Citation
Predicting maternal adherence and child health status in childhood epilepsy

Material Information

Title:
Predicting maternal adherence and child health status in childhood epilepsy
Creator:
Fritz, Sobha P., 1975-
Publication Date:

Subjects

Subjects / Keywords:
Child psychology ( jstor )
Demography ( jstor )
Diseases ( jstor )
Emergency departments ( jstor )
Emotional distress ( jstor )
Epilepsy ( jstor )
Health status ( jstor )
Medications ( jstor )
Pills ( jstor )
Seizures ( jstor )

Record Information

Rights Management:
Copyright [name of dissertation author]. Permission granted to the University of Florida to digitize, archive and distribute this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Resource Identifier:
30435808 ( ALEPH )
54703266 ( OCLC )

Downloads

This item has the following downloads:


Full Text










PREDICTING MATERNAL ADHERENCE AND CHILD HEALTH
STATUS IN CHILDHOOD EPILEPSY: AN EXPLORATORY MODEL














By

SOBHA P. FRITZ


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

UNIVERSITY OF FLORIDA


2003


























This document is dedicated to my husband, parents, and sisters for their support and
unconditional love. My family's emphasis regarding the importance of education and
helping others has guided me into a career for which I am well suited. I would like to
especially thank my husband, Jason Jon Fritz, for always believing in me and providing a
secure foundation of warmth and encouragement.














ACKNOWLEDGMENTS

This investigation could not have been carried out without the mentorship and

direction of Robert Glueckauf, Ph.D., the valuable contributions ofNicole Jagusztyn, the

dedication of Paul Carney, M.D., and the staff at the University of Florida Pediatric

Neurology Clinic. I would like to acknowledge my committee members for their

scholarly assistance and guidance. I would additionally like to thank my family for their

support and encouragement throughout this project.














TABLE OF CONTENTS
Page

ACKN O W LED GM EN TS ............................................................................................ iii

ABSTRA CT...................................................................................................................... vii

CHAPTER

1 IN TRODUCTION ........................................................................................................ 1

Epilepsy: An Overview ................................................................................................. I
Etiology.........................................................................................................................2
D iagnosing Epilepsy..................................................................................................... 3
Classification of the Epilepsies..................................................................................... 4
Partial Seizures ...................................................................................................... 4
Generalized Seizures ............................................................................................. 6
Treatm ent......................................................................................................................8
Psychological Im pact of Epilepsy ................................................................................ 9
A dherence to the Regim en: An Overview .................................................................. 10
Adherence to the Regim en: Epilepsy ......................................................................... 14
M aternal Em otional Functioning................................................................................ 18

2 A IM S AN D STUD Y JU STIFICATION S .................................................................. 21

Purpose .......................................................................................................................21
Study M odel................................................................................................................22
Hypotheses..................................................................................................................23
Maternal Emotional Distress, Adherence, Child Health Status .......................... 23
Maternal Perceptions of the Treatment Regimen, Adherence, Child Health
Status................................................................................................................ 23
Condition and Regimen-related Variables, Adherence, Child Health Status......23
Demographic Characteristics, Adherence, Child Health Status.......................... 24

3 M ETH OD S ................................................................................................................ 25

Study Design...............................................................................................................25
Participants .................................................................................................................26
M measures .....................................................................................................................29
Participant Background Inform ation................................................................... 29
M medical Chart Inform ation.................................................................................. 30








Child Health Status and Adherence..................................................................... 31
Maternal Emotional Distress............................................................................... 32

4 RESULTS ................................................................................................................... 35

Preliminary Analyses.................................................................................................. 35
Primary Analyses........................................................................................................ 38
Correlations among Background Variables, Predictors, Criterion Variables......38
Hypothesis 1: Maternal Emotional Distress, Adherence, Child Health Status ...40
Hypothesis 2: Maternal Perceptions of the Treatment Regimen, Adherence,
Child Health Status .......................................................................................... 42
Hypothesis 3: Condition and Regimen-related Variables, Adherence, Child
Health Status .................................................................................................... 46
Hypothesis 4: Demographic Characteristics, Adherence, Child Health Status...49
Model Testing...................................................................................................... 51
Prediction of adherence................................................................................ 51
Prediction of health status ............................................................................ 52
Summary of Findings ................................................................................................. 55

5 DISCUSSION............................................................................................................. 57

Sample Characteristics of Demographics and Maternal Distress............................... 57
Hypotheses.................................................................................................................. 58
Hypothesis 1: Maternal Emotional Distress, Adherence, Child Health Status ...58
Hypothesis 2: Maternal Perceptions of the Treatment Regimen, Adherence,
Child Health Status .......................................................................................... 60
Hypothesis 3: Condition and Regimen-related Variables, Adherence, Child
Health Status .................................................................................................... 61
Hypothesis 4: Demographic Characteristics, Adherence, Child Health Status...63
Model Testing............................................................................................................. 64
Study Strengths and Limitations................................................................................. 65
Future Directions ........................................................................................................ 67
Clinical Implications................................................................................................... 69








APPENDIX

A BACKGROUND INFORMATION FORM TIME 1 .............................................. 71

B BACKGROUND INFORMATION FORM TIME 2 .............................................. 75

C MEDICAL CHART INFORMATION FORM TIME 1 ......................................... 77

D MEDICAL CHART INFORMATION FORM TIME 2 ......................................... 78

LIST O F REFEREN CES................................................................................................... 79

BIOGRAPHICAL SKETCH............................................................................................. 83














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

PREDICTING MATERNAL ADHERENCE AND CHILD HEALTH STATUS IN
CHILDHOOD EPILEPSY: AN EXPLORATORY MODEL

By

Sobha P. Fritz

December 2003

Chair: Robert L. Glueckauf
Major Department: Clinical and Health Psychology

Epilepsy is a chronic illness that typically begins in early childhood. Poor

adherence to the medication regimen can result in increased seizure frequency and

severity. Young children are reliant on their parents for their daily routine and

functioning. This issue is particularly salient for children with seizure disorders, who

rely on their parents to take anti-convulsant medicationss. Depressive symptomatology

and maternal stress have been associated with negative child status. Maternal emotional

factors such as depression and stress may play an important role in adherence to

medication for younger children with epilepsy. In the current study, we examined the

relationship between maternal adherence to medication (pill count percentage) and child

health status (seizure frequency, seizure-related emergency room visits) for children with

epilepsy and the following factors: maternal emotional distress (maternal depression,

maternal stress), demographics (family annual income, race, maternal educational level),

condition-related and regimen-related variables (disease type, disease duration, type and








complexity of the medication regimen, concordance between prescribed regimen and

reported regimen), and maternal perceptions of the treatment regimen (perceived

efficacy, perceived knowledge, perceived difficulty).

Research participants included children, ages one through eleven, diagnosed with

epilepsy and were approached for study participation at the University of Florida

Pediatric Neurology Clinic. Eighty-two mother-child dyads completed the study.

Results indicated that as maternal stress, parent-child dysfunction, and maternal

perceived regimen difficulty increased, child seizure frequency significantly increased.

As maternal perceived regimen efficacy levels increased, child seizure frequency

significantly decreased. Results revealed that as maternal perceived regimen difficulty

levels increased, seizure-related emergency room (ER) visits significantly increased. As

regimen concordance rates increased, seizure-related ER visits significantly decreased.

Results demonstrated that as concordance rates increased, pill count increased.

Final hierarchical regression analyses suggested that parent-child dysfunction was

predictive of seizure frequency. Maternal perceived regimen difficulty ratings and

concordance ratings between the prescribed and reported regimen were predictive of

seizure-related ER visits. Concordance ratings between the prescribed and reported

regimen were predictive of pill count percentage. These findings suggest that the

identification of specific variables that influence maternal adherence and child health

status may be particularly helpful for children with epilepsy.













CHAPTER 1
INTRODUCTION

Epilepsy: An Overview

Epilepsy is a chronic illness that typically begins in early childhood in which the

patient is prone to experience recurrent epileptic seizures, where two or more attacks

have occurred. Epilepsy is defined as a manifestation of an abnormal and excessive

synchronized discharge of a set of cerebral neurons (Shorvon, 2000). Clinical

manifestations are sudden and transient and can include a wide variety of motor, psychic,

and sensory phenomena with or without alteration in consciousness or awareness. The

specific symptoms that occur depend on the part of the brain involved. Status epilepticus

is defined as a condition in which epileptic seizures continue, or are repeated without

consciousness, for a period of 30 minutes or more and this occurrence often requires a

visit to the emergency room. Untreated status epilepticus can be fatal.

Provoked seizures have an obvious and immediate preceding cause, such as an

acute systemic, metabolic, or toxic insult or an acute cerebral event (stroke, trauma, and

infection). Provoked seizures are often isolated events and do not tend to recur once the

cause is removed. A person is described as having active epilepsy when at least one

seizure has occurred in the preceding 2-5 years. Epilepsy is said to be in remission when

no seizures have occurred in the preceding 2-5 years (Appleton & Gibbs, 1998). There

are approximately 80 cases of epilepsy per 100,000 people per year, with a range of 50-

120 cases. Nearly 75 per 100,000 adolescents per year go on to develop some form of

epilepsy. The cumulative incidence, risk of an individual developing epilepsy in his or








her lifetime, is between 3% and 5%. The highest incidence rates are observed in

neonates and young children with a second peak in old age. Greater than 50% of seizures

start in childhood and begin before age 20 (Browne & Holmes, 2000).

The prevalence of active epilepsy is 5-10 cases per 1,000 people per year. Isolated

seizures (first and only) occur in approximately 20 people per 100,000 each year. The

prevalence of epilepsy in school age children (aged 5-17 years) is 0.7%-0.8%. The

prevalence of epilepsy in adolescents is 6-7 per 1,000 adolescents. One-third of all

epilepsies that begin in childhood will have spontaneously remitted by puberty and this

change is usually sustained throughout adult life. Epilepsy is relatively static after early

childhood but shows a tendency to rise in old age. Epilepsy is more common in

developing countries and is thought to be associated with poorer nutrition, poorer

hygiene, higher prevalence of infectious diseases, and a higher proportion of children in

the population (Appleton & Gibbs, 1998).

Etiology

Congenital and genetic conditions are most commonly related to early childhood

epilepsy. Genetics are thought to play an important role in seizure disorders. Hereditary

and genetic factors are related to idiopathic or primary epilepsy syndromes and possibly

in some secondary epilepsies, but known single-gene disorders are related to epilepsy in

only 1-2% of cases. In older children and young adults, common causes of epilepsy

include inherited predisposition, hippocampal sclerosis, alcohol or drug abuse, and

trauma. Vascular disease is a predominant cause of epilepsy in the elderly. Tumors and

infections can occur at any age but malignant tumors are more likely to occur after age 30

(Shorvon, 2000). In 70-75% of childhood cases of epilepsy, a specific cause may not be








noted, but the etiology of the disease may be apparent from clinical information and test

results (Appleton & Gibbs, 1998).

Patients with specific syndromes such as Fragile X syndrome, Sturge-Weber

syndrome, Huntington's, Niemann-Pick disease type C, tuberous sclerosis (Boumrneville's

syndrome), neurofibromatosis (type 1 and type 2), and Down's syndrome may develop

epilepsy. Infectious diseases that involve the brain such as viral and bacterial meningitis

also lead to epilepsy. Cerebral tuberculosis frequently causes epilepsy. Malaria, strep

infections, and encephalitis are all thought to be related epilepsy. Cerebral tumors are

related to approximately 6% of newly diagnosed cases of epilepsy. In children, tumors

are not typical causes of epilepsy. Head injuries account for approximately 2-12% of all

cases of epilepsy (traffic accidents and falls). Drugs, alcohol, toxins, and metabolic

disorders have been associated with seizure disorders. Additional causes of epilepsy

include brain malformations, unruptured aneurysms, intracranial hemorrhages, lesions,

perinatal injury, and hippocampal sclerosis. Hippocampal sclerosis is the most common

cause of temporal lobe epilepsy (Shorvon, 2000).

Diagnosing Epilepsy

A physician needs detailed information to render an epilepsy diagnosis and a

diagnosis of epilepsy is dependent almost entirely on the history given. During the

assessment the doctor will typically gain the following information: a very detailed

history of the seizure(s), an eyewitness account of the seizure(s), a history of previous

illnesses or traumatic events, family history, and history of any co-occurring diseases)

and treatmentss. Noteworthy is the fact that an accurate eyewitness account is essential

and may be the only history available for young children. A physical examination will be

conducted. Video recordings, sleep studies, and EEGs of suspected seizures are useful








aids for diagnosis. Some characteristics that may mimic epilepsy include headaches,

jitters, tic disorder, shuddering, shivering, repetitive ritualistic movements, vacant stares,

daydreaming, inattention, psychological disturbances, and excessive sleep (Browne &

Holmes, 2000).

Classification of the Epilepsies

Seizures are classified into two main categories: partial and generalized. These two

categories are further classified. Partial seizures include simple partial, complex partial,

and partial seizures evolving to secondary generalized. Generalized seizures are

classified as typical absence (petit mal), atypical absence, myoclonic, clonic, tonic, tonic-

clonic (grand mal), atonic, and unclassifiable (Appleton & Gibbs, 1998).

Partial Seizures

Partial seizures are further classified as "simple," in which consciousness is

retained, or "complex," in which consciousness is impaired or lost. Simple partial

seizures (beginning locally) can occur with motor symptoms, sensory symptoms,

autonomic symptoms, and psychic symptoms. Simple partial seizures with sensory,

autonomic, or psychic symptoms may be easily overlooked in younger children, who may

be unable to describe such symptoms. Partial seizures may become secondary

generalized, resulting in tonic-clonic (grand mal) convulsions (Browne & Holmes, 2000).

Simple partial seizures last only a few seconds and the symptoms that occur depend

on the part of the cortex involved. Simple partial seizures can include motor

manifestations such as jerking or spasms, which usually occur from frontal or central

region epilepsies. Simple partial seizures can include somatosensory manifestations such

as simple hallucinations that can take the form of tingling or numbness, an electrical

shock-like feeling, burning, pain, or a feeling of heat. The epileptic focus is usually in








the central or parietal regions. Simple visual phenomena such as flashing lights and

colors usually occur if the calcarine cortex is affected. A rising epigastric sensation is the

most common manifestation reported, which typically occurs when the mesial temporal

lobe is affected. Simple partial seizures can also include autonomic manifestations such

as changes in skin color, blood pressure, heart rate, and pupil size and these changes

usually are related to generalized or complex partial seizures of frontal or temporal

origins. Psychic manifestations are common as well. Psychic auras can take various

forms and are more common in complex partial seizures. They usually occur from

epilepsy arising from a temporal, frontal, or parietal focus. There are six principal

categories of auras: dysphasic symptoms (cortical or speech areas are effected),

dysmnestic symptoms (disturbance of memory), cognitive symptoms (dreamy states and

depersonalization), affective symptoms (fear, depression, anger, irritability, and elation),

illusions of size, shape, weight, distance or sound, or structural hallucinations (visual,

auditory, gustatory, or olfactory forms) (Appleton & Gibbs, 1998).

Complex partial seizures have three main components: auras, altered

consciousness, and automatisms. The auras are equivalent to those of simple partial

seizures and usually last a few seconds, but in rare cases a prolonged aura persists for

minutes, hours, or even days. Altered consciousness may follow the aura and can include

motor arrest, during which the patient is motionless. The patient may be described as

vacant or glazed and often associated is a spasm, posturing, or mild jerking.

Automatisms are involuntary motor actions that occur during or in the aftermath of

epileptic seizures, in a state of unconsciousness where there is total amnesia for the

events.








Partial seizures (simple, complex, or simple evolving to complex) may spread to

become generalized. The partial seizure is often experienced as an aura in the seconds

before the generalized seizure. The generalized seizure is usually tonic-clonic, tonic, or

atonic (Browne & Holmes, 2000).

Generalized Seizures

There are two types of absence seizures: typical absence (petit mal seizures) and

atypical absence. In typical absence seizures there is an abrupt and sudden loss of

consciousness, cessation of all motor activity, tone is usually preserved, there is no fall,

the patient is unaware and inaccessible, the patient often appears glazed or vacant, the

attack ends as abruptly as it started, and previous activity is resumed as if nothing

happened. In atypical absence seizures, the duration is longer, loss of awareness is often

incomplete, associated tone changes are more severe, and the onset and cessation of

attacks are not as abrupt (Appleton & Gibbs, 1998).

With myoclonic seizures, there is a brief contraction of a muscle, muscle group, or

several muscle groups that can be single or repetitive, varying in severity from a nearly

imperceptible twitch to a severe jerking which can result in a sudden fall or the

propulsion of handheld objects (flying saucer syndrome). Clonic seizures consist of

asymmetrical and irregular jerking. Clonic seizures are most frequent in neonates,

infants, and young children. Tonic seizures are associated with muscle contractions with

altered consciousness. In tonic seizures, the contraction causes extension of the neck,

contraction of facial muscles, widening of the eyes, up-turning of the eyeballs,

contraction of respiratory muscles, and spasms of the upper limb muscles (Browne &

Holmes, 2000).








Tonic-clonic seizures are most associated with the type of epilepsy that the public

imagines. The tonic phase is sometimes preceded by a period in which an attack is

anticipated (prodromal period) defined by a vague feeling or by the occurrence of

increasing myoclonic jerking. If an aura occurs, this indicates that the tonic-clonic

seizure is secondary generalized. The seizure is initiated by loss of consciousness where

the patient will fall if standing and there is a brief period of tonic flexing and then a

longer phase of rigidity, with the eyes rolled up, the jaw clamped shut, the limbs stiff, and

the fists clenched, then respiration ceases. The clonic phase of tonic-clonic seizures has a

tonic stage that lasts 10-30 seconds and is followed by a clonic phase, during which

convulsive movements occur, usually of all four limbs and the jaw and facial muscles.

Breathing may be labored and saliva may froth from the mouth containing blood from

tongue-biting. The convulsive movements decrease in frequency, typically to about 4

clonic jerks per second, and increase in amplitude as the attack progresses. Autonomic

features (flushing, changes in BP, changes in pulse, and increased salivation) are

common. The clonic phase lasts 30-60 seconds and is followed by further brief tonic

contraction of the muscles, sometimes accompanied by incontinence. The final phase

lasts between 2-30 minutes and is characterized by flaccidity of muscles and then

consciousness is slowly regained (Browne & Holmes, 2000).

The most severe form of the atonic seizure is the classic drop attack in which all

postural tone is suddenly lost, causing the person to collapse. In atonic seizures, tone can

be more restricted, resulting in bobbing of the head, a bowing movement, or sagging at

the knee. These seizures are short and are usually followed by immediate recovery.








At least 1/3 of all seizure disorders are not classifiable and fall into the

unclassifiable category. Seizures are described as unclassifiable because typical clinical

and EEG patterns do not conform to the defined categories of seizures (Browne &

Holmes, 2000).

Treatment

Various drug treatments for epilepsy are available. One of the most difficult

aspects of drug treatment is adherence to medication. Some issues that arise are when to

start drug therapy, which drug to use, what dose to give, when to change the drug and/or

add another drug, when to stop a drug, and when to measure blood levels of the drug.

Drug treatments can vary from dosing once or twice a day to several times a day and drug

types include pills, liquids, sprinkles, chewables, suppositories, and inhalers (Appleton &

Gibbs, 1998).

Most physicians would recommend beginning drug treatment after a cluster of

seizures. Once a drug regimen is started, the goal is to achieve seizure control without

extreme side effects. Before adding additional medications, the dose of the initial drug

should be increased to the maximum tolerated level. Currently, the recommended first-

line of drugs in treating childhood epilepsy is sodium valproate for generalized

epilepsies, and carbamazepine for partial seizures. The second line of defense includes

carbamazepine, gabapentin, lamotrigine, and topiramate for generalized seizures, and

vigabatrin, gabapentin, lamotrigine, and topiramate for partial seizures. The third line of

defense includes phenytoin, ethosuximide, clonazepam, phenobarbitone, clobazam,

carbamazepine, acetazolamide, and topiramate for generalized seizures, and sodium

valproate, clobazam, and phenytoin for partial seizures. Many physicians take a polydrug








treatment approach of epilepsy treatment, specifically for those patients who do not

respond well to first, second, and third line treatment approaches (Shorvon, 2000).

Surgical treatment is an option for patients with severe intractable epilepsy that

fails to respond to medication and other forms of treatment. Alternative methods of

treatment include the ketogenic diet, steroids, intravenous immunoglobulins, behavior

therapy, acupuncture, herbal therapy, vagus nerve stimulation, and seizure alert dogs.

Psychological Impact of Epilepsy

Psychiatric disorders occur more frequently with epilepsy than with any other

chronic disorder (Appleton & Gibbs, 1998). Children and adolescents with poorly

controlled epilepsy have more restrictions imposed upon them to protect them from

getting injured during a seizure. They may be unable to participate in activities with their

peers such as swimming, driving, cycling, climbing, and sports. These restrictions can

result in fewer contacts with peers. Social stigma concerns such as teasing, isolation, and

poor social contact have been associated with increased restrictions. Restrictions,

depression, and isolation have been associated with poorer adherence in children and

adolescents with epilepsy (Kyngas, 1990). Poorer adherence to medication is associated

with increased seizure frequency, resulting in a larger number of epilepsy-related injuries

and deaths. One preventable cause of increased seizure frequency and its associated

injuries is taking the anti-epileptic medication as prescribed. Adherence to the medical

regimen becomes an important factor in decreasing seizure frequency.

Epilepsy is a disorder that is associated with a high degree of uncertainty. This

uncertainty and sudden loss of control have been associated with anxiety and depression

in patients with epilepsy including adults, children, and caregivers (Kyngas, 2000).

Increased stress levels and anxiety have been demonstrated as related to an increase in








seizure frequency and seizure severity (Spector et al., 1999). The effects of maternal

emotional adjustment and adherence with young children with epilepsy have received

limited attention.

Adherence to the Regimen: An Overview

One-third of all patients and 50-55% of chronically ill patients are non-adherent to

their prescribed medical regimens (Rapoff, 1999). Non-adherence can negatively impact

health status and health-related quality of life and can lead to increased risks of morbidity

and mortality. Non-adherence is also related to increased health care costs, where the

cost of non-adherence in the United States is nearly $100 billion each year (Rapoff,

1999). Additionally, non-adherence may inadvertently negatively impact clinical

decisions made by physicians. Physicians who attribute disease-related problems to

ineffective or incorrect prescriptions and not to the patient's adherence problems may

change the patient's medical regimen unnecessarily or inappropriately.

Identifying correlates of adherence is important because they facilitate the

development of risk profiles, medical regimens, and future research using matched

samples. Patient and family factors, disease-related factors, and treatment regimen

factors have been identified as correlates of adherence (Rapoff, 1999). The majority of

studies have focused exclusively on patient and family factors and their relationship to

adherence.

Patient/family correlates include demographics (child age, parental education level,

work status, number of children in the home, and socioeconomic status), knowledge

about disease course and treatment, as well as adjustment and coping. Previous research

has shown that patient and family demographic variables have a significant influence on

child and maternal adherence. Younger children with cancer, cystic fibrosis, diabetes,








and renal disease are more likely to be adherent to their medication regimen compared to

adolescents (Anderson, Auslander, Jung, Miller, & Santiago, 1990; Beck et al., 1980;

Bond, Aiken, & Somerville, 1992; Brownbridge & Fielding, 1994; Gudas, Koocher, &

Wyplj, 1991; Jacobsen et al., 1987; LaGreca, Follansbee, & Skyler, 1990; Patterson,

1985; Smith, Rosen, Truworthy, & Lowman, 1979; Tebbi et al., 1986). Boys tend to be

less adherent than girls to regimens for cystic fibrosis and diabetes (Lorenz, Christensin,

& Pichert, 1985; Patterson, 1985). Lower socioeconomic status (SES) and lower parental

education levels have been related to non-adherence to regimens for asthma, cystic

fibrosis, diabetes, and renal disease. Patients in larger families or whose mothers work

outside of the home tend to be less adherent to cancer and cystic fibrosis regimens

(Patterson, 1985; Tebbi et al., 1986).

Knowledge about the treatment regimen is an interesting and important factor in

predicting adherence to treatment. However, studies that have examined this factor have

shown an inconsistent pattern of results. Several investigators have reported that patients

who are less knowledgeable about their specific condition and treatment regimen tend to

be less adherent to regimens for cancer, cystic fibrosis, and diabetes (Gudas et al., 1991;

LaGreca et al., 1990; Tebbi et al., 1986) whereas knowledge appears to be unrelated to

adherence for patients with renal disease (Beck et al., 1980). Maternal knowledge for

children with diabetes was positively related to adherence for preadolescents and

unrelated to adherence for adolescents (LaGreca et al., 1990).

The relationship of adjustment and coping to adherence has been examined across a

variety of childhood disorders. Poor parental coping has been associated with non-

adherence to medical regimens for children with juvenile rheumatoid arthritis and renal








disease. Increased parental depression has been related to poor adherence to renal disease

regimens (Brownbridge & Fielding, 1994; Wynn & Eckel, 1986). Greater parental

anxiety and perceived stress have been associated with lower adherence to anti-epileptic

medications for children with epilepsy (Hazzard, 1990). Parental compliance behavior is

related to perceptions of the medications' efficacy as well as its side effects (Shope,

1988). Austin (1989) suggests that attitudes toward giving anti-epileptic medications and

behavioral intentions are important variables in the prediction of parental compliance

behavior. Parental knowledge about the efficacy and side effects of medications

influences parental attitudes towards medications and improves their medication-giving

behavior (Austin, 1989).

Patients who lack parental monitoring of regimen procedures or have families

where the role of the parental monitor is ambiguous tend to be less adherent to their

prescribed medical regimen. Parental monitoring has been found to be a key factor in

improving medication adherence for adolescents with chronic illness (Rapoff, 1999).

Condition-specific correlates include duration, course, symptoms, and perceived

severity of the condition. Diseases with a longer duration are related to poorer adherence.

Poorer adherence to regimens for diabetes, juvenile rheumatoid arthritis, and renal

disease has been related to longer illness duration. The relationship of disease course and

adherence has not been specifically examined for children with chronic illness (Rapoff,

1999). Condition-specific symptoms are an important factor to examine in their

relationship to adherence. Increased health problems and hospitalizations are related to

poorer adherence for patients with renal disease (Beck et al., 1980). Greater seizure

activity has been related to decreased adherence to anticonvulsant medications for








children with epilepsy (Hazzard, 1990). Maternal perceptions of higher disease severity

have been associated with better adherence to asthma medications (Radius et al., 1978).

Medication regimen-related correlates include type and complexity, costs, side

effects, and perceived efficacy of the medication regimen. Patients with more complex

regimens tend to be less adherent to their prescribed regimen in that more complex

regimens are more often associated with polydrug treatments, with higher numbers of

medication types and dosages. This is particularly true for patients with cystic fibrosis,

diabetes, and juvenile rheumatoid arthritis. Treatment costs can be related to non-

adherence especially for those patients and families who are economically disadvantaged.

Those medical regimens that produce a greater number of negative side effects tend to be

related to poorer adherence to the regimen. Patient and parental perceptions of

medication efficacy are especially related to adherence. Higher levels of perceived

benefits rated by patients and parents are associated with better adherence to regimens for

asthma and diabetes (Bobrow, Avruskin, & Siller, 1985; Bond et al., 1992; McCaul,

Glasgow, & Schafer, 1987; Radius et al., 1978).

In summary, important correlates of adherence for children and adolescents with

chronic illness include patient/family factors (demographics, knowledge, adjustment and

coping, parental monitoring), disease factors (duration, course, symptoms, perceived

severity), and regimen factors (type and complexity, costs, side effects, efficacy). These

summarized findings demonstrate that correlates of adherence to medical regimens are

critical in improving adherence. The majority of studies conducted to identify correlates

of adherence to medical regimens for children with chronic disease have focused on the

following: asthma, cancer, cystic fibrosis, diabetes, and renal disease. Studies examining








the specific correlates of adherence for children and adolescents with epilepsy are limited

in number and require further examination.

Adherence to the Regimen: Epilepsy

The vast majority of people with epilepsy begin having their seizures before the age

of 20, and more than 50% of epilepsy cases begin in childhood. Thus, most people

experience their first seizure at a time that is critical to the acquisition and development

of basic cognitive and social competencies, which are crucial for long-term academic,

interpersonal, and vocational adjustment. Due to the long-term nature of epilepsy and its

related medication regimen, medication adherence to anti-epileptic drugs (AEDs) is

extremely important from the time of diagnosis and throughout the course of treatment.

Although correlates of adherence have been identified in a wide range of pediatric

chronic illnesses, only a few studies have been conducted to date on adherence to

treatment among children and adolescents with seizure disorders. We continue to lack

basic information about the psychological, environmental, and familial factors that

influence adherence in children with epilepsy. Furthermore, there is only a modicum of

outcome research on intervention approaches to improve adherence in this at-risk

population. A few factors have been identified that are specifically related to individuals

with seizure disorders.

Well-controlled seizure patients are asymptomatic, which may gradually contribute

to decreased adherence to the medication regimen. Furthermore, nonadherence does not

necessarily lead to immediate symptomatology for patients with epilepsy. Pediatric

seizure patients are typically on long-term medication regimens, for which adherence is

worse than for short-term regimens (Becker & Maiman, 1975; Davis, 1966; Rapoff&

Christopherson, 1982).








A demographic factor generally associated with parental non-adherence is lower

socioeconomic status. Additionally, ethnic minority patients have been found to be more

likely to be nonadherent to their child's medical regimen (Becker & Maiman, 1975;

Rapoff & Christopherson, 1982).

Conducting research on adherence with pediatric seizure patients is made difficult

by the methodological problems involved in assessing patient's adherence. Although

blood levels of anticonvulsant medications can be obtained, it is difficult to translate

blood levels into comparable measures of compliance for patients on different

medications and doses. Therapeutic ranges vary for different medications and vary

according to dosage, time since the medication was ingested, and individual metabolic

rates, and hormonal changes (which are especially salient for adolescent females).

Furthermore, for patients taking more than one medication, there are medication

interactions, with some medications altering the blood level concentrations of other

medications. Because of these complications, studies examining pediatric patient

adherence to anticonvulsant medication regimens have tended to assess only one

medication level or have used group differences in blood levels to assess the validity of

other compliance measures (self report, pill counts). With these limitations, estimates of

parental adherence to anticonvulsant medication regimens for pediatric patients ranges

from 25% to 54% (Zysset, et al., 1981). These estimates are generally consistent with

adherence rates from other pediatric populations where a long-term medication has been

prescribed (Sackett & Haynes, 1976).

Only three studies have assessed factors related to parental medication adherence

with pediatric seizure patients (Freidman et al., 1986; Peterson et al., 1982). The studies








assessed at least some factors from the demographic, illness, treatment regimen, and

attitudinal domains. However, the Peterson et al. study focused primarily on adults, with

only 18 adolescents in the sample. Demographic variables were not demonstrated to be

associated to adherence in the pediatric seizure disorder populations in the two studies

(Freidman et al., 1986; Peterson et al., 1982). In terms of illness variables, Peterson et al.

(1982) found that patients experiencing more seizures in the past year, more recent

seizures, or having tonic-clonic seizures were more likely to be adherent. The only

treatment-related variable related to adherence for children who had an active role in

taking their medication, particularly adolescents, was perceived personal independence

(Freidman et al., 1986). Research in children with other medical conditions has also

shown that children with greater personal freedom (or fewer behavioral restrictions) were

more likely to be adherent with their medication regimen. Peterson et al. (1982) found

that satisfaction with the physician's explanation was unrelated to adherence. However,

in studies with other populations, patient satisfaction and good patient-doctor

communication have been related to good adherence. Peterson et al. (1982) also found

that worry about one's own health was related to self-reported adherence.

Shope reviewed a study that closely followed the Health Belief Model (HBM)

framework for pediatric epilepsy patients (Shope, 1992). Parents of a random sample of

patients drawn from an urban medical center pediatric seizure clinic were seen at their

regular appointment and interviewed from January of 1975 through May of 1975.

Questions asked included variables from the HBM as well as a wide range of medical,

psychosocial, and demographic variables. Serum antiepileptic levels were taken at the

clinic visit on 65 of the 90 study participants. Data presented were based on 65 patients.








Patients were judged to be compliant or non-compliant by an experienced clinician's

relating the prescribed dosage of medication to the obtained blood level. This process

was used, rather than therapeutic blood levels, because some patients had less than

therapeutic dosages prescribed.

The patient sample was 55% female and 78% African American with a mean age of

10 years. For 55% of the children, seizure disorder was the only medical problem

identified. Age of onset ranged from birth to 11.5 years and duration of the seizure

disorder ranged from 6 months to 19 years. Forty-eight percent of the children were

prescribed one medication, 39% were prescribed two, and 13% were prescribed three.

Phenobarbital was prescribed for most of the patients. The assessment of compliance, by

looking at blood levels, revealed that 57% of the parents gave their children their

medication as prescribed, 25% were giving less than the prescribed dose, and 18% were

giving no medication. Forty-three percent of the patients were not receiving their

medication as prescribed.

Results indicated that parents who reported at least one occurrence of running out

of their child's medication for several days were more likely to have children whose

blood levels revealed a discrepancy between the obtained and prescribed dosage (p < .05)

and parents who reported that they missed giving their child a medication more than once

a week were less likely to be compliant, as measured by blood levels (p < .001)

(motivational factors of the HBM). Recency of seizures was significantly positively

correlated with compliance (p < .05) (value of illness threat reduction of the HBM).

Child's enrollment in school was significantly and negatively correlated with compliance








(p < .01). Parents who reported that their child did not have any additional medical

problems were more compliant (p < .05).

In summary, factors that appear to influence parental non-adherence to anti-

epileptic medications for children with epilepsy include lower socioeconomic status and

ethnic minority group membership. However, the criteria for assessing socioeconomic

status have not been well defined nor have they been applied consistently across studies,

thus rendering comparability of socioeconomic findings somewhat suspect. Research

findings indicate that parents whose children had increased seizure frequency, more

recent seizure activity, and tonic-clonic seizures were more adherent to their child's anti-

epileptic medication regimen. Parents who perceived an increased level of disease

severity (more severe seizure events with greater frequency and intensity) were more

adherent to their child's medication regimen. The current study examined factors that are

related to medication adherence in pediatric epilepsy populations, specifically with

younger children with epilepsy.

Maternal Emotional Functioning

Depression is a highly prevalent mental disorder, especially among those women of

childbearing age. Downey and Coyne (1990) found that there is a high rate of serious

psychological problems in children of depressed mothers and that depressed mothers

have a host of difficulties as parents. Murray, Sinclair, Cooper, Ducoumau, Turner, and

Stein (1999) reported that exposure to maternal depression in the early postpartum

months may have an enduring influence on childhood psychological adjustment.

Cummings and Davies (1994) indicated that children of depressed parents are at

increased risk for the development of psychopathology and that they are two to five times

more likely to develop behavior problems than are children of normal parents. Depressed








mothers view their parenting role less positively than non-depressed mothers, experience

negative feelings towards parenting roles, and have feelings of rejection and hostility

towards their child (Downey & Coyne, 1990). Depressed mothers also tend to

demonstrate difficulties interacting with their children and displaying positive emotions

towards their children.

Hodes and Garralda (1999) examined the effects of maternal expressed emotion on

the psychological adjustment of children with epilepsy. Maternal expressed emotion and

risk for psychiatric disorders (for children and mothers) were assessed. The researchers

found that mothers showed significantly more emotional overinvolvement and a trend for

more hostility towards their children with epilepsy than they displayed towards sibling

controls. Maternal emotional overinvolvement was unrelated to child behavior problems.

High levels of criticism and hostility also were shown to covary with child behavior

problems (Hodes & Garralda, 1999). The effects of maternal expressed emotion on

adherence to medication were not assessed. However, the results suggest that maternal

expressed emotion is related to negative psychological functioning in children with

epilepsy. Thus, it seems highly plausible that maternal expressed emotion may have a

substantial impact on adherence to anti-epileptic medications for children with epilepsy.

Otero and Hodes (2000) assessed the relationship between maternal expressed

emotion and adherence to treatment in children with epilepsy. The researchers reported

that good treatment compliance was related to less maternal hostility and criticism for

epileptic children and their mothers. Children and mothers in the good compliance group

had fewer reported psychiatric symptoms. Poor treatment compliance for children with

epilepsy was associated with greater maternal psychological disturbance. Children with a








greater number of behavioral problems were less compliant. A relationship between

maternal depression and poorer compliance was reported (Otero & Hodes, 2000).

The relationship between maternal depression and behavior problems has been

investigated intensively across a variety of pediatric medical populations. Poor maternal

emotional functioning appears to exert a consistent negative impact on children with

disabilities and their healthy siblings. Research findings demonstrate a clear relationship

between maternal emotional functioning and childhood behavior problems.

Spector et al. (1999) found that increased maternal stress and anxiety levels were

related to an increase in seizure frequency and severity. Hazzard (1990) reported that

increased levels of parental anxiety and perceived stress were related to lower adherence

to anti-epileptic medications for children with a seizure disorder.

It is feasible that a true relationship exists between poor maternal adjustment and

difficulty with adherence to medication regimens for younger children. We examined the

relationship between maternal emotional distress variables (maternal depression and

maternal stress) and maternal adherence to anti-epileptic drugs (pill count percentage)

and child health status variables (seizure frequency, seizure-related emergency room

visits) for children diagnosed with epilepsy.













CHAPTER 2
AIMS AND STUDY JUSTIFICATIONS

Purpose

In summary, epilepsy is a chronic illness that typically begins in early childhood.

Poor adherence to the medication regimen can result in increased seizure frequency and

severity, and possibly, death. Young children are reliant on their parents, usually their

mothers, for their daily routine and functioning. This issue is particularly salient for

children with seizure disorders, who rely on their mothers to take their anti-convulsant

medicationss. It is well known that depression has a negative effect on daily

functioning. Additionally, depressive symptomatology and stress in mothers have been

associated with negative child health status and maternal adherence outcomes. Maternal

emotional factors such as depression and maternal stress may play an important role in

adherence to medication for younger children with epilepsy. To our knowledge, the

influence of maternal emotional distress on adherence to medication regimens of children

with epilepsy has not yet been examined.

An improved understanding of the impact of maternal emotional functioning on the

health status of children with epilepsy would enhance our ability to effectively intervene

with distressed mothers, and aid in developing interventions to prevent the possible

complications associated with poorer adherence, and improve overall psychological

funimctioning in both mother and child. In the current study, we examined the relationship

between maternal adherence to medication (pill count percentage) and child health status

(seizure frequency and seizure-related emergency room visits) for children with epilepsy








and the following factors: maternal emotional distress (maternal depression, maternal

stress), demographics (family annual income, race, maternal educational level),

condition-related and regimen-related variables (disease type, disease duration, type and

complexity of the medication regimen, concordance between prescribed regimen and

reported regimen), and maternal perceptions of the treatment regimen (perceived

efficacy, perceived knowledge, perceived difficulty).

Study Model

The purpose of the current study was to test a one-month predictive model of

maternal adherence and child health status to prescribed anti-epileptic medication in a

sample of 1-11 year old children diagnosed with epilepsy, focusing on the influence of

four predictor factors. The four predictor factors are defined as (1) maternal emotional

distress, (2) maternal perceptions of the treatment regimen, (3) condition-related and

regimen-related variables, and (4) demographic variables. The three outcome variables

are defined as (1) seizure frequency, (2) seizure-related emergency room visits, and (3)

pill count percentage, where (1) and (2) assess child health status and (3) assesses

maternal adherence. The current study extends earlier research in several ways. First, we

will integrate the four key sets of predictor factors identified in previous research in a

unifying predictive framework. Moreover, we strengthen the model by examining the

roles of the four, predictor factors at Time 1 (assessment) in predicting subsequent Time

2 (1-month follow-up) adherence and health status, controlling for Time 2 (1-month

follow-up) predictor factors (see Figure 1). This controls for the possibility that the

predictive role of adherence may be affected by maternal emotional distress, maternal

perceptions of the treatment regimen, condition-related and regimen-related variables,








and demographic variables at Time 2. The present study examined the following specific

hypotheses.

Hypotheses

Maternal Emotional Distress, Adherence, and Child Health Status

1. It was hypothesized that Time 1 maternal emotional distress would be negatively
related to Time 2 adherence to the child's medication regimen and positively with
child health status.
a. Time 1 maternal depression scores would be negatively associated with Time
2 adherence to the child's medication regimen and positively with health
status.
b. Time 1 maternal stress scores would be negatively associated with Time 2
adherence to the child's medication regimen and positively with health status.

Maternal Perceptions of the Treatment Regimen, Adherence, and Child Health
Status

2. We predicted that Time 1 ratings of maternal perceptions of the treatment regimen
(perceived knowledge, perceived difficulty, and perceived efficacy), and Time 1
maternal distress and Time 1 maternal perceptions of the treatment regimen would be
associated with Time 2 adherence to the child's medication regimen and child health
status.
a. Maternal ratings of perceived knowledge regarding the child's disease and
medication regimen at Time 1 would be positively associated with Time 2
adherence to the child's medication regimen and negatively with health status.
b. Maternal ratings of perceived difficulty in giving the child the prescribed
medication at Time 1 would be negatively associated with Time 2 adherence
to the child's medication regimen and positively with health status.
c. Maternal ratings of perceived efficacy regarding the child's medication
regimen at Time 1 would be positively associated with Time 2 adherence to
the child's medication regimen and negatively with health status.
d. Maternal ratings of maternal distress and maternal perceptions of the
treatment regimen at Time 1 would be associated with Time 2 adherence to
the child's medication regimen and health status.

Condition and Regimen-related Variables, Adherence, and Child Health Status

3. We predicted that Time 1 condition and regimen-related variables would be
associated with Time 2 adherence to the child's medication regimen and child health
status.
a. A tonic-clonic seizure diagnosis would be positively related to adherence to
the child's medication regimen and health status.
b. Time since diagnosis at Time 1 would be negatively associated with Time 2
adherence to the child's medication regimen and positively with health status.








c. Complexity of the medication regimen (higher doses, multiple medications,
more frequent doses) at Time 1 would be negatively associated with Time 2
adherence to the child's medication regimen and positively with health status.
d. Percentage of concordance between the physician-prescribed medication
regimen and self-reported regimen at Time 1 would be positively associated
with Time 2 adherence to the child's medication regimen and negatively with
health status.

Demographic Characteristics, Adherence, and Child Health Status

4. We predicted that Time 1 demographic characteristics would be positively related to
Time 2 adherence to the child's medication regimen and negatively with child health
status.
a. Total household income at Time 1 would be positively associated with Time 2
adherence to the child's medication regimen and negatively with health status.
b. Maternal educational level at Time 1 would be positively associated with
Time 2 adherence to the child's medication regimen and negatively with
health status.

















CHAPTER 3
METHODS


Study Design

The current study is a prospective study using a time lag design, in which we


examined the influence of predictor variables at Time 1 on outcome variables at Time 2,


while controlling for predictor variables at Time 2 (see Figure 1).


Time 1


Time 2


Depression

Maternal Stress




Efficacy
K now ledge
Difficulty


Com p lexity

Time Since Dx -
Regimen Type* -
Disease Type -


M aternal Perceptions
of the )
Treatment Regime n No s----.Seizure Frequency
---Health Sa ER Visits

---o--- Pill Counts
Adherence


Education Level
Annual Income
Race


Figure 1: Study Model








After obtaining informed consent, mothers were given a background questionnaire

(which included questions about condition and regimen-related variables), a

questionnaire assessing maternal stress, and a questionnaire assessing depressive

symptomatology to complete. The questionnaires were completed during a regularly

scheduled pediatric neurology clinic visit. Participants were asked to participate in a one-

month follow-up phone call. Approximately one month later, the investigator contacted

all participants at home by phone to re-assess condition and regimen specific-variables,

maternal perceptions of the treatment regimen, and maternal emotional distress. The one-

month delay in the assessment was designed to minimize sampling of temporary behavior

change that may have occurred immediately after a clinic visit and to re-assess maternal

ratings of emotional distress, condition and regimen-related measures, and maternal

perceptions regarding the treatment regimen, and their relationship to maternal adherence

to the child's medication regimen and child health outcomes.

Participants

Participants with a diagnosed seizure disorder were recruited from the Pediatric

Neurology clinic at Shands Hospital at the University of Florida's Health Science Center.

Based on data obtained from the pediatric neurologist, there are over 300 children with a

diagnosed seizure disorder between the ages of 1 and 11 who were eligible to participate

in the study. Young children are the focus of this study because: 1) epilepsy typically

begins to develop in early childhood, 2) mothers of younger children are more likely to

be highly involved in their child's medication regimen, and 3) one-third of epilepsies

spontaneously remit by adolescence.

Inclusionary criteria included: (1) An epilepsy diagnosis from the University of

Florida's pediatric neurologist which was noted in the patient's medical chart; (2) 1-11








years of age; (3) English-speaking; (4) Ability to read and write; (5) Telephone available

in the home. Exclusionary criteria included: (1) < 1 years of age or > 11 years of age; (2)

Non-English speaking; (3) Inability to read and write; (4) No telephone in the home.

A total of 82 participants completed all study measures. Data indicated the sample

consisted of 36 male and 46 female patients, aged 1-11 years (M = 5.76; SD = 2.93). The

mean education of mothers in the sample was 13.18 years (SD = 2.69). Mothers were

married (67.1%) and were homemakers (29.3%) or working full-time (35.4%). Average

family income was $44,450 (SD = 41,993). The ethnic makeup of the maternal sample

was 78.1% White, 11% African American, 6.1% Hispanic American, 2.4% Native

American, and 2.4% Other. The ethnic makeup of the child sample was 75.6% White,

9.8% African American, 7.3% Hispanic American, 2.4% Native American, and 4.9%

Biracial/Other. Approximately 93% of children in the sample were taking medication in

pill form, with seven percent taking medication in liquid form. Approximately 32% of

children in the study were diagnosed with partial seizures, 66% with generalized seizures,

and 2% with a diagnosis of other type. Study sample demographic characteristics are

presented in Table 1.

Table 1

Demographic and Medical Characteristics

Mean age of children in years (SD) 5.76 (2.93)

Sex of the child
Boys 36 (43.9%)
Girls 46(56.1%)

Mean education of mothers in years (SD) 13.18 (2.69)








Table 1 Continued


Mother Marital Status
Single/ Never Married
Married
Separated/Divorced

Mother Employment Status
Unemployed
Volunteer
Part-time
Full-time
Homemaker
Student
Other


Mean annual family income in dollars (SD)


Child race
White
African American
Hispanic American
Native American
Biracial/Other

Mom race
White
African American
Hispanic American
Native American
Biracial/Other


44,450 (41,993)


62 (75.6%)
8 (9.8%)
6 (7.3%)
5 (2.4%)
3 (4.9%)


64(78.1%)
9(11%)
5 (6.1%)
2 (2.4%)
2 (2.4%)


76 (92.7%)
6 (7.3%)


Regimen Type
Pills
Liquid


Seizure Type
Partial
Generalized
Other
Note. (n = 82).


26(31.7%)
54.9 (65.9%)
2 (2.4%)


Participants were recruited from the University of Florida's Pediatric Neurology

Outpatient Clinic at Shands Hospital during the patient's usual clinic visit. This clinic


15(18.3%)
55(67.1%)
12(14.6%)


13(15.9%)
1(1.2%)
8 (9.8%)
29 (35.4%)
24 (29.3%)
5 (6.0%)
2 (2.4%)








treats over 300 children a year diagnosed with epilepsy. We offered the study to all

patients who met the study's eligibility criteria.

The researcher and/or research assistants approached participants individually and

presented the rationale for the study. Parental informed consent and child assent was

obtained for children (if possible). Questions concerning subject obligation and

confidentiality were addressed at this time. After obtaining informed consent/assent,

participants completed a brief background questionnaire including questions regarding

maternal perceptions of the treatment regimen, the CES-D, and the PSI-SF. The entire

process took approximately 30-40 minutes.

The investigator recorded the following information from the participant's clinic

visit or from the medical chart review conducted to gather other pertinent information:

disease-specific information including date of diagnosis, time since diagnosis, type of

diagnosis, type of medication, dose of medication, seizure frequency, blood serum levels,

pill counts, and emergency room visits. Approximately one month after the clinic visit,

the mother was contacted by phone to re-assess condition-related information, maternal

perceptions regarding the treatment regimen, and maternal emotional distress.

Measures

Participant Background Information

A background questionnaire was completed at the time of initial assessment in

order to gather the following variables: age of child, age of mother, disease type, age of

onset, time since initial diagnosis, race, mother's level of education, family income, and

marital status. Time since initial diagnosis was divided into two categories: (1) those

who were newly diagnosed, and (2) those with a diagnosis of epilepsy greater or equal to

one year prior to the study. Mothers were asked to describe their child's seizure








diagnosis, seizure frequency, and medication regimen. The type(s) of seizure disorders)

the mothers reported were noted. Maternal report of the child's monthly seizure

frequency for the past month was also assessed. Furthermore, complexity of the

medication regimen reported by mothers was evaluated based on 4 indicators (i.e., type of

medication, number of medications per day, dosing, and timing). A complexity rating for

the medication regimen was calculated by multiplying the number of anti-epileptic

medications by the total number of times per day each medication is prescribed.

Maternal ratings of perceived difficulty (0 not difficult to 10 extremely difficult) and

perceived efficacy (0 not good to 10 extremely good) in carrying out the child's

medication regimen were assessed. Perceived difficulty ratings included items about

maternal perceptions of difficulty of medication type, medication dosage, and medication

timings. Perceived efficacy ratings included items about maternal perceptions of efficacy

of the child's medications and efficacy of maternal management of the child's medication

regimen. Mothers were asked to rate how they perceived their level of knowledge about

their child's medical condition. The knowledge ratings also ranged from 0 to 10 (0 not

knowledgeable to 10 extremely knowledgeable). Knowledge ratings included items

about maternal perceptions of knowledge of the child's seizure disorder and medications

(type, timing, number, and dosage) (see Appendix A & Appendix B).

Medical Chart Information

Condition-specific information including date of diagnosis, time since diagnosis,

type of diagnosis, type of medication, dose of medication, seizure frequency, blood

levels, pill counts, and emergency room (ER) visits were collected by a review of the

medical chart. The child's medication regimen was noted from the chart (type of

medication, number of medications per day, dosage, and timing) and compared to the








mothers' report of the child's medication regimen. Concordance ratings were calculated

between physician prescribed anti-epileptic medications) and maternal reports of

prescribed medications) by determining the percentage correct of maternal reports

compared to physician reports (see Appendix C & Appendix D). Specifically, points for

concordance were assigned for a correct match between the prescribed and reported

regimen as follows: 50 (name), 30 (dose), and 20 (timing). More weight was given to the

name of the medication and dosage of the medication, with the rationale that if the name

and dosage of the medication is incorrect, the mother might not be appropriately

administering the correct medication and/or dosage.

Child Health Status and Adherence

Child health status included two components: current seizure frequency and

seizure-related ER visits. Current monthly seizure frequency was assessed by maternal

self-report and by chart review over the last month. Seizure-related ER visits were

identified by conducting a chart review. Chart reviews consisted of a systematic

examination of the child's medical chart for the most recent data regarding child health

status. Written prescriptions and medication refill information were not noted

systematically in each medical chart and thus, this information was not available to be

collected.

Pill counts were assessed during the clinic visit if medication bottles were brought

to the clinic appointment. If mothers did not have the child's medication bottle at the

clinic visit, they were contacted at home on the same day of the appointment in order to

report a current pill count. Mothers were telephoned at one-month follow-up and asked

to report a pill count. Maternal adherence to the child's medical regimen was assessed by

percentage of pills used after one month [(Time 1-Time 2) / (Time 1)] x 100%.








Adherence was defined as an 80% pill count percentage. Blood serum levels were not

included as a measure of adherence because these values were not comparable across

different medication types; additionally, many participants had no blood serum levels

drawn.

Maternal Emotional Distress

The Center for Epidemiologic Studies Depression (CES-D; Radloff, 1977) is a

self-report measure developed to be used in epidemiological studies of depression and

was designed to measure current levels of symptoms which accompany depression in a

general adult population. The CES-D is a 20-item self-report measure of depressive

symptoms. Each item provides a statement representing a symptom characteristic of

depression (e.g. "I had crying spells"), followed by a 4-point Likert-type response scale

ranging from "rarely or none of the time" (less than 1 day) to "most all of the time" (5-7

days). Participants were instructed to bubble in the number of each statement which best

described how often they felt or behaved that way during the past week. Sixteen of the

scales range from 1 to 4 and the remaining four scales range from 4 to 1.

Four major factors have been derived based on a principal components factor

analysis of data from the general population. These factors have been described as

depressed affect, positive affect, somatic and retarded activity, and an interpersonal

factor. Total scores were used for the purpose of the current study. The CES-D is scored

by summing the ratings of the 20 items and scores range from 0 to 60. Total severity is

calculated by reversing scores for items 4, 8, 12, and 16 (the items that control for

response bias). Higher scores indicate higher levels of depressive symptoms experienced

during the past week. Investigators have designated a cutoff score of 16 as a suitable

indicator to determine depressed from nondepressed patients (Nezu, 2000).








Regarding internal consistency, Cronbach's alpha is high across a variety of

populations and ranges from 0.85 for the general population and 0.90 for the clinical

population. Split-half reliability is also high, ranging from 0.77 to 0.92. Test-retest

reliability studies ranging over 2-8 weeks show moderate correlations (r = 0.51-0.67).

Studies examining effects between African Americans, Caucasians, and Mexican

Americans found no differences in measures of internal consistency reliability.

In samples of outpatients with depression, alcoholism, drug addiction, or

schizophrenia, correlation coefficients between CES-D scores and Symptom Checklist-90

Depression Subscale Scores were high, ranging from 0.73-0.89. Correlations with the

Hamilton Rating Scale for Depression and Raskin Scale ranged from 0.45 and 0.28 for

patients with acute depression to 0.85 and 0.79 for patients with schizophrenia,

respectively. Scores on the CES-D have been shown to decrease significantly following

treatment for depression. For major depression, sensitivity was 79.5% and specificity

was 71.1%. The positive predictive value was 27.9%. (American Psychiatric

Association, 2000).

The Parenting Stress Index Short Form (PSI-SF; Abidin, 1995) is a 36-item short

form directly derived from the long form of the PSI. The PSI-SF was developed for use

in a variety of primary health care settings where it could be administered in less than 10

minutes. Participants were asked to respond to questions based on the five-point-scale.

The PSI-SF parent self-report scale contains three factor-analytically-derived

subscales (Parental Distress, Parent-Child Dysfunctional Interaction, and Difficult Child)

and a Total Stress Score. The Parental Distress (PD) scale determines the distress a

parent is experiencing in their role as a parent as a function of personal factors that are








directly related to parenting. The Parent-Child Dysfunctional Interaction (P-CDI)

subscale focuses on the parent's perception that their child does not meet their

expectations, and the interactions with their child are not reinforcing to them as a parent.

The Difficult Child (DC) subscale focuses on some of the basic behavioral characteristics

of children that make them either easy or difficult to manage. These characteristics can

include defiant, noncompliant, and demanding behavior. The Total Stress score is

designed to provide an indication of the overall level of parenting stress an individual is

experiencing. Parents who obtain a Total Stress score above a raw score of 90 are

experiencing clinically significant levels of parenting stress.

The three PSI-SF subscales differ from that of the longer PSI, but correlate in the

expected direction with the domains of the longer form. The PSI and PSI-SF total scores

are highly correlated with one another (0.94). The short form subscales have shown

Cronbach's alphas of 0.80 to 0.91 and 6-month test-retest reliabilities of 0.68 to 0.85.

Positive predictive values of the PSI-SF range from 25% to 45%.













CHAPTER 4
RESULTS

The overall findings of the data analyses are presented as follows: 1) descriptive

statistics of the predictor and criterion variables included in the conceptual model, 2)

internal consistency reliabilities of the questionnaire measures and the distributional

properties of key predictor and criterion variables, 3) correlations between the

background variables not included in the conceptual model and those between key

predictor variables and change scores on the criterion variables, 4) regression analyses

that control for the effects of Time 2 predictors on Time 2 criterion variables, and 5)

exploratory model testing to identify goodness of fit of the specific regression models.

Due to small sample size, structural equation modeling (SEM) was not conducted

as anticipated. The statistical approach shifted from SEM to variable-level model testing,

in which the goodness of fit of variables located under the four primary factors of the

conceptual framework was assessed (see Figure 1).

Preliminary Analyses

Medical characteristics of the child with epilepsy were variables collected at Time

1, except for seizure frequency. The latter was also collected at Time 2, since this time

point was of main interest in our primary analyses. Study sample medical characteristics

are displayed in Table 2.








Table 2

Mean Scores and Standard Deviations for Child Medical Characteristics


Measures M (SD)

Complexity of the regimen (uncorrected) 8.7(13.1)
Complexity of the regimen (corrected) 1.13 (0.5)
Time Since Diagnosis in years 4.0(2.8)
Concordance between reported and prescribed regimen-Time 2 72.8 (32.7)
Seizure Frequency-Time 2 (uncorrected) 50.0 (87.2)
Seizure Frequency-Time 2 (corrected) 2.5 (1.1)
Note. (n = 82). Corrected scores are corrected for skewness to ensure analyses integrity.

Means and standard deviations for maternal self-reports of maternal stress

(measured by the PSI-SF), depression (measured by the CES-D), and maternal

perceptions of the treatment regimen (measured by maternal reports of difficulty,

efficacy, and knowledge) for Time 1 and Time 2 are presented in Table 3 below.

Table 3

Mean Scores and Standard Deviations for Study Measures at Time 1 and Time 2

Measures M(SD) M (SD)
Time I Time 2
PSI-SF, Total score 85.2 (28.5) 65.5 (8.9)
PSI-SF, Parental distress 27.5(10.5) 24.5(3.5)
PSI-SF, Parent-child dysfunction 25.9(9.9) 21.6(5.1)
PSI-SF, Difficult child 31.8(11.5) 19.4 (3.7)
CES-D, Severity score 7.8(3.3) 12.4(2.2)
Difficulty, Total score uncorrected 6.0 (5.8) 12.9 (4.5)
Difficulty, corrected 2.1 (1.5) 1.5 (0.8)
Efficacy, Total score 14.3 (4.5) 11.0 (3.5)
Knowledge, Total score 39.4 (13.84) 39.5 (13.82)
Note. PSI-SF = Parenting Stress Index-Short Form; CES-D = Center for Epidemiologic
Studies-Depression. Corrected scores are corrected for skewness to ensure analyses
integrity.

Cronbach's alphas were calculated to assess the reliability (i.e. internal consistency)

of mother-completed self-report measures. Coefficients ranged from .68 to .96 for the

total sample at Time 1, suggesting adequate internal consistency reliability for all








measures used. Coefficients ranged from -.002 to .58 for the total sample at Time 2,

suggesting lower levels of internal consistency reliability for all measures used compared

to Time 1. Time 1 and Time 2 reliability statistics are presented in Table 4 below.

Table 4

Alpha Coefficients of Parent-completed Measures for Total Sample at Time 1 and Time 2

Measure N a a
Time 1 Time 2

PSI-SF, Total score 82 .96 .56
CES-D, Total score 82 .94 .35
Difficulty, Total score 82 .86 -.18
Efficacy, Total score 82 .68 -.002
Knowledge, Total score 82 .92 .08
Note. PSI-SF = Parenting Stress Index-Short Form; CES-D = Center for Epidemiologic
Studies-Depression.

We also analyzed the distribution properties of demographic, medical, predictor,

and outcome variables. Skewed distributions were corrected with the creation of new

variables. Annual family income was highly positively skewed and ranged from $1500

to $250,000 and was recorded into three income categories with a value of one

representing poverty level of $1500 to $18,000, a value of two representing middle level

of $18,001 to $116,000, and a value of three representing an upper middle level of

$116,001 to $250,000. The poverty level cutoff was determined using the United States

Census Bureau poverty guidelines (United States Census Bureau, 2001). The means and

standard deviations for uncorrected and corrected annual income were $44,450 (41,993)

and 1.2 (.62), respectively. The original scores for complexity of the regimen, seizure

frequency, and maternal perceptions of regimen difficulty were also highly positively

skewed. In order to correct for data skewness, the data was collapsed into groups that

were equally distributed, creating new complexity, seizure frequency, and difficulty








variables. Original complexity scores ranged from 1 to 80 and were recorded into three

categories, with one representing a complexity score of 1 to 26.7, two representing a

complexity score of 26.8 to 53.4, and three representing a complexity score of 53.5 to 80.

Means and standard deviations for uncorrected and corrected complexity scores were 8.7

(13.1) and 1.13 (0.5), respectively. Original seizure frequency scores ranged from 0 to

320 and were recorded into four categories with equal distributions of the sample in each

group, with one representing 0 seizures, two representing 1 to 7 seizures, three

representing 8 to 60 seizures, and four representing 61 to 320 seizures. The means and

standard deviations for uncorrected and corrected seizure frequency scores were 50.0

(87.2) and 2.5 (1.1), respectively. Difficulty scores were positively skewed, ranged from

0 to 24, and were recorded into four categories, with one representing a score of 0 to 6,

two representing a score of 6.1 to 12, three representing a score of 12.1 to 18, and four

representing a score of 18.1 to 24. The means and standard deviations for uncorrected

and corrected difficulty scores were 6.0 (5.8) and 2.1 (1.5) for Time 1 and 12.9 (4.5) and

1.5 (0.8) for Time 2, respectively.

Primary Analyses

Correlations among Background Variables, Predictors, and Criterion Variables

Correlational analyses were conducted between background variables not included

in the model (child sex, marital status, number of years married, employment status,

number of children in the home) and Time 1 key predictor variables of the following

factors: maternal emotional distress, maternal perceptions of the regimen, condition and

regimen-related variables, and demographic variables. No significant associations were

found between the background variables not included in the model and Time 1 predictors








(p > .05). No significant associations were found between the background variables not

included in the model and criterion variable change scores (see Table 5 below).

Table 5

Correlations between Background Variables not in the Model and Specific Time 2
Outcomes Change Scores

Background Factors Seizure Seizure-related ER
Frequency Visits
Child sex .160' -.006+
Marital status -.027+ .083+
Years married .085+ .060+
Employment status -.098+ -.072+
Number of children in home .062+ -.015+
Note. +p > .05.

Correlational and hierarchical regression analyses were conducted to examine the

relationships between Time 1 predictors (maternal emotional distress, maternal

perceptions of the treatment regimen, condition and regimen-related factors, and

demographic factors), and Time 2 outcome measures (seizure frequency, emergency

room visits, and pill count percentages). Note that the effects of Time 2 predictor

measures were partialed out in regression analyses to control for Time 2 predictor

variable effects on the criterion measures. The order of entry into the regression analyses

were as follows: (1) Time 2 predictors were entered in a block in a forced-entry fashion,

(2) Time 1 predictors were entered in blocks in a stepwise fashion. Note that Time 2

predictor measures were not included in correlational analyses due to low internal

consistency reliability. New variables that were created as corrections for data skewness

were used in all subsequent analyses; analyses conducted with skewed data did not reveal

any significant relationships. Partial correlation and regressions analyses were used in

determining which predictors to enter into the final hierarchical regression models.








Hypothesis 1: Maternal Emotional Distress, Adherence, and Child Health Status

We hypothesized that Time 1 maternal emotional distress (measured by the PSI-SF

and CES-D) would be negatively related to Time 2 adherence to the child's medication

regimen (pill count percentage) and positively associated with child health status (seizure

frequency and seizure-related emergency room visits). Specifically, it was predicted that

Time 1 maternal stress and depression scores would be negatively associated with Time 2

adherence to the child's medication regimen and that Time 1 ratings of maternal stress

and depressive affect would be positively associated with the child's health status. These

hypotheses were partially supported. Hierarchical regression analyses were conducted to

assess the association between Time 2 outcomes and Time 1 maternal emotional distress

variables, while controlling for the effects of Time 2 maternal emotional distress

variables. Time 1 PSI-SF overall mean scores and PSI-SF parent-child dysfunction

subscale scores significantly predicted Time 2 seizure frequency, while controlling for

Time 2 predictor variables. No substantial associations were found between Time 1

ratings of maternal stress and Time 2 seizure-related emergency room visits or Time 2

maternal adherence. No significant relationships were found between Time 1 ratings of

maternal depression and child health status or maternal adherence. Next, Time 1

maternal emotional distress variables did not predict Time 2 seizure-related emergency

room visits or pill count percentages, while controlling for Time 2 maternal emotional

distress. The results of the two sets of regression analyses for Time 1 maternal emotional

distress variables on Time 2 outcomes are presented in Tables 6 and 7.








Table 6

Summary of Hierarchical Regression Analyses for Time 1 PSI-SF Overall Score
Predicting Child Seizure Frequency (N = 82)


Variables Included P3 t R R2 Adj.R2 AR2 F AF
Controlling for T2 .091 .008 -.017 .008 .332 .332
CES-D mean T2 .055 .466
PSI-SF mean T2 -.093 -.784

Model 1 .253 .064 .028 .056 1.78 4.64*
CES-D mean T2 .090 .773
PSI-SF mean T2 -.121 -1.04
PSI-SF mean T1 .239 2.16*
Note. CES-D = Center for Epidemiologic Studies-Depression; PSI-SF = Parenting
Stress Index-Short Form; T1 = Time 1; T2 = Time 2.
*E < .05

Table 7

Summary of Hierarchical Regression Analyses for Time 1 PSI-SF Subscale Variables
Predicting Child Seizure Frequency (N = 82)


Variables Included t R R2 Adj. R2 AR2 F AF
Controlling for T2 .084 .007 -.031 .007 .183 .183
PD mean T2 -.048 .466
P-C dys T2 -.007 -.784
DC T2 -.059

Model 1 .278 .077 .029 .070 1.62 5.88*
PD mean T2 -.060 -.535
P-C dys T2 -.015 -.124
DC T2 -.078 -.629
P-C dys T1 .267 2.42*
Note. PSI-SF = Parenting Stress Index-Short Form; PD = Parental Distress; DC =
Difficult Child; P-C dys. = Parent-Child Dysfunction; T1 = Time 1; T2 = Time 2.
*p < .05

The data revealed significant relationships (partial correlations) between Time 1

PSI-SF overall mean scores and Time 2 seizure frequency (r = .24, p < .05), as well as

between Time 1 PSI-SF parent-child dysfunction subscale scores and Time 2 seizure








frequency (r = .27, p < .05). As hypothesized, overall Time 1 maternal stress levels were

positively related to Time 2 child seizure. Further analysis of the PSI-SF subscales

indicated that Time 1 PSI-SF parent-child dysfunction was positively related to Time 2

child seizure frequency. As maternal stress levels and parent-child dysfunction levels

increased, seizure frequency concomitantly increased. Contrary to prediction, no other

significant correlational relationships were found with Time 1 maternal emotional

distress and Time 2 outcomes. Correlational data between Time 1 maternal emotional

distress, Time 2 maternal adherence, and Time 2 child health status are presented in

Table 8 below.

Table 8

Partial Correlations between Time 1 Maternal Emotional Distress and Time 2 Outcomes

Health Status Adherence

Maternal Emotional Distress Seizure Seizure-related ER Pill Count
Frequency Visits Percentage
PSI-SF Total .237* .061 -.104
PSI-SF, Parental Distress .014 -.026 -.045
PSI-SF, Parent-child dysfunction .266* .056 -.165
PSI-SF, Difficult child -.096 .025 -.078
CES-D Total .045 .101 -.099
Note. PSI-SF = Parenting Stress Index; CES-D = Center for Epidemiologic Studies-
Depression.
*p < .05.

Hypothesis 2: Maternal Perceptions of the Treatment Regimen, Adherence and
Child Health Status

We predicted that Time 1 ratings of maternal perceptions of the treatment regimen

(measured by perceived knowledge, perceived difficulty, and perceived efficacy), and

Time 1 maternal distress (measured by the PSI-SF) and Time 1 maternal perceptions of

the treatment regimen would be associated with Time 2 adherence to the child's








medication regimen (pill count percentage) and child health status (seizure frequency and

seizure-related emergency room visits). Specifically, maternal ratings of perceived

knowledge and efficacy regarding their child's medical condition and medication

regimen at Time 1 would be positively associated with Time 2 adherence to the child's

medication regimen and negatively with health status. In addition, maternal ratings of

perceived difficulty in giving the child the prescribed medication at Time 1 would be

negatively associated with Time 2 adherence to the child's medication regimen and

positively associated with child health status, and last, maternal ratings of maternal

distress and maternal perceptions of the treatment regimen at Time 1 would be associated

with Time 2 adherence to the child's medication regimen and health status. These

hypotheses received mixed support.

Time 2 outcomes of child health status and maternal adherence were regressed on

Time 1 maternal perceptions of the treatment regimen variables, while controlling for the

effects of Time 2 maternal perceptions of the treatment regimen predictor variables.

Time 1 difficulty and efficacy scores successfully predicted Time 2 seizure frequency,

while controlling for Time 2 difficulty and efficacy predictor variables. Time 1 difficulty

scores also successfully predicted Time 2 seizure-related emergency room visits, while

controlling for Time 2 difficulty predictor variables. No other significant relationships

were found between Time 1 ratings of difficulty, efficacy, and knowledge and Time 2

outcomes of child health status and maternal adherence. Next, Time 1 maternal

perceptions of the treatment regimen variables were not successful in predicting Time 2

pill count percentages, while controlling for Time 2 maternal perceptions of the treatment

regimen. The findings of the two sets of regression analyses for Time 1 maternal








perceptions of the treatment regimen variables on Time 2 outcomes are presented in

Tables 9 and 10.

Table 9

Summary of Hierarchical Regression Analyses for Time 1 Maternal Perceptions of the
Treatment Regimen Predicting Child Seizure Frequency (N = 82)


Variables Included 3 t R R2 Adj.R2 AR2 F AF
Controlling for T2 .222 .049 .013 .049 1.35 1.35
Difficulty T2 -.025 -.229
Knowledge T2 .226 2.01*
Efficacy T2 -.046 -.406

Model 1 .334 .111 .065 .062 2.41 5.38*
Difficulty T2 -.034 -.312
Knowledge T2 .212 1.93
Efficacy T2 -.022 -.196
Difficulty T1 .250 2.32*

.459 .211 .159 .100 4.07 9.62**
Model 2
Difficulty T2 -.023 -.226
Knowledge T2 .406 3.35**
Efficacy T2 -.043 -.411
Difficulty T1 .172 1.63
Efficacy T1 -.376 -3.10**
Note. T1 = Time 1; T2 = Time 2.
*P <.05 **p<.l01

Table 10

Summary of Hierarchical Regression Analyses for Time 1 Maternal Perceptions of the
Treatment Regimen Predicting Seizure-related Emergency Room Visits (N = 82)


Variables Included 03 t R R2 Adj. R2 AR2 F AF
Controlling for T2 .227 .052 .015 .052 1.41 1.41
Difficulty T2 .060 .541
Knowledge T2 .211 1.87
Efficacy T2 -.110 -.972








Table 10- Continued

Model 1 .365 .134 .089 .082 2.97 7.29**
Difficulty T2 .051 .472
Knowledge T2 .195 1.80
Efficacy T2 -.082 -.750
Difficulty T1 .288 2.70**

Model 2 .377 .142 .085 .008 2.51 .729
Difficulty T2 .054 .449
Knowledge T2 .251 1.98
Efficacy T2 -.088 -.804
Difficulty T1 .265 2.41*
Efficacy T1 -.108 -.854
Note. T 1 = Time 1; T2 = Time 2.
*<.05 ** <.01

The data revealed significant relationships (partial correlations) between Time 1

difficulty scores and Time 2 seizure frequency (r = .26, p < .05), Time 1 efficacy scores

and Time 2 seizure frequency (r = -.34, p < .01), and Time 1 difficulty scores and Time 2

seizure-related emergency room visits (r = .29, p < .01). As hypothesized, Time 1

maternal perceptions of regimen difficulty were positively related to Time 2 child seizure

frequency and emergency room visits. As maternal perceptions of regimen difficulty

significantly increased, seizure frequency and seizure-related emergency room visits also

significantly increased. Time 1 maternal perceptions of regimen efficacy were negatively

related to Time 2 seizure frequency in that as efficacy significantly increased, seizure

frequency significantly decreased. Contrary to study hypotheses, no other significant

correlational relationships were found with Time 1 maternal perceptions of the treatment

regimen and Time 2 outcomes. Correlational data between Time 1 maternal perceptions

of the treatment regimen, Time 2 maternal adherence, and Time 2 child health status are

presented in Table 11.








Table 11

Partial Correlations between Time I Maternal Perceptions of the Treatment Regimen
and Time 2 Outcomes

Health Status Adherence

Maternal Perceptions of the Seizure Seizure-related ER Pill Count
Treatment Regimen Frequency Visits Percentage

Difficulty .255* .294** .037
Efficacy -.335** -.097 -.024
Knowledge .215 .194 .074
Note. *p <.05 **p <.01

Hypothesis 3: Condition and Regimen-related Variables, Adherence, and Child
Health Status

We predicted that condition and regimen-related variables (complexity, time since

diagnosis, regimen type, disease type, and concordance between the prescribed and

reported regimen) would be associated with Time 2 adherence to the child's medication

regimen (pill count percentage) and child health status (seizure frequency and seizure-

related emergency room visits. First, we hypothesized that a tonic-clonic seizure

diagnosis would be positively related to adherence to the child's medication regimen.

Second, we predicted that child health status and time since diagnosis and complexity of

the medication regimen (higher doses, multiple medications, more frequent doses) would

be negatively associated withTime 2 adherence to the child's medication regimen and

positively with health status. Third, we predicted that percentage of concordance

between the physician-prescribed medication regimen and self-reported regimen would

be positively associated with Time 2 adherence to the child's medication regimen and

negatively with health status. These hypotheses were partially confirmed. Hierarchical

regression analyses were conducted to predict Time 2 outcomes with condition and








regimen-related variables. Concordance scores successfully predicted Time 2 seizure-

related emergency room visits and Time 2 pill count percentage. No other significant

predictive relationships were obtained. Furthermore, condition and regimen-related

factors were not successful in predicting Time 2 seizure frequency. Regression data for

condition and regimen-related variables on Time 2 outcomes are presented in Tables 12

and 13.

Table 12

Summary of Hierarchical Regression Analyses for Condition and Regimen-related
Variables Predicting Seizure-related Emergency Room Visits (N = 82)


Variables Included 13 t R R2 Adj. R2 AR2 F AF
Model 1 .076 .006 -.020 .006 .222 .222
Seizure type -.074 -.653
Regimen type .019 .163

Model 2 .408 .166 .110 .161 2.96 4.76**
Seizure type .015 .140
Regimen type -.035 -.324
Concordance -.219 -1.97*
Complexity .210 1.80
Time since dx. .177 1.57
Note. dx = diagnosis.
* <.05 **'1<.01

Table 13

Summary of Hierarchical Regression Analyses for Condition and Regimen-related
Variables Predicting Pill Count Percentage (N = 82)


Variables Included 13 t R R2 Adj. R2 AR2 F AF

Model 1 .122 .015 -.011 .015 .586 .586
Seizure type -.074 -.653
Regimen type .019 .163








Table 13 Continued

Model 2 .755 .570 .541 .555 19.66 31.9**
Seizure type .015 .140
Regimen type -.035 -.324
Concordance -.219 -1.97*
Complexity .210 1.80
Time since dx. .177 1.57
Note. dx = diagnosis.
*E2<.05 **E<.01

The data revealed significant relationships (partial correlations) between

concordance scores and Time 2 seizure-related emergency room visits (r = -.22, p < .05),

as well as concordance scores and Time 2 pill count percentage (r = .69, p <.01). As

hypothesized, concordance scores were negatively related to Time 2 seizure-related

emergency room visits and positively related to Time 2 pill count percentage. As

concordance scores significantly increased, emergency room visits significantly

decreased and pill count percentages significantly increased. Contrary to study

hypotheses, a tonic-clonic seizure diagnosis was not associated with better adherence,

increased seizure frequency, or emergency room visits. No other significant relationships

were found with condition and regimen-related variables and Time 2 outcomes.

Correlational data between condition and regimen-related variables, Time 2 maternal

adherence, and Time 2 child health status are presented in Table 14.

Table 14

Partial Correlations between Time 1 Condition and Regimen-related Variables and Time
2 Outcomes

Health Status Adherence

Condition and Regimenb Seizure Seizure-related ER Pill Count
Variables Frequency Visits Percentage
Complexity .098 .205 -.108








Table 14 Continued

Time Since Diagnosisa .157 .180 -.076
Regimen Type" .089 -.038 .027
Disease Typea .114 -.016 -.083
Concordance"b -.160 -.223* .691**
Note. <.05 ** < .01

Hypothesis 4: Demographic Characteristics, Adherence, and Child Health Status

We predicted that demographic characteristics (total annual income, maternal

education level) would be positively related to Time 2 adherence to the child's medication

regimen (pill count percentage) and negatively associated with child health status (seizure

frequency and seizure-related emergency room visits). Specifically, total household

annual income and maternal education level would be positively associated with Time 2

adherence to the child's medication regimen and negatively associated with child health

status. These hypotheses were partially supported. Hierarchical regression analyses were

conducted to predict Time 2 outcomes with demographic variables. Family income

approached significance (i.e., trend in the predicted direction) for predicting Time 2 pill

count percentage. No other significant associations were found. Demographic variables

were not successful in predicting Time 2 seizure frequency or seizure-related emergency

room visits. Regression data for demographic variables on Time 2 outcomes is presented

in Table 15.








Table 15

Summary of Hierarchical Regression Analyses for Demographic Variables Predicting
Pill Count Percentage (N = 82)


Variables 13 t R R2 Adj. R2 AR2 F AF p
Included
Model 1 .343 .118 .064 .118 2.21 2.21 .078+
Family income .237 2.04 .058+
Mom education .132 1.13
Child race -.283 -1.17
Mom race .093 .388
Note. + > .05

The data revealed a significant relationship (partial correlation) between annual

family income and Time 2 pill count percentage (r = .24, p < .05). As hypothesized,

annual family income was positively related to Time 2 pill count percentage in that as

family income increased, pill count percentages increased. Contrary to study hypotheses,

no other significant correlational relationships were found between demographic

variables and Time 2 child health status or maternal adherence outcomes. Regression

analyses did not identify any additional variables that significantly predicted Time 2

outcomes. Correlational data between demographic variables, Time 2 maternal

adherence, and Time 2 child health status is presented in Table 16 below.

Table 16

Partial Correlations between Demographic Variables and Time 2 Outcomes

Health Status Adherence

Demographic Factors Seizure Seizure-related ER Pill Count
Frequency Visits Percentage
Total Annual Income .088 .089 .244*
Mother's Education -.057 -.101 .138
Child Race .037 -.130 -.143
Mother Race -.025 -.100 -.048
Note. *P < .05








Model Testing

Predictor variables for Time 2 outcomes were included based on the analyses

described above by selecting variables that showed both significant (3 weights and

accounted for significant R2 change. These variables were subsequently included in tests

of the overall goodness of fit to identify promising alternative regression models for use

in follow-up research (see Table 17). The statistical approach included two steps: (1)

Time 2 measures were entered in step one of the hierarchical regression analyses in a

forced-entry fashion and (2) Time 1 variables were entered in a stepwise fashion with

variables included in the model if r < .05.

Table 17

Identified Predictors of Time 2 Outcomes

Seizure Frequency Seizure-related ER Visits Pill Count Percentage
PSI-SF overall mean Difficulty Concordance
PSI-SF parent-child dysfunction Concordance Annual family income
Difficulty
Efficacy

Prediction of adherence

Hierarchical regression analyses were performed to test the hypotheses that the

variables of concordance and annual family income predicted Time 2 maternal adherence

(measured by pill count percentage). Concordance and annual family income were

entered in a stepwise manner with Time 2 pill count percentage as the outcome variable.

Only one model was generated, with the concordance variable retained as a significant

predictor. The concordance variable accounted for a significant proportion of the

variance in the Time 2 pill count percentage outcome measure (Adjusted R2 = .44; tp <

.001). The results suggested that the best predictor of maternal adherence was

concordance between mother and physician on the child's drug regimen. This measure








accounted for approximately 44% of the variance in the Time 2 pill count percentage

measure. In examining the significance of P3 weights for all predictor variables in the

model, only the concordance variable was significant (13 = .67; t = 7.4; p = .000) (see

Table 18).

Table 18

Summary of Regression Model for Predicting Maternal Adherence (N = 82)


Variables Included 13 t R R2 Adj. R2 AR2 F AF
Model l** .665 .443 .435 .443 54.8 54.8*
Concordance .665 7.4*
Note. *p < .01 **Best prediction model

Prediction of health status

Hierarchical regression analyses were performed to test the hypotheses that the

variables of maternal stress (measured by the PSI-SF), parent-child dysfunction

(measured by the subscale of the PSI-SF), and maternal perceptions of difficulty and

efficacy regarding the child's medication regimen successfully predicted Time 2 seizure

frequency. The PSI-SF overall score, parent-child dysfunction score, difficulty score,

and efficacy score for Time 2 were entered as a block in a forced-entry fashion to partial

out Time 2 effects on seizure frequency. The PSI-SF overall score, parent-child

dysfunction score, difficulty score, and efficacy score for Time 1 were then entered in a

stepwise fashion with Time 2 seizure frequency as the outcome variable. One model was

generated in which the Time 1 parent-child dysfunction subscale was a predictor. Parent-

child dysfunction accounted for a small but significant proportion of the variance in the

Time 2 seizure frequency outcome measure (Adjusted R2 = .022; p = .015). The results

suggested that the best predictor of Time 2 seizure frequency was parent-child








dysfunction, which accounted for approximately 2.2% of the variance in the Time 2

seizure frequency measure. In examining the significance of 13 weights for all predictor

variables in the model, only the parent-child dysfunction variable was significant (13P =

.29, t = 2.5, p = .015) (see Table 19).

Table 19

Summary of Regression Models for Predicting Seizure Frequency (N = 82)


Variables Included 3 t R R2 Adj. R2 AR2 F AF
Controlling for T2 .084 .007 -.045 .007 .135 .135
PSI-SFmeanT2 -.130 -.610
P-C dys. T2 .065 .306
Difficulty T2 -.001 -.010
Efficacy T2 -.006 -.049

Model 1** .287 .082 .022 .075 1.37 6.25*
PSI-SF mean T2 -.166 -.802
P-C dys. T2 .087 .421
Difficulty T2 -.057 -.497
Efficacy T2 .056 -.489
P-C dys. TI .287 2.50*
Note. PSI-SF = Parenting Stress Index; P-C dys. = Parent-Child Dysfunction; T1 =
Time 1; T2 = Time 2.
*" <.05 **Best prediction model

Hierarchical regression analyses were performed to test the hypotheses that the

variables of difficulty regarding the child's medication regimen and concordance between

the prescribed and reported regimen successfully predicted seizure-related emergency

room visits. The difficulty score for Time 2 was entered first to partial out Time 2 effects

on seizure-related emergency room visits. The difficulty score for Time 1 and

concordance were then entered as a block in a stepwise fashion with seizure-related

emergency room visits as the outcome variable. This resulted in the generation of two

models, the first of which included the difficulty variable as a predictor and the second








added concordance as a predictor. The first model accounted for a significant proportion

of the variance in the emergency room visits outcome measure (Adjusted R2 = .072; =

.006), with 7.2% of the variance accounted for by this block. The second model

accounted for a significant proportion of the variance in the emergency room visits

beyond model one (Adjusted R2 = .146; p = .007), with an additional 7.4% of the

variance accounted for by this model. The results suggested that the best model for

predicting seizure-related emergency room visits was Model 2, accounting for

approximately 15% of the variance in seizure-related emergency room visits. In

examining the significance of 13 weights for all factors in Model 2, two betas, difficulty (3

= .32, t = 3.1, p = .002) and concordance (13 = -.293, t = -2.8, p = .007), were statistically

significant (see Table 20).

Table 20

Summary of Regression Models for Predicting Seizure-related Emergency Room Visits (N
= 82)


Variables 13 t R R2 Adj.R2 AR2 F AF
Included
Controlling for T2 .062 .004 -.009 .004 .307 .307
Difficulty T2 .062 .554

Model 1 .308 .095 .072 .091 4.15 7.96*
Difficulty T2 .054 .505
Difficulty T1 .302 2.82*

Model 2** .421 .177 .146 .082 5.60 7.80*
Difficulty T2 -.003 -.026
Difficulty T1 .322 3.13*
Concordance -.293 -2.79*
Note. T1 = Time 1; T2 = Time 2.
*r < .01 **Best prediction model








Summary of Findings

Overall, we found mixed confirmation for the hypothesized relationships among

the predictor variables, health status, and adherence criterion variables.

First, the maternal stress measure of maternal emotional distress was found to be

significantly related to the child health status variable of seizure frequency. Maternal

stress and the subscale of parent-child dysfunction were significantly related to seizure

frequency and were significant predictors of seizure frequency in hypothesis testing.

Parent-child dysfunction was identified as a significant predictor of seizure frequency in

exploratory model testing. Note also that maternal emotional distress on measures of the

traditional psychological constructs of depression and maternal stress were not clinically

elevated in the study sample.

Second, the difficulty and efficacy variables of maternal perceptions of the

treatment regimen were identified as being significantly related to child health status

variables. Difficulty was significantly related to seizure frequency and seizure-related

emergency room visits. Difficulty was found to be a significant predictor of both seizure

frequency and seizure-related emergency room visits in hypothesis testing, while efficacy

was only a significant predictor of seizure frequency in hypothesis testing. Difficulty was

also identified as a significant predictor of seizure-related emergency room visits in

exploratory model testing.

Third, the concordance measure of condition and regimen-related variables were

identified as significantly related to the child health status variable of seizure-related

emergency room visits and maternal adherence. Concordance was found to be a

significant predictor of maternal adherence and seizure-related emergency room visits in

hypothesis testing, as well as in exploratory model testing.








Fourth, the demographic factor of annual family income was identified as being

related to maternal adherence. Annual family income was identified as a significant

predictor of maternal adherence in hypothesis testing. However, no significant predictive

relationships were identified between annual family income and outcomes in exploratory

model testing.

Finally, findings related to the predictive validity of Time 1 measures of maternal

emotional distress, maternal perceptions of the treatment regimen, condition and

regimen-related factors, and demographics on Time 2 child health status (seizure

frequency and seizure-related emergency room visits) and maternal adherence (pill count

percentage) outcomes suggest that several measures (PSI-SF parent-child dysfunction,

difficulty, concordance, and annual family income) play an important role in child health

status and maternal adherence outcomes.













CHAPTER 5
DISCUSSION

Sample Characteristics of Demographics and Maternal Distress

In regards to the study sample, mothers were highly educated, having attended an

average of one year of college, with their annual family income falling in the middle

income range ($18,001 to $116,000). The racial make-up of mothers completing this

study was predominately White (78.1%), with traditionally labeled minorities (e.g.,

African American, Hispanic, etc.) showing a fairly uniform distribution pattern. Two-

thirds of the mothers were married and nearly half worked outside the home at least part-

time. The racial profile of patients in this study is representative of the reported

distribution of children with epilepsy living in the United States (Adams, Hendershot, &

Marano, 1996). Patients were predominately White (75.6%), girls (56.1%), and between

five and six years of age. The racial heterogeneity observed in the current study suggests

that its findings are generalizable to the overall epilepsy population within the United

States.

Maternal stress and depression scores were below the cut-off for clinically

significant levels of 90 and 16, respectively, indicating that mothers were not reporting

heightened levels of maternal stress or depressive symptoms at Time 1 or Time 2. It was

surprising that the study sample did not meet criteria for significant levels of maternal

stress or depressive symptomatology since mothers in previous studies reported higher

levels of depressive symptomatology (Kyngas, 2000) and maternal stress (Spector et al.,

1999). One possible explanation for this finding may be that mothers of children with








epilepsy in this study may have experienced more stigma and emotional strain, as well as

daily pressures (e.g., frequent doctor's appointments) when the child was much younger

during the initial months following the child's diagnosis. Perhaps maternal stress and

depression scores were not significant because, compared to when the child was first

diagnosed, the current situation is a calmer and less stressful time for mothers. Studies

indicate that the initial phase following the child's diagnosis leads to negative emotional

consequences (e.g., increased levels of stress and depressive symptomatology) for

mothers of children with chronic illnesses (Brownbridge & Fielding, 1994; Austin, 1989).

It is also possible that only mothers of children who experienced low levels of stress or

depressive symptomatology opted to participate in the study.

Hypotheses

Hypothesis 1: Maternal Emotional Distress, Adherence, and Child Health Status

It was hypothesized that Time 1 maternal emotional distress variables would be

negatively related to Time 2 adherence to the child's medication regimen and positively

with child health status outcome measures. Overall, mixed support was obtained for the

study hypotheses. Hypothesis 1 was partially supported in that mothers who reported

high levels of maternal stress at Time 1, also reported higher seizure frequency levels for

their children at Time 2. No significant relationships were found between maternal

distress measures and seizure-related emergency room visits. Maternal emotional

distress variables were not significantly related to maternal adherence.

The maternal stress measure and parent-child dysfunction subscale scores on the

PSI-SF were positively related to seizure frequency of the child health status factor.

Mothers who reported higher stress levels and more dysfunctional interactions with their

children also reported increased seizure frequency at Time 2 for their children. The








results are consistent with previous findings that increased parental stress levels are

related to an increase in child seizure frequency and severity (Spector et al., 1999).

Spector et al. (1999) investigated the efficacy of a range of psychological interventions

and found that short-term group interventions combining supportive and cognitive-

behavioral components to improve coping strategies and decrease stress may be effective

for improving seizure control,

These findings may be attributable to the idea that mothers who have children with

increased levels of seizure frequency may be more stressed about their child's health and

well-being, which may carry over into increased levels of overall maternal stress. Of

note is that increased patient stress levels have been associated with increased seizure

frequency (Kyngas, 1990). Kyngas (1990) studied the compliance of adolescents with

epilepsy and found that good motivation, a strong sense of normality, subjective

outcome, energy and will-power, support from parents, physicians and nurses, a positive

attitude towards the disease and its treatment, no threat to social and emotional well-

being, and no fears of complications or seizures were related to good compliance and

decreased seizure frequency.

It is possible that dysfunctional mother-child relationships are related to increased

stress levels in children. Perhaps children involved in a dysfunctional parent-child

relationship are more likely to have increased stress levels and in turn, an increase in

seizure frequency as is suggested by Kyngas in 1990. Increased parent-child conflict has

been associated with poorer coping and increased behavioral and emotional problems in

children (Burt, Krueger, McGue, & lacono, 2003). Hodes and Garralda (1999) found that

mothers were more hostile towards their child with epilepsy than towards healthy siblings








and high levels of hostility and criticism were related to child behavior problems.

Additionally, mothers of children with higher levels of seizure frequency may have

reduced opportunities for positive interactions with their children and therefore, may

view their interactions with their child as more dysfunctional in nature.

Previous research has shown that greater parental stress and depression levels have

been associated with lower adherence to medications for children with epilepsy (Hazzard,

1990; Otero & Hodes, 2000). Hence, it was surprising that maternal emotional distress

variables were not significantly related to seizure-related emergency room visits or

maternal adherence. Note, however, that in these prior aforementioned studies, mothers

reported moderately high levels of maternal stress and depressive affect while in the

current study, mothers reported low levels of maternal stress and maternal depression

with the average number of depressive symptoms endorsed by mothers in our study

reported as two. This discrepancy suggests that low levels of maternal stress and/or

depressive affect may have accounted for the lack of a relationship between the maternal

emotional distress factor and the criterion variables.

Hypothesis 2: Maternal Perceptions of the Treatment Regimen, Adherence, and
Child Health Status

We predicted that Time 1 ratings of maternal perceptions of the treatment regimen

(perceived knowledge, perceived difficulty, and perceived efficacy), and Time 1 maternal

distress and Time 1 maternal perceptions of the treatment regimen would be associated

with Time 2 adherence to the child's medication regimen and child health status. Similar

to Hypothesis 1, a mixed pattern of results was obtained. Mothers who rated higher

levels of difficulty with their child's medical regimen at Time 1 concomitantly reported

increased seizure frequency levels and seizure-related emergency room visits for their








children at Time 2. Maternal perceptions regarding difficulty of the child's medical

regimen were positively related to child health status outcome variables. The relationship

between regimen complexity and maternal perceptions of regimen difficulty approached

significance (r = .205, p = .06). These findings may be related to the premise that

children with poorer health outcomes (e.g., increased levels of seizure frequency and

seizure-related emergency room visits) may have more difficult and complex regimens in

general, due to the nature of their illness. Furthermore, mothers who reported higher

levels of regimen efficacy had children with lower levels of seizure frequency.

No significant relationships were found between maternal perceptions of the

treatment regimen variables (e.g., regimen difficulty, efficacy, and knowledge) and

adherence. Overall findings suggest that maternal perceptions of the treatment regimen

are significantly related to poorer health outcomes for children with epilepsy but not

significantly related to maternal adherence. A possible explanation for this null finding is

that mothers of children with more difficult regimens and poorer health outcomes are

more adherent to their child's condition in an effort to minimize negative health outcomes

(e.g., increased seizure frequency and seizure-related emergency visits). Additionally,

relationships between maternal perceptions of the treatment regimen and adherence may

not have been identified because the questions comprising regimen perceptions are

preliminary in nature and are in need of further exploration, modification, and ultimately,

validation.

Hypothesis 3: Condition and Regimen-related Variables, Adherence, and Child
Health Status

We predicted that Time 1 condition and regimen-related variables would be

associated with Time 2 adherence to the child's medication regimen and child health








status. Hypothesis 3 received mixed confirmation. While no significant relationships

were found between condition and regimen-related variables and seizure frequency, one

condition and regimen-related variable was associated with adherence and child health

status outcome variables, partially supporting our hypothesis.

Relationships between condition and regimen variables and seizure frequency were

not related. Condition and regimen-related variables are more directly related to seizure-

related emergency room visits and maternal adherence to the medical regimen.

The concordance measure was related to the seizure-related emergency room visits

variable of child health status and pill count percentage indicator of adherence.

Concordance between the prescribed regimen and reported regimen was negatively

associated with seizure-related emergency room visits and positively associated with pill

count percentage. Mothers who were unable to consistently report their child's

medication regimen with the prescribed regimen may not have administered the child's

medication as prescribed, resulting in negative child health outcomes. The results also

demonstrate that mothers who were able to report their child's medication regimen were

more adherent to their child's medication regimen. It follows from these findings that

mothers who were unable to accurately report their child's medication regimen may not

have followed the prescribed regimen and in turn, reported increased levels of emergency

room visits related to their child's seizures. High concordance ratings may be

representative of good parent-physician communication. Peterson et al. (1982) found that

good patient-doctor communication has been related to better adherence. The significant

predictive relationship demonstrated between concordance ratings and outcomes (e.g.,

seizure frequency and maternal adherence) are of theoretical and practical importance in








that concordance can be easily assessed during each child's clinic visit and further

examined for its relationship to child health status outcomes and maternal adherence.

Hypothesis 4: Demographic Characteristics, Adherence, and Child Health Status

We predicted that Time 1 demographic characteristics would be positively related

to Time 2 adherence to the child's medication regimen and negatively with child health

status. Hypothesis 4 was only partially supported in that one demographic variable was

associated with adherence; family income was related to the pill count percentage

measure of adherence. Families with greater financial resources show better maternal

adherence to the child's medication regimen than those with less income. No significant

relationships were found between race, maternal education and Time 2 outcomes of child

health status and maternal adherence. A possible explanation for the lack of a significant

relationship between demographic variables and outcome measures is that the families in

the sample were not truly representative of families with children diagnosed with a

seizure disorder. Families in the present sample were wealthier than the average family

of children with epilepsy and mothers were more educated than average participants in

previous studies (Friedman et al., 1996). However, earlier research indicated that lower

socioeconomic status and lower parental education levels were correlated significantly

with non-adherence for asthma, cystic fibrosis, and renal disease (Patterson, 1985; Tebbi

et al., 1986) and lower socioeconomic status was related to parental non-adherence

(Becker & Maiman, 1975; Rapoff& Christopherson, 1982). One explanation for these

findings is that the population at the University of Florida's Health Science Center (UF

HSC) may not be representative of the larger demographic distribution. It is feasible that

wealthier and more educated families, because of their financial stability, are better

equipped to meet the demands of their child's medical appointments, including time off








work in order to travel to and from the UF HSC. Less wealthy families without such

resources may only be able to be seen closer to home in order to regularly attend medical

appointments. Second, it is plausible that wealthy and educated parents are more aware

of cutting-edge research and treatment available at the UF HSC and are more willing to

participate in clinical research. Finally, the small sample was from one study site and

may not have included sufficient participants to be representative of the true demographic

distribution.

Model Testing

Sets of predictors were identified as being related to the child health status

variables of seizure frequency and seizure-related emergency room visits and the

adherence measure of pill count percentage. Parent-child dysfunction appears to play a

major role in predicting seizure frequency. Improving the dysfunctional relationship

between parent and child may be beneficial in that a stronger, more positive parent-child

relationship may be developed that is related to decreased seizure frequency. However,

another interpretation of this finding is that parents of children with increased seizure

frequencies may have limited opportunities for positive interactions with their child.

Children with greater levels of seizure frequency may have less time to spend engaging in

non-medical interactions with their parents such as playtime and open communication.

In predicting seizure-related emergency room visits, the regimen-related variable of

concordance and maternal perceptions regarding difficulty of the child's medication

regimen emerge as important variables. These measures may be better predictors of

seizure-related emergency room visits compared to maternal emotional distress measures

because they specifically assess maternal understanding of the child's regimen and its

difficulty. It is suggested that more difficult regimens or poor concordance are more








closely related to seizure-related emergency room visits than maternal emotional distress

variables.

In predicting adherence, maternal emotional distress does not play a key role, but

the regimen-related variable of concordance indeed does. Concordance is likely to be

more predictive of maternal adherence than maternal emotional distress variables,

because concordance taps specifically into the mother's understanding of the child's

regimen. Our findings strongly indicate that incorrect reporting of the child's medication

regimen has a more negative effect on maternal adherence than maternal mood. In that

outcome measures are not solely related to maternal mood, it should be noted that these

variables warrant further investigation.

Study Strengths and Limitations

Studies identifying correlates and predictors of adherence have typically focused on

the following disease groups: diabetes, cystic fibrosis, juvenile rheumatoid arthritis,

asthma, and renal disease. A major strength of the current study was that it assessed

several variables in their relationship to child health status and maternal adherence for

children ages one to eleven diagnosed with epilepsy. Due to this study's multi-variable

approach, it is superior to research in earlier factor-based studies because it includes

individual variable relationships with outcome measures that may have been excluded or

overlooked by factor-based approaches, allowing for the reporting of new information in

the area of pediatric epilepsy research.

A significant methodological strength of the current study was the time-lag design.

This study design provided us with a better predictive model for Time 1 measures on

Time 2 outcomes, in that Time 2 measures were partialed out of Time 2 outcomes.








However, the time lag was relatively short (i.e., one month) and needs to be extended in

future studies to test the reliability of the results over a longer time interval.

The study also allowed us to tap into maternal perceptions of the treatment regimen

with the creation of questions to assess difficulty, efficacy, and knowledge regarding the

child's medical regimen. These types of questions are new to the area of epilepsy

research and can provide researchers with a great deal of information about maternal

perceptions of the treatment regimen and are important in their relationship to child

health status and maternal adherence outcomes.

However, several study limitations should be noted. Although attempts were made

to obtain a larger sample size, the final sample included 82 mother-child dyads. Small

sample size may have contributed to difficulties in detecting significant differences

between predictor and outcome variables. With this small sample size, statistical power

and prediction stability may have been methodological issues. The prediction models

may have been accurate in accounting for variance in the outcome measures but one

needs to examine the current results with caution due to this small sample size. The

correlations between some predictor and outcome variables were quite strong, and led us

to believe that a significant relationship does indeed exist between these measures.

However, a larger sample size would have resulted in greater power, which would have

enabled us to detect differences in other variables that may exist but were not apparent

due to low power leading to poor detectability. For example, difficulty scores

approached significance for inclusion in the final regression model predicting seizure

frequency (p = .056) as did the PSI-SF overall mean (p = .058). Having a larger sample

would also increase the power of prediction because variance on the measures would








increase, resulting in stronger predictive power. The generalizability of the results is

limited due to small sample size, higher than average family income, and higher than

average maternal education level.

Study limitations also include a lack of more concrete adherence measures such as

blood serum levels or prescription refill information. In the current study, we were

unable to report on blood serum levels because more than half of the patients did not have

these levels drawn consistently and some medications do not have normative data on

blood serum levels. Reporting of blood serum levels would have most likely resulted in

skewed findings related to this measure. Also, blood serum levels are not comparable

across different types of medications; thus, a coding system indicating therapeutic and

non-therapeutic blood level ranges suggesting adherence would be valuable in future

studies. Prescription refill information was also unavailable in the medical chart and was

not able to be obtained prospectively due to financial constraints of the current study.

Such information (e.g., blood levels and prescription refill information) is extremely

important to adherence research and should be included in future research.

Additional study limitations involve maternal perceptions of the treatment regimen.

The questions comprising the measures of difficulty, efficacy, and knowledge were

preliminary in nature and are not validated measures. In light of its strengths, the present

study can be used as a catalyst for future research in this area and serve as a model for

related fields.

Future Directions

In order for future studies to better assess predictors of maternal adherence and

child health status for young children with epilepsy, as well as outcome measures, the

following criterion should be met: (1) larger sample size; (2) assessment of child stress








levels; (3) validation of maternal perceptions of the treatment regimen questions; (4)

better measures of adherence; (5) inclusion of multiple study sites; (6) father data; and (7)

control group. With a larger sample size, structural equation modeling (SEM) procedures

would permit specific factor testing rather than the current variable-level testing model.

Multiple regression and other multivariate techniques can address a number of statistical

questions, but have a common limitation in that each technique can examine only a single

relationship at a time. SEM combats this problem and examines a series of predictive

relationships simultaneously. SEM techniques are distinguished by two characteristics:

(1) estimation of multiple and interrelated dependence relationships and (2) the ability to

represent unobserved concepts in these relationships and account for measurement error

in the estimation process. Relationships among predictor variables in each factor seem

promising for the SEM model given that there are significant relationships between the

specific variables comprising each factor of maternal emotional distress, maternal

perceptions of the treatment regimen, condition and regimen-related variables, and

demographic variables.

Assessment of child stress levels can assist in demonstrating a relationship between

maternal stress, child stress, and increased seizure frequency. It would be valuable to

empirically validate the specific questions assessing maternal perceptions of the treatment

regimen in the epilepsy population. Additional questions can then be created through this

validation process. This validation could result in the creation of an epilepsy-specific

measure that may have very important implications in child health status and maternal

adherence outcomes.








With proper funding, future studies could include blood levels, pharmacy refill

information, and electronic medication lid devices. Blood levels can be drawn as a part

of the study protocol, in order to have more consistent data regarding levels of

medication in the bloodstream for each study participant. The physician can aid in

identifying if the blood level falls into the therapeutic or non-therapeutic category. Study

protocol can also include permission to track the participant's pharmacy refill

information at their local pharmacy in order to have a more solid measure of adherence.

Electronic counts in which a device counts each time a medication bottle is opened and

closed, may be helpful in gaining additional adherence information.

The current study was exploratory in nature. Multi-site studies may serve in

gaining larger sample sizes and an overall distribution more representative of the general

population, especially in terms of annual family income and maternal education level.

Greater than 80% of patients were accompanied to the clinic visit by their mother.

However, attempts should also be made to include fathers in future studies by utilizing

mail-outs and/or semi-structured telephone interviews. A control group of children with

controlled seizures may be useful to compare to children that are currently experiencing

seizures. The consideration of additional variables and attempts at intervention studies in

this population is suggested for future research programs in order to gain more

information in this newly growing area of research.

Clinical Implications

The results of this study identified variables that were predictive of child health

status and maternal adherence to the child's medical regimen. An important clinical goal

is to assess maternal emotional distress, maternal perceptions of the treatment regimen,








condition and regimen-related, and demographic variables that may impact maternal

adherence to the child's medication regimen and child health status.

The PSI-SF could be utilized in pediatric clinical settings to identify mothers of

children with epilepsy who are in need of immediate interventions to assist them in

dealing with parenting stress and more specifically, parent-child dysfunction.

Therapeutic interventions could include stress management training, parent training,

progressive muscle relaxation training, and support groups for mothers of children with

seizure disorder. If maternal perceptions of the treatment regimen are identified as

particularly problematic, medical staff can assist with clarifying maternal concerns

regarding the regimen. Reported regimens can be compared with the prescribed regimen

to aid in identifying mothers who are in need of further explanation regarding their child's

medical regimen. Lastly, those mothers with lower education levels may need additional

assistance and additional time in order to properly understand the full nature of their

child's medication regimen. In conclusion, maternal stress, perceptions of the treatment

regimen, report of the regimen, and education level can be quickly assessed in the clinical

setting in order to have a beneficial effect on child health status and adherence outcomes.















APPENDIX A
BACKGROUND INFORMATION FORM TIME 1

Date_______


BACKGROUND INFORMATION

Child' s Name:

Mother's Name:

Address:


Phone #:( )


Child's Sex:


Male


Female


Child's Age: ____


Mother's Age: ____


Child's Birthdate:
m

Mother's Birthdate:


/ /
onth day year

/ /
month day year


Mother's Race:


White
African American
Hispanic American
Asian American
Native American
Biracial
Other:


1. White
2. African American
3. Hispanic American
4. Asian American
5. Native American
6. Biracial
7. Other:


Mother's Marital Status:

1. Single/Never married
2. Married
3. Separated
4. Divorced
5. Widowed


Number of Years Married (if applicable):

Child's Native Language:

Mother's Native Language:


Child's Race:






72



Mother's Occupation: _________

Current Employment Status:
(please circle all that apply)

1. Unemployed
2. Volunteer
3. Employed part-time
4. Employed full-time
5. Homemaker
6. Student
7. Retired
8. Other- Please describe ______

Years of mother's total formal education: (please circle number of years completed)

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

Post-secondary education: If attended college, number of years completed: ______

Degrees received:

Attended trade or technical school? Number of years completed: ______

How many children are in the family?______

What is the child's current grade in school?

Has this child ever had to repeat a grade in school?______ If so when?______

Please list the name(s) of the anti-seizure medications) your child is currently taking and the dose (how much
and how many times per day):


Please rate how difficult it is to manage which type of medication to give your child:

0 1 2 3 4 5 6 7 8
Not Somewhat
Difficult Difficult

Please rate how difficult it is to manage how much medication to give your child:

0 1 2 3 4 5 6 7 8
Not Somewhat
Difficult Difficult

Please rate how difficult it is to manage when to give medication to give your child:

0 1 2 3 4 5 6 7 8
Not Somewhat
Difficult Difficult


9 10
Extremely
Difficult



9 10
Extremely
Difficult



9 10
Extremely
Difficult









Please name your child's seizure diagnosis:___________

Over the past month, how many seizures did your child have per week?______
per month?______

Please rate how much you know about your child's seizure disorder:


0 1 2 3 4 5
Not Somewhat
Knowledgeable Knowledgeable


6 7 8 9 10
Extremely
Knowledgeable


Please rate how much you know about your child's seizure medicationss:


0 1
Not
Knowledgeable


2 3 4 5 6 7 8 9 10
Somewhat Extremely
Knowledgeable Knowledgeable


Please rate how much you know about the type of your child's seizure medicationss:


0 1
Not
Knowledgeable


2 3 4 5
Somewhat
Knowledgeable


6 7 8 9 10
Extremely
Knowledgeable


Please rate how much you know about the timing of your child's seizure medicationss:


0 1
Not
Knowledgeable


2 3 4 5 6 7 8 9 10
Somewhat Extremely
Knowledgeable Knowledgeable


If applicable, please rate how much you know about the number of your child's seizure medicationss:


0 1
Not
Knowledgeable


2 3 4 5
Somewhat
Knowledgeable


6 7 8 9 10
Extremely N/A
Knowledgeable


Please rate how much you know about the dosin2 of your child's seizure medicationss:


0 1
Not
Knowledgeable


2 3 4 5
Somewhat
Knowledgeable


6 7 8 9 10
Extremely
Knowledgeable


Please rate how good you feel your child's seizure medications are:


1 2 3 4 5
Somewhat
Good


6 7 8 9 10
Extremely
Good


Please rate how good you feel you are in managing your child's seizure medications:

0 1 2 3 4 5 6 7 8 9 10
Not Somewhat Extremely
Good Good Good


0
Not
Good









Please describe what type of information would be helpful to further your understanding of seizure
disorders:






Financial Status:

It is important for this study that we know something about your financial circumstances. We realize these are
extremely personal matters and we wish to assure you that your responses will be kept strictly confidential. Would
you please put an "X" next to the number that gives the best estimate of your total household income?


0. __ prefer not to answer

S1. ___ Under $10,000

2. ___ $10,000 to $20,000

3.___ $20,001 to 30,000


4. ___ $30,001 to $40,000

5. __ $40,001 to $50,000

6. ___ $50,001 to $60,000

7. ___ Over $60,000


Please do not answer the following, the researchers will fill out this portion.

Family ID: __________


Notes:















APPENDIX B
BACKGROUND INFORMATION FORM TIME 2

Date_______


BACKGROUND INFORMATION

Child' s Name: _________________

Mother's Name: ________________


Please list the name(s) of the anti-seizure medications) your child is currently taking and the dose (how much
and how many times per day):


Please rate how difficult it is to manage which type of medication to give your child:
0 1 2 3 4 5 6 7 8
Not Somewhat
Difficult Difficult

Please rate how difficult it is to manage how much medication to give your child:
0 1 2 3 4 5 6 7 8
Not Somewhat
Difficult Difficult

Please rate how difficult it is to manage when to give medication to give your child:
0 1 2 3 4 5 6 7 8
Not Somewhat
Difficult Difficult
Please name your child's seizure diagnosis:___________


9 10
Extremely
Difficult


9 10
Extremely
Difficult


9 10
Extremely
Difficult


Over the past month, how many seizures did your child have per week?______
per month?_____

Please rate how much you know about your child's seizure disorder:

0 1 2 3 4 5 6 7 8 9 10
Not Somewhat Extremely
Knowledgeable Knowledgeable Knowledgeable

Please rate much you know about your child's seizure medicationss:


0 1
Not
Knowledgeable


2 3 4 5 6 7 8 9 10
Somewhat Extremely
Knowledgeable Knowledgeable










Please rate how much you know about the type of your child's seizure medicationss:

0 1 2 3 4 5 6 7 8 9 10
Not Somewhat Extremely
Knowledgeable Knowledgeable Knowledgeable


Please rate how much you know about the timing of your child's seizure medicationss:


0 1
Not
Knowledgeable


2 3 4 5
Somewhat
Knowledgeable


6 7 8 9 10
Extremely
Knowledgeable


If applicable, please rate how much you know about the number of your child's seizure medicationss:

0 1 2 3 4 5 6 7 8 9 10
Not Somewhat Extremely N/A
Knowledgeable Knowledgeable Knowledgeable

Please rate how much you know about the dosing of your child's seizure medicationss:


0 1
Not
Knowledgeable


2 3 4 5
Somewhat
Knowledgeable


6 7 8 9 10
Extremely
Knowledgeable


Please rate how good you feel your child's seizure medications are:

0 1 2 3 4 5 6 7 8 9 10
Not Somewhat Extremely
Good Good Good

Please rate how good you feel you are in managing your child's seizure medications:


0 1 2 3 4 5
Not Somewhat
Good Good


6 7 8 9 10
Extremely
Good


Please describe what type of information would be helpful to further your understanding of seizure
disorders:


Please do not answer the following, the researchers will fill out this portion.

Family ID: __________


Notes:















APPENDIX C
MEDICAL CHART INFORMATION FORM TIME 1

MEDICAL CHART INFORMATION

Patient's Name:

Patient's DOB: ______ Age_____

Date of Diagnosis: _______

Newly Diagnosed: yes_____ no____
If no, time since diagnosis (months):_______

Seizure Frequency (over the past month):per week _____ per month ____

Seizure Type(s) {the type(s) of seizures your child is diagnosed with}:


Medication(s) and Dose Info:


ER Visits Dates:

Blood Levels:


Pharmacy Refill Information:


% Concordance between prescribed medication regimen and reported medication
regimen_______


Complexity_____















APPENDIX D
MEDICAL CHART INFORMATION FORM TIME 2

MEDICAL CHART INFORMATION

Patient's Name:

Patient's DOB: _____ Age_____

Date of Diagnosis: _______

Newly Diagnosed: yes_____ no____
If no, time since diagnosis (months):_______

Seizure Frequency (over the past month):per week ______ per month ______

Seizure Type(s) {the type(s) of seizures your child is diagnosed with}:





Medication(s) and Dose Info:______________


ER Visits Dates:

Blood Levels:


Pharmacy Refill Information:


% Concordance between prescribed medication regimen and reported medication
regimen______


Complexity______














LIST OF REFERENCES


Abidin, R.R. (1995). Parenting Stress Index.: Professional Manual (3rd ed). Odessa:
Psychological Assessment Resources.

Adams, P.F., Hendershot, G.E., & Marano, M.A. (1996). Current Estimates from the
National Health Interview Survey: National Center for Health Statistics. Vital
Health Statistics, 10(200), 85-86.

Anderson, B.J., Auslander, W.F., Jung, D.C., Miller, J.P., & Santiago, J.V. (1990).
Assessing family sharing of diabetes responsibilities. Journal ofPediatric
Psychology, 15, 477-492.

Appleton, R. & Gibbs, J. (1998). Epilepsy in Childhood and Adolescence. Malden:
Blackwell Science Incorporated.

Austin, J.K. (1989). Predicting parental anticonvulsant medication compliance using the
theory of reasoned action. Journal ofPediatric Nursing, 4(2), 88-95.

Beck, D.E., Fennell, R.S., Yost, R.L., Robinson, J.D., Geary, D., & Richards, G.A.
(1980). Evaluation of an educational program on compliance with medication
regimens in pediatric patients with renal transplants. The Journal of Pediatrics,
96, 1094-1097.

Bobrow, E.S., Avruskin, T.W., & Siller, J. (1985). Mother-daughter interaction and
adherence to diabetes regimens. Diabetes Care, 8, 146-151.

Bond, G.G., Aiken, L.S., & Somerville, S.C. (1992). The health belief model and
adolescents with insulin-dependent diabetes mellitus. Health Psychology, 11,
90-198.

Brownbridge,G. & Fielding, D.M. (1994). Psychosocial adjustment and adherence to
dialysis treatment regimens. Pediatric Nephrology, 8, 744-749.

Browne, T.R. & Holmes, G.L. (2000). Handbook of epilepsy (2nd edition).
Philadelphia: Lippincott, Williams, & Wilkins.

Burt, S.A., Krueger, R.F., McGue, M., & lacono, W. (2003). Archives of General
Psychiatry, 60(5), 505-513.








Cummings, E.M. & Davies, P.T. (1994). Maternal depression and child development.
Journal of Child Psychology and Psychiatry, 35(1), 73-112.

Downey, G. & Coyne, J. (1990). Children of depressed parents: An integrative review.
Psychological Bulletin, 108(1), 50-76.

Friedman, I.M., Litt, I.F., King, D., Henson, R., Holtzman, D., Halverson, D., &
Kraemer, H. (1986). Compliance with anticonvulsant therapy by epileptic youth.
Journal ofAdolescent Health Care, 7, 12-17.

Gudas, L.J., Kroocher, G.P., & Wyplj, D. (1991). Perceptions of medical compliance in
children and adolescents with cystic fibrosis. Developmental & Behavioral
Pediatrics, 12, 236-247.

Hazzard, A., Hutchinson, S.J., & Krawiecki, N. (1990). Factors related to adherence to
medication regimens in pediatric seizure patients. Journal ofPediatric
Psychology, 15(4), 543-555.

Hodes, M. & Garralda, M.E. (1999). Maternal expressed emotion and adjustment in
children with epilepsy. Journal of Child Psychology and Psychiatry, 7, 1083-
1093.

Jacobsen, A.M., Hauser, S.T., Wolfsdorf, J.I., Houlihan, J., Milley, J.E., Herskowitz,
R.D., Wertlieb, D., & Watt, E. (1987). Psychologic predictors of compliance in
children with recent onset diabetes mellitus. Journal of Pediatrics, 110, 805-811.

Kyngas, H. (2000). Compliance with health regimens of adolescents with epilepsy.
Seizure, 9, 598-604.

LaGreca, A.M., Follansbee, D., & Skyler, J.S. (1990). Developmental and behavioral
aspects of diabetes management in youngsters. Children's Health Care, 19, 132-
139.

Lorenz, R.A., Christensen, N.K., & Pichert, J.W. (1985). Diet-related knowledge, skill,
and adherence among children with insulin-dependent diabetes mellitus.
Pediatrics, 75, 872-876.

McCaul, K.D., Glasgow, R.E., & Schafer, L.C. (1987). Diabetes regimen behaviors:
Predicting adherence. Medical Care, 25, 868-881.

Murray, L., Sinclair, D., Cooper, P., Ducournau, P., Turner, P., & Stein, A. (1999). The
socioemotional development of 5-year-old children ofpostnatally depressed
mothers. Journal of Child Psychology and Psychiatry, 40(8), 1259-1271.

Myers, L.B. & Midence, K. (Eds.). (1998). Adherence to treatment in medical
conditions. New Delhi: Harwood Academic Publishers.









Nezu, A.M. (Ed.). (2000). Practitioner's guide to empirically based measures of
depression. New York: Kluwer Academic, Plenum Publishers.

Otero, S. & Hodes, M. (2000). Maternal expressed emotion and treatment compliance of
children with epilepsy. Developmental Medicine and Child Neurology, 42(9),
604-608.

Patterson, J.M. (1985). Critical factors affecting family compliance with home treatment
for children with cystic fibrosis. Family Relations, 34, 79-89.

Peterson, G.M., McLean, S., & Millingen, K.S. (1982). Determinants of patient
compliance with anticonvulsant therapy. Epilepsia, 23, 607-613.

Radius, S.M., Marshall, H.B., Rosenstock, I.M., Drachman, R.H., Schuberth, K.C., &
Teets, K.C. (1978). Factors influencing mothers' compliance with a medication
regimen for asthmatic children. Journal ofAsthma Research, 15, 133-149.

Radloff, L.S. (1977). The CES-D scale: A self-report depression scale for research in the
general population. Applied Psychological Measurement, 1, 385-401.

Rapoff, M.A. (1999). Adherence to pediatric medical regimens. New York: Kluwer
Academic & Plenum Publishers.

Rapoff, M.A. & Christophersen, E.R. (1982). Improving compliance in pediatric
practice. Pediatric Clinics of North America, 29(2), 339-357.

Shorvon, S. (2000). Handbook of epilepsy treatment. Oxford: Alden Press Limited.

Shope, J.T. (1988). Compliance in children and adults: review of studies. Epilepsy Res.
Suppl., 1, 23-47.

Smith, S.D., Rosen, D., Trueworthy, R.C., & Lowman, J.T. (1979). A reliable method
for evaluating drug compliance in children with cancer. Cancer, 43, 169-173.

Spector, S., Tranah, A., Cull, C., & Goldstein, L.H. (1999). Reduction in seizure
frequency following a short-term group intervention for adults with epilepsy.
Seizure, 8, 297-303.

Tebbi, C.K., Richards, M.E., Cummings, K.M., Zevon, M.A., & Mallon, J.C. (1986).
The role of parent-adolescent concordance in compliance with cancer
chemotherapy. Adolescence, 28, 599-611.

United States Census Bureau. (2001). Poverty in the United States: 2000. (U.S. Census
Publication No. P 60-214). Washington, DC: U.S. Government Printing Office.





82


Wynn, K.S. & Eckel, E.M. (1986). Juvenile rheumatoid arthritis and home physical
therapy program compliance. Physical & Occupational Therapy in Pediatrics, 6,
55-63.

Zysset, T., Redeberg, A., Vasella, F., Kupfer, A., & Bircher, J. (1981). Phenytoin
therapy for epileptic children: Evaluation of salivary and plasma concentrations
and of methods of assessing compliance. Developmental Medicine and Child
Neurology, 23, 66-75.














BIOGRAPHICAL SKETCH

Sobha P. Fritz was bom in St. Louis, Missouri in 1975 and raised in Tampa,

Florida. Sobha received a B.A. in psychology from the University of South Florida in

1997. Before coming to the University of Florida in 1998, she worked as a research

assistant for the Psychosocial Oncology Program at the H. Lee Moffitt Cancer Center.

Sobha married her high school sweetheart, Jason Jon Fritz, in December 1998, who is

completing his doctorate in molecular genetics and microbiology. Sobha received an

M.S. in clinical psychology from the University of Florida in 2000. She completed her

doctoral and internship training in the University of Florida's Department of Clinical and

Health Psychology, with a specialization in pediatric psychology, in 2003. Sobha will

begin a postdoctoral fellowship at Emory University and the Children's Healthcare of

Atlanta hospitals in Atlanta, Georgia in January 2004.








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


Robert Glueckatf, ClIaf.
Professor of Clinical and Health Psychology

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


Pssell Bauer
Professor of Clinical and Health Psychology

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


Paul Camrney
Professor of Neuroscience

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

Gary Gef en
Associate Professor of Clinical and Health
Psychology

I certify that I have read this study and that in my opinion it conforms to acceptable
standards of scholarly presentation and is fully adequate, in scope and quality, as a
dissertation for the degree of Doctor of Philosophy.
," _, -_ .... .__ ,. "--_.__..__
Claydell Home
Associate Professor of Nursing







This dissertation was submitted to the Graduate Faculty of the College of Health
Professions and to the Graduate School and was accepted as partial fulfillment of the
requirements for the degree of Doctor of Philosophy.

December 2003- _____' ....-
Dean, College of Health Professions


Dean, Graduate School








































UNIVERSITY OF FLORIDA
III1 1 IIIIIIIMt11 tl1 IIIIII11111
3 1262 08555 2866




Full Text
PREDICTING MATERNAL ADHERENCE AND CHILD HEALTH
STATUS IN CHILDHOOD EPILEPSY: AN EXPLORATORY MODEL
By
SOBHA P. FRITZ
A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA
2003

This document is dedicated to my husband, parents, and sisters for their support and
unconditional love. My family’s emphasis regarding the importance of education and
helping others has guided me into a career for which I am well suited. I would like to
especially thank my husband, Jason Jon Fritz, for always believing in me and providing
secure foundation of warmth and encouragement.

ACKNOWLEDGMENTS
This investigation could not have been carried out without the mentorship and
direction of Robert Glueckauf, Ph.D., the valuable contributions of Nicole Jagusztyn, the
dedication of Paul Carney, M.D., and the staff at the University of Florida Pediatric
Neurology Clinic. I would like to acknowledge my committee members for their
scholarly assistance and guidance. I would additionally like to thank my family for their
support and encouragement throughout this project.
iii

TABLE OF CONTENTS
Page
ACKNOWLEDGMENTS iii
ABSTRACT vii
CHAPTER
1 INTRODUCTION 1
Epilepsy: An Overview 1
Etiology 2
Diagnosing Epilepsy 3
Classification of the Epilepsies 4
Partial Seizures 4
Generalized Seizures 6
Treatment 8
Psychological Impact of Epilepsy 9
Adherence to the Regimen: An Overview 10
Adherence to the Regimen: Epilepsy 14
Maternal Emotional Functioning 18
2 AIMS AND STUDY JUSTIFICATIONS 21
Purpose 21
Study Model 22
Hypotheses 23
Maternal Emotional Distress, Adherence, Child Health Status 23
Maternal Perceptions of the Treatment Regimen, Adherence, Child Health
Status 23
Condition and Regimen-related Variables, Adherence, Child Health Status 23
Demographic Characteristics, Adherence, Child Health Status 24
3 METHODS 25
Study Design 25
Participants 26
Measures 29
Participant Background Information 29
Medical Chart Information 30
iv

Child Health Status and Adherence 31
Maternal Emotional Distress 32
4 RESULTS 35
Preliminary Analyses 35
Primary Analyses 38
Correlations among Background Variables, Predictors, Criterion Variables 38
Hypothesis 1: Maternal Emotional Distress, Adherence, Child Health Status ...40
Hypothesis 2: Maternal Perceptions of the Treatment Regimen, Adherence,
Child Health Status 42
Hypothesis 3: Condition and Regimen-related Variables, Adherence, Child
Health Status 46
Hypothesis 4: Demographic Characteristics, Adherence, Child Health Status...49
Model Testing 51
Prediction of adherence 51
Prediction of health status 52
Summary of Findings 55
5 DISCUSSION 57
Sample Characteristics of Demographics and Maternal Distress 57
Hypotheses 58
Hypothesis 1: Maternal Emotional Distress, Adherence, Child Health Status ...58
Hypothesis 2: Maternal Perceptions of the Treatment Regimen, Adherence,
Child Health Status 60
Hypothesis 3: Condition and Regimen-related Variables, Adherence, Child
Health Status 61
Hypothesis 4: Demographic Characteristics, Adherence, Child Health Status...63
Model Testing 64
Study Strengths and Limitations 65
Future Directions 67
Clinical Implications 69
v

APPENDIX
A BACKGROUND INFORMATION FORM - TIME 1 71
B BACKGROUND INFORMATION FORM - TIME 2 ....75
C MEDICAL CHART INFORMATION FORM - TIME 1 77
D MEDICAL CHART INFORMATION FORM - TIME 2 78
LIST OF REFERENCES 79
BIOGRAPHICAL SKETCH 83
vi

Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy
PREDICTING MATERNAL ADHERENCE AND CHILD HEALTH STATUS IN
CHILDHOOD EPILEPSY: AN EXPLORATORY MODEL
By
Sobha P. Fritz
December 2003
Chair: Robert L. Glueckauf
Major Department: Clinical and Health Psychology
Epilepsy is a chronic illness that typically begins in early childhood. Poor
adherence to the medication regimen can result in increased seizure frequency and
severity. Young children are reliant on their parents for their daily routine and
functioning. This issue is particularly salient for children with seizure disorders, who
rely on their parents to take anti-convulsant medication(s). Depressive symptomatology
and maternal stress have been associated with negative child status. Maternal emotional
factors such as depression and stress may play an important role in adherence to
medication for younger children with epilepsy. In the current study, we examined the
relationship between maternal adherence to medication (pill count percentage) and child
health status (seizure frequency, seizure-related emergency room visits) for children with
epilepsy and the following factors: maternal emotional distress (maternal depression,
maternal stress), demographics (family annual income, race, maternal educational level),
condition-related and regimen-related variables (disease type, disease duration, type and
vn

complexity of the medication regimen, concordance between prescribed regimen and
reported regimen), and maternal perceptions of the treatment regimen (perceived
efficacy, perceived knowledge, perceived difficulty).
Research participants included children, ages one through eleven, diagnosed with
epilepsy and were approached for study participation at the University of Florida
Pediatric Neurology Clinic. Eighty-two mother-child dyads completed the study.
Results indicated that as maternal stress, parent-child dysfunction, and maternal
perceived regimen difficulty increased, child seizure frequency significantly increased.
As maternal perceived regimen efficacy levels increased, child seizure frequency
significantly decreased. Results revealed that as maternal perceived regimen difficulty
levels increased, seizure-related emergency room (ER) visits significantly increased. As
regimen concordance rates increased, seizure-related ER visits significantly decreased.
Results demonstrated that as concordance rates increased, pill count increased.
Final hierarchical regression analyses suggested that parent-child dysfunction was
predictive of seizure frequency. Maternal perceived regimen difficulty ratings and
concordance ratings between the prescribed and reported regimen were predictive of
seizure-related ER visits. Concordance ratings between the prescribed and reported
regimen were predictive of pill count percentage. These findings suggest that the
identification of specific variables that influence maternal adherence and child health
status may be particularly helpful for children with epilepsy.
viii

CHAPTER 1
INTRODUCTION
Epilepsy: An Overview
Epilepsy is a chronic illness that typically begins in early childhood in which the
patient is prone to experience recurrent epileptic seizures, where two or more attacks
have occurred. Epilepsy is defined as a manifestation of an abnormal and excessive
synchronized discharge of a set of cerebral neurons (Shorvon, 2000). Clinical
manifestations are sudden and transient and can include a wide variety of motor, psychic,
and sensory phenomena with or without alteration in consciousness or awareness. The
specific symptoms that occur depend on the part of the brain involved. Status epilepticus
is defined as a condition in which epileptic seizures continue, or are repeated without
consciousness, for a period of 30 minutes or more and this occurrence often requires a
visit to the emergency room. Untreated status epilepticus can be fatal.
Provoked seizures have an obvious and immediate preceding cause, such as an
acute systemic, metabolic, or toxic insult or an acute cerebral event (stroke, trauma, and
infection). Provoked seizures are often isolated events and do not tend to recur once the
cause is removed. A person is described as having active epilepsy when at least one
seizure has occurred in the preceding 2-5 years. Epilepsy is said to be in remission when
no seizures have occurred in the preceding 2-5 years (Appleton & Gibbs, 1998). There
are approximately 80 cases of epilepsy per 100,000 people per year, with a range of 50-
120 cases. Nearly 75 per 100,000 adolescents per year go on to develop some form of
epilepsy. The cumulative incidence, risk of an individual developing epilepsy in his or
1

2
her lifetime, is between 3% and 5%. The highest incidence rates are observed in
neonates and young children with a second peak in old age. Greater than 50% of seizures
start in childhood and begin before age 20 (Browne & Holmes, 2000).
The prevalence of active epilepsy is 5-10 cases per 1,000 people per year. Isolated
seizures (first and only) occur in approximately 20 people per 100,000 each year. The
prevalence of epilepsy in school age children (aged 5-17 years) is 0.7%-0.8%. The
prevalence of epilepsy in adolescents is 6-7 per 1,000 adolescents. One-third of all
epilepsies that begin in childhood will have spontaneously remitted by puberty and this
change is usually sustained throughout adult life. Epilepsy is relatively static after early
childhood but shows a tendency to rise in old age. Epilepsy is more common in
developing countries and is thought to be associated with poorer nutrition, poorer
hygiene, higher prevalence of infectious diseases, and a higher proportion of children in
the population (Appleton & Gibbs, 1998).
Etiology
Congenital and genetic conditions are most commonly related to early childhood
epilepsy. Genetics are thought to play an important role in seizure disorders. Hereditary
and genetic factors are related to idiopathic or primary epilepsy syndromes and possibly
in some secondary epilepsies, but known single-gene disorders are related to epilepsy in
only 1-2% of cases. In older children and young adults, common causes of epilepsy
include inherited predisposition, hippocampal sclerosis, alcohol or drug abuse, and
trauma. Vascular disease is a predominant cause of epilepsy in the elderly. Tumors and
infections can occur at any age but malignant tumors are more likely to occur after age 30
(Shorvon, 2000). In 70-75% of childhood cases of epilepsy, a specific cause may not be

3
noted, but the etiology of the disease may be apparent from clinical information and test
results (Appleton & Gibbs, 1998).
Patients with specific syndromes such as Fragile X syndrome, Sturge-Weber
syndrome, Huntington's, Niemann-Pick disease type C, tuberous sclerosis (Boumeville's
syndrome), neurofibromatosis (type 1 and type 2), and Down’s syndrome may develop
epilepsy. Infectious diseases that involve the brain such as viral and bacterial meningitis
also lead to epilepsy. Cerebral tuberculosis frequently causes epilepsy. Malaria, strep
infections, and encephalitis are all thought to be related epilepsy. Cerebral tumors are
related to approximately 6% of newly diagnosed cases of epilepsy. In children, tumors
are not typical causes of epilepsy. Head injuries account for approximately 2-12% of all
cases of epilepsy (traffic accidents and falls). Drugs, alcohol, toxins, and metabolic
disorders have been associated with seizure disorders. Additional causes of epilepsy
include brain malformations, unruptured aneurysms, intracranial hemorrhages, lesions,
perinatal injury, and hippocampal sclerosis. Hippocampal sclerosis is the most common
cause of temporal lobe epilepsy (Shorvon, 2000).
Diagnosing Epilepsy
A physician needs detailed information to render an epilepsy diagnosis and a
diagnosis of epilepsy is dependent almost entirely on the history given. During the
assessment the doctor will typically gain the following information: a very detailed
history of the seizure(s), an eyewitness account of the seizure(s), a history of previous
illnesses or traumatic events, family history, and history of any co-occurring disease(s)
and treatment(s). Noteworthy is the fact that an accurate eyewitness account is essential
and may be the only history available for young children. A physical examination will be
conducted. Video recordings, sleep studies, and EEGs of suspected seizures are useful

4
aids for diagnosis. Some characteristics that may mimic epilepsy include headaches,
jitters, tic disorder, shuddering, shivering, repetitive ritualistic movements, vacant stares,
daydreaming, inattention, psychological disturbances, and excessive sleep (Browne &
Holmes, 2000).
Classification of the Epilepsies
Seizures are classified into two main categories: partial and generalized. These two
categories are further classified. Partial seizures include simple partial, complex partial,
and partial seizures evolving to secondary generalized. Generalized seizures are
classified as typical absence (petit mal), atypical absence, myoclonic, clonic, tonic, tonic-
clonic (grand mal), atonic, and unclassifiable (Appleton & Gibbs, 1998).
Partial Seizures
Partial seizures are further classified as “simple,” in which consciousness is
retained, or “complex,” in which consciousness is impaired or lost. Simple partial
seizures (beginning locally) can occur with motor symptoms, sensory symptoms,
autonomic symptoms, and psychic symptoms. Simple partial seizures with sensory,
autonomic, or psychic symptoms may be easily overlooked in younger children, who may
be unable to describe such symptoms. Partial seizures may become secondary
generalized, resulting in tonic-clonic (grand mal) convulsions (Browne & Holmes, 2000).
Simple partial seizures last only a few seconds and the symptoms that occur depend
on the part of the cortex involved. Simple partial seizures can include motor
manifestations such as jerking or spasms, which usually occur from frontal or central
region epilepsies. Simple partial seizures can include somatosensory manifestations such
as simple hallucinations that can take the form of tingling or numbness, an electrical
shock-like feeling, burning, pain, or a feeling of heat. The epileptic focus is usually in

5
the central or parietal regions. Simple visual phenomena such as flashing lights and
colors usually occur if the calcarine cortex is affected. A rising epigastric sensation is the
most common manifestation reported, which typically occurs when the mesial temporal
lobe is affected. Simple partial seizures can also include autonomic manifestations such
as changes in skin color, blood pressure, heart rate, and pupil size and these changes
usually are related to generalized or complex partial seizures of frontal or temporal
origins. Psychic manifestations are common as well. Psychic auras can take various
forms and are more common in complex partial seizures. They usually occur from
epilepsy arising from a temporal, frontal, or parietal focus. There are six principal
catergories of auras: dysphasic symptoms (cortical or speech areas are effected),
dysmnestic symptoms (disturbance of memory), cognitive symptoms (dreamy states and
depersonalization), affective symptoms (fear, depression, anger, irritability, and elation),
illusions of size, shape, weight, distance or sound, or structural hallucinations (visual,
auditory, gustatory, or olfactory forms) (Appleton & Gibbs, 1998).
Complex partial seizures have three main components: auras, altered
consciousness, and automatisms. The auras are equivalent to those of simple partial
seizures and usually last a few seconds, but in rare cases a prolonged aura persists for
minutes, hours, or even days. Altered consciousness may follow the aura and can include
motor arrest, during which the patient is motionless. The patient may be described as
vacant or glazed and often associated is a spasm, posturing, or mild jerking.
Automatisms are involuntary motor actions that occur during or in the aftermath of
epileptic seizures, in a state of unconsciousness where there is total amnesia for the
events.

6
Partial seizures (simple, complex, or simple evolving to complex) may spread to
become generalized. The partial seizure is often experienced as an aura in the seconds
before the generalized seizure. The generalized seizure is usually tonic-clonic, tonic, or
atonic (Browne & Holmes, 2000).
Generalized Seizures
There are two types of absence seizures: typical absence (petit mal seizures) and
atypical absence. In typical absence seizures there is an abrupt and sudden loss of
consciousness, cessation of all motor activity, tone is usually preserved, there is no fall,
the patient is unaware and inaccessible, the patient often appears glazed or vacant, the
attack ends as abruptly as it started, and previous activity is resumed as if nothing
happened. In atypical absence seizures, the duration is longer, loss of awareness is often
incomplete, associated tone changes are more severe, and the onset and cessation of
attacks are not as abrupt (Appleton & Gibbs, 1998).
With myoclonic seizures, there is a brief contraction of a muscle, muscle group, or
several muscle groups that can be single or repetitive, varying in severity from a nearly
imperceptible twitch to a severe jerking which can result in a sudden fall or the
propulsion of handheld objects (flying saucer syndrome). Clonic seizures consist of
asymmetrical and irregular jerking. Clonic seizures are most frequent in neonates,
infants, and young children. Tonic seizures are associated with muscle contractions with
altered consciousness. In tonic seizures, the contraction causes extension of the neck,
contraction of facial muscles, widening of the eyes, up-turning of the eyeballs,
contraction of respiratory muscles, and spasms of the upper limb muscles (Browne &
Holmes, 2000).

7
Tonic-clonic seizures are most associated with the type of epilepsy that the public
imagines. The tonic phase is sometimes preceded by a period in which an attack is
anticipated (prodromal period) defined by a vague feeling or by the occurrence of
increasing myoclonic jerking. If an aura occurs, this indicates that the tonic-clonic
seizure is secondary generalized. The seizure is initiated by loss of consciousness where
the patient will fall if standing and there is a brief period of tonic flexing and then a
longer phase of rigidity, with the eyes rolled up, the jaw clamped shut, the limbs stiff, and
the fists clenched, then respiration ceases. The clonic phase of tonic-clonic seizures has a
tonic stage that lasts 10-30 seconds and is followed by a clonic phase, during which
convulsive movements occur, usually of all four limbs and the jaw and facial muscles.
Breathing may be labored and saliva may froth from the mouth containing blood from
tongue-biting. The convulsive movements decrease in frequency, typically to about 4
clonic jerks per second, and increase in amplitude as the attack progresses. Autonomic
features (flushing, changes in BP, changes in pulse, and increased salivation) are
common. The clonic phase lasts 30-60 seconds and is followed by further brief tonic
contraction of the muscles, sometimes accompanied by incontinence. The final phase
lasts between 2-30 minutes and is characterized by flaccidity of muscles and then
consciousness is slowly regained (Browne & Holmes, 2000).
The most severe form of the atonic seizure is the classic drop attack in which all
postural tone is suddenly lost, causing the person to collapse. In atonic seizures, tone can
be more restricted, resulting in bobbing of the head, a bowing movement, or sagging at
the knee. These seizures are short and are usually followed by immediate recovery.

8
At least 1/3 of all seizure disorders are not classifiable and fall into the
unclassifiable category. Seizures are described as unclassifiable because typical clinical
and EEG patterns do not conform to the defined categories of seizures (Browne &
Holmes, 2000).
Treatment
Various drug treatments for epilepsy are available. One of the most difficult
aspects of drug treatment is adherence to medication. Some issues that arise are when to
start drug therapy, which drug to use, what dose to give, when to change the drug and/or
add another drug, when to stop a drug, and when to measure blood levels of the drug.
Drug treatments can vary from dosing once or twice a day to several times a day and drug
types include pills, liquids, sprinkles, chewables, suppositories, and inhalers (Appleton &
Gibbs, 1998).
Most physicians would recommend beginning drug treatment after a cluster of
seizures. Once a drug regimen is started, the goal is to achieve seizure control without
extreme side effects. Before adding additional medications, the dose of the initial drug
should be increased to the maximum tolerated level. Currently, the recommended first-
line of drugs in treating childhood epilepsy is sodium valproate for generalized
epilepsies, and carbamazepine for partial seizures. The second line of defense includes
carbamazepine, gabapentin, lamotrigine, and topiramate for generalized seizures, and
vigabatrin, gabapentin, lamotrigine, and topiramate for partial seizures. The third line of
defense includes phenytoin, ethosuximide, clonazepam, phenobarbitone, clobazam,
carbamazepine, acetazolamide, and topiramate for generalized seizures, and sodium
valproate, clobazam, and phenytoin for partial seizures. Many physicians take a polydrug

9
treatment approach of epilepsy treatment, specifically for those patients who do not
respond well to first, second, and third line treatment approaches (Shorvon, 2000).
Surgical treatment is an option for patients with severe intractable epilepsy that
fails to respond to medication and other forms of treatment. Alternative methods of
treatment include the ketogenic diet, steroids, intravenous immunoglobulins, behavior
therapy, acupuncture, herbal therapy, vagus nerve stimulation, and seizure alert dogs.
Psychological Impact of Epilepsy
Psychiatric disorders occur more frequently with epilepsy than with any other
chronic disorder (Appleton & Gibbs, 1998). Children and adolescents with poorly
controlled epilepsy have more restrictions imposed upon them to protect them from
getting injured during a seizure. They may be unable to participate in activities with their
peers such as swimming, driving, cycling, climbing, and sports. These restrictions can
result in fewer contacts with peers. Social stigma concerns such as teasing, isolation, and
poor social contact have been associated with increased restrictions. Restrictions,
depression, and isolation have been associated with poorer adherence in children and
adolescents with epilepsy (Kyngas, 1990). Poorer adherence to medication is associated
with increased seizure frequency, resulting in a larger number of epilepsy-related injuries
and deaths. One preventable cause of increased seizure frequency and its associated
injuries is taking the anti-epileptic medication as prescribed. Adherence to the medical
regimen becomes an important factor in decreasing seizure frequency.
Epilepsy is a disorder that is associated with a high degree of uncertainty. This
uncertainty and sudden loss of control have been associated with anxiety and depression
in patients with epilepsy including adults, children, and caregivers (Kyngas, 2000).
Increased stress levels and anxiety have been demonstrated as related to an increase in

10
seizure frequency and seizure severity (Spector et al., 1999). The effects of maternal
emotional adjustment and adherence with young children with epilepsy have received
limited attention.
Adherence to the Regimen: An Overview
One-third of all patients and 50-55% of chronically ill patients are non-adherent to
their prescribed medical regimens (Rapoff, 1999). Non-adherence can negatively impact
health status and health-related quality of life and can lead to increased risks of morbidity
and mortality. Non-adherence is also related to increased health care costs, where the
cost of non-adherence in the United States is nearly $100 billion each year (Rapoff,
1999). Additionally, non-adherence may inadvertently negatively impact clinical
decisions made by physicians. Physicians who attribute disease-related problems to
ineffective or incorrect prescriptions and not to the patient's adherence problems may
change the patient's medical regimen unnecessarily or inappropriately.
Identifying correlates of adherence is important because they facilitate the
development of risk profiles, medical regimens, and future research using matched
samples. Patient and family factors, disease-related factors, and treatment regimen
factors have been identified as correlates of adherence (Rapoff, 1999). The majority of
studies have focused exclusively on patient and family factors and their relationship to
adherence.
Patient/family correlates include demographics (child age, parental education level,
work status, number of children in the home, and socioeconomic status), knowledge
about disease course and treatment, as well as adjustment and coping. Previous research
has shown that patient and family demographic variables have a significant influence on
child and maternal adherence. Younger children with cancer, cystic fibrosis, diabetes,

11
and renal disease are more likely to be adherent to their medication regimen compared to
adolescents (Anderson, Auslander, Jung, Miller, & Santiago, 1990; Beck et al., 1980;
Bond, Aiken, & Somerville, 1992; Brownbridge & Fielding, 1994; Gudas, Koocher, &
Wyplj, 1991; Jacobsen et ah, 1987; LaGreca, Follansbee, & Skyler, 1990; Patterson,
1985; Smith, Rosen, Truworthy, & Lowman, 1979; Tebbi et ah, 1986). Boys tend to be
less adherent than girls to regimens for cystic fibrosis and diabetes (Lorenz, Christensin,
& Pichert, 1985; Patterson, 1985). Lower socioeconomic status (SES) and lower parental
education levels have been related to non-adherence to regimens for asthma, cystic
fibrosis, diabetes, and renal disease. Patients in larger families or whose mothers work
outside of the home tend to be less adherent to cancer and cystic fibrosis regimens
(Patterson, 1985; Tebbi et ah, 1986).
Knowledge about the treatment regimen is an interesting and important factor in
predicting adherence to treatment. However, studies that have examined this factor have
shown an inconsistent pattern of results. Several investigators have reported that patients
who are less knowledgeable about their specific condition and treatment regimen tend to
be less adherent to regimens for cancer, cystic fibrosis, and diabetes (Gudas et ah, 1991;
LaGreca et ah, 1990; Tebbi et ah, 1986) whereas knowledge appears to be unrelated to
adherence for patients with renal disease (Beck et ah, 1980). Maternal knowledge for
children with diabetes was positively related to adherence for preadolescents and
unrelated to adherence for adolescents (LaGreca et ah, 1990).
The relationship of adjustment and coping to adherence has been examined across a
variety of childhood disorders. Poor parental coping has been associated with non¬
adherence to medical regimens for children with juvenile rheumatoid arthritis and renal

12
disease. Increased parental depression has been related to poor adherence to renal disease
regimens (Brownbridge & Fielding, 1994; Wynn & Eckel, 1986). Greater parental
anxiety and perceived stress have been associated with lower adherence to anti-epileptic
medications for children with epilepsy (Hazzard, 1990). Parental compliance behavior is
related to perceptions of the medications' efficacy as well as its side effects (Shope,
1988). Austin (1989) suggests that attitudes toward giving anti-epileptic medications and
behavioral intentions are important variables in the prediction of parental compliance
behavior. Parental knowledge about the efficacy and side effects of medications
influences parental attitudes towards medications and improves their medication-giving
behavior (Austin, 1989).
Patients who lack parental monitoring of regimen procedures or have families
where the role of the parental monitor is ambiguous tend to be less adherent to their
prescribed medical regimen. Parental monitoring has been found to be a key factor in
improving medication adherence for adolescents with chronic illness (Rapoff, 1999).
Condition-specific correlates include duration, course, symptoms, and perceived
severity of the condition. Diseases with a longer duration are related to poorer adherence.
Poorer adherence to regimens for diabetes, juvenile rheumatoid arthritis, and renal
disease has been related to longer illness duration. The relationship of disease course and
adherence has not been specifically examined for children with chronic illness (Rapoff,
1999). Condition-specific symptoms are an important factor to examine in their
relationship to adherence. Increased health problems and hospitalizations are related to
poorer adherence for patients with renal disease (Beck et al., 1980). Greater seizure
activity has been related to decreased adherence to anticonvulsant medications for

13
children with epilepsy (Hazzard, 1990). Maternal perceptions of higher disease severity
have been associated with better adherence to asthma medications (Radius et al., 1978).
Medication regimen-related correlates include type and complexity, costs, side
effects, and perceived efficacy of the medication regimen. Patients with more complex
regimens tend to be less adherent to their prescribed regimen in that more complex
regimens are more often associated with polydrug treatments, with higher numbers of
medication types and dosages. This is particularly true for patients with cystic fibrosis,
diabetes, and juvenile rheumatoid arthritis. Treatment costs can be related to non¬
adherence especially for those patients and families who are economically disadvantaged.
Those medical regimens that produce a greater number of negative side effects tend to be
related to poorer adherence to the regimen. Patient and parental perceptions of
medication efficacy are especially related to adherence. Higher levels of perceived
benefits rated by patients and parents are associated with better adherence to regimens for
asthma and diabetes (Bobrow, Avruskin, & Siller, 1985; Bond et al., 1992; McCaul,
Glasgow, & Schafer, 1987; Radius et al., 1978).
In summary, important correlates of adherence for children and adolescents with
chronic illness include patient/family factors (demographics, knowledge, adjustment and
coping, parental monitoring), disease factors (duration, course, symptoms, perceived
severity), and regimen factors (type and complexity, costs, side effects, efficacy). These
summarized findings demonstrate that correlates of adherence to medical regimens are
critical in improving adherence. The majority of studies conducted to identify correlates
of adherence to medical regimens for children with chronic disease have focused on the
following: asthma, cancer, cystic fibrosis, diabetes, and renal disease. Studies examining

14
the specific correlates of adherence for children and adolescents with epilepsy are limited
in number and require further examination.
Adherence to the Regimen: Epilepsy
The vast majority of people with epilepsy begin having their seizures before the age
of 20, and more than 50% of epilepsy cases begin in childhood. Thus, most people
experience their first seizure at a time that is critical to the acquisition and development
of basic cognitive and social competencies, which are crucial for long-term academic,
interpersonal, and vocational adjustment. Due to the long-term nature of epilepsy and its
related medication regimen, medication adherence to anti-epileptic drugs (AEDs) is
extremely important from the time of diagnosis and throughout the course of treatment.
Although correlates of adherence have been identified in a wide range of pediatric
chronic illnesses, only a few studies have been conducted to date on adherence to
treatment among children and adolescents with seizure disorders. We continue to lack
basic information about the psychological, environmental, and familial factors that
influence adherence in children with epilepsy. Furthermore, there is only a modicum of
outcome research on intervention approaches to improve adherence in this at-risk
population. A few factors have been identified that are specifically related to individuals
with seizure disorders.
Well-controlled seizure patients are asymptomatic, which may gradually contribute
to decreased adherence to the medication regimen. Furthermore, nonadherence does not
necessarily lead to immediate symptomatology for patients with epilepsy. Pediatric
seizure patients are typically on long-term medication regimens, for which adherence is
worse than for short-term regimens (Becker & Maiman, 1975; Davis, 1966; Rapoff &
Christopherson, 1982).

15
A demographic factor generally associated with parental non-adherence is lower
socioeconomic status. Additionally, ethnic minority patients have been found to be more
likely to be nonadherent to their child's medical regimen (Becker & Maiman, 1975;
Rapoff & Christopherson, 1982).
Conducting research on adherence with pediatric seizure patients is made difficult
by the methodological problems involved in assessing patient's adherence. Although
blood levels of anticonvulsant medications can be obtained, it is difficult to translate
blood levels into comparable measures of compliance for patients on different
medications and doses. Therapeutic ranges vary for different medications and vary
according to dosage, time since the medication was ingested, and individual metabolic
rates, and hormonal changes (which are especially salient for adolescent females).
Furthermore, for patients taking more than one medication, there are medication
interactions, with some medications altering the blood level concentrations of other
medications. Because of these complications, studies examining pediatric patient
adherence to anticonvulsant medication regimens have tended to assess only one
medication level or have used group differences in blood levels to assess the validity of
other compliance measures (self report, pill counts). With these limitations, estimates of
parental adherence to anticonvulsant medication regimens for pediatric patients ranges
from 25% to 54% (Zysset, et al., 1981). These estimates are generally consistent with
adherence rates from other pediatric populations where a long-term medication has been
prescribed (Sackett & Haynes, 1976).
Only three studies have assessed factors related to parental medication adherence
with pediatric seizure patients (Freidman et al., 1986; Peterson et al., 1982). The studies

16
assessed at least some factors from the demographic, illness, treatment regimen, and
attitudinal domains. However, the Peterson et al. study focused primarily on adults, with
only 18 adolescents in the sample. Demographic variables were not demonstrated to be
associated to adherence in the pediatric seizure disorder populations in the two studies
(Freidman et al., 1986; Peterson et al., 1982). In terms of illness variables, Peterson et al.
(1982) found that patients experiencing more seizures in the past year, more recent
seizures, or having tonic-clonic seizures were more likely to be adherent. The only
treatment-related variable related to adherence for children who had an active role in
taking their medication, particularly adolescents, was perceived personal independence
(Freidman et al., 1986). Research in children with other medical conditions has also
shown that children with greater personal freedom (or fewer behavioral restrictions) were
more likely to be adherent with their medication regimen. Peterson et al. (1982) found
that satisfaction with the physician's explanation was unrelated to adherence. However,
in studies with other populations, patient satisfaction and good patient-doctor
communication have been related to good adherence. Peterson et al. (1982) also found
that worry about one's own health was related to self-reported adherence.
Shope reviewed a study that closely followed the Health Belief Model (HBM)
framework for pediatric epilepsy patients (Shope, 1992). Parents of a random sample of
patients drawn from an urban medical center pediatric seizure clinic were seen at their
regular appointment and interviewed from January of 1975 through May of 1975.
Questions asked included variables from the HBM as well as a wide range of medical,
psychosocial, and demographic variables. Serum antiepileptic levels were taken at the
clinic visit on 65 of the 90 study participants. Data presented were based on 65 patients.

17
Patients were judged to be compliant or non-compliant by an experienced clinician's
relating the prescribed dosage of medication to the obtained blood level. This process
was used, rather than therapeutic blood levels, because some patients had less than
therapeutic dosages prescribed.
The patient sample was 55% female and 78% African American with a mean age of
10 years. For 55% of the children, seizure disorder was the only medical problem
identified. Age of onset ranged from birth to 11.5 years and duration of the seizure
disorder ranged from 6 months to 19 years. Forty-eight percent of the children were
prescribed one medication, 39% were prescribed two, and 13% were prescribed three.
Phenobarbital was prescribed for most of the patients. The assessment of compliance, by
looking at blood levels, revealed that 57% of the parents gave their children their
medication as prescribed, 25% were giving less than the prescribed dose, and 18% were
giving no medication. Forty-three percent of the patients were not receiving their
medication as prescribed.
Results indicated that parents who reported at least one occurrence of running out
of their child's medication for several days were more likely to have children whose
blood levels revealed a discrepancy between the obtained and prescribed dosage (p < .05)
and parents who reported that they missed giving their child a medication more than once
a week were less likely to be compliant, as measured by blood levels (p < .001)
(motivational factors of the HBM). Recency of seizures was significantly positively
correlated with compliance (p < .05) (value of illness threat reduction of the HBM).
Child's enrollment in school was significantly and negatively correlated with compliance

18
(p < .01). Parents who reported that their child did not have any additional medical
problems were more compliant (p < .05).
In summary, factors that appear to influence parental non-adherence to anti¬
epileptic medications for children with epilepsy include lower socioeconomic status and
ethnic minority group membership. However, the criteria for assessing socioeconomic
status have not been well defined nor have they been applied consistently across studies,
thus rendering comparability of socioeconomic findings somewhat suspect. Research
findings indicate that parents whose children had increased seizure frequency, more
recent seizure activity, and tonic-clonic seizures were more adherent to their child's anti¬
epileptic medication regimen. Parents who perceived an increased level of disease
severity (more severe seizure events with greater frequency and intensity) were more
adherent to their child's medication regimen. The current study examined factors that are
related to medication adherence in pediatric epilepsy populations, specifically with
younger children with epilepsy.
Maternal Emotional Functioning
Depression is a highly prevalent mental disorder, especially among those women of
childbearing age. Downey and Coyne (1990) found that there is a high rate of serious
psychological problems in children of depressed mothers and that depressed mothers
have a host of difficulties as parents. Murray, Sinclair, Cooper, Ducoumau, Turner, and
Stein (1999) reported that exposure to maternal depression in the early postpartum
months may have an enduring influence on childhood psychological adjustment.
Cummings and Davies (1994) indicated that children of depressed parents are at
increased risk for the development of psychopathology and that they are two to five times
more likely to develop behavior problems than are children of normal parents. Depressed

19
mothers view their parenting role less positively than non-depressed mothers, experience
negative feelings towards parenting roles, and have feelings of rejection and hostility
towards their child (Downey & Coyne, 1990). Depressed mothers also tend to
demonstrate difficulties interacting with their children and displaying positive emotions
towards their children.
Hodes and Garralda (1999) examined the effects of maternal expressed emotion on
the psychological adjustment of children with epilepsy. Maternal expressed emotion and
risk for psychiatric disorders (for children and mothers) were assessed. The researchers
found that mothers showed significantly more emotional overinvolvement and a trend for
more hostility towards their children with epilepsy than they displayed towards sibling
controls. Maternal emotional overinvolvement was unrelated to child behavior problems.
High levels of criticism and hostility also were shown to covary with child behavior
problems (Hodes & Garralda, 1999). The effects of maternal expressed emotion on
adherence to medication were not assessed. However, the results suggest that maternal
expressed emotion is related to negative psychological functioning in children with
epilepsy. Thus, it seems highly plausible that maternal expressed emotion may have a
substantial impact on adherence to anti-epileptic medications for children with epilepsy.
Otero and Hodes (2000) assessed the relationship between maternal expressed
emotion and adherence to treatment in children with epilepsy. The researchers reported
that good treatment compliance was related to less maternal hostility and criticism for
epileptic children and their mothers. Children and mothers in the good compliance group
had fewer reported psychiatric symptoms. Poor treatment compliance for children with
epilepsy was associated with greater maternal psychological disturbance. Children with a

20
greater number of behavioral problems were less compliant. A relationship between
maternal depression and poorer compliance was reported (Otero & Hodes, 2000).
The relationship between maternal depression and behavior problems has been
investigated intensively across a variety of pediatric medical populations. Poor maternal
emotional functioning appears to exert a consistent negative impact on children with
disabilities and their healthy siblings. Research findings demonstrate a clear relationship
between maternal emotional functioning and childhood behavior problems.
Spector et al. (1999) found that increased maternal stress and anxiety levels were
related to an increase in seizure frequency and severity. Hazzard (1990) reported that
increased levels of parental anxiety and perceived stress were related to lower adherence
to anti-epileptic medications for children with a seizure disorder.
It is feasible that a true relationship exists between poor maternal adjustment and
difficulty with adherence to medication regimens for younger children. We examined the
relationship between maternal emotional distress variables (maternal depression and
maternal stress) and maternal adherence to anti-epileptic drugs (pill count percentage)
and child health status variables (seizure frequency, seizure-related emergency room
visits) for children diagnosed with epilepsy.

CHAPTER 2
AIMS AND STUDY JUSTIFICATIONS
Purpose
In summary, epilepsy is a chronic illness that typically begins in early childhood.
Poor adherence to the medication regimen can result in increased seizure frequency and
severity, and possibly, death. Young children are reliant on their parents, usually their
mothers, for their daily routine and functioning. This issue is particularly salient for
children with seizure disorders, who rely on their mothers to take their anti-convulsant
medication(s). It is well known that depression has a negative effect on daily
functioning. Additionally, depressive symptomatology and stress in mothers have been
associated with negative child health status and maternal adherence outcomes. Maternal
emotional factors such as depression and maternal stress may play an important role in
adherence to medication for younger children with epilepsy. To our knowledge, the
influence of maternal emotional distress on adherence to medication regimens of children
with epilepsy has not yet been examined.
An improved understanding of the impact of maternal emotional functioning on the
health status of children with epilepsy would enhance our ability to effectively intervene
with distressed mothers, and aid in developing interventions to prevent the possible
complications associated with poorer adherence, and improve overall psychological
functioning in both mother and child. In the current study, we examined the relationship
between maternal adherence to medication (pill count percentage) and child health status
(seizure frequency and seizure-related emergency room visits) for children with epilepsy
21

22
and the following factors: maternal emotional distress (maternal depression, maternal
stress), demographics (family annual income, race, maternal educational level),
condition-related and regimen-related variables (disease type, disease duration, type and
complexity of the medication regimen, concordance between prescribed regimen and
reported regimen), and maternal perceptions of the treatment regimen (perceived
efficacy, perceived knowledge, perceived difficulty).
Study Model
The purpose of the current study was to test a one-month predictive model of
maternal adherence and child health status to prescribed anti-epileptic medication in a
sample of 1-11 year old children diagnosed with epilepsy, focusing on the influence of
four predictor factors. The four predictor factors are defined as (1) maternal emotional
distress, (2) maternal perceptions of the treatment regimen, (3) condition-related and
regimen-related variables, and (4) demographic variables. The three outcome variables
are defined as (1) seizure frequency, (2) seizure-related emergency room visits, and (3)
pill count percentage, where (1) and (2) assess child health status and (3) assesses
maternal adherence. The current study extends earlier research in several ways. First, we
will integrate the four key sets of predictor factors identified in previous research in a
unifying predictive framework. Moreover, we strengthen the model by examining the
roles of the four, predictor factors at Time 1 (assessment) in predicting subsequent Time
2 (1-month follow-up) adherence and health status, controlling for Time 2 (1-month
follow-up) predictor factors (see Figure 1). This controls for the possibility that the
predictive role of adherence may be affected by maternal emotional distress, maternal
perceptions of the treatment regimen, condition-related and regimen-related variables,

23
and demographic variables at Time 2. The present study examined the following specific
hypotheses.
Hypotheses
Maternal Emotional Distress, Adherence, and Child Health Status
1. It was hypothesized that Time 1 maternal emotional distress would be negatively
related to Time 2 adherence to the child's medication regimen and positively with
child health status.
a. Time 1 maternal depression scores would be negatively associated with Time
2 adherence to the child's medication regimen and positively with health
status.
b. Time 1 maternal stress scores would be negatively associated with Time 2
adherence to the child's medication regimen and positively with health status.
Maternal Perceptions of the Treatment Regimen, Adherence, and Child Health
Status
2. We predicted that Time 1 ratings of maternal perceptions of the treatment regimen
(perceived knowledge, perceived difficulty, and perceived efficacy), and Time 1
maternal distress and Time 1 maternal perceptions of the treatment regimen would be
associated with Time 2 adherence to the child’s medication regimen and child health
status.
a. Maternal ratings of perceived knowledge regarding the child's disease and
medication regimen at Time 1 would be positively associated with Time 2
adherence to the child's medication regimen and negatively with health status.
b. Maternal ratings of perceived difficulty in giving the child the prescribed
medication at Time 1 would be negatively associated with Time 2 adherence
to the child's medication regimen and positively with health status.
c. Maternal ratings of perceived efficacy regarding the child's medication
regimen at Time 1 would be positively associated with Time 2 adherence to
the child's medication regimen and negatively with health status.
d. Maternal ratings of maternal distress and maternal perceptions of the
treatment regimen at Time 1 would be associated with Time 2 adherence to
the child's medication regimen and health status.
Condition and Regimen-related Variables, Adherence, and Child Health Status
3. We predicted that Time 1 condition and regimen-related variables would be
associated with Time 2 adherence to the child’s medication regimen and child health
status.
a. A tonic-clonic seizure diagnosis would be positively related to adherence to
the child's medication regimen and health status.
b. Time since diagnosis at Time 1 would be negatively associated with Time 2
adherence to the child’s medication regimen and positively with health status.

24
c. Complexity of the medication regimen (higher doses, multiple medications,
more frequent doses) at Time 1 would be negatively associated with Time 2
adherence to the child's medication regimen and positively with health status.
d. Percentage of concordance between the physician-prescribed medication
regimen and self-reported regimen at Time 1 would be positively associated
with Time 2 adherence to the child's medication regimen and negatively with
health status.
Demographic Characteristics, Adherence, and Child Health Status
4. We predicted that Time 1 demographic characteristics would be positively related to
Time 2 adherence to the child’s medication regimen and negatively with child health
status.
a. Total household income at Time 1 would be positively associated with Time 2
adherence to the child's medication regimen and negatively with health status.
b. Maternal educational level at Time 1 would be positively associated with
Time 2 adherence to the child's medication regimen and negatively with
health status.

CHAPTER 3
METHODS
Study Design
The current study is a prospective study using a time lag design, in which we
examined the influence of predictor variables at Time 1 on outcome variables at Time 2,
while controlling for predictor variables at Time 2 (see Figure 1).
T im e 1
T im e 2
Figure 1: Study Model
25

26
After obtaining informed consent, mothers were given a background questionnaire
(which included questions about condition and regimen-related variables), a
questionnaire assessing maternal stress, and a questionnaire assessing depressive
symptomatology to complete. The questionnaires were completed during a regularly
scheduled pediatric neurology clinic visit. Participants were asked to participate in a one-
month follow-up phone call. Approximately one month later, the investigator contacted
all participants at home by phone to re-assess condition and regimen specific-variables,
maternal perceptions of the treatment regimen, and maternal emotional distress. The one-
month delay in the assessment was designed to minimize sampling of temporary behavior
change that may have occurred immediately after a clinic visit and to re-assess maternal
ratings of emotional distress, condition and regimen-related measures, and maternal
perceptions regarding the treatment regimen, and their relationship to maternal adherence
to the child's medication regimen and child health outcomes.
Participants
Participants with a diagnosed seizure disorder were recruited from the Pediatric
Neurology clinic at Shands Hospital at the University of Florida's Health Science Center.
Based on data obtained from the pediatric neurologist, there are over 300 children with a
diagnosed seizure disorder between the ages of 1 and 11 who were eligible to participate
in the study. Young children are the focus of this study because: 1) epilepsy typically
begins to develop in early childhood, 2) mothers of younger children are more likely to
be highly involved in their child's medication regimen, and 3) one-third of epilepsies
spontaneously remit by adolescence.
Inclusionary criteria included: (1) An epilepsy diagnosis from the University of
Florida's pediatric neurologist which was noted in the patient's medical chart; (2) 1-11

27
years of age; (3) English-speaking; (4) Ability to read and write; (5) Telephone available
in the home. Exclusionary criteria included: (1) < 1 years of age or > 11 years of age; (2)
Non-English speaking; (3) Inability to read and write; (4) No telephone in the home.
A total of 82 participants completed all study measures. Data indicated the sample
consisted of 36 male and 46 female patients, aged 1-11 years (M = 5.76; SD = 2.93). The
mean education of mothers in the sample was 13.18 years (SD = 2.69). Mothers were
married (67.1%) and were homemakers (29.3%) or working full-time (35.4%). Average
family income was $44,450 (SD = 41,993). The ethnic makeup of the maternal sample
was 78.1% White, 11% African American, 6.1% Hispanic American, 2.4% Native
American, and 2.4% Other. The ethnic makeup of the child sample was 75.6% White,
9.8% African American, 7.3% Hispanic American, 2.4% Native American, and 4.9%
Biracial/Other. Approximately 93% of children in the sample were taking medication in
pill form, with seven percent taking medication in liquid form. Approximately 32% of
children in the study were diagnosed with partial seizures, 66% with generalized seizures,
and 2% with a diagnosis of other type. Study sample demographic characteristics are
presented in Table 1.
Table 1
Demographic and Medical Characteristics
Mean age of children in years (SD)
5.76 (2.93)
Sex of the child
Boys
36 (43.9%)
Girls
46 (56.1%)
Mean education of mothers in years (SD)
13.18(2.69)

28
Table 1 - Continued
Mother Marital Status
Single/ Never Married
Married
Separated/Divorced
15 (18.3%)
55 (67.1%)
12 (14.6%)
Mother Employment Status
Unemployed
Volunteer
Part-time
Full-time
Homemaker
Student
Other
13 (15.9%)
1 (1.2%)
8 (9.8%)
29 (35.4%)
24 (29.3%)
5 (6.0%)
2 (2.4%)
Mean annual family income in dollars (SD)
44,450(41,993)
Child race
White
African American
Hispanic American
Native American
Biracial/Other
62 (75.6%)
8 (9.8%)
6 (7.3%)
5 (2.4%)
3 (4.9%)
Mom race
White
African American
Hispanic American
Native American
Biracial/Other
64 (78.1%)
9(11%)
5 (6.1%)
2 (2.4%)
2 (2.4%)
Regimen Type
Pills
Liquid
76 (92.7%)
6 (7.3%)
Seizure Type
Partial
Generalized
Other
26 (31.7%)
54.9 (65.9%)
2 (2.4%)
Note, (n = 82).
Participants were recruited from the University of Florida’s Pediatric Neurology
Outpatient Clinic at Shands Hospital during the patient’s usual clinic visit. This clinic

29
treats over 300 children a year diagnosed with epilepsy. We offered the study to all
patients who met the study’s eligibility criteria.
The researcher and/or research assistants approached participants individually and
presented the rationale for the study. Parental informed consent and child assent was
obtained for children (if possible). Questions concerning subject obligation and
confidentiality were addressed at this time. After obtaining informed consent/assent,
participants completed a brief background questionnaire including questions regarding
maternal perceptions of the treatment regimen, the CES-D, and the PSI-SF. The entire
process took approximately 30-40 minutes.
The investigator recorded the following information from the participant’s clinic
visit or from the medical chart review conducted to gather other pertinent information:
disease-specific information including date of diagnosis, time since diagnosis, type of
diagnosis, type of medication, dose of medication, seizure frequency, blood serum levels,
pill counts, and emergency room visits. Approximately one month after the clinic visit,
the mother was contacted by phone to re-assess condition-related information, maternal
perceptions regarding the treatment regimen, and maternal emotional distress.
Measures
Participant Background Information
A background questionnaire was completed at the time of initial assessment in
order to gather the following variables: age of child, age of mother, disease type, age of
onset, time since initial diagnosis, race, mother's level of education, family income, and
marital status. Time since initial diagnosis was divided into two categories: (1) those
who were newly diagnosed, and (2) those with a diagnosis of epilepsy greater or equal to
one year prior to the study. Mothers were asked to describe their child's seizure

30
diagnosis, seizure frequency, and medication regimen. The type(s) of seizure disorder(s)
the mothers reported were noted. Maternal report of the child’s monthly seizure
frequency for the past month was also assessed. Furthermore, complexity of the
medication regimen reported by mothers was evaluated based on 4 indicators (i.e., type of
medication, number of medications per day, dosing, and timing). A complexity rating for
the medication regimen was calculated by multiplying the number of anti-epileptic
medications by the total number of times per day each medication is prescribed.
Maternal ratings of perceived difficulty (0 - not difficult to 10 - extremely difficult) and
perceived efficacy (0 - not good to 10 - extremely good) in carrying out the child's
medication regimen were assessed. Perceived difficulty ratings included items about
maternal perceptions of difficulty of medication type, medication dosage, and medication
timings. Perceived efficacy ratings included items about maternal perceptions of efficacy
of the child’s medications and efficacy of maternal management of the child’s medication
regimen. Mothers were asked to rate how they perceived their level of knowledge about
their child's medical condition. The knowledge ratings also ranged from 0 to 10 (0 - not
knowledgeable to 10 - extremely knowledgeable). Knowledge ratings included items
about maternal perceptions of knowledge of the child’s seizure disorder and medications
(type, timing, number, and dosage) (see Appendix A & Appendix B).
Medical Chart Information
Condition-specific information including date of diagnosis, time since diagnosis,
type of diagnosis, type of medication, dose of medication, seizure frequency, blood
levels, pill counts, and emergency room (ER) visits were collected by a review of the
medical chart. The child's medication regimen was noted from the chart (type of
medication, number of medications per day, dosage, and timing) and compared to the

31
mothers' report of the child's medication regimen. Concordance ratings were calculated
between physician prescribed anti-epileptic medication(s) and maternal reports of
prescribed medication(s) by determining the percentage correct of maternal reports
compared to physician reports (see Appendix C & Appendix D). Specifically, points for
concordance were assigned for a correct match between the prescribed and reported
regimen as follows: 50 (name), 30 (dose), and 20 (timing). More weight was given to the
name of the medication and dosage of the medication, with the rationale that if the name
and dosage of the medication is incorrect, the mother might not be appropriately
administering the correct medication and/or dosage.
Child Health Status and Adherence
Child health status included two components: current seizure frequency and
seizure-related ER visits. Current monthly seizure frequency was assessed by maternal
self-report and by chart review over the last month. Seizure-related ER visits were
identified by conducting a chart review. Chart reviews consisted of a systematic
examination of the child’s medical chart for the most recent data regarding child health
status. Written prescriptions and medication refill information were not noted
systematically in each medical chart and thus, this information was not available to be
collected.
Pill counts were assessed during the clinic visit if medication bottles were brought
to the clinic appointment. If mothers did not have the child’s medication bottle at the
clinic visit, they were contacted at home on the same day of the appointment in order to
report a current pill count. Mothers were telephoned at one-month follow-up and asked
to report a pill count. Maternal adherence to the child’s medical regimen was assessed by
percentage of pills used after one month [(Time 1-Time 2) / (Time 1)] x 100%.

32
Adherence was defined as an 80% pill count percentage. Blood serum levels were not
included as a measure of adherence because these values were not comparable across
different medication types; additionally, many participants had no blood serum levels
drawn.
Maternal Emotional Distress
The Center for Epidemiologic Studies - Depression (CES-D; Radloff, 1977) is a
self-report measure developed to be used in epidemiological studies of depression and
was designed to measure current levels of symptoms which accompany depression in a
general adult population. The CES-D is a 20-item self-report measure of depressive
symptoms. Each item provides a statement representing a symptom characteristic of
depression (e.g. "I had crying spells"), followed by a 4-point Likert-type response scale
ranging from "rarely or none of the time" (less than 1 day) to "most all of the time" (5-7
days). Participants were instructed to bubble in the number of each statement which best
described how often they felt or behaved that way during the past week. Sixteen of the
scales range from 1 to 4 and the remaining four scales range from 4 to 1.
Four major factors have been derived based on a principal components factor
analysis of data from the general population. These factors have been described as
depressed affect, positive affect, somatic and retarded activity, and an interpersonal
factor. Total scores were used for the purpose of the current study. The CES-D is scored
by summing the ratings of the 20 items and scores range from 0 to 60. Total severity is
calculated by reversing scores for items 4, 8, 12, and 16 (the items that control for
response bias). Higher scores indicate higher levels of depressive symptoms experienced
during the past week. Investigators have designated a cutoff score of 16 as a suitable
indicator to determine depressed from nondepressed patients (Nezu, 2000).

33
Regarding internal consistency, Cronbach's alpha is high across a variety of
populations and ranges from 0.85 for the general population and 0.90 for the clinical
population. Split-half reliability is also high, ranging from 0.77 to 0.92. Test-retest
reliability studies ranging over 2-8 weeks show moderate correlations (r = 0.51-0.67).
Studies examining effects between African Americans, Caucasians, and Mexican
Americans found no differences in measures of internal consistency reliability.
In samples of outpatients with depression, alcoholism, drug addiction, or
schizophrenia, correlation coefficients between CES-D scores and Symptom Checklist-90
Depression Subscale Scores were high, ranging from 0.73-0.89. Correlations with the
Hamilton Rating Scale for Depression and Raskin Scale ranged from 0.45 and 0.28 for
patients with acute depression to 0.85 and 0.79 for patients with schizophrenia,
respectively. Scores on the CES-D have been shown to decrease significantly following
treatment for depression. For major depression, sensitivity was 79.5% and specificity
was 71.1%. The positive predictive value was 27.9%. (American Psychiatric
Association, 2000).
The Parenting Stress Index - Short Form (PSI-SF; Abidin, 1995) is a 36-item short
form directly derived from the long form of the PSI. The PSI-SF was developed for use
in a variety of primary health care settings where it could be administered in less than 10
minutes. Participants were asked to respond to questions based on the five-point-scale.
The PSI-SF parent self-report scale contains three factor-analytically-derived
subscales (Parental Distress, Parent-Child Dysfunctional Interaction, and Difficult Child)
and a Total Stress Score. The Parental Distress (PD) scale determines the distress a
parent is experiencing in their role as a parent as a function of personal factors that are

34
directly related to parenting. The Parent-Child Dysfunctional Interaction (P-CDI)
subscale focuses on the parent's perception that their child does not meet their
expectations, and the interactions with their child are not reinforcing to them as a parent.
The Difficult Child (DC) subscale focuses on some of the basic behavioral characteristics
of children that make them either easy or difficult to manage. These characteristics can
include defiant, noncompliant, and demanding behavior. The Total Stress score is
designed to provide an indication of the overall level of parenting stress an individual is
experiencing. Parents who obtain a Total Stress score above a raw score of 90 are
experiencing clinically significant levels of parenting stress.
The three PSI-SF subscales differ from that of the longer PSI, but correlate in the
expected direction with the domains of the longer form. The PSI and PSI-SF total scores
are highly correlated with one another (0.94). The short form subscales have shown
Cronbach's alphas of 0.80 to 0.91 and 6-month test-retest reliabilities of 0.68 to 0.85.
Positive predictive values of the PSI-SF range from 25% to 45%.

CHAPTER 4
RESULTS
The overall findings of the data analyses are presented as follows: 1) descriptive
statistics of the predictor and criterion variables included in the conceptual model, 2)
internal consistency reliabilities of the questionnaire measures and the distributional
properties of key predictor and criterion variables, 3) correlations between the
background variables not included in the conceptual model and those between key
predictor variables and change scores on the criterion variables, 4) regression analyses
that control for the effects of Time 2 predictors on Time 2 criterion variables, and 5)
exploratory model testing to identify goodness of fit of the specific regression models.
Due to small sample size, structural equation modeling (SEM) was not conducted
as anticipated. The statistical approach shifted from SEM to variable-level model testing,
in which the goodness of fit of variables located under the four primary factors of the
conceptual framework was assessed (see Figure 1).
Preliminary Analyses
Medical characteristics of the child with epilepsy were variables collected at Time
1, except for seizure frequency. The latter was also collected at Time 2, since this time
point was of main interest in our primary analyses. Study sample medical characteristics
are displayed in Table 2.
35

36
Table 2
Mean Scores and Standard Deviations for Child Medical Characteristics
Measures
M(SD)
Complexity of the regimen (uncorrected)
8.7(13.1)
Complexity of the regimen (corrected)
1.13(0.5)
Time Since Diagnosis in years
4.0 (2.8)
Concordance between reported and prescribed regimen-Time 2
72.8 (32.7)
Seizure Frequency-Time 2 (uncorrected)
50.0 (87.2)
Seizure Frequency-Time 2 (corrected)
2.5 (1.1)
Note, (n = 82). Corrected scores are corrected for skewness to ensure analyses integrity.
Means and standard deviations for maternal self-reports of maternal stress
(measured by the PSI-SF), depression (measured by the CES-D), and maternal
perceptions of the treatment regimen (measured by maternal reports of difficulty,
efficacy, and knowledge) for Time 1 and Time 2 are presented in Table 3 below.
Table 3
Mean Scores and Standard Deviations for Study Measures at Time 1 and Time 2
Measures
M(SD)
Time 1
M (SD)
Time 2
PSI-SF, Total score
85.2 (28.5)
65.5 (8.9)
PSI-SF, Parental distress
27.5 (10.5)
24.5 (3.5)
PSI-SF, Parent-child dysfunction
25.9 (9.9)
21.6 (5.1)
PSI-SF, Difficult child
31.8(11.5)
19.4 (3.7)
CES-D, Severity score
7.8 (3.3)
12.4 (2.2)
Difficulty, Total score uncorrected
6.0 (5.8)
12.9 (4.5)
Difficulty, corrected
2.1 (1.5)
1.5 (0.8)
Efficacy, Total score
14.3 (4.5)
11.0 (3.5)
Knowledge, Total score
39.4(13.84)
39.5 (13.82)
Note. PSI-SF = Parenting Stress Index-Short Form; CES-D = Center for Epidemiologic
Studies-Depression. Corrected scores are corrected for skewness to ensure analyses
integrity.
Cronbach’s alphas were calculated to assess the reliability (i.e. internal consistency)
of mother-completed self-report measures. Coefficients ranged from .68 to .96 for the
total sample at Time 1, suggesting adequate internal consistency reliability for all

37
measures used. Coefficients ranged from -.002 to .58 for the total sample at Time 2,
suggesting lower levels of internal consistency reliability for all measures used compared
to Time 1. Time 1 and Time 2 reliability statistics are presented in Table 4 below.
Table 4
Alpha Coefficients of Parent-completed Measures for Total Sample at Time 1 and Time 2
Measure
N
a
Time 1
a
Time 2
PSI-SF, Total score
82
.96
.56
CES-D, Total score
82
.94
.35
Difficulty, Total score
82
.86
-.18
Efficacy, Total score
82
.68
-.002
Knowledge, Total score
82
.92
.08
Note. PSI-SF = Parenting Stress Index-Short Form; CES-D = Center for Epidemiologic
Studies-Depression.
We also analyzed the distribution properties of demographic, medical, predictor,
and outcome variables. Skewed distributions were corrected with the creation of new
variables. Annual family income was highly positively skewed and ranged from $1500
to $250,000 and was recoded into three income categories with a value of one
representing poverty level of $1500 to $18,000. a value of two representing middle level
of $18,001 to $116,000, and a value of three representing an upper middle level of
$116,001 to $250,000. The poverty level cutoff was determined using the United States
Census Bureau poverty guidelines (United States Census Bureau, 2001). The means and
standard deviations for uncorrected and corrected annual income were $44,450 (41,993)
and 1.2 (.62), respectively. The original scores for complexity of the regimen, seizure
frequency, and maternal perceptions of regimen difficulty were also highly positively
skewed. In order to correct for data skewness, the data was collapsed into groups that
were equally distributed, creating new complexity, seizure frequency, and difficulty

38
variables. Original complexity scores ranged from 1 to 80 and were recoded into three
categories, with one representing a complexity score of 1 to 26.7, two representing a
complexity score of 26.8 to 53.4, and three representing a complexity score of 53.5 to 80.
Means and standard deviations for uncorrected and corrected complexity scores were 8.7
(13.1) and 1.13 (0.5), respectively. Original seizure frequency scores ranged from 0 to
320 and were recoded into four categories with equal distributions of the sample in each
group, with one representing 0 seizures, two representing 1 to 7 seizures, three
representing 8 to 60 seizures, and four representing 61 to 320 seizures. The means and
standard deviations for uncorrected and corrected seizure frequency scores were 50.0
(87.2) and 2.5 (1.1), respectively. Difficulty scores were positively skewed, ranged from
0 to 24, and were recoded into four categories, with one representing a score of 0 to 6,
two representing a score of 6.1 to 12, three representing a score of 12.1 to 18, and four
representing a score of 18.1 to 24. The means and standard deviations for uncorrected
and corrected difficulty scores were 6.0 (5.8) and 2.1 (1.5) for Time 1 and 12.9 (4.5) and
1.5 (0.8) for Time 2, respectively.
Primary Analyses
Correlations among Background Variables, Predictors, and Criterion Variables
Correlational analyses were conducted between background variables not included
in the model (child sex, marital status, number of years married, employment status,
number of children in the home) and Time 1 key predictor variables of the following
factors: maternal emotional distress, maternal perceptions of the regimen, condition and
regimen-related variables, and demographic variables. No significant associations were
found between the background variables not included in the model and Time 1 predictors

39
(p > .05). No significant associations were found between the background variables not
included in the model and criterion variable change scores (see Table 5 below).
Table 5
Correlations between Background Variables not in the Model and Specific Time 2
Outcomes Change Scores
Background Factors
Seizure
Frequency
Seizure-related ER
Visits
Child sex
,160+
i
o
o
On
+
Marital status
-,027+
.083+
Years married
,085+
.060+
Employment status
-,098+
-,072+
Number of children in home
,062+
-,015+
viote. +p > .05.
Correlational and hierarchical regression analyses were conducted to examine the
relationships between Time 1 predictors (maternal emotional distress, maternal
perceptions of the treatment regimen, condition and regimen-related factors, and
demographic factors), and Time 2 outcome measures (seizure frequency, emergency
room visits, and pill count percentages). Note that the effects of Time 2 predictor
measures were partialed out in regression analyses to control for Time 2 predictor
variable effects on the criterion measures. The order of entry into the regression analyses
were as follows: (1) Time 2 predictors were entered in a block in a forced-entry fashion,
(2) Time 1 predictors were entered in blocks in a stepwise fashion. Note that Time 2
predictor measures were not included in correlational analyses due to low internal
consistency reliability. New variables that were created as corrections for data skewness
were used in all subsequent analyses; analyses conducted with skewed data did not reveal
any significant relationships. Partial correlation and regressions analyses were used in
determining which predictors to enter into the final hierarchical regression models.

40
Hypothesis 1: Maternal Emotional Distress, Adherence, and Child Health Status
We hypothesized that Time 1 maternal emotional distress (measured by the PSI-SF
and CES-D) would be negatively related to Time 2 adherence to the child's medication
regimen (pill count percentage) and positively associated with child health status (seizure
frequency and seizure-related emergency room visits). Specifically, it was predicted that
Time 1 maternal stress and depression scores would be negatively associated with Time 2
adherence to the child's medication regimen and that Time 1 ratings of maternal stress
and depressive affect would be positively associated with the child’s health status. These
hypotheses were partially supported. Hierarchical regression analyses were conducted to
assess the association between Time 2 outcomes and Time 1 maternal emotional distress
variables, while controlling for the effects of Time 2 maternal emotional distress
variables. Time 1 PSI-SF overall mean scores and PSI-SF parent-child dysfunction
subscale scores significantly predicted Time 2 seizure frequency, while controlling for
Time 2 predictor variables. No substantial associations were found between Time 1
ratings of maternal stress and Time 2 seizure-related emergency room visits or Time 2
maternal adherence. No significant relationships were found between Time 1 ratings of
maternal depression and child health status or maternal adherence. Next, Time 1
maternal emotional distress variables did not predict Time 2 seizure-related emergency
room visits or pill count percentages, while controlling for Time 2 maternal emotional
distress. The results of the two sets of regression analyses for Time 1 maternal emotional
distress variables on Time 2 outcomes are presented in Tables 6 and 7.

41
Table 6
Summary of Hierarchical Regression Analyses for Time 1 PSI-SF Overall Score
Predicting Child Seizure Frequency (N = 82)
Variables Included
P t
R
R2 Adj. R2
AR2
F
AF
Controlling for T2
CES-D mean T2
PSI-SF mean T2
.055 .466
-.093 -.784
.091
.008 -.017
.008
.332
.332
Model 1
CES-D mean T2
PSI-SF mean T2
PSI-SF mean T1
.090 .773
-.121 -1.04
.239 2.16*
.253
.064 .028
.056
1.78
4.64*
Note. CES-D = Center for Epidemiologic Studies-Depression; PSI-SF
Stress Index-Short Form; T1 = Time 1; T2 = Time 2.
*P < .05
= Parenting
Table 7
Summary of Hierarchical Regression Analyses for Time 1 PSI-SF Subscale Variables
Predicting Child Seizure Frequency (N = 82)
Variables Included
P t
R
R2 Adj. R2
AR2
F
AF
Controlling for T2
PD mean T2
P-C dys T2
DC T2
-.048 .466
-.007 -.784
-.059
.084
.007 -.031
.007
.183
.183
Model 1
PD mean T2
P-C dys T2
DC T2
P-C dys T1
-.060 -.535
-.015 -.124
-.078 -.629
.267 2.42*
.278
.077 .029
.070
1.62
5.88*
Note. PSI-SF = Parenting Stress Index-Short Form; PD = Parental Distress; DC =
Difficult Child; P-C dys. = Parent-Child Dysfunction; T1 = Time 1; T2 = Time 2.
*2 < .05
The data revealed significant relationships (partial correlations) between Time 1
PSI-SF overall mean scores and Time 2 seizure frequency (r = .24, p < .05), as well as
between Time 1 PSI-SF parent-child dysfunction subscale scores and Time 2 seizure

42
frequency (r = .27, £ < .05). As hypothesized, overall Time 1 maternal stress levels were
positively related to Time 2 child seizure. Further analysis of the PSI-SF subscales
indicated that Time 1 PSI-SF parent-child dysfunction was positively related to Time 2
child seizure frequency. As maternal stress levels and parent-child dysfunction levels
increased, seizure frequency concomitantly increased. Contrary to prediction, no other
significant correlational relationships were found with Time 1 maternal emotional
distress and Time 2 outcomes. Correlational data between Time 1 maternal emotional
distress, Time 2 maternal adherence, and Time 2 child health status are presented in
Table 8 below.
Table 8
Partial Correlations between Time 1 Maternal Emotional Distress and Time 2 Outcomes
Health Status
Adherence
Maternal Emotional Distress
Seizure
Seizure-related ER
Pill Count
Frequency
Visits
Percentage
PSI-SF Total
.237*
.061
-.104
PSI-SF, Parental Distress
.014
-.026
-.045
PSI-SF, Parent-child dysfunction
.266*
.056
-.165
PSI-SF, Difficult child
-.096
.025
-.078
CES-D Total
.045
.101
-.099
Note. PSI-SF = Parenting Stress Index; CES-D = Center for Epidemiologic Studies-
Depression.
*P < .05.
Hypothesis 2: Maternal Perceptions of the Treatment Regimen, Adherence and
Child Health Status
We predicted that Time 1 ratings of maternal perceptions of the treatment regimen
(measured by perceived knowledge, perceived difficulty, and perceived efficacy), and
Time 1 maternal distress (measured by the PSI-SF) and Time 1 maternal perceptions of
the treatment regimen would be associated with Time 2 adherence to the child’s

43
medication regimen (pill count percentage) and child health status (seizure frequency and
seizure-related emergency room visits). Specifically, maternal ratings of perceived
knowledge and efficacy regarding their child’s medical condition and medication
regimen at Time 1 would be positively associated with Time 2 adherence to the child's
medication regimen and negatively with health status. In addition, maternal ratings of
perceived difficulty in giving the child the prescribed medication at Time 1 would be
negatively associated with Time 2 adherence to the child's medication regimen and
positively associated with child health status, and last, maternal ratings of maternal
distress and maternal perceptions of the treatment regimen at Time 1 would be associated
with Time 2 adherence to the child's medication regimen and health status. These
hypotheses received mixed support.
Time 2 outcomes of child health status and maternal adherence were regressed on
Time 1 maternal perceptions of the treatment regimen variables, while controlling for the
effects of Time 2 maternal perceptions of the treatment regimen predictor variables.
Time 1 difficulty and efficacy scores successfully predicted Time 2 seizure frequency,
while controlling for Time 2 difficulty and efficacy predictor variables. Time 1 difficulty
scores also successfully predicted Time 2 seizure-related emergency room visits, while
controlling for Time 2 difficulty predictor variables. No other significant relationships
were found between Time 1 ratings of difficulty, efficacy, and knowledge and Time 2
outcomes of child health status and maternal adherence. Next, Time 1 maternal
perceptions of the treatment regimen variables were not successful in predicting Time 2
pill count percentages, while controlling for Time 2 maternal perceptions of the treatment
regimen. The findings of the two sets of regression analyses for Time 1 maternal

44
perceptions of the treatment regimen variables on Time 2 outcomes are presented in
Tables 9 and 10.
Table 9
Summary of Hierarchical Regression Analyses for Time 1 Maternal Perceptions of the
Treatment Regimen Predicting Child Seizure Frequency (N = 82)
Variables Included
P
t
R
R2
Adj.R2
AR2
F
AF
Controlling for T2
Difficulty T2
-.025
-.229
.222
.049
.013
.049
1.35
1.35
Knowledge T2
.226
2.01*
Efficacy T2
-.046
-.406
Model 1
Difficulty T2
-.034
-.312
.334
.111
.065
.062
2.41
5.38*
Knowledge T2
.212
1.93
Efficacy T2
-.022
-.196
Difficulty T1
.250
2.32*
.459
.211
.159
.100
4.07
9.62**
Model 2
Difficulty T2
-.023
-.226
Knowledge T2
.406
3.35**
Efficacy T2
-.043
-.411
Difficulty T1
.172
1.63
Efficacy T1
-.376
-3.10**
Note. T1 = Time 1; T2 = Time 2.
*P < .05 **p<.01
Table 10
Summary of Hierarchical Regression Analyses for Time 1 Maternal Perceptions of the
Treatment Regimen Predicting Seizure-related Emergency Room Visits (N = 82)
Variables Included
P
t
R R2
Adj.R2 AR2 F AF
Controlling for T2
.227 .052
.015 .052 1.41 1.41
Difficulty T2
.060
.541
Knowledge T2
.211
1.87
Efficacy T2
-.110
-.972

45
Table 10 - Continued
Model 1
Difficulty T2
.051
.472
.365 .134
.089
.082
2.97
7.29**
Knowledge T2
.195
1.80
Efficacy T2
-.082
-.750
Difficulty T1
.288
2.70**
Model 2
Difficulty T2
.054
.449
.377 .142
.085
.008
2.51
.729
Knowledge T2
.251
1.98
Efficacy T2
-.088
-.804
Difficulty T1
.265
2.41*
Efficacy T1
-.108
-.854
Note. T1 = Time 1; T2 = Time 2.
*£<.05 **£<.01
The data revealed significant relationships (partial correlations) between Time 1
difficulty scores and Time 2 seizure frequency (r = .26, p < .05), Time 1 efficacy scores
and Time 2 seizure frequency (r = -.34, p < .01), and Time 1 difficulty scores and Time 2
seizure-related emergency room visits (r = .29, p < .01). As hypothesized, Time 1
maternal perceptions of regimen difficulty were positively related to Time 2 child seizure
frequency and emergency room visits. As maternal perceptions of regimen difficulty
significantly increased, seizure frequency and seizure-related emergency room visits also
significantly increased. Time 1 maternal perceptions of regimen efficacy were negatively
related to Time 2 seizure frequency in that as efficacy significantly increased, seizure
frequency significantly decreased. Contrary to study hypotheses, no other significant
correlational relationships were found with Time 1 maternal perceptions of the treatment
regimen and Time 2 outcomes. Correlational data between Time 1 maternal perceptions
of the treatment regimen, Time 2 maternal adherence, and Time 2 child health status are
presented in Table 11.

46
Table 11
Partial Correlations between Time 1 Maternal Perceptions of the Treatment Regimen
and Time 2 Outcomes
Health Status Adherence
Maternal Perceptions of the
Treatment Regimen
Seizure
Frequency
Seizure-related ER
Visits
Pill Count
Percentage
Difficulty
.255*
.294**
.037
Efficacy
-.335**
-.097
-.024
Knowledge
.215
.194
.074
Note. *p < .05 **p<.01
Hypothesis 3: Condition and Regimen-related Variables, Adherence, and Child
Health Status
We predicted that condition and regimen-related variables (complexity, time since
diagnosis, regimen type, disease type, and concordance between the prescribed and
reported regimen) would be associated with Time 2 adherence to the child’s medication
regimen (pill count percentage) and child health status (seizure frequency and seizure-
related emergency room visits. First, we hypothesized that a tonic-clonic seizure
diagnosis would be positively related to adherence to the child's medication regimen.
Second, we predicted that child health status and time since diagnosis and complexity of
the medication regimen (higher doses, multiple medications, more frequent doses) would
be negatively associated with Time 2 adherence to the child’s medication regimen and
positively with health status. Third, we predicted that percentage of concordance
between the physician-prescribed medication regimen and self-reported regimen would
be positively associated with Time 2 adherence to the child's medication regimen and
negatively with health status. These hypotheses were partially confirmed. Hierarchical
regression analyses were conducted to predict Time 2 outcomes with condition and

47
regimen-related variables. Concordance scores successfully predicted Time 2 seizure-
related emergency room visits and Time 2 pill count percentage. No other significant
predictive relationships were obtained. Furthermore, condition and regimen-related
factors were not successful in predicting Time 2 seizure frequency. Regression data for
condition and regimen-related variables on Time 2 outcomes are presented in Tables 12
and 13.
Table 12
Summary of Hierarchical Regression Analyses for Condition and Regimen-related
Variables Predicting Seizure-related Emergency Room Visits (N = 82)
Variables Included
P
t
R
R2
Adj. R2
AR2
F
AF
Model 1
Seizure type
-.074
-.653
.076
.006
-.020
.006
.222
.222
Regimen type
.019
.163
Model 2
Seizure type
.015
.140
.408
.166
.110
.161
2.96
4.76**
Regimen type
-.035
-.324
Concordance
-.219
-1.97*
Complexity
.210 .
1.80
Time since dx.
.177
1.57
Note, dx = diagnosis.
*P<.05 **p<.01
Table 13
Summary of Hierarchical Regression Analyses for Condition and Regimen-related
Variables Predicting Pill Count Percentage (N
= 82)
Variables Included
P
t
R
R2
Adj. R2
AR2
F
AF
Model 1
Seizure type
-.074
-.653
.122
.015
-.011
.015
.586
.586
Regimen type
.019
.163

48
Table 13 - Continued
Model 2
Seizure type
Regimen type
Concordance
Complexity
Time since dx
Note, dx = diagnosis.
*g < .05 **p<.01
The data revealed significant relationships (partial correlations) between
concordance scores and Time 2 seizure-related emergency room visits (r = -.22, g < .05),
as well as concordance scores and Time 2 pill count percentage (r = .69, g < .01). As
hypothesized, concordance scores were negatively related to Time 2 seizure-related
emergency room visits and positively related to Time 2 pill count percentage. As
concordance scores significantly increased, emergency room visits significantly
decreased and pill count percentages significantly increased. Contrary to study
hypotheses, a tonic-clonic seizure diagnosis was not associated with better adherence,
increased seizure frequency, or emergency room visits. No other significant relationships
were found with condition and regimen-related variables and Time 2 outcomes.
Correlational data between condition and regimen-related variables, Time 2 maternal
adherence, and Time 2 child health status are presented in Table 14.
Table 14
Partial Correlations between Time 1 Condition and Regimen-related Variables and Time
2 Outcomes
.755 .570 .541 .555 19.66 31.9**
.015 .140
-.035 -.324
-.219 -1.97*
.210 1.80
.177 1.57
Health Status
Adherence
Condition3 and Regimen”
Seizure Seizure-related ER
Pill Count
Variables
Frequency Visits
Percentage
Complexity”
.098 .205
-.108

49
Table 14 - Continued
Time Since Diagnosis3
.157
.180
-.076
Regimen Typeb
.089
-.038
.027
Disease Type3
.114
-.016
-.083
Concordanceb
-.160
-.223*
.691**
Note. *2<.05 **p<.01
Hypothesis 4: Demographic Characteristics, Adherence, and Child Health Status
We predicted that demographic characteristics (total annual income, maternal
education level) would be positively related to Time 2 adherence to the child's medication
regimen (pill count percentage) and negatively associated with child health status (seizure
frequency and seizure-related emergency room visits). Specifically, total household
annual income and maternal education level would be positively associated with Time 2
adherence to the child's medication regimen and negatively associated with child health
status. These hypotheses were partially supported. Hierarchical regression analyses were
conducted to predict Time 2 outcomes with demographic variables. Family income
approached significance (i.e., trend in the predicted direction) for predicting Time 2 pill
count percentage. No other significant associations were found. Demographic variables
were not successful in predicting Time 2 seizure frequency or seizure-related emergency
room visits. Regression data for demographic variables on Time 2 outcomes is presented
in Table 15.

50
Table 15
Summary of Hierarchical Regression Analyses for Demographic Variables Predicting
Pill Count Percentage (N = 82)
Variables (3 t R R2 Adj. R2 AR2 F AF E
Included
Model 1 343 Ü8 364 .118 2.21 2.21 .078+
Family income .237 2.04 .058+
Mom education .132 1.13
Child race -.283 -1.17
Mom race .093 .388
Note. +e > .05
The data revealed a significant relationship (partial correlation) between annual
family income and Time 2 pill count percentage (r = .24, p < -05). As hypothesized,
annual family income was positively related to Time 2 pill count percentage in that as
family income increased, pill count percentages increased. Contrary to study hypotheses,
no other significant correlational relationships were found between demographic
variables and Time 2 child health status or maternal adherence outcomes. Regression
analyses did not identify any additional variables that significantly predicted Time 2
outcomes. Correlational data between demographic variables, Time 2 maternal
adherence, and Time 2 child health status is presented in Table 16 below.
Table 16
Partial Correlations between Demographic Variables and Time 2 Outcomes
Health Status
Adherence
Demographic Factors
Seizure
Seizure-related ER
Pill Count
Frequency
Visits
Percentage
Total Annual Income
.088
.089
.244*
Mother's Education
-.057
-.101
.138
Child Race
.037
-.130
-.143
Mother Race
-.025
-.100
-.048
Note. *p < -05

51
Model Testing
Predictor variables for Time 2 outcomes were included based on the analyses
described above by selecting variables that showed both significant P weights and
accounted for significant R2 change. These variables were subsequently included in tests
of the overall goodness of fit to identify promising alternative regression models for use
in follow-up research (see Table 17). The statistical approach included two steps: (1)
Time 2 measures were entered in step one of the hierarchical regression analyses in a
forced-entry fashion and (2) Time 1 variables were entered in a stepwise fashion with
variables included in the model if p < .05.
Table 17
Identified Predictors of Time 2 Outcomes
Seizure Frequency
Seizure-related ER Visits
Pill Count Percentage
PSI-SF overall mean
Difficulty
Concordance
PSI-SF parent-child dysfunction
Concordance
Annual family income
Difficulty
Efficacy
Prediction of adherence
Hierarchical regression analyses were performed to test the hypotheses that the
variables of concordance and annual family income predicted Time 2 maternal adherence
(measured by pill count percentage). Concordance and annual family income were
entered in a stepwise manner with Time 2 pill count percentage as the outcome variable.
Only one model was generated, with the concordance variable retained as a significant
predictor. The concordance variable accounted for a significant proportion of the
variance in the Time 2 pill count percentage outcome measure (Adjusted R2 = .44; p <
.001). The results suggested that the best predictor of maternal adherence was
concordance between mother and physician on the child’s drug regimen. This measure

52
accounted for approximately 44% of the variance in the Time 2 pill count percentage
measure. In examining the significance of (3 weights for all predictor variables in the
model, only the concordance variable was significant (P = .67; t = 7.4; p = .000) (see
Table 18).
Table 18
Summary of Regression Model for Predicting Maternal Adherence (N = 82)
Variables Included
P
t R
R2
Adj.R2
AR2
F
AF
Model 1**
Concordance
.665
.665
7.4*
.443
.435
.443
54.8
54.8*
Note. *p < .01 **Best prediction model
Prediction of health status
Hierarchical regression analyses were performed to test the hypotheses that the
variables of maternal stress (measured by the PSI-SF), parent-child dysfunction
(measured by the subscale of the PSI-SF), and maternal perceptions of difficulty and
efficacy regarding the child's medication regimen successfully predicted Time 2 seizure
frequency. The PSI-SF overall score, parent-child dysfunction score, difficulty score,
and efficacy score for Time 2 were entered as a block in a forced-entry fashion to partial
out Time 2 effects on seizure frequency. The PSI-SF overall score, parent-child
dysfunction score, difficulty score, and efficacy score for Time 1 were then entered in a
stepwise fashion with Time 2 seizure frequency as the outcome variable. One model was
generated in which the Time 1 parent-child dysfunction subscale was a predictor. Parent-
child dysfunction accounted for a small but significant proportion of the variance in the
Time 2 seizure frequency outcome measure (Adjusted R2 = .022; p = .015). The results
suggested that the best predictor of Time 2 seizure frequency was parent-child

53
dysfunction, which accounted for approximately 2.2% of the variance in the Time 2
seizure frequency measure. In examining the significance of P weights for all predictor
variables in the model, only the parent-child dysfunction variable was significant (P =
.29, t = 2.5, p = .015) (see Table 19).
Table 19
Summary of Regression Models for Predicting Seizure Frequency (N = 82)
Variables Included
P
t
R
R2
Adj.R2
AR2
F
AF
Controlling for T2
PSI-SF mean T2
-.130
-.610
.084
.007
-.045
.007
.135
.135
P-C dys. T2
.065
.306
Difficulty T2
-.001
-.010
Efficacy T2
-.006
-.049
Model 1**
PSI-SF mean T2
-.166
-.802
.287
.082
.022
.075
1.37
6.25*
P-C dys. T2
.087
.421
Difficulty T2
-.057
-.497
Efficacy T2
.056
-.489
P-C dys. T1
.287
2.50*
Note. PSI-SF = Parenting Stress Index; P-C dys. = Parent-Child Dysfunction; T1 =
Time 1; T2 = Time 2.
*P < .05 **Best prediction model
Hierarchical regression analyses were performed to test the hypotheses that the
variables of difficulty regarding the child's medication regimen and concordance between
the prescribed and reported regimen successfully predicted seizure-related emergency
room visits. The difficulty score for Time 2 was entered first to partial out Time 2 effects
on seizure-related emergency room visits. The difficulty score for Time 1 and
concordance were then entered as a block in a stepwise fashion with seizure-related
emergency room visits as the outcome variable. This resulted in the generation of two
models, the first of which included the difficulty variable as a predictor and the second

54
added concordance as a predictor. The first model accounted for a significant proportion
of the variance in the emergency room visits outcome measure (Adjusted R2 = .072; p =
.006), with 7.2% of the variance accounted for by this block. The second model
accounted for a significant proportion of the variance in the emergency room visits
beyond model one (Adjusted R2 = .146; p = .007), with an additional 7.4% of the
variance accounted for by this model. The results suggested that the best model for
predicting seizure-related emergency room visits was Model 2, accounting for
approximately 15% of the variance in seizure-related emergency room visits. In
examining the significance of p weights for all factors in Model 2, two betas, difficulty (P
= .32, t = 3.1, p = .002) and concordance (P = -.293, t = -2.8, p = .007), were statistically
significant (see Table 20).
Table 20
Summary of Regression Models for Predicting Seizure-related Emergency Room Visits (N
= 82)
Variables
Included
P
t R
R2
Adj.R2
AR2
F
AF
Controlling for T2
.062
.004
-.009
.004
.307
.307
Difficulty T2
.062
.554
Model 1
.308
.095
.072
.091
4.15
7.96*
Difficulty T2
.054
.505
Difficulty T1
.302
2.82*
Model 2**
.421
.177
.146
.082
5.60
7.80*
Difficulty T2
-.003
-.026
Difficulty T1
.322
3.13*
Concordance
-.293
-2.79*
Note. T1 = Time 1; T2 = Time 2.
*P < .01 * * Best prediction model

55
Summary of Findings
Overall, we found mixed confirmation for the hypothesized relationships among
the predictor variables, health status, and adherence criterion variables.
First, the maternal stress measure of maternal emotional distress was found to be
significantly related to the child health status variable of seizure frequency. Maternal
stress and the subscale of parent-child dysfunction were significantly related to seizure
frequency and were significant predictors of seizure frequency in hypothesis testing.
Parent-child dysfunction was identified as a significant predictor of seizure frequency in
exploratory model testing. Note also that maternal emotional distress on measures of the
traditional psychological constructs of depression and maternal stress were not clinically
elevated in the study sample.
Second, the difficulty and efficacy variables of maternal perceptions of the
treatment regimen were identified as being significantly related to child health status
variables. Difficulty was significantly related to seizure frequency and seizure-related
emergency room visits. Difficulty was found to be a significant predictor of both seizure
frequency and seizure-related emergency room visits in hypothesis testing, while efficacy
was only a significant predictor of seizure frequency in hypothesis testing. Difficulty was
also identified as a significant predictor of seizure-related emergency room visits in
exploratory model testing.
Third, the concordance measure of condition and regimen-related variables were
identified as significantly related to the child health status variable of seizure-related
emergency room visits and maternal adherence. Concordance was found to be a
significant predictor of maternal adherence and seizure-related emergency room visits in
hypothesis testing, as well as in exploratory model testing.

56
Fourth, the demographic factor of annual family income was identified as being
related to maternal adherence. Annual family income was identified as a significant
predictor of maternal adherence in hypothesis testing. However, no significant predictive
relationships were identified between annual family income and outcomes in exploratory
model testing.
Finally, findings related to the predictive validity of Time 1 measures of maternal
emotional distress, maternal perceptions of the treatment regimen, condition and
regimen-related factors, and demographics on Time 2 child health status (seizure
frequency and seizure-related emergency room visits) and maternal adherence (pill count
percentage) outcomes suggest that several measures (PSI-SF parent-child dysfunction,
difficulty, concordance, and annual family income) play an important role in child health
status and maternal adherence outcomes.

CHAPTER 5
DISCUSSION
Sample Characteristics of Demographics and Maternal Distress
In regards to the study sample, mothers were highly educated, having attended an
average of one year of college, with their annual family income falling in the middle
income range ($18,001 to $116,000). The racial make-up of mothers completing this
study was predominately White (78.1%), with traditionally labeled minorities (e.g.,
African American, Hispanic, etc.) showing a fairly uniform distribution pattern. Two-
thirds of the mothers were married and nearly half worked outside the home at least part-
time. The racial profile of patients in this study is representative of the reported
distribution of children with epilepsy living in the United States (Adams, Hendershot, &
Maraño, 1996). Patients were predominately White (75.6%), girls (56.1%), and between
five and six years of age. The racial heterogeneity observed in the current study suggests
that its findings are generalizable to the overall epilepsy population within the United
States.
Maternal stress and depression scores were below the cut-off for clinically
significant levels of 90 and 16, respectively, indicating that mothers were not reporting
heightened levels of maternal stress or depressive symptoms at Time 1 or Time 2. It was
surprising that the study sample did not meet criteria for significant levels of maternal
stress or depressive symptomatology since mothers in previous studies reported higher
levels of depressive symptomatology (Kyngas, 2000) and maternal stress (Spector et al.,
1999). One possible explanation for this finding may be that mothers of children with
57

58
epilepsy in this study may have experienced more stigma and emotional strain, as well as
daily pressures (e.g., frequent doctor’s appointments) when the child was much younger
during the initial months following the child’s diagnosis. Perhaps maternal stress and
depression scores were not significant because, compared to when the child was first
diagnosed, the current situation is a calmer and less stressful time for mothers. Studies
indicate that the initial phase following the child’s diagnosis leads to negative emotional
consequences (e.g., increased levels of stress and depressive symptomatology) for
mothers of children with chronic illnesses (Brownbridge & Fielding, 1994; Austin, 1989).
It is also possible that only mothers of children who experienced low levels of stress or
depressive symptomatology opted to participate in the study.
Hypotheses
Hypothesis 1: Maternal Emotional Distress, Adherence, and Child Health Status
It was hypothesized that Time 1 maternal emotional distress variables would be
negatively related to Time 2 adherence to the child's medication regimen and positively
with child health status outcome measures. Overall, mixed support was obtained for the
study hypotheses. Hypothesis 1 was partially supported in that mothers who reported
high levels of maternal stress at Time 1, also reported higher seizure frequency levels for
their children at Time 2. No significant relationships were found between maternal
distress measures and seizure-related emergency room visits. Maternal emotional
distress variables were not significantly related to maternal adherence.
The maternal stress measure and parent-child dysfunction subscale scores on the
PSI-SF were positively related to seizure frequency of the child health status factor.
Mothers who reported higher stress levels and more dysfunctional interactions with their
children also reported increased seizure frequency at Time 2 for their children. The

59
results are consistent with previous findings that increased parental stress levels are
related to an increase in child seizure frequency and severity (Spector et al., 1999).
Spector et al. (1999) investigated the efficacy of a range of psychological interventions
and found that short-term group interventions combining supportive and cognitive-
behavioral components to improve coping strategies and decrease stress may be effective
for improving seizure control,
These findings may be attributable to the idea that mothers who have children with
increased levels of seizure frequency may be more stressed about their child's health and
well-being, which may carry over into increased levels of overall maternal stress. Of
note is that increased patient stress levels have been associated with increased seizure
frequency (Kyngas, 1990). Kyngas (1990) studied the compliance of adolescents with
epilepsy and found that good motivation, a strong sense of normality, subjective
outcome, energy and will-power, support from parents, physicians and nurses, a positive
attitude towards the disease and its treatment, no threat to social and emotional well¬
being, and no fears of complications or seizures were related to good compliance and
decreased seizure frequency.
It is possible that dysfunctional mother-child relationships are related to increased
stress levels in children. Perhaps children involved in a dysfunctional parent-child
relationship are more likely to have increased stress levels and in turn, an increase in
seizure frequency as is suggested by Kyngas in 1990. Increased parent-child conflict has
been associated with poorer coping and increased behavioral and emotional problems in
children (Burt, Krueger, McGue, & Iacono, 2003). Hodes and Garralda (1999) found that
mothers were more hostile towards their child with epilepsy than towards healthy siblings

60
and high levels of hostility and criticism were related to child behavior problems.
Additionally, mothers of children with higher levels of seizure frequency may have
reduced opportunities for positive interactions with their children and therefore, may
view their interactions with their child as more dysfunctional in nature.
Previous research has shown that greater parental stress and depression levels have
been associated with lower adherence to medications for children with epilepsy (Hazzard,
1990; Otero & Hodes, 2000). Hence, it was surprising that maternal emotional distress
variables were not significantly related to seizure-related emergency room visits or
maternal adherence. Note, however, that in these prior aforementioned studies, mothers
reported moderately high levels of maternal stress and depressive affect while in the
current study, mothers reported low levels of maternal stress and maternal depression
with the average number of depressive symptoms endorsed by mothers in our study
reported as two. This discrepancy suggests that low levels of maternal stress and/or
depressive affect may have accounted for the lack of a relationship between the maternal
emotional distress factor and the criterion variables.
Hypothesis 2: Maternal Perceptions of the Treatment Regimen, Adherence, and
Child Health Status
We predicted that Time 1 ratings of maternal perceptions of the treatment regimen
(perceived knowledge, perceived difficulty, and perceived efficacy), and Time 1 maternal
distress and Time 1 maternal perceptions of the treatment regimen would be associated
with Time 2 adherence to the child’s medication regimen and child health status. Similar
to Hypothesis 1, a mixed pattern of results was obtained. Mothers who rated higher
levels of difficulty with their child's medical regimen at Time 1 concomitantly reported
increased seizure frequency levels and seizure-related emergency room visits for their

61
children at Time 2. Maternal perceptions regarding difficulty of the child's medical
regimen were positively related to child health status outcome variables. The relationship
between regimen complexity and maternal perceptions of regimen difficulty approached
significance (r = .205, g = .06). These findings may be related to the premise that
children with poorer health outcomes (e.g., increased levels of seizure frequency and
seizure-related emergency room visits) may have more difficult and complex regimens in
general, due to the nature of their illness. Furthermore, mothers who reported higher
levels of regimen efficacy had children with lower levels of seizure frequency.
No significant relationships were found between maternal perceptions of the
treatment regimen variables (e.g., regimen difficulty, efficacy, and knowledge) and
adherence. Overall findings suggest that maternal perceptions of the treatment regimen
are significantly related to poorer health outcomes for children with epilepsy but not
significantly related to maternal adherence. A possible explanation for this null finding is
that mothers of children with more difficult regimens and poorer health outcomes are
more adherent to their child's condition in an effort to minimize negative health outcomes
(e.g., increased seizure frequency and seizure-related emergency visits). Additionally,
relationships between maternal perceptions of the treatment regimen and adherence may
not have been identified because the questions comprising regimen perceptions are
preliminary in nature and are in need of further exploration, modification, and ultimately,
validation.
Hypothesis 3: Condition and Regimen-related Variables, Adherence, and Child
Health Status
We predicted that Time 1 condition and regimen-related variables would be
associated with Time 2 adherence to the child’s medication regimen and child health

62
status. Hypothesis 3 received mixed confirmation. While no significant relationships
were found between condition and regimen-related variables and seizure frequency, one
condition and regimen-related variable was associated with adherence and child health
status outcome variables, partially supporting our hypothesis.
Relationships between condition and regimen variables and seizure frequency were
not related. Condition and regimen-related variables are more directly related to seizure-
related emergency room visits and maternal adherence to the medical regimen.
The concordance measure was related to the seizure-related emergency room visits
variable of child health status and pill count percentage indicator of adherence.
Concordance between the prescribed regimen and reported regimen was negatively
associated with seizure-related emergency room visits and positively associated with pill
count percentage. Mothers who were unable to consistently report their child's
medication regimen with the prescribed regimen may not have administered the child's
medication as prescribed, resulting in negative child health outcomes. The results also
demonstrate that mothers who were able to report their child's medication regimen were
more adherent to their child's medication regimen. It follows from these findings that
mothers who were unable to accurately report their child's medication regimen may not
have followed the prescribed regimen and in turn, reported increased levels of emergency
room visits related to their child's seizures. High concordance ratings may be
representative of good parent-physician communication. Peterson et al. (1982) found that
good patient-doctor communication has been related to better adherence. The significant
predictive relationship demonstrated between concordance ratings and outcomes (e.g.,
seizure frequency and maternal adherence) are of theoretical and practical importance in

63
that concordance can be easily assessed during each child's clinic visit and further
examined for its relationship to child health status outcomes and maternal adherence.
Hypothesis 4: Demographic Characteristics, Adherence, and Child Health Status
We predicted that Time 1 demographic characteristics would be positively related
to Time 2 adherence to the child’s medication regimen and negatively with child health
status. Hypothesis 4 was only partially supported in that one demographic variable was
associated with adherence; family income was related to the pill count percentage
measure of adherence. Families with greater financial resources show better maternal
adherence to the child’s medication regimen than those with less income. No significant
relationships were found between race, maternal education and Time 2 outcomes of child
health status and maternal adherence. A possible explanation for the lack of a significant
relationship between demographic variables and outcome measures is that the families in
the sample were not truly representative of families with children diagnosed with a
seizure disorder. Families in the present sample were wealthier than the average family
of children with epilepsy and mothers were more educated than average participants in
previous studies (Friedman et al., 1996). However, earlier research indicated that lower
socioeconomic status and lower parental education levels were correlated significantly
with non-adherence for asthma, cystic fibrosis, and renal disease (Patterson, 1985; Tebbi
et al., 1986) and lower socioeconomic status was related to parental non-adherence
(Becker & Maiman, 1975; Rapoff & Christopherson, 1982). One explanation for these
findings is that the population at the University of Florida's Health Science Center (UF
HSC) may not be representative of the larger demographic distribution. It is feasible that
wealthier and more educated families, because of their financial stability, are better
equipped to meet the demands of their child's medical appointments, including time off

64
work in order to travel to and from the UF HSC. Less wealthy families without such
resources may only be able to be seen closer to home in order to regularly attend medical
appointments. Second, it is plausible that wealthy and educated parents are more aware
of cutting-edge research and treatment available at the UF HSC and are more willing to
participate in clinical research. Finally, the small sample was from one study site and
may not have included sufficient participants to be representative of the true demographic
distribution.
Model Testing
Sets of predictors were identified as being related to the child health status
variables of seizure frequency and seizure-related emergency room visits and the
adherence measure of pill count percentage. Parent-child dysfunction appears to play a
major role in predicting seizure frequency. Improving the dysfunctional relationship
between parent and child may be beneficial in that a stronger, more positive parent-child
relationship may be developed that is related to decreased seizure frequency. However,
another interpretation of this finding is that parents of children with increased seizure
frequencies may have limited opportunities for positive interactions with their child.
Children with greater levels of seizure frequency may have less time to spend engaging in
non-medical interactions with their parents such as playtime and open communication.
In predicting seizure-related emergency room visits, the regimen-related variable of
concordance and maternal perceptions regarding difficulty of the child's medication
regimen emerge as important variables. These measures may be better predictors of
seizure-related emergency room visits compared to maternal emotional distress measures
because they specifically assess maternal understanding of the child’s regimen and its
difficulty. It is suggested that more difficult regimens or poor concordance are more

65
closely related to seizure-related emergency room visits than maternal emotional distress
variables.
In predicting adherence, maternal emotional distress does not play a key role, but
the regimen-related variable of concordance indeed does. Concordance is likely to be
more predictive of maternal adherence than maternal emotional distress variables,
because concordance taps specifically into the mother’s understanding of the child’s
regimen. Our findings strongly indicate that incorrect reporting of the child’s medication
regimen has a more negative effect on maternal adherence than maternal mood. In that
outcome measures are not solely related to maternal mood, it should be noted that these
variables warrant further investigation.
Study Strengths and Limitations
Studies identifying correlates and predictors of adherence have typically focused on
the following disease groups: diabetes, cystic fibrosis, juvenile rheumatoid arthritis,
asthma, and renal disease. A major strength of the current study was that it assessed
several variables in their relationship to child health status and maternal adherence for
children ages one to eleven diagnosed with epilepsy. Due to this study's multi-variable
approach, it is superior to research in earlier factor-based studies because it includes
individual variable relationships with outcome measures that may have been excluded or
overlooked by factor-based approaches, allowing for the reporting of new information in
the area of pediatric epilepsy research.
A significant methodological strength of the current study was the time-lag design.
This study design provided us with a better predictive model for Time 1 measures on
Time 2 outcomes, in that Time 2 measures were partialed out of Time 2 outcomes.

66
However, the time lag was relatively short (i.e., one month) and needs to be extended in
future studies to test the reliability of the results over a longer time interval.
The study also allowed us to tap into maternal perceptions of the treatment regimen
with the creation of questions to assess difficulty, efficacy, and knowledge regarding the
child’s medical regimen. These types of questions are new to the area of epilepsy
research and can provide researchers with a great deal of information about maternal
perceptions of the treatment regimen and are important in their relationship to child
health status and maternal adherence outcomes.
However, several study limitations should be noted. Although attempts were made
to obtain a larger sample size, the final sample included 82 mother-child dyads. Small
sample size may have contributed to difficulties in detecting significant differences
between predictor and outcome variables. With this small sample size, statistical power
and prediction stability may have been methodological issues. The prediction models
may have been accurate in accounting for variance in the outcome measures but one
needs to examine the current results with caution due to this small sample size. The
correlations between some predictor and outcome variables were quite strong, and led us
to believe that a significant relationship does indeed exist between these measures.
However, a larger sample size would have resulted in greater power, which would have
enabled us to detect differences in other variables that may exist but were not apparent
due to low power leading to poor detectability. For example, difficulty scores
approached significance for inclusion in the final regression model predicting seizure
frequency (p = .056) as did the PSI-SF overall mean (p = .058). Having a larger sample
would also increase the power of prediction because variance on the measures would

67
increase, resulting in stronger predictive power. The generalizability of the results is
limited due to small sample size, higher than average family income, and higher than
average maternal education level.
Study limitations also include a lack of more concrete adherence measures such as
blood serum levels or prescription refill information. In the current study, we were
unable to report on blood serum levels because more than half of the patients did not have
these levels drawn consistently and some medications do not have normative data on
blood serum levels. Reporting of blood serum levels would have most likely resulted in
skewed findings related to this measure. Also, blood serum levels are not comparable
across different types of medications; thus, a coding system indicating therapeutic and
non-therapeutic blood level ranges suggesting adherence would be valuable in future
studies. Prescription refill information was also unavailable in the medical chart and was
not able to be obtained prospectively due to financial constraints of the current study.
Such information (e.g., blood levels and prescription refill information) is extremely
important to adherence research and should be included in future research.
Additional study limitations involve maternal perceptions of the treatment regimen.
The questions comprising the measures of difficulty, efficacy, and knowledge were
preliminary in nature and are not validated measures. In light of its strengths, the present
study can be used as a catalyst for future research in this area and serve as a model for
related fields.
Future Directions
In order for future studies to better assess predictors of maternal adherence and
child health status for young children with epilepsy, as well as outcome measures, the
following criterion should be met: (1) larger sample size; (2) assessment of child stress

68
levels; (3) validation of maternal perceptions of the treatment regimen questions; (4)
better measures of adherence; (5) inclusion of multiple study sites; (6) father data; and (7)
control group. With a larger sample size, structural equation modeling (SEM) procedures
would permit specific factor testing rather than the current variable-level testing model.
Multiple regression and other multivariate techniques can address a number of statistical
questions, but have a common limitation in that each technique can examine only a single
relationship at a time. SEM combats this problem and examines a series of predictive
relationships simultaneously. SEM techniques are distinguished by two characteristics:
(1) estimation of multiple and interrelated dependence relationships and (2) the ability to
represent unobserved concepts in these relationships and account for measurement error
in the estimation process. Relationships among predictor variables in each factor seem
promising for the SEM model given that there are significant relationships between the
specific variables comprising each factor of maternal emotional distress, maternal
perceptions of the treatment regimen, condition and regimen-related variables, and
demographic variables.
Assessment of child stress levels can assist in demonstrating a relationship between
maternal stress, child stress, and increased seizure frequency. It would be valuable to
empirically validate the specific questions assessing maternal perceptions of the treatment
regimen in the epilepsy population. Additional questions can then be created through this
validation process. This validation could result in the creation of an epilepsy-specific
measure that may have very important implications in child health status and maternal
adherence outcomes.

69
With proper funding, future studies could include blood levels, pharmacy refill
information, and electronic medication lid devices. Blood levels can be drawn as a part
of the study protocol, in order to have more consistent data regarding levels of
medication in the bloodstream for each study participant. The physician can aid in
identifying if the blood level falls into the therapeutic or non-therapeutic category. Study
protocol can also include permission to track the participant’s pharmacy refill
information at their local pharmacy in order to have a more solid measure of adherence.
Electronic counts in which a device counts each time a medication bottle is opened and
closed, may be helpful in gaining additional adherence information.
The current study was exploratory in nature. Multi-site studies may serve in
gaining larger sample sizes and an overall distribution more representative of the general
population, especially in terms of annual family income and maternal education level.
Greater than 80% of patients were accompanied to the clinic visit by their mother.
However, attempts should also be made to include fathers in future studies by utilizing
mail-outs and/or semi-structured telephone interviews. A control group of children with
controlled seizures may be useful to compare to children that are currently experiencing
seizures. The consideration of additional variables and attempts at intervention studies in
this population is suggested for future research programs in order to gain more
information in this newly growing area of research.
Clinical Implications
The results of this study identified variables that were predictive of child health
status and maternal adherence to the child's medical regimen. An important clinical goal
is to assess maternal emotional distress, maternal perceptions of the treatment regimen,

70
condition and regimen-related, and demographic variables that may impact maternal
adherence to the child's medication regimen and child health status.
The PSI-SF could be utilized in pediatric clinical settings to identify mothers of
children with epilepsy who are in need of immediate interventions to assist them in
dealing with parenting stress and more specifically, parent-child dysfunction.
Therapeutic interventions could include stress management training, parent training,
progressive muscle relaxation training, and support groups for mothers of children with
seizure disorder. If maternal perceptions of the treatment regimen are identified as
particularly problematic, medical staff can assist with clarifying maternal concerns
regarding the regimen. Reported regimens can be compared with the prescribed regimen
to aid in identifying mothers who are in need of further explanation regarding their child's
medical regimen. Lastly, those mothers with lower education levels may need additional
assistance and additional time in order to properly understand the full nature of their
child's medication regimen. In conclusion, maternal stress, perceptions of the treatment
regimen, report of the regimen, and education level can be quickly assessed in the clinical
setting in order to have a beneficial effect on child health status and adherence outcomes.

APPENDIX A
BACKGROUND INFORMATION FORM - TIME 1
Date
BACKGROUND INFORMATION
Child's Name:
Mother's Name:
Address:
Phone #: ( )
Child's Sex: Male Female
Child's Age: Child's Birthdate: I /
month day year
Mother's Age: Mother's Birthdate:
month day year
Child's Race:
1. White
2. African American
3. Hispanic American
4. Asian American
5. Native American
6. Biracial
7. Other:
Mother's Marital Status:
1. Single/Never married
2. Married
3. Separated
4. Divorced
5. Widowed
Number of Years Married (if applicable):
Child's Native Language:
Mother's Native Language:
Mother's Race:
1. White
2. African American
3. Hispanic American
4. Asian American
5. Native American
6. Biracial
7. Other:
71

72
Mother's Occupation:
Current Employment Status:
(please circle all that apply)
1. Unemployed
2. Volunteer
3. Employed part-time
4. Employed full-time
5. Homemaker
6. Student
7. Retired
8. Other- Please describe
Years of mother's total formal education: (please circle number of years completed)
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20+
Post-secondary education: If attended college, number of years completed:
Degrees received:
Attended trade or technical school? Number of years completed:
How many children are in the family?
What is the child’s current grade in school?
Has this child ever had to repeat a grade in school? If so when?
Please list the name(s) of the anti-seizure medication(s) your child is currently taking and the dose (how much
and how many times per day):
Please rate how difficult it is to manage which type of medication to give your child:
012345678
Not Somewhat
Difficult Difficult
Please rate how difficult it is to manage how much medication to give your child:
012345678
Not Somewhat
Difficult Difficult
Please rate how difficult it is to manage when to give medication to give your child:
012345678
Not Somewhat
Difficult Difficult
9 10
Extremely
Difficult
9 10
Extremely
Difficult
9 10
Extremely
Difficult

73
Please name your child's seizure diagnosis:
Over the past month, how many seizures did your child have per week?
per month?
Please rate how much you know about your child's seizure disorder:
012345678
Not Somewhat
Knowledgeable Knowledgeable
Please rate how much you know about your child's seizure medication(s):
012345678
Not Somewhat
Knowledgeable Knowledgeable
Please rate how much you know about the type of your child's seizure medication(s):
012345678
Not Somewhat
Knowledgeable Knowledgeable
9
9
9
10
Extremely
Knowledgeable
10
Extremely
Knowledgeable
10
Extremely
Knowledgeable
Please rate how much you know about the timing of your child's seizure medication(s):
0 1
2
3
4 5 6
7
8
9
10
Not
Somewhat
Extremely
Knowledgeable
Knowledgeable
Knowledgeable
If applicable, please rate how much you know about the number of your child's seizure medication(s):
0 12 3
Not
Knowledgeable
4 5 6
Somewhat
Knowledgeable
7
8 9 10
Extremely N/A
Knowledgeable
Please rate how much you know about the dosing of your child's seizure medication(s):
0 1
2
3
4 5 6
7
8
9
Not
Somewhat
Knowledgeable
Knowledgeable
10
Extremely
Knowledgeable
Please rate how good you feel your child's seizure medications are:
0123456789
Not Somewhat
Good Good
10
Extremely
Good
Please rate how good you feel you are in managing your child's seizure medications:
0123456789
Not Somewhat
Good Good
10
Extremely
Good

74
Please describe what type of information would be helpful to further your understanding of seizure
disorders:
Financial Status:
It is important for this study that we know something about your financial circumstances. We realize these are
extremely personal matters and we wish to assure you that your responses will be kept strictly confidential. Would
you please put an "X" next to the number that gives the best estimate of your total household income?
0. prefer not to answer
1. Under $ 10,000
2. $10,000 to $20,000
3. $20,001 to 30,000
4. $30,001 to $40,000
5. $40,001 to $50,000
6. $50,001 to $60,000
7. Over $60,000
Please do not answer the following, the researchers will fill out this portion.
Family ID:
Notes:

APPENDIX B
BACKGROUND INFORMATION FORM - TIME 2
Date
BACKGROUND INFORMATION
Child's Name:
Mother's Name:
Please list the name(s) of the anti-seizure medication(s) your child is currently taking and the dose (how much
and how many times per day):
Please rate
0
Not
Difficult
how difficult it is to manage which type of medication to give your child:
1 2 3 4 5 6 7 8
Somewhat
Difficult
Please rate how difficult it is to manage how much medication to give your child:
0 1 2 3 4 5 6 7 8
Not Somewhat
Difficult Difficult
Please rate how difficult it is to manage when to give medication to give your child:
012345678
Not Somewhat
Difficult Difficult
Please name your child's seizure diagnosis:
Over the past month, how many seizures did your child have per week?
per month?
9
9
9
10
Extremely
Difficult
10
Extremely
Difficult
10
Extremely
Difficult
Please rate how much you know about your child's seizure disorder:
0 1 2 3 4 5 6 7
Not Somewhat
Knowledgeable Knowledgeable
Please rate much you know about your child's seizure medication(s):
0 1 2 3 4 5 6 7
Not Somewhat
Knowledgeable Knowledgeable
8 9 10
Extremely
Knowledgeable
8 9 10
Extremely
Knowledgeable
75

76
Please rate how much you know about the type of your child's seizure medication(s):
0123456789 10
Not Somewhat Extremely
Knowledgeable Knowledgeable Knowledgeable
Please rate how much you know about the timing of your child's seizure medication(s):
0123456789 10
Not Somewhat Extremely
Knowledgeable Knowledgeable Knowledgeable
If applicable, please rate how much you know about the number of your child's seizure medication(s):
0 1
2 3 4
5 6
7 8
9
10
Not
Somewhat
Extremely N/A
Knowledgeable
Knowledgeable
Knowledgeable
Please rate how
much you know about the dosing of vour child's seizure medicationfs):
0 1
2 3 4
5 6
7 8
9
10
Not
Somewhat
Extremely
Knowledgeable
Knowledgeable
Knowledgeable
Please rate how
good you feel your child's seizure medications are:
0 1
2 3 4
5 6
7 8
9
10
Not
Somewhat
Extremely
Good
Good
Good
Please rate how
good you feel you are in managing your child's seizure medications:
0 1
2 3 4
5 6
7 8
9
10
Not
Somewhat
Extremely
Good
Good
Good
Please describe what type of information would be helpful to further your understanding of seizure
disorders:
Please do not answer the following, the researchers will fill out this portion.
Family ID:
Notes:

APPENDIX C
MEDICAL CHART INFORMATION FORM - TIME 1
MEDICAL CHART INFORMATION
Patient's Name:
Patient's DOB: Age
Date of Diagnosis:
Newly Diagnosed: yes no
If no, time since diagnosis (months):
Seizure Frequency (over the past month):per week per month
Seizure Type(s) {the type(s) of seizures your child is diagnosed with}:
Medication(s) and Dose Info:
ER Visits Dates:
Blood Levels:
Pharmacy Refill Information:
% Concordance between prescribed medication regimen and reported medication
regimen
Complexity
77

APPENDIX D
MEDICAL CHART INFORMATION FORM - TIME 2
MEDICAL CHART INFORMATION
Patient's Name:
Patient's DOB: Age
Date of Diagnosis:
Newly Diagnosed: yes no
If no, time since diagnosis (months):
Seizure Frequency (over the past month):per week per month
Seizure Type(s) {the type(s) of seizures your child is diagnosed with}:
Medication(s) and Dose Info:
ER Visits Dates:
Blood Levels:
Pharmacy Refill Information:
% Concordance between prescribed medication regimen and reported medication
regimen
Complexity
78

LIST OF REFERENCES
Abidin, R.R. (1995). Parenting Stress Index: Professional Manual (3rd ed). Odessa:
Psychological Assessment Resources.
Adams, P.F., Elendershot, G.E., & Maraño, M.A. (1996). Current Estimates from the
National Health Interview Survey: National Center for Health Statistics. Vital
Health Statistics, 10(200), 85-86.
Anderson, B.J., Auslander, W.F., Jung, D.C., Miller, J.P., & Santiago, J.V. (1990).
Assessing family sharing of diabetes responsibilities. Journal of Pediatric
Psychology, 15, 477-492.
Appleton, R. & Gibbs, J. (1998). Epilepsy in Childhood and Adolescence. Malden:
Blackwell Science Incorporated.
Austin, J.K. (1989). Predicting parental anticonvulsant medication compliance using the
theory of reasoned action. Journal of Pediatric Nursing, 4(2), 88-95.
Beck, D.E., Fennell, R.S., Yost, R.L., Robinson, J.D., Geary, D., & Richards, G.A.
(1980). Evaluation of an educational program on compliance with medication
regimens in pediatric patients with renal transplants. The Journal of Pediatrics,
96, 1094-1097.
Bobrow, E.S., Avruskin, T.W., & Siller, J. (1985). Mother-daughter interaction and
adherence to diabetes regimens. Diabetes Care, 8, 146-151.
Bond, G.G., Aiken, L.S., & Somerville, S.C. (1992). The health belief model and
adolescents with insulin-dependent diabetes mellitus. Health Psychology, 11,
90-198.
Brownbridge,G. & Fielding, D-M. (1994). Psychosocial adjustment and adherence to
dialysis treatment regimens. Pediatric Nephrology, 8, 744-749.
Browne, T.R. & Holmes, G.L. (2000). Handbook of epilepsy (2nd edition).
Philadelphia: Lippincott, Williams, & Wilkins.
Burt, S.A., Krueger, R.F., McGue, M„ & Iacono, W. (2003). Archives of General
Psychiatry, 60(5), 505-513.
79

80
Cummings, E.M. & Davies, P.T. (1994). Maternal depression and child development.
Journal of Child Psychology and Psychiatry, 35(1), 73-112.
Downey, G. & Coyne, J. (1990). Children of depressed parents: An integrative review.
Psychological Bulletin, 108(1), 50-76.
Friedman, I.M., Litt, I.F., King, D., Henson, R., Holtzman, D., Halverson, D., &
Kraemer, H. (1986). Compliance with anticonvulsant therapy by epileptic youth.
Journal of Adolescent Health Care, 7, 12-17.
Gudas, L.J., Kroocher, G.P., & Wyplj, D. (1991). Perceptions of medical compliance in
children and adolescents with cystic fibrosis. Developmental & Behavioral
Pediatrics, 12, 236-247.
Hazzard, A., Hutchinson, S.J., & Krawiecki, N. (1990). Factors related to adherence to
medication regimens in pediatric seizure patients. Journal of Pediatric
Psychology, 15(4), 543-555.
Hodes, M. & Garralda, M.E. (1999). Maternal expressed emotion and adjustment in
children with epilepsy. Journal of Child Psychology and Psychiatry, 7, 1083-
1093.
Jacobsen, A.M., Hauser, S.T., Wolfsdorf, J.I., Houlihan, J., Milley, J.E., Herskowitz,
R.D., Wertlieb, D., & Watt, E. (1987). Psychologic predictors of compliance in
children with recent onset diabetes mellitus. Journal of Pediatrics, 110, 805-811.
Kyngas, H. (2000). Compliance with health regimens of adolescents with epilepsy.
Seizure, 9, 598-604.
LaGreca, A.M., Follansbee, D., & Skyler, J.S. (1990). Developmental and behavioral
aspects of diabetes management in youngsters. Children’s Health Care, 19, 132-
139.
Lorenz, R.A., Christensen, N.K., & Pichert, J.W. (1985). Diet-related knowledge, skill,
and adherence among children with insulin-dependent diabetes mellitus.
Pediatrics, 75, 872-876.
McCaul, K.D., Glasgow, R.E., & Schafer, L.C. (1987). Diabetes regimen behaviors:
Predicting adherence. Medical Care, 25, 868-881.
Murray, L., Sinclair, D., Cooper, P., Ducournau, P., Turner, P., & Stein, A. (1999). The
socioemotional development of 5-year-old children of postnatally depressed
mothers. Journal of Child Psychology and Psychiatry, 40(8), 1259-1271.
Myers, L.B. & Midence, K. (Eds.). (1998). Adherence to treatment in medical
conditions. New Delhi: Harwood Academic Publishers.

81
Nezu, A.M. (Ed.). (2000). Practitioner's guide to empirically based measures of
depression. New York: Kluwer Academic, Plenum Publishers.
Otero, S. & Hodes, M. (2000). Maternal expressed emotion and treatment compliance of
children with epilepsy. Developmental Medicine and Child Neurology, 42(9),
604-608.
Patterson, J.M. (1985). Critical factors affecting family compliance with home treatment
for children with cystic fibrosis. Family Relations, 34, 79-89.
Peterson, G.M., McLean, S., & Millingen, K.S. (1982). Determinants of patient
compliance with anticonvulsant therapy. Epilepsia, 23, 607-613.
Radius, S.M., Marshall, H.B., Rosenstock, I.M., Drachman, R.H., Schuberth, K.C., &
Teets, K.C. (1978). Factors influencing mothers’ compliance with a medication
regimen for asthmatic children. Journal of Asthma Research, 15, 133-149.
Radloff, L.S. (1977). The CES-D scale: A self-report depression scale for research in the
general population. Applied Psychological Measurement, 1, 385-401.
Rapoff, M.A. (1999). Adherence to pediatric medical regimens. New York: Kluwer
Academic & Plenum Publishers.
Rapoff, M.A. & Christophersen, E.R. (1982). Improving compliance in pediatric
practice. Pediatric Clinics of North America, 29(2), 339-357.
Shorvon, S. (2000). Handbook of epilepsy treatment. Oxford: Alden Press Limited.
Shope, J.T. (1988). Compliance in children and adults: review of studies. Epilepsy Res.
Suppl, 1, 23-47.
Smith, S.D., Rosen, D., Trueworthy, R.C., & Lowman, J.T. (1979). A reliable method
for evaluating drug compliance in children with cancer. Cancer, 43, 169-173.
Spector, S., Tranah, A., Cull, C., & Goldstein, L.H. (1999). Reduction in seizure
frequency following a short-term group intervention for adults with epilepsy.
Seizure, 8, 297-303.
Tebbi, C.K., Richards, M.E., Cummings, K.M., Zevon, M.A., & Mallon, J.C. (1986).
The role of parent-adolescent concordance in compliance with cancer
chemotherapy. Adolescence, 28, 599-611.
United States Census Bureau. (2001). Poverty in the United States: 2000. (U.S. Census
Publication No. P 60-214). Washington, DC: U.S. Government Printing Office.

82
Wynn, K.S. & Eckel, E.M. (1986). Juvenile rheumatoid arthritis and home physical
therapy program compliance. Physical & Occupational Therapy in Pediatrics, 6,
55-63.
Zysset, T., Redeberg, A., Vasella, F., Kupfer, A., & Bircher, J. (1981). Phenytoin
therapy for epileptic children: Evaluation of salivary and plasma concentrations
and of methods of assessing compliance. Developmental Medicine and Child
Neurology, 23, 66-75.

BIOGRAPHICAL SKETCH
Sobha P. Fritz was born in St. Louis, Missouri in 1975 and raised in Tampa,
Florida. Sobha received a B.A. in psychology from the University of South Florida in
1997. Before coming to the University of Florida in 1998, she worked as a research
assistant for the Psychosocial Oncology Program at the H. Lee Moffitt Cancer Center.
Sobha married her high school sweetheart, Jason Jon Fritz, in December 1998, who is
completing his doctorate in molecular genetics and microbiology. Sobha received an
M.S. in clinical psychology from the University of Florida in 2000. She completed her
doctoral and internship training in the University of Florida's Department of Clinical and
Health Psychology, with a specialization in pediatric psychology, in 2003. Sobha will
begin a postdoctoral fellowship at Emory University and the Children’s Healthcare of
Atlanta hospitals in Atlanta, Georgia in January 2004.
83

I certify that I have read this study and that in my opinion it conforms to acceptable
standards of scholarly presentation and is fully adequate, in scope andjquality, as a
dissertation for the degree of Doctor of Philosophy.
Robert Gluecka
Professor of Clinical and Health Psychology
1 certify that I have read this study and that in my opinion it conforms to acceptable
standards of scholarly presentation and is fully adequate, in scope and quality, as a
dissertation for the degree of Doctor of Philosoj
tussell Bauer
3rofessor of Clinical and Health Psychology
I certify that I have read this study and that in my opinion it conforms to acceptable
standards of scholarly presentation and is fully adequate, in scope and quality, as a
dissertation for the degree of Doctor of Philosophy.
g
Cy/tA
Paul Carney
Professor of Neuroscience
I certify that I have read this study and that in my opinion it conforms to acc
standards of scholarly presentation and is fully adequate, in scope and quality, as
dissertation for the degree of Doctor of Philosophy.
Á.
ptable
Gary Geffy en
Associate Professor of Clinical and Health
Psychology
I certify that I have read this study and that in my opinion it conforms to acceptable
standards of scholarly presentation and is fully adequate, in scope and quality, as a
dissertation for the degree of Doctor of Philosophy.
/] -
Claydell Home
Associate Professor of Nursing

This dissertation was submitted to the Graduate Faculty of the College of Health
Professions and to the Graduate School and was accepted as partial fulfillment of the
requirements for the degree of Doctor of Philoss
December 2003
Dean, College of Health Professions
Dean, Graduate School

UNIVERSITY OF FLORIDA
3 1262 08555 2866



xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID E9CL4Z4EP_JG0VB4 INGEST_TIME 2014-06-25T15:34:05Z PACKAGE AA00022328_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES



PAGE 1

35(',&7,1* 0$7(51$/ $'+(5(1&( $1' &+,/' +($/7+ 67$786 ,1 &+,/'+22' (3,/(36< $1 (;3/25$725< 02'(/ %\ 62%+$ 3 )5,7= $ ',66(57$7,21 35(6(17(' 72 7+( *5$'8$7( 6&+22/ 2) 7+( 81,9(56,7< 2) )/25,'$ ,1 3$57,$/ )8/),//0(17 2) 7+( 5(48,5(0(176 )25 7+( '(*5(( 2) '2&725 2) 3+,/2623+< 81,9(56,7< 2) )/25,'$

PAGE 2

7KLV GRFXPHQW LV GHGLFDWHG WR P\ KXVEDQG SDUHQWV DQG VLVWHUV IRU WKHLU VXSSRUW DQG XQFRQGLWLRQDO ORYH 0\ IDPLO\fV HPSKDVLV UHJDUGLQJ WKH LPSRUWDQFH RI HGXFDWLRQ DQG KHOSLQJ RWKHUV KDV JXLGHG PH LQWR D FDUHHU IRU ZKLFK DP ZHOO VXLWHG ZRXOG OLNH WR HVSHFLDOO\ WKDQN P\ KXVEDQG -DVRQ -RQ )ULW] IRU DOZD\V EHOLHYLQJ LQ PH DQG SURYLGLQJ VHFXUH IRXQGDWLRQ RI ZDUPWK DQG HQFRXUDJHPHQW

PAGE 3

$&.12:/('*0(176 7KLV LQYHVWLJDWLRQ FRXOG QRW KDYH EHHQ FDUULHG RXW ZLWKRXW WKH PHQWRUVKLS DQG GLUHFWLRQ RI 5REHUW *OXHFNDXI 3K' WKH YDOXDEOH FRQWULEXWLRQV RI 1LFROH -DJXV]W\Q WKH GHGLFDWLRQ RI 3DXO &DUQH\ 0' DQG WKH VWDII DW WKH 8QLYHUVLW\ RI )ORULGD 3HGLDWULF 1HXURORJ\ &OLQLF ZRXOG OLNH WR DFNQRZOHGJH P\ FRPPLWWHH PHPEHUV IRU WKHLU VFKRODUO\ DVVLVWDQFH DQG JXLGDQFH ZRXOG DGGLWLRQDOO\ OLNH WR WKDQN P\ IDPLO\ IRU WKHLU VXSSRUW DQG HQFRXUDJHPHQW WKURXJKRXW WKLV SURMHFW LLL

PAGE 4

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

PAGE 5

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

PAGE 6

$33(1',; $ %$&.*5281' ,1)250$7,21 )250 7,0( % %$&.*5281' ,1)250$7,21 )250 7,0( & 0(',&$/ &+$57 ,1)250$7,21 )250 7,0( 0(',&$/ &+$57 ,1)250$7,21 )250 7,0( /,67 2) 5()(5(1&(6 %,2*5$3+,&$/ 6.(7&+ YL

PAGE 7

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
PAGE 8

FRPSOH[LW\ RI WKH PHGLFDWLRQ UHJLPHQ FRQFRUGDQFH EHWZHHQ SUHVFULEHG UHJLPHQ DQG UHSRUWHG UHJLPHQf DQG PDWHUQDO SHUFHSWLRQV RI WKH WUHDWPHQW UHJLPHQ SHUFHLYHG HIILFDF\ SHUFHLYHG NQRZOHGJH SHUFHLYHG GLIILFXOW\f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f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

PAGE 9

&+$37(5 ,1752'8&7,21 (SLOHSV\ $Q 2YHUYLHZ (SLOHSV\ LV D FKURQLF LOOQHVV WKDW W\SLFDOO\ EHJLQV LQ HDUO\ FKLOGKRRG LQ ZKLFK WKH SDWLHQW LV SURQH WR H[SHULHQFH UHFXUUHQW HSLOHSWLF VHL]XUHV ZKHUH WZR RU PRUH DWWDFNV KDYH RFFXUUHG (SLOHSV\ LV GHILQHG DV D PDQLIHVWDWLRQ RI DQ DEQRUPDO DQG H[FHVVLYH V\QFKURQL]HG GLVFKDUJH RI D VHW RI FHUHEUDO QHXURQV 6KRUYRQ f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f 3URYRNHG VHL]XUHV DUH RIWHQ LVRODWHG HYHQWV DQG GR QRW WHQG WR UHFXU RQFH WKH FDXVH LV UHPRYHG $ SHUVRQ LV GHVFULEHG DV KDYLQJ DFWLYH HSLOHSV\ ZKHQ DW OHDVW RQH VHL]XUH KDV RFFXUUHG LQ WKH SUHFHGLQJ \HDUV (SLOHSV\ LV VDLG WR EH LQ UHPLVVLRQ ZKHQ QR VHL]XUHV KDYH RFFXUUHG LQ WKH SUHFHGLQJ \HDUV $SSOHWRQ t *LEEV f 7KHUH DUH DSSUR[LPDWHO\ FDVHV RI HSLOHSV\ SHU SHRSOH SHU \HDU ZLWK D UDQJH RI FDVHV 1HDUO\ SHU DGROHVFHQWV SHU \HDU JR RQ WR GHYHORS VRPH IRUP RI HSLOHSV\ 7KH FXPXODWLYH LQFLGHQFH ULVN RI DQ LQGLYLGXDO GHYHORSLQJ HSLOHSV\ LQ KLV RU

PAGE 10

KHU OLIHWLPH LV EHWZHHQ b DQG b 7KH KLJKHVW LQFLGHQFH UDWHV DUH REVHUYHG LQ QHRQDWHV DQG \RXQJ FKLOGUHQ ZLWK D VHFRQG SHDN LQ ROG DJH *UHDWHU WKDQ b RI VHL]XUHV VWDUW LQ FKLOGKRRG DQG EHJLQ EHIRUH DJH %URZQH t +ROPHV f 7KH SUHYDOHQFH RI DFWLYH HSLOHSV\ LV FDVHV SHU SHRSOH SHU \HDU ,VRODWHG VHL]XUHV ILUVW DQG RQO\f RFFXU LQ DSSUR[LPDWHO\ SHRSOH SHU HDFK \HDU 7KH SUHYDOHQFH RI HSLOHSV\ LQ VFKRRO DJH FKLOGUHQ DJHG \HDUVf LV bb 7KH SUHYDOHQFH RI HSLOHSV\ LQ DGROHVFHQWV LV SHU DGROHVFHQWV 2QHWKLUG RI DOO HSLOHSVLHV WKDW EHJLQ LQ FKLOGKRRG ZLOO KDYH VSRQWDQHRXVO\ UHPLWWHG E\ SXEHUW\ DQG WKLV FKDQJH LV XVXDOO\ VXVWDLQHG WKURXJKRXW DGXOW OLIH (SLOHSV\ LV UHODWLYHO\ VWDWLF DIWHU HDUO\ FKLOGKRRG EXW VKRZV D WHQGHQF\ WR ULVH LQ ROG DJH (SLOHSV\ LV PRUH FRPPRQ LQ GHYHORSLQJ FRXQWULHV DQG LV WKRXJKW WR EH DVVRFLDWHG ZLWK SRRUHU QXWULWLRQ SRRUHU K\JLHQH KLJKHU SUHYDOHQFH RI LQIHFWLRXV GLVHDVHV DQG D KLJKHU SURSRUWLRQ RI FKLOGUHQ LQ WKH SRSXODWLRQ $SSOHWRQ t *LEEV f (WLRORJ\ &RQJHQLWDO DQG JHQHWLF FRQGLWLRQV DUH PRVW FRPPRQO\ UHODWHG WR HDUO\ FKLOGKRRG HSLOHSV\ *HQHWLFV DUH WKRXJKW WR SOD\ DQ LPSRUWDQW UROH LQ VHL]XUH GLVRUGHUV +HUHGLWDU\ DQG JHQHWLF IDFWRUV DUH UHODWHG WR LGLRSDWKLF RU SULPDU\ HSLOHSV\ V\QGURPHV DQG SRVVLEO\ LQ VRPH VHFRQGDU\ HSLOHSVLHV EXW NQRZQ VLQJOHJHQH GLVRUGHUV DUH UHODWHG WR HSLOHSV\ LQ RQO\ b RI FDVHV ,Q ROGHU FKLOGUHQ DQG \RXQJ DGXOWV FRPPRQ FDXVHV RI HSLOHSV\ LQFOXGH LQKHULWHG SUHGLVSRVLWLRQ KLSSRFDPSDO VFOHURVLV DOFRKRO RU GUXJ DEXVH DQG WUDXPD 9DVFXODU GLVHDVH LV D SUHGRPLQDQW FDXVH RI HSLOHSV\ LQ WKH HOGHUO\ 7XPRUV DQG LQIHFWLRQV FDQ RFFXU DW DQ\ DJH EXW PDOLJQDQW WXPRUV DUH PRUH OLNHO\ WR RFFXU DIWHU DJH 6KRUYRQ f ,Q b RI FKLOGKRRG FDVHV RI HSLOHSV\ D VSHFLILF FDXVH PD\ QRW EH

PAGE 11

QRWHG EXW WKH HWLRORJ\ RI WKH GLVHDVH PD\ EH DSSDUHQW IURP FOLQLFDO LQIRUPDWLRQ DQG WHVW UHVXOWV $SSOHWRQ t *LEEV f 3DWLHQWV ZLWK VSHFLILF V\QGURPHV VXFK DV )UDJLOH ; V\QGURPH 6WXUJH:HEHU V\QGURPH +XQWLQJWRQnV 1LHPDQQ3LFN GLVHDVH W\SH & WXEHURXV VFOHURVLV %RXPHYLOOHnV V\QGURPHf QHXURILEURPDWRVLV W\SH DQG W\SH f DQG 'RZQfV V\QGURPH PD\ GHYHORS HSLOHSV\ ,QIHFWLRXV GLVHDVHV WKDW LQYROYH WKH EUDLQ VXFK DV YLUDO DQG EDFWHULDO PHQLQJLWLV DOVR OHDG WR HSLOHSV\ &HUHEUDO WXEHUFXORVLV IUHTXHQWO\ FDXVHV HSLOHSV\ 0DODULD VWUHS LQIHFWLRQV DQG HQFHSKDOLWLV DUH DOO WKRXJKW WR EH UHODWHG HSLOHSV\ &HUHEUDO WXPRUV DUH UHODWHG WR DSSUR[LPDWHO\ b RI QHZO\ GLDJQRVHG FDVHV RI HSLOHSV\ ,Q FKLOGUHQ WXPRUV DUH QRW W\SLFDO FDXVHV RI HSLOHSV\ +HDG LQMXULHV DFFRXQW IRU DSSUR[LPDWHO\ b RI DOO FDVHV RI HSLOHSV\ WUDIILF DFFLGHQWV DQG IDOOVf 'UXJV DOFRKRO WR[LQV DQG PHWDEROLF GLVRUGHUV KDYH EHHQ DVVRFLDWHG ZLWK VHL]XUH GLVRUGHUV $GGLWLRQDO FDXVHV RI HSLOHSV\ LQFOXGH EUDLQ PDOIRUPDWLRQV XQUXSWXUHG DQHXU\VPV LQWUDFUDQLDO KHPRUUKDJHV OHVLRQV SHULQDWDO LQMXU\ DQG KLSSRFDPSDO VFOHURVLV +LSSRFDPSDO VFOHURVLV LV WKH PRVW FRPPRQ FDXVH RI WHPSRUDO OREH HSLOHSV\ 6KRUYRQ f 'LDJQRVLQJ (SLOHSV\ $ SK\VLFLDQ QHHGV GHWDLOHG LQIRUPDWLRQ WR UHQGHU DQ HSLOHSV\ GLDJQRVLV DQG D GLDJQRVLV RI HSLOHSV\ LV GHSHQGHQW DOPRVW HQWLUHO\ RQ WKH KLVWRU\ JLYHQ 'XULQJ WKH DVVHVVPHQW WKH GRFWRU ZLOO W\SLFDOO\ JDLQ WKH IROORZLQJ LQIRUPDWLRQ D YHU\ GHWDLOHG KLVWRU\ RI WKH VHL]XUHVf DQ H\HZLWQHVV DFFRXQW RI WKH VHL]XUHVf D KLVWRU\ RI SUHYLRXV LOOQHVVHV RU WUDXPDWLF HYHQWV IDPLO\ KLVWRU\ DQG KLVWRU\ RI DQ\ FRRFFXUULQJ GLVHDVHVf DQG WUHDWPHQWVf 1RWHZRUWK\ LV WKH IDFW WKDW DQ DFFXUDWH H\HZLWQHVV DFFRXQW LV HVVHQWLDO DQG PD\ EH WKH RQO\ KLVWRU\ DYDLODEOH IRU \RXQJ FKLOGUHQ $ SK\VLFDO H[DPLQDWLRQ ZLOO EH FRQGXFWHG 9LGHR UHFRUGLQJV VOHHS VWXGLHV DQG ((*V RI VXVSHFWHG VHL]XUHV DUH XVHIXO

PAGE 12

DLGV IRU GLDJQRVLV 6RPH FKDUDFWHULVWLFV WKDW PD\ PLPLF HSLOHSV\ LQFOXGH KHDGDFKHV MLWWHUV WLF GLVRUGHU VKXGGHULQJ VKLYHULQJ UHSHWLWLYH ULWXDOLVWLF PRYHPHQWV YDFDQW VWDUHV GD\GUHDPLQJ LQDWWHQWLRQ SV\FKRORJLFDO GLVWXUEDQFHV DQG H[FHVVLYH VOHHS %URZQH t +ROPHV f &ODVVLILFDWLRQ RI WKH (SLOHSVLHV 6HL]XUHV DUH FODVVLILHG LQWR WZR PDLQ FDWHJRULHV SDUWLDO DQG JHQHUDOL]HG 7KHVH WZR FDWHJRULHV DUH IXUWKHU FODVVLILHG 3DUWLDO VHL]XUHV LQFOXGH VLPSOH SDUWLDO FRPSOH[ SDUWLDO DQG SDUWLDO VHL]XUHV HYROYLQJ WR VHFRQGDU\ JHQHUDOL]HG *HQHUDOL]HG VHL]XUHV DUH FODVVLILHG DV W\SLFDO DEVHQFH SHWLW PDOf DW\SLFDO DEVHQFH P\RFORQLF FORQLF WRQLF WRQLF FORQLF JUDQG PDOf DWRQLF DQG XQFODVVLILDEOH $SSOHWRQ t *LEEV f 3DUWLDO 6HL]XUHV 3DUWLDO VHL]XUHV DUH IXUWKHU FODVVLILHG DV fVLPSOHf LQ ZKLFK FRQVFLRXVQHVV LV UHWDLQHG RU fFRPSOH[f LQ ZKLFK FRQVFLRXVQHVV LV LPSDLUHG RU ORVW 6LPSOH SDUWLDO VHL]XUHV EHJLQQLQJ ORFDOO\f FDQ RFFXU ZLWK PRWRU V\PSWRPV VHQVRU\ V\PSWRPV DXWRQRPLF V\PSWRPV DQG SV\FKLF V\PSWRPV 6LPSOH SDUWLDO VHL]XUHV ZLWK VHQVRU\ DXWRQRPLF RU SV\FKLF V\PSWRPV PD\ EH HDVLO\ RYHUORRNHG LQ \RXQJHU FKLOGUHQ ZKR PD\ EH XQDEOH WR GHVFULEH VXFK V\PSWRPV 3DUWLDO VHL]XUHV PD\ EHFRPH VHFRQGDU\ JHQHUDOL]HG UHVXOWLQJ LQ WRQLFFORQLF JUDQG PDOf FRQYXOVLRQV %URZQH t +ROPHV f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

PAGE 13

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f G\VPQHVWLF V\PSWRPV GLVWXUEDQFH RI PHPRU\f FRJQLWLYH V\PSWRPV GUHDP\ VWDWHV DQG GHSHUVRQDOL]DWLRQf DIIHFWLYH V\PSWRPV IHDU GHSUHVVLRQ DQJHU LUULWDELOLW\ DQG HODWLRQf LOOXVLRQV RI VL]H VKDSH ZHLJKW GLVWDQFH RU VRXQG RU VWUXFWXUDO KDOOXFLQDWLRQV YLVXDO DXGLWRU\ JXVWDWRU\ RU ROIDFWRU\ IRUPVf $SSOHWRQ t *LEEV f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

PAGE 14

3DUWLDO VHL]XUHV VLPSOH FRPSOH[ RU VLPSOH HYROYLQJ WR FRPSOH[f PD\ VSUHDG WR EHFRPH JHQHUDOL]HG 7KH SDUWLDO VHL]XUH LV RIWHQ H[SHULHQFHG DV DQ DXUD LQ WKH VHFRQGV EHIRUH WKH JHQHUDOL]HG VHL]XUH 7KH JHQHUDOL]HG VHL]XUH LV XVXDOO\ WRQLFFORQLF WRQLF RU DWRQLF %URZQH t +ROPHV f *HQHUDOL]HG 6HL]XUHV 7KHUH DUH WZR W\SHV RI DEVHQFH VHL]XUHV W\SLFDO DEVHQFH SHWLW PDO VHL]XUHVf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t *LEEV f :LWK P\RFORQLF VHL]XUHV WKHUH LV D EULHI FRQWUDFWLRQ RI D PXVFOH PXVFOH JURXS RU VHYHUDO PXVFOH JURXSV WKDW FDQ EH VLQJOH RU UHSHWLWLYH YDU\LQJ LQ VHYHULW\ IURP D QHDUO\ LPSHUFHSWLEOH WZLWFK WR D VHYHUH MHUNLQJ ZKLFK FDQ UHVXOW LQ D VXGGHQ IDOO RU WKH SURSXOVLRQ RI KDQGKHOG REMHFWV IO\LQJ VDXFHU V\QGURPHf &ORQLF VHL]XUHV FRQVLVW RI DV\PPHWULFDO DQG LUUHJXODU MHUNLQJ &ORQLF VHL]XUHV DUH PRVW IUHTXHQW LQ QHRQDWHV LQIDQWV DQG \RXQJ FKLOGUHQ 7RQLF VHL]XUHV DUH DVVRFLDWHG ZLWK PXVFOH FRQWUDFWLRQV ZLWK DOWHUHG FRQVFLRXVQHVV ,Q WRQLF VHL]XUHV WKH FRQWUDFWLRQ FDXVHV H[WHQVLRQ RI WKH QHFN FRQWUDFWLRQ RI IDFLDO PXVFOHV ZLGHQLQJ RI WKH H\HV XSWXPLQJ RI WKH H\HEDOOV FRQWUDFWLRQ RI UHVSLUDWRU\ PXVFOHV DQG VSDVPV RI WKH XSSHU OLPE PXVFOHV %URZQH t +ROPHV f

PAGE 15

7RQLFFORQLF VHL]XUHV DUH PRVW DVVRFLDWHG ZLWK WKH W\SH RI HSLOHSV\ WKDW WKH SXEOLF LPDJLQHV 7KH WRQLF SKDVH LV VRPHWLPHV SUHFHGHG E\ D SHULRG LQ ZKLFK DQ DWWDFN LV DQWLFLSDWHG SURGURPDO SHULRGf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f DUH FRPPRQ 7KH FORQLF SKDVH ODVWV VHFRQGV DQG LV IROORZHG E\ IXUWKHU EULHI WRQLF FRQWUDFWLRQ RI WKH PXVFOHV VRPHWLPHV DFFRPSDQLHG E\ LQFRQWLQHQFH 7KH ILQDO SKDVH ODVWV EHWZHHQ PLQXWHV DQG LV FKDUDFWHUL]HG E\ IODFFLGLW\ RI PXVFOHV DQG WKHQ FRQVFLRXVQHVV LV VORZO\ UHJDLQHG %URZQH t +ROPHV f 7KH PRVW VHYHUH IRUP RI WKH DWRQLF VHL]XUH LV WKH FODVVLF GURS DWWDFN LQ ZKLFK DOO SRVWXUDO WRQH LV VXGGHQO\ ORVW FDXVLQJ WKH SHUVRQ WR FROODSVH ,Q DWRQLF VHL]XUHV WRQH FDQ EH PRUH UHVWULFWHG UHVXOWLQJ LQ EREELQJ RI WKH KHDG D ERZLQJ PRYHPHQW RU VDJJLQJ DW WKH NQHH 7KHVH VHL]XUHV DUH VKRUW DQG DUH XVXDOO\ IROORZHG E\ LPPHGLDWH UHFRYHU\

PAGE 16

$W OHDVW RI DOO VHL]XUH GLVRUGHUV DUH QRW FODVVLILDEOH DQG IDOO LQWR WKH XQFODVVLILDEOH FDWHJRU\ 6HL]XUHV DUH GHVFULEHG DV XQFODVVLILDEOH EHFDXVH W\SLFDO FOLQLFDO DQG ((* SDWWHUQV GR QRW FRQIRUP WR WKH GHILQHG FDWHJRULHV RI VHL]XUHV %URZQH t +ROPHV f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t *LEEV f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

PAGE 17

WUHDWPHQW DSSURDFK RI HSLOHSV\ WUHDWPHQW VSHFLILFDOO\ IRU WKRVH SDWLHQWV ZKR GR QRW UHVSRQG ZHOO WR ILUVW VHFRQG DQG WKLUG OLQH WUHDWPHQW DSSURDFKHV 6KRUYRQ f 6XUJLFDO WUHDWPHQW LV DQ RSWLRQ IRU SDWLHQWV ZLWK VHYHUH LQWUDFWDEOH HSLOHSV\ WKDW IDLOV WR UHVSRQG WR PHGLFDWLRQ DQG RWKHU IRUPV RI WUHDWPHQW $OWHUQDWLYH PHWKRGV RI WUHDWPHQW LQFOXGH WKH NHWRJHQLF GLHW VWHURLGV LQWUDYHQRXV LPPXQRJOREXOLQV EHKDYLRU WKHUDS\ DFXSXQFWXUH KHUEDO WKHUDS\ YDJXV QHUYH VWLPXODWLRQ DQG VHL]XUH DOHUW GRJV 3V\FKRORJLFDO ,PSDFW RI (SLOHSV\ 3V\FKLDWULF GLVRUGHUV RFFXU PRUH IUHTXHQWO\ ZLWK HSLOHSV\ WKDQ ZLWK DQ\ RWKHU FKURQLF GLVRUGHU $SSOHWRQ t *LEEV f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f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f ,QFUHDVHG VWUHVV OHYHOV DQG DQ[LHW\ KDYH EHHQ GHPRQVWUDWHG DV UHODWHG WR DQ LQFUHDVH LQ

PAGE 18

VHL]XUH IUHTXHQF\ DQG VHL]XUH VHYHULW\ 6SHFWRU HW DO f 7KH HIIHFWV RI PDWHUQDO HPRWLRQDO DGMXVWPHQW DQG DGKHUHQFH ZLWK \RXQJ FKLOGUHQ ZLWK HSLOHSV\ KDYH UHFHLYHG OLPLWHG DWWHQWLRQ $GKHUHQFH WR WKH 5HJLPHQ $Q 2YHUYLHZ 2QHWKLUG RI DOO SDWLHQWV DQG b RI FKURQLFDOO\ LOO SDWLHQWV DUH QRQDGKHUHQW WR WKHLU SUHVFULEHG PHGLFDO UHJLPHQV 5DSRII f 1RQDGKHUHQFH FDQ QHJDWLYHO\ LPSDFW KHDOWK VWDWXV DQG KHDOWKUHODWHG TXDOLW\ RI OLIH DQG FDQ OHDG WR LQFUHDVHG ULVNV RI PRUELGLW\ DQG PRUWDOLW\ 1RQDGKHUHQFH LV DOVR UHODWHG WR LQFUHDVHG KHDOWK FDUH FRVWV ZKHUH WKH FRVW RI QRQDGKHUHQFH LQ WKH 8QLWHG 6WDWHV LV QHDUO\ ELOOLRQ HDFK \HDU 5DSRII f $GGLWLRQDOO\ QRQDGKHUHQFH PD\ LQDGYHUWHQWO\ QHJDWLYHO\ LPSDFW FOLQLFDO GHFLVLRQV PDGH E\ SK\VLFLDQV 3K\VLFLDQV ZKR DWWULEXWH GLVHDVHUHODWHG SUREOHPV WR LQHIIHFWLYH RU LQFRUUHFW SUHVFULSWLRQV DQG QRW WR WKH SDWLHQWnV DGKHUHQFH SUREOHPV PD\ FKDQJH WKH SDWLHQWnV PHGLFDO UHJLPHQ XQQHFHVVDULO\ RU LQDSSURSULDWHO\ ,GHQWLI\LQJ FRUUHODWHV RI DGKHUHQFH LV LPSRUWDQW EHFDXVH WKH\ IDFLOLWDWH WKH GHYHORSPHQW RI ULVN SURILOHV PHGLFDO UHJLPHQV DQG IXWXUH UHVHDUFK XVLQJ PDWFKHG VDPSOHV 3DWLHQW DQG IDPLO\ IDFWRUV GLVHDVHUHODWHG IDFWRUV DQG WUHDWPHQW UHJLPHQ IDFWRUV KDYH EHHQ LGHQWLILHG DV FRUUHODWHV RI DGKHUHQFH 5DSRII f 7KH PDMRULW\ RI VWXGLHV KDYH IRFXVHG H[FOXVLYHO\ RQ SDWLHQW DQG IDPLO\ IDFWRUV DQG WKHLU UHODWLRQVKLS WR DGKHUHQFH 3DWLHQWIDPLO\ FRUUHODWHV LQFOXGH GHPRJUDSKLFV FKLOG DJH SDUHQWDO HGXFDWLRQ OHYHO ZRUN VWDWXV QXPEHU RI FKLOGUHQ LQ WKH KRPH DQG VRFLRHFRQRPLF VWDWXVf NQRZOHGJH DERXW GLVHDVH FRXUVH DQG WUHDWPHQW DV ZHOO DV DGMXVWPHQW DQG FRSLQJ 3UHYLRXV UHVHDUFK KDV VKRZQ WKDW SDWLHQW DQG IDPLO\ GHPRJUDSKLF YDULDEOHV KDYH D VLJQLILFDQW LQIOXHQFH RQ FKLOG DQG PDWHUQDO DGKHUHQFH
PAGE 19

DQG UHQDO GLVHDVH DUH PRUH OLNHO\ WR EH DGKHUHQW WR WKHLU PHGLFDWLRQ UHJLPHQ FRPSDUHG WR DGROHVFHQWV $QGHUVRQ $XVODQGHU -XQJ 0LOOHU t 6DQWLDJR %HFN HW DO %RQG $LNHQ t 6RPHUYLOOH %URZQEULGJH t )LHOGLQJ *XGDV .RRFKHU t :\SOM -DFREVHQ HW DO /D*UHFD )ROODQVEHH t 6N\OHU 3DWWHUVRQ 6PLWK 5RVHQ 7UXZRUWK\ t /RZPDQ 7HEEL HW DO f %R\V WHQG WR EH OHVV DGKHUHQW WKDQ JLUOV WR UHJLPHQV IRU F\VWLF ILEURVLV DQG GLDEHWHV /RUHQ] &KULVWHQVLQ t 3LFKHUW 3DWWHUVRQ f /RZHU VRFLRHFRQRPLF VWDWXV 6(6f DQG ORZHU SDUHQWDO HGXFDWLRQ OHYHOV KDYH EHHQ UHODWHG WR QRQDGKHUHQFH WR UHJLPHQV IRU DVWKPD F\VWLF ILEURVLV GLDEHWHV DQG UHQDO GLVHDVH 3DWLHQWV LQ ODUJHU IDPLOLHV RU ZKRVH PRWKHUV ZRUN RXWVLGH RI WKH KRPH WHQG WR EH OHVV DGKHUHQW WR FDQFHU DQG F\VWLF ILEURVLV UHJLPHQV 3DWWHUVRQ 7HEEL HW DO f .QRZOHGJH DERXW WKH WUHDWPHQW UHJLPHQ LV DQ LQWHUHVWLQJ DQG LPSRUWDQW IDFWRU LQ SUHGLFWLQJ DGKHUHQFH WR WUHDWPHQW +RZHYHU VWXGLHV WKDW KDYH H[DPLQHG WKLV IDFWRU KDYH VKRZQ DQ LQFRQVLVWHQW SDWWHUQ RI UHVXOWV 6HYHUDO LQYHVWLJDWRUV KDYH UHSRUWHG WKDW SDWLHQWV ZKR DUH OHVV NQRZOHGJHDEOH DERXW WKHLU VSHFLILF FRQGLWLRQ DQG WUHDWPHQW UHJLPHQ WHQG WR EH OHVV DGKHUHQW WR UHJLPHQV IRU FDQFHU F\VWLF ILEURVLV DQG GLDEHWHV *XGDV HW DO /D*UHFD HW DO 7HEEL HW DO f ZKHUHDV NQRZOHGJH DSSHDUV WR EH XQUHODWHG WR DGKHUHQFH IRU SDWLHQWV ZLWK UHQDO GLVHDVH %HFN HW DO f 0DWHUQDO NQRZOHGJH IRU FKLOGUHQ ZLWK GLDEHWHV ZDV SRVLWLYHO\ UHODWHG WR DGKHUHQFH IRU SUHDGROHVFHQWV DQG XQUHODWHG WR DGKHUHQFH IRU DGROHVFHQWV /D*UHFD HW DO f 7KH UHODWLRQVKLS RI DGMXVWPHQW DQG FRSLQJ WR DGKHUHQFH KDV EHHQ H[DPLQHG DFURVV D YDULHW\ RI FKLOGKRRG GLVRUGHUV 3RRU SDUHQWDO FRSLQJ KDV EHHQ DVVRFLDWHG ZLWK QRQn DGKHUHQFH WR PHGLFDO UHJLPHQV IRU FKLOGUHQ ZLWK MXYHQLOH UKHXPDWRLG DUWKULWLV DQG UHQDO

PAGE 20

GLVHDVH ,QFUHDVHG SDUHQWDO GHSUHVVLRQ KDV EHHQ UHODWHG WR SRRU DGKHUHQFH WR UHQDO GLVHDVH UHJLPHQV %URZQEULGJH t )LHOGLQJ :\QQ t (FNHO f *UHDWHU SDUHQWDO DQ[LHW\ DQG SHUFHLYHG VWUHVV KDYH EHHQ DVVRFLDWHG ZLWK ORZHU DGKHUHQFH WR DQWLHSLOHSWLF PHGLFDWLRQV IRU FKLOGUHQ ZLWK HSLOHSV\ +D]]DUG f 3DUHQWDO FRPSOLDQFH EHKDYLRU LV UHODWHG WR SHUFHSWLRQV RI WKH PHGLFDWLRQVn HIILFDF\ DV ZHOO DV LWV VLGH HIIHFWV 6KRSH f $XVWLQ f VXJJHVWV WKDW DWWLWXGHV WRZDUG JLYLQJ DQWLHSLOHSWLF PHGLFDWLRQV DQG EHKDYLRUDO LQWHQWLRQV DUH LPSRUWDQW YDULDEOHV LQ WKH SUHGLFWLRQ RI SDUHQWDO FRPSOLDQFH EHKDYLRU 3DUHQWDO NQRZOHGJH DERXW WKH HIILFDF\ DQG VLGH HIIHFWV RI PHGLFDWLRQV LQIOXHQFHV SDUHQWDO DWWLWXGHV WRZDUGV PHGLFDWLRQV DQG LPSURYHV WKHLU PHGLFDWLRQJLYLQJ EHKDYLRU $XVWLQ f 3DWLHQWV ZKR ODFN SDUHQWDO PRQLWRULQJ RI UHJLPHQ SURFHGXUHV RU KDYH IDPLOLHV ZKHUH WKH UROH RI WKH SDUHQWDO PRQLWRU LV DPELJXRXV WHQG WR EH OHVV DGKHUHQW WR WKHLU SUHVFULEHG PHGLFDO UHJLPHQ 3DUHQWDO PRQLWRULQJ KDV EHHQ IRXQG WR EH D NH\ IDFWRU LQ LPSURYLQJ PHGLFDWLRQ DGKHUHQFH IRU DGROHVFHQWV ZLWK FKURQLF LOOQHVV 5DSRII f &RQGLWLRQVSHFLILF FRUUHODWHV LQFOXGH GXUDWLRQ FRXUVH V\PSWRPV DQG SHUFHLYHG VHYHULW\ RI WKH FRQGLWLRQ 'LVHDVHV ZLWK D ORQJHU GXUDWLRQ DUH UHODWHG WR SRRUHU DGKHUHQFH 3RRUHU DGKHUHQFH WR UHJLPHQV IRU GLDEHWHV MXYHQLOH UKHXPDWRLG DUWKULWLV DQG UHQDO GLVHDVH KDV EHHQ UHODWHG WR ORQJHU LOOQHVV GXUDWLRQ 7KH UHODWLRQVKLS RI GLVHDVH FRXUVH DQG DGKHUHQFH KDV QRW EHHQ VSHFLILFDOO\ H[DPLQHG IRU FKLOGUHQ ZLWK FKURQLF LOOQHVV 5DSRII f &RQGLWLRQVSHFLILF V\PSWRPV DUH DQ LPSRUWDQW IDFWRU WR H[DPLQH LQ WKHLU UHODWLRQVKLS WR DGKHUHQFH ,QFUHDVHG KHDOWK SUREOHPV DQG KRVSLWDOL]DWLRQV DUH UHODWHG WR SRRUHU DGKHUHQFH IRU SDWLHQWV ZLWK UHQDO GLVHDVH %HFN HW DO f *UHDWHU VHL]XUH DFWLYLW\ KDV EHHQ UHODWHG WR GHFUHDVHG DGKHUHQFH WR DQWLFRQYXOVDQW PHGLFDWLRQV IRU

PAGE 21

FKLOGUHQ ZLWK HSLOHSV\ +D]]DUG f 0DWHUQDO SHUFHSWLRQV RI KLJKHU GLVHDVH VHYHULW\ KDYH EHHQ DVVRFLDWHG ZLWK EHWWHU DGKHUHQFH WR DVWKPD PHGLFDWLRQV 5DGLXV HW DO f 0HGLFDWLRQ UHJLPHQUHODWHG FRUUHODWHV LQFOXGH W\SH DQG FRPSOH[LW\ FRVWV VLGH HIIHFWV DQG SHUFHLYHG HIILFDF\ RI WKH PHGLFDWLRQ UHJLPHQ 3DWLHQWV ZLWK PRUH FRPSOH[ UHJLPHQV WHQG WR EH OHVV DGKHUHQW WR WKHLU SUHVFULEHG UHJLPHQ LQ WKDW PRUH FRPSOH[ UHJLPHQV DUH PRUH RIWHQ DVVRFLDWHG ZLWK SRO\GUXJ WUHDWPHQWV ZLWK KLJKHU QXPEHUV RI PHGLFDWLRQ W\SHV DQG GRVDJHV 7KLV LV SDUWLFXODUO\ WUXH IRU SDWLHQWV ZLWK F\VWLF ILEURVLV GLDEHWHV DQG MXYHQLOH UKHXPDWRLG DUWKULWLV 7UHDWPHQW FRVWV FDQ EH UHODWHG WR QRQn DGKHUHQFH HVSHFLDOO\ IRU WKRVH SDWLHQWV DQG IDPLOLHV ZKR DUH HFRQRPLFDOO\ GLVDGYDQWDJHG 7KRVH PHGLFDO UHJLPHQV WKDW SURGXFH D JUHDWHU QXPEHU RI QHJDWLYH VLGH HIIHFWV WHQG WR EH UHODWHG WR SRRUHU DGKHUHQFH WR WKH UHJLPHQ 3DWLHQW DQG SDUHQWDO SHUFHSWLRQV RI PHGLFDWLRQ HIILFDF\ DUH HVSHFLDOO\ UHODWHG WR DGKHUHQFH +LJKHU OHYHOV RI SHUFHLYHG EHQHILWV UDWHG E\ SDWLHQWV DQG SDUHQWV DUH DVVRFLDWHG ZLWK EHWWHU DGKHUHQFH WR UHJLPHQV IRU DVWKPD DQG GLDEHWHV %REURZ $YUXVNLQ t 6LOOHU %RQG HW DO 0F&DXO *ODVJRZ t 6FKDIHU 5DGLXV HW DO f ,Q VXPPDU\ LPSRUWDQW FRUUHODWHV RI DGKHUHQFH IRU FKLOGUHQ DQG DGROHVFHQWV ZLWK FKURQLF LOOQHVV LQFOXGH SDWLHQWIDPLO\ IDFWRUV GHPRJUDSKLFV NQRZOHGJH DGMXVWPHQW DQG FRSLQJ SDUHQWDO PRQLWRULQJf GLVHDVH IDFWRUV GXUDWLRQ FRXUVH V\PSWRPV SHUFHLYHG VHYHULW\f DQG UHJLPHQ IDFWRUV W\SH DQG FRPSOH[LW\ FRVWV VLGH HIIHFWV HIILFDF\f 7KHVH VXPPDUL]HG ILQGLQJV GHPRQVWUDWH WKDW FRUUHODWHV RI DGKHUHQFH WR PHGLFDO UHJLPHQV DUH FULWLFDO LQ LPSURYLQJ DGKHUHQFH 7KH PDMRULW\ RI VWXGLHV FRQGXFWHG WR LGHQWLI\ FRUUHODWHV RI DGKHUHQFH WR PHGLFDO UHJLPHQV IRU FKLOGUHQ ZLWK FKURQLF GLVHDVH KDYH IRFXVHG RQ WKH IROORZLQJ DVWKPD FDQFHU F\VWLF ILEURVLV GLDEHWHV DQG UHQDO GLVHDVH 6WXGLHV H[DPLQLQJ

PAGE 22

WKH VSHFLILF FRUUHODWHV RI DGKHUHQFH IRU FKLOGUHQ DQG DGROHVFHQWV ZLWK HSLOHSV\ DUH OLPLWHG LQ QXPEHU DQG UHTXLUH IXUWKHU H[DPLQDWLRQ $GKHUHQFH WR WKH 5HJLPHQ (SLOHSV\ 7KH YDVW PDMRULW\ RI SHRSOH ZLWK HSLOHSV\ EHJLQ KDYLQJ WKHLU VHL]XUHV EHIRUH WKH DJH RI DQG PRUH WKDQ b RI HSLOHSV\ FDVHV EHJLQ LQ FKLOGKRRG 7KXV PRVW SHRSOH H[SHULHQFH WKHLU ILUVW VHL]XUH DW D WLPH WKDW LV FULWLFDO WR WKH DFTXLVLWLRQ DQG GHYHORSPHQW RI EDVLF FRJQLWLYH DQG VRFLDO FRPSHWHQFLHV ZKLFK DUH FUXFLDO IRU ORQJWHUP DFDGHPLF LQWHUSHUVRQDO DQG YRFDWLRQDO DGMXVWPHQW 'XH WR WKH ORQJWHUP QDWXUH RI HSLOHSV\ DQG LWV UHODWHG PHGLFDWLRQ UHJLPHQ PHGLFDWLRQ DGKHUHQFH WR DQWLHSLOHSWLF GUXJV $('Vf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t 0DLPDQ 'DYLV 5DSRII t &KULVWRSKHUVRQ f

PAGE 23

$ GHPRJUDSKLF IDFWRU JHQHUDOO\ DVVRFLDWHG ZLWK SDUHQWDO QRQDGKHUHQFH LV ORZHU VRFLRHFRQRPLF VWDWXV $GGLWLRQDOO\ HWKQLF PLQRULW\ SDWLHQWV KDYH EHHQ IRXQG WR EH PRUH OLNHO\ WR EH QRQDGKHUHQW WR WKHLU FKLOGnV PHGLFDO UHJLPHQ %HFNHU t 0DLPDQ 5DSRII t &KULVWRSKHUVRQ f &RQGXFWLQJ UHVHDUFK RQ DGKHUHQFH ZLWK SHGLDWULF VHL]XUH SDWLHQWV LV PDGH GLIILFXOW E\ WKH PHWKRGRORJLFDO SUREOHPV LQYROYHG LQ DVVHVVLQJ SDWLHQWnV DGKHUHQFH $OWKRXJK EORRG OHYHOV RI DQWLFRQYXOVDQW PHGLFDWLRQV FDQ EH REWDLQHG LW LV GLIILFXOW WR WUDQVODWH EORRG OHYHOV LQWR FRPSDUDEOH PHDVXUHV RI FRPSOLDQFH IRU SDWLHQWV RQ GLIIHUHQW PHGLFDWLRQV DQG GRVHV 7KHUDSHXWLF UDQJHV YDU\ IRU GLIIHUHQW PHGLFDWLRQV DQG YDU\ DFFRUGLQJ WR GRVDJH WLPH VLQFH WKH PHGLFDWLRQ ZDV LQJHVWHG DQG LQGLYLGXDO PHWDEROLF UDWHV DQG KRUPRQDO FKDQJHV ZKLFK DUH HVSHFLDOO\ VDOLHQW IRU DGROHVFHQW IHPDOHVf )XUWKHUPRUH IRU SDWLHQWV WDNLQJ PRUH WKDQ RQH PHGLFDWLRQ WKHUH DUH PHGLFDWLRQ LQWHUDFWLRQV ZLWK VRPH PHGLFDWLRQV DOWHULQJ WKH EORRG OHYHO FRQFHQWUDWLRQV RI RWKHU PHGLFDWLRQV %HFDXVH RI WKHVH FRPSOLFDWLRQV VWXGLHV H[DPLQLQJ SHGLDWULF SDWLHQW DGKHUHQFH WR DQWLFRQYXOVDQW PHGLFDWLRQ UHJLPHQV KDYH WHQGHG WR DVVHVV RQO\ RQH PHGLFDWLRQ OHYHO RU KDYH XVHG JURXS GLIIHUHQFHV LQ EORRG OHYHOV WR DVVHVV WKH YDOLGLW\ RI RWKHU FRPSOLDQFH PHDVXUHV VHOI UHSRUW SLOO FRXQWVf :LWK WKHVH OLPLWDWLRQV HVWLPDWHV RI SDUHQWDO DGKHUHQFH WR DQWLFRQYXOVDQW PHGLFDWLRQ UHJLPHQV IRU SHGLDWULF SDWLHQWV UDQJHV IURP b WR b =\VVHW HW DO f 7KHVH HVWLPDWHV DUH JHQHUDOO\ FRQVLVWHQW ZLWK DGKHUHQFH UDWHV IURP RWKHU SHGLDWULF SRSXODWLRQV ZKHUH D ORQJWHUP PHGLFDWLRQ KDV EHHQ SUHVFULEHG 6DFNHWW t +D\QHV f 2QO\ WKUHH VWXGLHV KDYH DVVHVVHG IDFWRUV UHODWHG WR SDUHQWDO PHGLFDWLRQ DGKHUHQFH ZLWK SHGLDWULF VHL]XUH SDWLHQWV )UHLGPDQ HW DO 3HWHUVRQ HW DO f 7KH VWXGLHV

PAGE 24

DVVHVVHG DW OHDVW VRPH IDFWRUV IURP WKH GHPRJUDSKLF LOOQHVV WUHDWPHQW UHJLPHQ DQG DWWLWXGLQDO GRPDLQV +RZHYHU WKH 3HWHUVRQ HW DO VWXG\ IRFXVHG SULPDULO\ RQ DGXOWV ZLWK RQO\ DGROHVFHQWV LQ WKH VDPSOH 'HPRJUDSKLF YDULDEOHV ZHUH QRW GHPRQVWUDWHG WR EH DVVRFLDWHG WR DGKHUHQFH LQ WKH SHGLDWULF VHL]XUH GLVRUGHU SRSXODWLRQV LQ WKH WZR VWXGLHV )UHLGPDQ HW DO 3HWHUVRQ HW DO f ,Q WHUPV RI LOOQHVV YDULDEOHV 3HWHUVRQ HW DO f IRXQG WKDW SDWLHQWV H[SHULHQFLQJ PRUH VHL]XUHV LQ WKH SDVW \HDU PRUH UHFHQW VHL]XUHV RU KDYLQJ WRQLFFORQLF VHL]XUHV ZHUH PRUH OLNHO\ WR EH DGKHUHQW 7KH RQO\ WUHDWPHQWUHODWHG YDULDEOH UHODWHG WR DGKHUHQFH IRU FKLOGUHQ ZKR KDG DQ DFWLYH UROH LQ WDNLQJ WKHLU PHGLFDWLRQ SDUWLFXODUO\ DGROHVFHQWV ZDV SHUFHLYHG SHUVRQDO LQGHSHQGHQFH )UHLGPDQ HW DO f 5HVHDUFK LQ FKLOGUHQ ZLWK RWKHU PHGLFDO FRQGLWLRQV KDV DOVR VKRZQ WKDW FKLOGUHQ ZLWK JUHDWHU SHUVRQDO IUHHGRP RU IHZHU EHKDYLRUDO UHVWULFWLRQVf ZHUH PRUH OLNHO\ WR EH DGKHUHQW ZLWK WKHLU PHGLFDWLRQ UHJLPHQ 3HWHUVRQ HW DO f IRXQG WKDW VDWLVIDFWLRQ ZLWK WKH SK\VLFLDQnV H[SODQDWLRQ ZDV XQUHODWHG WR DGKHUHQFH +RZHYHU LQ VWXGLHV ZLWK RWKHU SRSXODWLRQV SDWLHQW VDWLVIDFWLRQ DQG JRRG SDWLHQWGRFWRU FRPPXQLFDWLRQ KDYH EHHQ UHODWHG WR JRRG DGKHUHQFH 3HWHUVRQ HW DO f DOVR IRXQG WKDW ZRUU\ DERXW RQHnV RZQ KHDOWK ZDV UHODWHG WR VHOIUHSRUWHG DGKHUHQFH 6KRSH UHYLHZHG D VWXG\ WKDW FORVHO\ IROORZHG WKH +HDOWK %HOLHI 0RGHO +%0f IUDPHZRUN IRU SHGLDWULF HSLOHSV\ SDWLHQWV 6KRSH f 3DUHQWV RI D UDQGRP VDPSOH RI SDWLHQWV GUDZQ IURP DQ XUEDQ PHGLFDO FHQWHU SHGLDWULF VHL]XUH FOLQLF ZHUH VHHQ DW WKHLU UHJXODU DSSRLQWPHQW DQG LQWHUYLHZHG IURP -DQXDU\ RI WKURXJK 0D\ RI 4XHVWLRQV DVNHG LQFOXGHG YDULDEOHV IURP WKH +%0 DV ZHOO DV D ZLGH UDQJH RI PHGLFDO SV\FKRVRFLDO DQG GHPRJUDSKLF YDULDEOHV 6HUXP DQWLHSLOHSWLF OHYHOV ZHUH WDNHQ DW WKH FOLQLF YLVLW RQ RI WKH VWXG\ SDUWLFLSDQWV 'DWD SUHVHQWHG ZHUH EDVHG RQ SDWLHQWV

PAGE 25

3DWLHQWV ZHUH MXGJHG WR EH FRPSOLDQW RU QRQFRPSOLDQW E\ DQ H[SHULHQFHG FOLQLFLDQnV UHODWLQJ WKH SUHVFULEHG GRVDJH RI PHGLFDWLRQ WR WKH REWDLQHG EORRG OHYHO 7KLV SURFHVV ZDV XVHG UDWKHU WKDQ WKHUDSHXWLF EORRG OHYHOV EHFDXVH VRPH SDWLHQWV KDG OHVV WKDQ WKHUDSHXWLF GRVDJHV SUHVFULEHG 7KH SDWLHQW VDPSOH ZDV b IHPDOH DQG b $IULFDQ $PHULFDQ ZLWK D PHDQ DJH RI \HDUV )RU b RI WKH FKLOGUHQ VHL]XUH GLVRUGHU ZDV WKH RQO\ PHGLFDO SUREOHP LGHQWLILHG $JH RI RQVHW UDQJHG IURP ELUWK WR \HDUV DQG GXUDWLRQ RI WKH VHL]XUH GLVRUGHU UDQJHG IURP PRQWKV WR \HDUV )RUW\HLJKW SHUFHQW RI WKH FKLOGUHQ ZHUH SUHVFULEHG RQH PHGLFDWLRQ b ZHUH SUHVFULEHG WZR DQG b ZHUH SUHVFULEHG WKUHH 3KHQREDUELWDO ZDV SUHVFULEHG IRU PRVW RI WKH SDWLHQWV 7KH DVVHVVPHQW RI FRPSOLDQFH E\ ORRNLQJ DW EORRG OHYHOV UHYHDOHG WKDW b RI WKH SDUHQWV JDYH WKHLU FKLOGUHQ WKHLU PHGLFDWLRQ DV SUHVFULEHG b ZHUH JLYLQJ OHVV WKDQ WKH SUHVFULEHG GRVH DQG b ZHUH JLYLQJ QR PHGLFDWLRQ )RUW\WKUHH SHUFHQW RI WKH SDWLHQWV ZHUH QRW UHFHLYLQJ WKHLU PHGLFDWLRQ DV SUHVFULEHG 5HVXOWV LQGLFDWHG WKDW SDUHQWV ZKR UHSRUWHG DW OHDVW RQH RFFXUUHQFH RI UXQQLQJ RXW RI WKHLU FKLOGnV PHGLFDWLRQ IRU VHYHUDO GD\V ZHUH PRUH OLNHO\ WR KDYH FKLOGUHQ ZKRVH EORRG OHYHOV UHYHDOHG D GLVFUHSDQF\ EHWZHHQ WKH REWDLQHG DQG SUHVFULEHG GRVDJH S f DQG SDUHQWV ZKR UHSRUWHG WKDW WKH\ PLVVHG JLYLQJ WKHLU FKLOG D PHGLFDWLRQ PRUH WKDQ RQFH D ZHHN ZHUH OHVV OLNHO\ WR EH FRPSOLDQW DV PHDVXUHG E\ EORRG OHYHOV S f PRWLYDWLRQDO IDFWRUV RI WKH +%0f 5HFHQF\ RI VHL]XUHV ZDV VLJQLILFDQWO\ SRVLWLYHO\ FRUUHODWHG ZLWK FRPSOLDQFH S f YDOXH RI LOOQHVV WKUHDW UHGXFWLRQ RI WKH +%0f &KLOGnV HQUROOPHQW LQ VFKRRO ZDV VLJQLILFDQWO\ DQG QHJDWLYHO\ FRUUHODWHG ZLWK FRPSOLDQFH

PAGE 26

S f 3DUHQWV ZKR UHSRUWHG WKDW WKHLU FKLOG GLG QRW KDYH DQ\ DGGLWLRQDO PHGLFDO SUREOHPV ZHUH PRUH FRPSOLDQW S f ,Q VXPPDU\ IDFWRUV WKDW DSSHDU WR LQIOXHQFH SDUHQWDO QRQDGKHUHQFH WR DQWLn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nV DQWLn HSLOHSWLF PHGLFDWLRQ UHJLPHQ 3DUHQWV ZKR SHUFHLYHG DQ LQFUHDVHG OHYHO RI GLVHDVH VHYHULW\ PRUH VHYHUH VHL]XUH HYHQWV ZLWK JUHDWHU IUHTXHQF\ DQG LQWHQVLW\f ZHUH PRUH DGKHUHQW WR WKHLU FKLOGnV PHGLFDWLRQ UHJLPHQ 7KH FXUUHQW VWXG\ H[DPLQHG IDFWRUV WKDW DUH UHODWHG WR PHGLFDWLRQ DGKHUHQFH LQ SHGLDWULF HSLOHSV\ SRSXODWLRQV VSHFLILFDOO\ ZLWK \RXQJHU FKLOGUHQ ZLWK HSLOHSV\ 0DWHUQDO (PRWLRQDO )XQFWLRQLQJ 'HSUHVVLRQ LV D KLJKO\ SUHYDOHQW PHQWDO GLVRUGHU HVSHFLDOO\ DPRQJ WKRVH ZRPHQ RI FKLOGEHDULQJ DJH 'RZQH\ DQG &R\QH f IRXQG WKDW WKHUH LV D KLJK UDWH RI VHULRXV SV\FKRORJLFDO SUREOHPV LQ FKLOGUHQ RI GHSUHVVHG PRWKHUV DQG WKDW GHSUHVVHG PRWKHUV KDYH D KRVW RI GLIILFXOWLHV DV SDUHQWV 0XUUD\ 6LQFODLU &RRSHU 'XFRXPDX 7XUQHU DQG 6WHLQ f UHSRUWHG WKDW H[SRVXUH WR PDWHUQDO GHSUHVVLRQ LQ WKH HDUO\ SRVWSDUWXP PRQWKV PD\ KDYH DQ HQGXULQJ LQIOXHQFH RQ FKLOGKRRG SV\FKRORJLFDO DGMXVWPHQW &XPPLQJV DQG 'DYLHV f LQGLFDWHG WKDW FKLOGUHQ RI GHSUHVVHG SDUHQWV DUH DW LQFUHDVHG ULVN IRU WKH GHYHORSPHQW RI SV\FKRSDWKRORJ\ DQG WKDW WKH\ DUH WZR WR ILYH WLPHV PRUH OLNHO\ WR GHYHORS EHKDYLRU SUREOHPV WKDQ DUH FKLOGUHQ RI QRUPDO SDUHQWV 'HSUHVVHG

PAGE 27

PRWKHUV YLHZ WKHLU SDUHQWLQJ UROH OHVV SRVLWLYHO\ WKDQ QRQGHSUHVVHG PRWKHUV H[SHULHQFH QHJDWLYH IHHOLQJV WRZDUGV SDUHQWLQJ UROHV DQG KDYH IHHOLQJV RI UHMHFWLRQ DQG KRVWLOLW\ WRZDUGV WKHLU FKLOG 'RZQH\ t &R\QH f 'HSUHVVHG PRWKHUV DOVR WHQG WR GHPRQVWUDWH GLIILFXOWLHV LQWHUDFWLQJ ZLWK WKHLU FKLOGUHQ DQG GLVSOD\LQJ SRVLWLYH HPRWLRQV WRZDUGV WKHLU FKLOGUHQ +RGHV DQG *DUUDOGD f H[DPLQHG WKH HIIHFWV RI PDWHUQDO H[SUHVVHG HPRWLRQ RQ WKH SV\FKRORJLFDO DGMXVWPHQW RI FKLOGUHQ ZLWK HSLOHSV\ 0DWHUQDO H[SUHVVHG HPRWLRQ DQG ULVN IRU SV\FKLDWULF GLVRUGHUV IRU FKLOGUHQ DQG PRWKHUVf ZHUH DVVHVVHG 7KH UHVHDUFKHUV IRXQG WKDW PRWKHUV VKRZHG VLJQLILFDQWO\ PRUH HPRWLRQDO RYHULQYROYHPHQW DQG D WUHQG IRU PRUH KRVWLOLW\ WRZDUGV WKHLU FKLOGUHQ ZLWK HSLOHSV\ WKDQ WKH\ GLVSOD\HG WRZDUGV VLEOLQJ FRQWUROV 0DWHUQDO HPRWLRQDO RYHULQYROYHPHQW ZDV XQUHODWHG WR FKLOG EHKDYLRU SUREOHPV +LJK OHYHOV RI FULWLFLVP DQG KRVWLOLW\ DOVR ZHUH VKRZQ WR FRYDU\ ZLWK FKLOG EHKDYLRU SUREOHPV +RGHV t *DUUDOGD f 7KH HIIHFWV RI PDWHUQDO H[SUHVVHG HPRWLRQ RQ DGKHUHQFH WR PHGLFDWLRQ ZHUH QRW DVVHVVHG +RZHYHU WKH UHVXOWV VXJJHVW WKDW PDWHUQDO H[SUHVVHG HPRWLRQ LV UHODWHG WR QHJDWLYH SV\FKRORJLFDO IXQFWLRQLQJ LQ FKLOGUHQ ZLWK HSLOHSV\ 7KXV LW VHHPV KLJKO\ SODXVLEOH WKDW PDWHUQDO H[SUHVVHG HPRWLRQ PD\ KDYH D VXEVWDQWLDO LPSDFW RQ DGKHUHQFH WR DQWLHSLOHSWLF PHGLFDWLRQV IRU FKLOGUHQ ZLWK HSLOHSV\ 2WHUR DQG +RGHV f DVVHVVHG WKH UHODWLRQVKLS EHWZHHQ PDWHUQDO H[SUHVVHG HPRWLRQ DQG DGKHUHQFH WR WUHDWPHQW LQ FKLOGUHQ ZLWK HSLOHSV\ 7KH UHVHDUFKHUV UHSRUWHG WKDW JRRG WUHDWPHQW FRPSOLDQFH ZDV UHODWHG WR OHVV PDWHUQDO KRVWLOLW\ DQG FULWLFLVP IRU HSLOHSWLF FKLOGUHQ DQG WKHLU PRWKHUV &KLOGUHQ DQG PRWKHUV LQ WKH JRRG FRPSOLDQFH JURXS KDG IHZHU UHSRUWHG SV\FKLDWULF V\PSWRPV 3RRU WUHDWPHQW FRPSOLDQFH IRU FKLOGUHQ ZLWK HSLOHSV\ ZDV DVVRFLDWHG ZLWK JUHDWHU PDWHUQDO SV\FKRORJLFDO GLVWXUEDQFH &KLOGUHQ ZLWK D

PAGE 28

JUHDWHU QXPEHU RI EHKDYLRUDO SUREOHPV ZHUH OHVV FRPSOLDQW $ UHODWLRQVKLS EHWZHHQ PDWHUQDO GHSUHVVLRQ DQG SRRUHU FRPSOLDQFH ZDV UHSRUWHG 2WHUR t +RGHV f 7KH UHODWLRQVKLS EHWZHHQ PDWHUQDO GHSUHVVLRQ DQG EHKDYLRU SUREOHPV KDV EHHQ LQYHVWLJDWHG LQWHQVLYHO\ DFURVV D YDULHW\ RI SHGLDWULF PHGLFDO SRSXODWLRQV 3RRU PDWHUQDO HPRWLRQDO IXQFWLRQLQJ DSSHDUV WR H[HUW D FRQVLVWHQW QHJDWLYH LPSDFW RQ FKLOGUHQ ZLWK GLVDELOLWLHV DQG WKHLU KHDOWK\ VLEOLQJV 5HVHDUFK ILQGLQJV GHPRQVWUDWH D FOHDU UHODWLRQVKLS EHWZHHQ PDWHUQDO HPRWLRQDO IXQFWLRQLQJ DQG FKLOGKRRG EHKDYLRU SUREOHPV 6SHFWRU HW DO f IRXQG WKDW LQFUHDVHG PDWHUQDO VWUHVV DQG DQ[LHW\ OHYHOV ZHUH UHODWHG WR DQ LQFUHDVH LQ VHL]XUH IUHTXHQF\ DQG VHYHULW\ +D]]DUG f UHSRUWHG WKDW LQFUHDVHG OHYHOV RI SDUHQWDO DQ[LHW\ DQG SHUFHLYHG VWUHVV ZHUH UHODWHG WR ORZHU DGKHUHQFH WR DQWLHSLOHSWLF PHGLFDWLRQV IRU FKLOGUHQ ZLWK D VHL]XUH GLVRUGHU ,W LV IHDVLEOH WKDW D WUXH UHODWLRQVKLS H[LVWV EHWZHHQ SRRU PDWHUQDO DGMXVWPHQW DQG GLIILFXOW\ ZLWK DGKHUHQFH WR PHGLFDWLRQ UHJLPHQV IRU \RXQJHU FKLOGUHQ :H H[DPLQHG WKH UHODWLRQVKLS EHWZHHQ PDWHUQDO HPRWLRQDO GLVWUHVV YDULDEOHV PDWHUQDO GHSUHVVLRQ DQG PDWHUQDO VWUHVVf DQG PDWHUQDO DGKHUHQFH WR DQWLHSLOHSWLF GUXJV SLOO FRXQW SHUFHQWDJHf DQG FKLOG KHDOWK VWDWXV YDULDEOHV VHL]XUH IUHTXHQF\ VHL]XUHUHODWHG HPHUJHQF\ URRP YLVLWVf IRU FKLOGUHQ GLDJQRVHG ZLWK HSLOHSV\

PAGE 29

&+$37(5 $,06 $1' 678'< -867,),&$7,216 3XUSRVH ,Q VXPPDU\ HSLOHSV\ LV D FKURQLF LOOQHVV WKDW W\SLFDOO\ EHJLQV LQ HDUO\ FKLOGKRRG 3RRU DGKHUHQFH WR WKH PHGLFDWLRQ UHJLPHQ FDQ UHVXOW LQ LQFUHDVHG VHL]XUH IUHTXHQF\ DQG VHYHULW\ DQG SRVVLEO\ GHDWK
PAGE 30

DQG WKH IROORZLQJ IDFWRUV PDWHUQDO HPRWLRQDO GLVWUHVV PDWHUQDO GHSUHVVLRQ PDWHUQDO VWUHVVf GHPRJUDSKLFV IDPLO\ DQQXDO LQFRPH UDFH PDWHUQDO HGXFDWLRQDO OHYHOf FRQGLWLRQUHODWHG DQG UHJLPHQUHODWHG YDULDEOHV GLVHDVH W\SH GLVHDVH GXUDWLRQ W\SH DQG FRPSOH[LW\ RI WKH PHGLFDWLRQ UHJLPHQ FRQFRUGDQFH EHWZHHQ SUHVFULEHG UHJLPHQ DQG UHSRUWHG UHJLPHQf DQG PDWHUQDO SHUFHSWLRQV RI WKH WUHDWPHQW UHJLPHQ SHUFHLYHG HIILFDF\ SHUFHLYHG NQRZOHGJH SHUFHLYHG GLIILFXOW\f 6WXG\ 0RGHO 7KH SXUSRVH RI WKH FXUUHQW VWXG\ ZDV WR WHVW D RQHPRQWK SUHGLFWLYH PRGHO RI PDWHUQDO DGKHUHQFH DQG FKLOG KHDOWK VWDWXV WR SUHVFULEHG DQWLHSLOHSWLF PHGLFDWLRQ LQ D VDPSOH RI \HDU ROG FKLOGUHQ GLDJQRVHG ZLWK HSLOHSV\ IRFXVLQJ RQ WKH LQIOXHQFH RI IRXU SUHGLFWRU IDFWRUV 7KH IRXU SUHGLFWRU IDFWRUV DUH GHILQHG DV f PDWHUQDO HPRWLRQDO GLVWUHVV f PDWHUQDO SHUFHSWLRQV RI WKH WUHDWPHQW UHJLPHQ f FRQGLWLRQUHODWHG DQG UHJLPHQUHODWHG YDULDEOHV DQG f GHPRJUDSKLF YDULDEOHV 7KH WKUHH RXWFRPH YDULDEOHV DUH GHILQHG DV f VHL]XUH IUHTXHQF\ f VHL]XUHUHODWHG HPHUJHQF\ URRP YLVLWV DQG f SLOO FRXQW SHUFHQWDJH ZKHUH f DQG f DVVHVV FKLOG KHDOWK VWDWXV DQG f DVVHVVHV PDWHUQDO DGKHUHQFH 7KH FXUUHQW VWXG\ H[WHQGV HDUOLHU UHVHDUFK LQ VHYHUDO ZD\V )LUVW ZH ZLOO LQWHJUDWH WKH IRXU NH\ VHWV RI SUHGLFWRU IDFWRUV LGHQWLILHG LQ SUHYLRXV UHVHDUFK LQ D XQLI\LQJ SUHGLFWLYH IUDPHZRUN 0RUHRYHU ZH VWUHQJWKHQ WKH PRGHO E\ H[DPLQLQJ WKH UROHV RI WKH IRXU SUHGLFWRU IDFWRUV DW 7LPH DVVHVVPHQWf LQ SUHGLFWLQJ VXEVHTXHQW 7LPH PRQWK IROORZXSf DGKHUHQFH DQG KHDOWK VWDWXV FRQWUROOLQJ IRU 7LPH PRQWK IROORZXSf SUHGLFWRU IDFWRUV VHH )LJXUH f 7KLV FRQWUROV IRU WKH SRVVLELOLW\ WKDW WKH SUHGLFWLYH UROH RI DGKHUHQFH PD\ EH DIIHFWHG E\ PDWHUQDO HPRWLRQDO GLVWUHVV PDWHUQDO SHUFHSWLRQV RI WKH WUHDWPHQW UHJLPHQ FRQGLWLRQUHODWHG DQG UHJLPHQUHODWHG YDULDEOHV

PAGE 31

DQG GHPRJUDSKLF YDULDEOHV DW 7LPH 7KH SUHVHQW VWXG\ H[DPLQHG WKH IROORZLQJ VSHFLILF K\SRWKHVHV +\SRWKHVHV 0DWHUQDO (PRWLRQDO 'LVWUHVV $GKHUHQFH DQG &KLOG +HDOWK 6WDWXV ,W ZDV K\SRWKHVL]HG WKDW 7LPH PDWHUQDO HPRWLRQDO GLVWUHVV ZRXOG EH QHJDWLYHO\ UHODWHG WR 7LPH DGKHUHQFH WR WKH FKLOGnV PHGLFDWLRQ UHJLPHQ DQG SRVLWLYHO\ ZLWK FKLOG KHDOWK VWDWXV D 7LPH PDWHUQDO GHSUHVVLRQ VFRUHV ZRXOG EH QHJDWLYHO\ DVVRFLDWHG ZLWK 7LPH DGKHUHQFH WR WKH FKLOGnV PHGLFDWLRQ UHJLPHQ DQG SRVLWLYHO\ ZLWK KHDOWK VWDWXV E 7LPH PDWHUQDO VWUHVV VFRUHV ZRXOG EH QHJDWLYHO\ DVVRFLDWHG ZLWK 7LPH DGKHUHQFH WR WKH FKLOGnV PHGLFDWLRQ UHJLPHQ DQG SRVLWLYHO\ ZLWK KHDOWK VWDWXV 0DWHUQDO 3HUFHSWLRQV RI WKH 7UHDWPHQW 5HJLPHQ $GKHUHQFH DQG &KLOG +HDOWK 6WDWXV :H SUHGLFWHG WKDW 7LPH UDWLQJV RI PDWHUQDO SHUFHSWLRQV RI WKH WUHDWPHQW UHJLPHQ SHUFHLYHG NQRZOHGJH SHUFHLYHG GLIILFXOW\ DQG SHUFHLYHG HIILFDF\f DQG 7LPH PDWHUQDO GLVWUHVV DQG 7LPH PDWHUQDO SHUFHSWLRQV RI WKH WUHDWPHQW UHJLPHQ ZRXOG EH DVVRFLDWHG ZLWK 7LPH DGKHUHQFH WR WKH FKLOGfV PHGLFDWLRQ UHJLPHQ DQG FKLOG KHDOWK VWDWXV D 0DWHUQDO UDWLQJV RI SHUFHLYHG NQRZOHGJH UHJDUGLQJ WKH FKLOGnV GLVHDVH DQG PHGLFDWLRQ UHJLPHQ DW 7LPH ZRXOG EH SRVLWLYHO\ DVVRFLDWHG ZLWK 7LPH DGKHUHQFH WR WKH FKLOGnV PHGLFDWLRQ UHJLPHQ DQG QHJDWLYHO\ ZLWK KHDOWK VWDWXV E 0DWHUQDO UDWLQJV RI SHUFHLYHG GLIILFXOW\ LQ JLYLQJ WKH FKLOG WKH SUHVFULEHG PHGLFDWLRQ DW 7LPH ZRXOG EH QHJDWLYHO\ DVVRFLDWHG ZLWK 7LPH DGKHUHQFH WR WKH FKLOGnV PHGLFDWLRQ UHJLPHQ DQG SRVLWLYHO\ ZLWK KHDOWK VWDWXV F 0DWHUQDO UDWLQJV RI SHUFHLYHG HIILFDF\ UHJDUGLQJ WKH FKLOGnV PHGLFDWLRQ UHJLPHQ DW 7LPH ZRXOG EH SRVLWLYHO\ DVVRFLDWHG ZLWK 7LPH DGKHUHQFH WR WKH FKLOGnV PHGLFDWLRQ UHJLPHQ DQG QHJDWLYHO\ ZLWK KHDOWK VWDWXV G 0DWHUQDO UDWLQJV RI PDWHUQDO GLVWUHVV DQG PDWHUQDO SHUFHSWLRQV RI WKH WUHDWPHQW UHJLPHQ DW 7LPH ZRXOG EH DVVRFLDWHG ZLWK 7LPH DGKHUHQFH WR WKH FKLOGnV PHGLFDWLRQ UHJLPHQ DQG KHDOWK VWDWXV &RQGLWLRQ DQG 5HJLPHQUHODWHG 9DULDEOHV $GKHUHQFH DQG &KLOG +HDOWK 6WDWXV :H SUHGLFWHG WKDW 7LPH FRQGLWLRQ DQG UHJLPHQUHODWHG YDULDEOHV ZRXOG EH DVVRFLDWHG ZLWK 7LPH DGKHUHQFH WR WKH FKLOGfV PHGLFDWLRQ UHJLPHQ DQG FKLOG KHDOWK VWDWXV D $ WRQLFFORQLF VHL]XUH GLDJQRVLV ZRXOG EH SRVLWLYHO\ UHODWHG WR DGKHUHQFH WR WKH FKLOGnV PHGLFDWLRQ UHJLPHQ DQG KHDOWK VWDWXV E 7LPH VLQFH GLDJQRVLV DW 7LPH ZRXOG EH QHJDWLYHO\ DVVRFLDWHG ZLWK 7LPH DGKHUHQFH WR WKH FKLOGfV PHGLFDWLRQ UHJLPHQ DQG SRVLWLYHO\ ZLWK KHDOWK VWDWXV

PAGE 32

F &RPSOH[LW\ RI WKH PHGLFDWLRQ UHJLPHQ KLJKHU GRVHV PXOWLSOH PHGLFDWLRQV PRUH IUHTXHQW GRVHVf DW 7LPH ZRXOG EH QHJDWLYHO\ DVVRFLDWHG ZLWK 7LPH DGKHUHQFH WR WKH FKLOGnV PHGLFDWLRQ UHJLPHQ DQG SRVLWLYHO\ ZLWK KHDOWK VWDWXV G 3HUFHQWDJH RI FRQFRUGDQFH EHWZHHQ WKH SK\VLFLDQSUHVFULEHG PHGLFDWLRQ UHJLPHQ DQG VHOIUHSRUWHG UHJLPHQ DW 7LPH ZRXOG EH SRVLWLYHO\ DVVRFLDWHG ZLWK 7LPH DGKHUHQFH WR WKH FKLOGnV PHGLFDWLRQ UHJLPHQ DQG QHJDWLYHO\ ZLWK KHDOWK VWDWXV 'HPRJUDSKLF &KDUDFWHULVWLFV $GKHUHQFH DQG &KLOG +HDOWK 6WDWXV :H SUHGLFWHG WKDW 7LPH GHPRJUDSKLF FKDUDFWHULVWLFV ZRXOG EH SRVLWLYHO\ UHODWHG WR 7LPH DGKHUHQFH WR WKH FKLOGfV PHGLFDWLRQ UHJLPHQ DQG QHJDWLYHO\ ZLWK FKLOG KHDOWK VWDWXV D 7RWDO KRXVHKROG LQFRPH DW 7LPH ZRXOG EH SRVLWLYHO\ DVVRFLDWHG ZLWK 7LPH DGKHUHQFH WR WKH FKLOGnV PHGLFDWLRQ UHJLPHQ DQG QHJDWLYHO\ ZLWK KHDOWK VWDWXV E 0DWHUQDO HGXFDWLRQDO OHYHO DW 7LPH ZRXOG EH SRVLWLYHO\ DVVRFLDWHG ZLWK 7LPH DGKHUHQFH WR WKH FKLOGnV PHGLFDWLRQ UHJLPHQ DQG QHJDWLYHO\ ZLWK KHDOWK VWDWXV

PAGE 33

&+$37(5 0(7+2'6 6WXG\ 'HVLJQ 7KH FXUUHQW VWXG\ LV D SURVSHFWLYH VWXG\ XVLQJ D WLPH ODJ GHVLJQ LQ ZKLFK ZH H[DPLQHG WKH LQIOXHQFH RI SUHGLFWRU YDULDEOHV DW 7LPH RQ RXWFRPH YDULDEOHV DW 7LPH ZKLOH FRQWUROOLQJ IRU SUHGLFWRU YDULDEOHV DW 7LPH VHH )LJXUH f 7 LP H 7 LP H )LJXUH 6WXG\ 0RGHO

PAGE 34

$IWHU REWDLQLQJ LQIRUPHG FRQVHQW PRWKHUV ZHUH JLYHQ D EDFNJURXQG TXHVWLRQQDLUH ZKLFK LQFOXGHG TXHVWLRQV DERXW FRQGLWLRQ DQG UHJLPHQUHODWHG YDULDEOHVf D TXHVWLRQQDLUH DVVHVVLQJ PDWHUQDO VWUHVV DQG D TXHVWLRQQDLUH DVVHVVLQJ GHSUHVVLYH V\PSWRPDWRORJ\ WR FRPSOHWH 7KH TXHVWLRQQDLUHV ZHUH FRPSOHWHG GXULQJ D UHJXODUO\ VFKHGXOHG SHGLDWULF QHXURORJ\ FOLQLF YLVLW 3DUWLFLSDQWV ZHUH DVNHG WR SDUWLFLSDWH LQ D RQH PRQWK IROORZXS SKRQH FDOO $SSUR[LPDWHO\ RQH PRQWK ODWHU WKH LQYHVWLJDWRU FRQWDFWHG DOO SDUWLFLSDQWV DW KRPH E\ SKRQH WR UHDVVHVV FRQGLWLRQ DQG UHJLPHQ VSHFLILFYDULDEOHV PDWHUQDO SHUFHSWLRQV RI WKH WUHDWPHQW UHJLPHQ DQG PDWHUQDO HPRWLRQDO GLVWUHVV 7KH RQH PRQWK GHOD\ LQ WKH DVVHVVPHQW ZDV GHVLJQHG WR PLQLPL]H VDPSOLQJ RI WHPSRUDU\ EHKDYLRU FKDQJH WKDW PD\ KDYH RFFXUUHG LPPHGLDWHO\ DIWHU D FOLQLF YLVLW DQG WR UHDVVHVV PDWHUQDO UDWLQJV RI HPRWLRQDO GLVWUHVV FRQGLWLRQ DQG UHJLPHQUHODWHG PHDVXUHV DQG PDWHUQDO SHUFHSWLRQV UHJDUGLQJ WKH WUHDWPHQW UHJLPHQ DQG WKHLU UHODWLRQVKLS WR PDWHUQDO DGKHUHQFH WR WKH FKLOGnV PHGLFDWLRQ UHJLPHQ DQG FKLOG KHDOWK RXWFRPHV 3DUWLFLSDQWV 3DUWLFLSDQWV ZLWK D GLDJQRVHG VHL]XUH GLVRUGHU ZHUH UHFUXLWHG IURP WKH 3HGLDWULF 1HXURORJ\ FOLQLF DW 6KDQGV +RVSLWDO DW WKH 8QLYHUVLW\ RI )ORULGDnV +HDOWK 6FLHQFH &HQWHU %DVHG RQ GDWD REWDLQHG IURP WKH SHGLDWULF QHXURORJLVW WKHUH DUH RYHU FKLOGUHQ ZLWK D GLDJQRVHG VHL]XUH GLVRUGHU EHWZHHQ WKH DJHV RI DQG ZKR ZHUH HOLJLEOH WR SDUWLFLSDWH LQ WKH VWXG\
PAGE 35

\HDUV RI DJH f (QJOLVKVSHDNLQJ f $ELOLW\ WR UHDG DQG ZULWH f 7HOHSKRQH DYDLODEOH LQ WKH KRPH ([FOXVLRQDU\ FULWHULD LQFOXGHG f \HDUV RI DJH RU \HDUV RI DJH f 1RQ(QJOLVK VSHDNLQJ f ,QDELOLW\ WR UHDG DQG ZULWH f 1R WHOHSKRQH LQ WKH KRPH $ WRWDO RI SDUWLFLSDQWV FRPSOHWHG DOO VWXG\ PHDVXUHV 'DWD LQGLFDWHG WKH VDPSOH FRQVLVWHG RI PDOH DQG IHPDOH SDWLHQWV DJHG \HDUV 0 6' f 7KH PHDQ HGXFDWLRQ RI PRWKHUV LQ WKH VDPSOH ZDV \HDUV 6' f 0RWKHUV ZHUH PDUULHG bf DQG ZHUH KRPHPDNHUV bf RU ZRUNLQJ IXOOWLPH bf $YHUDJH IDPLO\ LQFRPH ZDV 6' f 7KH HWKQLF PDNHXS RI WKH PDWHUQDO VDPSOH ZDV b :KLWH b $IULFDQ $PHULFDQ b +LVSDQLF $PHULFDQ b 1DWLYH $PHULFDQ DQG b 2WKHU 7KH HWKQLF PDNHXS RI WKH FKLOG VDPSOH ZDV b :KLWH b $IULFDQ $PHULFDQ b +LVSDQLF $PHULFDQ b 1DWLYH $PHULFDQ DQG b %LUDFLDO2WKHU $SSUR[LPDWHO\ b RI FKLOGUHQ LQ WKH VDPSOH ZHUH WDNLQJ PHGLFDWLRQ LQ SLOO IRUP ZLWK VHYHQ SHUFHQW WDNLQJ PHGLFDWLRQ LQ OLTXLG IRUP $SSUR[LPDWHO\ b RI FKLOGUHQ LQ WKH VWXG\ ZHUH GLDJQRVHG ZLWK SDUWLDO VHL]XUHV b ZLWK JHQHUDOL]HG VHL]XUHV DQG b ZLWK D GLDJQRVLV RI RWKHU W\SH 6WXG\ VDPSOH GHPRJUDSKLF FKDUDFWHULVWLFV DUH SUHVHQWHG LQ 7DEOH 7DEOH 'HPRJUDSKLF DQG 0HGLFDO &KDUDFWHULVWLFV 0HDQ DJH RI FKLOGUHQ LQ \HDUV 6'f f 6H[ RI WKH FKLOG %R\V bf *LUOV bf 0HDQ HGXFDWLRQ RI PRWKHUV LQ \HDUV 6'f f

PAGE 36

7DEOH &RQWLQXHG 0RWKHU 0DULWDO 6WDWXV 6LQJOH 1HYHU 0DUULHG 0DUULHG 6HSDUDWHG'LYRUFHG bf bf bf 0RWKHU (PSOR\PHQW 6WDWXV 8QHPSOR\HG 9ROXQWHHU 3DUWWLPH )XOOWLPH +RPHPDNHU 6WXGHQW 2WKHU bf bf bf bf bf bf bf 0HDQ DQQXDO IDPLO\ LQFRPH LQ GROODUV 6'f f &KLOG UDFH :KLWH $IULFDQ $PHULFDQ +LVSDQLF $PHULFDQ 1DWLYH $PHULFDQ %LUDFLDO2WKHU bf bf bf bf bf 0RP UDFH :KLWH $IULFDQ $PHULFDQ +LVSDQLF $PHULFDQ 1DWLYH $PHULFDQ %LUDFLDO2WKHU bf bf bf bf bf 5HJLPHQ 7\SH 3LOOV /LTXLG bf bf 6HL]XUH 7\SH 3DUWLDO *HQHUDOL]HG 2WKHU bf bf bf 1RWH Q f 3DUWLFLSDQWV ZHUH UHFUXLWHG IURP WKH 8QLYHUVLW\ RI )ORULGDfV 3HGLDWULF 1HXURORJ\ 2XWSDWLHQW &OLQLF DW 6KDQGV +RVSLWDO GXULQJ WKH SDWLHQWfV XVXDO FOLQLF YLVLW 7KLV FOLQLF

PAGE 37

WUHDWV RYHU FKLOGUHQ D \HDU GLDJQRVHG ZLWK HSLOHSV\ :H RIIHUHG WKH VWXG\ WR DOO SDWLHQWV ZKR PHW WKH VWXG\fV HOLJLELOLW\ FULWHULD 7KH UHVHDUFKHU DQGRU UHVHDUFK DVVLVWDQWV DSSURDFKHG SDUWLFLSDQWV LQGLYLGXDOO\ DQG SUHVHQWHG WKH UDWLRQDOH IRU WKH VWXG\ 3DUHQWDO LQIRUPHG FRQVHQW DQG FKLOG DVVHQW ZDV REWDLQHG IRU FKLOGUHQ LI SRVVLEOHf 4XHVWLRQV FRQFHUQLQJ VXEMHFW REOLJDWLRQ DQG FRQILGHQWLDOLW\ ZHUH DGGUHVVHG DW WKLV WLPH $IWHU REWDLQLQJ LQIRUPHG FRQVHQWDVVHQW SDUWLFLSDQWV FRPSOHWHG D EULHI EDFNJURXQG TXHVWLRQQDLUH LQFOXGLQJ TXHVWLRQV UHJDUGLQJ PDWHUQDO SHUFHSWLRQV RI WKH WUHDWPHQW UHJLPHQ WKH &(6' DQG WKH 36,6) 7KH HQWLUH SURFHVV WRRN DSSUR[LPDWHO\ PLQXWHV 7KH LQYHVWLJDWRU UHFRUGHG WKH IROORZLQJ LQIRUPDWLRQ IURP WKH SDUWLFLSDQWf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nV OHYHO RI HGXFDWLRQ IDPLO\ LQFRPH DQG PDULWDO VWDWXV 7LPH VLQFH LQLWLDO GLDJQRVLV ZDV GLYLGHG LQWR WZR FDWHJRULHV f WKRVH ZKR ZHUH QHZO\ GLDJQRVHG DQG f WKRVH ZLWK D GLDJQRVLV RI HSLOHSV\ JUHDWHU RU HTXDO WR RQH \HDU SULRU WR WKH VWXG\ 0RWKHUV ZHUH DVNHG WR GHVFULEH WKHLU FKLOGnV VHL]XUH

PAGE 38

GLDJQRVLV VHL]XUH IUHTXHQF\ DQG PHGLFDWLRQ UHJLPHQ 7KH W\SHVf RI VHL]XUH GLVRUGHUVf WKH PRWKHUV UHSRUWHG ZHUH QRWHG 0DWHUQDO UHSRUW RI WKH FKLOGfV PRQWKO\ VHL]XUH IUHTXHQF\ IRU WKH SDVW PRQWK ZDV DOVR DVVHVVHG )XUWKHUPRUH FRPSOH[LW\ RI WKH PHGLFDWLRQ UHJLPHQ UHSRUWHG E\ PRWKHUV ZDV HYDOXDWHG EDVHG RQ LQGLFDWRUV LH W\SH RI PHGLFDWLRQ QXPEHU RI PHGLFDWLRQV SHU GD\ GRVLQJ DQG WLPLQJf $ FRPSOH[LW\ UDWLQJ IRU WKH PHGLFDWLRQ UHJLPHQ ZDV FDOFXODWHG E\ PXOWLSO\LQJ WKH QXPEHU RI DQWLHSLOHSWLF PHGLFDWLRQV E\ WKH WRWDO QXPEHU RI WLPHV SHU GD\ HDFK PHGLFDWLRQ LV SUHVFULEHG 0DWHUQDO UDWLQJV RI SHUFHLYHG GLIILFXOW\ QRW GLIILFXOW WR H[WUHPHO\ GLIILFXOWf DQG SHUFHLYHG HIILFDF\ QRW JRRG WR H[WUHPHO\ JRRGf LQ FDUU\LQJ RXW WKH FKLOGnV PHGLFDWLRQ UHJLPHQ ZHUH DVVHVVHG 3HUFHLYHG GLIILFXOW\ UDWLQJV LQFOXGHG LWHPV DERXW PDWHUQDO SHUFHSWLRQV RI GLIILFXOW\ RI PHGLFDWLRQ W\SH PHGLFDWLRQ GRVDJH DQG PHGLFDWLRQ WLPLQJV 3HUFHLYHG HIILFDF\ UDWLQJV LQFOXGHG LWHPV DERXW PDWHUQDO SHUFHSWLRQV RI HIILFDF\ RI WKH FKLOGfV PHGLFDWLRQV DQG HIILFDF\ RI PDWHUQDO PDQDJHPHQW RI WKH FKLOGfV PHGLFDWLRQ UHJLPHQ 0RWKHUV ZHUH DVNHG WR UDWH KRZ WKH\ SHUFHLYHG WKHLU OHYHO RI NQRZOHGJH DERXW WKHLU FKLOGnV PHGLFDO FRQGLWLRQ 7KH NQRZOHGJH UDWLQJV DOVR UDQJHG IURP WR QRW NQRZOHGJHDEOH WR H[WUHPHO\ NQRZOHGJHDEOHf .QRZOHGJH UDWLQJV LQFOXGHG LWHPV DERXW PDWHUQDO SHUFHSWLRQV RI NQRZOHGJH RI WKH FKLOGfV VHL]XUH GLVRUGHU DQG PHGLFDWLRQV W\SH WLPLQJ QXPEHU DQG GRVDJHf VHH $SSHQGL[ $ t $SSHQGL[ %f 0HGLFDO &KDUW ,QIRUPDWLRQ &RQGLWLRQVSHFLILF LQIRUPDWLRQ LQFOXGLQJ GDWH RI GLDJQRVLV WLPH VLQFH GLDJQRVLV W\SH RI GLDJQRVLV W\SH RI PHGLFDWLRQ GRVH RI PHGLFDWLRQ VHL]XUH IUHTXHQF\ EORRG OHYHOV SLOO FRXQWV DQG HPHUJHQF\ URRP (5f YLVLWV ZHUH FROOHFWHG E\ D UHYLHZ RI WKH PHGLFDO FKDUW 7KH FKLOGnV PHGLFDWLRQ UHJLPHQ ZDV QRWHG IURP WKH FKDUW W\SH RI PHGLFDWLRQ QXPEHU RI PHGLFDWLRQV SHU GD\ GRVDJH DQG WLPLQJf DQG FRPSDUHG WR WKH

PAGE 39

PRWKHUVn UHSRUW RI WKH FKLOGnV PHGLFDWLRQ UHJLPHQ &RQFRUGDQFH UDWLQJV ZHUH FDOFXODWHG EHWZHHQ SK\VLFLDQ SUHVFULEHG DQWLHSLOHSWLF PHGLFDWLRQVf DQG PDWHUQDO UHSRUWV RI SUHVFULEHG PHGLFDWLRQVf E\ GHWHUPLQLQJ WKH SHUFHQWDJH FRUUHFW RI PDWHUQDO UHSRUWV FRPSDUHG WR SK\VLFLDQ UHSRUWV VHH $SSHQGL[ & t $SSHQGL[ 'f 6SHFLILFDOO\ SRLQWV IRU FRQFRUGDQFH ZHUH DVVLJQHG IRU D FRUUHFW PDWFK EHWZHHQ WKH SUHVFULEHG DQG UHSRUWHG UHJLPHQ DV IROORZV QDPHf GRVHf DQG WLPLQJf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fV PHGLFDO FKDUW IRU WKH PRVW UHFHQW GDWD UHJDUGLQJ FKLOG KHDOWK VWDWXV :ULWWHQ SUHVFULSWLRQV DQG PHGLFDWLRQ UHILOO LQIRUPDWLRQ ZHUH QRW QRWHG V\VWHPDWLFDOO\ LQ HDFK PHGLFDO FKDUW DQG WKXV WKLV LQIRUPDWLRQ ZDV QRW DYDLODEOH WR EH FROOHFWHG 3LOO FRXQWV ZHUH DVVHVVHG GXULQJ WKH FOLQLF YLVLW LI PHGLFDWLRQ ERWWOHV ZHUH EURXJKW WR WKH FOLQLF DSSRLQWPHQW ,I PRWKHUV GLG QRW KDYH WKH FKLOGfV PHGLFDWLRQ ERWWOH DW WKH FOLQLF YLVLW WKH\ ZHUH FRQWDFWHG DW KRPH RQ WKH VDPH GD\ RI WKH DSSRLQWPHQW LQ RUGHU WR UHSRUW D FXUUHQW SLOO FRXQW 0RWKHUV ZHUH WHOHSKRQHG DW RQHPRQWK IROORZXS DQG DVNHG WR UHSRUW D SLOO FRXQW 0DWHUQDO DGKHUHQFH WR WKH FKLOGfV PHGLFDO UHJLPHQ ZDV DVVHVVHG E\ SHUFHQWDJH RI SLOOV XVHG DIWHU RQH PRQWK >7LPH 7LPH f 7LPH f@ [ b

PAGE 40

$GKHUHQFH ZDV GHILQHG DV DQ b SLOO FRXQW SHUFHQWDJH %ORRG VHUXP OHYHOV ZHUH QRW LQFOXGHG DV D PHDVXUH RI DGKHUHQFH EHFDXVH WKHVH YDOXHV ZHUH QRW FRPSDUDEOH DFURVV GLIIHUHQW PHGLFDWLRQ W\SHV DGGLWLRQDOO\ PDQ\ SDUWLFLSDQWV KDG QR EORRG VHUXP OHYHOV GUDZQ 0DWHUQDO (PRWLRQDO 'LVWUHVV 7KH &HQWHU IRU (SLGHPLRORJLF 6WXGLHV 'HSUHVVLRQ &(6' 5DGORII f LV D VHOIUHSRUW PHDVXUH GHYHORSHG WR EH XVHG LQ HSLGHPLRORJLFDO VWXGLHV RI GHSUHVVLRQ DQG ZDV GHVLJQHG WR PHDVXUH FXUUHQW OHYHOV RI V\PSWRPV ZKLFK DFFRPSDQ\ GHSUHVVLRQ LQ D JHQHUDO DGXOW SRSXODWLRQ 7KH &(6' LV D LWHP VHOIUHSRUW PHDVXUH RI GHSUHVVLYH V\PSWRPV (DFK LWHP SURYLGHV D VWDWHPHQW UHSUHVHQWLQJ D V\PSWRP FKDUDFWHULVWLF RI GHSUHVVLRQ HJ KDG FU\LQJ VSHOOVf IROORZHG E\ D SRLQW /LNHUWW\SH UHVSRQVH VFDOH UDQJLQJ IURP UDUHO\ RU QRQH RI WKH WLPH OHVV WKDQ GD\f WR PRVW DOO RI WKH WLPH GD\Vf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f +LJKHU VFRUHV LQGLFDWH KLJKHU OHYHOV RI GHSUHVVLYH V\PSWRPV H[SHULHQFHG GXULQJ WKH SDVW ZHHN ,QYHVWLJDWRUV KDYH GHVLJQDWHG D FXWRII VFRUH RI DV D VXLWDEOH LQGLFDWRU WR GHWHUPLQH GHSUHVVHG IURP QRQGHSUHVVHG SDWLHQWV 1H]X f

PAGE 41

5HJDUGLQJ LQWHUQDO FRQVLVWHQF\ &URQEDFKnV DOSKD LV KLJK DFURVV D YDULHW\ RI SRSXODWLRQV DQG UDQJHV IURP IRU WKH JHQHUDO SRSXODWLRQ DQG IRU WKH FOLQLFDO SRSXODWLRQ 6SOLWKDOI UHOLDELOLW\ LV DOVR KLJK UDQJLQJ IURP WR 7HVWUHWHVW UHOLDELOLW\ VWXGLHV UDQJLQJ RYHU ZHHNV VKRZ PRGHUDWH FRUUHODWLRQV U f 6WXGLHV H[DPLQLQJ HIIHFWV EHWZHHQ $IULFDQ $PHULFDQV &DXFDVLDQV DQG 0H[LFDQ $PHULFDQV IRXQG QR GLIIHUHQFHV LQ PHDVXUHV RI LQWHUQDO FRQVLVWHQF\ UHOLDELOLW\ ,Q VDPSOHV RI RXWSDWLHQWV ZLWK GHSUHVVLRQ DOFRKROLVP GUXJ DGGLFWLRQ RU VFKL]RSKUHQLD FRUUHODWLRQ FRHIILFLHQWV EHWZHHQ &(6' VFRUHV DQG 6\PSWRP &KHFNOLVW 'HSUHVVLRQ 6XEVFDOH 6FRUHV ZHUH KLJK UDQJLQJ IURP &RUUHODWLRQV ZLWK WKH +DPLOWRQ 5DWLQJ 6FDOH IRU 'HSUHVVLRQ DQG 5DVNLQ 6FDOH UDQJHG IURP DQG IRU SDWLHQWV ZLWK DFXWH GHSUHVVLRQ WR DQG IRU SDWLHQWV ZLWK VFKL]RSKUHQLD UHVSHFWLYHO\ 6FRUHV RQ WKH &(6' KDYH EHHQ VKRZQ WR GHFUHDVH VLJQLILFDQWO\ IROORZLQJ WUHDWPHQW IRU GHSUHVVLRQ )RU PDMRU GHSUHVVLRQ VHQVLWLYLW\ ZDV b DQG VSHFLILFLW\ ZDV b 7KH SRVLWLYH SUHGLFWLYH YDOXH ZDV b $PHULFDQ 3V\FKLDWULF $VVRFLDWLRQ f 7KH 3DUHQWLQJ 6WUHVV ,QGH[ 6KRUW )RUP 36,6) $ELGLQ f LV D LWHP VKRUW IRUP GLUHFWO\ GHULYHG IURP WKH ORQJ IRUP RI WKH 36, 7KH 36,6) ZDV GHYHORSHG IRU XVH LQ D YDULHW\ RI SULPDU\ KHDOWK FDUH VHWWLQJV ZKHUH LW FRXOG EH DGPLQLVWHUHG LQ OHVV WKDQ PLQXWHV 3DUWLFLSDQWV ZHUH DVNHG WR UHVSRQG WR TXHVWLRQV EDVHG RQ WKH ILYHSRLQWVFDOH 7KH 36,6) SDUHQW VHOIUHSRUW VFDOH FRQWDLQV WKUHH IDFWRUDQDO\WLFDOO\GHULYHG VXEVFDOHV 3DUHQWDO 'LVWUHVV 3DUHQW&KLOG '\VIXQFWLRQDO ,QWHUDFWLRQ DQG 'LIILFXOW &KLOGf DQG D 7RWDO 6WUHVV 6FRUH 7KH 3DUHQWDO 'LVWUHVV 3'f VFDOH GHWHUPLQHV WKH GLVWUHVV D SDUHQW LV H[SHULHQFLQJ LQ WKHLU UROH DV D SDUHQW DV D IXQFWLRQ RI SHUVRQDO IDFWRUV WKDW DUH

PAGE 42

GLUHFWO\ UHODWHG WR SDUHQWLQJ 7KH 3DUHQW&KLOG '\VIXQFWLRQDO ,QWHUDFWLRQ 3&',f VXEVFDOH IRFXVHV RQ WKH SDUHQWnV SHUFHSWLRQ WKDW WKHLU FKLOG GRHV QRW PHHW WKHLU H[SHFWDWLRQV DQG WKH LQWHUDFWLRQV ZLWK WKHLU FKLOG DUH QRW UHLQIRUFLQJ WR WKHP DV D SDUHQW 7KH 'LIILFXOW &KLOG '&f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f 7KH VKRUW IRUP VXEVFDOHV KDYH VKRZQ &URQEDFKnV DOSKDV RI WR DQG PRQWK WHVWUHWHVW UHOLDELOLWLHV RI WR 3RVLWLYH SUHGLFWLYH YDOXHV RI WKH 36,6) UDQJH IURP b WR b

PAGE 43

&+$37(5 5(68/76 7KH RYHUDOO ILQGLQJV RI WKH GDWD DQDO\VHV DUH SUHVHQWHG DV IROORZV f GHVFULSWLYH VWDWLVWLFV RI WKH SUHGLFWRU DQG FULWHULRQ YDULDEOHV LQFOXGHG LQ WKH FRQFHSWXDO PRGHO f LQWHUQDO FRQVLVWHQF\ UHOLDELOLWLHV RI WKH TXHVWLRQQDLUH PHDVXUHV DQG WKH GLVWULEXWLRQDO SURSHUWLHV RI NH\ SUHGLFWRU DQG FULWHULRQ YDULDEOHV f FRUUHODWLRQV EHWZHHQ WKH EDFNJURXQG YDULDEOHV QRW LQFOXGHG LQ WKH FRQFHSWXDO PRGHO DQG WKRVH EHWZHHQ NH\ SUHGLFWRU YDULDEOHV DQG FKDQJH VFRUHV RQ WKH FULWHULRQ YDULDEOHV f UHJUHVVLRQ DQDO\VHV WKDW FRQWURO IRU WKH HIIHFWV RI 7LPH SUHGLFWRUV RQ 7LPH FULWHULRQ YDULDEOHV DQG f H[SORUDWRU\ PRGHO WHVWLQJ WR LGHQWLI\ JRRGQHVV RI ILW RI WKH VSHFLILF UHJUHVVLRQ PRGHOV 'XH WR VPDOO VDPSOH VL]H VWUXFWXUDO HTXDWLRQ PRGHOLQJ 6(0f ZDV QRW FRQGXFWHG DV DQWLFLSDWHG 7KH VWDWLVWLFDO DSSURDFK VKLIWHG IURP 6(0 WR YDULDEOHOHYHO PRGHO WHVWLQJ LQ ZKLFK WKH JRRGQHVV RI ILW RI YDULDEOHV ORFDWHG XQGHU WKH IRXU SULPDU\ IDFWRUV RI WKH FRQFHSWXDO IUDPHZRUN ZDV DVVHVVHG VHH )LJXUH f 3UHOLPLQDU\ $QDO\VHV 0HGLFDO FKDUDFWHULVWLFV RI WKH FKLOG ZLWK HSLOHSV\ ZHUH YDULDEOHV FROOHFWHG DW 7LPH H[FHSW IRU VHL]XUH IUHTXHQF\ 7KH ODWWHU ZDV DOVR FROOHFWHG DW 7LPH VLQFH WKLV WLPH SRLQW ZDV RI PDLQ LQWHUHVW LQ RXU SULPDU\ DQDO\VHV 6WXG\ VDPSOH PHGLFDO FKDUDFWHULVWLFV DUH GLVSOD\HG LQ 7DEOH

PAGE 44

7DEOH 0HDQ 6FRUHV DQG 6WDQGDUG 'HYLDWLRQV IRU &KLOG 0HGLFDO &KDUDFWHULVWLFV 0HDVXUHV 06'f &RPSOH[LW\ RI WKH UHJLPHQ XQFRUUHFWHGf f &RPSOH[LW\ RI WKH UHJLPHQ FRUUHFWHGf f 7LPH 6LQFH 'LDJQRVLV LQ \HDUV f &RQFRUGDQFH EHWZHHQ UHSRUWHG DQG SUHVFULEHG UHJLPHQ7LPH f 6HL]XUH )UHTXHQF\7LPH XQFRUUHFWHGf f 6HL]XUH )UHTXHQF\7LPH FRUUHFWHGf f 1RWH Q f &RUUHFWHG VFRUHV DUH FRUUHFWHG IRU VNHZQHVV WR HQVXUH DQDO\VHV LQWHJULW\ 0HDQV DQG VWDQGDUG GHYLDWLRQV IRU PDWHUQDO VHOIUHSRUWV RI PDWHUQDO VWUHVV PHDVXUHG E\ WKH 36,6)f GHSUHVVLRQ PHDVXUHG E\ WKH &(6'f DQG PDWHUQDO SHUFHSWLRQV RI WKH WUHDWPHQW UHJLPHQ PHDVXUHG E\ PDWHUQDO UHSRUWV RI GLIILFXOW\ HIILFDF\ DQG NQRZOHGJHf IRU 7LPH DQG 7LPH DUH SUHVHQWHG LQ 7DEOH EHORZ 7DEOH 0HDQ 6FRUHV DQG 6WDQGDUG 'HYLDWLRQV IRU 6WXG\ 0HDVXUHV DW 7LPH DQG 7LPH 0HDVXUHV 06'f 7LPH 0 6'f 7LPH 36,6) 7RWDO VFRUH f f 36,6) 3DUHQWDO GLVWUHVV f f 36,6) 3DUHQWFKLOG G\VIXQFWLRQ f f 36,6) 'LIILFXOW FKLOG f f &(6' 6HYHULW\ VFRUH f f 'LIILFXOW\ 7RWDO VFRUH XQFRUUHFWHG f f 'LIILFXOW\ FRUUHFWHG f f (IILFDF\ 7RWDO VFRUH f f .QRZOHGJH 7RWDO VFRUH f f 1RWH 36,6) 3DUHQWLQJ 6WUHVV ,QGH[6KRUW )RUP &(6' &HQWHU IRU (SLGHPLRORJLF 6WXGLHV'HSUHVVLRQ &RUUHFWHG VFRUHV DUH FRUUHFWHG IRU VNHZQHVV WR HQVXUH DQDO\VHV LQWHJULW\ &URQEDFKfV DOSKDV ZHUH FDOFXODWHG WR DVVHVV WKH UHOLDELOLW\ LH LQWHUQDO FRQVLVWHQF\f RI PRWKHUFRPSOHWHG VHOIUHSRUW PHDVXUHV &RHIILFLHQWV UDQJHG IURP WR IRU WKH WRWDO VDPSOH DW 7LPH VXJJHVWLQJ DGHTXDWH LQWHUQDO FRQVLVWHQF\ UHOLDELOLW\ IRU DOO

PAGE 45

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f 7KH PHDQV DQG VWDQGDUG GHYLDWLRQV IRU XQFRUUHFWHG DQG FRUUHFWHG DQQXDO LQFRPH ZHUH f DQG f UHVSHFWLYHO\ 7KH RULJLQDO VFRUHV IRU FRPSOH[LW\ RI WKH UHJLPHQ VHL]XUH IUHTXHQF\ DQG PDWHUQDO SHUFHSWLRQV RI UHJLPHQ GLIILFXOW\ ZHUH DOVR KLJKO\ SRVLWLYHO\ VNHZHG ,Q RUGHU WR FRUUHFW IRU GDWD VNHZQHVV WKH GDWD ZDV FROODSVHG LQWR JURXSV WKDW ZHUH HTXDOO\ GLVWULEXWHG FUHDWLQJ QHZ FRPSOH[LW\ VHL]XUH IUHTXHQF\ DQG GLIILFXOW\

PAGE 46

YDULDEOHV 2ULJLQDO FRPSOH[LW\ VFRUHV UDQJHG IURP WR DQG ZHUH UHFRGHG LQWR WKUHH FDWHJRULHV ZLWK RQH UHSUHVHQWLQJ D FRPSOH[LW\ VFRUH RI WR WZR UHSUHVHQWLQJ D FRPSOH[LW\ VFRUH RI WR DQG WKUHH UHSUHVHQWLQJ D FRPSOH[LW\ VFRUH RI WR 0HDQV DQG VWDQGDUG GHYLDWLRQV IRU XQFRUUHFWHG DQG FRUUHFWHG FRPSOH[LW\ VFRUHV ZHUH f DQG f UHVSHFWLYHO\ 2ULJLQDO VHL]XUH IUHTXHQF\ VFRUHV UDQJHG IURP WR DQG ZHUH UHFRGHG LQWR IRXU FDWHJRULHV ZLWK HTXDO GLVWULEXWLRQV RI WKH VDPSOH LQ HDFK JURXS ZLWK RQH UHSUHVHQWLQJ VHL]XUHV WZR UHSUHVHQWLQJ WR VHL]XUHV WKUHH UHSUHVHQWLQJ WR VHL]XUHV DQG IRXU UHSUHVHQWLQJ WR VHL]XUHV 7KH PHDQV DQG VWDQGDUG GHYLDWLRQV IRU XQFRUUHFWHG DQG FRUUHFWHG VHL]XUH IUHTXHQF\ VFRUHV ZHUH f DQG f UHVSHFWLYHO\ 'LIILFXOW\ VFRUHV ZHUH SRVLWLYHO\ VNHZHG UDQJHG IURP WR DQG ZHUH UHFRGHG LQWR IRXU FDWHJRULHV ZLWK RQH UHSUHVHQWLQJ D VFRUH RI WR WZR UHSUHVHQWLQJ D VFRUH RI WR WKUHH UHSUHVHQWLQJ D VFRUH RI WR DQG IRXU UHSUHVHQWLQJ D VFRUH RI WR 7KH PHDQV DQG VWDQGDUG GHYLDWLRQV IRU XQFRUUHFWHG DQG FRUUHFWHG GLIILFXOW\ VFRUHV ZHUH f DQG f IRU 7LPH DQG f DQG f IRU 7LPH UHVSHFWLYHO\ 3ULPDU\ $QDO\VHV &RUUHODWLRQV DPRQJ %DFNJURXQG 9DULDEOHV 3UHGLFWRUV DQG &ULWHULRQ 9DULDEOHV &RUUHODWLRQDO DQDO\VHV ZHUH FRQGXFWHG EHWZHHQ EDFNJURXQG YDULDEOHV QRW LQFOXGHG LQ WKH PRGHO FKLOG VH[ PDULWDO VWDWXV QXPEHU RI \HDUV PDUULHG HPSOR\PHQW VWDWXV QXPEHU RI FKLOGUHQ LQ WKH KRPHf DQG 7LPH NH\ SUHGLFWRU YDULDEOHV RI WKH IROORZLQJ IDFWRUV PDWHUQDO HPRWLRQDO GLVWUHVV PDWHUQDO SHUFHSWLRQV RI WKH UHJLPHQ FRQGLWLRQ DQG UHJLPHQUHODWHG YDULDEOHV DQG GHPRJUDSKLF YDULDEOHV 1R VLJQLILFDQW DVVRFLDWLRQV ZHUH IRXQG EHWZHHQ WKH EDFNJURXQG YDULDEOHV QRW LQFOXGHG LQ WKH PRGHO DQG 7LPH SUHGLFWRUV

PAGE 47

S f 1R VLJQLILFDQW DVVRFLDWLRQV ZHUH IRXQG EHWZHHQ WKH EDFNJURXQG YDULDEOHV QRW LQFOXGHG LQ WKH PRGHO DQG FULWHULRQ YDULDEOH FKDQJH VFRUHV VHH 7DEOH EHORZf 7DEOH &RUUHODWLRQV EHWZHHQ %DFNJURXQG 9DULDEOHV QRW LQ WKH 0RGHO DQG 6SHFLILF 7LPH 2XWFRPHV &KDQJH 6FRUHV %DFNJURXQG )DFWRUV 6HL]XUH )UHTXHQF\ 6HL]XUHUHODWHG (5 9LVLWV &KLOG VH[ L R R D? 0DULWDO VWDWXV
PAGE 48

+\SRWKHVLV 0DWHUQDO (PRWLRQDO 'LVWUHVV $GKHUHQFH DQG &KLOG +HDOWK 6WDWXV :H K\SRWKHVL]HG WKDW 7LPH PDWHUQDO HPRWLRQDO GLVWUHVV PHDVXUHG E\ WKH 36,6) DQG &(6'f ZRXOG EH QHJDWLYHO\ UHODWHG WR 7LPH DGKHUHQFH WR WKH FKLOGnV PHGLFDWLRQ UHJLPHQ SLOO FRXQW SHUFHQWDJHf DQG SRVLWLYHO\ DVVRFLDWHG ZLWK FKLOG KHDOWK VWDWXV VHL]XUH IUHTXHQF\ DQG VHL]XUHUHODWHG HPHUJHQF\ URRP YLVLWVf 6SHFLILFDOO\ LW ZDV SUHGLFWHG WKDW 7LPH PDWHUQDO VWUHVV DQG GHSUHVVLRQ VFRUHV ZRXOG EH QHJDWLYHO\ DVVRFLDWHG ZLWK 7LPH DGKHUHQFH WR WKH FKLOGnV PHGLFDWLRQ UHJLPHQ DQG WKDW 7LPH UDWLQJV RI PDWHUQDO VWUHVV DQG GHSUHVVLYH DIIHFW ZRXOG EH SRVLWLYHO\ DVVRFLDWHG ZLWK WKH FKLOGf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

PAGE 49

7DEOH 6XPPDU\ RI +LHUDUFKLFDO 5HJUHVVLRQ $QDO\VHV IRU 7LPH 36,6) 2YHUDOO 6FRUH 3UHGLFWLQJ &KLOG 6HL]XUH )UHTXHQF\ 1 f 9DULDEOHV ,QFOXGHG 3 W 5 5 $GM 5 $5 ) $) &RQWUROOLQJ IRU 7 &(6' PHDQ 7 36,6) PHDQ 7 0RGHO &(6' PHDQ 7 36,6) PHDQ 7 36,6) PHDQ 7 r r 1RWH &(6' &HQWHU IRU (SLGHPLRORJLF 6WXGLHV'HSUHVVLRQ 36,6) 6WUHVV ,QGH[6KRUW )RUP 7 7LPH 7 7LPH r3 3DUHQWLQJ 7DEOH 6XPPDU\ RI +LHUDUFKLFDO 5HJUHVVLRQ $QDO\VHV IRU 7LPH 36,6) 6XEVFDOH 9DULDEOHV 3UHGLFWLQJ &KLOG 6HL]XUH )UHTXHQF\ 1 f 9DULDEOHV ,QFOXGHG 3 W 5 5 $GM 5 $5 ) $) &RQWUROOLQJ IRU 7 3' PHDQ 7 3& G\V 7 '& 7 0RGHO 3' PHDQ 7 3& G\V 7 '& 7 3& G\V 7 r r 1RWH 36,6) 3DUHQWLQJ 6WUHVV ,QGH[6KRUW )RUP 3' 3DUHQWDO 'LVWUHVV '& 'LIILFXOW &KLOG 3& G\V 3DUHQW&KLOG '\VIXQFWLRQ 7 7LPH 7 7LPH r 7KH GDWD UHYHDOHG VLJQLILFDQW UHODWLRQVKLSV SDUWLDO FRUUHODWLRQVf EHWZHHQ 7LPH 36,6) RYHUDOO PHDQ VFRUHV DQG 7LPH VHL]XUH IUHTXHQF\ U S f DV ZHOO DV EHWZHHQ 7LPH 36,6) SDUHQWFKLOG G\VIXQFWLRQ VXEVFDOH VFRUHV DQG 7LPH VHL]XUH

PAGE 50

IUHTXHQF\ U e f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r 36,6) 3DUHQWDO 'LVWUHVV 36,6) 3DUHQWFKLOG G\VIXQFWLRQ r 36,6) 'LIILFXOW FKLOG &(6' 7RWDO 1RWH 36,6) 3DUHQWLQJ 6WUHVV ,QGH[ &(6' &HQWHU IRU (SLGHPLRORJLF 6WXGLHV 'HSUHVVLRQ r3 +\SRWKHVLV 0DWHUQDO 3HUFHSWLRQV RI WKH 7UHDWPHQW 5HJLPHQ $GKHUHQFH DQG &KLOG +HDOWK 6WDWXV :H SUHGLFWHG WKDW 7LPH UDWLQJV RI PDWHUQDO SHUFHSWLRQV RI WKH WUHDWPHQW UHJLPHQ PHDVXUHG E\ SHUFHLYHG NQRZOHGJH SHUFHLYHG GLIILFXOW\ DQG SHUFHLYHG HIILFDF\f DQG 7LPH PDWHUQDO GLVWUHVV PHDVXUHG E\ WKH 36,6)f DQG 7LPH PDWHUQDO SHUFHSWLRQV RI WKH WUHDWPHQW UHJLPHQ ZRXOG EH DVVRFLDWHG ZLWK 7LPH DGKHUHQFH WR WKH FKLOGfV

PAGE 51

PHGLFDWLRQ UHJLPHQ SLOO FRXQW SHUFHQWDJHf DQG FKLOG KHDOWK VWDWXV VHL]XUH IUHTXHQF\ DQG VHL]XUHUHODWHG HPHUJHQF\ URRP YLVLWVf 6SHFLILFDOO\ PDWHUQDO UDWLQJV RI SHUFHLYHG NQRZOHGJH DQG HIILFDF\ UHJDUGLQJ WKHLU FKLOGfV PHGLFDO FRQGLWLRQ DQG PHGLFDWLRQ UHJLPHQ DW 7LPH ZRXOG EH SRVLWLYHO\ DVVRFLDWHG ZLWK 7LPH DGKHUHQFH WR WKH FKLOGnV PHGLFDWLRQ UHJLPHQ DQG QHJDWLYHO\ ZLWK KHDOWK VWDWXV ,Q DGGLWLRQ PDWHUQDO UDWLQJV RI SHUFHLYHG GLIILFXOW\ LQ JLYLQJ WKH FKLOG WKH SUHVFULEHG PHGLFDWLRQ DW 7LPH ZRXOG EH QHJDWLYHO\ DVVRFLDWHG ZLWK 7LPH DGKHUHQFH WR WKH FKLOGnV PHGLFDWLRQ UHJLPHQ DQG SRVLWLYHO\ DVVRFLDWHG ZLWK FKLOG KHDOWK VWDWXV DQG ODVW PDWHUQDO UDWLQJV RI PDWHUQDO GLVWUHVV DQG PDWHUQDO SHUFHSWLRQV RI WKH WUHDWPHQW UHJLPHQ DW 7LPH ZRXOG EH DVVRFLDWHG ZLWK 7LPH DGKHUHQFH WR WKH FKLOGn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

PAGE 52

SHUFHSWLRQV RI WKH WUHDWPHQW UHJLPHQ YDULDEOHV RQ 7LPH RXWFRPHV DUH SUHVHQWHG LQ 7DEOHV DQG 7DEOH 6XPPDU\ RI +LHUDUFKLFDO 5HJUHVVLRQ $QDO\VHV IRU 7LPH 0DWHUQDO 3HUFHSWLRQV RI WKH 7UHDWPHQW 5HJLPHQ 3UHGLFWLQJ &KLOG 6HL]XUH )UHTXHQF\ 1 f 9DULDEOHV ,QFOXGHG 3 W 5 5 $GM5 $5 ) $) &RQWUROOLQJ IRU 7 'LIILFXOW\ 7 .QRZOHGJH 7 r (IILFDF\ 7 0RGHO 'LIILFXOW\ 7 r .QRZOHGJH 7 (IILFDF\ 7 'LIILFXOW\ 7 r rr 0RGHO 'LIILFXOW\ 7 .QRZOHGJH 7 rr (IILFDF\ 7 'LIILFXOW\ 7 (IILFDF\ 7 rr 1RWH 7 7LPH 7 7LPH r3 rrS 7DEOH 6XPPDU\ RI +LHUDUFKLFDO 5HJUHVVLRQ $QDO\VHV IRU 7LPH 0DWHUQDO 3HUFHSWLRQV RI WKH 7UHDWPHQW 5HJLPHQ 3UHGLFWLQJ 6HL]XUHUHODWHG (PHUJHQF\ 5RRP 9LVLWV 1 f 9DULDEOHV ,QFOXGHG 3 W 5 5 $GM5 $5 ) $) &RQWUROOLQJ IRU 7 'LIILFXOW\ 7 .QRZOHGJH 7 (IILFDF\ 7

PAGE 53

7DEOH &RQWLQXHG 0RGHO 'LIILFXOW\ 7 rr .QRZOHGJH 7 (IILFDF\ 7 'LIILFXOW\ 7 rr 0RGHO 'LIILFXOW\ 7 .QRZOHGJH 7 (IILFDF\ 7 'LIILFXOW\ 7 r (IILFDF\ 7 1RWH 7 7LPH 7 7LPH re rre 7KH GDWD UHYHDOHG VLJQLILFDQW UHODWLRQVKLSV SDUWLDO FRUUHODWLRQVf EHWZHHQ 7LPH GLIILFXOW\ VFRUHV DQG 7LPH VHL]XUH IUHTXHQF\ U S f 7LPH HIILFDF\ VFRUHV DQG 7LPH VHL]XUH IUHTXHQF\ U S f DQG 7LPH GLIILFXOW\ VFRUHV DQG 7LPH VHL]XUHUHODWHG HPHUJHQF\ URRP YLVLWV U S f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

PAGE 54

7DEOH 3DUWLDO &RUUHODWLRQV EHWZHHQ 7LPH 0DWHUQDO 3HUFHSWLRQV RI WKH 7UHDWPHQW 5HJLPHQ DQG 7LPH 2XWFRPHV +HDOWK 6WDWXV $GKHUHQFH 0DWHUQDO 3HUFHSWLRQV RI WKH 7UHDWPHQW 5HJLPHQ 6HL]XUH )UHTXHQF\ 6HL]XUHUHODWHG (5 9LVLWV 3LOO &RXQW 3HUFHQWDJH 'LIILFXOW\ r rr (IILFDF\ rr .QRZOHGJH 1RWH rS rrS +\SRWKHVLV &RQGLWLRQ DQG 5HJLPHQUHODWHG 9DULDEOHV $GKHUHQFH DQG &KLOG +HDOWK 6WDWXV :H SUHGLFWHG WKDW FRQGLWLRQ DQG UHJLPHQUHODWHG YDULDEOHV FRPSOH[LW\ WLPH VLQFH GLDJQRVLV UHJLPHQ W\SH GLVHDVH W\SH DQG FRQFRUGDQFH EHWZHHQ WKH SUHVFULEHG DQG UHSRUWHG UHJLPHQf ZRXOG EH DVVRFLDWHG ZLWK 7LPH DGKHUHQFH WR WKH FKLOGfV PHGLFDWLRQ UHJLPHQ SLOO FRXQW SHUFHQWDJHf DQG FKLOG KHDOWK VWDWXV VHL]XUH IUHTXHQF\ DQG VHL]XUH UHODWHG HPHUJHQF\ URRP YLVLWV )LUVW ZH K\SRWKHVL]HG WKDW D WRQLFFORQLF VHL]XUH GLDJQRVLV ZRXOG EH SRVLWLYHO\ UHODWHG WR DGKHUHQFH WR WKH FKLOGnV PHGLFDWLRQ UHJLPHQ 6HFRQG ZH SUHGLFWHG WKDW FKLOG KHDOWK VWDWXV DQG WLPH VLQFH GLDJQRVLV DQG FRPSOH[LW\ RI WKH PHGLFDWLRQ UHJLPHQ KLJKHU GRVHV PXOWLSOH PHGLFDWLRQV PRUH IUHTXHQW GRVHVf ZRXOG EH QHJDWLYHO\ DVVRFLDWHG ZLWK 7LPH DGKHUHQFH WR WKH FKLOGfV PHGLFDWLRQ UHJLPHQ DQG SRVLWLYHO\ ZLWK KHDOWK VWDWXV 7KLUG ZH SUHGLFWHG WKDW SHUFHQWDJH RI FRQFRUGDQFH EHWZHHQ WKH SK\VLFLDQSUHVFULEHG PHGLFDWLRQ UHJLPHQ DQG VHOIUHSRUWHG UHJLPHQ ZRXOG EH SRVLWLYHO\ DVVRFLDWHG ZLWK 7LPH DGKHUHQFH WR WKH FKLOGnV PHGLFDWLRQ UHJLPHQ DQG QHJDWLYHO\ ZLWK KHDOWK VWDWXV 7KHVH K\SRWKHVHV ZHUH SDUWLDOO\ FRQILUPHG +LHUDUFKLFDO UHJUHVVLRQ DQDO\VHV ZHUH FRQGXFWHG WR SUHGLFW 7LPH RXWFRPHV ZLWK FRQGLWLRQ DQG

PAGE 55

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f 9DULDEOHV ,QFOXGHG 3 W 5 5 $GM 5 $5 ) $) 0RGHO 6HL]XUH W\SH 5HJLPHQ W\SH 0RGHO 6HL]XUH W\SH rr 5HJLPHQ W\SH &RQFRUGDQFH r &RPSOH[LW\ 7LPH VLQFH G[ 1RWH G[ GLDJQRVLV r3 rrS 7DEOH 6XPPDU\ RI +LHUDUFKLFDO 5HJUHVVLRQ $QDO\VHV IRU &RQGLWLRQ DQG 5HJLPHQUHODWHG 9DULDEOHV 3UHGLFWLQJ 3LOO &RXQW 3HUFHQWDJH 1 f 9DULDEOHV ,QFOXGHG 3 W 5 5 $GM 5 $5 ) $) 0RGHO 6HL]XUH W\SH 5HJLPHQ W\SH

PAGE 56

7DEOH &RQWLQXHG 0RGHO 6HL]XUH W\SH 5HJLPHQ W\SH &RQFRUGDQFH &RPSOH[LW\ 7LPH VLQFH G[ 1RWH G[ GLDJQRVLV rJ rrS 7KH GDWD UHYHDOHG VLJQLILFDQW UHODWLRQVKLSV SDUWLDO FRUUHODWLRQVf EHWZHHQ FRQFRUGDQFH VFRUHV DQG 7LPH VHL]XUHUHODWHG HPHUJHQF\ URRP YLVLWV U J f DV ZHOO DV FRQFRUGDQFH VFRUHV DQG 7LPH SLOO FRXQW SHUFHQWDJH U J f $V K\SRWKHVL]HG FRQFRUGDQFH VFRUHV ZHUH QHJDWLYHO\ UHODWHG WR 7LPH VHL]XUHUHODWHG HPHUJHQF\ URRP YLVLWV DQG SRVLWLYHO\ UHODWHG WR 7LPH SLOO FRXQW SHUFHQWDJH $V FRQFRUGDQFH VFRUHV VLJQLILFDQWO\ LQFUHDVHG HPHUJHQF\ URRP YLVLWV VLJQLILFDQWO\ GHFUHDVHG DQG SLOO FRXQW SHUFHQWDJHV VLJQLILFDQWO\ LQFUHDVHG &RQWUDU\ WR VWXG\ K\SRWKHVHV D WRQLFFORQLF VHL]XUH GLDJQRVLV ZDV QRW DVVRFLDWHG ZLWK EHWWHU DGKHUHQFH LQFUHDVHG VHL]XUH IUHTXHQF\ RU HPHUJHQF\ URRP YLVLWV 1R RWKHU VLJQLILFDQW UHODWLRQVKLSV ZHUH IRXQG ZLWK FRQGLWLRQ DQG UHJLPHQUHODWHG YDULDEOHV DQG 7LPH RXWFRPHV &RUUHODWLRQDO GDWD EHWZHHQ FRQGLWLRQ DQG UHJLPHQUHODWHG YDULDEOHV 7LPH PDWHUQDO DGKHUHQFH DQG 7LPH FKLOG KHDOWK VWDWXV DUH SUHVHQWHG LQ 7DEOH 7DEOH 3DUWLDO &RUUHODWLRQV EHWZHHQ 7LPH &RQGLWLRQ DQG 5HJLPHQUHODWHG 9DULDEOHV DQG 7LPH 2XWFRPHV rr r +HDOWK 6WDWXV $GKHUHQFH &RQGLWLRQ DQG 5HJLPHQf 6HL]XUH 6HL]XUHUHODWHG (5 3LOO &RXQW 9DULDEOHV )UHTXHQF\ 9LVLWV 3HUFHQWDJH &RPSOH[LW\f

PAGE 57

7DEOH &RQWLQXHG 7LPH 6LQFH 'LDJQRVLV 5HJLPHQ 7\SHE 'LVHDVH 7\SH &RQFRUGDQFHE r rr 1RWH r rrS +\SRWKHVLV 'HPRJUDSKLF &KDUDFWHULVWLFV $GKHUHQFH DQG &KLOG +HDOWK 6WDWXV :H SUHGLFWHG WKDW GHPRJUDSKLF FKDUDFWHULVWLFV WRWDO DQQXDO LQFRPH PDWHUQDO HGXFDWLRQ OHYHOf ZRXOG EH SRVLWLYHO\ UHODWHG WR 7LPH DGKHUHQFH WR WKH FKLOGnV PHGLFDWLRQ UHJLPHQ SLOO FRXQW SHUFHQWDJHf DQG QHJDWLYHO\ DVVRFLDWHG ZLWK FKLOG KHDOWK VWDWXV VHL]XUH IUHTXHQF\ DQG VHL]XUHUHODWHG HPHUJHQF\ URRP YLVLWVf 6SHFLILFDOO\ WRWDO KRXVHKROG DQQXDO LQFRPH DQG PDWHUQDO HGXFDWLRQ OHYHO ZRXOG EH SRVLWLYHO\ DVVRFLDWHG ZLWK 7LPH DGKHUHQFH WR WKH FKLOGnV PHGLFDWLRQ UHJLPHQ DQG QHJDWLYHO\ DVVRFLDWHG ZLWK FKLOG KHDOWK VWDWXV 7KHVH K\SRWKHVHV ZHUH SDUWLDOO\ VXSSRUWHG +LHUDUFKLFDO UHJUHVVLRQ DQDO\VHV ZHUH FRQGXFWHG WR SUHGLFW 7LPH RXWFRPHV ZLWK GHPRJUDSKLF YDULDEOHV )DPLO\ LQFRPH DSSURDFKHG VLJQLILFDQFH LH WUHQG LQ WKH SUHGLFWHG GLUHFWLRQf IRU SUHGLFWLQJ 7LPH SLOO FRXQW SHUFHQWDJH 1R RWKHU VLJQLILFDQW DVVRFLDWLRQV ZHUH IRXQG 'HPRJUDSKLF YDULDEOHV ZHUH QRW VXFFHVVIXO LQ SUHGLFWLQJ 7LPH VHL]XUH IUHTXHQF\ RU VHL]XUHUHODWHG HPHUJHQF\ URRP YLVLWV 5HJUHVVLRQ GDWD IRU GHPRJUDSKLF YDULDEOHV RQ 7LPH RXWFRPHV LV SUHVHQWHG LQ 7DEOH

PAGE 58

7DEOH 6XPPDU\ RI +LHUDUFKLFDO 5HJUHVVLRQ $QDO\VHV IRU 'HPRJUDSKLF 9DULDEOHV 3UHGLFWLQJ 3LOO &RXQW 3HUFHQWDJH 1 f 9DULDEOHV W 5 5 $GM 5 $5 ) $) ( ,QFOXGHG 0RGHO )DPLO\ LQFRPH 0RP HGXFDWLRQ &KLOG UDFH 0RP UDFH 1RWH H 7KH GDWD UHYHDOHG D VLJQLILFDQW UHODWLRQVKLS SDUWLDO FRUUHODWLRQf EHWZHHQ DQQXDO IDPLO\ LQFRPH DQG 7LPH SLOO FRXQW SHUFHQWDJH U S f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r 0RWKHUnV (GXFDWLRQ &KLOG 5DFH 0RWKHU 5DFH 1RWH rS

PAGE 59

0RGHO 7HVWLQJ 3UHGLFWRU YDULDEOHV IRU 7LPH RXWFRPHV ZHUH LQFOXGHG EDVHG RQ WKH DQDO\VHV GHVFULEHG DERYH E\ VHOHFWLQJ YDULDEOHV WKDW VKRZHG ERWK VLJQLILFDQW 3 ZHLJKWV DQG DFFRXQWHG IRU VLJQLILFDQW 5 FKDQJH 7KHVH YDULDEOHV ZHUH VXEVHTXHQWO\ LQFOXGHG LQ WHVWV RI WKH RYHUDOO JRRGQHVV RI ILW WR LGHQWLI\ SURPLVLQJ DOWHUQDWLYH UHJUHVVLRQ PRGHOV IRU XVH LQ IROORZXS UHVHDUFK VHH 7DEOH f 7KH VWDWLVWLFDO DSSURDFK LQFOXGHG WZR VWHSV f 7LPH PHDVXUHV ZHUH HQWHUHG LQ VWHS RQH RI WKH KLHUDUFKLFDO UHJUHVVLRQ DQDO\VHV LQ D IRUFHGHQWU\ IDVKLRQ DQG f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f &RQFRUGDQFH DQG DQQXDO IDPLO\ LQFRPH ZHUH HQWHUHG LQ D VWHSZLVH PDQQHU ZLWK 7LPH SLOO FRXQW SHUFHQWDJH DV WKH RXWFRPH YDULDEOH 2QO\ RQH PRGHO ZDV JHQHUDWHG ZLWK WKH FRQFRUGDQFH YDULDEOH UHWDLQHG DV D VLJQLILFDQW SUHGLFWRU 7KH FRQFRUGDQFH YDULDEOH DFFRXQWHG IRU D VLJQLILFDQW SURSRUWLRQ RI WKH YDULDQFH LQ WKH 7LPH SLOO FRXQW SHUFHQWDJH RXWFRPH PHDVXUH $GMXVWHG 5 S f 7KH UHVXOWV VXJJHVWHG WKDW WKH EHVW SUHGLFWRU RI PDWHUQDO DGKHUHQFH ZDV FRQFRUGDQFH EHWZHHQ PRWKHU DQG SK\VLFLDQ RQ WKH FKLOGfV GUXJ UHJLPHQ 7KLV PHDVXUH

PAGE 60

DFFRXQWHG IRU DSSUR[LPDWHO\ b RI WKH YDULDQFH LQ WKH 7LPH SLOO FRXQW SHUFHQWDJH PHDVXUH ,Q H[DPLQLQJ WKH VLJQLILFDQFH RI ZHLJKWV IRU DOO SUHGLFWRU YDULDEOHV LQ WKH PRGHO RQO\ WKH FRQFRUGDQFH YDULDEOH ZDV VLJQLILFDQW W S f VHH 7DEOH f 7DEOH 6XPPDU\ RI 5HJUHVVLRQ 0RGHO IRU 3UHGLFWLQJ 0DWHUQDO $GKHUHQFH 1 f 9DULDEOHV ,QFOXGHG 3 W 5 5 $GM5 $5 ) $) 0RGHO rr &RQFRUGDQFH r r 1RWH rS rr%HVW SUHGLFWLRQ PRGHO 3UHGLFWLRQ RI KHDOWK VWDWXV +LHUDUFKLFDO UHJUHVVLRQ DQDO\VHV ZHUH SHUIRUPHG WR WHVW WKH K\SRWKHVHV WKDW WKH YDULDEOHV RI PDWHUQDO VWUHVV PHDVXUHG E\ WKH 36,6)f SDUHQWFKLOG G\VIXQFWLRQ PHDVXUHG E\ WKH VXEVFDOH RI WKH 36,6)f DQG PDWHUQDO SHUFHSWLRQV RI GLIILFXOW\ DQG HIILFDF\ UHJDUGLQJ WKH FKLOGn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f 7KH UHVXOWV VXJJHVWHG WKDW WKH EHVW SUHGLFWRU RI 7LPH VHL]XUH IUHTXHQF\ ZDV SDUHQWFKLOG

PAGE 61

G\VIXQFWLRQ ZKLFK DFFRXQWHG IRU DSSUR[LPDWHO\ b RI WKH YDULDQFH LQ WKH 7LPH VHL]XUH IUHTXHQF\ PHDVXUH ,Q H[DPLQLQJ WKH VLJQLILFDQFH RI 3 ZHLJKWV IRU DOO SUHGLFWRU YDULDEOHV LQ WKH PRGHO RQO\ WKH SDUHQWFKLOG G\VIXQFWLRQ YDULDEOH ZDV VLJQLILFDQW 3 W S f VHH 7DEOH f 7DEOH 6XPPDU\ RI 5HJUHVVLRQ 0RGHOV IRU 3UHGLFWLQJ 6HL]XUH )UHTXHQF\ 1 f 9DULDEOHV ,QFOXGHG 3 W 5 5 $GM5 $5 ) $) &RQWUROOLQJ IRU 7 36,6) PHDQ 7 3& G\V 7 'LIILFXOW\ 7 (IILFDF\ 7 0RGHO rr 36,6) PHDQ 7 r 3& G\V 7 'LIILFXOW\ 7 (IILFDF\ 7 3& G\V 7 r 1RWH 36,6) 3DUHQWLQJ 6WUHVV ,QGH[ 3& G\V 3DUHQW&KLOG '\VIXQFWLRQ 7 7LPH 7 7LPH r3 rr%HVW SUHGLFWLRQ PRGHO +LHUDUFKLFDO UHJUHVVLRQ DQDO\VHV ZHUH SHUIRUPHG WR WHVW WKH K\SRWKHVHV WKDW WKH YDULDEOHV RI GLIILFXOW\ UHJDUGLQJ WKH FKLOGn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

PAGE 62

DGGHG FRQFRUGDQFH DV D SUHGLFWRU 7KH ILUVW PRGHO DFFRXQWHG IRU D VLJQLILFDQW SURSRUWLRQ RI WKH YDULDQFH LQ WKH HPHUJHQF\ URRP YLVLWV RXWFRPH PHDVXUH $GMXVWHG 5 S f ZLWK b RI WKH YDULDQFH DFFRXQWHG IRU E\ WKLV EORFN 7KH VHFRQG PRGHO DFFRXQWHG IRU D VLJQLILFDQW SURSRUWLRQ RI WKH YDULDQFH LQ WKH HPHUJHQF\ URRP YLVLWV EH\RQG PRGHO RQH $GMXVWHG 5 S f ZLWK DQ DGGLWLRQDO b RI WKH YDULDQFH DFFRXQWHG IRU E\ WKLV PRGHO 7KH UHVXOWV VXJJHVWHG WKDW WKH EHVW PRGHO IRU SUHGLFWLQJ VHL]XUHUHODWHG HPHUJHQF\ URRP YLVLWV ZDV 0RGHO DFFRXQWLQJ IRU DSSUR[LPDWHO\ b RI WKH YDULDQFH LQ VHL]XUHUHODWHG HPHUJHQF\ URRP YLVLWV ,Q H[DPLQLQJ WKH VLJQLILFDQFH RI S ZHLJKWV IRU DOO IDFWRUV LQ 0RGHO WZR EHWDV GLIILFXOW\ 3 W S f DQG FRQFRUGDQFH 3 W S f ZHUH VWDWLVWLFDOO\ VLJQLILFDQW VHH 7DEOH f 7DEOH 6XPPDU\ RI 5HJUHVVLRQ 0RGHOV IRU 3UHGLFWLQJ 6HL]XUHUHODWHG (PHUJHQF\ 5RRP 9LVLWV 1 f 9DULDEOHV ,QFOXGHG 3 W 5 5 $GM5 $5 ) $) &RQWUROOLQJ IRU 7 'LIILFXOW\ 7 0RGHO r 'LIILFXOW\ 7 'LIILFXOW\ 7 r 0RGHO rr r 'LIILFXOW\ 7 'LIILFXOW\ 7 r &RQFRUGDQFH r 1RWH 7 7LPH 7 7LPH r3 r r %HVW SUHGLFWLRQ PRGHO

PAGE 63

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

PAGE 64

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f DQG PDWHUQDO DGKHUHQFH SLOO FRXQW SHUFHQWDJHf RXWFRPHV VXJJHVW WKDW VHYHUDO PHDVXUHV 36,6) SDUHQWFKLOG G\VIXQFWLRQ GLIILFXOW\ FRQFRUGDQFH DQG DQQXDO IDPLO\ LQFRPHf SOD\ DQ LPSRUWDQW UROH LQ FKLOG KHDOWK VWDWXV DQG PDWHUQDO DGKHUHQFH RXWFRPHV

PAGE 65

&+$37(5 ',6&866,21 6DPSOH &KDUDFWHULVWLFV RI 'HPRJUDSKLFV DQG 0DWHUQDO 'LVWUHVV ,Q UHJDUGV WR WKH VWXG\ VDPSOH PRWKHUV ZHUH KLJKO\ HGXFDWHG KDYLQJ DWWHQGHG DQ DYHUDJH RI RQH \HDU RI FROOHJH ZLWK WKHLU DQQXDO IDPLO\ LQFRPH IDOOLQJ LQ WKH PLGGOH LQFRPH UDQJH WR f 7KH UDFLDO PDNHXS RI PRWKHUV FRPSOHWLQJ WKLV VWXG\ ZDV SUHGRPLQDWHO\ :KLWH bf ZLWK WUDGLWLRQDOO\ ODEHOHG PLQRULWLHV HJ $IULFDQ $PHULFDQ +LVSDQLF HWFf VKRZLQJ D IDLUO\ XQLIRUP GLVWULEXWLRQ SDWWHUQ 7ZR WKLUGV RI WKH PRWKHUV ZHUH PDUULHG DQG QHDUO\ KDOI ZRUNHG RXWVLGH WKH KRPH DW OHDVW SDUW WLPH 7KH UDFLDO SURILOH RI SDWLHQWV LQ WKLV VWXG\ LV UHSUHVHQWDWLYH RI WKH UHSRUWHG GLVWULEXWLRQ RI FKLOGUHQ ZLWK HSLOHSV\ OLYLQJ LQ WKH 8QLWHG 6WDWHV $GDPV +HQGHUVKRW t 0DUDR f 3DWLHQWV ZHUH SUHGRPLQDWHO\ :KLWH bf JLUOV bf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f DQG PDWHUQDO VWUHVV 6SHFWRU HW DO f 2QH SRVVLEOH H[SODQDWLRQ IRU WKLV ILQGLQJ PD\ EH WKDW PRWKHUV RI FKLOGUHQ ZLWK

PAGE 66

HSLOHSV\ LQ WKLV VWXG\ PD\ KDYH H[SHULHQFHG PRUH VWLJPD DQG HPRWLRQDO VWUDLQ DV ZHOO DV GDLO\ SUHVVXUHV HJ IUHTXHQW GRFWRUfV DSSRLQWPHQWVf ZKHQ WKH FKLOG ZDV PXFK \RXQJHU GXULQJ WKH LQLWLDO PRQWKV IROORZLQJ WKH FKLOGfV GLDJQRVLV 3HUKDSV PDWHUQDO VWUHVV DQG GHSUHVVLRQ VFRUHV ZHUH QRW VLJQLILFDQW EHFDXVH FRPSDUHG WR ZKHQ WKH FKLOG ZDV ILUVW GLDJQRVHG WKH FXUUHQW VLWXDWLRQ LV D FDOPHU DQG OHVV VWUHVVIXO WLPH IRU PRWKHUV 6WXGLHV LQGLFDWH WKDW WKH LQLWLDO SKDVH IROORZLQJ WKH FKLOGfV GLDJQRVLV OHDGV WR QHJDWLYH HPRWLRQDO FRQVHTXHQFHV HJ LQFUHDVHG OHYHOV RI VWUHVV DQG GHSUHVVLYH V\PSWRPDWRORJ\f IRU PRWKHUV RI FKLOGUHQ ZLWK FKURQLF LOOQHVVHV %URZQEULGJH t )LHOGLQJ $XVWLQ f ,W LV DOVR SRVVLEOH WKDW RQO\ PRWKHUV RI FKLOGUHQ ZKR H[SHULHQFHG ORZ OHYHOV RI VWUHVV RU GHSUHVVLYH V\PSWRPDWRORJ\ RSWHG WR SDUWLFLSDWH LQ WKH VWXG\ +\SRWKHVHV +\SRWKHVLV 0DWHUQDO (PRWLRQDO 'LVWUHVV $GKHUHQFH DQG &KLOG +HDOWK 6WDWXV ,W ZDV K\SRWKHVL]HG WKDW 7LPH PDWHUQDO HPRWLRQDO GLVWUHVV YDULDEOHV ZRXOG EH QHJDWLYHO\ UHODWHG WR 7LPH DGKHUHQFH WR WKH FKLOGn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

PAGE 67

UHVXOWV DUH FRQVLVWHQW ZLWK SUHYLRXV ILQGLQJV WKDW LQFUHDVHG SDUHQWDO VWUHVV OHYHOV DUH UHODWHG WR DQ LQFUHDVH LQ FKLOG VHL]XUH IUHTXHQF\ DQG VHYHULW\ 6SHFWRU HW DO f 6SHFWRU HW DO f LQYHVWLJDWHG WKH HIILFDF\ RI D UDQJH RI SV\FKRORJLFDO LQWHUYHQWLRQV DQG IRXQG WKDW VKRUWWHUP JURXS LQWHUYHQWLRQV FRPELQLQJ VXSSRUWLYH DQG FRJQLWLYH EHKDYLRUDO FRPSRQHQWV WR LPSURYH FRSLQJ VWUDWHJLHV DQG GHFUHDVH VWUHVV PD\ EH HIIHFWLYH IRU LPSURYLQJ VHL]XUH FRQWURO 7KHVH ILQGLQJV PD\ EH DWWULEXWDEOH WR WKH LGHD WKDW PRWKHUV ZKR KDYH FKLOGUHQ ZLWK LQFUHDVHG OHYHOV RI VHL]XUH IUHTXHQF\ PD\ EH PRUH VWUHVVHG DERXW WKHLU FKLOGnV KHDOWK DQG ZHOOEHLQJ ZKLFK PD\ FDUU\ RYHU LQWR LQFUHDVHG OHYHOV RI RYHUDOO PDWHUQDO VWUHVV 2I QRWH LV WKDW LQFUHDVHG SDWLHQW VWUHVV OHYHOV KDYH EHHQ DVVRFLDWHG ZLWK LQFUHDVHG VHL]XUH IUHTXHQF\ .\QJDV f .\QJDV f VWXGLHG WKH FRPSOLDQFH RI DGROHVFHQWV ZLWK HSLOHSV\ DQG IRXQG WKDW JRRG PRWLYDWLRQ D VWURQJ VHQVH RI QRUPDOLW\ VXEMHFWLYH RXWFRPH HQHUJ\ DQG ZLOOSRZHU VXSSRUW IURP SDUHQWV SK\VLFLDQV DQG QXUVHV D SRVLWLYH DWWLWXGH WRZDUGV WKH GLVHDVH DQG LWV WUHDWPHQW QR WKUHDW WR VRFLDO DQG HPRWLRQDO ZHOOn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t ,DFRQR f +RGHV DQG *DUUDOGD f IRXQG WKDW PRWKHUV ZHUH PRUH KRVWLOH WRZDUGV WKHLU FKLOG ZLWK HSLOHSV\ WKDQ WRZDUGV KHDOWK\ VLEOLQJV

PAGE 68

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t +RGHV f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f DQG 7LPH PDWHUQDO GLVWUHVV DQG 7LPH PDWHUQDO SHUFHSWLRQV RI WKH WUHDWPHQW UHJLPHQ ZRXOG EH DVVRFLDWHG ZLWK 7LPH DGKHUHQFH WR WKH FKLOGfV PHGLFDWLRQ UHJLPHQ DQG FKLOG KHDOWK VWDWXV 6LPLODU WR +\SRWKHVLV D PL[HG SDWWHUQ RI UHVXOWV ZDV REWDLQHG 0RWKHUV ZKR UDWHG KLJKHU OHYHOV RI GLIILFXOW\ ZLWK WKHLU FKLOGnV PHGLFDO UHJLPHQ DW 7LPH FRQFRPLWDQWO\ UHSRUWHG LQFUHDVHG VHL]XUH IUHTXHQF\ OHYHOV DQG VHL]XUHUHODWHG HPHUJHQF\ URRP YLVLWV IRU WKHLU

PAGE 69

FKLOGUHQ DW 7LPH 0DWHUQDO SHUFHSWLRQV UHJDUGLQJ GLIILFXOW\ RI WKH FKLOGnV PHGLFDO UHJLPHQ ZHUH SRVLWLYHO\ UHODWHG WR FKLOG KHDOWK VWDWXV RXWFRPH YDULDEOHV 7KH UHODWLRQVKLS EHWZHHQ UHJLPHQ FRPSOH[LW\ DQG PDWHUQDO SHUFHSWLRQV RI UHJLPHQ GLIILFXOW\ DSSURDFKHG VLJQLILFDQFH U J f 7KHVH ILQGLQJV PD\ EH UHODWHG WR WKH SUHPLVH WKDW FKLOGUHQ ZLWK SRRUHU KHDOWK RXWFRPHV HJ LQFUHDVHG OHYHOV RI VHL]XUH IUHTXHQF\ DQG VHL]XUHUHODWHG HPHUJHQF\ URRP YLVLWVf PD\ KDYH PRUH GLIILFXOW DQG FRPSOH[ UHJLPHQV LQ JHQHUDO GXH WR WKH QDWXUH RI WKHLU LOOQHVV )XUWKHUPRUH PRWKHUV ZKR UHSRUWHG KLJKHU OHYHOV RI UHJLPHQ HIILFDF\ KDG FKLOGUHQ ZLWK ORZHU OHYHOV RI VHL]XUH IUHTXHQF\ 1R VLJQLILFDQW UHODWLRQVKLSV ZHUH IRXQG EHWZHHQ PDWHUQDO SHUFHSWLRQV RI WKH WUHDWPHQW UHJLPHQ YDULDEOHV HJ UHJLPHQ GLIILFXOW\ HIILFDF\ DQG NQRZOHGJHf DQG DGKHUHQFH 2YHUDOO ILQGLQJV VXJJHVW WKDW PDWHUQDO SHUFHSWLRQV RI WKH WUHDWPHQW UHJLPHQ DUH VLJQLILFDQWO\ UHODWHG WR SRRUHU KHDOWK RXWFRPHV IRU FKLOGUHQ ZLWK HSLOHSV\ EXW QRW VLJQLILFDQWO\ UHODWHG WR PDWHUQDO DGKHUHQFH $ SRVVLEOH H[SODQDWLRQ IRU WKLV QXOO ILQGLQJ LV WKDW PRWKHUV RI FKLOGUHQ ZLWK PRUH GLIILFXOW UHJLPHQV DQG SRRUHU KHDOWK RXWFRPHV DUH PRUH DGKHUHQW WR WKHLU FKLOGnV FRQGLWLRQ LQ DQ HIIRUW WR PLQLPL]H QHJDWLYH KHDOWK RXWFRPHV HJ LQFUHDVHG VHL]XUH IUHTXHQF\ DQG VHL]XUHUHODWHG HPHUJHQF\ YLVLWVf $GGLWLRQDOO\ UHODWLRQVKLSV EHWZHHQ PDWHUQDO SHUFHSWLRQV RI WKH WUHDWPHQW UHJLPHQ DQG DGKHUHQFH PD\ QRW KDYH EHHQ LGHQWLILHG EHFDXVH WKH TXHVWLRQV FRPSULVLQJ UHJLPHQ SHUFHSWLRQV DUH SUHOLPLQDU\ LQ QDWXUH DQG DUH LQ QHHG RI IXUWKHU H[SORUDWLRQ PRGLILFDWLRQ DQG XOWLPDWHO\ YDOLGDWLRQ +\SRWKHVLV &RQGLWLRQ DQG 5HJLPHQUHODWHG 9DULDEOHV $GKHUHQFH DQG &KLOG +HDOWK 6WDWXV :H SUHGLFWHG WKDW 7LPH FRQGLWLRQ DQG UHJLPHQUHODWHG YDULDEOHV ZRXOG EH DVVRFLDWHG ZLWK 7LPH DGKHUHQFH WR WKH FKLOGfV PHGLFDWLRQ UHJLPHQ DQG FKLOG KHDOWK

PAGE 70

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nV PHGLFDWLRQ UHJLPHQ ZLWK WKH SUHVFULEHG UHJLPHQ PD\ QRW KDYH DGPLQLVWHUHG WKH FKLOGnV PHGLFDWLRQ DV SUHVFULEHG UHVXOWLQJ LQ QHJDWLYH FKLOG KHDOWK RXWFRPHV 7KH UHVXOWV DOVR GHPRQVWUDWH WKDW PRWKHUV ZKR ZHUH DEOH WR UHSRUW WKHLU FKLOGnV PHGLFDWLRQ UHJLPHQ ZHUH PRUH DGKHUHQW WR WKHLU FKLOGnV PHGLFDWLRQ UHJLPHQ ,W IROORZV IURP WKHVH ILQGLQJV WKDW PRWKHUV ZKR ZHUH XQDEOH WR DFFXUDWHO\ UHSRUW WKHLU FKLOGnV PHGLFDWLRQ UHJLPHQ PD\ QRW KDYH IROORZHG WKH SUHVFULEHG UHJLPHQ DQG LQ WXUQ UHSRUWHG LQFUHDVHG OHYHOV RI HPHUJHQF\ URRP YLVLWV UHODWHG WR WKHLU FKLOGnV VHL]XUHV +LJK FRQFRUGDQFH UDWLQJV PD\ EH UHSUHVHQWDWLYH RI JRRG SDUHQWSK\VLFLDQ FRPPXQLFDWLRQ 3HWHUVRQ HW DO f IRXQG WKDW JRRG SDWLHQWGRFWRU FRPPXQLFDWLRQ KDV EHHQ UHODWHG WR EHWWHU DGKHUHQFH 7KH VLJQLILFDQW SUHGLFWLYH UHODWLRQVKLS GHPRQVWUDWHG EHWZHHQ FRQFRUGDQFH UDWLQJV DQG RXWFRPHV HJ VHL]XUH IUHTXHQF\ DQG PDWHUQDO DGKHUHQFHf DUH RI WKHRUHWLFDO DQG SUDFWLFDO LPSRUWDQFH LQ

PAGE 71

WKDW FRQFRUGDQFH FDQ EH HDVLO\ DVVHVVHG GXULQJ HDFK FKLOGnV FOLQLF YLVLW DQG IXUWKHU H[DPLQHG IRU LWV UHODWLRQVKLS WR FKLOG KHDOWK VWDWXV RXWFRPHV DQG PDWHUQDO DGKHUHQFH +\SRWKHVLV 'HPRJUDSKLF &KDUDFWHULVWLFV $GKHUHQFH DQG &KLOG +HDOWK 6WDWXV :H SUHGLFWHG WKDW 7LPH GHPRJUDSKLF FKDUDFWHULVWLFV ZRXOG EH SRVLWLYHO\ UHODWHG WR 7LPH DGKHUHQFH WR WKH FKLOGfV PHGLFDWLRQ UHJLPHQ DQG QHJDWLYHO\ ZLWK FKLOG KHDOWK VWDWXV +\SRWKHVLV ZDV RQO\ SDUWLDOO\ VXSSRUWHG LQ WKDW RQH GHPRJUDSKLF YDULDEOH ZDV DVVRFLDWHG ZLWK DGKHUHQFH IDPLO\ LQFRPH ZDV UHODWHG WR WKH SLOO FRXQW SHUFHQWDJH PHDVXUH RI DGKHUHQFH )DPLOLHV ZLWK JUHDWHU ILQDQFLDO UHVRXUFHV VKRZ EHWWHU PDWHUQDO DGKHUHQFH WR WKH FKLOGf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f +RZHYHU HDUOLHU UHVHDUFK LQGLFDWHG WKDW ORZHU VRFLRHFRQRPLF VWDWXV DQG ORZHU SDUHQWDO HGXFDWLRQ OHYHOV ZHUH FRUUHODWHG VLJQLILFDQWO\ ZLWK QRQDGKHUHQFH IRU DVWKPD F\VWLF ILEURVLV DQG UHQDO GLVHDVH 3DWWHUVRQ 7HEEL HW DO f DQG ORZHU VRFLRHFRQRPLF VWDWXV ZDV UHODWHG WR SDUHQWDO QRQDGKHUHQFH %HFNHU t 0DLPDQ 5DSRII t &KULVWRSKHUVRQ f 2QH H[SODQDWLRQ IRU WKHVH ILQGLQJV LV WKDW WKH SRSXODWLRQ DW WKH 8QLYHUVLW\ RI )ORULGDnV +HDOWK 6FLHQFH &HQWHU 8) +6&f PD\ QRW EH UHSUHVHQWDWLYH RI WKH ODUJHU GHPRJUDSKLF GLVWULEXWLRQ ,W LV IHDVLEOH WKDW ZHDOWKLHU DQG PRUH HGXFDWHG IDPLOLHV EHFDXVH RI WKHLU ILQDQFLDO VWDELOLW\ DUH EHWWHU HTXLSSHG WR PHHW WKH GHPDQGV RI WKHLU FKLOGnV PHGLFDO DSSRLQWPHQWV LQFOXGLQJ WLPH RII

PAGE 72

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nV PHGLFDWLRQ UHJLPHQ HPHUJH DV LPSRUWDQW YDULDEOHV 7KHVH PHDVXUHV PD\ EH EHWWHU SUHGLFWRUV RI VHL]XUHUHODWHG HPHUJHQF\ URRP YLVLWV FRPSDUHG WR PDWHUQDO HPRWLRQDO GLVWUHVV PHDVXUHV EHFDXVH WKH\ VSHFLILFDOO\ DVVHVV PDWHUQDO XQGHUVWDQGLQJ RI WKH FKLOGfV UHJLPHQ DQG LWV GLIILFXOW\ ,W LV VXJJHVWHG WKDW PRUH GLIILFXOW UHJLPHQV RU SRRU FRQFRUGDQFH DUH PRUH

PAGE 73

FORVHO\ UHODWHG WR VHL]XUHUHODWHG HPHUJHQF\ URRP YLVLWV WKDQ PDWHUQDO HPRWLRQDO GLVWUHVV YDULDEOHV ,Q SUHGLFWLQJ DGKHUHQFH PDWHUQDO HPRWLRQDO GLVWUHVV GRHV QRW SOD\ D NH\ UROH EXW WKH UHJLPHQUHODWHG YDULDEOH RI FRQFRUGDQFH LQGHHG GRHV &RQFRUGDQFH LV OLNHO\ WR EH PRUH SUHGLFWLYH RI PDWHUQDO DGKHUHQFH WKDQ PDWHUQDO HPRWLRQDO GLVWUHVV YDULDEOHV EHFDXVH FRQFRUGDQFH WDSV VSHFLILFDOO\ LQWR WKH PRWKHUfV XQGHUVWDQGLQJ RI WKH FKLOGfV UHJLPHQ 2XU ILQGLQJV VWURQJO\ LQGLFDWH WKDW LQFRUUHFW UHSRUWLQJ RI WKH FKLOGf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n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

PAGE 74

+RZHYHU WKH WLPH ODJ ZDV UHODWLYHO\ VKRUW LH RQH PRQWKf DQG QHHGV WR EH H[WHQGHG LQ IXWXUH VWXGLHV WR WHVW WKH UHOLDELOLW\ RI WKH UHVXOWV RYHU D ORQJHU WLPH LQWHUYDO 7KH VWXG\ DOVR DOORZHG XV WR WDS LQWR PDWHUQDO SHUFHSWLRQV RI WKH WUHDWPHQW UHJLPHQ ZLWK WKH FUHDWLRQ RI TXHVWLRQV WR DVVHVV GLIILFXOW\ HIILFDF\ DQG NQRZOHGJH UHJDUGLQJ WKH FKLOGf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f DV GLG WKH 36,6) RYHUDOO PHDQ S f +DYLQJ D ODUJHU VDPSOH ZRXOG DOVR LQFUHDVH WKH SRZHU RI SUHGLFWLRQ EHFDXVH YDULDQFH RQ WKH PHDVXUHV ZRXOG

PAGE 75

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f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f ODUJHU VDPSOH VL]H f DVVHVVPHQW RI FKLOG VWUHVV

PAGE 76

OHYHOV f YDOLGDWLRQ RI PDWHUQDO SHUFHSWLRQV RI WKH WUHDWPHQW UHJLPHQ TXHVWLRQV f EHWWHU PHDVXUHV RI DGKHUHQFH f LQFOXVLRQ RI PXOWLSOH VWXG\ VLWHV f IDWKHU GDWD DQG f FRQWURO JURXS :LWK D ODUJHU VDPSOH VL]H VWUXFWXUDO HTXDWLRQ PRGHOLQJ 6(0f SURFHGXUHV ZRXOG SHUPLW VSHFLILF IDFWRU WHVWLQJ UDWKHU WKDQ WKH FXUUHQW YDULDEOHOHYHO WHVWLQJ PRGHO 0XOWLSOH UHJUHVVLRQ DQG RWKHU PXOWLYDULDWH WHFKQLTXHV FDQ DGGUHVV D QXPEHU RI VWDWLVWLFDO TXHVWLRQV EXW KDYH D FRPPRQ OLPLWDWLRQ LQ WKDW HDFK WHFKQLTXH FDQ H[DPLQH RQO\ D VLQJOH UHODWLRQVKLS DW D WLPH 6(0 FRPEDWV WKLV SUREOHP DQG H[DPLQHV D VHULHV RI SUHGLFWLYH UHODWLRQVKLSV VLPXOWDQHRXVO\ 6(0 WHFKQLTXHV DUH GLVWLQJXLVKHG E\ WZR FKDUDFWHULVWLFV f HVWLPDWLRQ RI PXOWLSOH DQG LQWHUUHODWHG GHSHQGHQFH UHODWLRQVKLSV DQG f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

PAGE 77

:LWK SURSHU IXQGLQJ IXWXUH VWXGLHV FRXOG LQFOXGH EORRG OHYHOV SKDUPDF\ UHILOO LQIRUPDWLRQ DQG HOHFWURQLF PHGLFDWLRQ OLG GHYLFHV %ORRG OHYHOV FDQ EH GUDZQ DV D SDUW RI WKH VWXG\ SURWRFRO LQ RUGHU WR KDYH PRUH FRQVLVWHQW GDWD UHJDUGLQJ OHYHOV RI PHGLFDWLRQ LQ WKH EORRGVWUHDP IRU HDFK VWXG\ SDUWLFLSDQW 7KH SK\VLFLDQ FDQ DLG LQ LGHQWLI\LQJ LI WKH EORRG OHYHO IDOOV LQWR WKH WKHUDSHXWLF RU QRQWKHUDSHXWLF FDWHJRU\ 6WXG\ SURWRFRO FDQ DOVR LQFOXGH SHUPLVVLRQ WR WUDFN WKH SDUWLFLSDQWf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b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nV PHGLFDO UHJLPHQ $Q LPSRUWDQW FOLQLFDO JRDO LV WR DVVHVV PDWHUQDO HPRWLRQDO GLVWUHVV PDWHUQDO SHUFHSWLRQV RI WKH WUHDWPHQW UHJLPHQ

PAGE 78

FRQGLWLRQ DQG UHJLPHQUHODWHG DQG GHPRJUDSKLF YDULDEOHV WKDW PD\ LPSDFW PDWHUQDO DGKHUHQFH WR WKH FKLOGn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nV PHGLFDO UHJLPHQ /DVWO\ WKRVH PRWKHUV ZLWK ORZHU HGXFDWLRQ OHYHOV PD\ QHHG DGGLWLRQDO DVVLVWDQFH DQG DGGLWLRQDO WLPH LQ RUGHU WR SURSHUO\ XQGHUVWDQG WKH IXOO QDWXUH RI WKHLU FKLOGnV PHGLFDWLRQ UHJLPHQ ,Q FRQFOXVLRQ PDWHUQDO VWUHVV SHUFHSWLRQV RI WKH WUHDWPHQW UHJLPHQ UHSRUW RI WKH UHJLPHQ DQG HGXFDWLRQ OHYHO FDQ EH TXLFNO\ DVVHVVHG LQ WKH FOLQLFDO VHWWLQJ LQ RUGHU WR KDYH D EHQHILFLDO HIIHFW RQ FKLOG KHDOWK VWDWXV DQG DGKHUHQFH RXWFRPHV

PAGE 79

$33(1',; $ %$&.*5281' ,1)250$7,21 )250 7,0( 'DWH %$&.*5281' ,1)250$7,21 &KLOGnV 1DPH 0RWKHUnV 1DPH $GGUHVV 3KRQH f &KLOGnV 6H[ 0DOH )HPDOH &KLOGnV $JH &KLOGnV %LUWKGDWH PRQWK GD\ \HDU 0RWKHUnV $JH 0RWKHUnV %LUWKGDWH PRQWK GD\ \HDU &KLOGnV 5DFH :KLWH $IULFDQ $PHULFDQ +LVSDQLF $PHULFDQ $VLDQ $PHULFDQ 1DWLYH $PHULFDQ %LUDFLDO 2WKHU 0RWKHUnV 0DULWDO 6WDWXV 6LQJOH1HYHU PDUULHG 0DUULHG 6HSDUDWHG 'LYRUFHG :LGRZHG 1XPEHU RI
PAGE 80

0RWKHUnV 2FFXSDWLRQ &XUUHQW (PSOR\PHQW 6WDWXV SOHDVH FLUFOH DOO WKDW DSSO\f 8QHPSOR\HG 9ROXQWHHU (PSOR\HG SDUWWLPH (PSOR\HG IXOOWLPH +RPHPDNHU 6WXGHQW 5HWLUHG 2WKHU 3OHDVH GHVFULEH
PAGE 81

3OHDVH QDPH \RXU FKLOGnV VHL]XUH GLDJQRVLV 2YHU WKH SDVW PRQWK KRZ PDQ\ VHL]XUHV GLG \RXU FKLOG KDYH SHU ZHHN" SHU PRQWK" 3OHDVH UDWH KRZ PXFK \RX NQRZ DERXW \RXU FKLOGnV VHL]XUH GLVRUGHU 1RW 6RPHZKDW .QRZOHGJHDEOH .QRZOHGJHDEOH 3OHDVH UDWH KRZ PXFK \RX NQRZ DERXW \RXU FKLOGnV VHL]XUH PHGLFDWLRQVf 1RW 6RPHZKDW .QRZOHGJHDEOH .QRZOHGJHDEOH 3OHDVH UDWH KRZ PXFK \RX NQRZ DERXW WKH W\SH RI \RXU FKLOGnV VHL]XUH PHGLFDWLRQVf 1RW 6RPHZKDW .QRZOHGJHDEOH .QRZOHGJHDEOH ([WUHPHO\ .QRZOHGJHDEOH ([WUHPHO\ .QRZOHGJHDEOH ([WUHPHO\ .QRZOHGJHDEOH 3OHDVH UDWH KRZ PXFK \RX NQRZ DERXW WKH WLPLQJ RI \RXU FKLOGnV VHL]XUH PHGLFDWLRQVf 1RW 6RPHZKDW ([WUHPHO\ .QRZOHGJHDEOH .QRZOHGJHDEOH .QRZOHGJHDEOH ,I DSSOLFDEOH SOHDVH UDWH KRZ PXFK \RX NQRZ DERXW WKH QXPEHU RI \RXU FKLOGnV VHL]XUH PHGLFDWLRQVf 1RW .QRZOHGJHDEOH 6RPHZKDW .QRZOHGJHDEOH ([WUHPHO\ 1$ .QRZOHGJHDEOH 3OHDVH UDWH KRZ PXFK \RX NQRZ DERXW WKH GRVLQJ RI \RXU FKLOGnV VHL]XUH PHGLFDWLRQVf 1RW 6RPHZKDW .QRZOHGJHDEOH .QRZOHGJHDEOH ([WUHPHO\ .QRZOHGJHDEOH 3OHDVH UDWH KRZ JRRG \RX IHHO \RXU FKLOGnV VHL]XUH PHGLFDWLRQV DUH 1RW 6RPHZKDW *RRG *RRG ([WUHPHO\ *RRG 3OHDVH UDWH KRZ JRRG \RX IHHO \RX DUH LQ PDQDJLQJ \RXU FKLOGnV VHL]XUH PHGLFDWLRQV 1RW 6RPHZKDW *RRG *RRG ([WUHPHO\ *RRG

PAGE 82

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

PAGE 83

$33(1',; % %$&.*5281' ,1)250$7,21 )250 7,0( 'DWH %$&.*5281' ,1)250$7,21 &KLOGnV 1DPH 0RWKHUnV 1DPH 3OHDVH OLVW WKH QDPHVf RI WKH DQWLVHL]XUH PHGLFDWLRQVf \RXU FKLOG LV FXUUHQWO\ WDNLQJ DQG WKH GRVH KRZ PXFK DQG KRZ PDQ\ WLPHV SHU GD\f 3OHDVH UDWH 1RW 'LIILFXOW KRZ GLIILFXOW LW LV WR PDQDJH ZKLFK W\SH RI PHGLFDWLRQ WR JLYH \RXU FKLOG 6RPHZKDW 'LIILFXOW 3OHDVH UDWH KRZ GLIILFXOW LW LV WR PDQDJH KRZ PXFK PHGLFDWLRQ WR JLYH \RXU FKLOG 1RW 6RPHZKDW 'LIILFXOW 'LIILFXOW 3OHDVH UDWH KRZ GLIILFXOW LW LV WR PDQDJH ZKHQ WR JLYH PHGLFDWLRQ WR JLYH \RXU FKLOG 1RW 6RPHZKDW 'LIILFXOW 'LIILFXOW 3OHDVH QDPH \RXU FKLOGnV VHL]XUH GLDJQRVLV 2YHU WKH SDVW PRQWK KRZ PDQ\ VHL]XUHV GLG \RXU FKLOG KDYH SHU ZHHN" SHU PRQWK" ([WUHPHO\ 'LIILFXOW ([WUHPHO\ 'LIILFXOW ([WUHPHO\ 'LIILFXOW 3OHDVH UDWH KRZ PXFK \RX NQRZ DERXW \RXU FKLOGnV VHL]XUH GLVRUGHU 1RW 6RPHZKDW .QRZOHGJHDEOH .QRZOHGJHDEOH 3OHDVH UDWH PXFK \RX NQRZ DERXW \RXU FKLOGnV VHL]XUH PHGLFDWLRQVf 1RW 6RPHZKDW .QRZOHGJHDEOH .QRZOHGJHDEOH ([WUHPHO\ .QRZOHGJHDEOH ([WUHPHO\ .QRZOHGJHDEOH

PAGE 84

3OHDVH UDWH KRZ PXFK \RX NQRZ DERXW WKH W\SH RI \RXU FKLOGnV VHL]XUH PHGLFDWLRQVf 1RW 6RPHZKDW ([WUHPHO\ .QRZOHGJHDEOH .QRZOHGJHDEOH .QRZOHGJHDEOH 3OHDVH UDWH KRZ PXFK \RX NQRZ DERXW WKH WLPLQJ RI \RXU FKLOGnV VHL]XUH PHGLFDWLRQVf 1RW 6RPHZKDW ([WUHPHO\ .QRZOHGJHDEOH .QRZOHGJHDEOH .QRZOHGJHDEOH ,I DSSOLFDEOH SOHDVH UDWH KRZ PXFK \RX NQRZ DERXW WKH QXPEHU RI \RXU FKLOGnV VHL]XUH PHGLFDWLRQVf 1RW 6RPHZKDW ([WUHPHO\ 1$ .QRZOHGJHDEOH .QRZOHGJHDEOH .QRZOHGJHDEOH 3OHDVH UDWH KRZ PXFK \RX NQRZ DERXW WKH GRVLQJ RI YRXU FKLOGnV VHL]XUH PHGLFDWLRQIVf 1RW 6RPHZKDW ([WUHPHO\ .QRZOHGJHDEOH .QRZOHGJHDEOH .QRZOHGJHDEOH 3OHDVH UDWH KRZ JRRG \RX IHHO \RXU FKLOGnV VHL]XUH PHGLFDWLRQV DUH 1RW 6RPHZKDW ([WUHPHO\ *RRG *RRG *RRG 3OHDVH UDWH KRZ JRRG \RX IHHO \RX DUH LQ PDQDJLQJ \RXU FKLOGnV VHL]XUH PHGLFDWLRQV 1RW 6RPHZKDW ([WUHPHO\ *RRG *RRG *RRG 3OHDVH GHVFULEH ZKDW W\SH RI LQIRUPDWLRQ ZRXOG EH KHOSIXO WR IXUWKHU \RXU XQGHUVWDQGLQJ RI VHL]XUH GLVRUGHUV 3OHDVH GR QRW DQVZHU WKH IROORZLQJ WKH UHVHDUFKHUV ZLOO ILOO RXW WKLV SRUWLRQ )DPLO\ ,' 1RWHV

PAGE 85

$33(1',; & 0(',&$/ &+$57 ,1)250$7,21 )250 7,0( 0(',&$/ &+$57 ,1)250$7,21 3DWLHQWnV 1DPH 3DWLHQWnV '2% $JH 'DWH RI 'LDJQRVLV 1HZO\ 'LDJQRVHG \HV QR ,I QR WLPH VLQFH GLDJQRVLV PRQWKVf 6HL]XUH )UHTXHQF\ RYHU WKH SDVW PRQWKfSHU ZHHN SHU PRQWK 6HL]XUH 7\SHVf ^WKH W\SHVf RI VHL]XUHV \RXU FKLOG LV GLDJQRVHG ZLWK` 0HGLFDWLRQVf DQG 'RVH ,QIR (5 9LVLWV 'DWHV %ORRG /HYHOV 3KDUPDF\ 5HILOO ,QIRUPDWLRQ b &RQFRUGDQFH EHWZHHQ SUHVFULEHG PHGLFDWLRQ UHJLPHQ DQG UHSRUWHG PHGLFDWLRQ UHJLPHQ &RPSOH[LW\

PAGE 86

$33(1',; 0(',&$/ &+$57 ,1)250$7,21 )250 7,0( 0(',&$/ &+$57 ,1)250$7,21 3DWLHQWnV 1DPH 3DWLHQWnV '2% $JH 'DWH RI 'LDJQRVLV 1HZO\ 'LDJQRVHG \HV QR ,I QR WLPH VLQFH GLDJQRVLV PRQWKVf 6HL]XUH )UHTXHQF\ RYHU WKH SDVW PRQWKfSHU ZHHN SHU PRQWK 6HL]XUH 7\SHVf ^WKH W\SHVf RI VHL]XUHV \RXU FKLOG LV GLDJQRVHG ZLWK` 0HGLFDWLRQVf DQG 'RVH ,QIR (5 9LVLWV 'DWHV %ORRG /HYHOV 3KDUPDF\ 5HILOO ,QIRUPDWLRQ b &RQFRUGDQFH EHWZHHQ SUHVFULEHG PHGLFDWLRQ UHJLPHQ DQG UHSRUWHG PHGLFDWLRQ UHJLPHQ &RPSOH[LW\

PAGE 87

/,67 2) 5()(5(1&(6 $ELGLQ 55 f 3DUHQWLQJ 6WUHVV ,QGH[ 3URIHVVLRQDO 0DQXDO UG HGf 2GHVVD 3V\FKRORJLFDO $VVHVVPHQW 5HVRXUFHV $GDPV 3) (OHQGHUVKRW *( t 0DUDR 0$ f &XUUHQW (VWLPDWHV IURP WKH 1DWLRQDO +HDOWK ,QWHUYLHZ 6XUYH\ 1DWLRQDO &HQWHU IRU +HDOWK 6WDWLVWLFV 9LWDO +HDOWK 6WDWLVWLFV f $QGHUVRQ %$XVODQGHU :) -XQJ '& 0LOOHU -3 t 6DQWLDJR -9 f $VVHVVLQJ IDPLO\ VKDULQJ RI GLDEHWHV UHVSRQVLELOLWLHV -RXUQDO RI 3HGLDWULF 3V\FKRORJ\ $SSOHWRQ 5 t *LEEV f (SLOHSV\ LQ &KLOGKRRG DQG $GROHVFHQFH 0DOGHQ %ODFNZHOO 6FLHQFH ,QFRUSRUDWHG $XVWLQ -. f 3UHGLFWLQJ SDUHQWDO DQWLFRQYXOVDQW PHGLFDWLRQ FRPSOLDQFH XVLQJ WKH WKHRU\ RI UHDVRQHG DFWLRQ -RXUQDO RI 3HGLDWULF 1XUVLQJ f %HFN '( )HQQHOO 56
PAGE 88

&XPPLQJV (0 t 'DYLHV 37 f 0DWHUQDO GHSUHVVLRQ DQG FKLOG GHYHORSPHQW -RXUQDO RI &KLOG 3V\FKRORJ\ DQG 3V\FKLDWU\ f 'RZQH\ t &R\QH f &KLOGUHQ RI GHSUHVVHG SDUHQWV $Q LQWHJUDWLYH UHYLHZ 3V\FKRORJLFDO %XOOHWLQ f )ULHGPDQ ,0 /LWW ,) .LQJ +HQVRQ 5 +ROW]PDQ +DOYHUVRQ t .UDHPHU + f &RPSOLDQFH ZLWK DQWLFRQYXOVDQW WKHUDS\ E\ HSLOHSWLF \RXWK -RXUQDO RI $GROHVFHQW +HDOWK &DUH *XGDV /.URRFKHU *3 t :\SOM f 3HUFHSWLRQV RI PHGLFDO FRPSOLDQFH LQ FKLOGUHQ DQG DGROHVFHQWV ZLWK F\VWLF ILEURVLV 'HYHORSPHQWDO t %HKDYLRUDO 3HGLDWULFV +D]]DUG $ +XWFKLQVRQ 6t .UDZLHFNL 1 f )DFWRUV UHODWHG WR DGKHUHQFH WR PHGLFDWLRQ UHJLPHQV LQ SHGLDWULF VHL]XUH SDWLHQWV -RXUQDO RI 3HGLDWULF 3V\FKRORJ\ f +RGHV 0 t *DUUDOGD 0( f 0DWHUQDO H[SUHVVHG HPRWLRQ DQG DGMXVWPHQW LQ FKLOGUHQ ZLWK HSLOHSV\ -RXUQDO RI &KLOG 3V\FKRORJ\ DQG 3V\FKLDWU\ -DFREVHQ $0 +DXVHU 67 :ROIVGRUI -, +RXOLKDQ 0LOOH\ -( +HUVNRZLW] 5' :HUWOLHE t :DWW ( f 3V\FKRORJLF SUHGLFWRUV RI FRPSOLDQFH LQ FKLOGUHQ ZLWK UHFHQW RQVHW GLDEHWHV PHOOLWXV -RXUQDO RI 3HGLDWULFV .\QJDV + f &RPSOLDQFH ZLWK KHDOWK UHJLPHQV RI DGROHVFHQWV ZLWK HSLOHSV\ 6HL]XUH /D*UHFD $0 )ROODQVEHH t 6N\OHU -6 f 'HYHORSPHQWDO DQG EHKDYLRUDO DVSHFWV RI GLDEHWHV PDQDJHPHQW LQ \RXQJVWHUV &KLOGUHQfV +HDOWK &DUH /RUHQ] 5$ &KULVWHQVHQ 1. t 3LFKHUW -: f 'LHWUHODWHG NQRZOHGJH VNLOO DQG DGKHUHQFH DPRQJ FKLOGUHQ ZLWK LQVXOLQGHSHQGHQW GLDEHWHV PHOOLWXV 3HGLDWULFV 0F&DXO .' *ODVJRZ 5( t 6FKDIHU /& f 'LDEHWHV UHJLPHQ EHKDYLRUV 3UHGLFWLQJ DGKHUHQFH 0HGLFDO &DUH 0XUUD\ / 6LQFODLU &RRSHU 3 'XFRXUQDX 3 7XUQHU 3 t 6WHLQ $ f 7KH VRFLRHPRWLRQDO GHYHORSPHQW RI \HDUROG FKLOGUHQ RI SRVWQDWDOO\ GHSUHVVHG PRWKHUV -RXUQDO RI &KLOG 3V\FKRORJ\ DQG 3V\FKLDWU\ f 0\HUV /% t 0LGHQFH (GVf f $GKHUHQFH WR WUHDWPHQW LQ PHGLFDO FRQGLWLRQV 1HZ 'HOKL +DUZRRG $FDGHPLF 3XEOLVKHUV

PAGE 89

1H]X $0 (Gf f 3UDFWLWLRQHUnV JXLGH WR HPSLULFDOO\ EDVHG PHDVXUHV RI GHSUHVVLRQ 1HZ
PAGE 90

:\QQ .6 t (FNHO (0 f -XYHQLOH UKHXPDWRLG DUWKULWLV DQG KRPH SK\VLFDO WKHUDS\ SURJUDP FRPSOLDQFH 3K\VLFDO t 2FFXSDWLRQDO 7KHUDS\ LQ 3HGLDWULFV =\VVHW 7 5HGHEHUJ $ 9DVHOOD ) .XSIHU $ t %LUFKHU f 3KHQ\WRLQ WKHUDS\ IRU HSLOHSWLF FKLOGUHQ (YDOXDWLRQ RI VDOLYDU\ DQG SODVPD FRQFHQWUDWLRQV DQG RI PHWKRGV RI DVVHVVLQJ FRPSOLDQFH 'HYHORSPHQWDO 0HGLFLQH DQG &KLOG 1HXURORJ\

PAGE 91

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nV 'HSDUWPHQW RI &OLQLFDO DQG +HDOWK 3V\FKRORJ\ ZLWK D VSHFLDOL]DWLRQ LQ SHGLDWULF SV\FKRORJ\ LQ 6REKD ZLOO EHJLQ D SRVWGRFWRUDO IHOORZVKLS DW (PRU\ 8QLYHUVLW\ DQG WKH &KLOGUHQfV +HDOWKFDUH RI $WODQWD KRVSLWDOV LQ $WODQWD *HRUJLD LQ -DQXDU\

PAGE 92

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ƒ SWDEOH *DU\ *HII\ HQ $VVRFLDWH 3URIHVVRU RI &OLQLFDO DQG +HDOWK 3V\FKRORJ\ FHUWLI\ WKDW KDYH UHDG WKLV VWXG\ DQG WKDW LQ P\ RSLQLRQ LW FRQIRUPV WR DFFHSWDEOH VWDQGDUGV RI VFKRODUO\ SUHVHQWDWLRQ DQG LV IXOO\ DGHTXDWH LQ VFRSH DQG TXDOLW\ DV D GLVVHUWDWLRQ IRU WKH GHJUHH RI 'RFWRU RI 3KLORVRSK\ f &OD\GHOO +RPH $VVRFLDWH 3URIHVVRU RI 1XUVLQJ

PAGE 93

7KLV GLVVHUWDWLRQ ZDV VXEPLWWHG WR WKH *UDGXDWH )DFXOW\ RI WKH &ROOHJH RI +HDOWK 3URIHVVLRQV DQG WR WKH *UDGXDWH 6FKRRO DQG ZDV DFFHSWHG DV SDUWLDO IXOILOOPHQW RI WKH UHTXLUHPHQWV IRU WKH GHJUHH RI 'RFWRU RI 3KLORVV 'HFHPEHU 'HDQ &ROOHJH RI +HDOWK 3URIHVVLRQV 'HDQ *UDGXDWH 6FKRRO

PAGE 94

81,9(56,7< 2) )/25,'$