85.5 <1


Tests of significance used the Wilks Lambda

DO (7.8 and 4.2). Support for a difference in the mean

scores on the DyO subscale (-2.0 and -5.0) was marginal

(Table 14). There was no evidence that the two death-related

treatments, compared to each other or compared to the stress

management control condition, produced any difference in

death orientation. A post hoc examination of the univariate

tests of the five death orientation variables (Table 15) also

revealed no evidence of differences. The only F test larger

than 1 was associated with mean TI scores (Table 12), but

even this apparent difference was not significant. Overall,

the results did not support the effectiveness of either death-

related experience for producing a more positive death orien-


In order to prevent an interpretation of the apparently

lower TI scores for the treatment groups compared to the con-

trol group (Table 12) as an effect of the treatment conditions,

a multivariate analysis of covariance was conducted on the

posttest death orientation scores adjusted for possible pre-

test differences in TI scores (Table 16). This analysis also

failed to support a treatment effect or an effect for the

interaction of treatment and sex of participant. Of the

post hoc univariate F tests comparing the effect of the two

treatments to the control condition on the five death orienta-

tion scores, only the F statistic for DO scores was greater

than 1 [F(1,88) = 1.05, p < .31]. As indicated by the pre-

vious analysis, sex differences remained in the posttest

death orientation scores. Neither the multivariate analysis

Table 14
Anovas for Sex Differences in Posttest
Death Orientation Scores

Variable MS F(1,89) p

Fear of death of self (DS) 457.64 5.18 <.025

Fear of dying of self (DyS) 540.35 13.77 <.001

Fear of death of others (DO) 255.32 5.26 <.024

Fear of dying of others (DyO) 207.20 2.28 <.135

Death threat (TI) 164.10 2.77 <.100

Table 15
Anovas for Treatment Differences in
Posttest Death Orientation Scores

Contrast between
control and average Contrast between
of two treatments two treatments
Variable rMS F(1,89) E Mrs F(1,89) p

DS 6.19 <1 6.80 <1

DyS .44 <1 1.99 <1

DO 26.45 <1 116.38 2.40 <.125

DyO .22 <1 150.79 1.66 <.200

TI 126.38 2.13 <.15 47.95 <1

Table 16
Multivariate Analysis of Covariance of
Posttest Death Orientation Scores with
Pretest Death Threat Scores as Covariate

Soreof variation df F p-

Cvrate (TI) 5,84 13.10 <.001

Se participant 5,84 2.42 <.043

Tetent condition
Cotrast between control and
garage of two treatments 5,84 <1
Cotrast between two treatments 5,84 <1

Tetent by sex
Rots 1 and 2 10,168 1.48 <.15
Rot 2 4,84.5 <1

of variance nor the multivariate analysis of covariance re-

vealed any evidence that the two death-related experiences,

compared to the stress management experience, reduced the

participants' level of death threat and fear. The results

failed to support the second hypothesis.

Third Hypothesis

A higher level of optimal functioning for participants

in the death-related treatments compared to the control group

members was predicted in the third hypothesis. As indicated

above, analysis of the pretest scores did not provide evidence

of pretest differences in optimal functioning. A multivariate

analysis of variance of the posttest scores on the optimal

functioning variables similarly failed to support either a

significant interaction between participants' sex and treat-

ment condition, or a significant difference due to sex (Table

17). Contrary to expectation, however, there was no indication

of the increased optimal functioning of participants in the

death-related experiences. A post hoc examination of the

univariate tests of each of the five optimal functioning

variables revealed the same support found in the pretest

scores for lower integration in the treatment groups, with

higher chi-square scores and marginally lower ordination

scores (Tables 18 and 12).

In order to prevent an interpretation of the post hoc

indication of lower integration as an effect of the death-related

Table 17
Manova Summary Table for Posttest
Optimal Functioning Scores

Source of variation df F p

Sex of participant 5,85 1.50 <.20

Treatment condition
Contrast between control and
average of two treatments 5,85 1.16 <.33
Contrast between two treatments 5,85 1.14 <.35

Treatment by sex
Roots 1 and 2 10,170 1.16 <.32
Root 2 4,85.5 <1

Note. Tests of significance used the Wilks Lambda

Table 18
Anovas for Treatment Effects on
Posttest Optimal Functioning Scores

Contrast between
control and average Contrast between
of two treatments two treatments
Variable MS F(1,89) p MSS F(1,89) p

PIL 28.03 <1 2.61 <1

FIC 14.46 <1 45.64 1.59 <.21

ROC 590.04 <1 2903.79 2.25 <.14

CHI 7099.95 4.52 <.036 11.05 <1

ORD 351.62 3.57 <.062 129.65 1.32 <.25

experiences, a multivariate analysis of covariance was con-

ducted on the posttest optimal functioning scores adjusted

for possible pretest differences in ordination and chi-square

scores (Table 19). As in the previous analysis of the post-

test optimal functioning scores, neither the treatment effect

nor the interaction between treatment and sex was significant.

Even the univariate F tests for the five optimal functioning

variables failed to indicate any post hoc evidence for treat-

ment differences in optimal functioning scores. Finally,

although the multivariate test indicated a significant sex

difference in the adjusted optimal functioning scores, the

follow-up univariate tests did not reveal the nature of this

difference. Only the test for the adjusted chi-square scores

approached significance [F(1,87) = 3.81, p < .054].

To summarize, neither the multivariate analysis of post-

test optimal functioning scores nor the multivariate analysis

of covariance supported the significant effect of the death-

related treatment experiences predicted in the third hypothesis.

Additional Analyses

Besides the analyses linked to the three central hypothe-

ses, further analyses were performed regarding relationships

about which no specific hypotheses were generated. After

briefly reporting the results of these analyses, this section

will conclude with an overall summary of the results.

Table 19
M.ultivariate Analysis of Covariance of Posttest
Optimal Functioning Scores with Pretest
Ordination and Chi-square Scores as Covariates

Source of variation df F p

Roots 1 and 2 10,166 21.27 <.001
Root 2 4,83.5 9.57 <.001

Sex of participant 5,83 2.60 <.031

Treatment condition
Contrast between control and
average of two treatments 5,83 <1
Contrast between two treatments 5,83 1.09 <.37

Treatment by sex
Roots 1 and 2 10,166 1.72 <.08
Root 2 4,83.5 <1


Tests of significance used the Wilks Lambda

Peligious Beliefs and Practices

Previous research (Rigd3n et al., 1979) indicated the

potential usefulness of the TI for determining the relation-

ship between religious beliefs and death orientation. Al-

thoughy no a priori hypotheses were established for the present

study, relevant data were available to examine this relation-

ship. One-half of the participants reported they were

Protestant, 28 were Catholic, and 10 were Jewish. The

other 10 claimed other religious affiliations, including

agnosticism and atheism. In addition to indicating their

religious affiliation, participants responded to two items

regarding their overall level of religious devotion and their

belief in life after death. Three other items assessed the

frequency of attending church, saying grace, and reading the

Bible or other religious literature. When an examination of

the intercorrelations between these five items revealed a

high association between devotion and church attendance (.82),

the latter was retained in the analyses because it appeared

to be a more behavioral indication of religious devotion.

The first analysis, a canonical correlation of death orien-

tation variables and religious practice variables, revealed

a significant, moderate relationship (R = .45, R2 = .20, p <

.01) between death orientation and religious belief, with

more devout participants having more positive death orienta-

tion (Table 20). High scores on the TI, the DyS and DyO

subscales were related to a tendency not to believe in an

afterlife and to less frequent church attendance, religious

I ru

-Hr N c
rfl N3 C) )

4J -A 1

CD 4 -4
-Hf a *H -U

U 0 0-4
t44 .-4
4J >1 0 r-
4-U U D'
(y -I r=- II X 0

4 r 0) .N

U II O *

Q> Q 4 4
1 0 .- 4) r;

o 0 o >1,u- a)

S.i W0 a-n 0
.0 0 -. I II 0

r- >
Q 0 m N .C '-'a

Ul -0 O O *H( (

U -H C > 4- 4 M
0 >i0 *^-* *H -H X H

M 4 -o .)-J *

( I-. U t 0i
S- < ra n O

( o I o U H1)
S c ( u- o r J
O-3 W: T U Hc m in

r--i 4) 0 Q-.,4 ty,
4) 4 J Q)r p4-1 4 E -

0- fO C4 -O *4Qt
Z00 i (a 0 *EM

> O *>a) *I
00 -j U s

U) 0 0 0 a)

Q4J*H 0C **

UU 4-J C%
C --1 0 OI 0 1 En

S- -a 0 :- j

S 0 4 -o U D0. O i

C0 $4 Q-H CD
O < 0) 0
C I C30-4(N


reading, and saying grace. Because of the theoretical impor-

tance of the TI for the present study, the relationship between

death threat and religious practice was analyzed in particular.

A multiple regression analysis revealed a significant, moder-

ate relationship [R = .39, R2 = .15, F(4,91) = 4.02, p < .0051.

Because church attendance [F(1,91) = .021 and saying grace

[F(1,92) = .081 did not make a significant contribution to the

amount of explained variance, they were eliminated from the

analysis, resulting in a significant relationship of low death

threat to more frequent religious reading and a stronger belief

in a life after death (Table 21).

Previous Experience with Death and Dying

Of the 96 participants, 22 responded that there was a time

when they genuinely believed they were going to die, whether

from an accident (16), an illness (4), or a suicide attempt (2).

One might predict that such individuals, as a result of their

"close brush" with death, would develop a more positive death

orientation. A multivariate analysis of variance of the death

orientation scores failed to support such a hypothesis, al-

though the overall test approached significance [F(5,90) =

2.17, p < .0641 and the follow-up univariate test of the TI

scores indicated marginally lower death threat [F(1,94) = 2.70,

E < .10] for those who experienced a near-death experience

(mean = 13.23) compared to those who had not had such an

Table 21
Multiple Regression of Pretest Death Threat
Scores on Religious Beliefs/Practices Variables

Source df MS F p

Regression 2,93 416.30 8.15 <.0005

IAD 1,93 243.72 4.77 <.03

Bible 1,93 354.06 6.93 <.01

Note. LAD = Belief in life after death. Bible =
Frequency of reading bible or religious literature.

University of Florida Home Page
© 2004 - 2010 University of Florida George A. Smathers Libraries.
All rights reserved.

Acceptable Use, Copyright, and Disclaimer Statement
Last updated October 10, 2010 - - mvs