Anger

INFORMATION TO USERS

This manuscript has been reproduced from the microfilm master. UMI films the

text directly from the original or copy submitted. Thus, some thesis and

dissertation copies are in typewriter face, while others may be from any type of

computer printer.

The quality of this reproduction is dependent upon th e quality o f the copy

subm itted. Broken or indistinct print, colored or poor quality illustrations and

photographs, print bleedthrough, substandard margins, and improper alignment

can adversely affect reproduction.

In the unlikely event that the author did not send UMI a complete manuscript and

there are missing pages, these will be noted. Also, if unauthorized copyright

material had to be removed, a note will indicate the deletion.

Oversize materials (e.g., maps, drawings, charts) are reproduced by sectioning

the original, beginning at the upper left-hand comer and continuing from left to

right in equal sections with small overlaps. Each original is also photographed in

one exposure and is included in reduced form at the back of the book.

Photographs included in the original manuscript have been reproduced

xerographically in this copy. Higher quality 6” x 9” black and white photographic

prints are available for any photographs or illustrations appearing in this copy for

an additional charge. Contact UMI directly to order.

Bell & Howell Information and Learning 300 North Zeeb Road, Ann Arbor, Ml 48106-1346 USA

800-521-0600

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

Cognitive Distortions in the Experience and Expression of Anger

A Doctoral Dissertation

by

Deborah Azoulay

Submitted to:

Dr. Mark Stone, Ed.D, Committee Chairperson

Dr. John Newbauer, Ed.D, Committee Member

In Partial Fulfilment of the

Requirements For the Degree

Of Doctor of Psychology

Adler Education of Professional Psychology

Chicago, IL

USA

June, 1999

R eprod u ced with perm ission o f the copyright ow ner. Furth er reproduction prohibited w ithout perm ission.

UMI Number: 9944410

UMI Microform 9944410 Copyright 1999, by UMI Company. All rights reserved.

This microform edition is protected against unauthorized copying under Title 17, United States Code.

UMI 300 North Zeeb Road Ann Arbor, MI 48103

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

Adler School of Professional Psychology

Dissertation Final Submission Form

This dissertation has been successfully defended and approved for final submission.

Certified by: ^ ̂ Chair Date

icu jz_______ Meinber Date

Member Date

Accepted by: 9 /^ y /9 r Dean Date

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

ABSTRACT

This study examined cognitive distortions as predictors of the experience and

expression of anger as measured by the State Trait Anger Expression Inventory

(STAXI). It also compared subjects' self ratings with ratings of the subjects by live-in

partners for trait anger (and its subscales of angry temperament and angry reaction),

anger-in, anger-out, and anger-control. Using subscales of the Survey of Personal

Beliefs and the Common Beliefs Survey III as independent variables, stepwise

multiple regressions were performed for each of the anger scales. Results did not

support the hypothesis that different cognitive distortions would predict anger-in and

anger-out. The cognitive distortions most strongly predictive of all aspects of anger

were Low Frustration Tolerance (the belief that difficulties can not be withstood) and

Self Downing (self blame for problem situations). The relationship between Low

Frustration Tolerance and anger-control was inverse. In addition, results of this

study showed that self ratings and ratings of subjects by live-in partners have

moderate to high correlations, with the correlation highest for trait anger and lowest

for anger-in. However, scores of the two ratings were significantly different for all the

scales, with subjects' self ratings higher for trait anger and anger expression (in and

out) and lower for anger control. This study shows that an element of self criticism

appears to be associated with self ratings of anger, both in comparison to the ratings

by partners and in relation to the cognitive distortions which predict anger scores.

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

ACKNOWLEDGMENTS iv

I wish to thank the members of my committee, Dr. Mark Stone, Ed.D,

(Chairperson) and Dr. John Newbauer, Ed.D, for their willingness to share their

expertise and time to assist me with this dissertation. Their readiness to answer

questions and provide feedback at each step of the dissertation process was greatly

appreciated.

Thanks also to Dr. Gary Challis. Ph.D, and Dr. Joanne Coutts, Ph.D, Training

Directors at Whitby Mental Health Centre, Whitby Ontario. Their encouragement and

support greatly facilitated this work. I am extremely grateful to Dr. Challis for his

incredibly patient and generous guidance and assistance. It was invaluable and very

much appreciated.

Thank you to Dr. H. Kassinove, Dr. E. Nottingham, Dr. C. D. Spielberger, and

Dr. G. Thorpe for their assistance in facilitating my use of the instruments employed

in this study.

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

V

This D issertation is Dedicated

W ith Love and Gratitude

To My Family

My parents, Dr. Arnold and Mrs. Edith Lewin Their help, support, and

confidence in me made possible the completion of this dissertation and the

doctoral program.

My children, David and Karen A zoulay Their expressions of pride in their mother

strengthened my resolve to continue.

And Especially

My husband, Armand Azoulay He believed in me. He encouraged me. He did extra

housework without complaint. He readily gave up financial gain. He was my

chauffeur to the airport and he spent tim e alone while I was away. During five years

he did not waiver once in his com m itm ent to my studies. Without him this doctorate

would still be a dream.

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

DEBORAH AZOULAY 135 GlenmanorWay

Thornhill, Ontario, L4J 3A3

EDUCATION:

Doctoral Candidate in Clinical Psychology (Psy.D) Adler School of Professional Psychology (Chicago,IL)

Master of Arts (MA) in School Psychology University of Toronto (Toronto, ON) 1973

Additional Qualifications:

Registration as Psychological Associate (College of Psychologists of Ontario); 1994 Theraplay Therapy (Introduction and Intermediate courses) Certified Capacity Assessor (Under Substitute Decisions Act)

AWARDS

Claudine Bock Scholarship for Academic Excellence Adler School of Professional Psychology (September 1997)

DOCTORAL INTERNSHIP:

Whitby Mental Health Centre (Dr. J. Coutts and Dr. G. Challis, Training Directors) Whitby, Ontario September 1998-September 1999

Responsibilities:work with adult and adolescent psychiatric hospital inpatients and outpatients; outpatients in community clinic; psychological assessments (intellectual and personality); individual psychotherapy; group counselling/therapy including forensic anger management group; intake assessments; presentations to staff; intake and stabilization; forensic; trauma; psychotic, mood, and anxiety disorders; assessment and intake reports

W ORK EXPERIENCE

T o ro n to D istrict School Board Psycho-Educational Consultant District Wide Program R esponsibilities : intellectual/emotional/ Developmental Handicap behavioural assessments; development and

implementation of behaviour management Toronto, ON and social skills programs; teacher training and 1974-September, 1998 consultation; intake assessments; (students

with mental/physical handicaps, PDD, ADHD)

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

Deborah Azoulay

York Region Roman Catholic Psvchometrist (part time) vii Separate School Board Responsibilities: intellectual/emotional Aurora, ON assessments: consultation with teachers / parents; 1990 (6 months) development of remedial strategies for learning and

emotional/social difficulties

DOCTORAL STUDENT PRACTICUM EXPERIENCE

Baycrest Centre for Geriatric Care Toronto, ON summer 1997

Jewish Family and Child Service Toronto, ON April 1997-April 1998 (2 evenings per week)

Clinical Assessment Practicum Student Responsibilities: administration and scoring of neuropsychological, intellectual and personality instruments; geriatric and also younger adults presenting with dementia, MVA injury, stroke rehabilitation, depression.

Clinical Therapy Practicum Student Responsibilities: therapy for individuals, couples and groups with a variety of presenting problems including stress management; relationship problems, mood disorders, difficulties with child management; co-therapist of group assisting women with expression of anger.

Markham-Stouffville Hospital Clinical Therapy Practicum Student Mental Health Unit Responsibilities: short term Markham, ON treatment of out-patient and in-patient adults summer 1996 with a variety of mood, anxiety and eating

disorders; intake assessments of newly referred patients; some formal assessment and report writing; experience as co-therapist in day treatment group for in-patients;

Bellwood Health Services Clinical Therapy Practicum Student Scarborough, ON Responsibilities: co-therapist of the April -Dec. 1996 therapy group for adult children of alcoholics;

counselling group for clients recovering from substance abuse; support group for families of individuals with substance abuse problems

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

Deborah Azoulay

Aurora, On 1993-1995 (part time)

Blue Hills Child and Family Services

Theraplav Therapy Extern Responsibilities: family theraplay therapy with children and parents; parent counselling; intake assessment experience: Marschak Interaction Method

VII!

PUBLICATIONS AND PRESENTATIONS:

Azoulay, D. (June 1997). Multicultural aspects of the parent teacher relationship. Presented to Beverley Resource Centre, Toronto District School Board, Toronto, ON.

Azoulay, D. ( accepted for publication). Encouragement and positive reinforcement vs logical consequences and punishment: A re-examination. Journal of Individual Psychology.

Azoulay, D. (accepted for publication). Theraplay therapy with physically handicapped and with developmentally delayed children. In E. Munns (Ed.) Innovations in theraplav. New Jersey: Jason Aaronson Publishers.

Azoulay, D., Ali, J., Lawrence, H. & Munns, E. (1994). Increasing the bond between a mother and her child: A case study of family theraplay. Playground.. 5-6 & 11. Kingston, ON: Canadian Association of Child and Play Therapy.

Azoulay, D., Blurton, J. & Smith, P. (1985).lncreasinq appropriate peer interaction during leisure time of moderately developmentally handicapped adolescents. Presented at the annual conference of the American Association on Mental Deficiency (Ont. Chapter), Toronto, ON.

Azoulay, D. & Maniacci, M. (1996). Posttraumatic stress disorder: An Adlerian view of treatment. Individual Psychology 52 f4L 420-430.

Gouse-sheese, J., Azoulay, D., Armes, J., Botai, A., Oullette, M. & Luftspring, F. (May 1997). Models for teaching positive interactions to students with developmental handicaps and challenging behaviours. Presented at The Canadian National Conference on Educating Students with Behavioural Disorders., St. Catherines, Ontario.

Munns, E., Azoulay, D., Ali, J., & Wood, S. (1994, June). Theraplav therapy with a single abused mother and her abused children. Presented at the national convention of the Canadian Association for Child and Play Therapy, Toronto, ON.

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

ix TABLE OF CONTENTS

Approval II

Abstract iii

Acknowledgements iv

Curriculum Vitae vi

Table of Contents ix

Chapter I: Introduction 1

Chapter I I : Review of Literature 10

Chapter III: Methodology 70

Chapter IV: Results 82

Chapter V: Discussion 102

References 121

Appendix A: Names and Definitions of Variables 145

Appendix B: Means and SD and N of Subjects for Variables 147

Appendix C: Demographic Questionnaire 156

Appendix D: Sample Consent Form 157

Appendix E: Ethics in Research With Human Subjects Form 158

Appendix F: Sample Letter of Instructions to Subjects 159

List of Tables

Table 1: Relationship of Sex and Anger Variables

Table 2: Relationship of School and Anger Variables

Table 3: Relationship of Age and Anger Variables

R e p rod u ced with perm ission o f the copyright ow ner. F u rth e r reproduction prohibited w ithout perm ission.

X

Table 4: Relationship of Time and Anger Variables

Table 5: Relationship of Sex and Cognitive Distortion Scales

(Including Univeriate F Scores for Subscales of CBS)

Table 6: Relationship of School and Cognitive Distortion Scales

(Including Univeriate F Scores for Subscales of CBS and SPB)

Table 7: Tukey Post Hoc Test

(School and "Other Directed Shoulds")

Table 8: Tukey Post Hoc Test

(School Level and "Perfectionism")

Table 9: Relationship of Age and Cognitive Distortion Scales

Table 10: Relationship of Time and Cognitive Distortion Scales

Table 11: Anger Correlation Matrix

Table 12: Correlation Between Trait Anger and Its Subscales

Table 13: Correlation Between Trait Anger Subscales

Table 14: Correlation Between Trait Anger and

Anger Expression Styles

Table 15: Intercorrelation of Cognitive Distortion Subscales

Table 16: Relationship Between Cognitive Distortions

and Anger Scores (Full Sample)

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

Table 17: Relationship Between Cognitive Distortions

and Anger Scores (Higher Trait anger)

List of Figures

Rgure 1: Number of Subjects at Each Location

Figure 2: Education of Subjects

Rgure 3: Cognitive Distortions Predicting Trait Anger

R gure 4: Cognitive Distortions Predicting Angry Temperament

Rgure 5: Cognitive Distortions Predicting Angry Reaction

Figure 6: Cognitive Distortions Predicting Anger-ln

R gure 7: Cognitive Distortions Predicting Anger-Out

R gure 8: Cognitive Distortions Predicting Anger-Control

Figure 9: Cognitive Distortions Predicting Trait Anger

(Higher Anger Group)

Figure 10: Cognitive Distortions Predicting Angry Temperament

(Higher Anger Group)

R e p rod u ced with perm ission o f the copyright owner. F u rth e r reproduction prohibited w ithout perm ission.

R gure 11: Cognitive Distortions Predicting Angry Reaction

(Higher Anger Group)

R gure 12: Cognitive Distortions Predicting Anger-ln

(Higher Anger Group)

R gure 13: Cognitive Distortions Predicting Anger-Out

(Higher Anger Group)

R gure 14: Cognitive Distortions Predicting Anger-Contro!

(Higher Anger Group)

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

Chapter 1

INTRODUCTION

Anger is a common emotion that serves an adaptive function,

energizing behavior that is oriented to overcoming obstacles, correcting

injustices and solving problems. However, disruptive consequences can

also result from high trait anger or an intense expression of anger.

Deffenbacher, Lynch, Oetting, & Kemper (1996) state that high trait anger

. (anger experienced frequently over time) correlates with health problems,

especially cardiovascular disease. They also report that people with high

trait anger tend to express their anger in more dysfunctional ways which

negatively impact on education, work, relationships, self esteem and

cause internal turmofi and distress.

It has also been shown that anger that is not expressed but is rather

turned in can lead to difficulties. According to Kassinove & Sukhodolsky

(1995), "Suppressed anger is related to a number of medical conditions

including essential hypertension, coronary artery disease, and cancer"

(p.1). Novaco & W elsh (1989) state that the "inability to regulate anger

constitutes a risk factor for both harm-doing to others and for multiple

impairments affecting health, performance, and relationships” (p. 39).

R e prod u ced with perm ission of the copyright ow ner. F u rth e r reproduction prohibited w ithout perm ission.

2

Kassinove & Sukhodolsky (1995) go on to comment, "Clearly, it is

important to understand the causes, correlates, and outcomes of anger,

with the goal of developing effective remediation programs when anger is

excessive and disruptive" (p.2). Yet, "relatively little research has been

carried out in this area" (Kassinove & Eckhardt, 1995, p. 203). Only in the

last few decades has there been a growing interest in the study of anger.

Three research trends are noted (Ben-Zur & Breznitz, 1991): a) the

investigation of individual differences in anger proneness, b) the study of

methods to teach anger management, and c) exploration of "the observed,

inferred, or attributed characteristics of the situation which affects

anger...." (p. 1). To this list one more category can be added: the

investigation of the different modes of anger expression and their

correlates.

Definition of Anger

There has been debate about w hat anger actually is. Current

definitions "regard anger as a multidimensional construct consisting of

physiological..., cognitive..., and behavioral....variables" (Eckhardt,

Barbour, & Stuart, 1997, p. 335). Novaco (1995) agrees but also sees

anger as a stress reaction. Kassinove & Sukhodolsky (1995) add a social

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

3

constructive aspect while stressing cognitive distortions:

We define anger as a negative, phenomenological (or internal)

feeling state associated with specific cognitive and perceptual

distortions and deficiencies (e.g., misappraisals, errors, and

attributions o f blame, injustice, preventability and/or intentionality),

subjective labelling, physiological changes, and action tendencies

to engage in socially constructed and reinforced organized

behavioral scripts (original italics). ( Kassinove & Sukhodolsky,

1995, p. 7)

Anger, however, is not always based on cognitive and perceptual

distortions but may also reflect an accurate understanding of events.

Tavris (1989) stresses the social basis of all anger, except possibly

anger caused by organic abnormalities. Angry episodes 'assume meaning

only in terms of the social contract between participants. The beliefs we

have about anger, and the interpretations we give to the experience, are

as important to its understanding as anything intrinsic to the emotion

itself" (p. 19). She goes on to state that our "emotions may emerge and

differ primarily because of the situations in which they occur and because

of the interpretations that we give to our bodily states-psychological and

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

4

social matters, not solely biological ones" (p. 73).

Kassinove & Eckhardt (1995) describe a model of anger. There is

first of all an activating event that is perceived as negative. The same

event may cause anger in one person and happiness in another. At this

level, interventions using cognitive therapy are important.

According to this model, next to emerge is a state of arousal and

negative affect. They can be brought about by situational or environmental

forces but are also impacted by individual differences that predispose

people to become aroused in this way. Once the person is aroused,

memories and previously learned scripts for behavior are activated and

evaluated for appropriateness. The result can be positive coping (e.g.,

assertion) or negative coping (e.g., aggression). Positive coping leads to

anger control and negative coping leads to Anger-ln (suppressed) or

Anger-Out (expressed aggressively).

In addition, the experience of anger has been conceptualized

(Spielberger, 1996) as having two major components--state and trait.

State anger is an emotional state that varies in intensity and over time as

a result of the perceptions and frustration of the individual. Trait anger is

seen as “the disposition to perceive a wide range of situations as

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

5

annoying or frustrating and the tendency to respond to such situations

with more frequent elevations in state anger" (p. 1).

Spielberger (1996) also views the expression of anger has having

three components: Anger-Out (toward other people or objects), Anger-ln

(suppression of angry feelings), and Anger-Control (frequency witn which

the individual attempts to control the expression of anger).

A cognitive component is present in generally accepted definitions

of anger in the literature. Novaco (1995) states that, “The 'subjective

affect' element of anger is a cognitive labeling of the emotional state as

anger or something semantically proximate, such as 'annoyed'..."

(p. 321). Treatment programs for anger disorders often use cognitive

restructuring techniques (Deffenbacher, eta!., 1996; Novaco, 1977),

based on the assumption that cognitive distortions can occur in the

experience and expression of anger.

There are theories about which distortions or irrational thoughts

might be present (Hogg & Deffenbacher, 1986; Mizes, Morgan, & Buder,

1990; Thorpe, Parker, & Barnes, 1992). Thinking has centered on

distinguishing anger from other emotions. Beck (1976) suggests that

anger focuses externally and deals with thoughts of transgression. Many

R eprod u ced with perm ission o f the copyright ow ner. Furth er reproduction prohibited w ithout perm ission.

6

researchers have emphasized the importance of appraisal and attributions

(Fitness & Fletcher, 1993; Levine, 1996; Smith, Haynes, Lazarus, & Pope,

1993).

In summary, anger is defined (modifying the definition of Kassinove

& Sukhodolsky, 1995) as a feeling state that varies in intensity and is

associated with specific perceptions and cognitions, some degree of

physiological arousal, subjective labeling, and the tendency to engage in

socially constructed and reinforced behavior scripts. Anger can occur for

one episode (state) or in frequently repeated episodes over time (trait). It

can be expressed (aggressively or assertively), suppressed, or controlled

(modulated and regulated). For the purposes of this study, anger is

measured by the State-Trait Anger Expression Inventory (Spielberger,

1996).

Statement of the Problem

Cognitive theories state that emotions and behavior often follow

thoughts (Beck, 1963, 1976). Different patterns of behavior may result

from different ways of thinking. Yet, in the literature on anger research, the

few studies that have actually examined cognitive distortions have

generally compared them only to over-all scores of trait anger (Mizes,

R eprod u ced with perm ission o f the copyright ow ner. Furth er reproduction prohibited w ithout perm ission.

7

Morgan, & Buder, 1990; Muran, Kassinove, Ross, & Muran, 1989; Thorpe,

Parker, & Barnes, 1992).

In a review of the literature, no study has been found that

systematically examined the relationship between different categories of

cognitive distortions and the types of anger expression (Anger-Out, Anger-

in, and Anger-Control). That is what this study has done. Specific

hypotheses concerned the relationship of the cognitive distortions and

expressions of anger. Different behavior patterns may result from

different ways of thinking.

Kassinove & Eckhardt (1995) ask, "How can we better understand

the causes and consequences of anger to alleviate the suffering

associated with the felt experience and the expression of anger, including

Anger-Out or aggression?" (p. 204). Knowledge of which, if any, specific

cognitive distortions are frequently found in people who present different

types of anger expression can add to our understanding of anger and its

expression, clarifying some of the differences inherent in the cognitive

patterns underlying various behavior patterns.

The purpose of this study was to assist clinicians who are trying to

help clients change maladaptive angry behavior. They will gain information

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

8

that may allow them to be more specific about which cognitive distortions

to target or not target. "These empirically validated links would be

particularly helpful in designing group programs when individual

assessment is not readily available so that interventions can be focused

on those beliefs that have the greatest empirical relationship to anger"

(Hazaleus & Deffenbacher, 1985, p. 47).

Assumptions and Limitations

This study was based mainly on the results of self report

instruments. However, the procedure incorporated some behavioral

observation of anger expression by using the report of a live-in partner as

well as the subject's report. Also, confidentiality was maintained with

actual names and identifying information not recorded (except on the

consent form). There was, nevertheless, an assumption that the subjects

would be accurate in their observations and honest in their reports.

Partners were given instructions to complete their questionnaires

privately and place the finished scales in sealed envelopes. This was

done so that the scores would not be affected by apprehension among

subjects that their ratings would be seen by their partners. However,

questionnaires were completed at home by the subjects and it is possible

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

9

that directions were not always followed and that some ratings were

influenced by the fear that a partner would see the results.

Although an attempt was made to maintain heterogeneity in subject

selection, the actual demographic features of subjects depended to some

extent on subject availability. This may limit generalization of the results

and might result in some restriction of range (which would negatively

impact the correlation coefficients).

The instruments chosen have been selected because studies are

available attesting to their reliability and validity. However, there is less

information in this regard for the measures of cognitive distortion than for

the anger measurement scale.

Each questionnaire package contained a seif evaluation of anger, a

partner rating, and two tests of cognitive distortions. These packages were

arranged so that the order of the tests was randomized in an attempt to

avoid possible bias caused by the order of testing. However, once the

packages were distributed, only some were actually completed by

subjects. It is possible that in the packages actually used in this study the

tests were not in random order and that this may have affected responses.

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

Chapter 2

REVIEW OF THE LITERATURE

History of Anger Measurement and Research on Anger

Little research was conducted on anger until the last few decades

but anger is beginning to receive greater attention in psychology

(Deffenbacher, Oetting, et al., 1996). The contemporary literature on this

topic has grown dramatically, assisted by the development of more

precise definitions and more reliable and valid measurement instruments.

Nevertheless, many research findings continue to be either inconsistent or

unexplainable (Kassinove & Eckhardt, 1995).

Historically, in the literature, until recently, there has been

confusion concerning the definition of anger. In addition, anger has often

been confused with other concepts such as hostility and aggression. The

current definition of anger is given in chapter 1. It is distinguished from

hostility, which is generally defined as a set of negative evaluative

attitudes toward other people that often but not always include angry

feelings and that may motivate a person toward aggression (Spielberger,

1996). In contrast to anger and hostility, which refer to feelings and

attitudes, aggression implies actual behavior, usually destructive and

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

11

punitive in nature (Spielberger, 1996). Until recently these concepts were

often used interchangeably with that of anger, confusing the meaning of

research results.

Recently, anger has also been more clearly understood to be

experienced along a continuum of intensity and to be expressed in

different ways (Anger-ln, or suppressed anger; Anger-Out, or expressed

anger; Anger-Control, or regulated anger). It is seen as a state and also

as a trait. State anger "refers to a transitory emotional-physiological

condition consisting of subjective feelings of anger and activation of the

autonomic nervous system, either at a particular moment or over a short

period of time" (Deffenbacher, 1992, p. 177) and can vary in intensity and

fluctuate over time. Trait anger "refers to a relatively stable personality

dimension of anger proneness" (Deffenbacher, 1992, p. 177).

In this chapter we will look at the history of anger measurement and

the development of current views about anger. An overview of current

research will then show how this understanding of anger has been

applied. The cognitive view of anger will be discussed and information

given about research that focuses on anger and cognition. This last

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

12

section will also include an overview of the development of measures of

cognitive distortions.

The Development of Anger Measurement Instruments

Earlier Measures of Anger

"The earliest efforts to assess anger and hostility were based on

clinical interviews, behavioral observations, and projective techniques..."

(Spielberger & Sydeman, 1994, p. 300). In addition, early psychometric

measures often did not clearly differentiate the concept of anger from that

of hostility and/or aggression. Spielberger and associates have referred

collectively to the overlap in definitions and procedures used to assess

the constructs as the AHA Syndrome (i.e., anger, hostility, and

aggression) with anger being at the core. (Spielberger, Johnson, et al.,

1985; Spielberger, Krasner, & Soloman, 1988; Spielberger, Ritterband,

Sydeman, Reheiser, & Unger, 1995). Anger was often seen to be part of

aggression, which was the focus of many studies.

This problem appears to date back to several theoretical positions.

These include those of a) early psychodynamic theorists who believed

that instinctive aggressive drives were the primary motivator of all

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

13

destructive acts; b) ethnologists who supported this notion by postulating

a universal aggressive instinct that humans had immense difficulty

inhibiting; and c) logical positivists who eschewed the self-report of

internal constructs (e.g., anger, hostility) opting instead for their

observable equivalent (e.g., aggression) (E ckh a rd t, Barbour, & Stuart,

1997, p. 335).

In the 1950s a number of measures were developed to assess

hostility. Among them, the Buss-Durkee Hostility Inventory (BDHI) "is

generally considered the most carefully constructed measure of hostility”

(Spielberger, Ritterband, et al., 1995, p. 46.). In many studies the

physiological and behavioral aspects of anger and hostility were studied

but the subjective experience of angry feelings was not a focus of

research. Also, much of the work confounded angry feelings with the

direction of anger expression (Spielberger, et al., 1988; Spielberger &

Sydeman, 1994).

By the 1970s, however, several measures of anger (as opposed to

hostility) were constructed. Spielberger, Ritterband, etal., 1995, discuss

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

14

several of these: the Reaction Inventory (Rl) by Evans and Stangeland in

1971; the Anger Inventory (NI) by Novaco in 1975; the Anger Self-Report

(ASR) by Zelin, Adler, & Myerson in 1972. The Rl and the NI assessed the

degree to which anger was provoked in different situations. The ASR

assessed awareness of anger and anger expression modes. The NI has

been found to be significantly correlated with a number of other anger

measures (Huss, Leak, & Davis, 1993).

Spielberger, Ritterband, et al. (1995) state that two common

problems with the older measures of hostility and anger were that they did

not take into account the state-trait distinction and they confounded the

experience and expression of anger with situational factors.

Biaggio (1980) and Biaggio, Supplee, & Curtis (1981) examined the

BDHI, the Rl, the ASR, and the NI. They looked at the reliability,

concurrent and predictive validity, and correlates of the scales. They found

the BDHI and Rl showed good test-retest reliability, the ASR fair reliability,

and the NI poor reliability. Many ASR subscales showed good predictive

validity and the Rl and BDHI had some predictive validity. However, most

of the BDHI subscales did not measure specific states or behavior. The NI

did not correlate with any of the criterion measures. None of the scales

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

15

were thought to have proven research validity although more validity

studies were suggested. Sharkin (1988) states that the ASR apparently

confounds the expression of anger with aggression.

Other measures of anger were the Multidimensional Anger

Inventory (MAI) developed by Siegel (1985), the Subjective Anger Scale

(SAS) developed by Knight, Ross, Collins, & Paramenter (1985) and the

Awareness and Expression of Anger Indicator (AEAI) designed by

Catchlove & Braha (1985). The AEAI assesses discrepancies between a

person's awareness of angry feelings and the individual's expression of

anger. The SAS measures the affective experience of anger.

Sharkin (1988) states that the MAI may try to assess too many

measures of anger in one measure. On the other hand, according tc

Spielberger et al. (1985), the MAI does not appear to measure the

intensity of anger and only indirectly assesses the frequency of angry

feelings. Riley & Treiber (1989) state that the Anger-In/Brood subscale of

the MAI shows no relationship to anger suppression.

One other anger measure is the Framingham Anger In/Out and

Discuss Scales (Haynes, Feinleib, Levine, Scotch, & Kannel, 1978).

According to Riley & Treiber (1989), the Framingham Anger-Out scale

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

16

shows questionable validity, correlating with measures of anger

experience and hostility but not with measures of anger expression.

Development of the State-Trait Anaer Expression Inventory fSTAXH

The STAXI was developed by Charles D. Spielberger

(a) to provide a method of assessing components of anger that

could be used for detailed evaluations of normal and abnormal

personality, and

(b) to provide a means of measuring the contributions of various

components of anger to the development of medical

conditions...(Spielberger, 1996, p. 1)

The development of this scale emerged from earlier work by Spielberger

in the fields of personality assessment and also on the etiology of medical

disorders. During the course of his research, he realized that there was a

need for an instrument that was designed with a clear definition of anger

that distinguished it from hostility and aggression and that differentiated

state and trait anger. It later also became clear that mode of expression of

anger was also an important variable to be investigated (Spielberger,

1996). His instrument was systematically developed in stages to finally

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

17

reflect the components that he had identified as the AHA! Syndrome of

the anger emotion. Hostility was seen to be reflected in Trait Anger, and

aggression in Anger-Out.

Spielberger, who had previously constructed state and trait anxiety

scales, first developed a State-Trait-Anger Scale (STAS) in the early

1980s (Spielberger, 1996; Spielberger, Jacobs, Russell, & Crane, 1983;

Spielberger, Reheiser, & Sydeman, 1995). A pool of items was generated

to assess state and trait anger and administered to sample subjects.

Factor analysis of the S (State)-Anger items for a sample of 550 young

adults identified a single factor. However, factor analysis o f the T (Trait)-

Anger Scale indicated two factors. These were labeled Angry

Temperament (T-Anger/T) and Angry Reaction (T-Anger/R). The first

describes disposition to the experience of anger without provoking

circumstances and the second describes anger in situations that involve

frustration and/or negative evaluation.

Evidence supporting the state-trait distinction has been provided by

other researchers as well as Spielberger and associates (Deffenbacher,

1992; Deffenbacher, Oetting, etal., 1996).

In addition, evidence for the 2 factor Trait Anger score was

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

18

provided in a study in which T-Anger hypertensive patients were higher

than that of a control group of medical patients with normal blood

pressure. The difference was due entirely to higher T-Anger/R scores with

no difference found in T-Anger/T scores (Spielberger, Reheiser et al.,

1995)

Other research demonstrated good internal consistency,

convergent and divergent validity of the STAS and also that the STAS and

measures of hostility assessed related but different constructs

(Spielberger, Reheiser et al., 1995). Final items were also selected to

eliminate those with high correlation with anxiety measures (Spielberger,

1996).

The need for a measure of anger expression resulted from

research on medical illnesses which suggested that the way in which

anger is expressed is a critical variable (Spielberger, 1996). In fact, the

importance of modes of anger expression on the cardiovascular system

was shown by work in the 1950s (Funkenstein, King, & Drolette, 1954). In

that work students classified as showing high Anger-ln showed much

higher pulse rates than those who showed high Anger-Out. More work has

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

19

been done in this area (see section on anger research and medical

conditions).

Anger-ln is differentiated from anger that is repressed or denied. “In

contrast, suppressed anger is consciously experienced as an emotional

state, S-Anger, which may vary in intensity and fluctuate over time as a

function of the provoking circumstances' (Spielberger, e ta l., 1988, p. 94).

This research generally defined anger expression as a

unidimensional bipolar variable and Spielberger and his associates

attempted to construct an Anger-Expression (AX) Scale to measure it.

"Contrary to expectation, factor analyses indicated that the anger-

expression items were tapping independent Anger-ln (AX/ln) and Anger-

Out (AX/Out) factors, rather than a unidimensional, bipolar variable”

(Spielberger, Reheiser et al., 1995, p. 58). The loadings on the factors

were used to derive two scales to measure Anger-ln and Anger-Out.

Research has shown that the two scales "are empirically independent and

factorially orthogonal” (Spielberger, 1996, p. 58).

This scale was found to be psychometrically sound by other

researchers also. Knight, Chisholm, Paulin, & Waal-Manning (1988), for

example, reported a confirmatory factor analysis showing the

R eprod u ced with perm ission o f the copyright ow ner. Furth er reproduction prohibited w ithout perm ission.

20

independence of the Anger-ln and Anger-Out subscales and satisfactory

levels of reliability.

It has been found (Spielberger, Reheiser et al., 1995) that

individuals who are high in Angry Temperament are more likely to express

their anger outwardly than to suppress it; those high in Anger Reactivity

are equally likely to suppress or express anger. Individuals with high AX/ln

or AX/out experience anxiety more often than individuals with low scores

on those scales.

The final scale developed was Anger-Control. The first items used

were some control items that had been in the original AX scale and had

substantial loadings on both Anger/ln and Anger/Out (Spielberger,

Reheiser et al., 1995; Spielberger, 1996). Further items were written and,

following a factor analysis, those items with the highest loading on control

were retained. In a separate factor analysis of all three anger expression

scales, the control factor was the strongest to emerge and the three

factors were each defined by the items constituting their scales.

Although earlier versions of the measure also contained an Anger

Expression Scale (AX/ EX), this scale only combined expression out and

R eprod u ced with perm ission o f the copyright ow ner. Furth er reproduction prohibited w ithout perm ission.

21

Expression Scale (AX/ EX), this scale only combined expression out and

in. Modifications were made to adjust for Anger-Control (by subtracting

that score and adding a constant of 16 to prevent the possibility of

negative scores) (Spielberger, 1996).

Finally, in 1988, the STAS and the AX were combined to form the

State-Trait Anger Expression Inventory (STAXI) (see section on

instruments in methodology). Factor analysis of this instrument has shown

first a 6 factor solution and recently a 7 factor model (Forgays,

Spielberger, Ottaway, & Forgays, 1998; Fuqua, Leonard, Masters, Smith,

Campbell, & Fischer, 1991). The factors are State Anger, Trait Anger

(Temperament), Trait Anger (Reactivity), Anger-ln, Anger-Out, Anger-

control, and Feel-Like-Expressing-Anger (e.g., "I feel like breaking

things."). Fuqua, et al. (1991) report that "these results lend substantial

credibility to the multidimensional theoretical treatment of the anger

construct represented by the STAXI" (p. 445).

Spielberger is currently standardizing a STAXI-2 which was not

available in time for use in this study. T his new scale is essentially the

same as the old but some of the scales are expanded.

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

22

Overview of Current Research on Anger

In this section, an overview will be given of research on anger in

three areas: medical conditions, gender, and depression. The results of

these studies show the importance of differentiating high and low trait

anger as well as mode of anger expression in subjects because research

effects may be correlated specifically with a particular level of anger or

with a type of expression. It is also seen that anger interacts with other

variables, such as gender role and cognitive appraisals, when producing

an effect.

Research on Anaer and Medical Conditions

There is a growing literature on the effects of anger on medical

conditions and it is not possible to examine it all in this context. Some

important trends will be presented, focusing on anger and its affect on

blood pressure since much research has been completed on this subject.

It has often been shown in studies that blood pressure is higher in

individuals who suppress anger (Esler, et al., 1977; Gentry, Chesney,

Gary, Hall, & Harburg, 1982; Harburg, Blackelock, & Roeper, 1979).

According to Spielberger, Crane, Kearns, Pellegin, Bickman, &

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

23

Johnson (1991) hypertensive patients have been shown to have high

Trait Anger. They report experiencing more intense anger more frequently

than a normal control group and they tend to suppress anger rather than

express it (p. 279). According to Feshbach (1986) studies of

"hypertension reflect a consistent relationship between elevated blood

pressure and suppressed anger" (p. 129). Suls, Costa, & Wan, (1995), in

their meta analysis of studies looking at anger and blood pressure, report

that anger and systolic BP showed a small but significant positive

relationship which was mostly based on results of the Spielberger AX

Anger-ln scale and the Harburg Anger-ln/Anger-Out scale.

In addition to the work on blood pressure, other research has

looked at cardiovascular disease. Dembroski, MacDougall, Williams,

Haney, & Blumenthal (1985) have shown that Anger-ln is significantly and

positively associated with angiographically documented severity of

coronary atherosclerosis (CAD). The frequent experience of anger is one

important com ponent of the traits found to be associated with high

mortality rates in cardiovascular disease (Williams & Williams, 1993).

Frasure-Smith, Lesperance, & Talajic (1995) looked at the impact

of negative emotions on prognosis following myocardial infarction. They

R e p rod u ced with perm ission o f the copyright ow ner. F u rth e r reproduction prohibited w ithout perm ission.

24

found Anger-Out was unrelated but Anger-ln was associated with

increased risk for coronary arrhythmic events. They noted that it was

possible that the very angry (who expressed Anger-Out) did not survive

long enough to be studied or did not participate because of unwillingness

to discuss anger directly.

Deffenbacher (1992) reported that in his studies suppressed anger

was the best single predictor of all health variables except severity of

illness. He stated: "Assessing how people deal with their anger

contributes to a richer, more differentiated, understanding of the influence

of various components of anger on coping and health" (p. 190).

Although trends are apparent in the research on anger and blood

pressure and heart disease, there is also some contradiction in results.

For example, Spielberger (1996) reports the work of Boyle and Siegman

(1992) who found that Anger-Out correlated positively with systolic BP,

diastolic BP, and heart rate reactivity for participants in whom anger was

provoked. No correlation was found between cardiovascular reactivity

(CVR) and Anger-ln. Spicer & Chamberlain (1996) found in their research

that resting blood pressure was higher in men and women who were less

likely to suppress anger and in women who are more cynically hostile.

R eprod u ced with perm ission o f the copyright ow ner. Furth er reproduction prohibited w ithout perm ission.

25

There may be a number of reasons for this inconsistency. First, as

Spicer & Chamberlain (1996) stated, much of the previous research

confused the terms "hostility" and "anger." However, much of the more

recent work (including theirs) uses Spielberger’s anger measure, which,

as discussed elsewhere in this dissertation, is a reliable and valid

psychometric measure of anger as both a transitory emotional state and

as a relatively stable personality trait. Yet, some contradictory results are

still found.

A number of possible reasons have been suggested. Feshbach

(1986) states that what is most important is not whether anger is

expressed or suppressed but the frequency with which it is experienced.

Forgay, et al. (1998) state "It appears that the strength of the anger-CHD

link can vary by anger dimension and, at times, by gender" (p. 142).

Engebretson, Matthews, & Scheier (1989) designed a study to

clarify the inconsistent effects reported. Psychophysiological responses

varied as a function of whether or not provoked subjects were able to use

their preferred anger expression style. Those who were high in Anger-Out

had lower BP when allowed, after being provoked, to write a negative

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

26

evaluation of a confederate. Those who were high in Anger-in were lower

in BP when allowed to write a positive evaluation.

Gender effects have also been found in this area of research.

Faber & Burns (1996) found that following provocation systolic BP levels

remained elevated for high Anger-Out men (who expressed anger) while

low Anger-Out men showed a gradual recovery. However, high Anger-Out

women returned to baseline BP levels quickly while low Anger-Out women

look much longer to recover.

One study which attempted to reconcile inconsistent results.

Keinan, Ben-Zur, Zilka, and Carel (1992) found that intensity of anger

expression was negatively correlated with ill health while frequency of

anger expression was positively correlated with ill health. One possible

explanation is given for this. It may be that people who do not chronically

express anger but, when they do express it, do so firmly, may have the

biggest impact on their environment, more readily achieve their aims, and

feel more satisfaction. This results in lower risk for health problems.

These researchers did not find evidence that anger expression relates

differentially to different types of illnesses.

It was also found in this study that people with low intensity of

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

27

anger expression and low interest in social approval have a high level of

health problems. It is suggested that difficulties with emotional expression

and preoccupation with self (as opposed to others) might produce a high

level of arousal and thus cause health problems. It is noteworthy that this

study treats Anger-ln and Anger-Out as opposite ends of one continuum

in contrast to many studies (Spielberger, et al., 1985) that show they are

independent dimensions.

Tavris (1989) suggests that contradictory results in the area of

anger and hypertension may occur because we are looking for causes of

the disease when anger is actually a result in some cases. She proposes

that hypertension provides a non-specific physiological arousal and, given

a provocation, the person may then become angry more easily than would

have been the case without the prior physical arousal. When looking at

the relationship between anger and blood pressure, it is important to

remember that correlation does not necessarily mean causation. It is just

as possible that high blood pressure may be causing anger as opposed to

anger causing high blood pressure.

In addition to all the above research results, it is clear (Suls, et al.,

1995) that essential hypertension is a multicausal disorder. Anger

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

28

suppression or anger expression "may have to co-occur with other

attributes, such as a family history of hypertension or chronic stress"

(p. 454) for the disease to develop.

Anaer and Gender

Studies looking at the relationship between anger and gender have

produced different results. According to Sharkin (1993 & 1996), much

theory focuses on a common theme that women "are emotionally

expressive, with the exception of anger" (p.386). In contrast, "men are

generally viewed as emotionally inexpressive with the exception of anger"

(p.386). It is often believed that socialization processes led people to

believe that the expression of anger is more acceptable for men.

In a review of literature on the subject, Sharkin found few

differences between men and women. He does note the finding of

Deffenbacher, Oetting, et al. (1996) that women may be more prone to

associate other negative feelings, like shame, with anger. He also

believes that results are not conclusive. The theories that purport to find

gender differences are based on clinical work with clients. The research

showing no difference is often done with non-clinical populations and

uses self report. According to Sharkin (1993), the result is that theories

R eprod u ced with perm ission o f the copyright ow ner. Furth er reproduction prohibited w ithout perm ission.

29

that view anger as being a function of gender have not really been tested.

Averill (1983), using self reports, found that women becam e angry

as often as men, as intensely, for the same reasons, and they expressed

their anger as openly. The only major difference found was that women

reported crying more often.

Deffenbacher (1992) reported that his work showed consistently

that gender had very little impact on the correlates and consequences of

trait anger. He stated that men and women may differ in the expression of

anger but they were "similar in the frequency and intensity of their anger

experiences" (p. 192).

Work in this area has also been done by Kopper (1993), and

Kopper and Epperson (1991 & 1996). Using the Spielberger scales for

anger and the Bern Sex Role Inventory, they found no difference in

identified sex but did find differences for sex role. Femininity was

negatively correlated with Anger-Out and showed the greatest Anger-ln.

The male sex role was highest in trait anger and higher in Anger-Out. It

showed less Anger-ln and less Anger-Control. Those with androgynous

roles were lowest in Anger-ln. Undifferentiated had lowest scores in

Anger-Control.

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

30

On the other hand, important differences in gender for children and

adolescents were found by Clay. Hagglund, Kashani, and Frank (1996).

Their research showed that anger expression significantly predicted

sadness but not aggression for females and was important in predicting

aggression but not sadness in males. They used the Differential Emotions

Scale-Version IV to measure anger and depressed mood and the

Pediatric Anger Expression Scale (3rd ed.)

Spielberger, Reheiser, and Sydeman (1995) report results of STAXI

scores for a sample of 1010 young men and women. Males scored

significantly higher than fem ales on trait anger (due to higher scores on

the T-Anger/Temperament subscale). They were also significantly higher

in Anger-ln and Anger-Out (but more so in Anger-Out). No gender

differences were found on scales for state anger, T-Anger/Reaction or

Anger-Control measures. Differences in female and male patterns of

answers to specific questions s u g g e s t" a stronger disposition for men to

express their angry feelings in physically or verbally aggressive behavior"

(p. 62).

More recently Spielberger and associates (Forgays, et al., 1998)

report gender differences in response patterns on the newly identified

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

31

State Anger and Feel-Like-Expressing Anger subscales of the State Anger

Scale. The latter subscale is considered to represent "a cognitive

dimension of the angry feeling, a moderator between the anger

experience and angry behavior” (p. 153). It was found to be present in

middle age men and middle age women but only young women and not

young men. These researchers believe it may represent women's

knowledge that acting on anger has great social costs for them which men

do not face.

Thomas (1993) did not find gender differences in Anger-ln or out

but found that women had more anger related physical symptoms and that

trait anger was strongly related to perceived stress. Self-esteem was a

modifying factor for women as females with high self-esteem had a

decreased tendency to become angry. Low self-esteem was related both

to venting anger and suppressing it. Thomas states that "it makes sense

to us that women's anger would have unique aspects. Women and men

are biologically different and socialized differently as well..." (p. 15).

As Tavris (1989) states “of course there are differences between

the sexes; it is just that the differences have less to do with innate or even

learned ways of expressing anger and more to do with the intricate

R eprod u ced with perm ission o f the copyright ow ner. Furth er reproduction prohibited w ithout perm ission.

32

differences between their cultures and their circumstances" (p. 219).

According to this author, men and women "do not always feel angry about

the same thing" (p. 210).

Anger and Depression

It has often been theorized that depression is related to anger

turned inward (Clay, Anderson, & Dixon, 1993; Riley, Treiber, & Woods

(1989). In the literature that examines this theory, as in other areas of

anger research, anger is sometimes used interchangeably with the

concept of hostility. One study, by Moore and Paolillo (1984), used the

Buss-Durkee Hostility Inventory with measures of depression and

hopelessness. It was found that 50% of the variance of depression was

explained by hopelessness but of the remaining variance, covert hostility

accounted for 7%. Covert hostility was determined by scores for that

variable on the BDHI and is not further defined. However, the authors cite

the Abraham-Freud psychoanalytic theory of depression "in which

depression is viewed as the inward turning of aggressive instinct which is

not directed at the appropriate object” (p. 880). Subjects were outpatients

at a mental health center.

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

33

Biaggio and Goodwin (1987) used measures of hostility and anger

(including the Spielberger Anger Expression Inventory) and found a

correlation between depression and inwardly directed hostility but not

Anger-ln (which actually correlated inversely with depression).

However, Clay, Anderson, and Dixon (1993) found in their study of

anger (cited earlier) that stressful life events and anger directed inward

each were significantly related to depression. In a hierarchical multiple

regression analysis each added to the prediction of depression

independently. Anger-ln (measured by Spielberger’s Anger Expression

Scale) accounted for 10% of the variance above and beyond the stressful

events. Anger-ln and stressful events were uncorrelated and there was no

significant interaction effect between these two variables. It was found that

Anger-Out and Anger-Control were not significantly related to depression.

However, this study was correlational in nature so cause and effect

relationships cannot be determined and gender effects were not

examined. Also, subjects were college students who reported some

depressive symptoms but who were non-depressed for the most part.

Riley, Treiber, and Woods (1989) looked at the relationship

between depression and anger using multiple measures of anger and

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

34

hostility and comparing results with those of two other groups, normal

subjects and a sample of people diagnosed with Posttraumatic Stress

Disorder. Although many instruments were used, only one actually

measured Anger-ln (a subscale of the Multidimensional Anger Inventory).

However, this scale did show a significant difference between the groups

with the depressed group showing more suppressed anger. On most of

the measures the depressed sample reported lower or equivalent levels of

anger expression compared to the normal group and the PTSD group

expressed the most anger.

The authors note that gender differences were not considered and

may have contributed to the results. Also, results are correlational.

Results also showed that it is possible that level of anger is important with

moderate anger being suppressed by depressed clients and severe anger

being expressed.

More recently Kopper and Epperson (1996), using a sample of

male and female college students and a number of measures including

Spielbergeris Anger Expression Scale, found a significant positive

correlation between anger suppression (which is Anger-ln) and

depression. This was true for both men and women. "Gender did not

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

35

uniquely contribute to the prediction of any of the mental health variables

(examined in their research), nor did it moderate the relationship of these

variables with other predictors" (p. 158). However, others (see section of

anger and gender) have found gender to be a modifying factor in anger

and depression.

Others have challenged the Anger-ln/depression hypothesis.

Robbins and Tanck (1997) report that work done "suggests that anger is

related to depressed affect though much about the relation remains

unclear" (p. 490). They state that other variables, such as source of anger

and context may affect the relationship between anger and depression.

In their work, Robbins and Tanck (1997) found a positive

relationship between anger and depressed affect. There was also a

positive correlation between anger suppression and depression. However,

their data showed that this might reflect trait anger as opposed to the

general level of felt anger. Findings suggested that expression of anger

was associated with less depressed affect. In addition, individuals who

directed anger at the self and had higher levels of depressed affect than

individuals who did not direct anger inward.

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

36

Thomas (1993) cites studies that found that depressed women

expressed more anger in interpersonal relationships than did the control

subjects (Folkman & Lazarus, 1980) and women who ventilated their

anger were more depressed than those who did not (Weissman & Paykel,

1974). Eckhardt and Deffenbacher (1995) state that the hypothesis of

depression resulting from self directed anger has received little empirical

support (although they cite only Beck, 1963). However they state that “it is

often the case that individuals experience anger and depression

simultaneously" but "it is an oversimplification to assume a linear

relationship between such complex emotional experiences" (p. 36).

Fava, et al. (1993) showed that a high percentage of depressed

outpatients (21%) also met the criteria for anger attacks. Their self report

questionnaire does not rate subjective anger but rather physiological

arousal and aggression. Hostility decreased after treatment with

antidepressant medication.

Goldman and Haaga (1995) state that observational research on

married couples shows that depressed people “express more anger

toward their spouses than do non-depressed people" (p. 505). They state

that this discrepancy might result because of the specific group being

R eprod u ced with perm ission o f the copyright ow ner. Furth er reproduction prohibited w ithout perm ission.

37

examined (marital couples) or it might derive from the observational

nature of data collection as opposed to self report often used elsewhere.

In research designed to evaluate these possibilities, Goldman and

Haaga (1995) found that depressed subjects scored significantly higher on

self report measures of anger and anger suppression (Spielberger’s

STAXI). This was true whether the target of the anger was the spouse or

others. Therefore, the differences in results seen previously are not due to

the group sampled (spouses) but may be due to the use of behavioral

observation instead of a self rating scale.

In addition, Goldman & Haaga (1995) asked the depressed patients

why they suppressed anger. The interviews suggested 4 main themes:

fear of retaliation, of rejection, of hurting others, and of the futility of

expression.

Development of the Cognitive Theory of Anger

According to Novaco (1979), the idea that anger arousal results

from cognitive appraisal is an old one. It was found, for example, in the

Roman Stoic philosophers, Marcus Aurelius and Epictetus. For William

James and Carl Lange, "emotions were mental states resulting from the

perception or consciousness of physiological manifestations that had

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

38

been excited by external events. Emotion was the feeling of bodily

changes" (p. 255). However, Novaco adds that James did note in 1890

that anger had cognitive determents as he recognized that in thinking over

one's insult one m ight become angrier than when originally receiving it.

G. Stanley Hall (1899), studied anger and attempted to shift the

analysis from philosophical to empirical inquiry. His account of anger

includes personal appraisals of provoking circumstances taken from 2.184

questionnaires asking for detailed information on the experience of anger.

Many categories of anger were found but the strongest anger developed in

situations when one person was mistreated by another.

Novaco (1979) states "The interests of experimental psychology in

observable events resuited in a preference for the study of aggressive

behavior over the emotion of anger" (p.255). From this work came the

frustration-aggression hypothesis (Miller, Sears, Mowrer, Doob, & Dollard,

1941). This theory stated that frustration often leads to aggression and

where there is aggression there is likely also frustration. It did not

emphasize "anger" but did use the term “instigation" and there were

cognitive elements in the instigation process.

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

39

Arnold (1960) is considered by Thomas (1993) to be "the first

modern theorist to propose that appraisals determine the particular

emotion that a person will experience" (p. 28). Arnold believed that anger

arises when a threat is perceived and it is seen as difficult to overcome.

Berkowitz (1990), using the term cognitive-neoassociationistic for

his model, integrated arousal and cognitive concepts like appraisal. He

stated that anger can "be affected by both the associative linkages

connecting negative affect with aggression-related ideas, memories, and

expressive-motor reactions and also by people's schemas regarding the

nature of anger" (P. 494). Appraisals and attributions intensify or weaken

the anger experience.

Bandura (1973) developed the "social learning theory of

aggression in which cognitive appraisal mechanisms have an important

role in the arousal of anger" (Novaco, 1979, p. 257). Beck (1963,1976)

and Ellis (1977) as well as Ellis and Tafrate (1998) have placed great

emphasis on the importance of cognitions in the role of emotions,

including anger.

Novaco and Welsh (1989) comment that contemporary ideas about

the cognitive mediation of anger "is too often restricted to an event-

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

40

interpretation tandem process. This misses the automaticity of cognitive

appraisal as embedded in the perception of events..." (p. 40). From this it

follows that cognitive distortions already present in a person’s thinking will

impact on the development of the experience of anger.

Research on Anger and Cognition

Anger is viewed as having several components: emotion,

physiological arousal, and cognition. Novoco (1977) states "At the

cognitive level, anger is a function of appraisals, attributions,

expectations, and self-statements that occur in the context of provocation"

(p. 600). This section will focus on the cognitive aspect of anger, looking

first at research that attempts to assess which cognitions are associated

with anger as opposed to other emotions. Then, work that looks at anger,

aggression and cognitions will be addressed. Finally, we will look at

previous research on specific cognitive distortions/irrational thoughts and

anger.

Research on Appraisals. Attributions. Expectancies

Research that has been carried out to determine which cognitions

are specifically associated with each emotion has focused on appraisals,

attributions, and expectancies. The more global category is that of

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

41

appraisal. Eckhardt and Deffenbacher (1995) discuss primary and

secondary appraisal in connections with anger. Primary appraisal

"involves an evaluation of an event and its context in terms of its potential

relevance and threat or harm value" (p. 32). On the other hand, secondary

appraisal "involves a judgement about the person’s coping capacities" (p.

32). "Anger has been found to be associated with the belief that

something can be done to reinstate one's goal...or with a sense of

power...” (Levine, 1996, p. 341). It is the combination of arousal and

interpretation that determines how the person labels the feeling. Appraisal

has been part of the theories of Berkowitz (1990), Bandura (Bandura,

Adams, & Berger, 1977), and Beck (1976) among others.

Attributions are the appraisals that are concerned with agency.

Much research has shown that people are angry when they believe

someone else harmed them intentionally or through negligence (Averill,

1982; Levine, 1996; Novaco, 1979). Thus the cause of harm is seen to be

external and controllable.

Novaco (1979) states that less attention has been paid to

expectations. "Expectations, as subjective probabilities about events, are

based on previous appraisals of related circumstances. Appraisals are a

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

42

function of the expectations one holds regarding oneself and others”

(p.253). Novaco's model states that there are three types of expectations

that can affect anger. An expectation that does not materialize can

produce aversive arousal that is experienced as anger. An expectation of

aversive events can lead to selective attention to negative cues. Also,

when anger is expected to be instrumental in obtaining something, that

expectation can lead to anger.

There have been many attempts in research to assess the most

relevant categories of appraisal that are associated with anger. W ork by

Weiner, Graham and Chandler (1982) and extended by McAuley and

Shaffer (1993) investigated anger as a function of causal cognitions

(locus, stability, and controllability). Anger was greater when the cause

was perceived as controllable, stable and internal to another individual

and uncontrollable by the subject.

According to Averill (1982), the major issue for people in the

development of anger is "the perceived justification for the instigator’s

behavior" (p. 1149). In his research, over 85% of episodes described by

angry people involved "either an act that they considered voluntary and

unjustified (59%), or else a potentially avoidable accident (e.g., due to

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

43

negligence or lack of foresight, 28%)" (p. 1150). He states that "anger is a

response to some perceived misdeed" (p. 1150). It involves a value

judgement and an attribution of blame.

Using factor analysis and multidimensional scaling, Smith &

Ellsworth (1985) identified cognitive dimensions associated with different

emotions. Anger was seen by subjects as associated with

unpleasantness, anticipated effort, a fair degree of certainly about the

situation, and strong attributions of human agency and other-

responsibility/control.

Wickless & Kirsch (1988) had undergraduates record their thoughts

when feeling angry. They found that, in line with Beck's theory, anger is

associated more with thoughts of transgression (attribution of harm to

them by others) than loss and more with loss than threat. But these

thoughts did tend to occur in combination.

The effects of nine event dimensions on self-reported anger were

investigated by Ben-Zur and Breznitz (1991). Their results showed level of

damage (appraisal of harm) to be the most important but intentionality and

preventability of the damage (attribution of agency) also consistently

increased anger. Also involved were the expectation that damage would

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

44

occur and the ability to correct it. The authors found three main

factors: extent of damage (combining damage and correctability), causes

of damage (combining intentionality and preventability) and likelihood of

damage occurrence (expectedness and agent).

Fitness and Fletcher (1993), examining the appraisals of subjects

angry with their spouses along a number of dimensions, found that angry

partners reported believing they had been treated unfairly (appraisal of

harm). Anger events were appraised as predictable and the spouses' fault

(attribution of agency). The subjects did believe that they themselves had

more control over the situation than when other emotions were expressed

(secondary appraisal).

Smith, Haynes, Lazarus and Pope (1993) found that appraisal

serves as a mediator between attributions and emotional responses,

including anger. They also found that external attributions and

controllability were associated with anger. The core relational them e for

anger was "other blame" (appraisal or attribution of other accountability).

Levine (1996) looked at the extent to which appraisal models

accurately capture responses to a naturally occurring event (in this case

Perot's withdrawal from the 1992 presidential campaign). This research

R eprod u ced with perm ission o f the copyright ow ner. Furth er reproduction prohibited w ithout perm ission.

45

evaluated the cognitive processes associated with anger and found the

importance of the dimensions of agency (attribution) and coping potential

(secondary appraisal) in differentiating anger from sadness during the

experience of goal failure. Angry subjects were likely to blame another

person (Perot) and also to believe that desired substitute goals could still

be obtained.

In a somewhat different approach, Smucker, Martin and Wilt (1993)

assess values and anger. They found that women who scored higher in

angry temperament placed more value on social recognition (respect,

admiration by others) and women who were low on angry temperament

were higher on self respect. It is suggested that perhaps the angry women

do not receive the admiration they seek. Other explanations are that

women with higher self esteem can better control their emotions or that

the display of anger may be considered demeaning and lower self

respect. In this research it was also found that women having difficulties

enacting their values were more likely to suppress their anger. Also,

women who felt unsupported had high anger scores.

In summary, research has shown that anger is related to cognitive

appraisals of the potential relevance or harm of the event, the

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

46

intentionality or preventability of the event, and the belief that a goal can

be re-instated. There is often an attribution of blame to another person.

The situation is seen as controllable and stable. There may be

expectations that do not materialize or that produce selective attention to

aversive stimuli.

Research on Anaer. Aggression, and Cognitions

Work on the cognitive variables involved with anger has also been

done in the area of aggression. Kassinove and Sukhodolsky (1995)

discuss the difference between anger and aggression. They state that

aggression "refers to motor behavior carried out with the intent to hurt

someone through physical contact" (p. 11). They go on to state that

aggression can be instrumental (carried out for an extrinsic purpose) or

emotional (with the intent to attack someone with whom one is angry).

Instrumental aggression may occur without anger. Anger may also occur

without aggression.

Averill (1983) states that his research showed that most anger

episodes are not aggressive and only 10% of the episodes reported by

subjects involved physical aggression or punishment. However, at times

the two overlap.

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

47

There are other slightly different definitions of aggression. It has

been viewed as destructive behavior directed at other people or objects

(Spielberger, Jacobs, Russell, & Crane, 1983; Spielberger, Reheiser, &

Sydeman, 1995). Aggression must include actual behavior whether

motoric or verbal and there must be intent to harm. Actual physical

contact is not necessary.

It appears that a crucial element in the development of aggression

is the accompanying cognition of the aggressor. Current research has

attempted to develop a more comprehensive analysis of the thoughts that

are associated with aggression. "Contemporary theories of aggression

emphasize the role of social cognitive processes in mediating the

relationship between environmental stressors and aggressive

interpersonal behavior" (Fondacaro & Heller, 1990, p. 76). W hile it is

beyond the scope of this work to outline fully the research that has looked

at hostile attributions, aggression and anger, some examples and trends

in findings will be highlighted.

Much work has focused on the association between hostile

interpersonal attributions and aggressive behavior. Numerous studies in

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

48

the past decade have demonstrated that children with hostile attributional

tendencies are likely to experience behavioral difficulties in interacting

with peers. Aggressive children display a hostile attributional bias: They

are up to 50% more likely than average children to attribute hostile intent

to a hypothetical peer after an ambiguous provocation by the peer

(Dodge, Price, Bachorowski, & Newman, 1990, p. 385).

Fodacaro and Heller, 1990, state that this type of cognition "may

fuel anger arousal that energizes the enactment of retaliatory responses"

(p. 77). They investigated attributional style in three groups of adolescent

boys, nonoffenders, nonaggressive offenders and aggressive offenders.

Results showed that both aggressive and nonaggressive offenders were

more likely than nonoffenders to blame others for difficulties in ambiguous

interpersonal problem situations. In addition, however, the degree of

aggressiveness in behavior increased as a function of the strength of

other-blame attributions. There were no significant differences in

nonambiguous situations.

Similar studies have been conducted by many others. Slaby and

Guerra, 1988, examined cognitive mediators of male and female

adolescents. They found anti-social aggressive individuals were more

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

49

likely to define social problems in hostile ways, adopt hostile goals, hold a

set of beliefs supporting the use of aggression.

Dodge and others have completed important work that examines

the social-cognitive difficulties of aggressive children and adolescents. In

one study (Dodge, et al., 1990) incarcerated juvenile offenders were

administered a task to assess hostile attributional biases. The biases were

positively correlated with undersocialized aggressive conduct disorder,

with reactive-aggressive (angry) behavior, and with the number of

interpersonally violent crimes committed. There was no relationship

between hostile attributional bias and nonviolent crimes or socialized

aggressive behavior disorder. The findings held even when intelligence,

SES and race were statistically controlled. The authors state that most

previous studies examined less disturbed younger children selected from

Education populations. The correlational nature of the work prevents

conclusions related to causation.

Another study (Lochman & Dodge, 1994) examined social cognition

in a cross sectional sample of boys outside of a correctional institution.

There were three levels of behavior (severely aggressive, moderately

aggressive and nonaggressive) and two age levels (preadolescent and

R eprod u ced with perm ission o f the copyright ow ner. Furth er reproduction prohibited w ithout perm ission.

50

adolescent). A large number of information processing and schema

variables were analyzed. Results showed that a wide variety of social-

cognitive processes are distorted and deficient for violent and moderately

aggressive children. "Attributional biases, problem-solving deficiencies,

outcome evaluations favoring aggression, and low perceived self-worth

emerge as key characteristics of aggressive children, and the degree of

these difficulties is directly related to the severity of the aggressive

behavior problems" (Lochman & Dodge, 1994, p. 372).

Specific Studies of Anger and Cognitive Distortions

Appraisals, attributions, and expectancies may be accurate, formed

as a result of an empirical review of the evidence. They may also be

inaccurate or only partially accurate and based on cognitive distortions or

poor logic. In this section we will look at work that assesses the

relationship between anger and cognitions which are distortions or

irrational and also at instruments used for the assessment of distortions

The measurement of cognitive distortions

A number of instruments have been developed to assess irrational

thinking/cognitive distortions. Information about the Survey of Personal

Beliefs and the Common Beliefs Survey (used in this research) can be

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

51

found in the methodology section. Some additional information about

those scales will be presented here and a number of other assessment

tools will be discussed as well.

Much research investigating irrational thinking and affect uses the

Irrational Beliefs Test (IBT) (Jones. 1969). This is a 100-item self report

inventory which uses a 5-point scale of level of endorsement. There are

10 subscales corresponding to Ellis' early categories of irrational beliefs.

Jones (1969) reported interna! consistency reliability of items ranging from

.66 to .80. Test-retest reliability over 24 hours were .68 to .87 for

subscales and .92 for the total inventory. Trexler and Karst (1973)

reported further test-retest reliability. Lohr & Bonge (1982) replicated

Jones' original factor analysis, modifying the factor structure and scoring

system.

However, the validity of this instrument has been questioned (Smith

& Zurawski, 1983; Zurawski & Smith, 1987). They offer evidence that total

scores from this instrument can not be differentiated from many measures

of emotional distress. Both types of measures assess the more general

dimension of negative reactivity. The authors state that research using

this scale has questionable results. According to Ramanaiah, Heerboth

R eprod u ced with perm ission o f the copyright ow ner. Furth er reproduction prohibited w ithout perm ission.

52

and Schill (1987), 50% of the items on the IBT are not stated as beliefs

and 30% of the items are not clearly stated as rational or irrational beliefs.

The Rational Behavior Inventory , developed by Shorkey and

Whiteman in 1977, contains 37 irrational statements with 5 point rating

scales.lt has been criticized on the same grounds as the Jones Irrational

Belief Test (Wertheim & Poulakis, 1992; Zurawski & Smith, 1987). In

addition, these scales have been described as outdated as they derive

from Ellis' earlier writings that have since been revised (Wertheim &

Poulakis, 1992). Ramanaiah et al. (1987) report that 50% of the items are

not stated as beliefs.

In order to solve some of these problems some newer measures

have been developed. The General Attitude and Belief Scale (GABS),

later called the Attitudes and Beliefs Scale II (A&BS-II), was developed by

DiGiuseppe and others in 1988. According to one study (Wertheim &

Poulakis, 1992) the GABS shows good internal consistency and construct

validity.

Also, the Irrational Belief Scale (IBS) was designed by Malouff and

Schutte, who claimed evidence for good internal consistency, test-retest

reliability, and construct validity (1986). They stated that this scale asked

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

53

only about irrational beliefs, in contrast to other instruments of cognitive

distortions that actually asked questions about anxiety reactions. Robb

and Warren (1990) evaluated a number of tests of irrational beliefs. They

found that the Irrational Belief Scale emphasized “facts of life" but

neglected "rules for living". This last category would be represented by a

statement such as "The world MUST treat me fairly" (p. 308).

Two other measures examined have been used to investigate

depression. The Cognitive Error Questionnaire (CEQ) (Lefebvre, 1981) is

a test which measures frequency of information processing errors which

lead to inaccurate interpretation and vulnerability to depression. One other

measure is the Dysfunctional Attitude Scale (DAS) developed by

Weissman in 1979 to assess the presence of beliefs postulated by Beck

to characterize the thinking of depressed people.

In their analysis, Robb and Warren (1990) state that the IBS, the

A&BS-II and the DAS contain only statements of belief. However, the

tests varied in content, with most containing mostly "facts of life" (how the

world works) questions as opposed to also including "rules for living"

(which often contain words like "should"). Of the measures examined only

the A&BS-II had both types of questions.

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

54

Howard Kassinove (1986) and Andrew Berger first developed the

Survey of Personal Beliefs in 1981 to assess irrational thinking according

to Rational-Emotive personality theory. In this theory "the experience of

distress is thought to be sharply heightened by adherence to a number of

rather specific beliefs or 'irrational ideas' which depart from logical and

reality based assumptions and which lead to self-defeating behavior" (p.

119). At that time an earlier list of 11 core irrational ideas had been

reduced to 4 categories: Awfulizing, ideas reflecting low frustration

tolerance, ideas reflecting demanding shoulds, oughts, and musts, and

self worth ideas that represent overgeneralizations.

Although a number of assessment instruments had been previously

developed for the same purpose. Kassinove (1986) wished to make a tool

that would be free of some of the common errors found in predecessors.

These problems “included insufficient demonstrations of reliability and

validity, lack of normative data, difficulty in distinguishing cognitive and

philosophical change from simply learning a rational vocabulary, etc" (p.

121). Perhaps the most important mistake in previous scales was the

wording of questions to include not only an irrational belief but also an

emotion and/or action. For example, one such sentence stated "I get

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

55

upset and angry when my plans go wrong". This includes an emotion as

well as an event and no actual belief. Kassinove stated that flaws in

measures used affected the validity of prior studies.

A slight revision to the scale was made by Berger, Caerula and

Kassinove in 1982 ( Kassinove & Berger, 1994). At that time the

questions, originally meant for parents, were made into a more general

form. The latest version (1994) has reduced the number of questions from

.60 to 50 and one subscale (called basic principles) was dropped as factor

analysis showed it to be unreliable. Kassinove and Berger (1994) have

provided some normative data (see Appendix B).

In addition to the studies cited in the methodology section,

Kassinove (1986) used this instrument to obtain moderate correlations

with neuroticism and self-reported negative affect. There was a

nonsignificant association with positive affect. These results were in

support of RET theory. Kassinove and Eckhardt (1994) used the

instrument in a cross cultural study to obtain a moderate overall

relationship of rationality to self-reported affect. (See section of specific

cognitions and anger.)

Information on the Common Beliefs Survey III (Bessai 1977;

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

56

revised 1997 and 1998) is provided in the methodology section. In

addition, Thorpe and Frey (1996) report on a short form which they

developed.

Studies of anger and cognitive distortions/irrational thinking

A few studies have been found which looked at the relationship

between trait anger and irrational thinking in general, without specifying

irrational beliefs. For example, W oods (1987) gave stress management

workshops to corporation employees that focused on reducing irrational

thinking. He obtained pre-measures and post-measures of a number of

traits including anger (Spielberger's State-Trait Anger Scale was used)

and irrational beliefs (using Jones' Irrational Beliefs Inventory). Although

this did not prove to be a very angry group of subjects, there was still a

significant decline in anger scores in relationship to a decline in irrational

beliefs.

Many studies have looked at overall anger in relationship to specific

cognitive distortions. Most have used the Novaco Anger Scale (NAS) that

is a self report measure of anger proneness and which has been criticized

(see earlier section on anger measurement). Several used the Jones

Irrational Beliefs Test (IBT), which has also been criticized (see section on

R e prod u ced with perm ission o f the copyright owner. F urth er reproduction prohibited w ithout perm ission.

57

measures of cognitive distortions). Zwemer and Deffenbacher (1984) used

college students as subjects and the NAS and IBT and found that

Personal Perfection, Anxious Overconcem, Blame Proneness, and

Catastophizing were predictors of general anger and high anger. Beliefs

were even more relevant for the more angry. Findings were applicable to

both sexes.

Hazaleus and Deffenbacher (1985), also using college students

and the NAS and IBT found anger to be generally related to endorsement

of irrational beliefs. However, regression analysis showed that only

Anxious Overconcern and Blame Proneness were significant predictors of

anger arousal. However, Blame Proneness was a factor only when anger

was moderate. There was no difference on the measure of anger between

results of men and women. The authors suggest that, since work has

often shown that men are more likely to be aggressive than women,

"similar anger levels of the sexes suggest that women may be more likely

to follow anger arousal with behaviors other than aggression" (p. 51),

possibly turning it in and becoming depressed. An interesting finding was

that women high in anger tended to be less likely to endorse the need to

be dependent on others.

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

58

Hogg and Deffenbacher (1986) again using students and the same

measures as used above, found Catastophizing, Personal Perfection, and

Demand for Approval predicted anger. These occur when people set

unrealistic demands on themselves and others and the demands are not

met. Blame Proneness was not a factor in this study.

Story and Deffenbacher (1985) as cited in Deffenbacher (1992)

administered the Trait Anger scale of Spielberger's State-Trait Anger

Scale (STAS) and the Irrational Beliefs Test to students. They found that

the irrational beliefs of Catastrophizing, Need for Personal Perfection,

Blame Proneness, and Dependency entered a stepwise regression in that

order for predicting Trait Anger scores. Failure and the belief that one

must be perfect appeared to place unrealistic pressure on individuals,

according to the authors. Blame Proneness and Dependency appeared to

increase frustration.

Lohr and Bonge (1982) developed scoring keys for the Irrational

Beliefs Test. In a study using this scoring method, Lohr, Hamberger and

Bonge (1988) used the IBT and the NAS and found sex differences. All

the NAS subscales and total score showed a significant association with

Anxious Overconcem for both genders. For males, Demand for Approval

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

59

accounted for significant increments in variance (above Anxious

Overconcern) on 3 of 6 anger scales. For females, the same was true of

Blame Proneness on 5 of 6 scales. Dependency and Problem Avoidance

were negatively correlated with anger proneness. The authors stated

"Perhaps feminine sex-role socialization involves the acquisition of

irrational beliefs that serve to suppress anger expression because it may

be perceived as gender-inappropriate" (p. 182). They also reported that

the association between irrational beliefs and anger proneness is not

large, accounting for less than 20% of the variance on any of the anger

scales.

Mizes, Morgan and Buder (1990) criticize earlier work that looked

at cognitions and anger using a scoring system of the Irrational Beliefs

Test by Jones (1969) that has been questioned. Mizes, et al. used the

same test with the newer scoring version by Lohr and Bonge, as well as

the Rational Behavior Inventory and the Novaco Provocation Inventory.

They found that irrational cognitions were correlated with anger. They

failed to find anger related to assertion specific cognitions. They did

find modest but significant correlations between overall anger

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

60

and Frustration, Guilt, Blame Proneness, Problem Avoidance, Negative

Evaluation, Projected Misfortune, and Self Control.

Other tests of irrational beliefs have also been employed. Zwerdling

and Thorpe (1987) used the NAS and a number of measures including

The Common Beliefs Survey with undergraduate students. High and

moderate anger groups scored significantly higher than low anger subjects

on the measure of irrational thinking. High anger subjects scored higher

than low anger subjects on a number of traits including hostility and

criticism of themselves and others.

Ford (1991) used prison inmates with a history of violence as

subjects. He found that irrational beliefs, as measured by the Belief Scale

(Malouff & Schutte, 1986), correlated significantly with Trait Anger, Angry

Temperament, and interpersonal provocability, using Spielbergeris State-

Trait Personality (Anger) Inventory. This was true even when anxiety was

partialled out.

Thorpe, Parker and Barnes (1992) looked at the Common Beliefs

Survey III and the Novaco Anger Scale. Again, only total anger scores

were used. The total anger score correlated significantly with several

types of cognitive distortions: total CBS (.26); Importance of History (.26);

R eprod u ced with perm ission o f the copyright ow ner. Furth er reproduction prohibited w ithout perm ission.

61

Self Downing (.28); Need for Approval (.19); and Perfectionism (.26).

There was no significant correlation with Blaming or Control of Emotions.

Kassinove and Eckhardt (1994) tested college students in both

America and Russia to investigate levels of irrational thinking using the

Survey of Personal Beliefs and self-reported negative and positive affects

(using the Multiple Affect Adjective Checklist-Revised). Their results

showed a small but significant inverse association between anger and

total rationality. There were significant relationships between anger and

both Awfulizing and Low Frustration Tolerance. Russian students reported

more rational beliefs and less negative affectivity than Americans and

various explanatory hypotheses are discussed but the cross cultural

nature of the research may have had an impact on results. Subjects were

all students. In addition, the measure of emotion was not one specific to

anger.

Woods and Coggin ((1985) used the IBT and the trait scale of

Spielberger's State-Trait Anger Scale to see which irrational beliefs would

best distinguish high anger from low anger groups. They found High Self

Expectations, Anxious Overconcern, Frustration Reaction and Demand for

Approval to be most strongly positively correlated with level of anger.

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

62

Muran, Kassinove, Ross and Muran (1989) used the Survey of

Personal Beliefs and Spielberger's State Trait Anger Inventory to look at

irrational thoughts and anger. They compared a student and a clinical

population. There was a significant age difference in the two groups with

the clinical group being on average 10 years older. The clinical subjects

scored significantly higher on many measures but no significant difference

between the populations was found for anger. The SPB produced only

one significant difference with the student population scoring higher in

tolerance for frustration. Total irrational beliefs was significantly related to

Trait Anger and Low Frustration Tolerance was related to Trait Anger.

Other subscales were not related to anger. This study did not look at the

expression of anger, only Trait Anger.

Only a few studies have looked at anger expression and irrational

cognitions and they did not do so in a manner that systematically

examined anger expression and various irrational beliefs. Lopez and

Thuman (1986) investigated how high angry students differed from low

angry students in their endorsement of irrational beliefs and in the manner

of anger expression. Spielberger’s Trait Anger Scale and his Anger

Expression Scale were used to measure anger and the Irrational Beliefs

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

63

Questionnaire was used to assess irrational beliefs. This last instrument

consists only of 11 statements, 1 for each belief, and therefore can not be

considered a comprehensive assessment measure of irrational belief

categories.

In this study, trait anger correlated .35 with irrational beliefs; Anger-

In correlated .36 with irrational beliefs; Anger-Out correlated .25 with

irrational beliefs. High Trait Angry group endorsed 7 of the 11 beliefs

significantly more than the low Trait Angry group. The two groups most

differed on the statements “I become more upset than I should when

things are not the way I want them to be” and “I believe I need another

person stronger than myself on whom to rely.“

An analysis was done of Anger-ln and Anger-Out subscales and

high and low Trait Anger also. On Anger-ln items and Anger-Out items

subjects in the high Trait Anger group scored significantly higher than

subjects in the low Trait Anger group with one exception ("I boil inside but

don't show it").

In more recent research, Stuckless, Ford and Vitelli (1995) looked

at anger, vengeance and irrational beliefs in an adult male inmate

population. The used the Spielberger's Anger Expression Scale as well as

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

64

the Anger Expression Scale and the Belief Scale of Malouff and Schutte.

Both Trait Anger and Expressed Anger were found to be correlated

significantly with scores on total irrational beliefs.

Ferrari (1995) found that perfectionism cognitions significantly

related to anger suppression and inversely related to anger expression

(with approximately the same coefficients). However, no other cognitions

were examined. He used Spielberger's Anger Expression Scale and a

measure of perfectionism cognitions. College student subjects were used.

Summary of Literature Review

There has been a development of measures to assess both anger

and cognitive distortions. Early measures used to investigate anger were

affected by imprecision in the definition of anger as well as psychometric

inadequacies. However, the STAXI, developed by Spielberger, provided

an instrument that has been shown to have good reliability and validity

and is based on a precise definition of anger and on accepted theoretical

constructs. A number of scales have been used to assess irrational

thinking and cognitive distortions but most have been criticized

psychometrically. However, the Survey of Personal Beliefs and the

R eprod u ced with perm ission o f the copyright ow ner. Furth er reproduction prohibited w ithout perm ission.

65

Common Beliefs Survey have been formulated to correct deficiencies in

previous measures and appear to be useful for research.

An overview of research that examines anger and medical

conditions, gender, and depression shows the importance of

differentiating high and low Trait Anger and also of viewing separately

each of the modes of anger expression. It shows th a t gender of subjects

may affect results in some cases. The importance o f cognitions in the

experience and expression of anger is emphasized. It becomes clear, as

well, that research on anger and other variables generally involves many

factors and can be complex.

The cognitive theory of anger dates back to ancient times.

However, recently theoretical work in this area has been done Arnold

(1960), Bandura (1973), Beck (1963,1976), and Novaco (Novaco &

Welsh, 1989) among others. Research on the im portance of appraisals,

attributions, and expectations has also been completed in regard to anger

(Averill, 1982; Eckhardt & Deffenbacher, 1995; Novaco, 1979). Results

have focused on appraisal of potential threat, controllability of the event,

stability of consequences. Attribution of blame to others is an important

factor. Expectations are often thwarted or they selectively focus on the

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

66

negative. Research that has looked at anger and aggression has

emphasized hostile attribution bias (Dodge, etal., 1990).

Studies of anger and cognitive distortions have often used

instruments that have since been criticized. Results have frequently been

contradictory. Many studies have examined overall irrational thinking with

anger (and found a positive correlations between irrationality and trait

anger). Other studies have looked at overall trait anger and specific

cognitive distortions. Although many distortions have been implicated the

ones most frequently found to correlate are personal perfection,

catastrophizing, blame proneness, demand for approval, and low

frustration level. However, these studies have not usually looked

separately at different modes of anger expression and cognitive

distortions. In one study perfectionism was found to correlate with anger

suppression.

Research suggests that specific cognitive distortions may lead to

specific anger expression styles. For example, much research has looked

at the relationship between hostile attributions and aggression (Dodge,

Price, Bachorowski, & Newman, 1990). Kassinove & Eckhardt (1995)

state that "aggressive children have distorted and deficient social

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

67

information-processing mechanisms, including hostile attribution biases

and cue-detection deficits" (p. 198).

On the other hand, beginning with Freud (1917), there has been

theorizing that depression is anger turned in. Clay, Anderson, & Dixon

(1993) cite studies that have supported this contention. They found in their

own study of anger, stressful life events, and depression that

Anger-ln adds to the prediction of depression independent of other

factors. Studies by other researchers have challenged the Anger-ln

depression hypothesis (cited in Thomas, 1993). It may be that the

connection between Anger-ln and depression, when it exists, is mediated

by the specific cognitive distortions involved.

Hypotheses

This current research is based on the following hypotheses:

1. Anger-ln will be predicted by the cognitive distortions involving: Self

Directed Shoulds, Low Self W orth, Self Downing (perhaps suggesting the

belief that there is a lack of entitlem ent to angry feelings), the Importance

of Approval of Others, and Perfectionism.

2. Anger-Out will be predicted by cognitive distortions involving: O ther

Directed Shoulds and Blame Proneness and Low Frustration Tolerance.

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

68

3. Subjects high in Anger-Control will show significantly fewer cognitive

distortions than those who are high in Anger-Out or Anger-ln.

In addition, this study is a validation study of the STAXI (see

instruments, below). This scale has always been used as a self report.

This study used it that way also. However, live in partners of the subjects

also rated them (using a version that replaced pronouns) on trait anger

and expression of anger. Thus, there was a behavioral observation

component.

4. A comparison will be made between self rating of anger and the

partner's rating. Various results are possible: a) A positive correlation

between all the ratings of anger expression and no significant differences

between the means of the subscales of partner and self ratings. This

would provide more evidence of the validity of the STAXI. b) No

correlation and a significant difference between means, thus bringing

into question the validity of a self rating scale alone in accurately

recording a person's expression of anger, c) A pattern of positive,

negative or no correlations (with correlation between some

measures and the partner's ratings and no correlation

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

69

between the rest) and/or a patterns of significance or nonsignificance in

the comparison of means.

For example, it may be that people who rate themselves higher in

regard to expressing Anger-Out are also seen to have higher scores on

this subscale by partners, resulting in a positive correlation between self

and partner ratings of Anger-Out. People who rate themselves high on

suppressing anger may be seen by partners to not be experiencing anger

(resulting in a negative correlation or no correlation between self and

partner ratings of Anger-ln).

On the other hand, it may be that a people who rate themselves

high on keeping anger in are perceived to be high on Anger-ln by partners

(resulting in a positive correlation). However, a person who is seen by a

partner to express anger out may be less able to accurately perceive this

and may give a self rating of much lower Anger-Out (resulting in a

negative correlation or no correlation for Anger-Out).

It may also be that some of the differences between mean

subscale scores for self ratings and partner ratings of subjects will be

significant and some will not be. This would show that subjects and

partners give sim ilar ratings for some aspects of anger but not for others.

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

Chapter 3

METHODOLOGY

Sample Employed

A total of 98 subjects participated in this research. There were 47

couples (94 people) and four additional women whose partners did not

complete the questionnaires. Subjects were selected in couples so that

each person could be a subject but also give a report on observed

behavior in regard to anger for the partner. It was required that the

couples be currently living together but there were no other criteria for

participation. An attempt was made, by approaching possible subjects in a

variety of settings, to obtain a sample that was diverse in background.

Sex of Subjects: 47 subjects were male and 51 were female. AH couples

were heterosexual.

Age of Subjects: The ages ranged from 17 to 84, with a mean of 43.58

(standard deviation = 13.5).

Length of Time With Partner: This varied from under one year to 63

years, with a mean of 17.1 years (standard deviation = 13.92)

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

7 1

Education of Subjects: Two subjects did not respond to this question.

For the others, highest level of education completed was grade school

(3); high school (30); community college (31); college/university (16);

graduate school (16). (See Figure 2)

Most subjects answered all the questionnaires but in some cases

not all instruments were completed. As a result, there is data on the

following number of subjects:

Cognitive Distortion Questionnaires: 90

Self Anger Rating (STAXI): 92

Subject Rates Partner's Anger: 92

Partner Rates Subject's Anger 89

Subjects were sought in a variety of situations. I offered to give a

presentation on anger in exchange for participation in my study and was

able to obtain the subjects from a parent child center in this manner.

Subjects were obtained from the following locations (See Figure 1):

1. a suburban parent child center attended by children and their mothers:

35 subjects

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

72

2. individuals in a major Canadian city (and its suburbs) who were known

to friends of the researcher

22 subjects

3. staff members and their partners at a large psychiatric hospital where

the researcher was doing an internship:

19 subjects

4. individuals in a small city of 20,000 people in the United States who

were known to friends of the researcher:

22 subjects

Every effort was made to comply with ethical standards for

research with human subjects. Research participants were informed of the

nature of the research and asked for informed consent to participate,

using language that was reasonably understandable. Subjects were

informed that they were free to decline participation or withdraw at any

time. As participation only involved filling out some questionnaires and this

was entirely voluntary, no adverse consequences were expected.

Nevertheless, subjects were informed that the researcher could be

contacted to answer questions. If any subject, as a result of participation,

experienced distress that could not be abated by informal discussion, then

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

73

the researcher was prepared to assist the subject to obtain help

elsewhere. No subjects contacted the researcher for more information or

stated that participation was in any way stressful.

Instruments

A Demographic Questionnaire asked age, sex, am ount of time

the subject had lived with the partner, and level of education completed

(see Appendix C).

There were two measures of cognitive distortions. These were the

independent variables used to predict the results on the anger rating

instruments. One, the Survey of Personal Beliefs (Kassinove, 1986), was

originally developed by Andrew Berger and Howard Kassinove in 1981

and has undergone several revisions. It is a 50 item test of irrational

beliefs (based on rational emotive theory). It produces a total score as

well as subscales for 5 types of distortions (Awfulizing, Self Directed

Shoulds, Other Directed Shoulds, Low Frustration Tolerance, Self Worth).

There is a 6 point scoring scale for each item. Low Frustration Tolerance

is a measure of the subject's belief about ability to tolerate unpleasant

situations. For these scales, subjects with higher scores endorse fewer

cognitive distortions than those with lower scores.

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

74

Several studies have provided psychometric data on this scale.

Demaria, Kassinove and Dill (1989) performed a test of internal

consistency and confirmatory factor analysis. Subjects included a

nonclinical sample of 130 men and 150 women. Results indicated that the

SPB had satisfactory total and scale reliability. It supported a higher order

factor model including 5 first order factors and one second order general

factor.

In another study (Muran, Kassinove, Ross, & Muran, 1989) the

alpha coefficient reported for the total test was .89. Test-retest reliability

coefficients for the factors ranged from .74 to .87. Significant correlations

were found between overall irrational ideation and trait anger as well as

between irrationality and high depression scores. Nottingham (1992)

found significant correlation with measures of depression, hopelessness,

and anxiety. He also found decreases in irrationality scores for clients who

had been treated with rational emotive therapy.

The other measure of cognitive distortions was The Common

Beliefs Survey III (Bessai, 1977) which was revised in 1997 and 1998).

Also based on Ellis’s theory of irrational thinking, this instrument consists

of 54 items, with each scored on a 5 point Likert scale. There are 6 nine

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

75

item subscales derived from factor analysis. They are: Importance of the

Past, Blame Proneness, Self Downing, Importance of Approval,

Perfectionism, and Control of Emotions. The last subscale measure the

subject's beliefs about the possibility that people can control and influence

feelings. For these scales, subjects with higher scores endorse fewer

cognitive distortions than those with lower scores.

In a replication study by Tosi, Forman, Rudy and Murphy (1986)

the original factor structure was replicated in a sample of 264 medical

patients and the alpha coefficient was .85. Another study by Thorpe,

Parker and Barnes (1992) examined discriminative validity. The alpha

coefficient was also .85. Self Downing and Perfectionism produced more

irrational scores in patients than in the non-clinical sample. Only 2

subscales failed to show significant correlation with a questionnaire on

emotional states (Blaming and Control). However, the CBS did show a

stronger correlation with another questionnaire of cognitions than with the

one on affective states (showing discriminant validity).

Dependent measures were two versions of the State-Trait Anger

Expression Inventory (STAXI) (Spielberger, 1996). This instrument is

published by Psychological Assessment Resources, Inc. and is used as a

R e p rod u ced with perm ission o f the copyright ow ner. F u rth e r reproduction prohibited w ithout perm ission.

76

self rating scale to measure the experience and expression of anger. It

comes with a manual that gives normative data based on responses of

more than 9,000 subjects from three populations, adults, adolescents and

college students. It gives T scores and percentiles. It also describes

reliability and validity and current research. It is based on two earlier

instruments developed by Spielberger, The State-Trait Anger Scale and

the Anger Expression Scale. The older and the newer versions have been

widely used in research. The STAXI has been given good reviews in The

Thirteenth Mental Measurements Yearbook (Pittenger, 1998; Raphael,

1998). This test is for ages 13 and above and can be administered

individually or in groups. It consists of 44 items that form 6 scales and 2

subscales, with responses for each statement coded on a 4 point Likert

scale. The scales are as follows:

1)State Anaer (S-Anger) (This is a 10 item scale that measures the

intensity of angry feelings at a particular time)

2) Trait Anaer (T-Anger) (This is a 10 item scale that measures individual

differences in the disposition to experience anger. It has 2 subscales)

R e p rod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

77

a) Anarv Temperament fr-Anaer/~n (This is a 4 item scale that

measures a general propensity to experience and express

anger without specific provocation)

b) Anarv Reaction (T-Anger/R) (This is a 4 items scale that

measures individual differences in the disposition to express

anger when criticized or treated unfairly by other individuals)

-3) Anaer-ln (AX/ln) (This is an 8 item anger expression scale that

measures the frequency with which angry feelings are held in or

suppressed)

4) Anger-Out (AX/Out) (This is an 8 item anger expression scale that

measures how often an individual expresses anger toward other people or

objects in the environment)

5) Anaer Control (AX/Con^ (This is an 8 item scale that measures the

frequency with which an individual attempts to control or modulate the

expression of anger)

6) Anger Expression CAX/Ex) (This is a scale, based on Anger-ln, Anger-

Out, and Anger-Control, that provides a general index of the frequency

with which anger is expressed, regardless of the direction of expression)

Reliability for the STAXI is provided with coefficient alphas

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

78

and remainder coefficients for all the scales for normative populations of

adults, college students, and adolescents. The alphas ranged from .65 to

.93 and are uniformly large except for the 4 item T-Anger/R subscale.

With some exceptions gender differences have been minimal. Age related

differences are present and separate norms are provided for different age

groups. In general, AX/ln, AX/Out, and AX/Ex scores decrease and

AX/Con scores increase as a function of age. There is no test-retest

information. Anger-Out and Anger-ln have consistently been found to be

orthogonal. Correlations are given for the STAXI and other measures of

anger and psychopathology. According to Raphael (1998) "These data

provide strong information on the convergent and divergent validity of the

scales, which appear to hold up well" (p. 950).

Another version of the State-Trait Anger Expression Inventory

(STAXI), in which the pronoun “I" had been changed to "your partner",

was also used. With this instrument subjects rated their partner's anger for

the same dimensions used in the self rating.

For information for each variable in the instruments used (number

of subjects, mean and standard deviation) as well as comparisons to

normative data when it is available, see Appendix B.

R eprod u ced with perm ission o f the copyright ow ner. Furth er reproduction prohibited w ithout perm ission.

79

Procedures

Packages were prepared consisting of the demographics form, the

questionnaires, a letter giving an explanation of the research, instructions

for those who volunteer to participate, and a consent form (see

Appendix D). The instructions varied slightly according to the location. The

instruments were placed in the packages in randomized order in an

attempt to prevent the introduction of bias in the results based on order in

which the measures were completed. (However, many packages Were

distributed that were not completed. Therefore, it is possible that the order

of instruments in the completed packages may not have been random.)

The demographics form and four questionnaires were completed by the

subjects.

At the first location (parent child center), packages were distributed

by the center coordinator and returned by the women who completed

them when they attended a presentation on anger given by the

researcher. At the second and fourth locations, packages were given to

friends of the researcher and they distributed the packages to people they

knew who were willing to participate.

At the mental health center, the third location, permission was first

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

80

obtained from the chief psychiatrist and the committee on ethics in

research. A staff psychologist served as on-site coordinator of the

research. Permission was also obtained from the program managers of

each ward where staff were asked if they would participate. Packages

were placed in the mail boxes of staff members (with each staff member

receiving two packages, one for the employee and one for that person's

partner).

It was made clear to everyone who received a package that

participation was voluntary. Subjects were assured that no names or

identifying information would be placed on the packages and their

responses would be completely confidential (see section on ethical

considerations). Subjects were informed that they could contact the

researcher if they had any questions or concerns.

Each partner was instructed to complete the questionnaires without

discussing the answers with anyone. Each partner was asked to seal the

finished questionnaires in an envelope provided for that purpose. Consent

forms were included with each package and were also signed but were

not placed in the sealed envelopes. The sealed packages and the

separate consent forms were collected by the coordinator of the parent

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

81

child center, the on site supervisor at the mental health center, and by the

friends of the researcher and returned to the researcher separately.

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

Chapter 4

RESULTS

Comparison of Results with Normative Data

See Appendix B for means and standard deviations of all the

test variables in this study as well as a comparison with normative data

when it is available. Kassinove and Berger (1994) give means for the

total score of the Survey of Personal Beliefs and for each scale, but

only proved a SD for the total score. Results from this study show that

means of subjects’ scores were well within one SD of the means

published by Kassinove and Berger (using the SDs from the current

study). Some means were almost identical to those published. No

normative data is available for the Common Beliefs Survey.

Appendix B also shows all the means and standard deviations for

the results of the anger scales (self rating, rating of partner by subject

and rating of subject by partner). Since norms have been published in

the STAXI manual (Spielberger, 1996), it was possible to also provide

t-scores for males and females in the current study. This is also in

Appendix B. Results of the t-scores show that means for all the

variables are well within one SD of the normative means.

R e p rod u ced with perm ission o f the copyright ow ner. F u rth e r reproduction prohibited w ithout perm ission.

83

Demographics

Information about four demographic variables was obtained from

subjects: age of subject (age), level of education completed (education),

gender (sex), and length of time in years subject has lived with current

partner (time). Each of these variables was checked for possible

significance in relation to the subscales of the anger and cognitive

distortion measures.

Demographic Variables and Anger Scales

Sex and anger scales

MANOVAs were performed to examine the relationship between

sex and the self rating for anger, between sex and the rating by the

partner of the subject's anger, and between sex and the subject's rating of

the partner's anger. Each of these measures included scales for Trait

Anger, Angry Temperament, Angry Reaction, Anger-ln, Anger-Out, and

Anger-Control. Nonsignificant F scores were obtained in each of these

MANOVAs (see Table 1).

Education and anger scales

MANOVAs were performed to examine the relationship between

education and the self rating for anger, between education and the rating

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

84

by the partner of the subject's anger, and between education and the

subject's rating of the partner's anger. Each of these measures included

the scales listed in the previous section. Nonsignificant F scores were

obtained in each of these MANOVAs (see Table 2).

Age and anger scales

Age of subject and time subject has lived with partner are highly

correlated (Pearson product-moment correlation r =.83). Therefore, the

relationship of age to each or the anger measures was calculated with

time covaried out.

MANCOVAs were performed to examine the relationship between

age and each of the anger ratings: self rating, rating of partner , and rating

of self by partner. Each measure consisted of subscales, as described

above. Nonsignificant F scores were calculated for each anger rating (see

Table 3).

Time and anger scales

As described above, since tim e subject has lived with the partner

and age of subject are highly correlated (Pearson product-moment

correlation r =.83) the relationship of time to each or the anger measures

was calculated with age covaried out.

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

85

MANCOVAs were performed to examine the relationship between

time and each of the anger ratings: self rating, rating of partner, and rating

of self by partner. Each measure consisted of subscales, as described

above. Nonsignificant F scores were calculated for each anger rating (see

Table 4).

Demographic Variables and Cognitive Distortion Scales

Sex and cognitive distortion scales

MANOVAs were performed to examine the relationship between

sex and the Survey of Personal Beliefs (SPB) as well as the relationship

between sex and the Common Beliefs Survey III (CBS). The SPB has the

following subscales: Awfulizing, Self Directed Shoulds, Other Directed

Shoulds, Low Frustration Tolerance, and Self Worth. The CBS has the

following subscales: Importance of the Past, Blame Proneness, Self

Downing, Importance of Approval, Perfectionism, and Control of

Emotions.

The relationship between sex and the SPB was nonsignificant.

However, the relationship between sex and the CBS was significant,

overall F (6, 83) = 3.63, q < .01 (see Table 5).

Univeriate F tests identified that the only significance was for the

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

86

subscale "Blame Proneness" (M males = 25.33; M females = 28.76), F

(1, 88) = 11.40, 2 = <.01 (see Table 5).

Education and cognitive distortion scales

MANOVAs were performed to examine the relationship between

education and the Survey of Personal Beliefs (SPB) as well as the

relationship between education and the Common Beliefs Survey III (CBS).

The subscales contained in these measures are listed in the previous

section.

The relationship between education and the SPB was significant,

with overall F (25, 298) = 1.54, 2 =.05 (see Table 6). Subsequent

univeriate F tests showed the only significance to be on the subscale

"Other Directed Shoulds", F (5, 84) = 2.72, 2 <-05 (see Table 6).

A subsequent Tukey post hoc test (see Table 7) showed that the

differences did not follow a predictable pattern in regard to level of

education. The significance occurred between level one of education

(grade school completed) and other levels of education (high school,

community college, and university). Since there were only three subjects

in level one, it was decided that they were outliers and they were excluded

R eprod u ced with perm ission o f the copyright ow ner. Furth er reproduction prohibited w ithout perm ission.

87

from subsequent analysis of the cognitive distortions.

The relationship between education and the CBS was also

significant, overall F (30, 318) = 1.80, g < .01 (see Table 6). Subsequent

univeriate F tests showed the significance to be for the subscale

"Perfectionism", F (4, 83) = 4.05, g <.01 (see Table 6).

A subsequent Tukey post hoc test (see Table 8) showed that the

only significant difference, for this one subscale, was between high school

education and post graduate education.

Age and cognitive distortion scales

As described above, since age of subject and time subject has

lived with the partner are highly correlated (Pearson product-moment

correlation r =.83) the relationship of age to each cognitive distortion

measures was calculated with tim e covaried out.

MANCOVAs were performed to examine the relationship between

age of subject and each cognitive distortion scale, the Common Beliefs

Survey III (CBS) and the Survey of Personal Beliefs (SPB). Each

measure consisted of subscales, as described above. Nonsignificant F

scores were calculated for each cognitive distortion measure (see

Table 9).

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

88

Time and cognitive distortion scales

As described above, since time subject has lived with the partner

and age of subject are highly correlated (Pearson product-moment

correlation r =.83) the relationship of time to each of the cognitive

distortion measures was calculated with age covaried out.

MANCOVAs were performed to examine the relationship between

time with partner and each cognitive distortion scale, the Common Beliefs

Survey III (CBS) and the Survey of Personal Beliefs (SPB). Each

measure consisted of subscaies, as described above. Nonsignificant F

scores were calculated for each cognitive distortion measure (see

Table 10).

Prediction of Trait Anger and Anger Expression

by Cognitive Distortions

It was hypothesized that trait anger and anger expression style

would be predicted by different cognitive distortions.

Intercorrelation of Cognitive Distortion Subscales

As a first step in the data analysis of this part of the study, a

correlation matrix (using Pearson Product Moment r ) was prepared for the

R eprod u ced with perm ission o f the copyright ow ner. Furth er reproduction prohibited w ithout perm ission.

89

subscales of the two measures of cognitive distortion, the Survey of

Personal Beliefs (SPB) and the Common Beliefs Survey III (CBS). This

was done to assess these measures for multicollinearity. Low to moderate

correlations were found (see Table 15 for the correlations).

Statistical Analysis of Cognitive Distortions Predicting Anger

Stepwise multiple regressions were performed. The subscales of

the SPB were entered as independent (predictor) variables in separate

calculations for each of the dependent variables: Trait anger, Angry

Temperament, Angry Reaction, Anger-ln, Anger-Out, and Anger-Control

as subjects rated themselves. Stepwise multiple regressions were also

performed using the subscales of the CBS as independent variables for

each of the dependent variables listed above. (See Table 16 and

Figures 3-8.)

For the multiple regressions, the three subjects with grade school

education were removed from the sample (see education and cognitive

distortions in demographics section above).

Cognitive distortions as predictors of trait anaer

Results of the stepwise multiple regression for the subscales of the

SPB showed that "Low Frustration" accounted for 23% of the variance of

R eprod u ced with perm ission o f the copyright ow ner. Furth er reproduction prohibited w ithout perm ission.

90

Trait Anger {F (1, 85) = 25.61, 2 <»01). The correlation was negative. This

means that people who have a lo w e r score on Low Frustration (and thus

are less rational and have m ore cognitive distortions about their ability to

tolerate unpleasant situations) have higher Trait Anger. The multiple

regression for the CBS showed that “Self Downing" accounted for 8% of

the variance of Trait Anger (F (1, 85) = 7.32, p = .01). The correlation was

also negative. (See Tables 16 and Figure 3.)

Cognitive distortions as predictors of anqrv temperament

Results of the stepwise multiple regression for the subscales of the

SPB showed that "Low Frustration" accounted for 13% of the variance of

Angry Temperament (F (1, 85) = 12.77, p <.01). The correlation was

negative. The stepwise regression for the CBS showed that "Control of

Emotions" accounted for 4% of the variance (with a positive correlation)

and "Control of Emotions" and “Self Downing" combined accounted for

8% for the variance (F (2, 84) = 3.42, p <.05). The correlation for "Self

Downing" was negative. (See Table 16 and Figure 4.)

Cognitive distortions as predictors of anarv reaction

Results of the stepwise multiple regression for the subscales of the

SPB showed that "Low Frustration" accounted for 22% of the variance of

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

91

Angry Reaction. The correlation was negative. "Awfulizing" and “Low

Frustration” combined accounted for 26% of the variance (F (2. 84) =

14.51, £ <.01). The stepwise regression for the CBS showed that "Self

Downing" accounted for 8% for the variance (F (1, 85) = 7.21, £ =.01). All

the correlations were negative. (See Table 16 and Figure 5.)

Cognitive distortions as predictors of Anger-ln

Results of the stepwise multiple regression for the subscales of the

SPB showed that "Low Frustration" accounted for 18% of the variance of

Anger-ln (F (1, 85) = 18.10, £ .01). The correlation was negative. The

stepwise regression for the CBS showed that "Self Downing" accounted

for 10% for the variance (F (1, 85) = 9.10, £ .01). The correlation was

negative. (See Table 16 and Figure 6.)

Cognitive distortions as predictors of Anger-Out

Results of the stepwise multiple regression for the subscales of the

SPB showed that "Low Frustration" accounted for 6% of the variance of

Anger-Out. The correlation was negative. The combination of "Low

Frustration" and "Other Directed Shoulds" accounted for 10% of the

variance (F (2, 84) = 4.59, £ =.01). However, the correlation for "Other

Directed Shoulds" was positive. The stepwise regression for the CBS

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

92

showed that "Self Downing" accounted for 5% for the variance (F (1, 85)

= 4.09, £ =.05). The correlation was negative. (See Table 16 and

Figure 7.)

Cognitive distortions as predictors of Anqer-Control

Results of the stepwise multiple regression for the subscales of the

SPB showed that “Low Frustration" accounted for 12% of the variance of

Anger-Control. The correlation was positive. The combination of "Low

Frustration" and "Awfulizing" accounted for 15% of the variance (F (2, 84)

= 7.16, g .01). The correlation for "Awfulizing" was negative. The stepwise

regression for the CBS showed that "Self Downing" accounted for 3% for

the variance. The correlation was positive. However, the F score was

nonsignificant. (See Table 16 and Figure 8.)

Total Cognitive Distortion Scores as Predictors of Trait Anger

Since it has been found previously that trait anger is associated

with higher overall scores of irrational thinking (Woods, 1987) a stepwise

multiple regression was performed using the SPB total score (SPB-tot)

and the CBS total score (CBS-tot) as independent variables and the self

report of Trait Anger as the dependent variable. Only the SPB total score

entered the regression equation with a correlation of r = -.34 (F (1, 88) =

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

93

11.59, <.01.

Cognitive Distortions as Predictors of Anger Subscales Using Subjects

With Higher Trait Anger

Zwerdling & Thorpe (1987) found that subjects with high and

moderate anger results scored significantly higher on measures of

irrational thinking. To see whether or not this would apply to the present

study stepwise multiple regression procedures were also calculated, as

described above, using only the 46 subjects who had the higher half of

scores for Trait Anger.

To choose these subjects a median split was used. The 46

subjects all had Trait Anger scores above the median of 16.5. According

to the norms published in the STAXI manual (Spielberger, 1996), a raw

score of 17 on Trait Anger is equal to a t -score of 50 for male

adults and a t-score of 47 for female adults. Thus the subjects used in

this study for the analysis of high Trait Anger scored approximately in the

upper half of the population used for the standardization of the STAXI.

Cognitive distortions as predictors of trait anger

Results of the stepwise multiple regression for the subscales of the

SPB showed that "Low Frustration" accounted for 13% of the variance of

R eprod u ced with perm ission o f the copyright ow ner. Furth er reproduction prohibited w ithout perm ission.

94

Trait Anger (F (1, 44) = 7.51, £ .01). The correlation was negative. The

multiple regression for the CBS showed that "Self Downing" accounted for

10% of the variance of trait anger. The correlation was negative. The

combination of "Self Downing" and "Importance of Approval From Others"

accounted for 16% of the variance (F (2, 41) = 3.78, g .05). The

correlation for "Approval" was positive. (See Table 17 and Figure 9.)

Cognitive distortions as predictors of angry temperament

Results of the stepwise multiple regression for the subscales of the

SPB showed that "Low Frustration" accounted for 6% of the variance of

Angry Temperament. The correlation was negative. The combination of

"Awfulizing" and “Low Frustration" accounted for 13% of the variance. The

correlation for "Awfulizing" was positive. However, the F score was

nonsignificant. The stepwise regression for the CBS showed that "Control

o f Emotions" accounted for 12% of the variance (with a positive

correlation) and "Control of Emotions" and "Self Downing" combined

accounted for 17% for the variance (F (2, 41) = 4.17, g .05). The

correlation for “Self Downing" was negative. (See Table 17 and

Figure 10.)

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

95

Cognitive distortions as predictors of angry reaction

Results of the stepwise multiple regression for the subscales of the

SPB showed that "Awfulizing" accounted for 14% of the variance of Angry

Reaction (F (1, 42) = 6.82t g =.01). The correlation was negative. The

stepwise regression for the CBS showed that “Self Downing" accounted

for 5% for the variance. The correlation was negative. However, the F

score was nonsignificant. (See Table 17 and Figure 11.)

Cognitive distortions as predictors of anger-in

Results of the stepwise multiple regression for the subscales of the

SPB showed that "Low Frustration" accounted for 15% of the variance of

Anger-ln (F (1, 42) = 7.65, g =.01). The correlation was negative. The

stepwise regression for the CBS showed that "Self Downing" accounted

for 29% for the variance (F (1, 42) = 16.84, g .01. The correlation was

negative. (See Table 17 and Figure 12.)

Cognitive distortions as predictors of anger-out

Results of the stepwise multiple regression for the subscales of the

SPB showed that "Low Frustration" accounted for 8% of the variance of

Anger-Out The correlation was negative. The combination of "Low

Frustration" and "Other Directed Shoulds" accounted for 14% of the

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

96

variance (F (2, 41) = 3.42, £ .05). However, the correlation for "Other

Directed Shoulds" was positive. The stepwise regression for the CBS

showed that "Self Downing" accounted for 11% for the variance. The

correlation was negative. The combination of "Importance of Approval of

Others” and “Self Downing” accounted for 16% of the variance (F (2, 41) =

3.95, £ .05). The correlation of "Approval" was positive. (See Table 17 and

Figure 13.)

Cognitive distortions as predictors of anoer-control

Results of the stepwise multiple regression for the subscales of the

SPB showed that "Low Frustration" accounted for 10% of the variance of

Anger-Control. The correlation was positive. The combination of "Low

Frustration" and "Awfulizing" accounted for 15% of the variance (F (2, 41)

= 3.98, £ .05). The correlation for "Awfulizing" was negative. The stepwise

regression for the CBS showed that "Self Downing" accounted for 5% for

the variance. The correlation was positive. However, the F score was

nonsignificant. (See Table 17 and Figure 14.)

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

97

Total scores of cognitive distortions as predictor of anoer-control

A stepwise multiple regression was performed using the SPB total

score (SPB-tot) and the CBS total score (CBS-tot) as independent

variables and self rating of Anger-Control as the dependent variable. The

resulting F score was nonsignificant.

Comparison Between Anger Self Ratings

and Anger Ratings of Subject by Partner

In order to provide a behavioral observation of each subject's level

of Trait Anger and anger expression style, in addition to the subject's self

report of anger, each subject was rated by a live-in partner.1

1A correction was made for an error in the scales showing subject rating of partner and partner rating of subject, in the Anger-Out subscale of these measures, one of 8 questions (#39) was inadvertently left out for 46 subjects. It was included in the subscale for 48 other subjects. (Four of the 98 subjects did not complete these scales.)The STAXI manual states that when an item in an anger expression scale is left out, 2 should be added to replace it (Spielberger, 1996,p. 4). However, in an effort to assure accuracy, a more exacting procedure was followed. First, the mean scores of the two groups (46 and 48 subjects) was compared (based on 7 out of 8 questions for each group) and the difference between the rating scales was found to not be significant. (For the scale showing rating of subject by partner was, t=.45, p=>.05. The means were 12.9 and 13.3. For the ratings of partners by subjects, t=.89, p=>.05. The means were 12.7 and 13.4.) Since no significant difference was found, the average score on question # 39 was calculated for the 48 subjects who completed it and found to be 1.7. This number was added to all the anger out scales of the subjects missing that question.

R e p rod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

98

Correlations Between Self Ratings of Anger and Anger Ratings of Subject

Bv Partner

Pearson Product Moment correlations were calculated2 between

each self rating and the rating by the partner. Moderate to high

correlations were found. Correlations were as follows:

Trait Anger: .64

Trait Anger (Temperament): .59

Trait Anger (Reaction): .42

Anger-ln: .33

Anger-Out: .49

Anger Control: .60

Correlations for State Anger are not reported as the subjects and

their partners did not necessarily complete the rating scales at the same

time and, as a result, there is no reason to expect the ratings for State

Anger to be similar. In addition, most self reports rated State Anger as

very low. This probably results from the fact that most subjects chose to

2For the entire correlation matrix, see Table 11.

R eprod u ced with perm ission o f the copyright ow ner. Furth er reproduction prohibited w ithout perm ission.

99

complete the questionnaires when they were not in an angry state.

Correlations Between Self Ratings of Anger and Ratings of Partner bv

Subject

To examine the possibility of rater bias (a tendency for a subject to

be non-discriminating and rate a partner in a manner very similar to the

self rating) Pearson Product Moment correlations were calculated

between self ratings and ratings subjects made of their partners. These

correlations were all low. The correlations are as follows:

Trait Anger: .22

Trait Anger (Temperament): .16

Trait Anger (Reaction): .27

Anger-ln: .16

Anger-Out: .01

Anger-Control: .14

Significance of the Difference Between Means of Self Anger Ratings and

Anger Ratings of Subject Bv Partner

Compared sample t tests were performed to examine the

difference between the means. All differences were significant at jd <.01.

The

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

100

t scores were as follows:

Trait Anger: t = 7.23 Trait Anger (Temperament): t = 3.91

Trait Anger (Reaction): t = 2.94

Anger-ln: t = 3.46

Anger-Out: t = 5.15

Anger-Control: t = 6.60

For all subscales except anger control, partners gave subjects lower

scores than subjects gave themselves. The opposite was true for anger

control with partner's giving higher scores than subjects gave themselves.

Correlations Between Anaer Subscales

Correlations between trait anger and anger subscales

Angry Temperament and Angry Reaction are subscales of Trait

Anger. High correlations (r ranges from .82 to .94) are found for the

relationship between Trait Anger and their subscales for all three rating

scales (see Table 12). The correlation between the two subscales,

AngryTemperament and Angry Reaction, fo r each of the three scales is

more variable, with r ranging from (.42 to .73) (see Table 13).

Correlations between trait anger and anger expression style

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

101

High correlations (see Table 14) were found between Trait Anger

and Anger-Out for each of the rating scales. For self rating of Trait Anger

and Anger-Out r =.64. For ratings of subjects by their partners for Trait

Anger and Anger-Out r =.79. For ratings by subjects of their partners for

Trait Anger and Anger-Out r =.78.

Moderate correlations were found between Trait Anger and

Anger-ln for each of the rating scales. For self rating of Trait Anger and

Anger-ln r =.48. For ratings of subjects by their partners for Trait Anger

and Anger-ln r =.52. For ratings by subjects of their partners for Trait

Anger and Anger-ln r =.54.

High negative correlations were found between Trait Anger and

Anger-Control for each of the rating scales. For self rating of Trait Anger

and Anger-Control r = -.59. For ratings of subjects by their partners for

Trait Anger and Anger-Control r = -.71. For ratings by subjects of their

partners for Trait Anger and Anger-Control r = -.70.

R e p rod u ced with perm ission of the copyright ow ner. F u rth e r reproduction prohibited w ithout perm ission.

Chapter 5

DISCUSSION

The focus of this research was to examine cognitive distortions as

predictors of Trait Anger and of anger expression (Anger-ln or

Anger-Out). In general support was found for a significant relationship

between certain cognitive distortions and the experience and expression

of anger. However, little support was found for the differentiation of

distortions hypothesized to be associated with Anger-ln and Anger-Out. In

addition, it was found that partner's rated subjects significantly lower on

Trait Anger, Anger-ln and Anger-Out and higher on Anger-Control than

subjects rated themselves. However, subjects and their partners, overall,

presented fairly similar patterns in their ratings.

Cognitive Distortions as Predictors of Anger

Experience and Expression

Cognitive Distortions and Trait Anger

No specific hypotheses were made regarding use of cognitive

distortions and level of Trait Anger or its component subscales of Angry

Temperament or Angry Reaction. An exploration of these factors to

discern relationships that have been reported elsewhere indicate,

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

103

however, partial support for an association between overall level of

cognitive distortions and Trait Anger. A significant relationship existed for

only one of the measures of cognitive distortions used (the SPB).

The relationship of individual cognitive distortions to T rait Anger

was also explored. Results showed the presence of a fairly strong

association of Low Frustration Tolerance (beliefs that one can not stand

difficult situations) to Trait Anger and its subscale Angry Reaction and a

smaller relationship with the subscale Angry Temperament. Awfulizing

(which is similar to catastophizing) was slightly related to Angry Reaction.

The relationship found between Low Frustration Tolerance and

Trait Anger is supportive of the work by Muran, et al. (1989), who also

utilized the SPB. Kassinove and Eckhardt (1994), using the SPB, also

found Low Frustration Tolerance and Awfulizing to be important.

Current results, using the CBS, showed Self Downing was also a

predictor of Trait Anger and its subscales but accounted for a sm aller

amount of variance. Control of Emotions showed a small, positive

correlation with Angry Temperament, which indicates that greater belief in

the ability to control emotions (fewer cognitive distortions) is associated

with a higher score for Angry

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

104

Temperament. However, the amount of variance accounted for by this last

cognitive distortion was negligible (4%).

Results of this study were similar to the work of Thorpe, Parker, &

Barnes (1992) who used the Common Beliefs Survey III and found that

Self Downing was most highly correlated to Trait Anger.

It has been found in the literature that Trait Anger has been

associated variously with perfectionism (Hogg & Deffenbacher, 1986;

Storey and Deffenbacher, 1985; Thorpe, Parker, & Barnes, 1992),

catastrophizing (Hogg & Deffenbacher, 1986; Storey and Deffenbacher,

1985), demand for approval (Hogg & Deffenbacher, 1986; Thorpe, Parker,

& Barnes, 1992; Woods & Coggin, 1985), blame proneness and

dependency (Storey and Deffenbacher, 1985), self downing and

importance of the Past (Thorpe, Parker, & Barnes, 1992), and frustration

tolerance ( Kassinove and Eckhardt, 1993; Muran, et al., 1989; Woods &

Coggin, 1985).

However, this study did not support a relationship between Trait

Anger and its subscales and the cognitive distortions of Self and Other

Directed Shoulds, Self Worth, Importance of the Past, Blame Proneness,

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

105

Importance of Approval, and Perfectionism. It is important to note that the

studies cited above did not use the SPB or CBS in their analyses. The

difference in instruments may account for the contrast in results.

Cognitive Distortions and Anger Expression

It was hypothesized that different types of cognitive distortions

would be identified as predictors of Anger-ln and of Anger-Out. This was

not supported. Again, results showed that the cognitive distortions most

likely to predict both Anger-ln and Anger-Out were Low Frustration

Tolerance and Self Downing. This may show that there is a relationship

between self reported anger and self criticism (including the beliefs that

one can not stand difficulties and that one is to blame when problems

occur).

Cognitive distortions and Anger-ln

It was hypothesized that Anger-ln would be predicted by the

following cognitive distortions: Self Directed Shoulds, Low Self Worth, Self

Downing, the Importance of the Approval of Others, and Perfectionism.

Results of the present study showed that Low Frustration Tolerance and

Self Downing were predictors of Anger-ln. Thus, while the hypothesis was

supported for Self Downing, none of the other distortions hypothesized to

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

106

be important entered into the regression equation. In addition, Low

Frustration Tolerance, which was not predicted to be important, accounted

for the greatest percentage of variance of all the distortions involved.

These results are in contrast to Ferrari (1995) who found that

perfectionism was related to anger suppression. However, he used a

different measure of perfectionism than this study and his sam ple was

restricted to college students. It is possible that sample or instrument

differences account for the dissim ilarity in results.

Cognitive distortions and Anger-Out

It was hypothesized that Anger-Out would be predicted by the

following cognitive distortions: O ther Directed Shoulds, Blame Proneness,

and Low Frustration Tolerance. This hypothesis was only partially

supported. Although a relatively small proportions of variance was

accounted for by Low Frustration and Other Directed Shoulds, these

distortions were found to be predictors of Anger-Out. However, Self

Downing also entered into the regression equation, accounting for a small

portion of the variance (less than for Anger-ln). Interestingly, the

correlation of Anger-Out and O ther Directed Shoulds was positive rather

than negative. This means that a higher score on Other Directed Shoulds

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

107

(which means a more rational belief system in regard to this cognitive

distortion and fewer distortions) was associated with a higher score for

Anger-Out. This is opposite to what was predicted in the hypothesis. The

amount of variance accounted for by this distortion was small. Blame

Proneness did not enter into the regression.

These results are in contrast to many other studies (Dodge, et al.,

1990; Fodacaro & Heller, 1991) that found that hostile attributions (blame

proneness) are present in children who exhibit aggression. However,

those studies did not employ adults as subjects.

Cognitive distortions and Anaer-Control

It was hypothesized that subjects high in Anger-Control would show

significantly fewer distortions than those who were high in Anger-Out and

Anger-ln. Results showed no significant relationship between overall

scores on cognitive distortion scales and Anger-Control and thus did not

support the hypothesis in terms of overall level of cognitive distortions.

However, two specific cognitive distortions were significant predictors of

Anger-Control. Subjects who had fewer distortions relating to Low

Frustration Tolerance had higher scores on anger control, thus

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

108

supporting the hypothesis for the one cognitiva distortion most associated

with Anger-Control. For Awfulizing, subjects with more distortions had

higher scores in Anger-Control.

These results may show that subjects with higher Anger-Control

believe that they can tolerate unpleasant situations. This may make

control of emotions easier, or, conversely, they may have this positive

view because they know how to modulate their emotions. However, they

do have some tendency to catastrophize bad situations (and may

therefore be more highly motivated to control their feelings when angry to

not make difficulties worse). The amount of variance accounted for by

Awfulizing was minimal, however. The most important predictor was Low

Frustration.

Cognitive Distortions and Subjects With Higher Trait Anger

Generally the differences in correlation scores between the higher

anger sample and the full sample were not great. It may be that there are

no significant differences for people with high Trait Anger over those with

low Trait Anger. However, this part of the analysis also used a smaller

sample and restricted the range of anger scores employed. This may have

lowered correlations. As well, the “high” Trait Anger subjects

R eprod u ced with perm ission o f the copyright ow ner. Furth er reproduction prohibited w ithout perm ission.

109

consisted of people who scored in approximately the top 50 % for

Trait Anger in comparison with the published norms for adults. The

results might have been quite different with a group of subjects consisting

of people who all scored much higher in Trait Anger.

In this analysis, in addition to Low Frustration Tolerance and Self

Downing, Importance of Approval was found to be a predictor of Trait

Anger in combination with Self Downing. However, Approval was inversely

related; that is, presence o f fewer distortions in this area was related to

higher Trait Anger. Perhaps people who care more about obtaining the

approval of others manage to reduce their anger or, alternatively, that

people who have high anger do not care about the approval of others.

Control of Emotions and Control combined with Self Downing were

the only significant predictors of Angry Temperament. However, fewer

cognitive distortions regarding control of emotions was associated with

Angry Temperament. It may be that people who believe that they are to

blame for failure and but also believe that emotions can be controlled are

more likely to have an Angry Temperament. Angry Reaction was

predicted by Awfulizing only.

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

110

As in the entire sample, Anger-ln was predicted by Low Frustration

Tolerance and Self Downing but in the more angry sample, Self Downing

accounted for a much higher percentage of the variance (and thus results

were more in support of the original hypothesis).

For Anger-Out, Low Frustration Tolerance, Other Directed Shoulds

and also Self Downing were slightly more important predictors than with

the full sample of subjects. Other Directed Shoulds was again positively

correlated (fewer cognitive distortions in that area in relation to Anger-

Out). In the higher anger sample, Importance of Approval also combined

with Self Downing to predict Anger-Out with Approval being positively

correlated with Anger-Out. Subjects who are more self critical but have

few er distortions about the need for approval of others and few er

distortions about what others should do, have higher Anger-Out scores.

For Anger-Control, only Low Frustration and Low Frustration

combined with Awfulizing were predictors. As in the full sample, Low

Frustration Tolerance was positively correlated and Awfulizing was

negatively correlated. People with fewer cognitive distortions related to

their frustration tolerance but more cognitive distortions related to

awfulizing have more Anger-Control.

R eprod u ced with perm ission o f the copyright ow ner. Furth er reproduction prohibited w ithout perm ission.

111

Cognitive Distortions and Anger: Interpretation of Results

Results were only partially supportive of the original hypotheses.

What was shown in this study is that the cognitive distortions most likely

to predict anger were Low Frustration Tolerance and Self Downing. These

were the main distortions that accounted for variance in all of the anger

measures except Anger-Control. With the exception of Anger-Control,

subjects with more of these two cognitive distortions had higher anger

scores. However, for Anger-Control subjects with fewer cognitive

distortions in the area of frustration tolerance had higher scores on

Anger-Control. Self Downing was not found to be significant for that one

anger scale.

Correlations to examine the relationship between the scales of the

two measures of cognitive distortions showed that Low Frustration

Tolerance and Self Downing were more highly related than any of the

other subscales (see Table 15). The common element between them may

be what is associated with anger.

Low Frustration Tolerance is a measure of a subject's belief that it

is not possible to endure unpleasantness. Self Downing is a measure of

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

112

the subject’s belief that failure must result in self blame. It is possible to

speculate that the commonality may be a tendency to be self critical.

People who endorse these distortions may believe that they do not have

the ability to tolerate difficulties and when they fail it is a personal failure.

Being self critical might be what is associated with high Trait Anger and

anger expression (in or out). The idea that one can stand unpleasant

situations might be associated with Anger-Control.

It could be that people who have anger and do not control it are

then self critical. It could be that people who have a lower opinion of

themselves do not believe they have the ability to control anger and

therefore do not try to control it

Comparison Between Subjects' Self Rating of Anger and

Rating of Subject by Partner

To assess the validity of a self report of anger, subjects were also

rated by their live in partners. The correlations between the self report and

partner report were positive and moderate to high (See Table 11). This .

shows that subjects and their partners, overall, followed the same patterns

in rating the subjects for anger. For example, when subjects rated

themselves higher in Anger-Out, partners also gave higher ratings. When

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

113

subjects rated themselves lower in Anger-Control, partners also gave

lower ratings

The highest correlations were for Trait Anger, Angry Temperament

and Anger-Control, showing very similar patterns for these aspects of

anger. The lowest correlation was for Anger-ln. It may be that Anger-ln is

less visible to other people than to the subjects themselves and thus it is

more likely that partners will rate differently than the subjects for this

aspect of anger.

Results also showed that the actual ratings by the partners were

significantly different than the ratings by subjects. The partners rated the

subjects lower on Trait Anger and on anger expression (In and Out) than

the subjects rated themselves. The partners rated sQbjects higher on

Anger-Control than subjects rated themselves. Thus, subjects appear to

be more critical of themselves in their ratings of their anger than their

partners are of them.

Comparison of Self Rating of Anger and Rating of Partner by Subject

An analysis of the relationship between a person's rating of self and

the person's rating of partner showed low correlations. These results give

R eprod u ced with perm ission o f the copyright ow ner. Furth er reproduction prohibited w ithout perm ission.

114

support for the validity of self ratings of anger as well as for the validity of

partner ratings. They provide evidence that subjects were discriminative in

their ratings and did not have a set approach that resulted in rating others

with the same pattern used when they rated themselves. In other words,

subjects who rated themselves high or low in anger did not necessarily

also rate their partners similarly.

Correlations Between Anger Scales

Very high correlations were found between Trait Anger and each of

its subscales (Angry Temperament and Angry Reaction) for each of the

anger measures (self rating, rating by partner and rating of partner).

Moderate to high correlations were found between the subscales, with

lowest correlations for the self rating. Subjects rated themselves higher,

overall, on Angry Reaction (M= 8.09) than on Angry Temperament

(M= 6.30). It may be that being angry when there is a provocation seemed

more justified and less self critical to subjects than it did to rate

themselves as having an Angry Temperament, which is more enduring.

Correlations between Trait Anger and Anger-ln and Anger-Out were

moderate to high. There is a larger correlation with Anger-Out. Spielberger

(1996) reported the sam e trend and stated that people who experience

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

115

anger more frequently are more likely to express Anger-Out (p. 20}.

However, the correlations found in this study are higher than those given

by Spielberger (1996) in his manual. Those correlations are based on

scores of students only and the sample was much larger than that in the

current study. Either of those factors may explain the difference.

Current results show correlations between Trait Anger and

Anger-Control were moderate to high and negative. Subjects who

experience anger frequently are less likely to control their feelings (or,

perhaps, those who know how to modulate their emotions are less likely to

experience anger frequently).

In summary, the results of this study show that Trait Anger and

anger expression (Out or In) are all predicted by cognitive distortions that

show self criticism. Subjects higher in Anger-Control have fewer of those

cognitive distortions. In addition, subjects' self ratings of anger are less

positive than the ratings given them by partners. This may show that

subjects are self critical when they think about their experience and

expression of anger. It appears that there is an association between self

criticism and self rating of anger and that this association is important for

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

116

all the subscales of experience and expression of anger (the relationship

with Anger-Control being inverse).

Assumptions and Limitations

There are a number of limitations and assumptions inherent in this

study. First of all, in regard to the instruments used, the two measures of

cognitive distortion were selected because they appeared to have the best

psychometric properties of those available. Yet, reliability and validity data

.is weak. Also, an attempt was made to place all the instruments in the

response booklets in random order. However, most booklets given out

were not used and it is not known whether or not the tests were in random

order in the packages completed by the subjects. If they were not, there

may have been a response bias caused by the order of the measures. Ali

of the instruments depended on subject self report except the rating by the

live-in partner. Even that rating depended on the report given by that

person. It is assumed that people were honest and perceptive in their

responding but it is possible that they were not.

An attempt was made to use subjects from a variety of settings

who differed in age, time with partner, and education level. Although the

subjects appear to have been fairly heterogeneous in those respects, they

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

117

may have been homogeneous in other ways. It is not known what their

socioeconomic status was, for example, and it is possible that most if not

all were middle class. This may limit the possibility of generalizing the

results to people of a different SES.

There are also a number of limitations in regard to the statistical

analysis. An attempt was made to control for Type I error by using overall

MANOVAs for an instrument as a whole and only further analyzing the

results of the individual scales if the overall F was significant. However,

since there were still quite a number of statistical tests used in this study,

it is possible that some results were significant due to chance.

Another limitation results from the multicollinearity of the subscales

in the two measures of cognitive distortions. It is possible that significant

relations were not found due to this problem.

Also, analysis showed that there is a significant gender difference

for Blame Proneness. However, this was not taken into account in

subsequent analysis involving cognitive distortions. The decision was

made to essentially ignore this difference because the analysis chosen,

stepwise multiple regression, involved using all of the subscales of the

CBS at once as independent variables. It was not possible to do separate

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

118

analysis for each gender of only the Blame Proneness subscale alone.

However, it is possible that the results for Blame Proneness were affected

by the gender difference.

As well, a significant difference was found in the CBS

Perfectionism scale for level of education. Since analysis showed that the

significance did not seem to reflect a consistent effect for increase in

education level, this was also ignored in subsequent analysis using the

CBS. However, it is possible that the Perfectionism scale was affected by

differences in education level.

Analysis of cognitive distortions predicting anger were completed

twice, using both the full sample and also the half of the sample with

higher anger scores. However, the number of subjects in the higher anger

group was small (n=46). This small sample size may have resulted in

nonsignificant scores when significance would have been found for a

larger n. In addition, by taking only the higher anger scores, a restriction of

range was imposed for anger and this also may have reduced the

correlation. As well, subjects used in the “high” Trait Anger group had

scores in the top half of the norms for this scale. However, different

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

119

results might have been obtained if these subjects had all scored very

high in Trait Anger.

Implications and Future Directions

As we attempt to understand factors underlying anger, it is

important to explore associated cognitions. The results of this study

suggest that cognitive distortions related to self criticism are associated

with the experience and expression of this emotion. Clinicians working

with people who have difficulties with the expression or management of

anger should be aware that it is important to explore in the counselling and

therapy cognitions relating to self blame and ability to withstand and

overcome problem situations.

In regard to the STAXI, this is the first known study where a live-in

partner has also rated subjects and scores were compared. Results

suggest that a self rating of anger may be different from a rating of the

same person by an observer. This research shows that people tend to

rate themselves more harshly. Current results can not tell us whose rating

is more accurate. Researchers and others using the STAXI should be

aware of this discrepancy.

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

120

Future work in this area should include replication of this study

using a larger sample of people with high Trait Anger. In addition, it would

be of interest to look separately at subjects with high and low scores for

the cognitive distortions of Self Blame and Low Frustration Tolerance. Do

those groups rate themselves differently for anger, as would be expected

from current results? Are people with fewer distortions in those areas

more likely to rate themselves as their partners rate them? More work on

the STAXI comparing different behavioral observations of anger with self

ratings would be useful as we attempt to understand exactly what is

measured by this test. Finally, it is important to further develop

instruments to measure cognitive distortions. More reliability and validity

studies are needed for these measures.

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

121

References

Arnold, M. B. (1960). Emotion and personality. New York: Columbia

University Press.

Averill, J. R. (1982). Anger and aggression: An essav on emotion.

New York: Springer-Verlag.

Averill, J. R. (1983). Studies on anger and aggression: Implications

for theories of emotion. American Psychologist. 38.1145-1160.

Bandura, A. (1973). Aggression: A social learning analysis.

Englewood Cliffs, NJ : Prentice-Hall.

Bandura, A., Adams, N. E., & Berger, J. (1977). Cognitive

processes mediating behavior change. Journal of Personality and Social

Psychology. 35. 125-139.

Beck, A. T. (1963). Thinking and depression. Archives of General

Psychiatry. 9. 324-333.

Beck, A. T. (1976). Cognitive therapy and the emotional disorders.

New York: International Universities Press.

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

122

Ben-Zur, H., & Breznitz, S. (1991). W hat makes people angry:

Dimensions of anger-evoking events. Journal of Research in Personality.

25 1-22.

Berkowitz, L. (1990). On the formation and regulation of anger and

aggression: A cognitive-neoassociationistic analysis. American

Psvcholoaist.45 (4), 494-503.

Bessai, J. L. (1977, June). A factored measure of irrational

beliefs. Paper presented at the Second National Conference on

Rational Emotive Therapy, Chicago.

Biaggio, M. K. (1980). Assessment of anger arousal. Journal of

Personality Assessment. 44. 289-298.

Biaggio, M. K., & Godwin, W. H. (1987). Relation of depression to

anger and hostility constructs. Psychological Reports. 61. 87-90.

Biaggio, M. K., Supplee, K., & Curtis, N. (1981). Reliability and

validity of four anger scales. Journal of Personality Assessment. 4 5 . 639-

648.

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

123

Boyle, S., & Siegman, A. W. (1992). Dimensions of anger

expression and CVR in angered men. Unpublished manuscript. Baltimore

County: University of Maryland. (Cited in Spielberger, C. K. (1996). State-

Trait Anger Expression Inventory professional manual. Odessa, FL:

Psychological Assessment Resources, Inc.).

Catchlove, R. F., & Braha, R. E. (1985). A test to measure the

awareness and expression of anger. Psychotherapy and Psvchosomatics.

43, 113-119.

Chemtob, C. M., Hamada, R. S., Roitblat, H. L., & Muraoka, M. Y.

(1994). Anger, impulsivity, and anger control in combat-related

posttraumatic stress disorder. Journal of Consulting and Clinical

Psychology. 62 (4), 827-832.

Chemtob, C. M, Novaco, R. W., Hamada, R. S., & Gross, D. M.

(1997). Cognitive-behavioral treatment for severe anger in posttraumatic

stress disorder. Journal of Consulting and Clinical Psychology. 65 (1),

184-189.

Clay, D. L., Anderson, W. P., & Dixon, W. A. (1993). Relationship

between anger expression and stress in predicting depression. Journal of

Counseling and Development. 72. 91-94

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

124

Clay, D. L., Hagglund, K. J., Kashani, J. H., & Frank, R. G. (1996).

Sex differences in anger expression, depressed mood, and aggression in

children and adolescents. Journal of Clinical Psychology in Medical

Settings. 3 (1), 79-93.

Craigie, F. C., Jr. (1992). Evaluation and six-year follow up of a

community-based Christian cognitive-behavioral stress management

program. Journal of Psychology and Christianity. 11 (3), 269-276.

Cramer, D „ & Fong, J. (1991). Effect of rational and irrational

beliefs on intensity and "inappropriateness" of feelings: A test of rational-

emotive theory. Cognitive Therapy and Research. 15 (4), 319-329.

Deffenbacher, J. L. (1992). Trait anger: Theory, findings, and

implications. In C. D. Spielberger & J.N. Butcher (Eds.). Advances in

personality assessment. (Vol. 9) (pp. 177-201). Hillsdale, NJ: Lawrence

Erlbaum Associates, Publishers.

Deffenbacher, J. L., Lynch, R. S., Oetting, E. R, & Kemper, C. C.

(1996). Anger reduction in early adolescents. Journal of Counseling

Psychology. 43 (2), 149-157.

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

125

Deffenbacher, J. L., Oetting, E. R., Thwaites, G. A. Lynch, R. S.,

Baker, D. A., Stark, R. S. Thacker, S., & Eiswerth-Cox, L. (1996). State-

trait anger theory and the utility of the Trait Anger Scale. Journal of

Counseling Psychology. 43 (2). 131-148.

Demaria, T. P., Kassinove, H., & Dill, C. A. (1989). Psychometric

properties of the Survey of Personal Beliefs: A rational-emotive measure

of irrational thinking. Journal of Personality Assessment. 53(2T 329-341.

Dembroski, T. M., MacDougall, J. M., Williams, R. B., Haney, T. L.,

& Blumenthal, J. A. (1985). Components of type A, hostility, and Anger-ln:

Relationship to angiographic findings. Psychosomatic Medicine. 4 7 . (3),

219-233.

Dodge, K. A., Price, J. M., Bachorowski, J. & Newman, J.P. (1990).

Hostile attribution biases in severely aggressive adolescents. Journal of

Abnormal Psychology. 99 (4), 385-392.

Eckhardt, C. I., Barbour, K. A., & Stuart, G. L. (1997). Anger and

hostility in maritally violent men: Conceptual distinctions, measurement

issues, and literature review. Clinical Psychology Review. 17 (4), 333-358.

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

126

Eckhardt, C. I., & Deffenbacher, J.L. (1995). Diagnosis of anger

disorders. In H. Kassinove (Ed.) Anger disorders. Definition, diagnosis and

treatment (pp. 27-47). Washington, D. C.:Taylor & Francis.

Ellis, A. E. (1977). Anger: How to live with it and without it. New

York: Citadel Press.

Ellis, A. E. & Tafrate, R. C. (1998). How to control vour anger

before it controls you. Secaucus, NJ: Carol Publishing Group. ~s<r ̂ € I e i<\

Engebretson, T. O., Matthews, K. A., & Schei/er, M. F. (1989).

Relations between anger expression and cardiovascular reactivity:

Reconciling inconsistent findings through a matching hypothesis. Journal

of Personality and Social Psychology. 57 (3). 513-521.

Esler, M., Julius, S., Zwefler, A., Randall, O., Harburg, E., Gardiner,

H., & DeQuattro, V. (1977). Mild high-renin essential hypertension. New

England Journal of Medicine. 296. 405-411.

Faber, S. D., & Burns, J. W. (1996). Anger management style,

degree of expressed anger, and gender influence cardiovascular recovery

from interpersonal harassment. Journal of Behavioral Medicine. 19 (1),

31-53.

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

127

Fava, M., Rosenbaum, J. F., Pava, J., McCarthy, M. K., Steingard,

R. J., & Bouffides, E. (1993). Anger attacks in unipolar depression. Part I

Clinical correlates and response to Duoxetine treatment. American

Journal of Psychiatry. 150. 1158-1163.

Ferrari, J. R. (1995). Perfectionism cognitions with nonclinical and

clinical samples. Journal of Social Behavior and Personality. 10(1), 143-

156.

Feshbach, S. (1986). Reconceptualizations of anger: Some

research perspectives. Journal of Social and Clinical Psychology. 4 (2),

123-132.

Fitness, FJ., & Fletcher, G. J. O. (1993). Love, hate, anger, and

jealousy in close relationships: A prototype and cognitive appraisal

analysis. Journal of Personality and Social psychology. 65 (5), 942-958.

Folkman, S., & Lazarus, R. S. (1980). An analysis of coping in a

middle-aged community sample. Journal of Health and Social Behavior.

21, 219-239.

Fondacaro, M. R., & Heller, K. (1990). Attributional style in

aggressive adolescent boys. Journal of Abnormal Child Psychology, 18

(1), 75-89.

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

128

Ford, B. D. (1991). Anger and irrational beliefs in violent inmates.

Personality and Individual Differences. 12 (3), 211-215.

Forgays, D. K., Spielberger, C. D., Ottaway, S. A., & Forgays, D. G.

(1998). Factor structure of the state-trait anger expression inventory for

middle-aged men and women. Psychological Assessment. 5 (2), 141-155.

Frasure-Smith, N., Lesperance, F., & Talajic, M. (1995). The impact

of negative emotions on prognosis following myocardial infarction: Is it

more than depression? Health Psychology. 14 (5), 388-398.

Freud, S. (1917/1961). Mourning and melancholia. In J. Strachey

(Ed.), The standard edition of the complete psychological works of

Sigmund Freud (Vol 14, pp. 237-258). London: Hogarth.

Friedman, A. S. (1970). Hostility factors and clinical improvement in

depressed patients. Archives of General Psychiatry. 23, 524-537.

Funkenstein, D. H . , King, S. H., & Drolette, M. E. (1954). The

direction of anger during a laboratory stress-inducing situation.

Psychosomatic Medicine. 16. 404-413.

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

129

Fuqua, D. R., Leonard, E., Masters, M. A., Smith, R. J., Campbell,

J. L., & Fischer, P. C. (1991). A structural analysis of the state-trait anger

expression inventory. Educational and Psychological Measurement. 51.

439-446.

Gentry, W. D., Chesney, A. P., Gary, H. G., Hall, R. P. & Herburg.

E. (1982). Habitual anger-coping styles: I. Effect on mean blood pressure

and risk for essential hypertension. Psychosomatic Medicine. 44, 195-202

Goldman, L., & Haaga, D. A. F. (1995). Depression and the

experience and expression of anger in marital and other relationships. The

Journal of Nervous and Mental Disease. 183 (8), 505-509.

Hall, G. S. (1899). A study of anger. American Journal of

Psychology. 10. 516-591.

Harburg, E., Blackelock, E. H., & Roeper, P. J. (1979). Resentful

and reflective coping with arbitrary authority and blood pressure: Detroit.

Psychosomatic Medicine. 51. 189-202.

Haynes, S. G., Feinieib, M., Levine, S., Scotch, N., & Kannel, W. B.

(1978). The relationship of psychosocial factors to coronary heart disease

in the Framingham Study: II. Prevalence of coronary heart disease.

American Journal of Epidemiology. 107. 384-401.

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

130

Hazaleus, S. L., & Deffenbacher, J. L. (1985). Irrational beliefs and

anger arousal. Journal of College Student Personnel. 26. 47-52.

Hogg, J. A., & Deffenbacher, J. L. (1986). Irrational beliefs,

depression, and anger among college students. Journal of College

Student Personnel. 27. 349-353.

Huss, M. T., Leak, G. K., & Davis, S. F. (1993). A validation study

of the Novaco Anger Inventory. Bulletin of the Psvchonomic Society. 31

(4), 279-281.

Jones, R. G. (1969). A factored measure of Ellis' irrational belief

system, with personality and maladjustment correlates. Dissertation

Abstracts International, 29, 4379-B (University Microfilms No. 69-6443).

Kassinove, H. (1986). Self-reported affect and core irrational

thinking: A preliminary analysis. Journal of Rational-Emotive Therapy. 4

(2), 119-130.

Kassinove, H. (Ed.). (1995). Anaer disorders: Definition, diagnosis

& treatment. Washington. DC: Taylor & Francis.

R e p rod u ced with perm ission o f the copyright ow ner. F u rth e r reproduction prohibited w ithout perm ission.

131

Kassinove, H., & Berger, A. (1994). Survey of Personal Beliefs.

(Unpublished Test). Hofstra University.

Kassinove, H., & Eckhardt, C. I. (1994). Irrational beliefs and self-

reported affect in Russia and America. Personality and Individual

Differences. 16 (1), 133-142.

Kassinove, H., & Eckhardt, C. I. (1995). An anger model and a look

to the future. In H. Kassinove (Ed.) Anger disorders: Definition, diagnosis

& treatment (pp. 197-204). Washington, DC: Taylor & Francis.

Kassinove, H., & Sukhodolsky, D. G. (1995). Anger disorders:

Basic science and practice issues. In H. Kassinove (Ed.) Anger disorders:

Definition, diagnosis & treatment (pp. 1-26). Washington, DC: Taylor &

Francis.

Keinan, G., Ben-Zur, H., Zilka, M., & Carel, R. S. (1992). Anger in

or out, which is healthier? An attempt to reconcile inconsistent findings.

Psvcholoov and Health. 7. 83-98.

Knight, R. G., Chisholm, B. J., Paulin, J. M., & Waal-Manning, H. J.

(1988). The Spielberger Anger Expression Scale: Some psychometric

data. British Journal of Clinical Psvcholoov. 27. 279-281.

R e p rod u ced with perm ission o f the copyright owner. F u rth e r reproduction prohibited w ithout perm ission.

132

Knight, R. G., Ross. R. A., Collins, J . !_., & Paramenter, S. A.

(1985). Some norms, reliability, and preliminary validity data for an s-r

inventory of anger: The Subjective Anger Scale (SAS). Personality and

Individual Differences. 6 , 331-339.

Kopper, B. A. (1993). Role of gender, sex role identity, and hype A

behavior in anger expression and mental health functioning. Journal of

Counseling Psychology. 40. 232-237.

Kopper, B. A., & Epperson, D. L.(1991). Women and anger: Sex

and sex-role comparisons in the expression of anger. Psychology of

Women Quarterly. 15. 7-14.

Kopper, B. A., & Epperson, D. L. (1996). The Experience and

expression of anger: Relationships with gender, gender role socialization,

depression, and mental health functioning. Journal of Counseling

Psychology. 43 (2L 158-165.

Lazarus, R. S., & Folkman, S. (1984). Stress, appraisal and coping.

New York: Springer Publishing Company.

Lefebvre, M. F. (1981). Cognitive distortion and cognitive errors in

depressed psychiatric and low back pain patients. Journal of Consulting

and Clinical Psychology 49(4). 517-525.

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

133

Levine, L. J. (1996). The anatom y of disappointment: A naturalistic

test of appraisal models of sadness, anger, and hope. Cognition and

Emotion. 10 (4), 337-359.

Lochman, J. E., & Dodge, K. (1994). Social-cognitive processes of

severely violent, moderately aggressive and nonaggressive boys. Journal

of Consulting and Clinical Psychology. 62 (2), 366-374.

Lohr, J. M., & Bonge, D. (1982). The factorial validity of the

Irrational Beliefs Test: A psychometric investigation. Cognitive Therapy

and Research. 6. 225-230.

Lohr, J. M., Hamberger, L. K., & Bonge, D. (1988). The relationship

of factorially validated measures of anger-proneness and irrational beliefs.

Motivation and Emotion. 12 (2), 171-182.

Lopez, F. G., & Thurman, C. W . (1986). A cognitive-behavioral

investigation of anger among college students. Cognitive Therapy and

Research. 10 (2), 245-256.

Malouff, J. M., & Schutte, N. S. (1986). Development and validation

of a measure of irrational belief. Journal of Consulting and Clinical

Psychology. 54 (6), 860-862.

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

134

McAuley, E., & Shaffer, S. (1993). Affective responses to externally

and personally controllable attributions. Basic and Applied Social

Psychology. 14 (4), 475-485.

Merluzzi, T. V., & Boltwood, M. D. (1989). Cognitive assessment. In

A. Freeman, D. M. Simon, L. E. Beutler, & H. Arkowitz (Eds.).

Comprehensive handbook of cognitive therapy (pp.249-266). New York:

Plenum Press.

Miller, N. E., Sears, R. R., Mowrer, O. H., Doob, L. W., &

Dollard, J. (1941). The frustration-aggression hypothesis. The

Psychological Review. 48 (1), 337-342.

Mizes, J. S., & Christiano, B. A. (1995). Assessment of cognitive

variables relevant to cognitive behavioral perspectives on anorexia

nervosa and bulimia nervosa. Behavior Research and Therapv.33(1). 95-

105.

Mizes, J. S., Morgan, G. D., & Buder, J. (1990). The Relationship of

cognitions, assertion, and anger arousal. Journal of Cognitive

Psychotherapy: An International Quarterly. 4 (4), 369-376.

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

135

Moore, T. W., & Paolillo, J. G. P. (1984). Depression: Influence of

hopelessness, locus of control, hostility and length of treatment.

Psychological Report. 54. 875-881.

Muran, J.C., Kassinove, H., Ross, S.,& Muran, E. (1989). Irrational

thinking and negative emotionality in college students and applicants for

mental health services. Journal of Clinical Psychology 45(2). 188-193.

Neimeyer, R. A., & Feixas, G. (1992). Cognitive assessment in

depression: A comparison o f some existing measures. European Journal

of Psychological Assessment.8f1), 47-56.

Nottingham, E. F. J., IV (1992). Use of the Survey of Personal

Beliefs Scale: Further validation of a measure of irrational beliefs with

psychiatric inpatients. Journal of Rational-Emotive & Coanitive-Behavior

Therapy. 10 (4), 207-217.

Novaco, R. W. (1977). Stress inoculation: A Cognitive therapy for anger

and its application to a case of depression. Journal of Consulting and

Clinical Psvcholoav.45 (4), 600-698.

R e p rod u ced with perm ission o f the copyright ow ner. F u rth e r reproduction prohibited w ithout perm ission.

136

Novaco, R. W. (1979). The Cognitive regulation of anger and

stress. In P.C. Kendall & S. D. Hollon (Eds.). Cognitive-behavioral

interventions: Theory, research, and procedures (pp. 241-285). New York:

Academic Press, Inc.

Novaco, R. W. (1995). Clinical problems of anger and its

assessment and regulation through a stress coping skills approach. In W.

O'Donohue & L. Kresner (Eds.). Handbook of psychological skills training:

Clinical technigues & applications (pp. 320-338). Boston: Allyn & Bacon.

Novaco, R. W., & Welsh, W. N. (1989). Anger disturbances:

Cognitive mediation and clinical prescriptions. In K. Howells & C. R. Hollin

(Eds.). Clinical approaches to violence (pp. 39-60). New York: John Wiley

& Sons.

Pittenger, D.J. (1998). Review of the State-Trait Anger Expression

Inventory, Revised Research Edition. In J.C. Impara & B.S. Plake (Eds.)

The thirteenth mental measurement yearbook (pp. 948-949). Lincoln,

Nebraska: The University of Nebraska Press.

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

137

Ramanaiah, N. V., Heerboth, J. R., & Schill, T. R. (1987).

Measurement of irrational beliefs: A critical review. In C. D. Spielberger &

J.N. Butcher (Eds.), Advances in personality assessment (Vol. 6) (pp. 49-

67). Hillsdale, NJ: Lawrence Erlbaum Associates, Publishers.

Raphael, A.J.(1998). Review of the State-Trait Anger Expression

Inventory, Revised Research Edition. In J.C. Impara & B.S. Plake (Eds.)

The thirteenth mental measurement yearbook (pp. 949-951). Lincoln,

Nebraska: The University of Nebraska Press.

Riley, W. T., & Treiber, F. A. (1989). The validity of

multidimensional self-report anger and hostility measures. Journal of

Clinical Psychology. 45 (3), 397-404.

Riley, W. T., Treiber, F. A., & Woods, M. G. (1989). Anger and

hostility in depression. The Journal of Nervous and Mental Disease. 177

(11), 668-674.

Robb, H. B. Ill, & Warren, R. (1990). Irrational belief tests: New

insights, new directions. Journal of Cognitive Psychotherapy: An

International Quarterly. 4 (3). 303-311.

R e prod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

138

Robbins, P. R., & Tanck, R. H. (1997). Anger and depressed affect:

Interindividual and intraindividual perspectives. The Journal of

Psychology. 13 (5), 489-500.

Sharkin, B. S. (1993). Anger and gender: Theory, research, and

implications. Journal of Counseling and Development. 7 1 . 386-389.

Sharkin, B. S. (1988). The measurement and treatment of client

anger in counseling. Journal of Counseling and Development. 66. 361-

365.

Siegel, HJ. M. (1985). The measurement of anger as a

multidimensional construct. In M. A. Chesney & R. H. Rosenman (Eds.).

Anaer and hostility in cardiovascular and behavior disorders (pp. 59-82).

Washington, DC: Hemisphere Publishing Corp.

Slaby, R. G., & Guerra, N. G. (1988). Cognitive mediators of

aggression in adolescent offenders: 1. Assessment. Developmental

Psychology. 24 (4), 580-588.

Smith, C. A., & Ellsworth, P. C. (1985). Patterns of cognitive

appraisal in emotion. Journal of Personality and Social Psychology. 48

(4), 813-836.

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

139

Smith, C. A., Haynes, K. N., Lazarus, R. S., & Pope, L. K. (1993). In

search of the "hot" cognitions: Attributions, appraisals, and their relation to

emotion. Journal of Personality and Social Psychology. 65 (5), 916-929.

Smith, T. W „ & Zurawski, R. M. (1983). Assessment of irrational

beliefs: The question of discriminant validity. Journal of Clinical

Psychology. 39 (6), 976-979.

Smucker, C., Martin, J. & Wilt, D. (1993) Values and anger. In S. P.

Thomas (Ed.) Women and anger (pp. 129-153). New York: Springer

Publishing Company.

Spicer, J., & Chamberlain, K. (1996). Cynical hostility, anger, and

resting blood pressure. Journal of Psychosomatic Research. 40 (4), 359-

368.

Spielberger, C. K. (1996). State-Trait Anger Expression Inventory

professional m anual. Odessa, FL: Psychological Assessment Resources,

Inc.

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

140

Spielberger, C. D., Crane, R. S., Kearns, W. D., Pellegin, K. L.,

Bickman, R. L., & Johnson, E. H. (1991). Anger and anxiety in essential

hypertension. In C. D. Spielberger. I. G. Sarason, Z. Kulcsar, & G. L. Van

Heck (Eds.) Stress and emotion: Anxiety, anaer and curiosity. (Vol. 14)

(pp. 265-283). New York: Hemisphere Publishing Corporation.

Spielberger, C. D., Jacobs, G., Russell, S., & Crane, R. S. (1983).

Assessment of anger: The State-Trait Anger Scale. In C. D. Spielberger &

J.N. Butcher (Eds.). Advances in personality assessment. (Vol. 2) (pp.

161-189). Hillsdale, NJ: Lawrence Erlbaum Associates, Publishers.

Spielberger, C. D., Johnson, E. H., Russell, S. F., Crane, R. J.,

Jacobs, G. A., & W orden, T. J. (1985). The experience and expression of

anger: Construction and validation of an anger expression scale. In M. A.

Chesney & R. H. Rosenman (Eds.), Anaer and hostility in cardiovascular

and behavioral disorders (pp. 5-30). New York: Hemisphere/McGraw-Hill.

Spielberger, C. D „ Krasner, S. S., & Solomon, E. P. (1988). The

experience, expression, and control of anger. In M. P. Janisse (Ed.).

Individual differences, stress, and health psychology (pp. 89-108). New

York: Springer-Verlag.

R eprod u ced with perm ission o f the copyright ow ner. Furth er reproduction prohibited w ithout perm ission.

141

Spielberger, D. D., Reheiser, E. C., & Sydeman, S. J. (1995).

Measuring the experience, expression, and control of anger. In H.

Kassinove (Ed.) Anaer disorders definition, diagnosis, and treatment (pp.

49-67). Washington, D.C.: Taylor & Francis.

Spielberger, C. D., Ritterband, L. M., Sydeman, S. J., Reheiser, E.

C., & Unger, K. K. (1995). Assessment of emotional states and personality

traits: Measuring psychological vital signs. In J. N. Butcher (Ed.) Clinical

personality assessment: Practical approaches (pp. 42-58). New York:

Oxford University Press.

Spielberger, C. D. & Sydeman, S. J. (1994). State-Trait Anxiety

Inventory and State-Trait Anger Expression Inventory. In M. E. Maruish

(Ed.) The use of psychological tests for treatment planning and outcome

assessment (pp. 292-321). Hillsdale, NJ: LEA.

Story, D., & Deffenbacher, J.L. (1985 April). Trait anger and

irrational beliefs. Paper presented at Rocky Mountain Psychological

Association. Tucson, AZ.

Straton, D. (1990). Catharsis reconsidered. Australian and New

Zealand Journal of Psychiatry. 24. 543-551.

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

142

Stuckless, N., Ford, B. D., & Vitelli, R. (1995). Vengeance, anger

and irrational beliefs in inmates: A caveat regarding social desirability.

Personality and Individual Differences. 18 (1). 1-6.

Suls, J., Costa, P. T., Jr., & Wan, C. K. (1995). Relationship of trait

anger to resting blood pressure: a meta analysis. Health Psychology. 14

(5), 444-456.

Tavris, C. (1989). Anger the misunderstood emotion (rev, ed.). New

York: Simon & Schuster, Inc.

Thomas, S. P. (Ed.). (1993). Women and anger. New York:

Springer Publishing Company.

Thorpe, G. L., & Frey, R. B. (1996). A short form of the common

Beliefs Survey III. Journal of Rational-Emotive & Cocnitive-Behavior

Therapy. 14(3). 193-198.

Thorpe, G. L., Parker, J. D., & Barnes, G. S. (1992). The Common

Beliefs Survey III and its subscales: Discriminant validity in clinical and

nonclinical subjects. Journal of Rational-Emotive & Cognitive-Behavior

Therapy. 10 (2), 95-104.

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

143

Tosi, D. I., Forman, M.A., Rudy, D. R., & Murphy, M.A. (1986).

Factor analysis of the Common Beliefs Survey III: A replication study.

Journal of Consulting & Clinical Psychology. 54 (3), 404-405.

Trexier, L., & Karst, J. (1973). Further validation for a new measure

of irrational cognitions. Journal of Personality Assessment. 37. 150-155.

Warren, R., & Kurlychek, R. T. (1981). Treatment of maladaptive

anger and aggression: Catharsis vs behavior therapy. Corrective & Social

Psychology and Journal of Behavior Technology Methods and Therapy.

27(3), 135-139.

Weiner, B., Graham, S., & Chandler, C. (1982). Pity, anger, and

guilt: An attributional analysis. Personality and Social Psychology Bulletin.

8 (2), 226-232.

Weissman, M., & Paykel, E. (1974). The depressed woman.

Chicago: The University of Chicago Press.

Wertheim, E. D. H., & Poulakis, Z. (1992). The relationships among

the General Attitude and Belief Scale, other dysfunctional cognition

measures, and depressive or bulimic tendencies. Journal of Rational-

Emotive & Cognitive-Behavior Therapy. 10 (4), 219-233.

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

144

Wickiess, C., & Kirsch. I. (1988). Cognitive correlates of anger,

anxiety, and sadness. Cognitive Therapy and Research. 12, (4), 367-377.

Williams, R., & Williams, V. (1993). Anger kills. New York: Random

House.

Woods, P. J. (1987). Reductions in type A behavior, anxiety, anger,

and physical illness as related to changes in irrational beliefs: Results of a

demonstration project in industry. Journal of Rational-Emotive Therapy. 5

(4), 213-237.

Woods, P. J., & Coggin, S. K. (1985). Irrationality profiles for anger

and anxiety. Journal of Rationai-Emotive Theraov. 3 (2), 124-129.

Zurawski, R. M., & Smith, T. W . (1987). Assessing irrational beliefs

and emotional distress: Evidence and implications of limited discriminant

validity. Journal of Counseling Psychology. 34 (2), 224-277.

Zwemer, W. A., & Deffenbacher, J. L. (1984). Irrational beliefs,

anger, and anxiety. Journal of Counseling Psychology. 31 (3), 391-393.

Zwerdling, A. B., & Thorpe, G. L. (1987). Anger and irrational

thoughts: Rational-Emotive Therapy. 5 (2), 108-117.

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

145

APPENDIX A

REGRESSION COEFFICIENTS FOR COGNITIVE DISTORTIONS

PREDICTING ANGER

(FULL SAMPLE)

Survey of Personal Beliefs

Trait Anger Low Frustration Tolerance -0.36 Angry Temperament Low Frustration Tolerance -0.12 Angry Reaction Awfulizing -0.11

Low Frustration Tolerance -0.14 Anger-ln Low Frustration Tolerance -0.26 Anger-Out Low Frustration Tolerance -0.20

Other Shoulds 0.15 Anger-Control Awfulizing -0.18

Low Frustration Tolerance 0.35

Common Beliefs Survey 111

Trait Anger Self Downing -0.26 Angry Temperament Self Downing -0.07

Control of Emotions 0.08 Angry Reaction Self Downing -0.13 Anger-ln Self Downing -0.24 Anger-Out Self Downing -0.15 Anger-Control not significant

R eprod u ced with perm ission o f the copyright ow ner. Furth er reproduction prohibited w ithout perm ission.

146

Appendix A (continued)

(HIGH TRAIT ANGER SAMPLE)

Survey of Personal Beliefs

Trait Anger Low Frustration Tolerance -0.25 Angry Temperament Not Significant Angry Reaction Awfulizing -0.19 Anger-ln Low Frustration Tolerance -0.27 Anger-Out Low Frustration Tolerance -0.28

Other Shoulds 0.20 Anger-Control Awfulizing -0.29

Low Frustration Tolerance 0.32

Common Beliefs Survey III

Trait Anger Self Downing -0.24 Need for Approval 0.21

Angry Temperament Self Downing -0.08 Control of Emotions 0.13

Angry Reaction Not Significant Anger-ln Self Downing -0.43 Anger-Out Self Downing -0.23

Need for approval 0.19 Anger-Control not significant

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

147

APPENDIX B

MEANS AND STANDARD DEVIATIONS OF VARIABLES

(With Normative comparison when Available)

NUMBER OF SUBJECTS FOR VARIABLES

N of Subjects:

Total s 98; SPB and CBS = 90

Anger Self Rating = 92

Anger Rating of Partner by Subject = 92;

Anger Rating of Subject by Partner = 89

AGE OF SUBJECTS AND TIME WITH PARTNER

Age of Subjects Time in years with Partner

Mean 43.58 17.06

SD 13.50 13.92

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

148

Appendix B (continued)

SURVEY OF PERSONAL BELIEFS

(Numbers in parenthesis are means and SD in norms provided by Kassinove & Berger, 1994

for normal adults) SPBTotai Awfulizing Self Directed

Shoulds

Mean 145.38 (139.8) 24.82 (25.1) 26.83 (25.6)

SD 20.37 (25.2) 5.12 (NA) 5.19 (NA)

Other Directed Low Frustration Self Worth

Mean

SD

Shoulds

30.42 (29.0)

5.74 (NA)

Tolerance

32.83 (30.3)

6.70 (NA)

30.46 (30.0)

5.84 (NA)

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

149

Mean

SD

Mean

SD

Mean

SD

Appendix B (continued)

COMMON BELIEFS SURVEY

(No norais available.)

CBSTotal Importance Blame Proneness

Of the Past

170.16 28.96 27.04

15.25 5.20 5.08

Self downing Approval Perfectionism

Of Others

24.36 30.30 29.68

5.47 4.56 6.10

Control of

Emotions

29.82

4.27

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

150

Mean

SD

Mean

SD

Mean

SD

Appendix B (continued)

ANGER SELF RATING

State Anger Trait Anger Angry Temperament

11.47 17.53 6.30

3.55 4.92 2.27

Angry Reaction Anger-ln

8.09 14.86

2.46 4.17

Anger-Out

15.00

3.83

Anger-Control Anger Expression

22.57 23.00

5.77 10.25

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

151

Appendix B (continued)

Mean T-Scores for Males and Females Self Rating of Anaer

Based on Norms Provided for Adults in STAXl Manual

(Spielberger. 19961

State Anger

Males: 53.26

Females: 51.70

Trait Anger Angry Temperament

47.96 53.78

47.07 51.87

Males:

Angry Reaction Anger-ln Anger-Out

45.80 49.59 51.63

Females: 46.28 48.54 51.35

Anger-Control Anger Expression

Males: 54.35 54.35

Females: 50.30 29.67

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

152

Mean

SD

Mean

SD

Mean

SD

Appendix B (continued’)

ANGER RATING OF PARTNER BY SUBJECT

State Anger

11.66

4.27

Trait Anger Angry Temperament

17.49 6.39

6.18 2.89

Angry Reaction Anger-ln

8.15

2.82

13.89

3.84

Anger-Out

(7 items of scale)

13.09

3.91

Anger-Out

(all 8 items of scale)

14.93

4.14

Anger-Control

23.46

6.27

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

153

Appendix B (continued)

Mean T-Scores for Males and Females

Anger Rating of Partner by Subject

Based on Norms Provided for Adults in STAXl Manual

(Spielberger. 1996)

State Anger Trait Anger

Males: 52.16 46.91

Females: 53.04 46.77

Angry Reaction Anger-ln

Males: 46.00 46.16

Females: 46.62 47.77

Anger-Control

Males: 50.49

Females: 49.19

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

Angry Temperament

53.00

52.60

Anger-Out

51.75

50.80

154

Mean

SD

Mean

SD

Mean

SD

Appendix B (continued)

ANGER RATING OF SUBJECT BY PARTNER

State Anger

11.70

4.33

Trait Anger Angry Temperament

17.49 6.35

6.19 2.92

Angry Reaction Anger-ln

8.15

2.75

13.99

3.78

Anger-Out

(7 items of scale)

13.13

3.94

Anger-Out

(all 8 items of scale)

14.98

4.18

Anger-Control

23.51

6.32

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

155

Appendix B (continued)

Mean T-Scores for Males and Females

Partners' Rating of Subjects' Anger

Based on Norms Provided for Adults in STAXl Manual

(Spielberger. 1996)

State Anger Trait Anger Angry Temperament

Males: 53.11 47.22

Females: 52.16 46.60

Angry Reaction Anger-ln

Males: 47.04 47.93

Females: 45.58 46.07

Anger-Control

Males: 48.52

Females: 50.86

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

52.76

52.72

Anger-Out

50.80

51.62

156

APPENDIX C

Demographics Questionnaire

Code Number:

Please fill in the correct answers for the following questions:

Sex (please circle one): male female

Age: _______

Length of time you have lived with current partner (indicate years or months):

Education I graduated from (please circle one):

grade school high school college university master's/doctorate

R eprod u ced with perm ission o f the copyright ow ner. Furth er reproduction prohibited w ithout perm ission.

157

APPENDIX D

INFORMED CONSENT FORM

I,_______________________________ give my consent to participate in the research study of Deborah Azoulay.

I understand that this research is being conducted to learn more about the types of thoughts that people have when they are angry. My participation involves answering questionnaires about myself and my partner and will take approximately 30 minutes.

Subject's Rights

‘ I understand that:

1. This participation is voluntary and without force and I am free to withdraw from this agreement at any time.

2. This research is for the purposes of completing a requirement toward the student's attainment of her doctoral degree.

3 . 1 am free to ask questions of the student and receive explanations about the research study and my rights as a subject.

4. Ail my answers to the questionnaires will be completely confidential. I will not be identified by name in this project. All information that might lead to my identity will be disguised.

Deborah Azoulay has explained the above rights to me as a subject and has informed me that this consent will remain in a confidential file.

Subject's Name P rin te d:________________________________ Signature:___________________________________ D a te :__

Witness' Name P rinted:_________________________________ Signature:_____________________________________D a te :__ Research Student's Signature:______________________ Date:

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

APPENDIX E

S ta te m e n t of C o m p lia n c e w ith H u m a n S u b ject U s e R e q u ir e m e n ts for the Use o f P s y c h o lo g ic a l Testing in R e s e a rc h

I hereby certify that, in compliance with the Human Subject Use Requirements. I have secured written c o n se n t o f all su b je cts w h o have voluntarily agreed to participate in m y research on:

_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ Q ~P / 4 ~ k 5} g ^ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _

A ll subjects have signed the attached sam ple Informed Consent Form. These signed forms w ill remain in m \ pos> ession.

- 5 - e { ? e r ^ . l \ / 4 Z c u /Student's N am e Printed: ■ f

S tu d en t's Signature: ■ w M . ! ? ? ?

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

159

APPENDIX F

INSTRUCTIONS

Thank you for participating in this research project. You will be asked to rate yourself and your partner for anger and also to rate yourself fo r some common beliefs.

1. In every couple, each partner will fill out one consent form and one set of questionnaires. Please fill out the questionnaires in the order you find them.

2. Before beginning the questionnaires, each partner must fill out a consent form. (Your partner can be your witness). The signed consent form should not be placed in the envelope with the questionnaires. You will be asked to turn in the consent forms separately.

3. Each partner should fill out one set of questionnaires on his/her own. Please do not discuss the questions with anyone while you are filling them out. Answer the questions as best you can.

4. Confidentiality will be maintained. Your identity will not be recorded on the forms.

5. If you have any questions please call me at: Tel (number here)

PLEASE MAKE SURE THAT YOU HAVE COMPLETED ALL THE PAGES.

Thank you.

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

TABLE I

RELATIONSHIP BETWEEN SEX AND ANGER VARIABLES

ANGER VARIABLES F SCORE

S e lf A nger R ating F (8, 83) = 1.33 R atin g of Partner's A nger by Subject F ( 7 , 8 2 ) = .41 R atin g of Subject’s A nger by Partner F ( 7 , 8 0 ) = .45

TA BLE 2

RELATIONSHIP BETWEEEN EDUCATION AND ANGER VARIABLES

ANGER VARIABLES F SCORE

S e lf A nger R ating F (32, 289) = 1.29 R atin g of Partner's A nger by Subject F (28, 279) = .74 R atin g of Subject's A nger by Partner F (28, 271) = 1.10

R e p rod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

TABLE 3

RELATIONSHIP BETWEEN AGE AND ANGER VARIABLES

ANGER VARIABLES F SCORE

S e lf A nger R ating F (312, 363) = .97 R ating o f Partner's A nger by Sub ject F (266, 3 1 6 ) = 1.07 R ating o f Subject's A nger by P artner F (266, 3 0 2 ) = 1.16

T A B L E 4

RELATIONSHIP BETWEEN TIME AND ANGER VARIABLES

ANGER VARIABLES F SCORE

S e lf A nger Rating F (248, 420) =1.02 R ating o f Partner's A nger by Sub ject F (203, 373) = .97 R ating of Sub ject's A nger b y Partner F (203, 360) = .90

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

TABLES

RELATIONSHIP BETW EEN SEX AND COGNITIVE DISTORTION MEASURES

COGNITIVE DISTORTION MEASURES F SCORE n f va ria n c e

SURVEY OF PERSONAL BELIEFS F (5,84) = 1.42 COMMON BELIEFS SURVEY III 1 F (6, 83) = 3.63* i 15%

*I> < .01

Univariate F Scores for Subscales of CBS CBS Subscale F Score % cif va ria nce

Im portance o f P ast F (1, 88) = 1.49 B lam e Proneness F ( 1 ,8 8 ) = 11.40* -■= 10% S e lf D ow ning F (1, 88) = 2.29 Im portance o f A p proval F (1, 88) = 2.01 Im portance o f P erfection F (1, 88) = 1.11 C ontrol of E m otions F (1, 88) = .64

* 2 < .01

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

TABLE 6

RELATIONSHIP BETWEEN EDUCATION AND COGNITIVE DISTORTION MEASURES

COGNITIVE DISTORTION MEASURES F SCORE o f v a ria n c e

SURVEY OF PERSONAL BELIEFS F (25, 298) = 1.54* ' 4% COMMON BELIEFS SURVEY III F (30, 318) = 1.80** 6%

*j> = .05 **g < .01

Univariate F Scores for Subscales of CBS CBS Subscales F Scores % of variance

Im portance o f Past F ( 4 , 8 3 ) = .75 B lam e Proneness F ( 4 , 83) = 1.33 S elf D ow ning F ( 4 , 83) = 2.33 Im portance of A pproval F (4, 83) = .59 Im portance of Perfection F ( 4 , 8 3 ) = 4.05* d - 12% C ontrol o f Emotions F ( 4 , 8 3 ) = 1.03

* £ < .01

Univariate F Scores for Subscales of SPB

SPB Subscales F Scores % o f variance

A w fulizing F (5, 84) = .99 S elf D irected Shoulds F (5, 84) = 1.49 O ther D irected Shoulds F (5, 84) = 2 .7 2 * c ^9% L ow Frustration F (5, 84) = 1.08 S elf W orth F (5, 84) = 1.45

£<.05

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

TABLE 7

TUKEY POST HOC TEST FO R EDUCATION LEVEL AND "OTHER DIRECTED SHOULDS"

EDUCATION LEVEL COMPLETED MEAN SCORE

GRADE SCHOOL 40.67 HIGH SCHOOL 29.64* COMMUNITY COLLEGE 29.43** COLLEGE/UNIVERSITY 30.64* POST GRADUATE DEGREE 31.87

Differences between grade school and these means are significant at * £ < .05 **£ < .01

Note: In C anada "com munity college" is called "college" and "college" is called "university".

R eprod u ced with perm ission o f the copyright ow ner. Furth er reproduction prohibited w ithout perm ission.

TABLE 8

TUKEY POST HOC TEST FO R EDUCATION LEVEL AND "PERFECTIONISM"

EDUCATION LEVEL COMPLETED MEAN SCORE

GRADE SCHOOL 29.67 HIGH SCHOOL 27.36* COMMUNITY COLLEGE 29.57 COLLEGE/UNIVERSITY 30.93 POST GRADUATE DEGREE 34.20*

Difference between high school and post grad is significant < .01

N ote: In Canada "com m unity college" is called "college" and "college" is called "university".

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

TABLE 9

RELATIONSHIP BETWEEN AGE AND COGNITIVE DISTORTION MEASURES

COGNITIVE DISTORTION MEASURES F SCORE

SURVEY OF PERSONAL BELIEFS F (190, 233) = 1.07 COMMON BELIEFS SURVEY III F (228, 275) = 1.16

TA B LE 10

RELATIONSHIP BETWEEN TIME AND COGNITIVE DISTORTION MEASURES

COGNITIVE DISTORTION MEASURES F SCORE

SURVEY OF PERSONAL BELIEFS F (155, 267) = 1.23 COMMON BELIEFS SURVEY III F (186, 315) = 1.16

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

TABLE 11

A n s e r Correlation M atrix

S E L F RATING

R A TIN G

TR A IT T E M P E R A M E N T REACTION ANGER-IN ANGER-OUT ANGER -CONTROL

T R A IT LOO TEMPERAMENT .84 1.00 REACTION _ .82 .42 1.00 ANGER-IN .48 .26 .26 1.00 ANGER-OUT .64 .70 .68 .20 1.00 ANGER- CONTROL -.59 -.65 -.68 -.18 -.56 1.00 SUBJECT RATES PARTNER T R A IT .22 .17 .18 .05 .03 -.08 TEMPERAMENT .14 .16 .04 .06 0 -.06 REACTION .28 .18 - .27 .05 .07 -.09 ANGER-IN .23 .20 .24 .16 -.02 -.18 ANGER-OUT .06 .09 -.03 -.03 .01 .02 ANGER- CONTROL -.07 -.09 -.01 -.05 .02 .14 P AR TN ER RATES SUBJECT T R A IT .64 .61 .41 .20 .53 -.56 TEMPERAMENT .57 .59 .30 .10 .51 -.56 REACTION .59 .55 .42 .26 .43 -.51 ANGER-IN .35 .26 .32 .33 .22 -.25 ANGER-OUT .46 .49 .22 .08 .49 -.45 ANGER- CONTROL -.52 -.49 -.30 -.16 -.40 .60

R eprod u ced with perm ission o f the copyright ow ner. Furth er reproduction prohibited w ithout perm ission.

T A B L E 11 continued

A n g e r C orrelation M atrix

p a r t n e r r a t e s SUBJECT

S U B JE C T RA TES PA R T N E R

TRAIT T E M P E R A M E N T REACTION AN GER-IN ANGER-OUT ANGER- CO.NTROI.

T R A IT .18 .15 I .18 .28 0 -.05 TEM PERAM EN T .14 .12 | .15 .23 .01 -.02 REACTION .20 .17 i .19 .31 0 -.11 AN GER -IN .29 .24 .31 .35 .08 -.10 AN GER -OU T -.02 -.01 -.03 .07 -.05 .04 ANGER- CONTROL -.06 -.04 -.09 -.07 .03 .02

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

TABLE 12

CORRELATIONS BETWEEN TRAIT ANGER AND ITS SUBSCALES

TRAIT ANGER

SELF RATING SUBJECT RATES PARTNER

PARTNER RATES SUBJECT

TR A IT ANGER SUBSCALES

SELF — r a t i n g T E M PER A M EN T

.84

REA CTIO N ci00

s u b j e c t TEM PER A M EN T .94 REACTION .91

p a r t n e r TEM PER A M EN T .94 RA t ES S U R IF fiT REA CTIO N .91

T A B L E 13

CORRELATIONS BETW EEN TRAIT ANGER SUBSCALES

REACTION

SELF RATING SUBJECT RATES PARTNER

PARTNER RATES SUBJECT

TEMPERAMENT

S E L F R A T IN G .42 SUBJECT RATES P A R T N E R .73 P A R T N E R R A T E S S U B J E C T

.73

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

TABLE 14

CORRELATION BETW EEN TRAIT ANGER AND ANGER EXPRESSION STYLES

ANGER STYLE

OUT ]N CONTROL

TRAIT AN GER

SELF RATING .64 .48 -.59 SUBJECT RATES PARTNER .78 .54

1 • o

PARTNER RATES SUBJECT .79 .52 -.71

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

TABLE 15

INTERCORRELATION O F COGNITIVE DISTORTION SUBSCALES

SURVEY OF PERSONAL BELIEFS

AW FU LIZIN G SELF OTHERS LOW SELF WORTH SHOULD SHOULD FRUSTRATION

AW FU LIZIN G LOO 1 SELF SHOULD .35 1.00 OTHERS SHOULD .31 .38 LOO LO W FRUSTRATION .42 .34 .40 1.00 SEL F W ORTH .48 .40 .34 .19 1.00 I

COMMON BELIEFS SURVEY HI

PAST BLAM E SELF APPROVAL PERFECT CONTROL PRONE DOWN

IM PORTANCE O F PAST 1.00 | BLAM E PRONENESS .01 1.00 1 SELF DOW NING .33 .09 LOO IM PORTANCE APPROVAL .09 -.16 -.05 LOO PERFECTIONISM .27 .31 .57 -.23 1.00 CO N TR O L O F EMOTIONS -.08 .03 -.15 .48 -.38 1.00

R e p rod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

TABLE 15 (CONTINUED)

SURVEY OF PERSONAL BELIEFS AND COMMON BELIEFS SURVEY ID

S U R V E Y O F P E R S O N A L B E L IE F S

A W FU LIZIN G SELF OTHERS LOW SELF SHOULD SHOULD FRUSTRA TIO N WORTH

C O M M O N B E L IE F S S U R V E Y I I I

IMPORTANCE O F PAST .13 .06 .10 .19 .13 BLAM E PRONENESS .32

ooo .29 .21 .1 1 SELF DOW NING .32 .23 .24 .48 .32 IMPORTANCE A PPR O V A L .13 .19 .13 .04 .23 PERFECTIONISM .26 .26 .43 .46 .30 CONTROL O F EM OTIONS .06 -.04 -.09 .15 -.04

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

TABLE 16

R elationship Between C o gnitive D istortions and A nger Scores (Tull Sample)

Survey of Personal Beliefs

Anger Scale F Score Distortions % o f variance in Regression Equation

T rait A nger F (1, 85) = 25.61** Low Frustration £ _ 22<7c A ngry T em peram ent F (1, 85) = 12.77** Low Frustration ^ I2 7 r

A ngry Reaction F (2, 84) = 14.51** Low Frustration; Awfulizins- f 247c i A nger-In F (1, 85) = 18.10** Low Frustration y Y17c !

A nger-O ut F ( 2 , 84) = 4.59* Low Frustration Other ShouIds(+> 8 %

A n g er C ontrol F (2, 84) = 7.16** Low Frustration (+): Awfulizing 137c

* £ = . 0 1 * * £ < . 0 1 (+) = correlation positive

Common Beliefs Survey HI

Anger Scale F Score Distortions % of variance in Regression Equation

T ra it A nger F (1, 85) = 7.32*** Self Downing

A ngry T em peram ent F (2, 84) = 3.42** Control of Emotions (+) Self Downing 5* A ngry R eaction F ( 1 ,8 5 ) = 7.21*** Self Downing

A nger-In F (1, 85) =9.10**** Self Downing 9 *

A nger-O ut F (1, 85)=4.09* Self Downing £■ - 37c A n g er C ontrol F ( 1 ,8 5 ) = 2.95 Self Downing (+)

* £ =.05 **£ <.05 * * * 2 =.01 ****£ < o i (+) = correlation positive

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

TABLE 17

Relationship B etw een C ognitive Distortions and A nger Scores (H igher T rait Anger)

Survey of Personal Beliefs

Anger Scale F Score Distortions % of variance in Regression Equation

T ra it A nger F (1, 85) = 25.61*** Low Frustration .£ ; 137c

A n g ry Tem peram ent F (2, 41) = 2.88 Low Frustration : Awfuiizing (+) A n g ry Reaction F (1, 42) = 6.82** Awfulizing - 12 %

A nger-In F (1, 42) = 7.65** Low Frustration . 13% A nger-O ut F (2, 41) = 3.42* Low Frustration c- t

Other ShouIds(+) 107c

A n g e r Control F (2, 41) = 3.98* Low Frustration (+); Awfulizing ' 12%

* £ < .05 **£ = .0 1 < o I (+) = correlation positive

Common Beliefs Survey HI

Anger Scale F Score Distortions % of variance in Regression Equation

T ra it A nger F ( 2 , 41) = 3.78* Self Downing: Approval (+) c ll<7c A n g ry Tem peram ent F ( 2 ,4 1 ) = 4.17* Control o f Emotions (+)

Self Downing t - 13%

A n g ry Reaction F (1, 42) = 2.27 Self Downing A nger-In F ( 1 ,4 2 ) =16.84** Self Downing ^ = 27%

A nger-O ut F ( 2 , 4 1 )= 3.95* Self Downing: Approval (+) > r 127c A n g er Control F ( 1,4 2 ) = 2.36 Self Downing (+)

* £ < .05 **£ < .0 1 (+) = correlation positive

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

R eproduced

with perm

ission of

the copyright

ow ner.

F urther

reproduction prohibited

w ithout

perm issio

n .

□ P A R E N T C H IL D C E N T E RFIGURE 1: NUMBER OF SUBJECTS AT EACH LOCATION

35 □ O B T A IN E D T H R O U G H

F R IE N D S (C anada)

35-1 □ M E N T A L H E A LTH C E N T E R

E M P L O Y E E S

30- □ O B T A IN E D T H R O U G H

F R IE N D S (U S A )

25-

2 0 -

15-

10 -

5 -

LOCATIONS

R eproduced

with perm

ission of

the copyright

ow ner.

F urther

reproduction prohibited

w ithout

perm issio

n .

FIGURE 2: EDUCATION OF SUBJECTS

35

30

25

20

Number of Subjects

15

10

■•■I 111 2:;:::::::: ::::::::: a u iiu a i

Grade School High School Community College College/Universily Posl Grad

School Level Completed

^

R eproduced

with perm

ission of

the copyright

ow ner.

F urther

reproduction prohibited

w ithout

p e

rm issio

n .

FIGURE 3: COGNITIVE DISTORTIONS PREDICTING TRAIT ANGER

L o w Frustration S e lf D ow nin g

R eproduced

with perm

ission of

the copyright

ow ner.

F urther

reproduction prohibited

w ithout

perm issio

n .

FIGURE 4: COGNITIVE DISTORTIONS PREDICTING ANGRY TEMPERAMENT

30- '

25-

Low Frustration Control of Emotions C ontrol & S e lf D ow ning

FI G

UR E

5: C

O G

N IT

IV E

D IS

TO R

TI O

N S

PR ED

IC TI

N G

AN

G RY

R

E A

C TI

O N

■iiiiiihSiSiSiSi!

w h iW iw ! 1

i oco LOCM oCM LD in

aoueuBA s6eiuaojsd

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

Lo w

Fr us

tr at

io n

A w

fu liz

in g

& Lo

w Fr

us tr

at io

n S

el f

D o

w n

in g

0454373997

R eproduced

with perm

ission of

the copyright

ow ner.

Further reproduction

prohibited w

ithout p

erm issio

n .

FIGURE 6: COGNITIVE DISTORTIONS PREDICTING ANGER-IN

30-

25-

2 20 - c .52‘u(0 >

d)u> <0 Ca>ol_0) O.

15-

1 0 -

w m m m

Low Frustration S elf D ow ning

25015^511673^^

R eproduced

with perm

ission of

the copyright

ow ner.

Further reproduction

prohibited w

ithout p

erm issio

n .

FIGURE 7: COGNITIVE DISTORTIONS PREDICTING ANGER-OUT

Low Frustration Self Downing

FI G

UR E

8: C

O G

N IT

IV E

D IS

TO R

TI O

N S

PR ED

IC TI

N G

A

N G

E R

-C O

N T

R O

L

in

I o 00

LD CM

I o CM

LO

aoueuBA P a6Biuaojad

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

Lo w

Fr us

tr at

io n

Lo w

Fr us

tr at

io n

& A

w fu

liz in

g

77373079^39452

^

R eproduced

with perm

ission of

the copyright

ow ner.

Further reproduction

prohibited w

ithout p

erm issio

n .

FIGURE 9: COGNITIVE DISTORTIONS PREDICTING TRAIT ANGER (HIGHER ANGER GROUP)

30-

25-

W / / A

Low Frustration Sell Downing Sell Downing and Approval

R eproduced

with perm

ission of

the copyright

ow ner.

Further reproduction

prohibited w

ithout p

erm issio

n .

FIGURE 10: COGNITIVE DISTORTIONS PREDICTING ANGRY TEMPERAMENT (HIGHER ANGER GROUP)

3 0 -

2 5 -

• 2 ° -oc .55 TO > •R 1 5 -

ca>ob.o QL

1 0 -

5 -

0 -

C ontrol of E m otions S e lf D ow nin g an d C ontrol

R eproduced

with perm

ission of

the copyright

ow ner.

Further reproduction

prohibited w

ithout p

erm issio

n .

FIGURE 11: COGNITIVE DISTORTIONS PREDICTING ANGRY REACTION (HIGHER ANGER GROUP)

Awfulizing

R eproduced

with perm

ission of

the copyright

ow ner.

Further reproduction

prohibited w

ithout p

erm issio

n .

FIGURE 12: COGNITIVE DISTORTIONS PREDICTING ANGER-IN (HIGHER ANGER GROUP)

29

L ow Frustration S e lf D ow ning

R eproduced

with perm

ission of

the copyright

ow ner.

Further reproduction

prohibited w

ithout p

erm issio

n .

FIGURE 13: COGNITIVE DISTORTIONS PREDICTING ANGER-OUT (HIGHER ANGER GROUP)

3 0 - i

2 5 -

« 2 0 - oc .2u<0 >

5 15- O) ra<¥■*c.Q)O <u a 1 0 -

L o w Frustration O ther Shoulds & Low Frustration

S elf D ow ning A pp ro val & S e lf D ow nin g

FI G

U R

E 14

: C

O G

N IT

IV E

D

IS TO

R TI

O N

S

PR ED

IC TI

N G

A

N G

E R

-C O

N TR

O L

(H IG

H ER

A

N G

E R

G R

O U

P )

s r ~

' ~ &Ciu

o CO

LO C\J

LOs aoueuBA io a6ejuaojad

uo

R eprod u ced with perm ission o f the copyright ow ner. F urth er reproduction prohibited w ithout perm ission.

Lo w

Fr us

tr at

io n

A w

fu liz

in g

& Lo

w F

ru st

ra ti

o n