Citation
Forgiveness in Children: The Child/Adolescent Dispositional Forgiveness Inventory

Material Information

Title:
Forgiveness in Children: The Child/Adolescent Dispositional Forgiveness Inventory
Creator:
LEEVER, BROOKE A. ( Author, Primary )
Copyright Date:
2008

Subjects

Subjects / Keywords:
Adolescents ( jstor )
Child psychology ( jstor )
Factor analysis ( jstor )
Forgiveness ( jstor )
Psychological research ( jstor )
Psychology ( jstor )
Psychometrics ( jstor )
Social desirability ( jstor )
Social psychology ( jstor )
Transgression ( jstor )

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Copyright Brooke A. Leever. Permission granted to University of Florida to digitize and display this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Embargo Date:
8/31/2006
Resource Identifier:
476200338 ( OCLC )

Downloads

This item has the following downloads:


Full Text












FORGIVENESS IN CHILDREN: THE CHILD/ADOLESCENT DISPOSITIONAL
FORGIVENESS INVENTORY















By

BROOKE A. LEEVER


A THESIS PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
MASTER OF SCIENCE

UNIVERSITY OF FLORIDA


2006









































Copyright 2006

by

Brooke A. Leever















ACKNOWLEDGMENTS

I first wish to thank my family and friends who continually offer me support and

have never failed to believe in me. They mean more to me than they know.

I am extremely grateful to my mother, Brenda Leever, for her dedication and

helpful coordination efforts with this project. I would also like to thank my advisor, Dr.

Ken Rice, for all of his time and wealth of statistical knowledge without which this

project could not have been accomplished.
















TABLE OF CONTENTS



ACKNOWLEDGMENT S .......... .... ............. iii
A BLE..........CK N OW LED GM EN TS ....................................... ............................................. vi


ABSTRACT ............... ............................................ vii

CHAPTER


1 IN TRODU CTION ................................................. ...... .................

2 LITERATURE REVIEW ........................................................................8

Forgiveness and Forgiveness M easures.................................... ........................ 8
C children and A adolescents ................................................... ................... ............... 13
The H eartland Forgiveness Scale ........................................ .......... ............... 17
Research Questions and Hypotheses ........................................ ....... ............... 20
S social D esirab ility ...............................................................22

3 M E T H O D S ........................................................................................................... 2 4

P a rtic ip a n ts ........................................................................................................... 2 4
P ro c e d u re ..................................................................2 5
M e a su re s ............... ...... ............. .... ................ ... ... ... ........... ............... 2 5
Child/Adolescent Dispositional Forgiveness Inventory (CADFI) ...................26
Children's Attributional Style Questionnaire-Revised (CASQ-R) ...................26
Children's Depression Inventory-Short form (CDI-S) .......................................28
Lie Scale from the Revised Children's Manifest Anxiety Scale (RCMAS) .......28

4 R E S U L T S ............................................................................. 3 0

In stru m e n t ...................................................................................................... 3 0
D descriptive Statistics .............................................................30
A n a ly se s .................... ... ......................................... .................... ............. 3 0
Confirm atory Factor A analysis ................................................................... 31
E xploratory F actor A analysis .......................................................................... 32
T h ree F actor M o d el ....................................................................................... 32
Parallel A nalysis................................................... 34










Tw o Factor M odel .......................................... ............. .... ....... 35
Internal C consistency ................................................ .......... ..... .. .......... 36
Convergent, Discriminant, and Construct Validity .......................................... 36
Q ualitative R esponses............ ... ...................................................... ...... .... .....38

5 D ISC U S SIO N ............................................................................... 4 1

APPENDIX

A CADFI ITEMS AND FACTOR LOADINGS ................................. ...............45

B QUALITATIVE QUESTIONS ............................................................................46

C QUALITATIVE RESPONSES ............................................................................47

R E F E R E N C E S ................................................................ 5 1

B IO G R A PH IC A L SK E TCH ..................................................................... ..................57





































v
















TABLE

Table page

1 Correlations Between All Scale and Subscale Scores............ .............38















Abstract of Thesis Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Master of Science

FORGIVENESS IN CHILDREN: THE CHILD/ADOLESCENT DISPOSITIONAL
FORGIVENESS INVENTORY

By

Brooke A. Leever

August, 2006

Chair: Kenneth G. Rice
Major Department: Psychology

The current study investigated dispositional forgiveness in the child and adolescent

population. Specifically, the factor structure, internal consistency reliability, and validity

of a modified version of the Heartland Forgiveness Scale (HFS), entitled the

Child/Adolescent Dispositional Forgiveness Inventory (CADFI), were examined in a

sample of children and adolescents. Reliability coefficients were found to be adequate

with this sample, and relationships with constructs of validity were found to be solid as

well. In particular, the relationship of forgiveness with depressive symptoms (as

measured by the Child Depression Inventory-Short Form; CSI-S), attributional style (as

measured by the Children's Attributional Style Questionnaire Revised; CASQ-R), and

social desirability (as measured using the Lie Scale of the Revised Children's Manifest

Anxiety Scale) was explored. Results support relationships in the expected directions

between forgiveness and the chosen measures. Internal consistency reliabilities were









reported for all scales and factor structure was also examined and a proposed two

factor model provided. Directions for future research are included.














CHAPTER 1
INTRODUCTION

Though no clear-cut consensus exists among researchers as to the definition of

forgiveness, most definitions include a key component of renouncing anger and

resentment. Forgiveness of another does not necessarily involve reconciliation, though

that may occur as part of the process (see Thompson & Snyder, 2003 for descriptions).

McCullough et al. (1998) assert that forgiveness must include pro-social changes in

motivation, and they provide criteria for distinguishing when these changes have

occurred. Thus, they consider the important aspect to be forgiveness as a process

involving interpersonal relationships. Thompson and Snyder (2003) and Thompson et al.

(2005) emphasize that a forgiver relinquishes the negative attachment to a transgression,

transgressor, or sequence of events involved in a transgression. Negative thoughts,

feelings, or behaviors are all contained within the concept of negative attachment, with

the eventual goal of forgiveness being a more positive or at least neutral reframing, which

results in a weakening of the attachment. Hargrave and Sells' (1997) conceptualization

provides support within itself for the restorative, healthy component of forgiveness; they

describe it as an effort to restore love and trustworthiness for relationships in order for

victims and victimizers to put an end the cycle of destructive entitlement. Finally, Hill

(2001) offers what seems to be one of the most encompassing and noteworthy

descriptions, effectively articulating the essential depth of the forgiveness construct and

the beneficial nature of forgiveness by describing it as being more of a discovery rather

than an act of will.









Within the realm of dispositional measures, only one covers what appears to

encompass the three main aspects of the forgiveness construct: the Heartland Forgiveness

Scale (HFS), measuring Forgiveness of Self, Forgiveness of Others, and Forgiveness of

Situations. These authors use a definition of forgiveness in which forgiveness occurs

primarily through a weakening of the negative attachment to a transgressor,

transgression, or some compilation of the two (Thompson & Snyder, 2003; Thompson et

al., 2005). The negative attachment may then transform into positive or neutral, providing

a mechanism for the forgiveness process to unfold.

A review of current forgiveness literature reveals that most researchers view

forgiveness as adaptive (e.g. Mauger et al., 1992; McCullough, 2000; Thompson &

Snyder, 2003; Thompson et al., 2005). Forgiveness has been linked to psychological

health and well-being in general (e.g. Mauger et al, 1992; Enright, 2000; Scobie &

Scobie, 2000; Worthington, Sandage, & Berry, 2000; Ripley & Worthington, 2002;

Thompson & Snyder, 2003), while unforgiveness has been correlated with higher

psychopathology (e.g. Mauger et al., 1992; Maltby, Macaskill, & Day, 2001;

Worthington, Mazzeo, & Klewer, 2002). At the theoretical and philosophical level,

forgiveness has been seen as evidence of one's objectivity, a demonstration of acceptance

of one's embeddedness in the world and of freedom, with it forgiveness and self-

forgiveness come (Cavell, 2003). Unforgiveness has not been clearly defined but

typically has been thought of as a lack of forgiveness. Unforgiveness has been attributed

to a lack of letting go of negative attachment to events, exemplified by behaviors such as

repeated ruminations, or a pervasive negative affect. Attempts to reduce unforgiveness,

often due to repeated ruminations, can lead to the presence of addictive behaviors and









psychological defense mechanisms. Unforgiveness, as part of a person's internalized

negative affect, may become detrimental both by causing addictive behaviors as well as

by being a result of such addictive behaviors (Worthington et al., 2002). Contrarily, a

handful of researchers believe forgiveness may not be beneficial because it may leave the

forgiver open to further victimization or victim-blaming (e.g. Bass & Davis, 1994; Katz,

Street, & Arias, 1997). However, the forgiving of others is not the only aspect of

forgiveness argued to be beneficial. Witvliet, Ludwig, and Bauer (2002) demonstrated the

powerfully beneficial aspects of being forgiven by a victim as well. After imagery

conditions in which the subjects imagined being forgiven, significant improvements were

reported by participants in basic emotions such as anger and sadness, moral emotions

such as guilt, shame, gratitude and hope, as well as perceived interpersonal forgiveness

(Witvliet et al., 2002). Discrepancies as to the beneficial nature of forgiveness exist

mainly due to differences in the conceptualizations and definitions of forgiveness chosen.

Current research thoroughly conveys the multidimensionality of the construct of

forgiveness. The typical notion that forgiveness only occurs from one person to another

interpersonallyy) has lead to a predominance of research examining transgression specific

forgiveness, or the extent of one's forgiveness specifically toward another person (e.g.

Hargrave & Sells, 1997; Enright, Freedman, & Rique, 1998; Tangney et al., 1999;

McCullough, 2000; McCullough et al., 2000; Witvliet et al., 2000; Wade & Worthington,

2003). Transgression specific forgiveness relates to a particular event, and therefore may

not necessarily be predictive of a person's willingness to forgive across situations and

contexts. On the other hand, dispositional forgiveness measures attempt to remedy this

limitation by emphasizing that the development of a forgiveness measure holding









predictive qualities remains key. Knowledge as to a person's willingness to forgive

across various circumstances and events reveals more about the presence of that trait

within an overall person, viewing forgiveness more expansively.

Assessing dispositional forgiveness across contexts (e.g., Berry et al., 2001) provides

a beneficial alternative to transgression specific assessments. Scores on measures of

dispositional forgiveness tend to be specifically related to mental health and well-being,

whereas scores on transgression specific measures do not tend to be (Thompson &

Snyder, 2003). Furthermore, measures of forgiveness that are dispositional in nature,

rather than event specific, provide a better basis for comparison with psychological

correlates of the forgiveness construct (Thompson & Snyder, 2003). Dispositional

forgiveness, as opposed to transgression specific forgiveness, offers a broader window

into forgiveness as a whole, since it gives due credence to the multifaceted nature of the

forgiveness.

The dispositional forgiveness construct falls within differing domains, or subscales.

Forgiveness may occur interpersonally, as exemplified through the transgression specific

model, as well as intrapersonally, as revealed through one's willingness to forgive

oneself. More recently, interest has begun to arise in a third area of forgiveness, that of

forgiveness of situations. Forgiveness of situations has been introduced as specifically a

dispositional aspect of the forgiveness construct and has been assessed along with the

previously established interpersonal and intrapersonal domains. Measures with reliable

scores have been created to include forgiveness of oneself (Mauger et al., 1992;

Worthington et al., 2002; Macaskill, Maltby, & Day, 2002) and forgiveness of others,









along with forgiveness of situations beyond the control of either self or others (Thompson

& Snyder, 2003).

The distinction of a construct along interpersonal as well as intrapersonal lines leads

to a better understanding of the depth of this construct of interest and a more specific

view into its dimensional nature. The construct of forgiveness clearly contains

interpersonal as well as intrapersonal components, existing as distinct yet related

concepts. As would naturally seem to follow, the psychopathologies associated with

forgiveness and unforgiveness appear along similar interpersonal and intrapersonal lines.

A distinction in related pathologies exists depending on the type of forgiveness (or lack

of) being assessed. Mauger's Forgiveness of Self (FS) and Forgiveness of Others (FO)

scales have been correlated with the Minnesota Multiphasic Personality Inventory

(MMPI), revealing a significant association of failure to forgive with negative emotional

states, such as depression, anxiety, anger/distrust, and negative self-esteem. Lack of

forgiveness has also been specifically linked to eating disorders (Worthington et al.,

2002). Personality factors have been found to highly correlate with dispositional

forgiveness of both self and others in the expected directions, especially for the category

of neuroticism versus emotional stability, in which lower levels of forgiveness related to

higher neuroticism (Walker & Gorsuch, 2000). Failure to forgive self has been

connected to more intra-punitive pathologies, such as anxiety and depression, while

failure to forgive others has been associated with extra-punitive pathologies, such as

social alienation, social introversion, depression, and psychoticism (Maltby, Macaskill, &

Day, 2001). Mauger and colleagues (1992) also found deficits in forgiveness of self and









others to be related to higher amounts of overt psychopathologies, for example

schizophrenia.

As is apparent, depression has been previously found to be a significant related

pathology of failure to forgive self and others, and therefore depression may represent a

manifestation of both intra- and extra-punitive forgiveness failures. That is, depression

appears to be a main component of the ways in which failure to forgive may manifest and

influence a person's psychological well-being, and therefore depression is important to

consider, theoretically and practically, in forgiveness research. As will be discussed,

researchers who developed the Heartland Forgiveness Scale (HFS) report a similar

relationship regarding the association of depression with lack of dispositional

forgiveness, manifest as a significant negative correlation. Following this premise, the

current study plans to examine correlation of the Child/Adolescent Dispositional

Forgiveness Inventory with related constructs of interest such as a child measure of

depression.

Developments in adult forgiveness research include broadening the scope of the

forgiveness construct to be inclusive of forgiveness of self, forgiveness of others, and

forgiveness of situations domains (i.e. Thompson et al., 2005). Evaluation of forgiveness,

or lack thereof, has been increasing in emphasis on dispositional factors for measurement

rather than specific event examples. The use of a dispositional measure lessens the

difficulties and ambiguities of the relationship of forgiveness to similar psychological

constructs. Correlations and conclusions made under a transgression specific framework

by nature impose stricter circumstances under which forgiveness frequency and behavior

may be predicted. Consistent with these trends, a main purpose of the current study was









to create a measure to meet the needs of the younger age range while retaining the

integrity of the dispositional forgiveness construct and keeping up with current trends in

this area of research. The creation of a dispositional measure of forgiveness appropriate

for the child/adolescent population follows in the footsteps of previous adult dispositional

measures, allowing for a dynamic insight into the nature of forgiveness across a broad

age range. Such a measure is necessary to examine the psychological correlates of

forgiveness across a broader age span, as well as to open examination of the possible

developmental influences on this construct. The study of forgiveness development as a

dispositional trait provides a means to possible illumination of implications this construct

may have on later psychological functioning. Future research in this direction is

predicated on having a useful dispositional measure of child/adolescent forgiveness

available.














CHAPTER 2
LITERATURE REVIEW

Forgiveness and Forgiveness Measures

There has been much debate as to the conceptualization of forgiveness and its aspects.

Most consider forgiveness of transgressions interpersonally to be the forgiveness

representative. However, approaches to measuring forgiveness range from presenting

events of specific transgressions hypothetically committed towards a respondent and

asking their propensity to forgive such an event, to more broad statements intended to

gauge one's disposition to forgive things beyond one's control.

Mauger and colleagues (1992) created the Forgiveness of Self (FS) and Forgiveness

of Others (FO) scales, two 15-item scales endorsed as true-false by respondents. These

scales provide a solid basis for emphasis and distinction of two different subtypes, which

fall under the greater umbrella of the construct of forgiveness. These scales tap the failure

to forgive others and failure to forgive oneself, which may more accurately characterize

them as measures of unforgiveness than forgiveness. The Forgiveness of Others items

revolve around "taking revenge, justifying retaliation and revenge, holding grudges, and

seeing other people as apt to cause one hurt," while forgiveness of self items involve

"feelings of guilt over past acts, seeing oneself as sinful, and having a variety of negative

attitudes towards yourself' (Mauger et al., 1992). These measures generally take the

approach of measuring forgiveness from the negative standpoint, in that they appear to be

tapping unforgiveness as indication of one's lack of forgiveness of self or others,

implying unforgiveness and forgiveness would exist on a continuum in which one is said









to have more or less forgiveness regardless of the negative or positive standpoint of the

measure.

In the transgression-specific vein, Enright and his colleagues developed the Enright

Forgiveness Inventory (EFI; Subkoviak et al., 1995) and the Willingness to Forgive scale

(WTF; Hebl & Enright, 1993). Enright subscribed to the mindset that forgiveness is

something that occurs between two people, excluding the idea of forgiveness of

situations. Thus, their definition is "a willingness to abandon one's right to resentment,

negative judgment, and indifferent behavior toward one who unjustly hurt us, while

fostering the undeserved qualities of compassion, generosity, and even love toward him

or her" (Enright et al., 1998, pp. 46-47). This definition clearly exhibits the necessity for

benevolence in order for forgiveness to occur, and Enright and Zell (1989) also assert

that, "the fruition of forgiveness is entering into loving community with others" (p.99),

suggesting that reconciliation may be a part of forgiveness as well. The focus of this scale

is, again, one specific transgression. Though respondents are asked to respond to

hypothetical as well as one real situation, the narrow focus of a transgression-specific

scale makes the correlation of forgiveness with related psychological constructs.

The EFI consists of 60 items on a 6-point Likert scale assessing six forgiveness

dimensions including positive affect, behavior, and cognitions, and the lack of negative

affect, behavior and cognitions toward the offender. Respondents are supposed to think of

the most recent transgressor or transgression as such when responding to the specific

transgressions. The summed total of scores on the items of the subscales yields an overall

forgiveness score. Relatedly, the WTF is a 16-item scale intended to measure the

effectiveness of forgiveness as a therapeutic interpersonal intervention. The first 15 items









include specific scenarios, while the last item addresses the person's specific

transgression of focus. Respondents choose from ten responses to each scenario, one

choice of which is forgiveness. The scale is scored based on how many times forgiveness

is selected as a solution for when a person would respond with and prefer to respond with

forgiveness to a hypothetical scenario, as well as respond with and prefer to respond with

forgiveness to the real scenario.

Barry, Worthington, and colleagues also created the Transgression Narrative Test of

Forgiveness (2001), a five scenario scale purported to measure the dispositional tendency

to forgive interpersonal transgressions over time and across situations, which they term

forgivingnesss". Participants read each scenario and respond using a 5-point Likert scale

as to their likelihood of forgiving the transgression were they in that situation. The

situations include two intentional transgressions by acquaintances, two negligent

transgressions by friends, and one intentional transgression by a relative followed by an

apology (Barry et al., 2001).

Others have taken the approach of developing a scale to measure the precise

components of their definitions of forgiveness. Similarly to Enright, McCullough (2000)

and colleagues (McCullough et al., 1998) conceptualize forgiveness as a construct

emphasizing prosocial changes in motivations. Under their definition one specifically

experiences decreased motivation to avoid personal and psychological contact with the

offender, to seek revenge or to see harm come to the offender, and increased benevolence

motivation toward the offender, implying that forgiveness of others is the basic

component. Under this framework, again, forgiveness is emphasized as an interpersonal

process rather than one that may occur intrapersonally.









McCullough et al.'s measure is the Transgression-Related Interpersonal Motivations

Inventory (TRIM; 1998), consisting of 12 items on a 5-point Likert scale, with two

subscales measuring two components of their definition, motivations to avoid contact,

and motivations to seek revenge or see harm come to the transgressor. It does not

however measure motivation toward benevolence, and as such appears to indicate mostly

the negative aspects, namely unforgiveness, as representative of one's lack of

forgiveness, which is fairly similar to Mauger's approach.

Hargrave and Sells (1997) also follow the idea of creating a scale to measure aspects

of their specific definition. They offer a definition of forgiveness that includes the

component of love, "effort in restoring love and trustworthiness to relationships so that

victims and victimizers can put an end to destructive entitlement" (p.43). They assert two

broader levels of forgiveness, exonerating (involving insight and understanding) and

forgiving (involving the overt act of forgiving as well as giving the opportunity for

compensation, meaning the ability for interactions with the transgressor in away

perceived by that person as non-threatening and encouraging of emotional bonding).

The Hargrave and Sells (1997) Interpersonal Relationship Resolution scale (IRRS)

taps forgiveness of a specific person who has causes hurt to the respondent. It consists of

two scales, forgiveness and pain, and has 22 total items and four subscales including

insight, understanding, giving the opportunity for compensation, and the overt act of

forgiving.

Tangney and colleagues (1999) have developed another measure of forgiveness that

does not include the components of love or compassion as necessary for forgiveness.

Specifically, their lengthy definition mainly emphasizes that giving up the negative









emotions is what is sufficient to constitute forgiveness. Their measure, the

Multidimensional Forgiveness Inventory (MFI), is a scenario-based approach to

measuring forgiveness of self and others, which consists of 72 items on a 5-point Likert

scale using 16 different transgression scenarios. There are nine subscales, including

propensity to forgive others, propensity to forgive self, propensity to ask for forgiveness,

time to forgive others, time to forgive self, propensity to blame others, propensity to

blame self, sensitivity to hurt feelings, and anger-proneness. This transgression-specific

measure attempts to tap many facets of forgiveness and includes both forgiveness of self

and forgiveness of others components, creating a much wider scope under which to

examine forgiveness through a transgression-specific lens. However, it is still that -

transgression specific. Examination of the construct of forgiveness using specific

descriptions of events still does not allow for a broader generalizability of forgiveness as

attributed to the respondent as well as to be related to other psychological constructs.

The current study argues that the superior approach to measurement of forgiveness is

both through a non trangression-specific framework, and is incorporative of the self,

others, and situational components of the forgiveness construct, in line with Snyder and

Yamhure Thompson's definition of forgiveness:

We define forgiveness as the framing of a perceived transgression such
that one's attachment to the transgressor, transgression, and sequelae of
the transgression is transformed from negative to neutral or positive. The
source of a transgression, and therefore the object of forgiveness, may be
oneself, another person or persons, or a situation that one views as being
beyond anyone's control (e.g. an illness, "fate", or a natural disaster).
(Thompson & Snyder, 2003, p. 302)

This definition appears as the broadest and most encompassing definition so far in

that it includes the possibility of feeling transgressed upon in an impersonal manner, and









thus allows for the forgiveness of situations to exist as a distinct and important aspect of

dispositional forgiveness. The authors' most compelling argument for the understanding

of forgiveness as dispositional rather than transgression-specific is the assertion that

forgiveness itself occurs only intrapersonally. In other words, the target of forgiveness

does not matter; it is the motivation and behavior of the forgiver that exemplify

forgiveness dispositionally. This measure created by Yamhure Thompson and colleagues,

the Heartland Forgiveness scale (HFS; Thompson et al., 2005) that measures forgiveness

of self, others, and situations, is elaborated on in detail later in this section.

Children and Adolescents

The current breadth of literature in the area of dispositional forgiveness begs a more

balanced examination of the forgiveness construct across a broader age range. The

overwhelming majority of research on the construct of forgiveness focuses on adults.

Identification of dispositional failures to forgive in adults inevitably points to the

important question of developmental implications of forgiveness. In other words, the

questions of whether a younger age range will exhibit similar forgiveness trends as

adults, as well as how stable the construct may be with age are imperative to

understanding how forgiveness develops. Future research in this direction requires a

comprehensive scale of dispositional forgiveness appropriate for a younger age range. As

well, a focus on children and adolescents using a dispositional, rather than event-specific,

lens will help shed light on the developmental nature of the forgiveness construct.

Children's understanding of the concept of forgiveness under different circumstances

has been previously evaluated, and development of the concept of forgiveness is

generally considered to occur at a young age. Darby and Schlenker (1982) evaluated

children's understanding of forgiveness and such adult social judgments by examining









child judgments when observing an actor apologizing. Using children's reactions to

apologies, children as young as kindergarten age were shown to be able to comprehend

the major factors of importance (e.g., intentions, motives, apologies) for adult-like social

judgment involving forgiveness, and younger children were found to exhibit similar

judgments to older children (Darby & Schlenker, 1982). These apologies varied with

respect to complexity and circumstances (i.e., intentional versus accidental transgression,

good or bad motives, and high or low responsibility and consequences).

Though this study helps shed light on children's abilities to wrap their minds, and

hearts, around the construct of forgiveness, the experimental and transgression-specific

nature of the study precludes dispositional insights into the children's constructs of

forgiveness in a more predictive sense, and across various dimensions. Forgiveness of

others, though overwhelmingly the most frequently studied aspect of forgiveness, does

not represent the whole of the forgiveness construct. Furthermore, transgression

specificity limits the degree to which examination of the forgiveness construct as a whole

can occur.

Park and Enright (1997) proposed and supported a developmental progression of the

understanding of forgiveness with adolescents in Korea, using junior high students ages

13-14. Specifically, they outline the development from the most basic level, that of

revengeful forgiveness manifested as words or gestures indicating forgiveness while

internal hostility may still exist and be openly expressed, to external forgiveness in

which external gestures indicate forgiveness but awareness exists of inner frustrations

and continuing frustrations are suppressed, to the more advanced level, that of internal

forgiveness where one genuinely seeks to understand the motives and perspective of









the other person and reinterprets the event incorporating both perspectives. This last level

is again said by the authors to arise out of the principles of beneficence and love. Though

beneficence and love are generally articulated in the views of forgiveness, this model also

add credence to the argument of Yamhure Thompson and Snyder, that forgiveness at the

highest and most comprehensive level occur strictly intrapersonally, regardless of the

target of forgiveness. This is to say that unforgiveness as a result of feeling transgressed

upon is something that occurs within a person though may be a result of events from

one's self, another person, or a situation. Thus, motivation for forgiveness can only occur

from within a person in the same manner. This intrapersonal emphasis seems to be

exactly what Park and Enright have supported in their model, although their model

focused upon forgiveness of others.

They also report a significant positive correlation between an adolescent's

understanding of forgiveness and age, meaning an adolescents' understanding of

forgiveness was significantly and positively associated with age; older adolescents were

more forgiving than younger adolescents. This study also revealed that the higher the

level of understanding of forgiveness, the more likely a person was to use a proactive

restorative strategy in the actual conflictual relationship in the real life situation. The

degree to which the offering of social support was a factor followed a similar pattern, in

that the early adolescents appeared to be in transition to this pattern while the later ones

seemed either deeply rooted in or transitioning out of social support as a factor. Thus,

significance of social support seems to be related to level of understanding attained.

Though narrow in scope and sample, this study shows that developmental trends exist in

the concept of forgiveness. Developmental knowledge as well as measurement









understanding of forgiveness can in turn provide an important background for working

with forgiveness in the clinical settings with any age range.

In the realm of clinical practice, models have begun to be established for forgiveness.

Worthington and Wade (1999) have set forth a model incorporating forgiveness and its

related areas, in which the personal attributes of the participants and of the relationship at

hand before a transgression factor into the propensity and likelihood of forgiveness. First

the person's perception of the events take place, and their initial emotional reaction to the

events occurs. Then interpersonally active responses such as revenge/retaliation, pro-

relationship behavior, and perception of the offender's response would occur as well as

the interpersonally passive response of rumination. Within this model, an emotionally

dissonant event may occur that changes one's initial reaction to the event, leading to

emotional dissonance. The authors assert that this aspect is crucial. "Underlying

forgiveness is an emotional dissonant event. Thus, a victim's ability to forgive will be

influenced partially by his or her ability to comprehend and successfully resolve

incompatible emotions, which is the core of the hypothesized construct of emotional

intelligence" (Worthington & Wade, 1999, p. 395).

Other models of forgiveness have also been proposed within the psychotherapeutic

context. Rosenak and Harnden (1992) describe a model in which, after an offensive

event, victims experience hurt that leads to anger. After the realization and expression of

hurt and anger comes an information gathering stage that occurs in therapy in which the

client needs to glean more knowledge about the transgressor. This information is thought

to promote a better understanding from that person's perspective, leading to true empathy

and in turn true forgiveness.









The Heartland Forgiveness Scale

In developing a dispositional forgiveness measure appropriate for children and

adolescents, an existing well-rounded dispositional measure of forgiveness in the adult

population was sought. Modeling after a reliable pre-existing adult measure of the same

construct appears often as a useful approach to developing a complementary

child/adolescent measure for a construct of interest, especially when there is a compelling

reason to suspect construct equivalence between children and adults. For instance, in

related areas of psychological constructs, the Child Adolescent Perfectionism Scale

(CAPS; Flett et al., 2001) was modeled after the Multidimensional Perfectionism Scale

(MPS; Hewitt & Flett, 1991) that had been developed for adults, and the Children's

Depression Inventory (CDI; Kovacs, 1985) was patterned after the framework of the

Beck Depression Inventory (BDI; Beck & Steer, 1987). Similarly, the measure created in

this study, the Child/Adolescent Dispositional Forgiveness Inventory (CADFI) was

designed as an age-appropriate modification from the Heartland Forgiveness Scale

(HFS), targeted for a younger age range.

The Heartland Forgiveness Scale (HFS; Thompson & Snyder, 2003) was chosen

to serve as the model in this study mainly because of its encompassing nature for the

construct of forgiveness. Under this framework, someone can be relieved of the negative

attachment through transforming the negative cognitions, emotions, and/or behaviors to

either neutral or positive, as well as through weakening the attachment by releasing the

perception of a strong connection between oneself and the transgressor or transgression

(Thompson & Snyder, 2003). This does not necessitate a forgetting of the event, nor

preclude the person from taking any actions, legal or otherwise. The emphasis is on the

valence of the attachment, and forgiveness is said to occur as long as the motivation does









not involve negative attachment (e.g., vengefulness). Instead, forgiveness is understood

as a process "through which people synthesize their prior assumptions and the reality of

the transgression into a new understanding of the transgression, transgressor,

transgression sequelae, and potentially, of themselves, other people, or the world"

(Thompson et al., 2005, p. 318). The authors are therefore unique in their perspective that

only the shift from a negative to a neutral attachment is necessary and sufficient to

constitute forgiveness, rather than needing empathy or compassion to be demonstrated.

The HFS was developed with adults to assess the dispositional tendency to forgive.

The HFS includes 18 items, endorsed on a 7-point Likert Scale ranging from almost

always false of me to almost always true of me. Subscales included in the HFS are

forgiveness of self, forgiveness of others, and forgiveness of situations. Thus, for the

current purposes of developing a multi-dimensional forgiveness measure for children and

adolescents, this measure was considered the most comprehensive representational

instrument for the dispositional nature of forgiveness. This measure of forgiveness was

chosen because of its ability to tap interpersonal as well as intrapersonal forgiveness on a

dispositional level across different contexts; the HFS is the only existing forgiveness

scale which taps forgiveness of situations, one of the three main aspects of forgiveness

(self, other, and situational forgiveness).

Adequate internal consistency reliability of HFS scores has been supported, with

Cronbach coefficient alphas on the total scale ranging from .84 to .87, and alphas on the

subscales ranging from .71 to .83. Test-retest reliability of the HFS, on the basis of a 3-

week follow-up period, was .83 for the total score and ranged from .72 to .77 (Thompson

& Snyder, 2003) for the subscales. The psychometric properties of the HFS have been









evaluated and adequately demonstrated with student and nonstudent samples. The HFS

has been found to correlate in the expected directions with other dispositional measures

of forgiveness as well as measures of related psychological constructs. The HFS shows a

large positive correlation with the FS and FO scales, a smaller positive correlation with

the WTF scale, and a moderate/medium correlation with the MFI (Thompson & Snyder,

2003). The forgiveness of others subscale of the HFS is significantly negatively

correlated with one measure of forgiveness of a specific transgression, the TRIM, which

measures forgiveness of a transgression done specifically by another person (higher

scores mean higher levels of unforgiveness). The HFS was also found to significantly

positively correlate with the EFI transgression-specific measure. The HFS was not

significantly correlated with the IRRS, a measure of forgiveness towards a specific

person. Overall, the HFS has demonstrated much higher correlation with the dispositional

measures of forgiveness than the transgression-specific measures, although positive

correlation was found with all measures of forgiveness examined (Thompson, et al.,

2005).

The HFS was found to negatively correlate with scores on measures of some related

constructs as well. Yamhure Thompson and colleagues (2005) also found moderately

large negative correlations of the HFS with the Hostile Automatic Thoughts scale, the

rumination subscale of the Response Style Questionnaire (Thompson et al., 2005), and

the Beck Depression Inventory (Thompson & Snyder, 2003) as well as the Vengeance

Scale (Stuckless & Goranson, 1992), and the negative affect of the Positive and Negative

Affect Schedule (Watson, Clark, & Tellegen, 1988). Since forgiveness involves

transforming mental energy away from negative behaviors and thoughts, it makes logical









sense that such constructs as negative affect, vengeance, and rumination leading to

feeling depressed would correlate negatively with the HFS forgiveness measure.

The HFS was also evaluated to determine its predictive power of forgiveness in

romantic relationships. Participants were given the HFS as well as measures of hostility,

trust, relationship satisfaction, and relationship duration once and again nine months later.

Forgiveness as evaluated by the HFS was revealed as a stronger predictor of relationship

satisfaction than hostility, and satisfaction was also significantly predicted by how

trusting the person was of their partner and their perception of how trusting their partner

was of them (Thompson & Snyder, 2003, Thompson et al., 2005). All of these studies

support the premise that the HFS is a reliable, valid, and useful measure for evaluating

the dispositional tendency to grant forgiveness of self, others, and across situations.

Basing a children's measure of forgiveness on such a structurally and theoretically

supported measure as the HFS, which is defining current forgiveness research trends,

provides a sound foundation for examining forgiveness dispositionally with children.

Introducing a dispositional measure of forgiveness that is appropriate for children and

adolescents, when supported by adequate psychometric characteristics, contributes

significantly to the forgiveness literature by permitting future multidimensional,

dispositional forgiveness research with participants of a developmentally wider age range

than has historically been studied.

Research Questions and Hypotheses

Many parallels between the HFS and CADFI were expected. Specifically, it was

expected that the factor structure of the CADFI would reflect the HFS structure because

the CADFI will retain item content integrity of the HFS that is simply adjusted to be at a

lower, age-appropriate reading level. The underlying constructs of dispositional









forgiveness as evaluated by assessing forgiveness of self, forgiveness of others, and

forgiveness of situations were not expected to differ significantly for younger people.

It was expected that acceptable levels of internal consistency reliability and validity

would be reached, similar to the range of such alpha coefficients reported for the HFS.

The HFS has been positively correlated with scores from the Cognitive Flexibility

Scale (Martin & Rubin, 1995), the Dyadic Trust Scale (Larzelere & Huston, 1980), and

the distraction subscale of the Response Style Questionnaire (Nolen-Hoeksema &

Morrow, 1991).These measures are appropriate for adults but would not be appropriate

for children. Fortunately similar instruments have been developed for the childhood age

range and were targeted for validity purposes in the current study. Specifically, the

Children's Attributional Style Questionnaire-Revised (CASQ-R; Kaslow & Nolen-

Hoeksema, 1991) was used to assess the relation of the CADFI to styles of cognitive

attribution or explanation of positive and negative events, and the Children's Depression

Inventory (CDI; Kovacs, 1985) measuring depressive symptoms was also administered as

an expected negative correlate of CADFI scores.

Though the dimensions of the CASQ-R (global-specific, internal-external, and stable-

unstable) differ from the adult measure, higher levels of depression have been found for

those whose scores reflect a higher depressive attributional style, which on the CASQ-R

appears as more frequent endorsement of internal, stable, and global items for negative

events, and external, specific, and unstable items for positive events (Gladstone &

Kaslow, 1995; Thompson, Kaslow, Weiss, & Nolen-Hoeksema, 1998). In fact, Gladstone

and Kaslow (1995) report a negative correlation between the CASQ-R and the Vanderbilt

Depression Inventory (VDI) in non-clinical samples, indicating that the higher the









depressive symptoms reported the more depressive the attributional style. It was

therefore expected in the current study that higher scores on the CDI would correlate

similarly with depressive attributional style, as well as lower levels of forgiveness (low

scores on the CADFI).

Sex differences were not expected to differ significantly with regard to CADFI

scores.

Social Desirability

As was introduced by Crowne and Marlowe (1960; Marlowe & Crowne, 1961), a

tendency exists for respondents to report more socially desirable responses as a result of

being influenced by other factors besides the content of measures administered. Initially,

the bulk of the social desirability research fell within the adult age range. Crandall,

Crandall, and Katkovsky (1965) expanded social desirability research by developing the

Children's Social Desirability (CSD) questionnaires for a younger age range, targeting

children and adolescents. Social desirability has also been incorporated into measure of

other psychological constructs as well. Specifically, the Children's Manifest Anxiety

Scale (CMAS; Castenada, McCandless, & Palermo, 1956) and the Revised Children's

Manifest Anxiety Scale (RCMAS; Reynolds & Richmond, 1978) exemplify self-report

measures developed with the intent of measuring characteristics and manifestations of

anxiety in children, and include a Lie subscale specifically intended to address social

desirability.

Social desirability measures were established to evaluate increased predictability that

a respondent would endorse a higher degree of socially desirable responses on a measure.

Younger children, girls, non-white children, and less intelligent children have previously

been found to endorse socially desirable responses more frequently (Crandall et al., 1965;









Shriberg, 1974). These findings have been replicated and validated by Klein, Gould, and

Corey (1969), who found that younger children of both sexes had higher social

desirability scores than older children, and that this tendency was most marked when

respondents were younger than 11 years old. Moreover, they suggested this pattern

reflects an emergence of adult approval motivations in children beginning as young as 11

years old (Klein, Gould, & Corey, 1969).

Considering these variables, especially the finding that younger more than older

children report socially desirable answers, the inclusion of a social desirability measure in

the current study was necessary in order to ascertain whether answers regarding

dispositional forgiveness, as well as depression and attributional style, may be influenced

by factors other than substantive content. The current study employed the Lie subscale of

the RCMAS in order to address social desirability, while also keeping the number of total

items in the battery within reasonable length for efficient administration.














CHAPTER 3
METHODS

Participants

The total sample (N = 96) of participants consisted of boys and girls ages 8-17, with

the majority of the sample falling between the ages of 9 and 11 years old (77% of the

sample). The distribution of ages consisted of one 8-year old, twenty-one 9-year olds,

thirty 10-year olds, twenty-three 11-year olds, five 12-year olds, one 13-year old, two 14-

year olds, five 15-year olds, three 16-year olds, and one 17-year old (four participants

failed to report age). Sixty-five percent of the sample was female; ethnic background of

the sample was approximately 49% White, 20% Hispanic, 16% Asian, and 11% African

American. Consent forms were distributed to children attending public elementary

schools in a southern Florida county (approx. 86% of the sample), as well as a private

school within a central Florida County (approx. 14% of the sample). Response rate was

approximately 48%. The majority of the sample was representative of the population

sampled; the majority of the participants (73%) came from the fourth and fifth grades of

an elementary school with the following population demographics for those grades: 44%

male, 65% White, 11% African American, 10% Hispanic, and 2% Asian.

Rationale for inclusion of children and adolescents was twofold. Focus on

children/adolescents a) helps balance the level of attention given to forgiveness research

between adults and children, considering the comparative dearth of child/adolescent

research in the area, and b) provides context for future exploration of conception and

dispositional expression of the forgiveness construct with age.









Given that a predetermined theory existed, confirmatory factor analysis (CFA) was

initially utilized and the sample size guidelines for such analysis were targeted.

Thompson (2004), asserts that CFA necessitates a participant pool of 10-20 people per

measured variable, with no less than five per variable (Gorsuch, 1983). Eighteen

measured variables existed for the CADFI, therefore the necessary sample size needed to

be at least 90 (Bentler, 1990; Thompson, 2004).

Procedure

Recruitment of participants occurred from public schools, as well as after school

programs in two Florida communities. Testing occurred in small groups within the school

settings. For the purposes of paralleling administration method of the CADFI with the

HFS, as well as for higher internal consistency in the study, the items of the measure

were not read aloud to the participants, but were distributed for self-evaluation. Every

participant was given sufficient and similar amounts of time to complete the measures,

and most finished within 20-30 minutes. In order to avoid sequencing effects, the

measures included in the battery were counterbalanced, resulting in four different

sequences of forms. Lastly, a small sub sample of participants was briefly interviewed for

qualitative responses to questions indicative of forgiveness (see Appendix B).

Respondents were selected due to interview availability (i.e. were accessible in after-

school program or otherwise), were asked all questions and prompts, and their responses

were recorded.

Measures

The CADFI consists of 18 items, the CDI-S consists of 10 items, and the CASQ-R

includes 24 items. Therefore, there will be a total of 52 items administered for all three

scales included in the study to assess the reliability and validity of the new measure. The









Lie scale of the RCMAS (9 items) was included in order to address social desirability

concerns, bringing the total number of items in the battery to 61.

Child/Adolescent Dispositional Forgiveness Inventory (CADFI)

This measure was based upon the content of the adult HFS items, tailored to meet the

readability and developmental needs of the younger child/adolescent population. This

inventory contained three (3) subscales: Forgiveness of Self, Forgiveness of Others, and

Forgiveness of Situations. Six (6) items comprised each subscale, for a total of 18 items.

The same number of negatively worded items remained as in the HFS (nine). The HFS

has a general readability level of 6.3 (Flesch-Kincaid), meaning it is most appropriate for

those with at least a sixth grade reading level, and a reading Ease rating of 73.0. The

CADFI measure has a readability level of 4.8, with an ease rating of 81.9.

Children's Attributional Style Questionnaire-Revised (CASO-R)

The CASQ-R (Kaslow & Nolen-Hoeksema, 1991) evaluates a child's causal

associations for positive and negative life events. Twenty-four forced choice response

items consist of presentation of a hypothetical situation followed by two statements

regarding why the event happened, and children are asked to choose which of the

statements they feel represents how they would typically respond. An equal number of

items address both positive and negative events. Three dimensions are reflected within

the measure: internal-external, stable-unstable, and global-specific. Composite, negative,

and positive scores are calculated, and a lower composite score reflects a more depressive

attributional style. The CASQ-R taps attributional style in children, a construct found to

relate to forgiveness and negative affect as examined using the Response Styles

Questionnaire (Nolen-Hoeksema & Morrow, 1991) for adults.









The CASQ-R is a condensed version of the Child Attributional Style Questionnaire

(CASQ; Seligman, Peterson, Kaslow, Tanenbaum, Alloy & Abramson, 1984), which

originally contained 48 items. Understanding that the CASQ was often being used as part

of a larger battery of measures, the authors responded with the CASQ-R, a scale of more

practical utility considering its general use with children, who often have shorter attention

spans (Thompson, et al., 1998). The scale is appropriate for an age range of 9-12 years.

Through extensive psychometric analysis, Thompson et al. (1998) revealed adequate

internal consistency reliability; being reliable for both boys (alpha = .58) and girls (alpha

= .63), as well as older (12-14 years old; alpha = .58) and younger children (9-11 years

old; alpha = .64). No significant differences were found in criterion related validity

among boys and girls or among younger and older children (Thompson et al., 1998).

Previous research has revealed the relationship of attributional and response styles

with rumination and depressive symptoms. In terms of concurrent validity, comparisons

between scores on the CASQ-R and measures of depression have indicated that children

reporting greater numbers of depressive symptoms tend to attribute events to internal

causes, and consider them to be fairly stable over time, as well as generalizable across

situations (Gladstone & Kaslow, 1995; Thompson, et al., 1998). On the CASQ-R this is

reflected as higher scores on the internal rather than external (e.g. would respond to the

statement "You get a bad grade in school" with "I am not a good student", rather than

"Teachers give hard tests"), stable rather than unstable (e.g. "You do not get your chores

done at home" would elicit the response "Many days I am lazy" rather than "I was lazy

that day"), and global rather than specific dimensions (e.g. "You go to an amusement









park and you have a good time", would elicit "I usually enjoy myself in many activities"

instead of "I usually enjoy myself at amusement parks"), respectively.

Children's Depression Inventory-Short form (CDI-S)

The CDI-S (Kovacs, 1985, 1992) has 10 items, each consisting of three response

choices. The child/adolescent will be instructed to select one of the three sentences for

each item that best describes him or her over the past two weeks. Higher scores indicate

higher levels of depressive symptoms. The CDI-S is a short form measure that includes

10 of the 27 items of the original CDI, intended to assess depressive symptomatology

under circumstances when the time with a child may be limited or a quick screening

measure is desired. The CDI-S correlates at the .89 level with the long version of the CDI

(Kovacs, 1992), which has demonstrated adequate internal consistency (a = .71 to .89),

test-retest reliability (r = .74 to .83), and convergent and divergent validity (Kovacs,

1992). The CDI is appropriate for children 7-17 years of age.

The CDI was selected specifically because of its construct equivalence and item

similarity with the Beck Depression Inventory (BDI) for adults. The BDI was found to

negatively correlate with the HFS, a dispositional measure of willingness to forgive.

Thus, the CDI-S was expected to yield similar negative correlations with the CADFI.

Lie Scale from the Revised Children's Manifest Anxiety Scale (RCMAS)

The 9 items from the Lie subscale of the RCMAS (Reynolds & Richmond, 1978,

1985) were used to measure social desirability. The Lie subscale was developed from the

original CMAS subscale called the What I Think and Feel (WITF) subscale (Castaneda,

McCandless, & Palermo, 1956). These items are self-report format in which children are

asked to respond to a series of statements regarding themselves (e.g., "I am always kind",

or "I never get angry") with either "Yes" or "No". High scores on the Lie scale indicate









higher social desirability (7 or above is considered high), or a higher tendency to present

oneself in a favorable light. As was asserted by Reynolds and Richmond (1985), and has

been confirmed using extensive factor analysis, the Lie scale constitutes its own factor (or

two related factors distinct from the anxiety factors according to Reynolds & Paget,

1981), apart from the anxiety factors onto which the rest of the items load (Stark &

Laurent, 2001). The concurrent validity of the RCMAS lie scale has also been examined

and supported by relationship with the Marlowe-Crowne Social Desirability Scale

(Hagborg, 1991a) and has been subsequently used as a measure upon itself of social

desirability (Hagborg, 1991b). The RCMAS has impressive test-retest reliability

estimates as well, as confirmed by Pela and Reynolds (1982) (r = .94 for the Lie scale).

The current study included only the items from the Lie scale for use in assessing socially

desirable response tendencies. Like most other social desirability measures, Lie scores of

the RCMAS have been found to be higher for younger rather than older children (Pina,

Silverman, Saavedra, & Weems, 2001). No significant sex differences were generally

found for the Lie scale (Dadds, Perrin, & Yule, 1998; Pina, et al., 2001).














CHAPTER 4
RESULTS

Instrument

The Child/Adolescent Dispositional Forgiveness Inventory (CADFI) was modeled

after the HFS. The items of the CADFI consisted of modified wording of the HFS items

in order to reach appropriate reading level for children as young as fourth grade, thus the

same number of total items as well as items within each subscale was retained. The same

numbers of positively and negatively worded items were represented in the CADFI as are

within the HFS; nine of each valence

Descriptive Statistics

In the current sample, scores on the CDI-S ranged from 0 to 15, M= 3.5, SD = 3.49.

Scores on the CASQ-R positive composite ranged from 0 to 24, M= 15.86, SD = 4.00,

from 0 to 16 for the negative composite, M= 7.6, SD = 3.57, and -8 to 24 on the overall

composite, M= 8.22, SD = 7.06. Scores on the CADFI ranged from 30 to 68, M= 47.2,

SD = 8.2 for the total score, 8 to 30, M = 21.4, SD = 4.4 for the positively worded CADFI

scale, and 8 to 38, M= 22.1, SD = 6.37 for the negatively worded CADFI scale. Lie scale

scores ranged from 0 to 9, M= 3.09, SD = 2.55. Eleven participants scored 7 or greater

on the Lie scale, considered high scores by the scale's authors (Reynolds & Richmond,

1985).

Analyses

Factor analysis was used to examine the proposed model. Cronbach's coefficient

alphas were determined and reported for all scales and subscales of the measures. To









examine discriminant validity, correlation analyses were used to assess the relationship of

the CADFI scores to other measures known to relate to forgiveness: namely, the CDI-S,

the CASQ-R, and the Lie scale of the RCMAS.

Confirmatory Factor Analysis

Confirmatory factor analysis (CFA) was proposed with this study for three main

purposes: to test the overall adequacy of the specified measurement model; to determine

the adequacy with which items indicate intended factors, and to empirically inform any

decisions regarding scale revisions or further analyses. According to Thompson (2004),

CFA proves more useful when there is a presupposed theory and a model exists because

(a) the theory is directly tested in CFA analysis, and (b) the degree of the model fit can be

quantified in a variety of useful ways. The set of factor scores generated through CFA

remains useful in subsequent analyses of data and constructs regardless of the outcome of

the analyses. Although confirmatory factor analysis requires having a theoretical model at

hand, in this case a proposed three factor structure, it is generally asserted that the testing

of rival models in CFA remains a necessary step, as the possibility exists that several

models could fit a given data set (Thompson, 2004).

The proposed model in which three factors were specified (self, others, and

situations) using the corresponding items for each subscale proved to be less than

adequate. Confirmatory factor analyses revealed poor fit indices (e.g. Comparative Fit

Index = .51), and did not meaningfully converge on an interpretable model. Given that

significant problems were encountered in attempting to fit the initial solution and no clear

indication of the best direction for item or factor revision was available, exploratory

factor analysis was selected as most appropriate for these data.









Exploratory Factor Analysis

In order to examine the suitability of the sample data for factor analysis, the

correlation matrix was scanned for the presence of correlations at .30 or greater, and

Bartlett's test of sphericity was determined (which should be significant at p <.05), and

the Kaiser-Meyer-Olkin (KMO) value found (should be .6 or above if suitable). In the

current sample, correlations of .30 or higher were found, Bartlett's test of sphericity was

significant (p = .000), and the KMO value equaled .627, therefore the suitability of the

data for factor analyses was supported.

In each analysis, principal axis factoring was performed and oblique rotations using

the Promax procedure were conducted. Promax factoring allows for the model factors to

be correlated, and was chosen because it was expected that each of the underlying factors

would be related to the overall factor of forgiveness and so would be correlated.

Eigenvalues and pattern coefficients were examined to determine factor and item

retention, looking for 'simple structure' in which each of the variables loads strongly on

only one factor and each factor is represented by a number of strongly loading variables

(Thurstone, 1947). Items were retained if pattern coefficients were .30 or greater for no

more than one factor, and item-total correlations were positive. Subsequent factor

analyses were conducted after item deletions until the criteria for retention was upheld for

all factors and the eigenvalues of retained factors were at least 1.0 as recommended by

Kaiser (1960). Subscales were identified using all items remaining for each factor.

Three Factor Model

In the first analysis of the complete 18 item CADFI, seven factors meeting Kaiser's

criterion of eigenvalues greater than one emerged, and combined to account for 69% of

the variance. These eigenvalues ranged from 3.41 to 1.04. Catell's scree test (Catell,









1966) was also used, which recommends looking for a point at which the shape of the

curve changes to become more horizontal, then retaining only factors before that change

(or elbow) as they are considered to contribute most to the explanation of variance. The

elbow in the current plot appeared to occur at the three factor mark, and this three factor

model accounted for 41.8% of the variance. The results of an oblique rotation yield both

a Factor Structure matrix, indicating the correlation between each of the measures and the

factors extracted, and a Factor Pattern matrix, indicating the independent relationship

between each measure and the factors, similar to typical regression coefficients where

factors are used as predictors of each measure (Russell, 2002). The pattern coefficients

were used in making item retention decisions, given that they are the most useful in

interpreting the meaning of factors (Russell, 2002). In examining the pattern coefficients,

item retention criteria consisted of loading .30 or greater onto one factor and a positive

item-total correlation. In the first run, for example, item 11 did not load more than .30 on

any of the three factors, so was selected for deletion and item 12 loaded more than .30 on

factor one and two, and so was selected for deletion as it was considered nondistinct.

Several factor analyses were conducted using the criteria for item retention, resulting

in items 1, 8, 11, and 12 being selected for deletion. Three factors were indicated in this

model, having eigenvalues of 1.38, 2.32, and 3.09, accounting for 48.4% of the variance.

However, the items within these three factors were not clearly delineated along lines of

self, others, and situations. Factor one consisted of items 2, 4, 6, 15, and 17, which

reflected forgiveness of self with three of the items but also included two situation items.

The second factor contained the most items (3, 5, 10, 14, 16, and 18), and included items

initially considered to be situations items as well as two forgiveness of self items. The









third factor consisted of three items total, items 7, 9, and 13. Two of these items reflected

forgiveness of others, but the other reflected forgiveness of situations. Given that this

three factor model was inexplicable along theoretical lines, parallel analysis was

conducted to explore the meaningful factors to support.

Parallel Analysis

Many have called into question the traditional Kaiser's eigenvalue-greater-than-one

criterion for factor retention, asserting that it often retains too many factors (i.e.

Lautenschlager, 1989; Russell, 2002). Monte Carlo parallel analysis, initially put forth by

Horn (1965) as an adaptation to Kaiser's criterion has as its rationale that "meaningful"

components from actual sample data should have eigenvalues greater than those of

random data generated to be similar to the sample data in size and number of variables

(Lautenschlager, 1989). Due to the importance of examining disconfirmability in

confirmatory factor analysis, parallel analysis was utilized in the current sample as a

comparison for suggested factor retention.

Using the Monte Carlo parallel analysis program, simulated eigenvalues were

generated for random data with similar sample size and number of variables as the

current sample data set. Simulated eigenvalues generated by parallel analysis ranged from

1.83 to 0.39, while actual sample eigenvalues ranged from 3.41 to 0.25. Parallel analysis

suggests factor retention when actual eigenvalues of factors are larger than the simulated

eigenvalues, essentially indicating the meaningful factors to retain. In the current sample,

parallel analysis suggested retention of two significant factors (simulated values were

1.83 and 1.66, while actual eigenvalues were 3.41 and 2.70). Following the parallel

analysis a two factor model was then examined with the current data set using the same

criteria as before indicated.









Two Factor Model

The first factor analysis using the two-factor model resulted in eigenvalues of 3.41

and 2.69 for the two factors, and accounted for 33.9% of the variance. Item 1 again was

selected for deletion due to a negative correlation, and item 11 was not found to load

significantly onto any of the factors so was also selected. When items 1 and 11 were

deleted, the eigenvalues of the two factors were 3.25 and 2.44, accounting for 35.6% of

the variance. Item 12 did not have a value of .3 or higher for either factor, so was not

retained. Next factor analysis included all items except items 1, 11, and 12 and accounted

for 36.7% of the variance. Item 8 then did not load significantly onto either factor and so

was not retained. The final factor analysis included all items except 1, 8, 11, and 12, and

accounted for 38.6% of the variance (see Appendix A for pattern coefficients).

Upon examination of retained items and factors, it became apparent that the factors

were clearly representing the underlying item valence for positively and negatively

worded items. Items 2, 4, 6, 7, 9, 13, 15, and 17 represented the factor of negatively

worded items in this model, or items that have been suggested to tap unforgiveness

aspects (Thompson et al., 2005), while items 3, 5, 10, 14, 16, and 18 represented a factor

of positively worded items, tapping forgiveness itself. The simple structure of this two-

factor model, the clarity provided by the underlying factor distinction based on the

valence of the item wording, and the strong parallel analysis support of two factors all

contributed to the strength and acceptability of this model. Both the two and three factor

models suggested the same problematic items, lending further support for item deletions.

The two-factor model was more clearly reported and explained due to the clear

delineation of factors along the lines of item wording valence and was retained as the

strongest model of the current data.









Pilot label descriptors for the factors and corresponding scales found above in the

two-factor model suggested qualitative differences as well. The factor comprised of

negatively worded items suggested labels of unforgiveness, harsh expectations/standards,

high expectations for perfection, and unrealistic expectations for environment, whereas

the factor comprised of positively worded items suggested a label of forgiveness, life

lessons, and acceptance of lack of control over events. Label suggestions for the concept

that embodies the negatively worded items appears not to simply be the lack of the

embodiment of the positively worded factor items, as would be anticipated if the items

reflected merely opposing points on the same continuum.

Internal Consistency

Cronbach's coefficient alpha was calculated to measure the internal consistency of

the final CADFI scale score (see Appendix A) and each of the subscale scores. Alpha

levels for the HFS were .84-.87 for the total scale and .71-.83 for the subscales (Yamhure

Thompson & Snyder, 2003). Coefficient alphas for the current sample were .71 for the

total CADFI, .74 for the unforgiveness (negatively-worded items) subscale, and .67 for

the forgiveness (positively-worded items) subscale. Alphas for all other scales fell within

acceptable range; a = .82 for the CDI-S, a = .80 for the Lie scale, and K-R 20s for the

CASQ-R positive and negative composite scores were .72 and .65, respectively.

Convergent, Discriminant, and Construct Validity

Convergent and divergent validity of CADFI scores were examined using correlation

analyses. Correlation with child measures previously found to have a relationship with

the HFS offered the best options for validity examination with the CADFI. Previous

research has found the HFS to be significantly negatively correlated with the Hostile

Automatic Thoughts scale (Snyder, Crowson, Houston, Kurylo, & Poirier, 1997), the









rumination subscale of the Response Style Questionnaire (Nolen-Hoeksema & Morrow,

1991), and the Beck Depression Inventory (Beck & Steer, 1987).

The current study examined the relationship of forgiveness, as represented by CADFI

scores, with depressive symptoms using the CDI-S, negative (or depressive) attributional

response style using the CASQ-R, and social desirability using the Lie scale of the

RCMAS. It was expected that the CADFI would negatively correlate with the CDI-S, as

well as the negative composite score of the CASQ-R, measuring depressive attributional

style. Both were found to significantly and negatively correlate with CADFI scores, r = -

.58 and r = -.52, p < .001, respectively. It was also expected that the CADFI would

positively correlate with the positive composite of the CASQ-R, as well as the total

CASQ-R, which reflects healthy (rather than depressive) attributional style. Scores on

the CADFI were indeed found to be significantly positively related to the positive

composite of the CASQ-R, r = .41, as well as the total CASQ-R, r = .50, with <.001 for

both (see Table 1 for all correlations). Interestingly, the correlation of CADFI scores for

the forgiveness (positively-worded items) and unforgiveness (negatively-worded items)

subscales did not reach significance, r = .11, p < .30, indicating the subscales appear as

distinct constructs rather than as significantly related to each other.

No significant effects emerged for measurement sequence or sex differences.

However, a moderately significant negative correlation, r = -.26, p <.05, was found for

CADFI scores with age, in that younger participants reported higher total forgiveness

levels.

It was expected that findings would hold when controlling for social desirability,

indicated by scores on the Lie scale of the RCMAS. Lie scale scores were found to be











significantly positively correlated with scores on the CADFI, indicating higher social


desirability was related to higher forgiveness scores, r = .31, p <.001. Correlations


remained significant at thep <.001 level in the expected directions when controlling for


social desirability, r = -.54 for CDI-S, r = -.46 for negative CASQ-R, r = .34 for positive


CASQ-R, and r = .43 for overall CASQ-R scores, demonstrating moderate to medium


effect sizes (Cohen, 1988) for all relationships examined.


Table 1


Correlations Between All Scale and Subscale Scores

CDI-S CASQ-R Negative Positive CADFI CADFI CADFI Lie
CASQ-R CASQ-R Total Negatively Positively- Scale
subscale subscale -worded worded
items items


CDI-S

CASQ-R

Negative
CASQ-R
subscale
Positive
CASQ-R
subscale
CADFI
Total
CADFI
Negatively-
worded
items
CADFI
Positively-
worded
items
Lie Scale


1.00

-.59**

.60**


1.00

-.93**


-.74**


-.52**

.45**


.41**

-.41**


1.00

-.84**


-.42**


-.51** .94**


-.58**

.53**


.50**

-.46**


-.32** .26*



-.29** .43**


** Correlation is significant at the 0.01 level (2-tailed).

* Correlation is significant at the 0.05 level (2-tailed).

Qualitative Responses

Nine children selected at random provided qualitative answers to questions on


forgiveness (see Appendix C for responses) of self, of others, and of situations. Of these









nine respondents, three of them scored 7 or higher on the Lie scale (top 11% of sample),

indicating social desirability may significantly influence their responses. Two of the

respondents scored a 10 or higher on the CDI-S (top 6% of the sample), and those same

two participants scored 6 or below on the CASQ-R (lower 42% of sample).

Of the nine respondents first asked, "Is it hard for you to accept yourself when you

mess up?" three said no, four stated that sometimes it was, and one definitely said yes.

When next asked "Do you stay upset with others who have hurt you?" four indicated

"No" outright, one said "Not Really," two said "Sometimes" and one said "Yes." Finally,

when asked "Is it hard for you to accept negative things that are not anyone's fault?" four

stated unequivocally no, one indicated a qualified yes by saying "It's kinda hard," and

three said yes outright-that it was hard to accept.

Interestingly, though the words "forgive" or "forgiveness" were intentionally not used

in the qualitative questions and prompts, two of the respondents used such words in their

responses, both when asked what they do when hurt by others. One respondent who

answered "Forgive them" correspondingly scored in the top 8% on the CADFI scale, zero

on the CDI-S (lower 20%), and only one on the Lie scale (lower 35%). The other

respondent clearly stating "I forgave them" scored in the upper 36% on the CADFI, the

top 20% on the CASQ-R, and the lower 48% on the CDI-S, but scored in the top 2% on

the Lie scale, indicating fairly significant social desirability influence.

Respondents typically could and would provide an example for forgiveness of self,

others, and situations when asked, however one participant exhibited high reluctance to

give examples, instead offering "Yes" or "No" as sufficient responses. This respondent

was one of the two who also scored especially high on the CDI-S scale (and low on






40


CASQ-R) and also scored in the lower 13% for the CADFI. This respondent scored a

zero on the Lie scale. This respondent is also the only one to answer "yes" to all

questions asking whether it was hard to forgive, be it self, others, or situations, pointing

to the speculation that the hesitation in sharing may be due more to the difficulty

experienced in such situations involving forgiveness.














CHAPTER 5
DISCUSSION

Given the various approaches to defining forgiveness and differing aspects

emphasized among previous researchers, an efficient and compositional definition of

forgiveness was sought for developing a child/adolescent forgiveness measure to fill a

current gap in the literature. The current study investigated the relationship of

dispositional forgiveness with depressive symptoms, and depressive attributional style, in

children. Also, the utility and factor structure of the CADFI, a version of the HFS

modified for use with a younger age range, was examined.

Though the initial three factor structure of self, others, and situations was not

confirmed, exploratory factor analyses revealed a clear and highly explainable two factor

structure within the current sample, along the lines of valence of item wording. Parallel

analyses also suggested a strong two factor structure. Given that these two factors were

revealed to be highly uncorrelated, indicating they may in fact be two distinct entities,

this study introduces the idea of unforgiveness as distinct from forgiveness itself. In other

words, in children it may be more useful to think about the disposition for forgiveness as

a general propensity for 'forgiveness' (as revealed through positively worded items) and

unforgivenesss' (negatively worded items). The measurement of forgiveness, as a

construct, by emphasizing forgiveness as well as unforgiveness aspects is of course not

unprecedented. For example, Mauger's (1992) Forgiveness of Others and Forgiveness of

Self scales, as well as McCullough's TRIM scale (1998), actually tap unforgiveness as a

means of looking at forgiveness with adults. The current study provides emerging









evidence that such scales measuring unforgiveness may not merely be measuring the lack

of forgiveness, but a qualitatively distinct construct of unforgiveness. Though

theoretically related, empirically these two constructs in children and adolescents do not

appear to be significantly correlated.

When considered in the context of children's psychological understanding of persons

(Flavell, 2004), it is possible that younger children may not clearly understand or be

capable of precise differentiation among perspectives of forgiving self, forgiving others,

and forgiving situations. The ability to clearly understand one's relationship to self, to

others, and to situations beyond anyone's control could be a pre requisite to clear

distinctions (and decisions) of forgiveness along the same dimensions, which may not be

as highly attuned among children and young adolescents. Indeed, through the brief

qualitative interviews, evidence of a lack of concrete understanding between what is

considered a situation beyond anyone's control and what might be considered an accident

(i.e. participant 305's response). Given that measures used with children often employ a

format of forced-choice among two generally opposing options, the positive or negative

valence (reinforcing 'either/or' ways of thinking) may prove more salient with younger

children because of its structural simplicity. Further research able to replicate these

results with children of various populations would lend more strength to the two-factor

item valence model, as well as provide important insight into the relationship between

forgiveness and unforgiveness as possibly separate and distinct constructs.

Despite the difference in factor structure found in the current sample, the relationship

of forgiveness with constructs of well-being was found to hold strong. Forgiveness levels

significantly related to lower levels of depressive symptoms as well as lower depressive









attributional styles. Those children who typically attributed positive events to internal,

stable, and global aspects, and attributed negative events to external, unstable, and

specific dimensions correspondingly had higher levels of overall forgiveness. Those with

higher depressive symptoms and the opposite attributional style showed lower

forgiveness levels, indicating a highly positive relationship between healthy attributional

styles and forgiveness. Future research might consider the implications of this

relationship for counseling interventions, in that ways of increasing forgiveness could

lead to lower levels of depressive symptoms and depressive attributional style, or that a

reduction in depressive attributions and symptoms may lead to increased levels of

forgiveness.

Given that social desirability is more prevalent in younger age ranges, the possibility

of a significant correlation with this sample certainly existed and in fact was found.

However, it has been asserted that correlations between a social desirability measure and

a self-report measure do not necessarily jeopardize construct validity (Diener, Sandvik,

Pavot, & Gallagher, 1991; McCrae & Costa, 1983; Thompson et al., 2005).

Though the reliability and validity with this sample was supported, the

generalizability of the findings in the current study is of course limited by the sample size

and demographics. With a larger and more diverse sample, exploratory and confirmatory

factor analyses may provide a replicable and more solid structural two factor model. A

sample from a range of settings, as well as ages, would provide a deeper basis for

understanding forgiveness in the child and adolescent population. For example, a cross-

sectional study examining forgiveness with different age groups would provide

interesting developmental information regarding the progression of the factor structure of









forgiveness and unforgiveness with age. If the factor structure does change from a two

factor model in children to a three factor model in adults, such a study could reveal the

point at which this may occur. Examination of this measure in a more clinically

depressed population may be beneficial as well, considering the majority of this sample

did not report depressive symptoms in the clinical range.

This study serves as an insightful beginning for further exploration of dispositional

forgiveness in younger ages, both as it relates to other constructs of well-being, and

particularly the importance of understanding the factors that may make up forgiveness (or

unforgiveness) with such a population. Future studies may wish to use the current scale

with a larger sample for comparison, both to better understand the factor structure of

forgiveness as it presents in children/adolescents, as well as the relationship of

forgiveness to other constructs of well-being and positive psychology (i.e. altruism,

cognitive flexibility, openness to experiences, etc) and more pathological constructs such

as neuroticism, perfectionism, or negative affect. Furthermore, the brief qualitative

responses obtained in this study encourage more extensive, qualitative interviews with

younger populations in order to better inform future quantitative analyses and provide a

solid and sound basis for appropriate development of tools for measuring forgiveness.
















APPENDIX A
CADFI ITEMS AND FACTOR LOADINGS

CADFI item Factor 1 Factor 2

2. I dislike myself for negative things I have done. .703

3. Learning from bad things that I have done helps me get over them. .516

4. It is really hard for me to accept myself when I've messed up. .546

5. With time I am understanding of myself for mistakes I have made. .486

6. I get upset with myself for negative things I have felt, thought, said, or done. .629

7. I keep punishing a person who has done something that I think is wrong. .368

9. I stay upset with others who have hurt me. .370

10. Even though others have hurt me in the past,

with time I have been able to see them as good people. .582

13. When things go wrong for reasons that can't be controlled,

I keep having negative thoughts about it. .396

14. With time I can be understanding of bad things in my life. .627

15. I keep thinking negatively about things in my life that I cannot control. .581

16. With time I make peace with bad things in my life. .409

17. It is hard for me to accept negative things that are not anyone's fault. .554

18. With time I let go of negative thoughts about bad things

that are beyond anyone's control. .412














APPENDIX B
QUALITATIVE QUESTIONS

Is it hard for you to accept yourself when you mess up?
What is it like for you when you mess up? Can you think of a time when you did?

Do you stay upset with others who have hurt you?
What do you do? Can you think of an example?

Is it hard for you to accept negative things that are not anyone's fault?
What is it like for you when something bad happens and it's not anyone's fault?
Can you think of a situation? What happened?









APPENDIX C
QUALITATIVE RESPONSES

Participant 300:

Is it hard for you to accept yourself when you mess up? No.
What is it like for you when you mess up? Iget annoyed i iih myself.
Can you think of a time when you did? When I dropped my baton twice.

Do you stay upset with others who have hurt you? No.
What do you do? Play in i/h other friends.
Can you think of an example? A day my friends said mean iting\ We are still friends I
still forgave them.

Is it hard for you to accept negative things that are not anyone's fault? Ah man!
What is it like for you when something bad happens and it's not anyone's fault? I do
m uieltling else.
Can you think of a situation? What happened? When it rains when I want to go in the
pool.

Participant 12:

Is it hard for you to accept yourself when you mess up? Sometimes.
What is it like for you when you mess up? Can you think of a time when you did? Like
when remembering my words (for school).

Do you stay upset with others who have hurt you? No.
What do you do? Like, I just try to ignore them until we both have forgotten about it.
Can you think of an example? At computers, someone was annoying me and I ignore it.
Today, we were getting our backpacks, he let the lid down and it hit me on the cheek.
"You need to watch what you 're doing".

Is it hard for you to accept negative things that are not anyone's fault? No.
What is it like for you when something bad happens and it's not anyone's fault? Like, I
keep going on u/ i/h my life.
Can you think of a situation? What happened? Like a car accident, he wrecked the van, it
wasn 't anybody's fault.

Participant 11:

Is it hard for you to accept yourself when you mess up? No.
What is it like for you when you mess up? Can you think of a time when you did?

Do you stay upset with others who have hurt you? What do you do? No. When someone
hurt me physically?
Can you think of a situation? When someone pushed me off my bike and almost sprained
my ankle. That was 2 years ago. I forgave them.










Is it hard for you to accept negative things that are not anyone's fault? No.
What is it like for you when something bad happens and it's not anyone's fault? Can you
think of an example? What happened?

Participant 306:

Is it hard for you to accept yourself when you mess up? No.
What is it like for you when you mess up? It doesn 't bother me.
Can you think of a time when you did? Saturday, I missed my baton.

Do you stay upset with others who have hurt you? No.
What do you do? Iget upset, andl let it go.
Can you think of an example? My friends, when they didn't want to hear about my trip I
was so excited to go on.

Is it hard for you to accept negative things that are not anyone's fault? No.
What is it like for you when something bad happens and it's not anyone's fault? Iget
upset.
Can you think of a situation? Maggie moved away to Oklahoma.
What happened? I got sad and dealt i/h it.

Participant 305:

Is it hard for you to accept yourself when you mess up? Sometimes.
What is it like for you when you mess up? It is hard
Can you think of a time when you did? One time I messed up in horseback riding
lessons.

Do you stay upset with others who have hurt you? Not really.
What do you do? I walk away and try to be calm.
Can you think of an example? One time a person called me a name and I didn't like it.

Is it hard for you to accept negative things that are not anyone's fault? It's kinda hard.
What is it like for you when something bad happens and it's not anyone's fault? Ifeel
funny.
Can you think of a situation? One time a glass bowlfell off the table.
What happened? I was doing a dance and everyone was clapping and one moment the
bowl was there and one moment it fell.

Participant 304:

Is it hard for you to accept yourself when you mess up? Sometimes.
What is it like for you when you mess up? Hard.
Can you think of a time when you did? At school.









Do you stay upset with others who have hurt you? No.
What do you do? Igo tell the teacher.
Can you think of an example? When another kid hurt my feelings in music class. She
yelled at me.

Is it hard for you to accept negative things that are not anyone's fault? Yes.
What is it like for you when something bad happens and it's not anyone's fault?
Sometimes I feel scared.
Can you think of a situation? When my dog Belle got sick.
What happened? Sometimes I get really scared feeling is scary because you don't know
what's happening.

Participant 303:

Is it hard for you to accept yourself when you mess up? Yes.
What is it like for you when you mess up? I feel upset.
Can you think of a time when you did? Yes, I told a lie.

Do you stay upset with others who have hurt you? Sometimes.
What do you do? Tell or keep it to myself.
Can you think of an example? Today someone accused me for v,,ihilig

Is it hard for you to accept negative things that are not anyone's fault? Yes.
What is it like for you when something bad happens and it's not anyone's fault? Very
sad.
Can you think of a situation? What happened? My grandma got very sick.

Participant 302:

Is it hard for you to accept yourself when you mess up? Sometimes.
What is it like for you when you mess up? Ifeel ashamed.
Can you think of a time when you did? Yes.

Do you stay upset with others who have hurt you? Sometimes.
What do you do? Forgive them.
Can you think of an example? When my friend and I fought at school.

Is it hard for you to accept negative things that are not anyone's fault? No, it's sad
What is it like for you when something bad happens and it's not anyone's fault? Sad
Can you think of a situation? When my friend moved.
What happened? She movedfar away.

Participant 301:

Is it hard for you to accept yourself when you mess up? Yes.
What is it like for you when you mess up? Bad.






50


Can you think of a time when you did? No.

Do you stay upset with others who have hurt you? Yes.
What do you do? Can you think of an example? No.

Is it hard for you to accept negative things that are not anyone's fault? Yes.
What is it like for you when something bad happens and it's not anyone's fault? Bad.
Can you think of a situation? What happened? No.















REFERENCES


Bass, E., & Davis, L. (1994). The courage to heal. New York: Harper Perennial.

Berry, J. W., Worthington, E. J., Parrott, L., O'Connor, L. E., & Wade, N. G. (2001).
Dispositional forgivingness: Development and construct validity of the
Transgression Narrative Test of Forgivingness (TNTF). Personality and Social
Psychology Bulletin, 27 (10), 1277-1290.

Beck, A. T., & Steer, R. A. (1987). The BeckDepression Inventory. San Antonio, TX:
Psychological Corporation.

Bentler, P. M. (1990). Comparative Fit indices in structural models. Psychological
Bulletin, 107, 238-246.

Cohen, J. (1988). Statistical power analysis for the behavioral sciences. (2nd ed.)
Hillsdale, NJ: Erlbaum.

Carpendale, J. I. M., & Lewis, C. (2004). Constructing an understanding of mind: The
development of children's social interaction within social interaction. Behavioral
and Brain Sciences, 27, 79-151.

Castenada, A., McCandless, B. R., & Palermo, D. S. (1956). The children's form of the
manifest anxiety scale. ChildDevelopment, 27(3), 317-326.

Catell, R. B. (1966). The scree test for number of factors. Multivariate Behavioral
Research, 1, 245-276.

Cavell, M. (2003). Freedom and forgiveness. International Journal ofPsychoanalysis,
84(3), 515-531.

Crandall, V. C., Crandall, V. J., & Katkovsky, W. (1965). A children's social desirability
questionnaire. Journal of Consulting Psychology, 29 (1), 27-36.

Crowne, D. P., & Marlowe, D. (1960). A new scale of social desirability independent of
psychopathology. Journal of Consulting Psychology, 10, 3-31.

Dadds, M. R., Perrin, S., & Yule, W. (1998). Social desirability and self-reported anxiety
in children: An analysis of the RCMAS Lie scale. Journal ofAbnormal Child
Psychology, 26(4), 311-317.









Darby, B. W., & Schlenker, B.R. (1982). Children's reactions to apologies. Journal of
Social andPersonality Psychology, 43(4), 742-753.

Diener, E., Sandvik, E., Pavot, W., & Gallagher, D. (1991). Response artifacts in the
measurement of subjective well-being. Social Indicators Research, 24, 36-56.

Enright, R. D., & Zell, R. L. (1989). Problems encountered when we forgive another.
Journal of Psychology and Christianity, 8, 52-60.

Enright, R. D. (2000). Helping clients forgive: An empirical guide for resolving anger
and restoring hope. Washington, DC: American Psychological Association.

Enright, R. D., Freedman, S., & Rique, J. (1998). The psychology of interpersonal
forgiveness. In R. D. Enright & J. North (Eds.), Exploring forgiveness (pp.46-62).
Madison: University of Wisconsin Press.

Flavell, J. H. (2000). Development of children's knowledge about the mental world.
International Journal ofBehavioral Development, 24, 15-23.

Flett, G. L., Hewitt, P. L., Davidson, L., & Munro, Y. (2001). The Child-Adolescent
Perfectionism Scale: Development, validation, and association with adjustment
(submitted for publication).

Gladstone, T. R.G., & Kaslow, N. J. (1995). Depression and attributions in children and
adolescents: A meta-analytic review. Journal ofAbnormal Child Psychology,
23(5), 597-606.

Gorsuch, R. L. (1983). Factor analysis (2nd ed.). Hillsdale, NJ: Erlbaum.

Hagborg, W. J. (1991a). The Revised Children's Manifest Anxiety Scale and social
desirability. Educational and Psychological Measurement, 51(2), 423-427.

Hagborg, W. J. (1991b). Adolescent clients and perceived counselor characteristics: A
study of background characteristics, therapeutic progress, psychological distress,
and social desirability. Journal of Clinical Psychology, 47(1), 107-113.

Hargrave,T. D., & Sells, J. N. (1997). The development of a forgiveness scale. Journal of
Marital and Family Therapy, 23, 41-63.

Hebl, J. H., & Enright, R. D. (1993). Forgiveness as a psychotherapeutic goal with
elderly females. Py htheit,/t/1), 30, 658-667.

Hewitt, P. L., & Flett, G. L. (1991). Perfectionism in the self and social contexts:
Conceptualization, assessment, and association with psychopathology. Journal of
Personality and Social Psychology, 60, 456-470.









Hill, E.W. (2001). Understanding forgiveness as discovery: Implications for marital and
family therapy. Contemporary Family Therapy, 23(4), 369-384.

Kaslow, N. J., & Nolen-Hoeksema, S. (1991). Children's Attributional Style
Questionnaire- Revised. Unpublished manuscript, Emory University, Atlanta,
GA.

Katz, J., Street, A., & Arias, I. (1997). Individual differences in self-appraisals and
responses to dating violence scenarios. Violence and Victims, 12(3), 265-276.

Klein, E. B., Gould, L. J., & Corey, M. (1969). Social desirability in children: An
extension and replication. Journal of Consulting and Clinical Psychology, 33(1),
128.

Kovacs, M. (1985). The Children's Depression Inventory (CDI). Psychopharmacology
Bulletin, 21, 995-998.

Kovacs, M. (1992). The Children's Depression Inventory. Manual. Toronto, Canada:
Multi-Health Systems, Inc.

Larzelere, R. E., & Huston, T. L. (1980). The dyadic trust scale: Toward understanding
interpersonal trust in close relationships. Journal ofMarriage and the Family, 42,
595-604.

Lautenschlager, G. J. (1989). A comparison of alternatives to conducting monte carlo
analyses for determining parallel analysis criteria. Multivariate Behavioral
Research, 24(3), 365-395.

Lopez, S. J., & Snyder, C. R. (2003). Positive psychological assessment: A handbook of
models and measures. Washington, DC: American Psychological Association.

Macaskill, A., Maltby, J., & Day, L. (2002). Forgiveness of self and others and emotional
empathy. The Journal of Social Psychology, 142(5), 663-665.

Maltby, J., Macaskill, A., & Day, L. (2001). Failure to forgive self and others: A
replication and extension of the relationship between forgiveness, personality,
social desirability and general health. Personality and individual Differences, 30,
881-885.

Marlowe, D., & Crowne, D. P. (1961). Social desirability and response to perceived
situational demands. Journal of Consulting Psychology, 25, 109-115.
Martin, M. M., & Rubin, R. B. (1995). A new measure of cognitive flexibility.
Psychological Reports, 76, 623-626.









Mauger, P. A., Perry, J. E., Freeman, T., Grove, D. C., McBride, A. G., McKinney, K. E.
(1992). The measurement of forgiveness: Preliminary research. Journal of
Psychology and Christianity, 11, 170-180.

McCrae, R. R., & Costa, P. T., Jr., (1983). Social desirability scales: More substance than
style. Journal of Consulting and Clinical Psychology, 51, 882-888.

McCullough, M. E. (2000). Forgiveness as human strength: Theory, measurement, and
links to well-being. Journal of Social and Clinical Psychology, 19, 43-55.

McCullough, M. E., Pargament, K. I., & Thoresen, C. E. (Eds). (2000). The psychology
of forgiveness: History, conceptual issues, and overview. In M.E. McCullough,
K.I. Pargament, & C.E. Thoresen (Eds.), Forgiveness: Theory, research, and
practice (pp.1-14). New York: Guilford Press.

McCullough, M. E., Rachal, K. C., Sandage, S. J., Worthington, E. L., Jr., Brown, S. W.,
& Hight, T. L. (1998). Interpersonal forgiveness in close relationships II:
Theoretical elaboration and measurement. Journal ofPersonality and Social
Psychology, 75, 1586-1603.

Nolen-Hoeksema, S. & Morrow, J. (1991). A prospective study of depression and distress
following a natural disaster: The 1989 Loma Prieta earthquake. Journal of
Personality and Social Psychology, 61, 105-121.

Park, Y. O., & Enright, R. D. (1997). The development of forgiveness in the context of
adolescent friendship conflict in Korea. Journal ofAdolescence, 20(4), 393-402.

Pela, O. A., & Reynolds, C. R. (1982). Convergent and divergent validity of the Revised
Children's Manifest Anxiety Scale: Normative and reliability data for Nigerian
primary school children. Psychological Reports, 51, 1135-1138.

Pina, A., Silverman, W. K., Saavedra, L. M., & Weems, C. F. (2001). An analysis of the
RCMAS lie scale in a clinical sample of anxious children. Journal ofAnxiety
Disorders, 15(5), 443-457.

Reynolds, C. R., & Paget, K. D. (1981). Factor analysis of the revised Children's
Manifest Anxiety Scale for blacks, whites, males and females with a national
normative sample. Journal of Consulting and Clinical Psychology, 49(3), 352-
359.

Reynolds, C. R., & Richmond, B. 0. (1978). What I think and feel: A revised measure of
children's manifest anxiety. Journal ofAbnormal Child Psychology, 6, 271-280.

Reynolds, C. R., & Richmond, B. 0. (1985). Revised Children's Manifest Anxiety Scale
(RCMAS) manual. Los Angeles: Western Psychological Services.









Ripley, J. S., & Worthington, E. L. (2002). Hope-focused and forgiveness-based group
interventions to promote marital enrichment. Journal of Counseling &
Development, 80, 452-463.

Rosenak, C. M., & Harnden, G. M. (1992). Forgiveness in the Psychotherapeutic process:
Clinical applications. Journal of Psychology and Christianity, 11(2), 188-197.

Russell, D. W. (2002). In search of underlying dimensions: The use (and abuse) of factor
analysis in Personality and Social Psychology Bulletin. Personality and Social
Psychology Bulletin, 28(12), 1629-1646.

Scobie, G. E. W., & Scobie, E. D. (2000). A comparison of forgiveness and pro-social
development. Early Child Development and Care, 160, 33-45.

Seligman, M. E. P., Peterson, C., Kaslow, N. J., Tanenbaum, R. L., Alloy, L. B., &
Abramson, L. Y. (1984). Explanatory style and depressive symptoms among
school children. Journal ofAbnormal Psychology, 93, 235-238.

Shriberg, L.D. (1974). Descriptive statistics for two children's social desirability scales,
general and test anxiety, and locus of control in elementary school children.
Psychological Reports, 34, 863-870.

Snyder, C. R., Crowson, J. J., Jr., Houston, B. K., Kurylo, M., & Poirier, J. (1997).
Assessing hostile automatic thoughts: Development and validation of the HAT
Scale. Cognitive Therapy and Research, 21, 477-492.

Stark, K. D., & Laurent, J. (2001). Joint factor analysis of the Children's Depression
Inventory and the Revised children's Manifest Anxiety Scale. Journal of Clinical
Child Psychology, 30(4), 552-567.

Stuckless, N., & Goranson, R. (1992). The Vengeance Scale: Development of a measure
of attitudes towards revenge. Journal of Social Behavior & Personality, 7(1), 25-
42.

Subkoviak, M. J., Enright, R. D., Wu, C. R., Gassin, E. A., Freedman, S., & Olson, L. M.,
et al. (1995). Measuring interpersonal forgiveness in late adolescence and middle
adulthood. Journal ofAdolescence, 18, 641-655.

Tabachnik, B. G., & Fidell, L. S. (1996). Using multivariate statistics (3rd ed.) New
York: Harper Collins Publishers.

Tangney, J., Fee, R., Reinsmith, C., Boone, A. L., & Lee, N. (1999, August). Assessing
individual differences in the propensity to forgive. Paper presented at the annual
meeting of the American Psychological Association, Boston, MA.









Thompson, B. (2004). Exploratory and confirmatory factor analysis: Understanding
concepts and applications. Washington, DC: American Psychological
Association.

Thompson, M., Kaslow, N. J., Weiss, B., & Nolen-Hoeksema, S. (1998). Children's
Attributional Style Questionnaire-Revised: Psychometric examination.
Psychological Assessment, 10(2), 166-170.

Thompson, L. Y., & Snyder, C. R. (2003). Measuring forgiveness. In S. J. Lopez and C.
R. Snyder (Eds.), Positive psychological assessment: A handbook of models and
measures (pp. 301-312). Washington DC: American Psychological Association.

Thompson, L. Y., Snyder, C. R., Hoffman, L., Michael, S. T., Rasmussen, H. N., &
Billings, L., et al. (2005). Dispositional forgiveness of self, others, and situations.
Journal ofPersonality, 73(2), 313-360.

Wade, N. G., & Worthington, E. L. (2003). Overcoming interpersonal offenses: Is
forgiveness the only way to deal with unforgiveness? Journal of Counseling &
Development, 81, 343-353.

Walker, D. F., & Gorsuch, R. L. (2000). Forgiveness within the Big Five personality
model. Personality and Individual Differences, 32, 1127-1137.

Watson, D., Clark, L. A., & Tellegen, A. (1988). Development and validation of brief
measures of positive and negative affect: The PANAS scales. Journal of
Personality and Social Psychology, 54, 1063-1070.

Witvliet, C., Ludwig, T. E., & Bauer, D. J. (2002). Please Forgive Me: Transgressors'
Emotions and Physiology During Imagery of Seeking Forgiveness and Victim
Responses. Journal of Psychology and Christianity, 21(3), 219-233.

Worthington, E. L., Jr., & Wade, N. G. (1999). The psychology of unforgiveness and
forgiveness and implications for clinical practice. Journal of Social and Clinical
Psychology, 18(4), 385-418.

Worthington, E. L., Jr., Sandage, S. J., & Berry, J. W. (2000). Group interventions to
promote forgiveness. In M.E. McCullough, K.I. Pargament, & C.E. Thoresen
(Eds.), Forgiveness: Theory, research, andpractice (pp.228-253). New York:
Guilford Press.

Worthington, E. L., Jr., Mazzeo, S. E., & Klewer, W. L. (2002). Addictive and eating
disorders, unforgiveness, and forgiveness. Journal of Psychology and
Christianity, 21(3), 257-261.















BIOGRAPHICAL SKETCH

I was born in Tampa, Florida on July 5th, 1980, and grew up in Temple Terrace

and South Tampa until eighteen years of age, when I graduated from King High

School's International Baccalaureate Program.

I attended the University of Florida as an undergraduate, majoring in psychology

and supplementing my psychology major with concentrations in dance and biological

sciences. I completed my undergraduate senior honors thesis in an area of

developmental psychology, and graduated summa cum laude in 2003 with my

Bachelor of Science degree with a major in psychology.

I then joined the Department of Psychology at the University of Florida as a

graduate student in the Counseling Psychology Program. I completed my Master of

Science degree in the summer of 2006.




Full Text

PAGE 1

FORGIVENESS IN CHILDREN: THE CHILD/ADOLESCENT DISPOSITIONAL FORGIVENESS INVENTORY By BROOKE A. LEEVER A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE UNIVERSITY OF FLORIDA 2006

PAGE 2

Copyright 2006 by Brooke A. Leever

PAGE 3

iii ACKNOWLEDGMENTS I first wish to thank my family and frie nds who continually offer me support and have never failed to believe in me. They mean more to me than they know. I am extremely grateful to my mother Brenda Leever, for her dedication and helpful coordination efforts with this project. I would also like to thank my advisor, Dr. Ken Rice, for all of his time and wealth of statistical knowledge without which this project could not have been accomplished.

PAGE 4

iv TABLE OF CONTENTS page ACKNOWLEDGMENTS.................................................................................................iii TABLE.......................................................................................................................... .....vi ABSTRACT......................................................................................................................v ii CHAPTER 1 INTRODUCTION........................................................................................................1 2 LITERATURE REVIEW.............................................................................................8 Forgiveness and Forgiveness Measures........................................................................8 Children and Adolescents...........................................................................................13 The Heartland Forgiveness Scale...............................................................................17 Research Questions and Hypotheses..........................................................................20 Social Desirability......................................................................................................22 3 METHODS.................................................................................................................24 Participants.................................................................................................................24 Procedure....................................................................................................................25 Measures.....................................................................................................................25 Child/Adolescent Dispositional Forgiveness Inventory (CADFI)......................26 Childrens Attributional Style Questionnaire-Revised (CASQ-R).....................26 Childrens Depression Invent ory-Short form (CDI-S)........................................28 Lie Scale from the Revised Childrens Manifest Anxiety Scale (RCMAS).......28 4 RESULTS...................................................................................................................30 Instrument...................................................................................................................30 Descriptive Statistics..................................................................................................30 Analyses......................................................................................................................3 0 Confirmatory Factor Analysis.............................................................................31 Exploratory Factor Analysis................................................................................32 Three Factor Model.............................................................................................32 Parallel Analysis..................................................................................................34

PAGE 5

v Two Factor Model......................................................................................................35 Internal Consistency...................................................................................................36 Convergent, Discriminant, and Construct Validity....................................................36 Qualitative Responses.................................................................................................38 5 DISCUSSION.............................................................................................................41 APPENDIX A CADFI ITEMS AND FACTOR LOADINGS...........................................................45 B QUALITATIVE QUESTIONS..................................................................................46 C QUALITATIVE RESPONSES..................................................................................47 REFERENCES..................................................................................................................51 BIOGRAPHICAL SKETCH.............................................................................................57

PAGE 6

vi TABLE Table page 1 Correlations Between All S cale and Subscale Scores.........38

PAGE 7

vii Abstract of Thesis Presen ted to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Science FORGIVENESS IN CHILDREN: THE CHILD/ADOLESCENT DISPOSITIONAL FORGIVENESS INVENTORY By Brooke A. Leever August, 2006 Chair: Kenneth G. Rice Major Department: Psychology The current study investigated dispositional forgiveness in the child and adolescent population. Specifically, the factor structure, intern al consistency reliability, and validity of a modified version of the Heartland Forgiveness Scale (HFS), entitled the Child/Adolescent Dispositional Forgiveness Inventory (CADFI), were examined in a sample of children and adoles cents. Reliability coefficients were found to be adequate with this sample, and relationships with cons tructs of validity were found to be solid as well. In particular, the relationship of fo rgiveness with depressive symptoms (as measured by the Child Depression Inventory-Sh ort Form; CSI-S), attributional style (as measured by the Childrens Attributional St yle Questionnaire Revised; CASQ-R), and social desirability (as measured using the Lie Scale of the Revised Childrens Manifest Anxiety Scale) was explored. Results support relationships in the expected directions between forgiveness and the chosen measures. Internal consistency reliabilities were

PAGE 8

viii reported for all scales and factor structure was also examined and a proposed two factor model provided. Directions fo r future research are included.

PAGE 9

1 CHAPTER 1 INTRODUCTION Though no clear-cut consensus exists among researchers as to the definition of forgiveness, most definitions include a key component of renouncing anger and resentment. Forgiveness of another does not necessarily involve reconciliation, though that may occur as part of the process (s ee Thompson & Snyder, 2003 for descriptions). McCullough et al. (1998) assert that forgiv eness must include pro-social changes in motivation, and they provide criteria for distinguishing when these changes have occurred. Thus, they consider the important aspect to be forgiveness as a process involving interpersonal relationships. Thom pson and Snyder (2003) and Thompson et al. (2005) emphasize that a forgiver relinquishes the negative attachment to a transgression, transgressor, or sequence of events involved in a tran sgression. Negative thoughts, feelings, or behaviors are all contained within the concept of negative attachment, with the eventual goal of forgiveness being a more positive or at least neutral reframing, which results in a weakening of th e attachment. Hargrave and Se lls (1997) conceptualization provides support within itself for the restorative, healthy co mponent of forgiveness; they describe it as an effort to restore love a nd trustworthiness for relationships in order for victims and victimizers to put an end the cy cle of destructive enti tlement. Finally, Hill (2001) offers what seems to be one of the most encompassing and noteworthy descriptions, effectively articulating the esse ntial depth of the forg iveness construct and the beneficial nature of forgiveness by descri bing it as being more of a discovery rather than an act of will.

PAGE 10

2 Within the realm of dispositional measures, only one covers what appears to encompass the three main aspects of the fo rgiveness construct: the Heartland Forgiveness Scale (HFS), measuring Forgiven ess of Self, Forgiveness of Others, and Forgiveness of Situations. These authors use a definition of forgiveness in which forgiveness occurs primarily through a weakeni ng of the negative attachme nt to a transgressor, transgression, or some compilation of the two (Thompson & Snyder, 2003; Thompson et al., 2005). The negative attachment may then tr ansform into positive or neutral, providing a mechanism for the forgiveness process to unfold. A review of current forgiveness literatu re reveals that most researchers view forgiveness as adaptive (e.g. Mauger et al., 1992; McCullough, 2000; Thompson & Snyder, 2003; Thompson et al., 2005). Forg iveness has been linked to psychological health and well-being in general (e.g. Mauger et al, 1992; Enright, 2000; Scobie & Scobie, 2000; Worthington, Sandage, & Berry, 2000; Ripley & Worthington, 2002; Thompson & Snyder, 2003), while unforgive ness has been correlated with higher psychopathology (e.g. Mauger et al., 1992; Maltby, Macaskill, & Day, 2001; Worthington, Mazzeo, & Klewer, 2002). At the theoretical and philosophical level, forgiveness has been seen as evidence of ones objectivity, a demonstration of acceptance of ones embeddedness in the world and of freedom, with it forgiveness and selfforgiveness come (Cavell, 2003). Unforgiv eness has not been clearly defined but typically has been thought of as a lack of fo rgiveness. Unforgiveness has been attributed to a lack of letting go of ne gative attachment to events, ex emplified by behaviors such as repeated ruminations, or a pervasive negative affect. Attempts to reduce unforgiveness, often due to repeated ruminations, can lead to the presence of addictive behaviors and

PAGE 11

3 psychological defense mechanisms. Unforgivene ss, as part of a persons internalized negative affect, may become detrimental both by causing addictive behaviors as well as by being a result of such addictive behavi ors (Worthington et al ., 2002). Contrarily, a handful of researchers believe forgiveness ma y not be beneficial because it may leave the forgiver open to further victimization or victim-blaming (e.g. Bass & Davis, 1994; Katz, Street, & Arias, 1997). However, the forgivi ng of others is not the only aspect of forgiveness argued to be beneficial. Witvliet, Ludwig, and Bauer (2002) demonstrated the powerfully beneficial aspects of being forg iven by a victim as well. After imagery conditions in which the subjects imagined be ing forgiven, significant improvements were reported by participants in basic emotions such as anger and sadness, moral emotions such as guilt, shame, gratitude and hope, as well as perceived interpersonal forgiveness (Witvliet et al., 2002). Discrepancies as to th e beneficial nature of forgiveness exist mainly due to differences in the conceptualiz ations and definitions of forgiveness chosen. Current research thoroughly conveys the multidimensionality of the construct of forgiveness. The typical notion that forgiven ess only occurs from one person to another (interpersonally) has lead to a predominance of research examining transgression specific forgiveness, or the extent of ones forgiv eness specifically toward another person (e.g. Hargrave & Sells, 1997; Enright, Freedm an, & Rique, 1998; Tangney et al., 1999; McCullough, 2000; McCullough et al., 2000; Witv liet et al., 2000; Wade & Worthington, 2003). Transgression specific forgiveness relates to a particular event, and therefore may not necessarily be predictive of a persons willingness to forgive across situations and contexts. On the other hand, dispositional forgiveness measures attempt to remedy this limitation by emphasizing that the devel opment of a forgiveness measure holding

PAGE 12

4 predictive qualities remains key. Knowledge as to a persons willingness to forgive across various circumstances and events reve als more about the pr esence of that trait within an overall person, viewing forgiveness more expansively. Assessing dispositional forgiveness across co ntexts (e.g., Berry et al., 2001) provides a beneficial alternative to transgression specific assessments. Scores on measures of dispositional forgiveness tend to be specifica lly related to mental health and well-being, whereas scores on transgression specific m easures do not tend to be (Thompson & Snyder, 2003). Furthermore, measures of forg iveness that are dispositional in nature, rather than event specific, provide a bette r basis for comparison with psychological correlates of the forgiveness construc t (Thompson & Snyder, 2003). Dispositional forgiveness, as opposed to transgression sp ecific forgiveness, offers a broader window into forgiveness as a whole, since it gives due credence to the multifaceted nature of the forgiveness. The dispositional forgiveness construct falls within differing domains, or subscales. Forgiveness may occur inter personally, as exemplified through the transgression specific model, as well as intra personally, as revealed through ones willingness to forgive oneself. More recently, interest has begun to aris e in a third area of forgiveness, that of forgiveness of situations. Forg iveness of situations has been introduced as specifically a dispositional aspect of the forgiveness cons truct and has been a ssessed along with the previously established interp ersonal and intrapersonal domai ns. Measures with reliable scores have been created to include forgiveness of oneself (Mauger et al., 1992; Worthington et al., 2002; Macaskill, Maltby, & Day, 2002) and forgiveness of others,

PAGE 13

5 along with forgiveness of situations beyond the control of either self or others (Thompson & Snyder, 2003). The distinction of a construct along interper sonal as well as intrapersonal lines leads to a better understanding of th e depth of this construct of interest and a more specific view into its dimensional nature. The c onstruct of forgiveness clearly contains interpersonal as well as intrapersonal co mponents, existing as distinct yet related concepts. As would naturally seem to follo w, the psychopathologies associated with forgiveness and unforgiveness appear along sim ilar interpersonal and intrapersonal lines. A distinction in related pathologies exists de pending on the type of forgiveness (or lack of) being assessed. Maugers Forgiveness of Se lf (FS) and Forgiveness of Others (FO) scales have been correlated with the Minnesota Multiphasic Personality Inventory (MMPI), revealing a significant association of failure to forg ive with negative emotional states, such as depression, anxiety, anger/distrust, and ne gative self-esteem. Lack of forgiveness has also been specifically linke d to eating disorders (Worthington et al., 2002). Personality factors have been found to highly correlate with dispositional forgiveness of both self and others in the e xpected directions, esp ecially for the category of neuroticism versus emotiona l stability, in which lower levels of forgiveness related to higher neuroticism (Walker & Gorsuch, 2000). Failure to forgive self has been connected to more intra-pun itive pathologies, such as anxiety and depression, while failure to forgive others has been associated with extra-punitive pathologies, such as social alienation, social in troversion, depression, and psyc hoticism (Maltby, Macaskill, & Day, 2001). Mauger and colleagues (1992) also f ound deficits in forgiveness of self and

PAGE 14

6 others to be related to higher amounts of overt psychopathologies, for example schizophrenia. As is apparent, depression has been prev iously found to be a significant related pathology of failure to forgive self and othe rs, and therefore depression may represent a manifestation of both intraand extra-punitive forgiveness failures. That is, depression appears to be a main component of the ways in which failure to forgive may manifest and influence a persons psychologica l well-being, and therefore de pression is important to consider, theoretically and practically, in forgiveness research. As will be discussed, researchers who developed the Heartland Fo rgiveness Scale (HFS) report a similar relationship regarding the a ssociation of depression with lack of dispositional forgiveness, manifest as a significant ne gative correlation. Following this premise, the current study plans to examine correlation of the Child/Adolescent Dispositional Forgiveness Inventory with related constructs of interest such as a child measure of depression. Developments in adult forgiveness resear ch include broadening the scope of the forgiveness construct to be inclusive of forg iveness of self, forgiveness of others, and forgiveness of situations domains (i.e. Thomps on et al., 2005). Evaluation of forgiveness, or lack thereof, has been increasing in em phasis on dispositional factors for measurement rather than specific event examples. The us e of a dispositional measure lessens the difficulties and ambiguities of the relationship of forgiveness to similar psychological constructs. Correlations and conclusions ma de under a transgression specific framework by nature impose stricter circ umstances under which forgiveness frequency and behavior may be predicted. Consistent with these tre nds, a main purpose of the current study was

PAGE 15

7 to create a measure to meet the needs of the younger age range while retaining the integrity of the dispositional forgiveness cons truct and keeping up with current trends in this area of research. The creation of a dis positional measure of forgiveness appropriate for the child/adolescent population follows in th e footsteps of previous adult dispositional measures, allowing for a dynamic insight into the nature of forgiveness across a broad age range. Such a measure is necessary to examine the psychological correlates of forgiveness across a broader age span, as we ll as to open examination of the possible developmental influences on this construct. The study of forgiveness development as a dispositional trait prov ides a means to possible illuminati on of implications this construct may have on later psychological functioning. Future research in this direction is predicated on having a useful dispositi onal measure of child/adolescent forgiveness available.

PAGE 16

8 CHAPTER 2 LITERATURE REVIEW Forgiveness and Forgiveness Measures There has been much debate as to the concep tualization of forgiveness and its aspects. Most consider forgiveness of transgressions interpersonally to be the forgiveness representative. However, approaches to m easuring forgiveness range from presenting events of specific transgressions hypothetically committed towards a respondent and asking their propensity to forgive such an ev ent, to more broad statements intended to gauge ones disposition to forgiv e things beyond ones control. Mauger and colleagues (1992) created the Forg iveness of Self (FS) and Forgiveness of Others (FO) scales, two 15-item scales endorsed as true-false by respondents. These scales provide a solid basis for emphasis and distinction of two different subtypes, which fall under the greater umbrella of the construct of forgiveness. These scales tap the failure to forgive others and failure to forgive one self, which may more accurately characterize them as measures of unforgiveness than fo rgiveness. The Forgiveness of Others items revolve around taking revenge justifying retaliation and re venge, holding grudges, and seeing other people as apt to cause one hurt, while forgiven ess of self items involve feelings of guilt over past acts, seeing oneself as sinful, and having a variety of negative attitudes towards yourself (M auger et al., 1992). These m easures generally take the approach of measuring forgiveness from the nega tive standpoint, in that they appear to be tapping unforgiveness as indica tion of ones lack of forgiveness of self or others, implying unforgiveness and forgiveness would ex ist on a continuum in which one is said

PAGE 17

9 to have more or less forgiveness regardless of the negative or positive standpoint of the measure. In the transgression-specifi c vein, Enright and his colle agues developed the Enright Forgiveness Inventory (EFI; Subkoviak et al ., 1995) and the Willingness to Forgive scale (WTF; Hebl & Enright, 1993). Enright subscrib ed to the mindset that forgiveness is something that occurs between two people, excluding the idea of forgiveness of situations. Thus, their defin ition is a willingness to aba ndon ones right to resentment, negative judgment, and indifferent behavior toward one who unjustly hurt us, while fostering the undeserved qualities of compassi on, generosity, and even love toward him or her (Enright et al., 1998, pp. 46-47). This definition clearly exhibi ts the necessity for benevolence in order for forgiveness to occu r, and Enright and Ze ll (1989) also assert that, the fruition of forgiveness is enteri ng into loving community with others (p.99), suggesting that reconciliation may be a part of forgiveness as well. The focus of this scale is, again, one specific transgression. T hough respondents are asked to respond to hypothetical as well as one real situation, th e narrow focus of a transgression-specific scale makes the correlation of forgivene ss with related psychological constructs. The EFI consists of 60 items on a 6-point Likert scale assessing six forgiveness dimensions including positive affect, behavior, and cognitions, and the lack of negative affect, behavior and cognitions toward the o ffender. Respondents are supposed to think of the most recent transgressor or transgre ssion as such when responding to the specific transgressions. The summed total of scores on the items of the subsca les yields an overall forgiveness score. Relatedly, the WTF is a 16-item scale intended to measure the effectiveness of forgiveness as a therapeutic interpersonal intervention. The first 15 items

PAGE 18

10 include specific scenarios, while the la st item addresses the persons specific transgression of focu s. Respondents choose from ten re sponses to each scenario, one choice of which is forgiveness. The scale is scored based on how many times forgiveness is selected as a solution for when a person would respond with and pr efer to respond with forgiveness to a hypothetical scenario, as well as respond with and prefer to respond with forgiveness to the real scenario. Barry, Worthington, and colleagues also creat ed the Transgression Narrative Test of Forgiveness (2001), a five scenario scale purpo rted to measure the dispositional tendency to forgive interpersonal transgressions over ti me and across situations, which they term forgivingness. Participants re ad each scenario and respond using a 5-point Likert scale as to their likelihood of forg iving the transgression were they in that situation. The situations include two intentional transg ressions by acquainta nces, two negligent transgressions by friends, a nd one intentional transgressi on by a relative followed by an apology (Barry et al., 2001). Others have taken the approach of de veloping a scale to measure the precise components of their definitions of forgiveness. Similarly to Enright, McCullough (2000) and colleagues (McCullough et al., 1998) c onceptualize forgiveness as a construct emphasizing prosocial changes in motivations Under their definition one specifically experiences decreased motivation to avoid pe rsonal and psychologica l contact with the offender, to seek revenge or to see harm co me to the offender, and increased benevolence motivation toward the offender, implying th at forgiveness of others is the basic component. Under this framework, again, forgiv eness is emphasized as an interpersonal process rather than one that may occur intrapersonally.

PAGE 19

11 McCullough et al.s measure is the Transg ression-Related Interp ersonal Motivations Inventory (TRIM; 1998), consis ting of 12 items on a 5-point Likert scale, with two subscales measuring two components of their definition, motivations to avoid contact, and motivations to seek revenge or see ha rm come to the tran sgressor. It does not however measure motivation toward benevolence, and as such appears to indicate mostly the negative aspects, namely unforgiveness, as representative of ones lack of forgiveness, which is fairly similar to Maugers approach. Hargrave and Sells (1997) also follow the idea of creating a scale to measure aspects of their specific definition. They offer a definition of forgiveness that includes the component of love, effort in restoring love and trustworthiness to relationships so that victims and victimizers can put an end to dest ructive entitlement ( p.43). They assert two broader levels of forgiveness, exonera ting (involving insight and understanding) and forgiving (involving the overt act of forgiv ing as well as giving the opportunity for compensation, meaning the ability for intera ctions with the tr ansgressor in away perceived by that person as non-threaten ing and encouraging of emotional bonding). The Hargrave and Sells (1997) Interpers onal Relationship Resolution scale (IRRS) taps forgiveness of a specific person who has causes hurt to the respondent. It consists of two scales, forgiveness and pain, and has 22 total items and four subscales including insight, understanding, giving the opportunity for compensa tion, and the overt act of forgiving. Tangney and colleagues (1999) have develope d another measure of forgiveness that does not include the components of love or compassion as necessary for forgiveness. Specifically, their lengthy definition mainly emphasizes that giving up the negative

PAGE 20

12 emotions is what is sufficient to cons titute forgiveness. Their measure, the Multidimensional Forgiveness Inventory (MFI), is a scenario-based approach to measuring forgiveness of self and others, wh ich consists of 72 items on a 5-point Likert scale using 16 different transg ression scenarios. There ar e nine subscales, including propensity to forgive others, propensity to forg ive self, propensity to ask for forgiveness, time to forgive others, time to forgive self propensity to blame others, propensity to blame self, sensitivity to hurt feelings, and anger-proneness. This transgression-specific measure attempts to tap many facets of forgiv eness and includes both forgiveness of self and forgiveness of others components, cr eating a much wider scope under which to examine forgiveness through a transgression-spe cific lens. However, it is still that transgression specific. Examination of th e construct of forgiveness using specific descriptions of events still does not allow for a broader gene ralizability of forgiveness as attributed to the respondent as well as to be related to other ps ychological constructs. The current study argues that th e superior approach to meas urement of forgiveness is both through a non trangression -specific framework, and is incorporative of the self, others, and situational components of the forgiveness cons truct, in line with Snyder and Yamhure Thompsons definition of forgiveness: We define forgiveness as the frami ng of a perceived transgression such that ones attachment to the transg ressor, transgression, and sequelae of the transgression is transformed from negative to neutral or positive. The source of a transgression, and therefore the object of forgiveness, may be oneself, another person or persons, or a situation that one views as being beyond anyones control (e.g. an illness, fate, or a natural disaster). (Thompson & Snyder, 2003, p. 302) This definition appears as the broadest and most encompassing definition so far in that it includes the pos sibility of feeling tr ansgressed upon in an impersonal manner, and

PAGE 21

13 thus allows for the forgiveness of situations to exist as a distinct and important aspect of dispositional forgiveness. The authors mo st compelling argument for the understanding of forgiveness as dispositional rather than transgression-specific is the assertion that forgiveness itself occurs only intra personally. In other words, the target of forgiveness does not matter; it is the mo tivation and behavior of th e forgiver that exemplify forgiveness dispositionally. This measure created by Yamhure Thompson and colleagues, the Heartland Forgiveness scale (HFS; Thomps on et al., 2005) that measures forgiveness of self, others, and situations, is elaborat ed on in detail later in this section. Children and Adolescents The current breadth of literature in the area of dispositional forgiveness begs a more balanced examination of the forgiveness construct across a br oader age range. The overwhelming majority of research on the c onstruct of forgiveness focuses on adults. Identification of dispositional failures to forgive in adults inevitably points to the important question of developmental implicatio ns of forgiveness. In other words, the questions of whether a younger age range w ill exhibit similar forgiveness trends as adults, as well as how stable the constr uct may be with age are imperative to understanding how forgiveness develops. Future research in this direction requires a comprehensive scale of dispos itional forgiveness appropria te for a younger age range. As well, a focus on children and adolescents using a dispositional, rather than event-specific, lens will help shed light on the developmen tal nature of the forgiveness construct. Childrens understanding of the concept of forgiveness under different circumstances has been previously evaluated, and development of the concept of forgiveness is generally considered to occu r at a young age. Darby and Schlenker (1982) evaluated childrens understandin g of forgiveness and such adult social judgments by examining

PAGE 22

14 child judgments when observing an actor apologizing. Using childrens reactions to apologies, children as young as kindergarten age were shown to be able to comprehend the major factors of importanc e (e.g., intentions, motives, ap ologies) for adult-like social judgment involving forgiveness, and younger children were found to exhibit similar judgments to older children (Darby & Schle nker, 1982). These apologies varied with respect to complexity and circumstances (i.e ., intentional versus acc idental transgression, good or bad motives, and high or low responsibility and consequences). Though this study helps shed light on childre ns abilities to wrap their minds, and hearts, around the construct of forgiveness, the experimental a nd transgression-specific nature of the study precludes dispositional insights into the childrens constructs of forgiveness in a more predictive sense, a nd across various dimens ions. Forgiveness of others, though overwhelmingly th e most frequently studied aspect of forgiveness, does not represent the whole of the forgivene ss construct. Furthermore, transgression specificity limits the degree to which examination of the forgiveness construct as a whole can occur. Park and Enright (1997) proposed and suppor ted a developmental progression of the understanding of forgiveness w ith adolescents in Korea, us ing junior high students ages 13-14. Specifically, they outline the developmen t from the most basic level, that of revengeful forgiveness manifested as word s or gestures indicating forgiveness while internal hostility may still exist and be ope nly expressed, to external forgiveness in which external gestures indicate forgiveness but awareness exists of inner frustrations and continuing frustrations are suppressed, to the more advanced level, that of internal forgiveness where one genuinely seeks to understand the motives and perspective of

PAGE 23

15 the other person and reinterprets the event inco rporating both perspectiv es. This last level is again said by the authors to arise out of the principles of beneficence and love. Though beneficence and love are generally articulated in the views of forgiveness, this model also add credence to the argument of Yamhure Thompson and Snyde r, that forgiveness at the highest and most comprehensive level occur strictly intrapersonally, regardless of the target of forgiveness. This is to say that unforgiveness as a result of feeling transgressed upon is something that occurs within a person th ough may be a result of events from ones self, another person, or a situation. Thus, motivation for forgiveness can only occur from within a person in the same manner. This intrapersonal emphasis seems to be exactly what Park and Enright have suppor ted in their model, although their model focused upon forgiveness of others. They also report a significant positive correlation between an adolescents understanding of forgiveness and age, m eaning an adolescents understanding of forgiveness was significantly and positively associated with age; older adolescents were more forgiving than younger adolescents. Th is study also revealed that the higher the level of understanding of forgiveness, the more likely a person was to use a proactive restorative strategy in the act ual conflictual relationship in the real life situation. The degree to which the offering of social support was a factor followed a similar pattern, in that the early adolescents appeared to be in transition to this pattern while the later ones seemed either deeply rooted in or transiti oning out of social suppor t as a factor. Thus, significance of social support seems to be related to level of understanding attained. Though narrow in scope and sample, this study shows that developmental trends exist in the concept of forgiveness. Developmental knowledge as well as measurement

PAGE 24

16 understanding of forgiveness can in turn pr ovide an important background for working with forgiveness in the clinical settings with any age range. In the realm of clinical practice, models ha ve begun to be established for forgiveness. Worthington and Wade (1999) have set forth a model incorporating forgiveness and its related areas, in which the personal attributes of the participants and of the relationship at hand before a transgression fact or into the propensity and lik elihood of forgiveness. First the persons perception of the events take pl ace, and their initial emotional reaction to the events occurs. Then interpersonally active responses such as revenge/retaliation, prorelationship behavior, and percep tion of the offenders response would occur as well as the interpersonally passive re sponse of rumination. Within this model, an emotionally dissonant event may occur that changes ones initial reaction to the event, leading to emotional dissonance. The authors assert th at this aspect is crucial. Underlying forgiveness is an emotional dissonant event. Thus, a victims ability to forgive will be influenced partially by his or her abilit y to comprehend and successfully resolve incompatible emotions, which is the core of the hypothesized c onstruct of emotional intelligence (Worthington & Wade, 1999, p. 395). Other models of forgiveness have also b een proposed within the psychotherapeutic context. Rosenak and Harnden (1992) descri be a model in which, after an offensive event, victims experience hurt that leads to an ger. After the realizat ion and expression of hurt and anger comes an information gathering stage that occurs in therapy in which the client needs to glean more knowledge about th e transgressor. This in formation is thought to promote a better understanding from that pers ons perspective, leading to true empathy and in turn true forgiveness.

PAGE 25

17 The Heartland Forgiveness Scale In developing a dispositional forgiveness measure appropriate for children and adolescents, an existing well -rounded dispositional measure of forgiveness in the adult population was sought. Modeling after a reliable pre-existing adult measure of the same construct appears often as a useful ap proach to developing a complementary child/adolescent measure for a construct of interest, especially when there is a compelling reason to suspect construct e quivalence between children an d adults. For instance, in related areas of psychological constructs, the Child Adolescent Perfectionism Scale (CAPS; Flett et al., 2001) was modeled afte r the Multidimensional Perfectionism Scale (MPS; Hewitt & Flett, 1991) that had been developed for adults, and the Childrens Depression Inventory (CDI; Kovacs, 1985) wa s patterned after the framework of the Beck Depression Inventory (BDI; Beck & Stee r, 1987). Similarly, the measure created in this study, the Child/Adolescent Dispositional Forgiveness Inventory (CADFI) was designed as an age-appropriate modificati on from the Heartland Forgiveness Scale (HFS), targeted for a younger age range. The Heartland Forgivene ss Scale (HFS; Thompson & Snyder, 2003) was chosen to serve as the model in this study mainly because of its encomp assing nature for the construct of forgiveness. Under this framewo rk, someone can be relieved of the negative attachment through transforming the negative cognitions, emotions, and/or behaviors to either neutral or positive, as well as through weakening th e attachment by releasing the perception of a strong connection between onese lf and the transgresso r or transgression (Thompson & Snyder, 2003). This does not ne cessitate a forgetting of the event, nor preclude the person from taking any actions, le gal or otherwise. The emphasis is on the valence of the attachment, and forgiveness is said to occur as long as the motivation does

PAGE 26

18 not involve negative attachment (e.g., vengefu lness). Instead, forgiveness is understood as a process through which peopl e synthesize their prior assu mptions and the reality of the transgression into a new understandi ng of the transgress ion, transgressor, transgression sequelae, and potentially, of themselves other people, or the world (Thompson et al., 2005, p. 318). The authors are ther efore unique in thei r perspective that only the shift from a negative to a neutral attachment is necessary and sufficient to constitute forgiveness, rather than needi ng empathy or compassion to be demonstrated. The HFS was developed with adults to asse ss the dispositional tendency to forgive. The HFS includes 18 items, endorsed on a 7point Likert Scale ranging from almost always false of me to almost always true of me. Subscales in cluded in the HFS are forgiveness of self, forgiveness of others, and forgiveness of situations. Thus, for the current purposes of developing a multi-dimensional forgiveness measure for children and adolescents, this measure was considered the most comprehensive representational instrument for the dispositional nature of forgiveness. This measure of forgiveness was chosen because of its ability to tap interpers onal as well as intrapersonal forgiveness on a dispositional level across diffe rent contexts; the HFS is the only existing forgiveness scale which taps forgiveness of situations, one of the three main aspects of forgiveness (self, other, and situ ational forgiveness). Adequate internal consistency reliability of HFS scores has been supported, with Cronbach coefficient alphas on the total scale ranging from .84 to .87, and alphas on the subscales ranging from .71 to .83. Test-retest re liability of the HFS, on the basis of a 3week follow-up period, was .83 for the total score and ranged from .72 to .77 (Thompson & Snyder, 2003) for the subscales. The psyc hometric properties of the HFS have been

PAGE 27

19 evaluated and adequately dem onstrated with student and nonstudent samples. The HFS has been found to correlate in the expected di rections with other dispositional measures of forgiveness as well as measures of rela ted psychological constr ucts. The HFS shows a large positive correlation with the FS and FO scales, a smaller positive correlation with the WTF scale, and a moderate/medium co rrelation with the MF I (Thompson & Snyder, 2003). The forgiveness of others subscale of the HFS is significantly negatively correlated with one measure of forgiveness of a specific transgre ssion, the TRIM, which measures forgiveness of a transgression done specifical ly by another person (higher scores mean higher levels of unforgivene ss). The HFS was also found to significantly positively correlate with the EFI transgression-specific measure. The HFS was not significantly correlated with the IRRS, a m easure of forgiveness towards a specific person. Overall, the HFS has demonstrated much higher correlation wi th the dispositional measures of forgiveness than the transg ression-specific measures, although positive correlation was found with all measures of forgiveness examined (Thompson, et al., 2005). The HFS was found to negatively correlate wi th scores on measures of some related constructs as well. Yamhure Thompson and colleagues (2005) also found moderately large negative correlations of the HFS with the Hostile Automatic Thoughts scale, the rumination subscale of the Response Style Questionnaire (Thompson et al., 2005), and the Beck Depression Inventory (Thompson & Snyder, 2003) as well as the Vengeance Scale (Stuckless & Goranson, 1992), and the nega tive affect of the Positive and Negative Affect Schedule (Watson, Clark, & Telle gen, 1988). Since forgiveness involves transforming mental energy away from negativ e behaviors and thoughts, it makes logical

PAGE 28

20 sense that such constructs as negative affect, vengeance, and rumination leading to feeling depressed would correlate negativ ely with the HFS forgiveness measure. The HFS was also evaluated to determine its predictive power of forgiveness in romantic relationships. Participants were give n the HFS as well as measures of hostility, trust, relationship satisfacti on, and relationship duration once a nd again nine months later. Forgiveness as evaluated by the HFS was reveal ed as a stronger pred ictor of relationship satisfaction than hostility, and satisfacti on was also significantly predicted by how trusting the person was of thei r partner and their perception of how trusting their partner was of them (Thompson & Snyder, 2003, Thom pson et al., 2005). All of these studies support the premise that the HFS is a reliable valid, and useful measure for evaluating the dispositional tendency to grant forgiven ess of self, others, and across situations. Basing a childrens measure of forgiveness on such a structurally and theoretically supported measure as the HFS, which is defi ning current forgiveness research trends, provides a sound foundation for examining forg iveness dispositionally with children. Introducing a dispositional measure of forgiv eness that is appropriate for children and adolescents, when supported by adequate ps ychometric characteristics, contributes significantly to the forgiveness literatu re by permitting future multidimensional, dispositional forgiveness research with partic ipants of a developmentally wider age range than has historically been studied. Research Questions and Hypotheses Many parallels between the HFS and CADF I were expected. Specifically, it was expected that the factor structure of the CADFI would reflect the HFS structure because the CADFI will retain item content integrity of the HFS that is simply adjusted to be at a lower, age-appropriate reading level. The underlying constructs of dispositional

PAGE 29

21 forgiveness as evaluated by assessing forgiv eness of self, forgiveness of others, and forgiveness of situations we re not expected to differ si gnificantly for younger people. It was expected that acceptable levels of internal consistency reliability and validity would be reached, similar to the range of such alpha coefficients reported for the HFS. The HFS has been positively correlated w ith scores from the Cognitive Flexibility Scale (Martin & Rubin, 1995), the Dyadic Trust Scale (L arzelere & Huston, 1980), and the distraction subscale of the Response Style Questionnaire (Nolen-Hoeksema & Morrow, 1991).These measures are appropriate for adults but would not be appropriate for children. Fortunately similar instrument s have been developed for the childhood age range and were targeted for validity purposes in the current study. Specifically, the Childrens Attributional St yle Questionnaire-Revised (CASQ-R; Kaslow & NolenHoeksema, 1991) was used to assess the rela tion of the CADFI to styles of cognitive attribution or explanation of positive and ne gative events, and the Childrens Depression Inventory (CDI; Kovacs, 1985) me asuring depressive symptoms was also administered as an expected negative correlate of CADFI scores. Though the dimensions of the CASQ-R (globalspecific, internal-external, and stableunstable) differ from the adult measure, hi gher levels of depressi on have been found for those whose scores reflect a hi gher depressive attributional style, which on the CASQ-R appears as more frequent endorsement of in ternal, stable, and gl obal items for negative events, and external, specific, and unstab le items for positive events (Gladstone & Kaslow, 1995; Thompson, Kaslow, Weiss, & Nolen-Hoeksema, 1998). In fact, Gladstone and Kaslow (1995) report a negative correlation between th e CASQ-R and the Vanderbilt Depression Inventory (VDI) in non-clinical samples, indicating that the higher the

PAGE 30

22 depressive symptoms reported the more depr essive the attributional style. It was therefore expected in the current study th at higher scores on the CDI would correlate similarly with depressive attri butional style, as well as lowe r levels of forgiveness (low scores on the CADFI). Sex differences were not e xpected to differ significantly with regard to CADFI scores. Social Desirability As was introduced by Crowne and Marl owe (1960; Marlowe & Crowne, 1961), a tendency exists for respondents to report more socially desirable res ponses as a result of being influenced by other factors besides the content of measures ad ministered. Initially, the bulk of the social desira bility research fell within the adult age range. Crandall, Crandall, and Katkovsky (1965) expanded social desirability research by developing the Childrens Social Desirability (CSD) questi onnaires for a younger age range, targeting children and adolescents. Social desirability has also been incorporated into measure of other psychological constructs as well. Spec ifically, the Childrens Manifest Anxiety Scale (CMAS; Castenada, McCandless, & Palermo, 1956) and the Revised Childrens Manifest Anxiety Scale (RCMAS; Reynolds & Richmond, 1978) exemplify self-report measures developed with the intent of meas uring characteristics and manifestations of anxiety in children, and include a Lie subscal e specifically intended to address social desirability. Social desirability measures were establishe d to evaluate increased predictability that a respondent would endorse a hi gher degree of socially desi rable responses on a measure. Younger children, girls, non-white children, and less intelligent children have previously been found to endorse socially desirable respon ses more frequently (C randall et al., 1965;

PAGE 31

23 Shriberg, 1974). These findings have been re plicated and validated by Klein, Gould, and Corey (1969), who found that younger chil dren of both sexes had higher social desirability scores than olde r children, and that this tendency was most marked when respondents were younger than 11 years old. Mo reover, they suggested this pattern reflects an emergence of adult approval motiv ations in children beginning as young as 11 years old (Klein, Gould, & Corey, 1969). Considering these variables, especially the finding that younger more than older children report socially desirable answers, the inclusion of a social de sirability measure in the current study was necessary in order to ascertain whether answers regarding dispositional forgiveness, as well as depressi on and attributional style, may be influenced by factors other than substantive content. Th e current study employed the Lie subscale of the RCMAS in order to address social desirabi lity, while also keeping the number of total items in the battery within reasonabl e length for efficient administration.

PAGE 32

24 CHAPTER 3 METHODS Participants The total sample (N = 96) of participants consisted of boys and girls ages 8-17, with the majority of the sample falling between th e ages of 9 and 11 years old (77% of the sample). The distribution of ages consisted of one 8-year old, twen ty-one 9-year olds, thirty 10-year olds, twenty-t hree 11-year olds, five 12-year olds, one 13-year old, two 14year olds, five 15-year olds, three 16-year olds, and one 17-year ol d (four participants failed to report age). Sixty-fi ve percent of the sample was female; ethnic background of the sample was approximately 49% White, 20% Hispanic, 16% Asian, and 11% African American. Consent forms were distributed to children attending public elementary schools in a southern Florida county (approx. 86% of the sample), as well as a private school within a central Florida County (appr ox. 14% of the sample). Response rate was approximately 48%. The majority of the sa mple was representative of the population sampled; the majority of the participants (73% ) came from the fourth and fifth grades of an elementary school with the following population demographics for those grades: 44% male, 65% White, 11% African Americ an, 10% Hispanic, and 2% Asian. Rationale for inclusion of children and adolescents was twofold. Focus on children/adolescents a) helps balance the level of attention given to forgiveness research between adults and children, considering the comparative dearth of child/adolescent research in the area, and b) provides context for future exploration of conception and dispositional expression of the fo rgiveness construct with age.

PAGE 33

25 Given that a predetermined theory existed, confirmatory factor analysis (CFA) was initially utilized and the sample size guide lines for such analysis were targeted. Thompson (2004), asserts that CFA necessita tes a participant pool of 10-20 people per measured variable, with no less than fi ve per variable (Gorsuch, 1983). Eighteen measured variables existed for the CADFI, therefore the necessary sample size needed to be at least 90 (Bentler, 1990; Thompson, 2004). Procedure Recruitment of participants occurred fr om public schools, as well as after school programs in two Florida communities. Testing occurred in small groups within the school settings. For the purposes of paralleling ad ministration method of the CADFI with the HFS, as well as for higher internal consis tency in the study, the items of the measure were not read aloud to the participants, but were distributed for self-evaluation. Every participant was given sufficient and similar amounts of time to complete the measures, and most finished within 20-30 minutes. In order to avoid sequencing effects, the measures included in the battery were count erbalanced, resulting in four different sequences of forms. Lastly, a small sub sample of participants was briefly interviewed for qualitative responses to questions indi cative of forgiveness (see Appendix B). Respondents were selected due to interview av ailability (i.e. were accessible in afterschool program or otherwise), were asked all questions and prompts, and their responses were recorded. Measures The CADFI consists of 18 items, the CDI-S consists of 10 items, and the CASQ-R includes 24 items. Therefore, there will be a total of 52 items administered for all three scales included in the study to assess the re liability and validity of the new measure. The

PAGE 34

26 Lie scale of the RCMAS (9 items) was include d in order to address social desirability concerns, bringing the total number of items in the battery to 61. Child/Adolescent Dispositional Forgiveness Inventory (CADFI) This measure was based upon the content of th e adult HFS items, tailored to meet the readability and developmental needs of the younger child/adolescent population. This inventory contained three (3) subscales: Forgiveness of Self Forgiveness of Others, and Forgiveness of Situations. Six (6) items comp rised each subscale, for a total of 18 items. The same number of negatively worded items remained as in the HFS (nine). The HFS has a general readability level of 6.3 (Flesch-Kincaid), me aning it is most appropriate for those with at least a sixth grade reading level, and a reading Ease rating of 73.0. The CADFI measure has a readability leve l of 4.8, with an ease rating of 81.9. Childrens Attributional Style Questionnaire-Revised (CASQ-R) The CASQ-R (Kaslow & Nolen-Hoeksema, 1991) evaluates a childs causal associations for positive and negative life events. Twenty -four forced choice response items consist of presentati on of a hypothetical situati on followed by two statements regarding why the event happened, and child ren are asked to choose which of the statements they feel represents how they would typically respond. An equal number of items address both positive and negative even ts. Three dimensions are reflected within the measure: internal-external, stable-unsta ble, and global-specific. Composite, negative, and positive scores are calculated, and a lower composite score reflects a more depressive attributional style. The CASQ-R taps attri butional style in children, a construct found to relate to forgiveness and negative affect as examined using the Response Styles Questionnaire (Nolen-Hoeksema & Morrow, 1991) for adults.

PAGE 35

27 The CASQ-R is a condensed version of th e Child Attributiona l Style Questionnaire (CASQ; Seligman, Peterson, Kaslow, Ta nenbaum, Alloy & Abramson, 1984), which originally contained 48 items. Understanding that the CASQ wa s often being used as part of a larger battery of measures, the authors responded with the CASQ -R, a scale of more practical utility considering its general use with children, who often have shorter attention spans (Thompson, et al., 1998). The scale is ap propriate for an age range of 9-12 years. Through extensive psychometric analysis, Th ompson et al. (1998) revealed adequate internal consistency reliability ; being reliable for both boys (a lpha = .58) and girls (alpha = .63), as well as older (12-14 years old; al pha = .58) and younger children (9-11 years old; alpha = .64). No signif icant differences were found in criterion re lated validity among boys and girls or among younger and ol der children (Thompson et al., 1998). Previous research has revealed the relations hip of attributional and response styles with rumination and depressive symptoms. In terms of concurrent validity, comparisons between scores on the CASQ-R and measures of depression have i ndicated that children reporting greater numbers of depressive sympto ms tend to attribute events to internal causes, and consider them to be fairly stab le over time, as well as generalizable across situations (Gladstone & Kasl ow, 1995; Thompson, et al., 1998). On the CASQ-R this is reflected as higher scores on the internal ra ther than external (e.g. would respond to the statement You get a bad grade in school with I am not a good student, rather than Teachers give hard tests), stable rather than unstable (e.g. You do not get your chores done at home would elicit the response Many da ys I am lazy rather than I was lazy that day), and global rather than specif ic dimensions (e.g. You go to an amusement

PAGE 36

28 park and you have a good time, would elicit I usually enjoy myself in many activities instead of I usually enjoy myself at amusement parks), respectively. Childrens Depression Invent ory-Short form (CDI-S) The CDI-S (Kovacs, 1985, 1992) has 10 items, each consisting of three response choices. The child/adolescent will be instructed to select one of the three sentences for each item that best describes him or her over the past two weeks. Higher scores indicate higher levels of depressive symptoms. The CD I-S is a short form measure that includes 10 of the 27 items of the original CDI, in tended to assess depressive symptomatology under circumstances when the time with a child may be limited or a quick screening measure is desired. The CDI-S co rrelates at the .89 level with the long version of the CDI (Kovacs, 1992), which has demonstrated adequate internal consistency ( = .71 to .89), test-retest reliability ( r = .74 to .83), and convergent a nd divergent validity (Kovacs, 1992). The CDI is appropriate for children 7-17 years of age. The CDI was selected specifically because of its construct equivalence and item similarity with the Beck Depression Invent ory (BDI) for adults. The BDI was found to negatively correlate with the HFS, a dispositional measure of willingness to forgive. Thus, the CDI-S was expected to yield sim ilar negative correlations with the CADFI. Lie Scale from the Revised Children s Manifest Anxiet y Scale (RCMAS) The 9 items from the Lie subscale of the RCMAS (Reynol ds & Richmond, 1978, 1985) were used to measure social desirabilit y. The Lie subscale was developed from the original CMAS subscale called the What I Th ink and Feel (WITF) subscale (Castaneda, McCandless, & Palermo, 1956). These items are self-report format in which children are asked to respond to a series of statements re garding themselves (e.g., I am always kind, or I never get angry) with either Yes or No. High scores on th e Lie scale indicate

PAGE 37

29 higher social desirability (7 or above is cons idered high), or a highe r tendency to present oneself in a favorable light. As was asse rted by Reynolds and Richmond (1985), and has been confirmed using extensive factor analysis, the Lie scale constitutes its own factor (or two related factors distinct from the anxi ety factors according to Reynolds & Paget, 1981), apart from the anxiety factors onto whic h the rest of the items load (Stark & Laurent, 2001). The concurrent validity of the RCMAS lie scale has also been examined and supported by relationship with the Marl owe-Crowne Social Desirability Scale (Hagborg, 1991a) and has been subsequently used as a measure upon itself of social desirability (Hagborg, 1991b). The RCMAS ha s impressive test-r etest reliability estimates as well, as confirme d by Pela and Reynolds (1982) ( r = .94 for the Lie scale). The current study included only the items from the Lie scale for use in assessing socially desirable response tendencies. Li ke most other social desirab ility measures, Lie scores of the RCMAS have been found to be higher for younger rather than ol der children (Pina, Silverman, Saavedra, & Weems, 2001). No si gnificant sex differences were generally found for the Lie scale (Dadds, Perrin, & Yule, 1998; Pina, et al., 2001).

PAGE 38

30 CHAPTER 4 RESULTS Instrument The Child/Adolescent Dispositional Forg iveness Inventory (CADFI) was modeled after the HFS. The items of the CADFI consis ted of modified wording of the HFS items in order to reach appropriate reading level for children as young as fourth grade, thus the same number of total items as well as items within each subscale was retained. The same numbers of positively and negatively worded items were represented in the CADFI as are within the HFS; nine of each valence Descriptive Statistics In the current sample, scores on the CDI-S ranged from 0 to 15, M = 3.5, SD = 3.49. Scores on the CASQ-R positive composite ranged from 0 to 24, M = 15.86, SD = 4.00, from 0 to 16 for the negative composite, M = 7.6, SD = 3.57, and -8 to 24 on the overall composite, M = 8.22, SD = 7.06. Scores on the CADFI ranged from 30 to 68, M = 47.2, SD = 8.2 for the total score, 8 to 30, M = 21.4, SD = 4.4 for the positively worded CADFI scale, and 8 to 38, M = 22.1, SD = 6.37 for the negatively worded CADFI scale. Lie scale scores ranged from 0 to 9, M = 3.09, SD = 2.55. Eleven participants scored 7 or greater on the Lie scale, considered high scores by the scales authors (Reynolds & Richmond, 1985). Analyses Factor analysis was used to examine th e proposed model. Cronbachs coefficient alphas were determined and reported for all scales and subscales of the measures. To

PAGE 39

31 examine discriminant validity, correlation analys es were used to assess the relationship of the CADFI scores to other measures known to relate to forgiveness: namely, the CDI-S, the CASQ-R, and the Lie scale of the RCMAS. Confirmatory Factor Analysis Confirmatory factor analys is (CFA) was proposed with this study for three main purposes: to test the overall ad equacy of the specified meas urement model; to determine the adequacy with which items indicate intende d factors, and to empirically inform any decisions regarding scale revisions or furt her analyses. According to Thompson (2004), CFA proves more useful when there is a pr esupposed theory and a model exists because (a) the theory is directly tested in CFA anal ysis, and (b) the degree of the model fit can be quantified in a variety of us eful ways. The set of factor scores generated through CFA remains useful in subsequent analyses of data and constructs regardless of the outcome of the analyses. Although confirmatory factor anal ysis requires having a theoretical model at hand, in this case a proposed thre e factor structure, it is genera lly asserted that the testing of rival models in CFA remains a necessary st ep, as the possibility exists that several models could fit a given data set (Thompson, 2004). The proposed model in which three factor s were specified (self, others, and situations) using the corresponding items fo r each subscale proved to be less than adequate. Confirmatory factor analyses re vealed poor fit indices (e.g. Comparative Fit Index = .51), and did not meaningfully converg e on an interpretable model. Given that significant problems were encountered in attemp ting to fit the initia l solution and no clear indication of the best direction for item or factor revision was av ailable, exploratory factor analysis was selected as most appropriate for these data.

PAGE 40

32 Exploratory Factor Analysis In order to examine the suitability of th e sample data for factor analysis, the correlation matrix was scanned for the pres ence of correla tions at .30 or greater, and Bartletts test of sphericity was determined (which should be signi ficant at p <.05), and the Kaiser-Meyer-Olkin (KMO) value found (s hould be .6 or above if suitable). In the current sample, correlations of .30 or higher we re found, Bartletts test of sphericity was significant ( p = .000), and the KMO value equaled .627, therefore the suitability of the data for factor analyses was supported. In each analysis, principal axis factoring was performed and oblique rotations using the Promax procedure were conducted. Promax factoring allows for the model factors to be correlated, and was chosen because it was ex pected that each of the underlying factors would be related to the overall factor of forgiveness and so would be correlated. Eigenvalues and pattern coefficients were examined to determine factor and item retention, looking for simple structure in wh ich each of the variables loads strongly on only one factor and each fact or is represented by a number of strongly loading variables (Thurstone, 1947). Items were reta ined if pattern coefficients were .30 or greater for no more than one factor, and item-total correla tions were positive. Subsequent factor analyses were conducted after it em deletions until the criteria for retention was upheld for all factors and the eigenvalues of retained factors were at leas t 1.0 as recommended by Kaiser (1960). Subscales were identified using all items remaining for each factor. Three Factor Model In the first analysis of the complete 18 item CADFI, seven factors meeting Kaisers criterion of eigenvalues great er than one emerged, and combined to account for 69% of the variance. These eigenvalues ranged from 3.41 to 1.04. Cate lls scree test (Catell,

PAGE 41

33 1966) was also used, which recommends looki ng for a point at which the shape of the curve changes to become more horizontal, then retaining only factors before that change (or elbow) as they are consider ed to contribute most to th e explanation of variance. The elbow in the current plot appeared to occur at the three factor mark, and this three factor model accounted for 41.8% of the variance. Th e results of an obli que rotation yield both a Factor Structure matrix, indicating the corr elation between each of the measures and the factors extracted, and a Factor Pattern matrix, indicating the independent relationship between each measure and the factors, simila r to typical regression coefficients where factors are used as predictors of each measure (Russell, 2 002). The pattern coefficients were used in making item retention decisions, given that they are the most useful in interpreting the meaning of factors (Russell, 2002). In examining the pattern coefficients, item retention criteria consisted of loading .30 or greater onto one factor and a positive item-total correlation. In the fi rst run, for example, item 11 did not load more than .30 on any of the three factors, so was selected fo r deletion and item 12 load ed more than .30 on factor one and two, and so was selected fo r deletion as it was c onsidered nondistinct. Several factor analyses were conducted using the criteria for item retention, resulting in items 1, 8, 11, and 12 being selected for de letion. Three factors were indicated in this model, having eigenvalues of 1.38, 2.32, and 3.09, accounting for 48.4% of the variance. However, the items within these three factor s were not clearly delineated along lines of self, others, and situations. Factor one consisted of items 2, 4, 6, 15, and 17, which reflected forgiveness of self with three of the items but also included two situation items. The second factor contained the most items (3, 5, 10, 14, 16, and 18), and included items initially considered to be situations items as well as two forgiveness of self items. The

PAGE 42

34 third factor consisted of three items total, items 7, 9, and 13. Two of these items reflected forgiveness of others, but the other reflected forgiveness of situations. Given that this three factor model was inexplicable along th eoretical lines, para llel analysis was conducted to explore the mean ingful factors to support. Parallel Analysis Many have called into question the traditi onal Kaisers eigenvalue-greater-than-one criterion for factor retenti on, asserting that it often retains too many factors (i.e. Lautenschlager, 1989; Russell, 2002). Monte Carl o parallel analysis, initially put forth by Horn (1965) as an adaptation to Kaisers crit erion has as its rationa le that meaningful components from actual sample data should have eigenvalues great er than those of random data generated to be similar to the sa mple data in size and number of variables (Lautenschlager, 1989). Due to the importa nce of examining disconfirmability in confirmatory factor analysis, parallel analys is was utilized in the current sample as a comparison for suggested factor retention. Using the Monte Carlo parallel analysis program, simulated eigenvalues were generated for random data with similar sa mple size and number of variables as the current sample data set. Simulated eigenvalues generated by parallel analysis ranged from 1.83 to 0.39, while actual sample eigenvalues ra nged from 3.41 to 0.25. Parallel analysis suggests factor retention when actual eigenvalues of factors are larger than the simulated eigenvalues, essentially indicati ng the meaningful factors to re tain. In the current sample, parallel analysis suggested retention of tw o significant factors (sim ulated values were 1.83 and 1.66, while actual eigenvalues we re 3.41 and 2.70). Following the parallel analysis a two factor model was then examined with the current data set using the same criteria as before indicated.

PAGE 43

35 Two Factor Model The first factor analysis us ing the two-factor model resu lted in eigenvalues of 3.41 and 2.69 for the two factors, and accounted fo r 33.9% of the variance. Item 1 again was selected for deletion due to a negative corr elation, and item 11 was not found to load significantly onto any of the factors so was also selected. When items 1 and 11 were deleted, the eigenvalues of the two factors were 3.25 and 2.44, accounting for 35.6% of the variance. Item 12 did not have a value of .3 or higher for either factor, so was not retained. Next factor analysis included all items except items 1, 11, and 12 and accounted for 36.7% of the variance. Item 8 then did not load significantly onto ei ther factor and so was not retained. The final f actor analysis included all items except 1, 8, 11, and 12, and accounted for 38.6% of the variance (see A ppendix A for pattern coefficients). Upon examination of retained items and fact ors, it became appare nt that the factors were clearly representing the underlying item valence for positively and negatively worded items. Items 2, 4, 6, 7, 9, 13, 15, and 17 represented the factor of negatively worded items in this model, or items that have been suggested to tap unforgiveness aspects (Thompson et al., 2005), while items 3, 5, 10, 14, 16, and 18 represented a factor of positively worded items, tapping forgiveness itself. The simple structure of this twofactor model, the clarity provided by the underlying factor distinction based on the valence of the item wording, and the strong pa rallel analysis support of two factors all contributed to the strength and acceptability of this model. Both the two and three factor models suggested the same problematic ite ms, lending further support for item deletions. The two-factor model was more clearly re ported and explaine d due to the clear delineation of factors along the lines of item wording valence and was retained as the strongest model of the current data.

PAGE 44

36 Pilot label descriptors for the factors a nd corresponding scales found above in the two-factor model suggested qua litative differences as well. The factor comprised of negatively worded items suggested labels of unforgiveness, harsh e xpectations/standards, high expectations for perfection, and unrealisti c expectations for environment, whereas the factor comprised of positively worded item s suggested a label of forgiveness, life lessons, and acceptance of lack of control over events. Label suggestions for the concept that embodies the negatively worded items a ppears not to simply be the lack of the embodiment of the positively worded factor it ems, as would be anticipated if the items reflected merely opposing points on the same continuum. Internal Consistency Cronbachs coefficient alpha was calculated to measure the intern al consistency of the final CADFI scale score (see Appendix A) and each of the subscale scores. Alpha levels for the HFS were .84-.87 for the total scale and .71-.83 for the subscales (Yamhure Thompson & Snyder, 2003). Coefficient alphas for the current sample were .71 for the total CADFI, .74 for the unforgiveness (nega tively-worded items) subscale, and .67 for the forgiveness (positively-worded items) subscale. Alphas for all other scales fell within acceptable range; = .82 for the CDI-S, = .80 for the Lie scale, and K-R 20s for the CASQ-R positive and negative composite scores were .72 and .65, respectively. Convergent, Discriminant and Construct Validity Convergent and divergent validity of CADFI scores were examined using correlation analyses. Correlation with child measures pr eviously found to have a relationship with the HFS offered the best opti ons for validity examination with the CADFI. Previous research has found the HFS to be significan tly negatively correlated with the Hostile Automatic Thoughts scale (Snyder, Crowson, Houston, Kurylo, & Poirier, 1997), the

PAGE 45

37 rumination subscale of the Response Style Questionnaire (Nolen-Hoeksema & Morrow, 1991), and the Beck Depression In ventory (Beck & Steer, 1987). The current study examined the relationship of forgiveness, as represented by CADFI scores, with depressive sympto ms using the CDI-S, negative (or depressive) attributional response style using the CASQ-R, and social desirability using the Lie scale of the RCMAS. It was expected that the CADFI woul d negatively correlate with the CDI-S, as well as the negative composite score of the CASQ-R, measuring depr essive attributional style. Both were found to significantly and negatively correlate with CADFI scores, r = .58 and r = -.52, p < .001, respectively. It was also expected that the CADFI would positively correlate with the positive composite of the CASQ-R, as well as the total CASQ-R, which reflects healthy (rather than de pressive) attributional style. Scores on the CADFI were indeed found to be significantly positively related to the positive composite of the CASQ-R, r = .41, as well as the total CASQ-R, r = .50, with p <.001 for both (see Table 1 for all correlations). Interest ingly, the correlation of CADFI scores for the forgiveness (positively-worded items) and unforgiveness (negatively-worded items) subscales did not reach significance, r = .11, p < .30, indicating the subscales appear as distinct constructs rather than as significantly related to each other. No significant effects emerged for m easurement sequence or sex differences. However, a moderately si gnificant negative correlation, r = -.26, p <.05, was found for CADFI scores with age, in that younger par ticipants reported hi gher total forgiveness levels. It was expected that findings would hold wh en controlling for social desirability, indicated by scores on the Li e scale of the RCMAS. Lie scal e scores were found to be

PAGE 46

38 significantly positively correlated with scores on the CADFI, indicating higher social desirability was related to higher forgiveness scores, r = .31, p <.001. Correlations remained significant at the p <.001 level in the expected di rections when controlling for social desirability, r = -.54 for CDI-S, r = -.46 for negative CASQ-R, r = .34 for positive CASQ-R, and r = .43 for overall CASQ-R scores, demonstrating moderate to medium effect sizes (Cohen, 1988) for all relationships examined. Table 1 Correlations Between All S cale and Subscale Scores CDI-S CASQ-R Negative CASQ-R subscale Positive CASQ-R subscale CADFI Total CADFI Negatively -worded items CADFI Positivelyworded items Lie Scale CDI-S 1.00 CASQ-R -.59** 1.00 Negative CASQ-R subscale .60** -.93** 1.00 Positive CASQ-R subscale -.51** .94** -.74** 1.00 CADFI Total -.58** .50** -.52** .41** 1.00 CADFI Negativelyworded items .53** -.46** .45** -. 41** -.84** 1.00 CADFI Positivelyworded items -.32** .26* -.31** 19 .63** .11 1.00 Lie Scale -.29** .43** -.42** .38** .31** -.27** .20 1.00 ** Correlation is significant at the 0.01 level (2-tailed). Correlation is significant at the 0.05 level (2-tailed). Qualitative Responses Nine children selected at random provi ded qualitative answers to questions on forgiveness (see Appendix C for responses) of self, of others, and of situations. Of these

PAGE 47

39 nine respondents, three of them scored 7 or higher on the Lie scale (top 11% of sample), indicating social desirability may significan tly influence their responses. Two of the respondents scored a 10 or highe r on the CDI-S (top 6% of th e sample), and those same two participants scored 6 or below on the CASQ-R (lower 42% of sample). Of the nine respondents first asked, Is it hard for you to accept yourself when you mess up? three said no, four stated that some times it was, and one definitely said yes. When next asked Do you stay upset with ot hers who have hurt you? four indicated No outright, one said Not Really, two sa id Sometimes and one said Yes. Finally, when asked Is it hard for you to accept nega tive things that are not anyones fault? four stated unequivocally no, one indicated a qua lified yes by saying Its kinda hard, and three said yes outright-that it was hard to accept. Interestingly, though the words forgive or forgiveness were intentionally not used in the qualitative questions and prompts, two of the respondents used such words in their responses, both when asked what they do when hurt by others. One respondent who answered Forgive them correspondingly scor ed in the top 8% on the CADFI scale, zero on the CDI-S (lower 20%), and only one on the Lie scale (lower 35%). The other respondent clearly stating I forgave them scored in the upper 36% on the CADFI, the top 20% on the CASQ-R, and the lower 48% on the CDI-S, but scored in the top 2% on the Lie scale, indicating fairly significant social desirability influence. Respondents typically could a nd would provide an example for forgiveness of self, others, and situations when asked, however one participant exhibited high reluctance to give examples, instead offering Yes or N o as sufficient responses. This respondent was one of the two who also scored esp ecially high on the CDI-S scale (and low on

PAGE 48

40 CASQ-R) and also scored in the lower 13% for the CADFI. This respondent scored a zero on the Lie scale. This respondent is al so the only one to answer yes to all questions asking whether it was ha rd to forgive, be it self, ot hers, or situations, pointing to the speculation that the hesitation in sh aring may be due more to the difficulty experienced in such situati ons involving forgiveness.

PAGE 49

41 CHAPTER 5 DISCUSSION Given the various approaches to defi ning forgiveness and differing aspects emphasized among previous researchers, an efficient and compositional definition of forgiveness was sought for developing a child /adolescent forgiveness measure to fill a current gap in the literature. The current study investigated the relationship of dispositional forgiveness with depressive symp toms, and depressive attributional style, in children. Also, the utility and factor structure of the C ADFI, a version of the HFS modified for use with a younger age range, was examined. Though the initial three factor structure of self, others, and situations was not confirmed, exploratory factor an alyses revealed a clear and highly explainable two factor structure within the current sample, along th e lines of valence of item wording. Parallel analyses also suggested a str ong two factor structure. Given that these two factors were revealed to be highly uncorrelated, indicating th ey may in fact be two distinct entities, this study introduces the idea of unforgiveness as dis tinct from forgiveness itself. In other words, in children it may be more useful to think about the disposition for forgiveness as a general propensity for forgi veness (as revealed through positively worded items) and unforgiveness (negatively worded items). The measurement of forgiveness, as a construct, by emphasizing forgiveness as well as unforgiveness aspects is of course not unprecedented. For example, Maugers (1992) Fo rgiveness of Others and Forgiveness of Self scales, as well as McCulloughs TRIM s cale (1998), actually ta p unforgiveness as a means of looking at forgiveness with a dults. The current study provides emerging

PAGE 50

42 evidence that such scales measuring unforgive ness may not merely be measuring the lack of forgiveness, but a qualitatively dis tinct construct of unforgiveness. Though theoretically related, empirically these two c onstructs in children and adolescents do not appear to be signifi cantly correlated. When considered in the context of child rens psychological understanding of persons (Flavell, 2004), it is possible that younger children may not clearly understand or be capable of precise differentiation among perspect ives of forgiving self, forgiving others, and forgiving situations. The ability to clea rly understand ones rela tionship to self, to others, and to situations beyond anyones c ontrol could be a pre requisite to clear distinctions (and decisions) of forgiveness along the same dimensions, which may not be as highly attuned among children and young adolescents. Indeed, through the brief qualitative interviews, evidence of a lack of concrete understanding between what is considered a situation beyond anyones control a nd what might be considered an accident (i.e. participant 305s response). Given that measures us ed with children often employ a format of forced-choice among two generally opposing options, the positive or negative valence (reinforcing either/or ways of thinking) may pr ove more salient with younger children because of its structural simplicity. Further research able to replicate these results with children of vari ous populations would lend more strength to the two-factor item valence model, as well as provide impor tant insight into the relationship between forgiveness and unforgiveness as possibl y separate and distinct constructs. Despite the difference in fact or structure found in the curr ent sample, the relationship of forgiveness with construc ts of well-being was found to hold strong. Forgiveness levels significantly related to lower levels of depre ssive symptoms as well as lower depressive

PAGE 51

43 attributional styles. Those children who typica lly attributed positive events to internal, stable, and global aspects, and attributed negative events to external, unstable, and specific dimensions correspondingl y had higher levels of overall forgiveness. Those with higher depressive symptoms and the oppos ite attributional style showed lower forgiveness levels, indicating a highly positive relationship between healthy attributional styles and forgiveness. Future research might consider the implications of this relationship for counseling interventions, in that ways of increasing forgiveness could lead to lower levels of depres sive symptoms and depressive at tributional style, or that a reduction in depressive attri butions and symptoms may lead to increased levels of forgiveness. Given that social desirability is more pr evalent in younger age ra nges, the possibility of a significant correlation with this sample certainly existed and in fact was found. However, it has been asserted that correlations between a social desirability measure and a self-report measure do not necessarily je opardize construct validity (Diener, Sandvik, Pavot, & Gallagher, 1991; McCrae & Co sta, 1983; Thompson et al., 2005). Though the reliability and validity wi th this sample was supported, the generalizability of the findings in the current study is of course limited by the sample size and demographics. With a larger and more di verse sample, exploratory and confirmatory factor analyses may provide a replicable and more solid structural two factor model. A sample from a range of settings, as well as ages, would provide a deeper basis for understanding forgiveness in the child and adolescent population. For example, a crosssectional study examining forgiveness w ith different age groups would provide interesting developmental information regardi ng the progression of the factor structure of

PAGE 52

44 forgiveness and unforgiveness with age. If th e factor structure does change from a two factor model in children to a three factor model in adults, such a study could reveal the point at which this may occur. Examination of this measure in a more clinically depressed population may be bene ficial as well, considering th e majority of this sample did not report depressive symptoms in the clinical range. This study serves as an insightful beginni ng for further exploration of dispositional forgiveness in younger ages, both as it relate s to other constructs of well-being, and particularly the importance of understanding the factors that may make up forgiveness (or unforgiveness) with such a population. Future studies may wish to use the current scale with a larger sample for comparison, both to better understand the factor structure of forgiveness as it presents in children/adolescents, as well as the relationship of forgiveness to other constructs of wellbeing and positive psychology (i.e. altruism, cognitive flexibility, openness to experiences, et c) and more pathologi cal constructs such as neuroticism, perfectionism or negative affect. Furthermore, the brief qualitative responses obtained in this study encourage mo re extensive, qualitative interviews with younger populations in order to better inform future quantitative an alyses and provide a solid and sound basis for appropriate developm ent of tools for measuring forgiveness.

PAGE 53

45 APPENDIX A CADFI ITEMS AND FACTOR LOADINGS CADFI item Factor 1 Factor 2 2. I dislike myself for negative things I have done. .703 3. Learning from bad things that I have done helps me get over them. .516 4. It is really hard for me to accep t myself when Ive messed up. .546 5. With time I am understanding of myself for mistakes I have made. .486 6. I get upset with myself for negative things I have felt, thought, said, or done. .629 7. I keep punishing a person who has done something that I think is wrong. .368 9. I stay upset with others who have hurt me. .370 10. Even though others have hurt me in the past, with time I have been able to see them as good people. .582 13. When things go wrong for reasons that cant be controlled, I keep having negative thoughts about it. .396 14. With time I can be understanding of bad things in my life. .627 15. I keep thinking negatively about things in my life that I cannot control. .581 16. With time I make peace with bad things in my life. .409 17. It is hard for me to accept negative th ings that are not anyones fault. .554 18. With time I let go of negative thoughts about bad things that are beyond anyones control. .412

PAGE 54

46 APPENDIX B QUALITATIVE QUESTIONS Is it hard for you to accept yourself when you mess up? What is it like for you when you mess up? Can you think of a time when you did? Do you stay upset with others who have hurt you? What do you do? Can you think of an example? Is it hard for you to accept negative th ings that are not anyones fault? What is it like for you when something ba d happens and its not anyones fault? Can you think of a situation? What happened?

PAGE 55

47 APPENDIX C QUALITATIVE RESPONSES Participant 300 : Is it hard for you to accept yourself when you mess up? No. What is it like for you when you mess up? I get annoyed with myself. Can you think of a time when you did? When I dropped my baton twice. Do you stay upset with others who have hurt you? No. What do you do? Play with other friends. Can you think of an example? A day my friends said mean things. We are still friends I still forgave them. Is it hard for you to accept negative things that are not anyones fault? Ah man! What is it like for you when something ba d happens and its not anyones fault? I do something else. Can you think of a situation? What happened? When it rains when I want to go in the pool. Participant 12: Is it hard for you to accept yourself when you mess up? Sometimes. What is it like for you when you mess up? Can you think of a time when you did? Like when remembering my words (for school). Do you stay upset with others who have hurt you? No. What do you do? Like, I just try to ignore them until we both have forgotten about it. Can you think of an example? At computers, someone was annoying me and I ignore it. Today, we were getting our backpacks, he let the lid down and it hi t me on the cheek. You need to watch what youre doing. Is it hard for you to accept negative things that are not anyones fault? No. What is it like for you when something bad happens and its not anyones fault? Like, I keep going on with my life. Can you think of a situation? What happened? Like a car accident, he wrecked the van, it wasnt anybodys fault. Participant 11: Is it hard for you to accept yourself when you mess up? No. What is it like for you when you mess up? Can you think of a time when you did? Do you stay upset with others who have hurt you? What do you do? No. When someone hurt me physically? Can you think of a situation? When someone pushed me off my bike and almost sprained my ankle. That was 2 years ago. I forgave them.

PAGE 56

48 Is it hard for you to accept negative things that are not anyones fault? No. What is it like for you when something bad happens and its not anyones fault? Can you think of an example? What happened? Participant 306: Is it hard for you to accept yourself when you mess up? No. What is it like for you when you mess up? It doesnt bother me. Can you think of a time when you did? Saturday, I missed my baton. Do you stay upset with others who have hurt you? No. What do you do? I get upset, and I let it go. Can you think of an example? My friends, when they didnt w ant to hear about my trip I was so excited to go on. Is it hard for you to accept negative things that are not anyones fault? No. What is it like for you when something bad happens and its not anyones fault? I get upset. Can you think of a situation? Maggie moved away to Oklahoma. What happened? I got sad and dealt with it. Participant 305: Is it hard for you to accept yourself when you mess up? Sometimes. What is it like for you when you mess up? It is hard. Can you think of a time when you did? One time I messed up in horseback riding lessons. Do you stay upset with others who have hurt you? Not really. What do you do? I walk away and try to be calm. Can you think of an example? One time a person called me a name and I didnt like it. Is it hard for you to accept negative things that are not anyones fault? Its kinda hard. What is it like for you when something bad happens and its not anyones fault? I feel funny. Can you think of a situation? One time a glass bowl fell off the table. What happened? I was doing a dance and everyone was clapping and one moment the bowl was there and one moment it fell. Participant 304: Is it hard for you to accept yourself when you mess up? Sometimes. What is it like for you when you mess up? Hard. Can you think of a time when you did? At school.

PAGE 57

49 Do you stay upset with others who have hurt you? No. What do you do? I go tell the teacher. Can you think of an example? When another kid hurt my feel ings in music class. She yelled at me. Is it hard for you to accept negative things that are not anyones fault? Yes. What is it like for you when something bad happens and its not anyones fault? Sometimes I feel scared. Can you think of a situation? When my dog Belle got sick. What happened? Sometimes I get really scared feel ing is scary because you dont know whats happening. Participant 303: Is it hard for you to accept yourself when you mess up? Yes. What is it like for you when you mess up? I feel upset. Can you think of a time when you did? Yes, I told a lie. Do you stay upset with others who have hurt you? Sometimes. What do you do? Tell or keep it to myself. Can you think of an example? Today someone accused me for something. Is it hard for you to accept negative things that are not anyones fault? Yes. What is it like for you when something bad happens and its not anyones fault? Very sad. Can you think of a situation? What happened? My grandma got very sick. Participant 302: Is it hard for you to accept yourself when you mess up? Sometimes. What is it like for you when you mess up? I feel ashamed. Can you think of a time when you did? Yes. Do you stay upset with others who have hurt you? Sometimes. What do you do? Forgive them. Can you think of an example? When my friend and I fought at school. Is it hard for you to accept negative things that are not anyones fault? No, its sad. What is it like for you when something bad happens and its not anyones fault? Sad. Can you think of a situation? When my friend moved. What happened? She moved far away. Participant 301: Is it hard for you to accept yourself when you mess up? Yes. What is it like for you when you mess up? Bad.

PAGE 58

50 Can you think of a time when you did? No. Do you stay upset with others who have hurt you? Yes. What do you do? Can you think of an example? No. Is it hard for you to accept negative things that are not anyones fault? Yes. What is it like for you when something bad happens and its not anyones fault? Bad. Can you think of a situation? What happened? No.

PAGE 59

51 REFERENCES Bass, E., & Davis, L. (1994). The courage to heal. New York: Harper Perennial. Berry, J. W., Worthington, E. J., Parrott, L., OConnor, L. E., & Wade, N. G. (2001). Dispositional forgivingness: Develo pment and construct validity of the Transgression Narrative Test of Forgivingness (TNTF). Personality and Social Psychology Bulletin, 27 (10), 1277-1290. Beck, A. T., & Steer, R. A. (1987). The Beck Depression Inventory. San Antonio, TX: Psychological Corporation. Bentler, P. M. (1990). Comparative Fit indices in structural models. Psychological Bulletin, 107, 238-246. Cohen, J. (1988). Statistical power analysis for the behavioral sciences. (2nd ed.) Hillsdale, NJ: Erlbaum. Carpendale, J. I. M., & Lewis, C. (2004). Constructing an understanding of mind: The development of childrens social inte raction within social interaction. Behavioral and Brain Sciences, 27, 79-151. Castenada, A., McCandless, B. R., & Palerm o, D. S. (1956). The childrens form of the manifest anxiety scale. Child Development, 27 (3), 317-326. Catell, R. B. (1966). The scree test for number of factors. Multivariate Behavioral Research, 1, 245-276. Cavell, M. (2003). Freedom and forgiveness. International Journal of Psychoanalysis, 84 (3), 515-531. Crandall, V. C., Crandall, V. J., & Katkovsky, W. (1965). A childrens social desirability questionnaire. Journal of Consulti ng Psychology, 29 (1), 27-36. Crowne, D. P., & Marlowe, D. (1960). A new s cale of social desirabi lity independent of psychopathology. Journal of Consulting Psychology, 10, 3-31. Dadds, M. R., Perrin, S., & Yule, W. (1998). Soci al desirability and self-reported anxiety in children: An analysis of the RCMAS Lie scale. Journal of Abnormal Child Psychology, 26 (4), 311-317.

PAGE 60

52 Darby, B. W., & Schlenker, B.R. (1982) Childrens reactions to apologies. Journal of Social and Personality Psychology, 43(4), 742-753. Diener, E., Sandvik, E., Pavot, W., & Gallagh er, D. (1991). Response artifacts in the measurement of subjective well-being. Social Indicators Research, 24, 36-56. Enright, R. D., & Zell, R. L. (1989). Problem s encountered when we forgive another. Journal of Psychology and Christianity, 8, 52-60. Enright, R. D. (2000). Helping clients forgive: An em pirical guide for resolving anger and restoring hope. Washington, DC: American Psychological Association. Enright, R. D., Freedman, S., & Rique, J. (1998). The psychology of interpersonal forgiveness. In R. D. Enright & J. North (Eds.), Exploring forgiveness (pp.46-62). Madison: University of Wisconsin Press. Flavell, J. H. (2000). Development of ch ildrens knowledge about the mental world. International Journal of Behavioral Development, 24, 15-23. Flett, G. L., Hewitt, P. L., Davidson, L., & Munro, Y. (2001). The Child-Adolescent Perfectionism Scale: Development, valid ation, and association with adjustment (submitted for publication). Gladstone, T. R.G., & Kaslow, N. J. (1995). De pression and attributions in children and adolescents: A meta-analytic review. Journal of Abnormal Child Psychology, 23 (5), 597-606. Gorsuch, R. L. (1983). Factor analysis (2nd ed.). Hillsdale, NJ: Erlbaum. Hagborg, W. J. (1991a). The Revised Childre ns Manifest Anxiet y Scale and social desirability. Educational and Psychological Measurement, 51 (2), 423-427. Hagborg, W. J. (1991b). Adolescen t clients and perceived coun selor characteristics: A study of background characteristics, therap eutic progress, psychological distress, and social desirability. Journal of Clinical Psychology, 47 (1), 107-113. Hargrave,T. D., & Sells, J. N. (1997). Th e development of a forgiveness scale. Journal of Marital and Family Therapy, 23, 41-63. Hebl, J. H., & Enright, R. D. (1993). Forg iveness as a psychotherapeutic goal with elderly females. Psychotherapy, 30, 658-667. Hewitt, P. L., & Flett, G. L. (1991). Perfec tionism in the self and social contexts: Conceptualization, assessment, a nd association with psychopathology. Journal of Personality and Social Psychology, 60, 456-470.

PAGE 61

53 Hill, E.W. (2001). Understanding forgiveness as discovery: Implications for marital and family therapy. Contemporary Family Therapy, 23(4), 369-384. Kaslow, N. J., & Nolen-Hoeksema, S. (1991). Childrens Attributional Style QuestionnaireRevised. Unpublished manus cript, Emory University, Atlanta, GA. Katz, J., Street, A., & Arias, I. (1997). I ndividual differences in self-appraisals and responses to dating violence scenarios. Violence and Victims, 12 (3), 265-276. Klein, E. B., Gould, L. J., & Corey, M. (1969) Social desirability in children: An extension and replication. Journal of Consulting and Clinical Psychology, 33 (1), 128. Kovacs, M. (1985). The Children s Depression Inventory (CDI). Psychopharmacology Bulletin, 21, 995-998. Kovacs, M. (1992). The Childrens Depression Inventory. Manual. Toronto, Canada: Multi-Health Systems, Inc. Larzelere, R. E., & Huston, T. L. (1980). Th e dyadic trust scale: Toward understanding interpersonal trust in close relationships. Journal of Marriage and the Family, 42, 595-604. Lautenschlager, G. J. (1989). A comparison of alternatives to conducting monte carlo analyses for determining parallel analysis criteria. Multivariate Behavioral Research, 24(3), 365-395. Lopez, S. J., & Snyder, C. R. (2003). Positive psychological assessment: A handbook of models and measures. Washington, DC: American Psychological Association. Macaskill, A., Maltby, J., & Day, L. (2002). Forg iveness of self and others and emotional empathy. The Journal of Social Psychology, 142(5), 663-665. Maltby, J., Macaskill, A., & Day, L. (2001). Failure to forgive self and others: A replication and extension of the relationship between forgiveness, personality, social desirability and general health. Personality and Individual Differences, 30, 881-885. Marlowe, D., & Crowne, D. P. (1961). Social desirability and response to perceived situational demands. Journal of Consulting Psychology, 25, 109-115. Martin, M. M., & Rubin, R. B. (1995). A new measure of cognitive flexibility. Psychological Reports, 76, 623-626.

PAGE 62

54 Mauger, P. A., Perry, J. E., Freeman, T., Grove D. C., McBride, A. G., McKinney, K. E. (1992). The measurement of forgiv eness: Preliminary research. Journal of Psychology and Christianity, 11, 170-180. McCrae, R. R., & Costa, P. T., Jr., (1983). Soci al desirability scales: More substance than style. Journal of Consulting and Clinical Psychology, 51, 882-888. McCullough, M. E. (2000). Forgiveness as hum an strength: Theory, measurement, and links to well-being. Journal of Social and Clinical Psychology, 19, 43-55. McCullough, M. E., Pargament, K. I., & Thor esen, C. E. (Eds). (2000). The psychology of forgiveness: History, conceptual i ssues, and overview. In M.E. McCullough, K.I. Pargament, & C.E. Thoresen (Eds.), Forgiveness: Theory, research, and practice (pp.1-14). New York: Guilford Press. McCullough, M. E., Rachal, K. C., Sandage, S. J., Worthington, E. L., Jr., Brown, S. W., & Hight, T. L. (1998). Interpersonal fo rgiveness in close relationships II: Theoretical elaboration and measurement. Journal of Personality and Social Psychology, 75, 1586-1603. Nolen-Hoeksema, S. & Morrow, J. (1991). A pr ospective study of depression and distress following a natural disaster: Th e 1989 Loma Prieta earthquake. Journal of Personality and Social Psychology, 61, 105-121. Park, Y. O., & Enright, R. D. (1997). The de velopment of forgivene ss in the context of adolescent friendship conflict in Korea. Journal of Adolescence, 20 (4), 393-402. Pela, O. A., & Reynolds, C. R. (1982). Converg ent and divergent validity of the Revised Childrens Manifest Anxiety Scale: Normative and reliability data for Nigerian primary school children. Psychological Reports, 51, 1135-1138. Pina, A., Silverman, W. K., Saavedra, L. M., & Weems, C. F. (2001). An analysis of the RCMAS lie scale in a clinical sample of anxious children. Journal of Anxiety Disorders, 15 (5), 443-457. Reynolds, C. R., & Paget, K. D. (1981). Fact or analysis of the revised Childrens Manifest Anxiety Scale for blacks, white s, males and females with a national normative sample. Journal of Consulting and Clinical Psychology, 49 (3), 352359. Reynolds, C. R., & Richmond, B. O. (1978). What I think and feel: A revised measure of childrens manifest anxiety. Journal of Abnormal Child Psychology, 6, 271-280. Reynolds, C. R., & Richmond, B. O. (1985). Revised Childrens Manifest Anxiety Scale (RCMAS) manual. Los Angeles: Western Psychological Services.

PAGE 63

55 Ripley, J. S., & Worthington, E. L. (2002) Hope-focused and forgiveness-based group interventions to promot e marital enrichment. Journal of Counseling & Development, 80, 452-463. Rosenak, C. M., & Harnden, G. M. (1992). Forg iveness in the Psychot herapeutic process: Clinical applications. Journal of Psychology and Christianity, 11 (2), 188-197. Russell, D. W. (2002). In search of underlying dimensions: The use (and abuse) of factor analysis in Personality and Social Psychology Bulletin. Personality and Social Psychology Bulletin, 28(12), 1629-1646. Scobie, G. E. W., & Scobie, E. D. (2000). A comparison of forgiveness and pro-social development. Early Child Development and Care, 160, 33-45. Seligman, M. E. P., Peterson, C., Kaslow, N. J., Tanenbaum, R. L., Alloy, L. B., & Abramson, L. Y. (1984). Explanatory st yle and depressive symptoms among school children. Journal of Abnormal Psychology, 93, 235-238. Shriberg, L.D. (1974). Descriptive statistics fo r two childrens social desirability scales, general and test anxiety, and locus of control in elementary school children. Psychological Reports, 34, 863-870. Snyder, C. R., Crowson, J. J., Jr., Houst on, B. K., Kurylo, M., & Poirier, J. (1997). Assessing hostile automatic thoughts: De velopment and validation of the HAT Scale. Cognitive Therapy and Research, 21, 477-492. Stark, K. D., & Laurent, J. (2001). Joint f actor analysis of th e Childrens Depression Inventory and the Revised childre ns Manifest Anxiety Scale. Journal of Clinical Child Psychology, 30 (4), 552-567. Stuckless, N., & Goranson, R. (1992). The Vengeance Scale: Development of a measure of attitudes towards revenge. Journal of Social Behavior & Personality, 7 (1), 2542. Subkoviak, M. J., Enright, R. D., Wu, C. R ., Gassin, E. A., Freedman, S., & Olson, L. M., et al. (1995). Measuring inte rpersonal forgiveness in late adolescence and middle adulthood. Journal of Adolescence, 18, 641-655. Tabachnik, B. G., & Fidell, L. S. (1996). Using multivariate statistics (3rd ed.) New York: Harper Collins Publishers. Tangney, J., Fee, R., Reinsmith, C., Boone, A. L., & Lee, N. (1999, August). Assessing individual differences in the propensity to forgive. Paper presented at the annual meeting of the American Psychol ogical Association, Boston, MA.

PAGE 64

56 Thompson, B. (2004). Exploratory and confirmatory factor analysis: Understanding concepts and applications. Washington, DC: American Psychological Association. Thompson, M., Kaslow, N. J., Weiss, B., & Nolen-Hoeksema, S. (1998). Childrens Attributional Style Ques tionnaire-Revised: Psyc hometric examination. Psychological Assessment, 10(2), 166-170. Thompson, L. Y., & Snyder, C. R. (2003). Measur ing forgiveness. In S. J. Lopez and C. R. Snyder (Eds.), Positive psychological assessment: A handbook of models and measure s (pp. 301-312) Washington DC: American Psychological Association. Thompson, L. Y., Snyder, C. R., Hoffman, L., Michael, S. T., Rasmussen, H. N., & Billings, L., et al. (2005). Dispositional forg iveness of self, others, and situations. Journal of Personality, 73 (2), 313-360. Wade, N. G., & Worthington, E. L. (2003). Overcoming interpersonal offenses: Is forgiveness the only way to deal with unforgiveness? Journal of Counseling & Development, 81, 343-353. Walker, D. F., & Gorsuch, R. L. (2000). Forg iveness within the Big Five personality model. Personality and Individua l Differences, 32, 1127-1137. Watson, D., Clark, L. A., & Tellegen, A. ( 1988). Development and validation of brief measures of positive and negative affect: The PANAS scales. Journal of Personality and Social Psychology, 54, 1063-1070. Witvliet, C., Ludwig, T. E., & Bauer, D. J. (2002). Please Forgive Me: Transgressors Emotions and Physiology During Imager y of Seeking Forgiveness and Victim Responses. Journal of Psychology and Christianity, 21(3), 219-233. Worthington, E. L., Jr., & Wade, N. G. (1999). The psychology of unforgiveness and forgiveness and implications for clinical practice. Journal of Social and Clinical Psychology, 18 (4), 385-418. Worthington, E. L., Jr., Sandage, S. J., & Berry, J. W. (2000). Group interventions to promote forgiveness. In M.E. McCullough, K.I. Pargament, & C.E. Thoresen (Eds.), Forgiveness: Theory, research, and practice (pp.228-253). New York: Guilford Press. Worthington, E. L., Jr., Mazzeo, S. E., & Kl ewer, W. L. (2002). Addictive and eating disorders, unforgiveness, and forgiveness. Journal of Psychology and Christianity, 21(3) 257-261.

PAGE 65

57 BIOGRAPHICAL SKETCH I was born in Tampa, Florida on July 5th, 1980, and grew up in Temple Terrace and South Tampa until eighteen years of age, when I graduated from King High Schools International Baccalaureate Program. I attended the University of Florida as an undergraduate, majoring in psychology and supplementing my psychology major with concentrations in dance and biological sciences. I completed my undergraduate senior honors thesis in an area of developmental psychology, and graduate d summa cum laude in 2003 with my Bachelor of Science degree with a major in psychology. I then joined the Department of Psychol ogy at the University of Florida as a graduate student in the Counseling Psyc hology Program. I completed my Master of Science degree in the summer of 2006.