University of Southern California Dissertations and Theses

Communication and social support of parents of children treated at Childrens Hospital Los Angeles

(USC Thesis Other)

Communication and social support of parents of children treated at Childrens Hospital Los Angeles

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content COMMUNICATION AND SOCIAL SUPPORT OF PARENTS OF CHILDREN
TREATED AT CHILDRENS HOSPITAL LOS ANGELES

by

Jeffrey A. Hall

A Dissertation Presented to the
FACULTY OF THE GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(COMMUNICATION)

December 2007

Copyright 2007 Jeffrey A. Hall

ii
Dedication
For Amber, Mom and Dad, and the teachers who lead me here.

iii
Acknowledgements
I would like to acknowledge my dissertation committee for their helpful guidance,
invaluable support, and willing participation in my dissertation. To my chair, Prof.
Michael J. Cody, I am deeply thankful for his commitment to my achievement both on
this project and all other publications, manuscripts, and ideas. At Childrens Hospital Los
Angeles, I would like to thank Prof. Ernest Katz for his willingness to commit effort and
time into this project. I would also like to thank the parents who shared their stories with
me to make this project possible. Finally, without the daily support of my wife, Amber, I
could not have finished this or any other project, and for her love I consider myself
blessed.

iv
Table of Contents
Dedication ii
Acknowledgements iii
Table of Contents iv
List of Figures v
Abstract vi
Chapter 1: Literature Review 1
Section One: Clinical research findings 3
Section Two: Theory in family communication and health 10
Section Three: Social support 15
Section Four: Masculinity and gender role conflict 25
Section Five: Culture 32
Section Six: Summary 35

Chapter 2: Methods 37
Section One: Instrument development 37
Section Two: Recruitment procedure and sample 38
Section Three: Measures 40
Table One: All means and standard deviations by sex of respondent 44
Chapter 3: Results 45
Section One: Child’s treatment and parent outcomes 45
Section Two: Communication and perceived social support 46
Section Three: Received social support 48
Section Four: Gender role conflict 52
Section Five: Acculturation 54
Section Six: Testing the Complete Model 55

Chapter 4: Discussion 61
Section One: Communication and perceived social support 61
Section Two: Received social support 62
Section Three: Gender role conflict 69
Section Four: Acculturation 72
Section Five: Limitations, strengths, and directions for future research 73

References 80

v
List of Figures
Figure One: Family Communication Model 12
Figure Two: Unique and Shared Support Sources 49
Figure Three: Actor-Partner Independence Model for Communication and 56
Perceived Social Support

Figure Four: APIM for Complete Model 60

vi
Abstract
This investigation explored the communication and social support networks of
mothers and fathers as they coped with the challenges caused by childhood cancer. Data
was gathered using a survey instrument at an annual Childrens Hospital Los Angeles
survivorship event. Fifty-four parents completed the survey, and 40 of these were mother-
father pairs. Results confirmed that competent communication enables emotional and
instrumental social support within the relationship. To test the theory of optimal
matching, multilevel modeling was used to determine what types of social support were
most valuable. Instrumental and emotional support both uniquely predicted the overall
quality of social support. The use of network methodology collected from mother-father
pairs, allowed the identification of each source of support as either shared or unique. For
social support from shared social support sources, mothers and fathers received the same
quality and types of support, but from their unique sources of support, mothers received
higher quality instrumental support in comparison to fathers.
Two indicators of gender role conflict, emotional expression and career
achievement, were explored in relation to fathers’ outcomes. Results demonstrated that
fathers’ career achievement conflict reduced the amount of instrumental and emotional
support that mothers perceived, but did not affect their own support or anxiety. The other
measure of masculinity, restricted emotional expression, was not related to either
mothers’ or fathers’ outcomes.
Using structural equation modeling, a complete dyadic model of family
communication tested the findings controlling for the non-independence of parents’ data.

vii
This model demonstrated that social support was related to parents’ trait anxiety. The
quality of received social support from fathers’ and mothers’ networks reduced fathers’
anxiety. When fathers perceived greater social support from their spouses, mothers also
reported less anxiety.
The role of ethnicity and acculturation in family function was not related to any of
the variables explored in this study, and differences between Latino and white families
were not significant.
Limitations and strengths of the present investigation are discussed, and future
directions for research and intervention are described.

1
Chapter One: Literature Review
The diagnosis of cancer in a child is a devastating experience for the whole
family. Fortunately, scientific advances in the past 40 years have significantly improved
the prospects for sick children. In comparison to all cancer types, childhood cancer has
seen the greatest increases in survivorship. According to the National Cancer Institute
(2004), treatment of childhood cancer has increased the survival rate from less than 30
percent to 80 percent since the 1960s. Nonetheless, the numbers of families affected by
cancer is substantial. Nationwide, one in 330 children will get cancer before the age of 19
(National Cancer Institute, 2004). As the population of cancer survivors increases, the
challenges facing the child and the family have changed to a greater focus on
survivorship and coping.
For every incident of childhood cancer, parents and other family members suffer
from the stress and trauma of coping with the emotional, financial, and health challenges
inherent to treatment and recovery. Even after the treatment has ended, some parents
show symptoms of post-traumatic stress disorder (Kazak et al., 1997; Lavee & May Dan,
2003). Yet, there is considerable variation in parental well-being. Several reviews of
parental coping argue that not all parents of children with cancer show high levels of
distress (Kupst, 1993; Kupst & Schulman, 1988; Sloper, 2000). Sloper (2000) concluded
that some parents function much better than others and that researchers should invest
time and energy into discovering why this might be the case. This dissertation aims to
respond to Sloper’s challenge with an approach that emphasizes the role of
communication, social support, and gender.

2
There are two overarching goals of this project. First, this dissertation attempts to
link clinical research findings with communication theory. The short-comings of clinical
research will be addressed by employing theory, broadening the investigation into social
support using network methodology, and exploring the affect of gender role ideology.
These areas of research will be discussed in detail in the literature review, and will be
reflected in the methods, results, and discussion. Second, this dissertation addresses
specific research interests requested by representatives from Childrens Hospital Los
Angeles (CHLA). Specifically, CHLA believes that communication research could
improve its ability to “treat the whole family,” a central component of the hospital’s
mission (Childrens Hospital Los Angeles, 2007). In addition, CHLA hopes to learn more
about the role of fathers, to better address a perceived lack of participation. Although this
dissertation will not evaluate an intervention or specific program, it attempts to offer
some guidance about how social support, gender, and communication might be employed
in future research.
To accomplish the aims of this dissertation, the first chapter will review pertinent
literature on family health and communication during a child’s treatment for cancer. The
first chapter is divided into five sections. The first section reviews the clinical research
findings regarding the psychological outcomes of parents of children treated for cancer.
This section will focus on relevant predictors of parents’ health, including social support.
The benefits and limitations of the clinical approach will be addressed prior to moving to
the theoretical underpinnings of the present investigation. In the second section, the
theories germane to a communication approach to parents’ health will be introduced. In

3
this section, a model of health communication will be advanced in order to make explicit
the interconnections between the family and their social network. The third section will
discuss how social support plays a role for parents of children with cancer, and will
outline the various approaches to social support. In this section, the theory of optimal
matching (Cutrona & Russell, 1990) will be introduced, as will the advantages of
network methodology employed in this dissertation. The fourth section will discuss the
importance of the sex of the parent, which is one of the more consistent predictors of
health outcomes. In most health research, sex is treated as a control variable, but studying
within gender differences may help to clarify why fathers might have different outcomes
than mothers. In the fifth section, the importance of exploring cultural and socio-
economic factors will be addressed, including a focus on the Latino population of CHLA.
Section One: Clinical research findings
The illness of a child is a source of great anxiety and stress for parents, especially
one as frightening and harmful as cancer. One study revealed that nearly a third of
parents of children with cancer had symptoms of a major psychological illness (Brown,
Kaslow, Madan-Swain, Doepke, Sexson, & Hill, 1993). This stress affects the parents’
relationship with one another. Nearly half of marital problems reported by parents
appeared after their child’s diagnosis (Kalins, Churchill, & Terry, 1980). Parents’ mental
health plays an important role in their children’s health as well (Brown et al., 1993). It is
important to note that the majority of parents do not evidence symptoms of psychological
disorder (Brown et al., 1993; Kalins et al., 1980). Rather, parents’ responses are diverse.
Some parents are emotionally devastated while others are able to more positively cope

4
with the stress. Some of the predictors of poor family functioning, such as depression,
alcoholism, anxiety, marital discord, and the age of the parents (e.g., Brown et al., 1993;
Dahlquist, Czyzweski, & Jones, 1996; Hamovitch, 1963; Kalins et al., 1980; Morrow,
Carpenter, & Hoagland, 1984), are not specific to families of children with cancer. Two
studies comparing parents of children with cancer with control parents demonstrate that
parents with children treated for cancer reported no differences in levels of depression
(Speechly & Noh, 1992) and anxiety (Kazak et al., 1997; Speechly & Noh, 1992). This
suggests that although depression and anxiety predict poor psychological health, the
cancer status of children does not necessarily cause depression and trait-based anxiety.
Hamovitch’s (1963) pioneering work on family health found that parents who struggle in
marriage or experience employment difficulties fare worse than parents who do not, but
their child’s illness alone is not cause enough for poor outcomes. One of the most
consistent predictors of psychological function and an important moderator for measuring
the effects of coping, social support, and other interventions is the sex of the parent.
The foremost researchers on parental outcomes during children’s illness,
Hoekstra-Weebers and colleagues, consistently demonstrate that there are different
predictors of long-term psychological health for fathers and mothers (Hoekstra-Weebers,
Heuvel, Jaspers, Kamps, & Klip, 1998; Hoekstra-Weebers, Jaspers, & Kamps, 2000;
Hoekstra-Weebers, Jaspers, Kamps, & Klip, 1998; Hoekstra-Weebers, Jaspers, Kamps, &
Klip, 1999; Hoekstra-Weebers, Jaspers, Kamps, & Klip, 2001). Trait anxiety and coping
style are significant predictors of fathers’ general psychological well-being, but do not
predict mother’s health outcomes (Hoekstra-Weebers et al., 1999). For mothers, greater

5
assertiveness was associated with well-being (Hoekstra-Weebers et al., 1999). Sex also
appears to moderate the effects of coping style: “whereas fathers’ use of the coping styles
at diagnosis did not have an affect on their current psychological functioning, it did have
longer-lasting effects on their future psychological functioning. The reverse was found
for mothers” (Hoekstra-Weebers et al., 2000, p. 264). Overall, fathers’ use of coping
styles is less predictive of negative outcomes than social and personal resources, while
mothers’ coping styles are more consistently related to outcomes.
Sloper (2000) similarly found significant differences between fathers and mothers
in predicting malaise. For mothers, self-directed coping reduced malaise, as did perceived
higher efficacy in dealing with stressors related to illness. For fathers, employment
problems and the number of times the child was admitted to the hospital were the primary
predictors of malaise. Two earlier studies (e.g., Cook, 1984; Kalins et al., 1980) also
point to the importance of employment problems in predicting father’s psychological
well-being. Unemployment and job insecurity are risk factors of men’s health generally,
and in the context of children’s illness, fathers who are unable to provide financial
support were particularly affected (Sloper, 2000).
Coping style not only offers different benefits for mothers and fathers, but the
relative benefit of these styles must be sensitive to the temporal distance from the child’s
diagnosis. There are three dimensions to the impact of time: the accumulation of stress,
the family’s ability to endure the complications of treatment, and the exhausting of social
support. Parents report high levels of psychological distress at the time of diagnosis, but
distress varies greatly as the treatment continues (Hoekstra-Weebers et al., 1998;

6
Hoekstra-Weebers et al., 2000; Hoekstra-Weebers et al., 2001). As the stress of the
child’s treatment continues to build up, some parents begin to become emotionally taxed.
Time also interacts with social support in important ways. Social support from the family
and friends of the sick child tends to be most pronounced when the illness first occurs,
but many parents experience a considerable drop off of support around 6 months
(Hoekstra-Weebers et al., 2000; Hoekstra-Weebers et al., 2001), and the negative effects
of a lack of support become particularly pronounced if the child’s illness persists for
more than 4 years (Kupst, Natta, Richardson, & Schulman, 1995; Lavee & Mey Dan,
2003). Although the amount of support lessens as time from diagnosis increases, the need
for support does not (Hoekstra-Weebers et al., 2000; Sloper, 2000). When those
emotional and social resources are depleted, the risks of an increased strain on the
parents’ relationship with one another are greater (Lavee & Mey Dan, 2003).
1.2 Social support
Clinical research on parents of children with cancer has revealed that social
support is an important predictor of parents’ health (for review see Kupst, 1993). There is
a positive relationship between coping and the quality of social support perceived (Kupst
& Shulman, 1988). However, the use of social support to cope tends to differ by sex.
Mothers are more likely than fathers to use social support seeking in order to cope
(Hoeskstra-Weebers et al., 1998), while fathers’ outcomes were not related to social
support variables (Kazak et al., 1997). The differences in behavior between men and
women are somewhat explainable by the differences in social support they receive.

7
Because women experience social support at higher levels of both quality and quantity
compared to men, they also benefit more from social support in times of need.
Fathers lack a means to obtain the desired amount of social support, both in
quality and in quantity (Hoekstra-Weebers et al., 1999; Hoekstra-Weebers et al., 2001;
Sloper, 2000). The quality of social support appears to be particularly problematic for
fathers. Even if fathers have broad social support networks, they are often dissatisfied
with the quality of support received (Hoekstra-Weebers et al., 2000; Hoekstra-Weebers et
al., 2001). In addition, fathers who seek support at the beginning of their child’s illness as
a form of coping became more dissatisfied over time (Hoekstra-Weebers et al., 1999;
Hoekstra-Weebers et al., 2000). For fathers, but not for mothers, the effects of higher
levels of stressors combined with lower levels of support were additive, resulting in
fathers suffering from more emotional distress than mothers (Sloper, 2000). These
findings suggest that, independent of other stressors, fathers tend to receive lesser quality
social support, but stand much more to gain by greater quality social support. A minor
increase in social support received by fathers tended to have a more positive impact than
a similar increase for mothers (Speechly & Noh, 1992).
In addition, social support seeking as a coping style tends to have long-term
implications. Fathers who sought more social support at diagnosis were more distressed
12 months later. However, mothers who initially sought less social support were more
distressed later (Hoeskstra-Weebers et al., 2000). In fact, fathers who used avoidance,
seeking social support, and expression of emotion more frequently at time of the
diagnosis reported more distress a year later (Hoeskstra-Weebers et al., 2000). Two

8
interpretations of these findings are possible. Hoekstra-Weebers and colleagues (2000)
report that when fathers were more distressed, they expressed their emotions more
frequently. However, the degree to which these two factors – social support seeking and
expression of emotions – co-vary is unknown. It could be that fathers only seek social
support when the situation is dire. In effect, these social support seeking men are the most
distressed and are therefore more likely to report adverse outcomes. This hypothesis is
related to the perceived versus received support debate, which will be discussed in greater
detail in the following section. However, given the inadequate quality of social support
that men report, it could also be that social support seeking and expression of emotion is
an unhelpful solution to men’s stress. By seeking social support, men are calling upon
resources that are incapable of providing the support they need.
One final relevant research finding related to social support is that men report a
more positive attitude toward marriage (Dahlquist et al., 1996) and more positive change
in their relationships during their child’s treatment than do women (Lavee & Mey Dan,
2003). As a possible explanation of these findings Dahlquist and colleagues (1996)
speculate that “women have a broader social support system in which they would more
commonly discuss very distressing situations. Men may rely more on the marriage for
social support and therefore feel increased satisfaction from this one support source in
times of stress” (p. 552). This seems to suggest that men’s reliance on their wives for
support leads to a more positive attitude toward marriage, but this behavior does not
improve mothers’ attitude toward marriage or well-being. This suggests that social
support affects functioning within the couple as well as within the individual.

9
1.3 Strengths and limitations of clinical research
The clinical research on parents of children with cancer offers many important
findings for the present investigation. Clinical research benefits from its continuous and
direct contact with parents and children. Researchers can employ a host of health
professionals and support staff who can manage a clinical investigation and provide
assistance for families as needed. Furthermore, the clinical setting is most able to explore
possible interventions and receive research funds to implement and evaluate these
interventions. Many clinical studies require years of recruitment to accrue enough
participating parents, especially if the investigation is longitudinal.
There are also limitations to the clinical approach. Clinical research often
acknowledges that communication is important, but typically communication is
unmeasured (Streisand, Braniecki, Tercyak, & Kazak, 2001) and underemphasized in
health research (Segrin & Flora, 2005). Despite the evidence supporting the importance
of couple’s relationships in times of stress, the medical model does not take account of
personal relationships (Lyons, Langille, & Duck, 2006). New approaches to family health
have begun to incorporate communication into research (e.g., Christensen, 2004), and
CHLA has put theory and communication front and center in their recent research on
mother’s assertiveness training (see Varni et al., 1999). This leads to the second
limitation of most clinical research. Theory is rarely used to explain or predict outcomes.
Communication theory offers tremendous benefits to exploring family and interpersonal
functioning. Incorporating theory into clinical research will help to organize findings and

10
explore new territory. In the following section, a theoretical approach is advanced to
frame the present investigation.
Section Two: Theory in family communication and health
Contemporary research in family communication and health emphasizes an
ecological or systems approach to families, which begins with the assertion that all family
members affect one another (Street, 2003; Segrin & Flora, 2005). Kazak (1989), one of
the predominant researchers on family systems in childhood disease, identified the
systems model as an important heuristic to understand family functioning, and argued
persuasively that the best way to serve children with cancer is to first understand the
ecology of the family. When a child is diagnosed with cancer, the effects of the disease
are not contained within the child, but reach parents and other family members (Kupst,
1993). Furthermore, the study of the family ecology should consider employment
challenges, social support, and the extended family because all are affected by the
chronic condition. Communication between the parents and within the family lies at the
core of the family systems approach.
Medical researchers who use systems theory are beginning to acknowledge the
importance of communication in family health. Open communication between parents
has been linked to the adjustment of children and their parents (Hoekstra-Weebers et al.,
1998; Kupst, 1993; Kupst & Schulman, 1988), and is a predictor of long-term health
outcomes for parents (Ostoff, Ross, & Steinglass, 2000). On the other hand, the absence
of good communication is a common and significant concern for families in the context
of illness (Lyons et al., 2006). Efforts are being made to measure and evaluate

11
communication in a clinical setting. Streisand and colleagues (2001) developed the
Pediatric Inventory for Parents at Childrens Hospital Philadelphia specifically for
measuring long-term treatment coping, and it includes items on communication and
family role function. Despite these and other examples of the importance of
communication in this context, communication’s value is often more asserted than
measured. Fortunately, communication literature has a long history of measuring and
exploring how communication affects the family. To help in the discussion of how
communication is conceptualized in this dissertation, a model of the family ecology is
presented in Figure One.

12
Figure One: Family Communication Model

13
2.2 Family communication
When studying the parents of children who are treated for cancer, one useful
conceptualization of the family is to begin with the dyad. When parents can count on
stable support from their spouses, they are more able to provide for their children (Brown
et al., 1993), show better relationships with health practitioners (Hamovitch, 1963), and
have better mental health outcomes (Kupst, 1993). Although no research on family
functioning during cancer treatment has used communicator competence or satisfaction to
measure the effect that communication has on parents, using communicator competence
as a predictor variable is not without precedent in the literature on family communication
(Farrell & Barnes, 1993; Fitzpatrick, Feng, & Crawford, 2003; Koesten, 2004). One study
that directly tested communication outcomes using a family systems perspective found
that the better the communication between parents, the better each family member
functions (Farrell & Barnes, 1993). There are three ways that exploring and measuring
communication may advance the research on parents of children with cancer.
According to research in communication, the interaction patterns between parents
are very important. Segrin and Flora (2005) suggest that clear, direct, and open
communication between parents improve entire family functioning. Poor, indirect,
defensive, and/or minimal communication is destructive and ultimately contributes to the
accumulation of family stress (Lyons et al., 2006). Furthermore, parents who are
competent communicators are able to foresee potential obstacles and can adjust to
changing circumstances (Koesten, 2004). That is, competent communicators possess an
anticipatory mindset. Clinical research suggest that positive health behaviors that exist

14
before the cancer diagnosis can protect families from more negative consequences
(Kazak et al., 1997; Kupst, 1993; Lyons et al., 2006). Similarly, communication between
the parents is expected to be able to buffer families from many of the stresses of the
diagnosis of cancer in children.
Second, communication is important to marshal social support both between
parents and within the extended support network (Kunkel & Burleson, 1998). That is,
being a competent communicator enables the listener to listen effectively, create person-
centered responses, and help the speaker identify helpful responses to stressful events
(Burleson & MacGeorge, 2002). For parents enduring a particularly frightening and
challenging event, communication is even more important. Therefore, this model
suggests that communication enables social support, which reduces negative outcomes
such as anxiety. Communication also affects social support generation outside of the
family. Prior research on communication and social support have suggested that
communication competence is central to marshalling useful and targeted social support
(e.g., Burleson & MacGeorge, 2002; Lyons et al., 2006; Reinhardt, Boerner & Horowitz,
2006; Sarason, Sarason, & Pierce, 1990). However, these studies, like the literature on
family ecology, typically do not measure communication explicitly. Measuring
communication may offer new insight into how families successfully use social support.
Furthermore, by measuring communication and social support for both parents, the upper
part of the model will be better understood (see Figure One).
Finally, ecological models and family systems researchers point out that
communication within the family matters greatly for the transmission of health

15
information and the negotiation of healthy behaviors (Christensen, 2004). Theoretically,
the influence of any component of the systems model can only be measured in terms of
communication. That is, the relationship between resources and stressors is most
certainly mediated by communication quality. Research on families in non-clinical
settings suggests that the degree to which communication is effective within the family is
a valuable predictor of the behaviors of the extended family (Fitzpatrick et al., 2003;
Koesten, 2004). This research does not focus on interaction with hospital staff or specific
outcomes and predictor variables in children, but prior research suggests that parent’s
communication affects both (Brown et al., 1993; Hamovitch, 1963). The following three
sections will directly explore the remaining portions of the model in greater detail. To
begin, the next section will explore both how social support is affected by parent
communication, and explain how a social network analysis approach might help answer
questions about who provides support and who receives it.
Section Three: Social support
Social support is one of the most heavily investigated areas of social science
(Thoits, 1995). It has been explored in multiple traditions, using a host of theoretical
frameworks and methodologies (Cohen, Gottlieb, & Underwood, 2000; Thoits, 1995). It
is not the mission of this dissertation to explore all of the ways that social support has
been investigated, but instead to explain how communication is compatible with social
support research, how social support is related to parents’ health, and how network
methodology can be utilized to investigate issues specific to parents well-being in the
context of children’s cancer. Therefore, the following sections will explore a limited

16
range of possible social support issues. To begin, there will be a brief discussion of
communication and social support, with an emphasis on sex differences. Then, the
differences between perceived and received social support will be discussed, to explain
why both are utilized in this investigation. Third, the theory of optimal matching will be
reviewed. Fourth, a discussion of the benefits of network methodology will be advanced,
which will be followed by specific hypotheses and research questions.
3.1 Social support as an interpersonal process
The study of social support has multiple traditions. The interpersonal process
tradition is one of the three critical traditions of social support research outlined by
Cohen, Gottlieb, and Underwood (2000). The interpersonal tradition focuses on the
expression of and receipt of social support. How people gain social support, express
social support, and the investigation of what sort of social support is considered valuable
are all topics that communication scholars have explored (Albrecht & Goldsmith, 2003;
Burleson & Kunkel, 2006). From this perspective, social support is ultimately a
communication phenomenon. Communication clarifies the kind of support needed,
allows for the expression of fear, uncertainty, or frustration, allows others to perceive
emotional needs, and serves as a primary mechanism of several types of social support,
such as relaying health information (Albrecht & Goldsmith, 2003; Cohen et al., 1989). In
fact, communication competence may be one of the most critical factors in the ability to
obtain the desired support, and to negotiate appropriate boundaries and expectations
within an individuals’ support network (Burleson & MacGeorge, 2002; Lyons et al.,
2006; Reinhardt et al., 2006; Sarason et al., 1990). While communication is central to

17
social support, there are important differences in the use of and ability to give support.
One important difference is sex.
Sex differences in social support have been documented in multiple traditions and
domains (Thoits, 1995). Early research on social support in health contexts suggest that
women are more likely than men to seek social support (Defares, Brandjes, Nass, & van
der Ploeg, 1985), a finding supported in clinical research (Hoeskstra-Weebers et al.,
1998; Kazak et al., 1997). More recent investigations into enacted support (Burleson &
Kunkel, 2006; Kunkel & Burleson, 1998) confirm these differences. In their review of
communication and social support, Burleson and Kunkel (2006) suggest that women are
more likely than men to provide and seek emotional support, and women view supportive
and expressive skills as slightly more important than do men. Yet, both men and women
value emotional support over instrumental support, and men and women identify the
same qualities of ‘good’ social support. Burleson and Kunkel (2006) conclude that both
men and women want quality emotional support, women just give it better. This offers a
possible explanation for men’s lack of quality and quantity of social support. If a man has
a male-dominated social support system, he may be able to draw only limited emotional
support from this group. Clinical research on parents have not demonstrated that female
support givers are more desirable, but past research has confirmed that women have
better social support networks (Hoeskstra-Weebers et al., 2001).
3.2 Perceived versus received support
The difference between received and perceived social support is both a theoretical
and methodological question. Received social support is generally defined as tangible,

18
measurable, and enumerated social supports of various types, and typically includes
instrumental, material, or informational support (Burleson & MacGeorge, 2002).
Received social support is often measured using network methodology, diary or written
methods, or using a checklist (Wills & Shinar, 2000). Alternatively, perceived social
support is defined as the perceived availability of support if it were needed (Burleson &
MacGeorge, 2002). Perceived social support means support is believed to be available,
even it is not used (Sarason et al., 1990). Some researchers have speculated that
perceived social support is actually more similar to self-esteem or other personality
characteristics (Lakey & Cohen, 2000). Nonetheless, perceived support is more
consistently related to positive health outcomes than received social support (Sarason et
al., 1990; Watts & Crimmins, in press). While there is a strong relationship between
social support and health (Schwarzer & Leppin, 1989), researchers have recognized that
the most meaningful and useful measure of social support depends upon what the
researcher is most interested in knowing (Wills & Shinar, 2000). In the present
investigation, both perceived and received social support will be measured, but from
different sources.
Perceived social support will be measured in relation to communication
competence within the couple. Received social support has shown relevance in inquiries
into couples’ relationships (e.g., Reis, Clark, & Holmes, 2004). However, recent work on
relational maintenance suggests that documenting or enumerating actual received support
in a highly interdependent relationship, such as a marriage, is inappropriate and may even
uncover contradictory or inappropriate results (Canary & Stafford, 2007). In the context

19
of parents with children receiving treatment for cancer, the perception of being supported
by one’s spouse is one of the most important predictors of positive psychological
outcomes (Kupst, 1993). Support gained within the dyad cannot be replaced completely
by support from the social network (Coyne, Ellard, & Smith, 1990). Within the context of
the relationship, the level of perceived support is likely to be predicted by the perceived
communication competence of the spouse. Received support, alternatively, will be
explored in relation to the broader social support network.
While perceived social support is often related to positive coping with stress,
received social support has shown less consistent results. Some research has suggested
that in the context of parents with children with cancer, level of family support is related
to positive coping (Kupst, 1993). Other research has demonstrated that the reception of
large amounts social support is related to extreme health problems, having a
temperamental or needy disposition, and the experience of intense stress (Burrell, 2002;
Coyne et al., 1990; Reinhardt et al., 2006). Nonetheless, researchers maintain that
received social support is an important construct to measure, especially in the context of
exploring enacted support, the types of support which are valuable, and the sources of
social support (Burleson & MacGeorge, 2002; Sarason et al., 1990; Wills & Shinar,
2000). Furthermore, in order to test the theory of optimal matching received social
support must be measured.
3.3 Theory of optimal matching
The theory of optimal matching suggests that specific forms of social support are
most beneficial when they match the needs of stressful life events (Cutrona & Russell,

20
1990). That is, the degree to which received social support matches the needed social
support ought to predict the efficacy or value of social support. This perspective dovetails
neatly with the concept of enacted support in communication research (Albrecht &
Goldsmith, 2003; Burleson & MacGeorge, 2002). Other research on social support in
health contexts supports the predictions of the theory of optimal matching (Lakey &
Cohen, 2000). Furthermore, both Thoits (1995) and Burleson and MacGeorge (2002)
believe that functional matching of social support is a useful theory that is rarely
measured in research and needs further study. But the question remains, how does a
researcher know which type of social support is needed? Cutrona and Russell (1990)
provide two dimensions of social support, controllability and domain, and both of these
dimensions guide predictions about what specific types of social support will be valuable.
The needs of the individual are dependent upon the stressor encountered. Some
events are uncontrollable, especially where there is a threat of harm or loss. This type of
event will require social support components that will foster emotion-focused coping
(Cutrona & Russell, 1990). Controllable events, however, will require social support that
fosters problem-solving or provides instrumental aid. Problem-centered coping tends to
work best when one has mastery and control over the arena in which the problem arises,
but is relatively ineffective when the problem is uncontrollable and fearful (Thoits, 1995).
The diagnosis of cancer in a child could be seen as a highly uncontrollable event or a
more controllable one by parents.
The diagnosis of cancer in a child is a very frightening event for parents, and may
be best understood as an uncontrollable stress (Kazak et al., 1997). Given a great deal of

21
uncertainty and fear, emotion centered support may yield the greatest benefits because it
expresses affection, concern, and care in the face of uncertainty. Other research has
revealed a unique benefit of emotional centered coping from family members during
childhood cancer treatment (Kupst, 1993). In fact, emotional support from family was
more effective than emotional support from spouse in other situations of uncertainty
(Cutrona & Russell, 1990).
On the other hand, the nature of the treatment has changed dramatically in the
past 50 years, and has increasingly become a process of uncertainty management and
organization of the treatment regimen. As the population of survivors increase, we are
witnessing a new phase of coping encompassing a distinct set of psychological issues
(Ostoff et al., 2000). That is, childhood cancer is becoming more controllable as the
likelihood of survivorship increases. This suggests that problem-centered and
instrumental support may be more valuable than emotion focused support for parents of
cancer survivors. For example, in a review of five studies about support for mothers with
children, instrumental support and tangible aid was most predictive of positive outcomes
in comparison to other types of support (Cutrona & Russell, 1990). Helping to transport
and care for the child with cancer and provide material or logistical assistance for the
family or siblings may be particularly useful during cancer treatment. Which of these two
types of support are more predictive of the overall quality of the support source will be
investigated directly in this study.
Finally, Cutrona and Russell’s (1990) second dimension, domain, refers to the
event or circumstance in which the social support is needed. In the case of health care,

22
researchers have demonstrated that informational support is often very helpful (Albrecht
& Goldsmith, 2003). That is, domain specific information about health care may also
predict the overall quality of support. On the other hand, health care information may be
amply provided by the hospital, and therefore if it is not provided by a medical
professional, it may not be perceived as valuable. To uncover specific aspects of social
support sources it is valuable to use social network analysis. The next section will discuss
the particular benefits of using network methodology.
3.4 Social network analysis
The methods employed by any researcher should match the needs of the inquiry.
If a researcher is interested in knowing from whom the support comes, and what sorts of
support are most effective, measuring received social support through network measures
is best (Wills & Shinar, 2000). In fact, if researchers are interested in exploring network
structure and the quality of support sources, it is inappropriate to measure perceived
social support from the network (Lakey & Cohen, 2000). To capture the efficacy of
different types of social support, which is central to testing the optimal matching
hypotheses, researchers must measure the different functions of received support (Wills
& Shinar, 2000). The social network approach offers unique advantages in identifying the
source of support, the type of support, and the quality of support from each source. It
offers the researcher the ability to identify better what types of social support are more
effective and from which source (Wellman & Hiscott, 1985). Network methods have a
long history in social support literature and employ a unique set of terms and concepts
(Wellman & Hiscott, 1985).

23
Network analysis quantifies relationships based upon in-degree and out-degree
(Wasserman & Faust, 1994). In-degree ties represent others nominating or contacting the
central person, or the ego. An out-degree tie represents the ego nominating others, or
alters. Through these basic components, an entire social network can be formulated.
While it is unrealistic to perform a complete social network analysis of every alter in the
parents’ network, a limited network analysis offers tremendous promise. Network
analysis can document, illustrate, and identify how the social networks of mothers and
fathers are different or similar and how they might interact with one another. Brissette,
Cohen, and Seeman (2000) recommend using an ego-network to measure social
integration and support. This method asks a parent to identify up to five of the most
important sources of social support during their child’s treatment. Gathering data on
social support sources past five alters offers diminishing returns (Sarason et al., 1990).
This procedure is useful for studying where support comes from, especially which
specific family members play a role in providing help (Brissette et al., 2000). Studies that
have used social network analysis in the context of children’s illness have already
uncovered some interesting results.
Research on social support networks has documented important differences
between men and women in their networks. For example, the number of friends is
unrelated to satisfaction with social support for either men or women, but the quality of
friendships does predict psychological health outcomes (Hoekstra-Weebers et al., 2001).
Investigating the sources of social support, Wellman (1992) found that immediate kin are
most active network members. Morrow and colleagues (1984) suggest that the most

24
important social support sources are spouse, relatives, friends, and primary doctor, in that
order. The present study is also interested in how these networks differentially impact
mothers and fathers. It is possible that men would receive less emotional support from
family networks as a consequence of gender role expectations: “men don’t need
emotional support, but women do.” This methodology allows for the possibility of
exploring another research question: Do mothers receive more social support from family
members than fathers, and if kin networks are shared, are mothers the primary recipients
of emotional and men the primary recipients of instrumental support? In order words, is
the support received gendered?
3.5 Social support research questions
In this section, all of the research questions and hypotheses generated from the
communication and social support sections will be presented. Following the model in
Figure One, this section begins with hypotheses which describe the relationship between
communication and perceived support from spouse, and then offers hypotheses related to
perceived social support.
H
1
) Greater perceived communication competence of the spouse will predict perceived
emotional and instrumental support from the spouse.
No sex differences are anticipated in the relationship between communicator competence
and social support.
H
2
) Fathers will perceive more social support from their spouses than will mothers.
H
3
) Greater perceived communication competence will predict quality of received social
support.

25
H
4
) Perceived social support from spouse will be negatively related to received social
support from the network.
The relationships between received and perceived social support are more
complicated. Although health researchers have begun to recognize the important
difference between social support quantity and quality, a focus on enacted or received
support clarifies what sorts of social support predict the positive outcomes. To explain the
relationships expected, the following hypotheses and research questions are offered:
H
5
) Mothers will receive more social support than fathers.
H
6
) Fathers will report less quality social support than mothers.
H
7
) Social support will be of higher quality when it is received without asking.
H
8
) Trait anxiety will be positively related to received social support.
The following research questions explore the theory of optimal matching:
H
9
) Amount of instrumental support will predict overall quality of social support.
H
10
) Amount of emotional support will predict overall quality of social support.
H
11
) Amount of health information support will predict overall quality of social support.
RQ
2
) Are there gender differences in the reception of social support?
RQ
3
) Which sources of social support are most valuable?
Section Four: Masculinity and gender role conflict
One of the primary interests of representatives at Childrens Hospital was the
fathers’ role in the treatment and survivorship process. Fathers’ roles have been primarily
stereotyped as inexpressive, unavailable, and passive from the point of view of most
health practitioners (Kazak et al., 1997; McNeil, 2007). In accordance with its own

26
mission, CHLA recognizes the importance of understanding fathers’ perspectives. In fact,
research has begun to make concerted efforts to recruit more fathers in clinical studies,
and increase fathers’ involvement in the treatment processes (Hoeskstra-Weebers et al.,
2000). Despite these efforts to explore fathers’ roles, most pediatric oncology literature
continues to treat sex as a categorical variable. This perspective limits understanding
because it ignores within gender differences. For example, research has demonstrated that
parental coping is not strictly predicted by biological sex – it is not the case that all men
are unable to cope and check out emotionally and all women cope well and are
emotionally available (Greenberg & Meadows, 1991). It is important to recognize that
sex is not composed of two static categories, but are created from set of socially
constructed identities and ideologies that are made real through behavior (Connell, 2002;
Courtenay, 2000c). Within gender differences are likely to be important to improving the
understanding of men’s role in the healing process. Rather than exploring sex differences,
this section explores the way that gender roles affect fathers.
There has been a growing interest in the relationship between gender and health. It
is becoming clearer that health behaviors are embedded in social systems that can foster
or inhibit healthy behaviors (Courtenay, 2000b). Behaviors that are related to poor health
are often part of a larger construction of gender within a society, especially behaviors
about personal health maintenance and family health care (Courtenay, 2000b). How men
and women understand their own gender can play a crucial role during their child’s
treatment for cancer. For example, a father’s gender identity as “stoic supporter” can
interfere with his own psychological health, and may deter him from asking for help from

27
his social support network. Qualitative investigations into the role of fathers during times
of chronic illness in the family reveal that fathers’ responses to illness are informed by
what they believe to be their role in the family and how they feel that men should behave
(Matta & Knudson-Martin, 2006; McNeil, 2007; Seiffge-Krenke, 2002). For example,
men’s ability or inability to nurture and care for their children is related to their
internalized constructions of masculinity (Matta & Knudson-Martin, 2006). This means
that fatherhood, as it is constructed by society, affects how men respond to stress, the
problems and challenges within the family, and their role within. It is also important to
emphasize that gendered behaviors are not necessarily unhealthy – assertiveness and self-
directed coping, behaviors more commonly found in men – are beneficial in coping with
the stress of children’s treatment (Hoekstra-Weebers et al., 1998; Sloper, 2000), and are
being used in interventions at CHLA for mothers (Varni et al., 1999). Rather than explore
the negative aspects of masculinity, this dissertation intends to offer a more complete
analysis of how masculinity impacts family health during children’s treatment. There are
three components to this investigation into masculinity: restricted emotionality,
employment and provider stress, and social support. Each will be explored to attempt to
link this approach with past clinical findings.
4.1 Restricted emotional expression
Interest in the influence of masculine ideology on men’s psychological health and
behavior has developed dramatically in the past 20 years. One of the first attempts to
identify and measure the negative consequences resulting from men’s attempts to attain
the societal standards of masculinity is the gender role conflict scale (O’Neil, Good, &

28
Holmes, 1995). Gender role conflict is a measure of how the unrealistic standards of
masculinity are internalized (O’Neil et al., 1995; Thomspon, Pleck, & Ferrera, 1992).
When men develop a rigid, sexist, or restrictive gender role ideology, a conflict may arise
internally when a man is unable to meet those restrictive standards himself (Good &
Wood, 1995; O’Neil et al., 1995). Messages that men receive about being male
communicate an ideal male type that is often inherently problematic for men (Breiding,
2004). This internal conflict manifests itself in four domains: power and competition,
homophobia, restrictive emotionality, and career achievement (Walker, Tokar, Fischer,
2000). The final two components of these scales are particularly relevant to the present
investigation.
One characteristic of the social construction of masculinity that has been
documented in various measures of masculinity is restricted emotionality (Sharpe,
Heppner, Dixon, 1995). Men who are high in restricted emotionality experience more
distress and feel more open to attack when trying to express affection (Saurer & Eisler,
1990), and are less likely to express emotion and admit to emotional or psychological
pain because it is inconsistent with the gender role they have internalized (Good &
Wood, 1995). Restricted emotionality is related to a host of poor psychological outcomes.
It has been shown to be negatively related to well-being and is associated with greater
depression in men (Fragoso & Kashubeck, 2000; Sharpe et al., 1995). In the case of
children with cancer, it is easy to see how the demands of cancer treatment may affect
traditional men more severely than men who have less rigid gender roles. Qualitative
research suggests that men may further internalize emotional inexpressivity when their

29
child is ill (Cook, 1984). In these families, fathers are expected to be stoic and strong for
the rest of the family, which can be very taxing on fathers’ psychological state.
Researchers in clinical work on children with cancer have stipulated that men’s lack of
emotional expressivity may actually explain the differences in men’s functioning in
comparison to women (Hoeskstra-Weebers et al., 1998). This dissertation may be the first
to link these disparate lines of research.
4.2 Career achievement conflict
The second way that gender role conflict may explain familial outcomes is
through employment expectations. When a child becomes severely ill, a mother usually
focuses most of her energies on the ill child, and sometimes she will quit her job(s) in
order to fulfill this responsibility. Kalins and colleagues (1980) report that half of the
mothers quit their jobs to spend more time with the sick child. Many fathers continue
their responsibilities as economic providers and are often responsible for coordinating
financial concerns (Cook, 1984). As previously noted, if a father experiences
employment problems under these stressful conditions, it can adversely affect his
psychological health (Cook, 1984; Kalins et al., 1980; Sloper, 2000). However, research
into men’s gender role conflict suggests that men do not need to experience problems
with employment to have negative outcomes associated with their role of supporter. The
strain of living up to strict standards of masculinity may negatively affect fathers who
have not had negative employment experiences.
Along with restricted emotionality, career achievement is an important predictor
of negative outcomes for men. It occurs when a man believes that his worth and the

30
worth of other men is measured by the degree to which they can provide for their family.
Men are taught to emphasize occupational success in their role of father (Conger, Elder,
Simons, & Ge, 1993). One qualitative investigation into fathers of children with chronic
health conditions found that men internalize this belief to their detriment and that of their
families (McNeil, 2007). In fact, career achievement expectations run directly counter to
what the family needs from fathers.
When a child is seriously ill, fathers are also expected to increase their
participation in the care of the siblings of the ill child (Cook, 1984). Men who score high
on career success conflict are much more likely to leave childcare and housework to
women (Mintz & Mahalik, 1996). This means that fathers who have internalized career
achievement conflict are unlikely to embrace their new familial roles, and may not be
able to increase their participation at home. If so, men who are motivated by this aspect
of masculinity remain committed to an ideology that is counterproductive for the well-
being of the family. During their children’s chronic health problems, men actively
acknowledge that the financial supporter role often is in contradiction with the role
needed, but they feel pressured internally to be a provider before all other roles (McNeil,
2007). Fathers, drawn away by real or ideological financial responsibilities, may also
seem absent from the hospital. It may not be biological sex that is responsible for fathers’
detachment, but internalized gender roles that exacerbate the distance between fathers
and their children.

31
4.3 Masculinity and social support
Men who strongly internalize male gender role ideology are less likely to seek
help (Addis & Mahalik, 2003; Berger, Levant, McMillan, Kelleher, & Sellers, 2005;
Good & Wood, 1995; Robertson & Fitzgerald, 1992). This finding is of theoretical
importance for the investigation of the role of received social support, and of practical
concern for CHLA. Although fathers’ lack of interest in socio-emotional support
programs available at Childrens Hospital can be attributed in part to the economic and
instrumental duties reported in Cook (1984), it may also be attributable to gender role
conflict. First, men are more likely to seek help from their female intimates, especially
their wives (Burleson & Kunkel, 2006; Courtenay, 2000a; Dahlquist et al., 1996). As a
consequence of taking care of a sick child, wives may not be as available as usual
(Hoekstra-Weebers et al., 2001). The attenuation of support from their spouse is
exacerbated for men who endorse traditional masculine roles. Research suggests that
endorsing traditional masculine norms is negatively related to both perceived and
received social support (Thomspon et al., 1992), which means that the internalization of
gender roles can result in both an inability to attain the desired support and a perception
that it is unavailable. Men with less masculine gender role conflict are more likely to be
satisfied with the support provided by their networks (Saurer & Eisler, 1990). By
exploring how men internalize their gender role, the present study may be able to shed
new light on why and how gender differences revealed in past research exist.

32
4.4 Masculinity and social support hypotheses
The following hypotheses are offered to clarify the relationships expected
between different gender role stresses:
H
12
) Restricted emotional expression in fathers will predict trait anxiety.
H
13
) Career achievement conflict in fathers will predict trait anxiety.
H
14
) Fathers with greater restricted emotionality will have wives who report less
perceived social support.
H
15
) Fathers with greater career achievement conflict will report less perceived social
support
H
16
) Fathers with greater career achievement conflict will report less received social
support.
Section Five: Culture
One final consideration that should be included in any investigation into parental
health is the role of culture and ethnicity. The only study that considered demographic
characteristics of parents found no relationship between demographics and psychological
outcomes (e.g., Hoesktra-Weebers et al., 2001). However, this study took place in the
Netherlands, which has different ethnic and racial demographic characteristics than the
U.S. There are strong reasons to consider culture and ethnicity in health research. In
terms of theory, many models of family and health systems regard culture and ethnicity
as relevant to family function (Christensen, 2004). Furthermore, one serious concern for
the present study is the disproportionate health burden on members of lower
socioeconomic classes (Kupst, 1993; Vega & Amaro, 1994). Latino and Black Americans

33
are much more likely to suffer from serious illness than white Americans, and much less
likely to have health insurance (Kupst, 1993; Fiscella, Franks, Doescher, & Saver, 2002;
Vega & Amaro, 1994). This means that in universal health care systems, like the
Netherlands, the effects of demographics on health outcomes could be less significant
than in America. In Los Angeles, issues specific to the Latino population should be
considered.
The Latino population, especially very recent immigrants, tends to rely on a
family centered or collectivist mentality, where family is more important than the
individual (Blackhall, Murphy, Frank, Michel, & Azen, 1995). Differences between
ethnic groups in medical decision-making are also cultural, rather than socioeconomic
(Blackhall et al., 1995). This means that upper, middle, and lower class Latino families
are all likely to see medical decision making as a family decision not an individual one.
The strong familial relations and integrated process of decision-making may have a
positive effect on family health during times of stress. In a study conducted in Los
Angeles, Hamovitch (1963) found a strong relationship between link between being
Latino and positive health outcomes. Mexican-American families were rated as having
better coping mechanisms than white families, but also were more likely to experience
family problems than white families.
In contrast with direct comparisons between Latino populations and white
Americans, another approach to studying the effects of ethnicity is acculturation (Marin,
Marin, Otero-Sabogal, & Perez-Stable, 1987). This measure examines the degree to
which a Latino family identifies with the language and social group of his/her own

34
ethnicity, and that of the country in which he/she resides. In the case of Los Angeles
Latinos, acculturation explores the degree to which a Latino family speaks Spanish
versus English at home and with family, and whether social activities are with Americans
or Latinos. This measure has been used in prior studies on Latino populations in Los
Angeles, and offers researchers an opportunity to explore issues of ethnicity in a health
context (Blackhall et al., 1995; Lara, Kahramanian, Morales, & Bautista, 2005; Murphy,
Palmer, Azen, Frank, Michel, & Blackhall, 1996).
Finally, masculine gender ideologies within the Latino population should be given
unique consideration. Latino men often are described as adhering to a “machismo”
temperament (Murphy, 1994). There are many domains of machismo, with different ways
of expressing attitudes toward men and women, the role of both sexes in society, and the
relationship between them (Fragoso & Kashubeck, 2000; Torres, Solberg, & Carlstrom,
2002). This concept of masculinity might adversely impact the likelihood of help seeking
behavior or admitting to psychological suffering, but also might help to cope with the
situation in other ways (Courtenay, 2000b). This dissertation will explore the relationship
between acculturation and gender role stresses in Latino fathers to determine if Latino
fathers have unique relationships in this context.
There is a lack of research on the influence of culture and acculturation on
families of children who are treated for cancer. Since Latinos view communication
competence in ways that are commensurate with how it is traditionally measured
(Bradford, Meyers, & Kane, 1999), the role of communication between parents on social
support is not expected to be influenced by acculturation or ethnicity. In addition, there

35
are specific cultural issues in regard to social support (Wortman & Lehman, 1985).
However, there is not enough research in children’s cancer treatment to offer specific
hypotheses. To explore how acculturation and ethnicity affect parents, three research
questions are posed for study:
RQ
5
) What role does acculturation have on negative outcomes for parents?
RQ
6
) Are there any meaningful differences in Latino families and other race/ethnic
groups?
RQ
7
) Is the role of social support in family function affected by ethnicity or
acculturation?
Section Six: Summary
This dissertation intends to accomplish two goals. First, by integrating
communication theory, network methodology, and psychological issues particular to
fathers, this project intends to shed new light on the family process during children’s
treatment. Second, through a partnership with CHLA, this dissertation hopes to offer
exploratory and descriptive research that will help CHLA better serve parents. Using
systems and ecological theory (Kazak, 1989; Segrin & Flora, 2005; Street, 2003), this
research hopes to explore how parental communication affects the function of the entire
family, including their broader social support network. This research maintains that for
families, communication is the vehicle by which positive outcomes are enabled (Segrin &
Flora, 2005). Utilizing network methods, this research will test hypotheses based upon
the theory of optimal matching (Cutrona & Russell, 1990). By measuring two indicators
of gender role stress, emotional expression and career achievement, this research will

36
explore the psychological factors within fathers that may help to explain why gender
differences exist. Finally, acknowledging the importance of culture in systems theory and
gender role behaviors, this research takes into account ethnicity and acculturation to
explore the roles of each in family function. To accomplish these goals, this review of
past literature has inspired 16 hypotheses and 7 research questions. The following section
details the methods used to explore these issues, and the results sections provides answers
to the hypotheses and research questions.

37
Chapter Two: Methods
To test the hypotheses and research questions advanced in the literature review, a
research study was developed in coordination with CHLA. The instrument was translated
into Spanish by a translator experienced in social scientific data procedures. Institutional
Review Board (IRB) approval was obtained from the IRB of CHLA. Data was gathered
during an annual survivorship festival in Hollywood, CA on June 11, 2006. Parents of
children who had undergone or were undergoing cancer treatment completed a survey
instrument, and all participants were given a choice of completing the interviews in
Spanish or English. In this chapter, the exact research methods and instrumentation used
in this project is described in detail.
Section One: Instrument development
To develop the instrument, the author attended four parents’ support group
meetings held at CHLA. At these meetings, the author interviewed researchers and social
workers at CHLA, and discussed with staff what sorts of issues were relevant to the study
audience. Notes were taken during support group meetings about the social support
process and the ways in which couples coped with their children’s treatment and
recovery. CHLA staff was involved in the project development and IRB approval. When
a final instrument was developed, a translator experienced in social scientific data
procedures translated the instrument into Spanish. Finally, a native Spanish speaker
checked the instrument for consistency and clarity of the translation and found it
acceptable.

38
Section Two: Recruitment procedure and sample
The present study utilized a cross-sectional survey administered to parents of children
who have been treated at CHLA. Participants were recruited and administered the survey
at the annual Celebrate Life survivorship fair. Participants were approached by the
dissertation author or by graduate student assistants. Other participants approached the
stationary booth set up at the event. At this booth, parents were encouraged to share their
story with the author by completing a survey instrument. If parents were interested in
participating, parents completed the consent form, which was signed by the study
coordinator. The instrument took about 20 minutes to complete. Participants either
completed the instrument on their own or were interviewed by the author or graduate
assistants. A total of 53 parents completed the survey at the study cite.
Exclusion criteria. In order to be included in the study, participating parents must
have been either married or living together at the time of their child’s treatment.
However, due to concerns about strict exclusion criteria reducing the overall sample size,
parents who were interested in participating while their spouse was either not present or
not available at the event were allowed to complete the instrument.
Follow up. Parents who were unable to answer all of the instrument questions or
unwilling to answer any survey questions at the time of the event were asked if they
would be willing to be contacted again by telephone to complete the instrument.
Participants who refused to be re-contacted were thanked and excused. Contact
information was gathered from participants who were willing to be re-contacted. The
author re-contacted parents within two weeks of the Celebrate Life event to complete the

39
survey. One parent completed the survey over the telephone within a week of the event,
bringing the total sample size to 54.
Incentives for participation. Participants who completed the survey were entered
into a raffle for three prizes: $400 cash, and one of two footballs signed by Pete Carroll,
the University of Southern California football coach. The second prizes were chosen due
to local popularity of the USC football team among men in the Los Angeles area. The
winners of prizes were announced at the end of the Celebrate Life event. All winners
were present and received their raffle prizes.
Identifying information. No identifying information was gathered from participants
who completed the survey on site. Information for the raffle was obtained separately and
was not linked to completed surveys. Raffle entries were gathered by asking participants
to put a name and phone number on the back of a raffle ticket. Once the raffle winners
were chosen, all tickets were destroyed. Contact information was gathered from
participants who agreed to be contacted to complete the survey after the Celebrate Life
event. Once the data was collected via the telephone, contact information was destroyed
and was not linked to the completed survey.
Participant demographics. Twenty mother and father pairs were collected, and
fourteen unmatched participants who met study requirements completed surveys (2 male,
12 female). Fathers were 41 years old on average, and 95% were employed full time.
Nineteen percent of fathers had some high school education, 14% had a high school
diploma or equivalent, 33% had some college, and 23% had a 4-year college degree or
more. Fathers were 55% Latino, 31% white, 9% Black, and 5% other. Thirty-two percent

40
were foreign born. Mothers were 39 years old on average, and 50% were not employed,
15% were part-time employed, and 35% full time employed. Thirty-six percent of
mothers had some high school education, 20% had a high school diploma or equivalent,
20% had some college, and 23% had a 4-year college degree or more. Mothers were 65%
Latina, 22% white, 6% Black, and 6% other, and 53% were foreign born.
Section Three: Measures
Communication satisfaction: Seven communication satisfaction questions were
adapted from Wiemann’s (1977) communicator competence scale. This scale measures
the ability of individuals to achieve interpersonal goals within the constraints of a given
situation, and has shown good reliability and validity (Rubin, Palmgreen, & Sypher,
1994). Modifications of this scale have been used in past research on family
communication (Fitzpatrick et al., 2003; Koesten, 2004). All items were measured on a 7-
point strongly agree to strongly disagree Likert-type scale. Questions were written to
specifically reflect satisfaction with the communication of the respondent’s spouse. The
same questions were asked to husbands and to wives, with the object of communication
changed (Sample Husband Item: “I feel like I could talk about anything with her”;
Sample Wife Item: “My husband showed me that he understands what I said.”). The
measure was reliable (α = .94).
Perceived emotional and instrumental support. Items from the Berlin Social
Support Scale (Schultz & Schwartzer, 2004) were adapted for measurement of
instrumental (3-items) and emotional (4-items) support. Both were measured using a 7-
point strongly agree to strongly disagree Likert-type scale. Questions were written to

41
specifically reflect support from spouse. The same questions were asked to husbands and
to wives, with the source and type of support changed. The two measures were reliable
(emotional support α = .94; instrumental support α = .72).
Gender role conflict. Male participants completed two dimensions of the gender
roles conflict scale (O’Neil et al., 1995). Career achievement measured the degree to
which a man believes his worth is tied to employment success. This dimension was
measured on a 5-item, 6-point Likert-type scale, with an acceptable reliability (α = .71).
The emotional expression scale measured the degree to which a man feels that emotional
expression is risky or opens him to criticism. This dimension was measured on a 6-item,
6-point Likert-type scale, with an acceptable reliability (α = .71).
Trait-based anxiety. The short form of the trait-based anxiety scale was used
(Marteau & Bekker, 1992). This scale measured the degree of recalled anxiety the parent
experienced during time of child’s treatment. Trait-based anxiety has been used in past
research as a parent outcome during their children’s treatment for cancer (e.g., Dahlquist
et al., 1996; Kazak et al., 1997). Trait anxiety appears to be an important mediating
variable determining to some extent the way the cancer experience affects parents (Kazak
et al., 1997). For this instrument, a 6-item, 4-point scale was used. Reliability was good
(α = .82).
Acculturation. The 8-item Marin Acculturation (Marin et al., 1987) scale was used
to measure Latino respondents’ ethnic self-identification. Four items measured social
acculturation, and four items measured linguistic acculturation, and both were found to be
reliable measures (social α =.84; language α = .95).

42
Received social support. A received social support instrument was given to mothers
and fathers independently. This component of the instrument was modified from the
UCLA Social Support Interview (Wills & Shinar, 2000). This network instrument asked
parents to identify “the first names of the five most helpful people during your son or
daughter’s treatment.” Respondents could identify as few as zero helpful individuals or as
many as five. Both the gender and the relationship to this person were obtained to attempt
to link mothers’ and fathers’ responses. Although respondents were only asked for the
first name or initial of the other person, most identified the person specifically (i.e., my
mother, my aunt). This made it possible to link parents’ received support networks.
Type of social support received. For each person identified, respondents were asked
to identify the type and quality of support received. Using Wills and Shinar’s (2000)
definitions of instrumental, emotional, and informational support, descriptions of each
type of support accompanied each item. These three types of support are 3 of the 5 types
of support identified as the primary types of social support in past research (Cutrona &
Russell, 1990; Wills & Shinar, 2000). Respondents were asked to identify the amount of
instrumental or material support, such as, “taking care of other children, offer
transportation or money,” the amount of emotional support, “How often did this person
listen to your concerns or talk about how you were feeling,” and the amount of
“information about health care or health insurance or types of cancer treatment.” Items
were measured on a five-point scale (0 = None, 4 = A lot).
Support requested or received. For each person identified, the respondents were
asked to evaluate the nature of the received support from this person on a 3-point scale (3

43
= Help was mainly offered by this person, 2 = Sometimes offered, sometimes had to ask,
or 1= I mainly had to ask for help).
Satisfaction with support. For each person identified, respondents were asked to
identify their satisfaction with the support offered on a single semantic-differential 7-
point scale (1= Not Good, 7 = Very Good). A summary of all sum score means and
standard deviations for all measures can be seen in Table 1.
Demographic information. Parents were asked demographic questions, including
number of children in the family, their own age, race, country of origin, education, and
employment in the past year. Summaries of these variables were provided by sex of
participant in the participant section.
Child’s treatment questions. Parents were asked to report the age, gender, and year
and month of admission to CHLA for the child who had cancer. The children treated at
CHLA typically had 2 siblings, were 54% female, the average age of diagnosis was 5, but
their current age was 11 years. Fifty-two percent were admitted less than 4 years ago,
48% had been admitted longer than 4 years ago, and 85% admitted to the hospital more
than once.

44
Table One: All means and standard deviations by sex of respondent
Men Women
N Mean SD N Mean SD
Communication Satisfaction 22 5.95 0.89 32 5.61 1.55
Perceived Emotional Support from Spouse 22 6.08 1.11 32 5.74 1.55
Perceived Instrumental Support from Spouse 22 5.79 1.08 32 5.66 1.56
GRC - Emotional Expression 21 3.01 1.03 0
GRC - Career Achievement 21 3.51 1.16 0
Trait based anxiety 22 2.33 0.75 32 2.37 0.81
Acculturation – Language 12 3.15 1.39 20 2.66 1.67
Acculturation – Social 12 2.58 1.00 20 2.23 0.85
Number of Sources of Received Social Support 22 3.77 1.45 32 3.13 1.81
Average Emotional Support 22 2.13 0.85 32 2.09 1.08
Average Instrumental Support 22 2.36 0.92 32 2.39 0.97
Average Health Care Support 22 1.07 0.97 32 1.20 1.14
Average Approach or Asked for Support 22 1.61 0.72 31 1.30 0.68
Social Support Overall 22 6.23 1.17 32 6.11 1.80

45
Chapter Three: Results
To explore the relationships between the variables proposed in the hypotheses and
research questions, the data were analyzed using multiple statistical techniques. To
explore the hypotheses relating communication to perceived social support, regression
analyses and t-tests were performed. To explore the relationships between husbands’ and
wives’ received social support, multilevel modeling, regression analyses, and t-tests were
used. Hypotheses using the two measures of gender role conflict were tested using
regression analyses. To explore research questions posed regarding acculturation, t-tests
and correlations were conducted. Finally, the complete Actor-Partner Independence
Model (APIM) was tested using structural equation modeling (SEM). The results are
separated into seven sections.
Section One: Child’s treatment and parent outcomes
Past research suggests that a child’s treatment and type of cancer generally do not
predict social support outcomes. Using number of months since diagnosis, whether the
child was admitted multiple times, age when child was admitted to the hospital, and
child’s sex as predictors, two regressions were performed using perceived support as the
dependent variable (i.e., instrumental and emotional support). None of these variables
predicted instrumental or emotional social support. However, both the child’s age and
whether the child was admitted multiple times predicted trait anxiety. Parents reported
less anxiety when children were admitted to the hospital multiple times (ß = -.65, SE =
.28, t = 2.30, p < .05) and more anxiety for older children (ß = .05, SE = .02, t = 2.31, p <
.05).

46
Section Two: Communication and perceived social support
Communication. To explore how communication affects social support,
regressions analyses were performed, using the two social support variables (i.e.,
instrumental and emotional support) as outcomes. Parent sex, age, race, education,
employment, being foreign born, number of months since the child was diagnosed,
family size, and number of children were used as covariates. When sex differences were
predicted, parent sex was not used as a control variable. These control variables were
selected so possible alternative explanations for the relationships explored in the present
statistical analyses could be taken into account and controlled statistically.
Months since child’s diagnosis merits more discussion. Past research
demonstrates that time since diagnosis of cancer is an important variable in the prediction
of parents’ outcomes. No specific hypotheses were advanced to test the effects of time. In
addition, due to the small sample size, multiple categories could not be created.
Therefore, number of months since diagnosis was used as a covariate rather than as an
independent or moderating variable.
The results of the regression analyses suggest that spouse rated communication
competence positively predicts perceived instrumental support (ß = .77, SE = .17, t =
5.25, p < .001). None of the covariates predicted perceived instrumental support.
Controlling for the same covariates, spouse rated communication competence positively
predicts perceived emotional support (ß = 1.05, SE = .15, t = 7.09, p < .001). These two
regressions offer strong evidence in support of hypothesis one. Unlike instrumental
support, perceived emotional support from spouse was positively predicted by being

47
Latino, using white parents as a reference group (ß = .75, SE = .38, t = 1.96, p < .05), and
being foreign-born negatively predicted perceived emotional support (ß = -.72, SE = .36, t
= 1.99, p < .03).
Sex differences in social support and other outcomes were explored. There were
no sex differences in communication competence, perceived emotional, or perceived
instrumental support. These findings do not support hypothesis two, which stipulated that
fathers would perceive greater support than mothers. However, men seem to have been
over-benefited in all three categories. That is, husbands perceived more support than did
wives, and evaluated their wives as more competent communicators.
To explore the relationship between communication and received social support
(H
3
), the quality of received social support was used as an outcome variable with the
same covariates stated above. Results indicate that communication competence in the
relationship was positively related to the quality of the social support received and this
relationship was approaching significance (ß = .51, SE = .23, t = 1.90, p = .07). This
offers weak support for hypothesis three, which stated that communication within the
relationship should predict the quality of received support from the network. The
relationship between perceived social support from spouse and the quality of social
support was explored using two regression analyses, one exploring the number of social
support sources and the other exploring the quality of the social support sources. The data
do not support the fourth hypotheses. There was a non-significant relationship between
perceived social support within the relationship and the quality or quantity of social
support from the network.

48
Section Three: Received social support
The network methods attempted to capture aspects of received social support
regarding amount, source, and type. On average, parents identified three people from
whom they received social support (M = 3.4, SD = 1.69). Most of this support came from
family (63%), primarily grandparents, or friends (26%). Most sources of social support
were women (69%), which suggests that women were more often providers of social
support to families with children with cancer than were men.
To explore the fifth and sixth hypotheses as well as the second research question,
several t-tests were conducted. Comparisons of fathers’ and mothers’ total received social
support identified trends in the data. In comparison to mothers, fathers received less
social support, of less quality, and were more likely to have to ask for that social support
(see Table One). These three findings do not support the hypotheses, as the differences
were not significant. The lack of significance may be a result of a small number of
respondents. Differences in asking for help were approaching significance (t = 1.85, p =
.09). That is, men were more likely to have to ask for help from their entire social support
network than were women.
Unique social support. Network methods allowed for the identification of portion
of help shared or unique support. More than half of social support members (1.60 of 3.40)
were shared by both parents (See Figure Two). Shared support individuals were also
more likely to be family members, specifically grandparents. To explore research
question two, which asked whether there were sex differences in the sources of support,
the differences between mothers and fathers in shared and unique social support sources

49
were explored. Whether a source of support was shared or not could only be determined
when both parents completed the survey instrument. Therefore, the number of shared
sources in a couple was identical. However, differences in the unique sources of social
support could be explored. In comparison to men (M = 2.00, SD = 2.07), women receive
more instrumental support (M = 2.60, SD = 2.57, p < .01). The quality of social support
for men (M = 6.40, SD = .90) was significantly less than the support women received
from their unique sources of social support (M = 6.80, SD = .37, p < .05). Men (M = 1.50,
SD = .70) and women (M = 1.46, SD = .77) were equally likely to have to ask for support.
Men and women were similar on received emotional (Men M = 2.40, SD = .95, Women
M = 2.60, SD = .81), and health information support (Men M = 1.01, SD = 1.12, Women
M = 1.28, SD = .136) from their unique sources of social support.
Figure Two: Unique and Shared Support Sources

3.2 Theory of optimal matching
One-with-Many Design. The method of collecting received social support data
required a different approach to data analysis. Typically, social support networks are

50
aggregated by averaging the scores for all individuals in the focal actor’s ego network.
This technique is used in most of the studies using ego-networks to measure of social
support (Bissette et al., 2000). This process, however, loses a great deal of variance in the
descriptive variables of the social network members. For example, an individual with
moderate or weak support across all members will have the same average as a person
with a few very good and a few very poor sources of support. Theoretically, it is not
meaningful to suppose that every member of a social support network is equally valuable
or equally supportive. In addition, to test the hypotheses predicted by the theory of
optimal matching, the relative value of different types of social support needed to be
disentangled and tested individually rather than summed across sources. To explore
individual variation in social support network members, multilevel modeling and one-
with-many design will be employed. These techniques have not been employed in past
research in parent support because they are relatively new and may have been unknown
to past researchers. Prior to reporting results, a brief description of the one-perceiver-
many-targets design will be introduced.
Kenny, Kashy, and Cook (2006) identify the one-perceiver many-targets design as
the most common one-with-many design. The research design typically asks the
individual to evaluate the other members in their social group, which could be a family,
an office, or some other network defined by their relationship with the ego. This method
encounters the same problems associated with dyadic data – the data is non-independent.
In this case, the data are non-independent in that they share a common fate – the social
support they provided to ego (Kenny et al., 2006). This methodological design is best

51
served by multilevel modeling, in that the shared variance of the ego (i.e., the parent) is
best modeled treating the ego as a level two predictor, and the network members as level
one predictors.
Using multilevel modeling, the unique effects of the individual were controlled.
Each of the up to five social support sources were treated as level one observations, while
individual parents were treated as level two predictors. This method controls for the
dependence of the ego networks by treating individual sources of support as a
consequence of the individual. The data were analyzed using LISREL 8.8 (Jöreskog, &
Sörbom, 1996). One-hundred and eighty-nine social support sources were identified for
52 individuals (two parents did not complete this section).
To test hypotheses 7, 9, 10, and 11, which were informed by the theory of optimal
matching, the following analyses were conducted. Using overall quality of received
social support as a level one dependent variable, multilevel modeling treated the three
types of social support (instrument, emotional, and health information), the sex of the
source, whether the support was from a family member, and whether the support source
was shared with the spouse as fixed effects. Multilevel modeling is concerned with model
fit as well as the significance of the parameter estimates (Roberts, 2004). The best fitting
model only included instrumental and emotional support amounts and family
membership as fixed effects. Results indicated that amount of instrumental support (ß =
.21, SE = .06, WALD = 3.76, p < .001) and amount of emotional support (ß = .45, SE =
.08, WALD = 5.50, p < .001) significantly predicted overall quality of support. These
results support hypotheses 9 and 10. Hypothesis 7 predicted that support would be of

52
higher quality when it was received without asking, and eleven predicted that health
information support would be a predictor of overall support quality. The data do not
support hypotheses 11 and 7. Results also provided some information pertaining to
research question three, which asked which source of support was most valuable. Being a
family member was marginally related to overall support quality (ß = .20, SE = .11,
WALD = 1.80, p = .08). The lack of significant findings for the sex of the source of
assistance as well as the shared vs. not shared dimension should be mentioned. These
results suggest that the quality of the source of support is not influenced by the sex or
whether the couple shares the source.
Finally, when the individual is treated as a level-two predictor, such as in one-in-
many designs, multilevel modeling allows the research to explore inter-individual
differences. Results indicted that parents were significantly different from one another at
the level of social support they received (ß = 4.76, SE = .24, WALD = 19.92, p < .001),
indicating that on average parents rated the support received about 5 out of 7, but parents
differed significantly from one another.
Section Four: Gender role conflict
Two types of gender role conflict were explored: conflict over career achievement
and restricted emotional expression (O’Neil et al., 1995). To explore whether restricted
emotional expression or career achievement conflict predicted anxiety (H
12,
H
13
), the
effects of gender role conflict were tested on anxiety, using the same control variables
listed in the communication section. Neither hypothesis was supported. However, being
foreign-born predicted trait anxiety (ß = .82, SE = .03, t = 2.82, p < .01).

53
To explore the next hypotheses regarding the influence of masculine gender role
conflict on the perception of social support, both types of conflict were used to predict
fathers’ outcomes. Using the same control variables listed in the communication section,
the effects of gender role conflict were tested on emotional and instrumental support.
Results do not support hypothesis fourteen. Men with high scores on restricted emotional
expression did not perceive less instrumental or emotional support from their wives. In
addition, men with high scores on career achievement did not report less instrumental or
emotional support from their wives. These results do not support hypothesis fifteen.
Finally, hypothesis sixteen predicted that fathers with greater career achievement conflict
should report less received social support. This was not supported. Gender role conflict
did not predict the quality or quantity of received social support.
Another approach was taken to explore whether masculinity measures had a
greater effect on the father’s spouse than on himself. To do so, mother’s scores were used
as the dependent variable. To test whether men’s gender role conflict impacted their
wives perceived social support, men’s scores were used as predictors and their spouse’s
perceived emotional and instrumental support scores were used as outcomes. Controlling
for demographic covariates, men with high scores on career achievement were more
likely to have wives who reported receiving less emotional support (ß = -.54, SE = .27, t
= 2.04, p < .05), and less instrumental support (ß = -.65, SE = .21, t = 3.10, p < .01). Men
with more restricted emotional expression were not more likely to have wives perceiving
less instrumental or emotional support. These results will be further explored in the final

54
section that will test simultaneous actor and partner effects within the entire family
communication model.
Section Five: Acculturation
To explore the research questions advanced regarding acculturation, regression
analyses were performed on Latino respondents (N = 32) to determine if acculturation
predicted anxiety and social support. The first regressions analyses demonstrated that
acculturation had no effect on trait anxiety, controlling for the demographic covariates.
Research question seven asked whether acculturation had an effect on social support.
Analyses demonstrated that acculturation had no significant impact on either the quality
or quantity of received social support. The third set of analyses found an insignificant
effect of acculturation on emotional support, but uncovered a positive effect of language
acculturation on perceived instrumental support (ß = .40, SE = .14, t = 2.23, p < .05). That
is, Latino parents who were primarily English speakers were more likely to report greater
instrumental support from their spouse. The low number of Latino participants may have
limited the power of analyses.
To explore research question six, differences between Latino parents and white
parents were compared using t-tests. These analyses were mainly exploratory as they did
not predict specific differences. No significant differences in communication, social
support, gender role conflict, and trait anxiety were found. There are no differences
between Latino and white families in the variables of interest.

55
Section Six: Testing the complete model
Dyadic Analysis. When data are collected from both partners in a relationship,
partners’ responses are often highly related to one another. In the present study, each
partner’s communication satisfaction and perceived social support are dependent upon
one another. In the review of literature, only one study collected data from paired mothers
and fathers (Hoekstra-Weebers et al., 1998). The researchers did not acknowledge the
problem with non-independence in the sample, and performed correlations and t-tests to
explore their results. It is possible that the problem with the non-independence of
couples’ data is not widely known in clinical research. To this author’s knowledge, this
dissertation is the first explore relationships between parents of children with cancer
using dyadic procedures.
Kenny and colleagues (2006) recommend that any data analysis that works with
less than 48 pairs should treat the data as non-independent. Furthermore, the lack of
independence in these data is also reflected in the significant correlations between
partners’ scores, which ranged from .20 to .57 (Kenny et al., 2006). The relationship
between partners’ social support and communication is at the center of the present
investigation. Therefore, rather than analyzing these data at the individual level, we used
the dyad as the unit of analysis. Kenny and colleagues (2006) suggest that when dyadic
data are collected and the dyads are distinguishable, researchers should use SEM.
Specifically, the Actor-Partner Independence Model allows researchers to estimate the
impact of husbands’ communication and gender role conflict on their own level of
perceived social support (actor effect), and on their wives’ perceived social support

56
(partner effect), controlling for their wives’ communication satisfaction (see Figure 3).
The following section describes the significant paths estimated using APIM and SEM.

57
The model advanced in the literature review section placed communication at the
core of family function. To incorporate both partners’ communication competence on the
same model on the APIM, mothers’ and fathers’ satisfaction are predictor variables, but
allowed to covary. The two measures of gender role conflict are also expected to function
as predictor variables in that they are psychological measures of fathers, whose effects
can be modeled for the dyad. The first level outcomes in the SEM model include men’s
perceived emotional and instrumental support and women’s perceived emotional and
instrumental support. In addition, the quality of social support was considered to be an
outcome variable, in that it is conceived to be a product of communication in the
relationship. Finally, the level of trait anxiety of both parents represented the final set of
outcome measures in this model (see Figure Four for final model). This portion of the
model allowed for the testing of hypothesis eight, which stated that trait anxiety will be
positively related to the quality of received social support. To create a model in the
APIM, both actor and partner effects (mothers effect on her outcomes, and her effect on
father’s outcomes) are predicted and tested for significance.
The results of gamma path coefficients demonstrate strong actor effects. When
men rate their wives as being competent communicators, they perceive a greater amount
of emotional (ß = .89, SE =.11, t = 8.04, p < .001) and instrumental support (ß = .48, SE =
.17, t = 2.83, p < .01) from their wives. Similarly, when wives perceive their husbands to
be competent communicators, they perceive a greater amount of instrumental support (ß
= .81, SE = .13, t = 6.32, p < .001) from their husbands. Similar to analyses on gender
role conflict conducted previously, men’s career achievement was negatively related to

58
their partners’ social support. Men with high scores on career achievement were more
likely to have wives who reported receiving less emotional support (ß = -.18, SE = .08, t
= 2.09, p < .05), and less instrumental support (ß = -.36, SE = .13, t = 2.70, p < .05).
Neither communication competence nor gender role conflict predicted the quality of
received social support for either fathers or for mothers.
The second component of the model explores the relationship between social
support perceptions and trait anxiety for mothers and fathers. Perceptions of instrumental
support and emotional support were related. Perceived greater emotional support leads to
a greater perceived instrumental support for men (ß = .21, SE = .11, t = 1.96, p < .05) and
women (ß = .81, SE = .12, t = 6.70, p < .001). In one of the few partner effects in the
model, the quality of social support for wives was related to father’s perceived
instrumental support. That is, men who felt more supported by their wives were more
likely to have wives with higher quality social support (ß = .22, SE = .08, t = 2.73, p <
.01). For men, there was no unique predictor of social support quality.
The predictors of recalled trait anxiety were sex specific. For men, anxiety was
negatively related with their own quality of social support (ß = -.21, SE = .10, t = 2.03, p
< .05) and their wives’ quality of social support (ß = -.23, SE = .08, t = 2.73, p < .01).
That is, men were less anxious when either they or their wives had higher quality social
support. In addition, there was evidence of a partner effect for women’s anxiety. Fathers
who perceived greater instrumental support were more likely to have wives who felt less
anxiety (ß = -.37, SE = .06, t = 6.23, p < .001). That is, if men feel supported by their
wives, their wives are less likely to be anxious.

59
The global fit statistics indicated that the final model was a good fit to the data
(RMSEA = .05). In addition, the chi-squared test indicated that the model misfit was not
significant, (X
2
= 47.57, p < .37). Other fit indices, the CFI and GFI, indicate a weaker fit
to the data, but these indices are less sensitive to sample size considerations (Byrne,
1997). Although the number of dyads was small, the critical N value was only 21, which
indicates only 21 dyads were necessary to fit this model. In addition, when testing dyadic
data, the global fit of the model is less relevant than the significance of the particular
parameters, because the SEM technique in this situation allows for the testing of non-
independent samples and is less focused on model testing (Kenny et al., 2006).

60

61
Chapter Four: Discussion
This dissertation set out to achieve two overarching goals. To extend prior clinical
research on predictors of parents’ health, this dissertation employed communication
theory, network methodology, and indicators of gender role conflict. In addition, this
dissertation addressed the research interests of Childrens Hospital Los Angeles,
specifically to learn more about the role of fathers. Results suggest that communication
between parents is a meaningful indicator of family function, that network methodology
offers new ways of looking at parents’ social support, and gender role conflict –
particularly career achievement – plays an important role in the dynamics of the family.
This final chapter summarizes the findings of the present investigation, links the results to
past research, discusses the limitations and strengths of the project, and offers suggestions
for future research.
Section One: Communication and perceived social support
The results suggest that communication is at the core of family functioning. When
a parent has a partner who is an effective communicator, both parents experience greater
emotional and instrumental support. These findings reinforce past research on the
relationship between social support and communication (e.g., Burleson & MacGeorge,
2002; Lyons et al., 2006; Sarason et al., 1990), and reflect Reinhardt and colleagues
(2006) statement, “Providing targeted help requires good communication” (p. 127).
Furthermore, this project empirically tested the oft-asserted relationship between
communication and social support, which is sometimes acknowledged but typically
unmeasured in health research (Segrin & Flora, 2005; Streisand et al., 2001). The

62
strength of the relationships between communication and social support suggest that
communication competence may be at the root of perceived social support. Poor
communication begets poor support, while competent communication enables support.
Fathers and mothers perceived equivalent amounts of support from their
respective partners, but dyadic analyses suggest that sex may moderate the relationship
between communication and emotional support. Husbands and wives were perceived as
equally competent communicators and perceived equal levels of instrumental and
emotional support. This does not support hypotheses two, which tested sex differences in
perceived social support. However, the results of the dyadic analyses imply that the
relationship between communication and social support may be moderated by sex. While
both mothers and fathers perceived greater instrumental social support from an
effectively communicating partner, only fathers perceived greater emotional support. One
interpretation of this finding is that when mothers are effective communicators, they offer
both emotional and instrumental support to their spouses. When fathers are effective
communicators, they provide greater support in the instrumental domain only. This
project was the first in the clinical setting to use the actor-partner independence model, so
this moderation effect should be replicated in future work using dyadic analyses. A larger
sample size would be critical in future studies to achieve the statistical power to test for a
true moderation effect.
Section Two: Received social support
The model of family ecology introduced in the literature review suggested that
communication enables perceived and received social support, which reduces negative

63
outcomes such as anxiety. Results confirm that competent communication within the
couple increases the perception and reception of support. Couples that communicated
more effectively were more likely to receive higher quality social support. This
relationship should be interpreted with caution as the relationship was only approaching
significance. Furthermore, in testing the complete model, the relationship between
perceived and received support was examined. Past research has suggested that received
social support is sometimes an indicator of poor individual functioning (e.g., Coyne et al.,
1990; Reinhardt et al., 2006). The results of this study demonstrate that parents who are
well supported in the relationship are no more or less well supported by their social
network. That is, the relationship between communication and received social support is
not mediated by perceived social support. Instead, it appears that communication
precedes the perception and reception of social support.
The use of dyadic analyses revealed new insights into the relationship between
parents. The SEM analyses demonstrated that when mothers receive greater quality social
support from social networks, fathers perceive greater instrumental support from their
wives. This finding suggests an intuitive connection previously uncovered by past
research in this domain. When mothers are well supported by their social network, their
husbands feel more supported as well. This suggests that support from the extended
family channels through mothers, who then help their spouses. The results do not suggest
this act is to the detriment of mothers. In fact, when fathers perceive greater support from
their wives, mothers report less anxiety. As these data are cross-sectional, rather than
longitudinal, the directionality of this relationship cannot be determined. Therefore, it is

64
equally possible that mothers who have good social support from their network and who
experience less anxiety are able to support their husbands the most. It is not by supporting
the father that a mother benefits from less anxiety, it is only because a mother is
experiencing less anxiety that she can provide support. Either explanation is in line with
past research that states that in times of conflict, men often rely more on their wives than
wives do on their husbands (Dahlquist et al., 1994; Hoekstra-Weebers et al., 2000).
Finally, the model of family ecology suggested that through social support,
communication reduces negative outcomes. The complete model demonstrates that for
men, anxiety was negatively related to the quality of social support from their own
network and from the quality of social support from their wives’ network. This means
men were less anxious when either they or their wives received higher quality social
support. Men are benefited by the quality of the family’s entire social support network,
while mothers are more affected by the support within the couple. Taken together, it
appears that competent communication enables the reception and perception of social
support, which reduces anxiety in parents. Results should be replicated in future research
using dyadic methods.
2.2 Unique versus shared support
One of the innovations of the current project was the use of network methods.
Parents were asked to identify the first names of the five most helpful people during their
child’s treatment. Although respondents were asked only for the first name or initials of
the their source of support, most identified the person specifically (e.g., my mother, my
aunt). This serendipitous circumstance made it possible to link parents’ received support

65
networks to one another. All sources of received social support could be identified then
as either shared (identified by both parents) or unique (identified by only one parent).
Linking the identities of network alters to create a complete sociogram is a common
technique in social network analysis (Wasserman & Faust, 1994), but this dissertation
may be the first to link the ego networks of couples.
On average, parents identified three people from whom they received social
support. Nearly two-thirds of this support came from family, primarily grandparents, and
nearly two thirds of sources of social support were women. This outcome also replicates
past research that found family are the most active network members (Wellman, 1992),
and that women are more likely than men to fill the role of family support provider
(Conger et al., 1993). When comparing all sources of received support, differences
between fathers and mothers in received social support were not significant. Most
researchers using network methodology average the scores from multiple sources of
support, and do not explore the identities of these sources (Bissette et al., 2000). Further
investigation into support sources demonstrated why shared support identification
matters.
Past research has documented fathers’ lack of social support, both in quality and
in quantity (Hoekstra-Weebers et al., 1999; Hoekstra-Weebers et al., 2000; Hoekstra-
Weebers et al., 2001; Sloper, 2000). The results of this investigation suggest that this
relationship should be qualified. Men and women received equivalent support from
mutually shared sources in both quantity and quality, and differences emerged only when
separating unique from shared support sources. In comparison to men, women receive

66
more instrumental support of higher quality, and had to ask for support less from their
unique sources. These differences would have been overlooked if shared and unique
sources were not separated.
The present study was able to address the question of whether or not parents
received different types of social support from networks. Previous research (Burleson &
Kunkel, 2006; Hoeskstra-Weebers et al., 2001) supported the prediction that women
would provide more and better emotional support than men. However, results indicated
that both parents’ unique support sources provided equivalent emotional support, but
women received more instrumental support. In addition, fathers and mothers did not
differ in how much support they received from men or women. Instead, both were more
likely to receive support from women. This could explain why fathers and mothers
received equivalent emotional support. Women, who are more capable of delivering
emotion-focused messages, were equally active in both parents’ networks.
Relying on network methodology allowed for the possibility of exploring another
research question, from whom does this support come? Knowing the source of support
helps to overcome a gap in research on received social support (Thoits, 1995). Results
indicated that shared support individuals were more likely to be family members,
specifically grandparents. Even when considering unique sources of social support, it
appears that family and friends both play an equally important role for mothers and
fathers. This confirms previous research into received social support that relatives and
friends are second only to one’s spouse in the provision of support (Morrow et al., 1984).
Scholars of gerontology may be interested in future work exploring how grandparents

67
influence and assist families during cancer treatment. Kazak (1989) points out that
grandparents can be very helpful because their assistance is less encumbered by worry of
interfering in family affairs. On the other hand, their involvement may sometimes exceed
the needs or desires of the parents. Knowing who is supported by which sources may help
to extend future inquiries into the navigation of the role extended families can and should
play.
Additionally, network analysis may offer answers to future questions about
parents’ social support networks. Separating unique and shared support can directly
explore whether shared support is more effective, results from better communication and
coordination between parents, and whether unique support sources have hitherto
unknown properties. If shared sources of support, like family networks, are tightly
connected, does this allow them to diffuse responsibility and support the family of the
sick child more effectively? If sources of support are disconnected from one another,
supporters might not be able to accurately assess what support is needed or provide
targeted support to families. While one outside caregiver may fatigue over time, a more
densely connected network could share responsibility (Hoeskstra-Weebers et al., 2000).
Future research may explore how communication outside of the immediate family
coordinates support in the most effective manner.
2.3 Theory of optimal matching
To further explore the value of received social support, the theory of optimal
matching was utilized. The theory states that support that meets the needs of parents is of
higher quality than support that is not matched to those needs (Cutrona & Russell, 1990).

68
Three hypotheses were advanced to test this theory. Multilevel modeling made it possible
to directly test these hypotheses. This analysis technique allows for a simultaneous test of
the theory, and controls for the lack of independence of the ego networks by treating
individual sources of support as a consequence of the individual. That is, it overcomes
independence problems germane to one-perceiver many-targets designs (Kenny et al.,
2006). This method can also explore whether every member of a social support network
is equally valuable or equally supportive. Therefore, the relative value of different types
of social support were disentangled and tested individually rather than summed across
sources. This dissertation is the first to use multilevel modeling to study support for
parents of children with cancer.
The results suggest that greater instrumental and emotional support offers unique
contributions to parents (H
9
& H
10
), while health information support did not predict the
quality of received support (H
11
). This confirms the theory of optimal matching, which
suggests that support quality should match needs of family. However, it is difficult to
discern from these results whether the treatment of a child for cancer should be seen as an
uncontrollable or controllable event, the two dimensions of Cutrona and Russell’s theory.
Results demonstrate that help in childcare, cooking, or driving was a critical predictor of
quality of social support, which suggests that support that attended to quantifiable and
controllable stressors was helpful. However, emotional support was also related to overall
quality. This suggests that the uncontrollable fears and anxieties produced by cancer in a
child also require emotional support. Therefore, the domain of children’s cancer is both
controllable and uncontrollable and the best support attends to both needs. Past research

69
would identify this sort of support as multiplex (Albrecht & Goldsmith, 2003). Multiplex
sources of support can provide nurturing and caring and deliver tangible assistance.
In addition, multilevel modeling demonstrated that support provided by family is
significantly more valuable than support from non-family sources. Whether the support
source was shared or unique and male or female was also tested, but found to be
insignificant. However, the variance of shared support may have been accounted for by
including families in the equation. Families were rated as higher quality sources of
support, and since most shared support sources are family, the shared or not shared
variable showed no effect. Overall, results suggest that the best support is provided by a
family member who attends to both the instrumental and emotional needs of the family.
Section Three: Gender role conflict
Another goal of this dissertation was to more closely examine the role of fathers.
In the literature review, within sex differences were explored in relation to men’s
communication and social support behaviors. Results suggest that there is merit to an
approach that examines issues within men. That is, gender should be understood as not
just a dichotomy or a control variable, but something that is internalized by varying
degrees. Specifically, this dissertation explored two types of gender role conflict, career
achievement and restricted emotional expression (O’Neil et al., 1995). Results offer
strong support for the detrimental effects of career achievement conflict. In addition, it is
not men who are most affected by their adherence to strict standards for career success, it
is their wives.

70
To explore whether restricted emotional expression or career achievement conflict
predicted anxiety (H
12,
H
13
), the effects of gender role conflict were tested on trait
anxiety. Neither hypothesis was supported. The effects of gender role conflict were tested
on fathers’ perceived emotional and instrumental support, but also found no effects (H
14,
H
15
). It was only when mother’s scores were used as the dependent variable that the
effects of gender role conflict were uncovered. Men with high scores on career
achievement were more likely to have wives who reported receiving less emotional and
instrumental support. These results were duplicated in the dyadic model, which
controlled for actor and partner effects as well as the lack of independence in couple’s
data.
Two important questions arise from these analyses. First, why is career
achievement but not emotional expression predictive of negative outcomes? Second, why
does career achievement conflict affect wives, but not the husbands? Three earlier studies
(e.g., Cook, 1984; Kalins et al., 1980; Sloper, 2000) point to the importance of
employment problems in predicting fathers’ psychological well-being. This is partly due
to the division of labor within the family. Often, women will quit their jobs to help take
care of a child who is ill, while men continue to work. The present sample reflects these
past findings. Ninety-five percent of men in this present investigation worked full time,
while less than 50% of mothers were employed, and of the mothers who did work, nearly
a third were employed part-time. It is possible that men in these circumstances are more
exposed to work-related stresses. Furthermore, past research suggests that men feel even
more inclined to focus on their careers during the illness of a child (Cook, 1984; McNeil,

71
2007). They embrace the role of breadwinner because they see it as the best way to help
their family. Yet, this emphasis on career may have an unintended negative effect. Men
who believe that career success is the best measure of a man’s worth have wives who
perceive less support from their husbands. Rather than helping their families, emphasis
on career achievement is harmful to men’s families.
However, this does not account for the lack of predictive utility in emotional
expression. Lack of emotional expression from fathers has been linked to poor outcomes
in other research (Good & Wood, 1995; Sharpe et al., 1995). One explanation is that
stress arising from emotional expression does not affect men in this domain. When men
are faced with the fear and uncertainty of a child’s diagnosis of cancer, they may feel
more comfortable reporting distress, because they feel it is appropriate for them to do so,
despite gender role expectations (Hoekstra-Weebers et al., 2000). Men feel less
encumbered about expressing their emotions in this domain than they might in other
contexts. Therefore, indicators of emotional expression conflict would not predict anxiety
or perceptions of social support in this circumstance.
Finally, why are mothers, but not fathers, affected by men’s gender role conflict?
It is possible that these effects would have been demonstrated in past research, but most
studies do not collect data on both partners. Therefore, the effects of one partner on the
other could not be found. Only one previous study in this domain collected data from
paired mothers and fathers (Hoekstra-Weebers et al., 1998). These authors did not
acknowledge the problem with the non-independence of the sample, and did not test for
partner effects. To this author’s knowledge, this dissertation is the first explore

72
relationships within couples using dyadic procedures, therefore it is the first to be able to
document the effects of one partner on another. Another explanation is simply that
mothers are more affected by fathers’ behaviors than fathers’ are by their own behavior.
Past research on gender role conflict suggests that it affects women the most when it
manifests in men’s behavior (Brieding, 2004). Perhaps career achievement conflict is
manifested in men’s behavior in this domain in ways that influence mothers, but not
fathers. How this occurs is a topic of future research.
Section Four: Acculturation
This dissertation explored possible ethnic and cultural differences in parents of
children who are treated for cancer. Few studies had demonstrated meaningful results in
past literature (Hoesktra-Weebers et al., 2001). As such, research questions in this
dissertation probed for potential effects of acculturation and ethnicity. Results suggest
that acculturation has little to no effect on parents’ outcomes, such as trait anxiety, the
quality or quantity of received social support, and perceived emotional support. A
positive effect of language acculturation on perceived instrumental support was revealed.
Latino parents who were primarily English speakers were more likely to report greater
instrumental support from their spouse. It is difficult to explain why this effect but not
others was uncovered. This single finding may be spurious because differences between
Latino parents and white parents in communication, social support, gender role conflict,
and trait anxiety were not found.
One explanation for these results is that there are no differences between whites
and Latinos, so none should have been found. The ability to report these differences, if

73
they were present, however, may have been due to limited power caused by the low total
number of Latino participants (N = 32). Given the lack of past research on these
populations, other explanations are difficult to formulate. One difference that did emerge
is that being foreign-born negatively predicted perceived emotional support and
positively predicted trait anxiety. Potentially, those parents who are foreign-born may
have additional stressors in their lives that increase anxiety, such as employment
differences, legal problems, financial difficulties, and being separated from their
extended family (Kupst, 1993). However, these results should be interpreted with caution
given the small sample size and the lack of significant findings in other comparative data.
Overall, the data suggest that gender conflict and the role of communication have a
greater impact on family function than do ethnic differences.
Section Five: Limitations, strengths, and future directions
5.1 Limitations of this study
There are three limitations to this study: a small sample size, the use of cross-
sectional data, and a need to group parents at different stages of cancer treatment and
recovery into one sample. First, the small sample may have impacted statistical power to
find significant results, and it diminishes generalizability of these results to a larger
population. However, past publications have reported results on samples as small as 25
(Brown et al, 1993). This suggests that the sample size alone should not reduce the value
of these findings. In their review of literature on childhood cancer and social support,
Hoekstra-Weebers and colleagues (1998) note that childhood cancer is a rare disease
resulting in small study samples, which limits studies’ abilities to generalize, make broad

74
claims, and suggest strategies for intervention. Additionally, statistical tests on the sample
size in structural equation modeling suggest the sample was large enough to report
findings. The critical N value suggested in the dyadic model was only 21, which indicates
only 21 dyads were necessary to fit this model. Overall, the results demonstrate strong
relationships despite the size of the sample.
The second and third limitations are related. The reliance on cross-sectional data
and grouping of parents at different times of cancer treatment and remission were
necessary given time and resource constraints. It is difficult to gather data on parents at
CHLA because there are few circumstances where both parents are available. Hospital
staff report that both parents were likely to be at the hospital at the same time only when
serious medical procedures were undertaken. Surveying parents under these
circumstances would be ethically questionable and may skew results. The Celebrate Life
survivorship event offered the opportunity to recruit both parents as participants
simultaneously in a less stressful environment. However, given these restrictions this
study could not afford to be overly discriminating in the selection criteria. This project
could not employ longitudinal methods realistically given the temporal and financial
constraints on the author. Therefore, the findings of this project are correlational and
inferential, but are not causal. This means that any relationships between parents and
their social and psychological outcomes should be interpreted with caution.
5.2 Strengths of this study
There are three strengths to the present study: the integration of communication
theory and clinical research, the use of new statistical procedures, and the emphasis on

75
fathers. The illness of a child affects the entire family’s well-being (Kupst, 1993). This
dissertation recognized the importance of looking at the family as a mutually dependent
group, and used systems theory to model these lines of influence. This use of theory
asserted that one parent’s actions influence the behaviors and outcomes of all other
members of the family. As a consequence, the use of dyadic analysis reflected theory in
that it underscored the interdependence of parents’ outcomes. In addition, measuring
communication brought about new insights into family function. The results of this
dissertation reinforce the concept that communication between parents is at the core of
family functioning, and influences individuals both within and outside of the family.
The statistical methods used in this project helped to advance study into family
communication. Specifically, issues of non-independence that were previously ignored in
research in this domain were accounted for using multilevel modeling and structural
equation modeling (Kenny et al., 2006). This dissertation may be the first study in this
domain that employed these two techniques. Using multilevel modeling allowed for
direct testing of the theory of optimal matching, and controlled for the non-independence
of network data. Dyadic data analysis procedures tested the complete model, which
offered insights into how the behavior and attitudes of one parent affected the other
parent. In sum, these procedures allowed for family level variables to show effects on
individual level outcomes. Thereby, they more fully accounted for the assumptions of
systems theory, and as such offered important statistical insight into how mutually
dependent individuals can be measured. In addition, the network methods measuring
received social support were innovative. Although past research has used network

76
methodology, measuring the sources of support both parents at the same time offered
interesting insights about shared (both parents) versus unique (only one parent) sources of
social support.
Finally, the emphasis on fathers is noteworthy. Clinical researchers have
recognized that fathers’ place in the family is poorly understood (Kazak et al., 1997;
Hoekstra-Weebers et al., 1998). This investigation suggests that when fathers internalize
their role as an economic provider, it may have detrimental effects on families.
Understanding these types of within sex differences are useful to better understand how
men and their families are affected by the internalization of gender roles, and may better
account for why some families are less affected by the stress of childhood cancer than
others.
5.3 Directions for future research
Interventions. This dissertation has only begun to explore issues regarding
communication and social support. One of the goals of this project was to offer feedback
to Childrens Hospital Los Angeles about the potential value of communication and
outreach to fathers. The results of this research would support two possible interventions
for parents. First, CHLA may want to explore the possibility of offering a basic
communication intervention for parents. CHLA currently offers an impressive range of
psychosocial programs and professional assistance to parents. In fact, their work on
maternal problem-solving may be very useful in designing a communication intervention
(Varni et al., 1999). This intervention for mothers is theoretically grounded and employs
many of the principles of program evaluation. If CHLA was interested in implementing a

77
parents’ communication intervention, there are many possible communication
interventions available (Kouneski & Olson, 2004). The present research helps to guide
the development of future communication interventions by identifying the relevant
variables in predicting parents’ success. By identifying parents who are at greater risk for
poor functioning, hospital staff will be more capable of utilizing limited resources most
effectively.
With regard to fathers, research has suggested that interventions must be sensitive
to the effects of gender role conflict that impede men from seeking help. Men who report
a high gender role conflict are less likely to seek out counseling, to participate in support
groups, and even discuss emotional issues (Addis & Mahalik, 2003). This means men
who are the greatest risk of negative outcomes are the least likely to participate in a
hospital sponsored program. The use of normative attitudes has demonstrated utility in
guiding men’s health behaviors in the past (Courtenay, 2000c). A man who is surrounded
by a supportive group of other men that encouraged members to share their problems
might be more likely to seek help than a man whose only social network discouraged
such behaviors (Addis & Mahalik, 2003). As Addis and Mahalik (2003) suggest, a men’s
group that emphasizes reciprocity and acceptance could normalized healthy behaviors,
and would establish an active role for each man in establishing that norm.
Interventions must also be sensitive to what sort of help men will accept. Men are
more likely to respond positively to materials that are cognitive-behavioral and problem-
solving focused (Berger et al., 2005). Case studies or other story-telling techniques might
reduce the stigma and help to normalize men’s emotional reactions. In campaigns to

78
increase men’s attendance at socio-emotional services, identifying services as
‘workshops’ and ‘seminars’ are more appealing to men who are high in gender role
conflict (Robertson & Fitzgerald, 1992).
Research. In addition to working on possible interventions, future research
projects could extend and replicate the findings of the present investigation. A few
additional research ideas have emerged. Sarason and colleagues (1990) emphasize that in
the past advances in network methods have opened up new possibilities of theory
building in social support. The concept of unique or shared support may be one such
insight. This concept is more descriptive and ecologically valid in describing the received
social support of parents, and can allow researchers to predict or explain how sharing a
support network might be qualitatively different than receiving help from a unique source
of support. It also offers new possibilities of understanding the social context in which
couples reside.
Future longitudinal research may be able to better understand how received social
support and anxiety interact. The results of this investigation suggested that support from
the social network channels through the mother, which creates benefits for the father.
Future research might be able to determine if mothers are more able than fathers to
coordinate help for the entire family, even in times of great stress. Another possible
research topic could determine the directionality of the relationship between mothers’
anxiety and fathers’ perceptions of support. Is it the case that in giving support to fathers,
mothers reduce their own anxiety? Or, is it that less anxious mothers can provide the best
support to fathers? Future research could determine which is a more tenable hypothesis.

79
Scholars of gerontology may be interested in future work exploring how
grandparents influence and assist families during cancer treatment. Morrow and
colleagues (1984) found that support received from grandparents was positively related to
adjustment for parents. Kazak (1989) cautions that grandparent over-involvement can be
detrimental to a family with a chronically ill child. These investigations into specific
sources of social support would be able to use insights of communication research into
enacted support. As Albrecht and Goldsmith (2003) point out, asking for social support
requires face-maintenance behaviors. Knowing how to navigate asking and receiving
support, but also appropriately setting boundaries for that support to protect face, are all
issues that communication theory would be well-equipped to study (Lyons et al., 2006).
In fact, communication competence may be one of the most critical factors in the ability
to obtain the desired support, and to negotiate appropriate boundaries and expectations
within an individuals’ support network (Burleson & MacGeorge, 2002; Lyons et al.,
2006; Reinhardt et al., 2006; Sarason et al., 1990).

80
References
Addis, M. E., & Mahalik, J. R. (2003). Men, masculinity, and the contexts of help
seeking. American Psychologist, 58(1), 5-14.

Albrecht, T. L., & Goldsmith, D. J. (2003). Social support, social networks, and health. In
T. L. Thompson, A. M. Dorsey, K. I. Miller, & R. Parrott, Handbook of health
communication (pp. 263-284). Mahwah, NJ: Lawrence Erlbaum Associates.

Berger, J. M., Levant, R. F., McMillan, K. K., Kelleher, W., & Sellers, A. (2005). Impact
of gender role conflict, traditional masculinity ideology, alexithymia, and age on
men's attitudes toward psychological help seeking. Psychology of Men and
Masculinity, 6, 73-78.

Blackhall, L. J., Murphy, S. T., Grank, G., Michel, V., & Azen, S. (1995). Ethnicity and
attitudes toward patient autonomy. Journal of the American Medical Association,
274, 820-825.

Bradford, L., Meyers, R. A., & Kane, K. A. (1999). Latino expectations of
communicative competence: A focus group interview study. Communication
Quarterly, 47, 98-117.

Breiding, M. J. (2004) Observed hostility and observe dominance as mediators of the
relationship between husbands’ gender role conflict and wife outcomes. Journal
of Counseling Psychology, 51, 429-436.

Brissette, I., Cohen, S., & Seeman, T. E. (2000). Measuring social integration and social
networks. In S. Cohen, L. G. Underwood, & B. H. Gottlieb (Eds.), Social support
measurement and intervention: A guide for health and social scientists (pp. 53-
85). Oxford: University Press.

Brown, R. T., Kaslow, N. J., Madan-Swain, A., Doepke, K. J., Sexson, S. B., & Hill, L. J.
(1993). Parental psychopathology and children’s adjustment to leukemia. Journal
of American Academic Child Adolescent Psychiatry, 32, 554-561.

Burleson, B. R., & Kunkel, A. (2006). Revisiting the different cultures thesis: An
assessment of sex differences and similarities in supportive communication. In K.
Dindia & D. J. Canary (Eds.), Sex differences and similarities in communication
2
nd
Edition (pp. 117-159). Mahwah, NJ: Lawrence Erlbaum Associates.

Burleson, B. R., & MacGeorge, E. L. (2002). Supportive communication. In M. L. Knapp
& J. A. Daly (Eds.), Handbook of interpersonal communication 3
rd
Edition (pp.
374-424). Thousand Oaks, CA: Sage.

81
Burrell, N. A. (2002). Divorce: How spouses seek social support. In M. Allen, R. W.
Preiss, B. M. Gayle, & N. A. Burrell (Eds.), Interpersonal communication
research: Advances through meta-analysis (pp. 247-262). Mahwah, NJ: Lawrence
Erlbaum Associates.

Byrne, B. M. (1998). Structural equation modeling with LISREL, PRELIS, and SIMPLIS:
Basic concept, applications, and programming. London: Lawrence Erlbaum
Associates.

Canary, D. J., & Stafford, L. (2007). People want – and maintain – fair marriages: Reply
to Ragsdale and Brandau-Brown. Journal of Family Communication, 7, 61-68.

Childrens Hospital Los Angeles. (2007, August). About us: Mission statement. Available
FTP: http://www.childrenshospitalla.org/

Christensen, P. (2004). The health-promoting family: A conceptual framework for future
research. Social Science & Medicine, 49, 377-387.

Cohen, S., Gottlieb, B. H., & Underwood, L. G. (2000). Social relationships and health.
In S. Cohen, L. G. Underwood, & B. H. Gottlieb (Eds.), Social support
measurement and intervention: A guide for health and social scientists (pp. 29-
52). Oxford: University Press.

Conger, R. D., Elder, G. H., Simons, R. L., & Ge, X. (1993). Husband and wife
differences in response to undesirable life events. Journal of Health and Social
Behavior, 34, 71-88.

Connell, R. W. (2002). Gender. Malden, MA: Polity Press.

Cook, J. A. (1984). Influence of gender on the problems of parents of fatally ill children.
Journal of Psychosocial Oncology, 2, 71-91.

Courtenay, W. H. (2000a). Teaming up for the new men’s health movement. Journal of
Men’s Studies, 8, 387-92.

Courtenay, W, H. (2000b). Behavioral factors associated with disease, injury, and death
among men: Evidence and implications for prevention. Journal of Men’s Studies,
9, 81-142.

Courtenay, W. H. (2000c). Men, gender, and health: Toward an interdisciplinary
approach. Journal of American College Health, 48, 243-246.

82
Coyne, J. C., Ellard, J. H., & Smith, D. A. F. (1990). Social support, interdependence,
and the dilemmas of helping. In B. R. Sarason, I. G. Sarason, & G. R. Pierce
(Eds.), Social support: An interaction view (pp. 129-167). New York: Wiley.

Cutrona, C. E., & Russell, D. (1990). Type of social support and specific stress: Toward a
theory of optimal matching. In B. R. Sarason, I. G. Sarason, & G. R. Pierce
(Eds.), Social support: An interaction view (pp. 319-66). New York: Wiley.

Dahlquist, L. M., Czyzewski, D. I., & Jones, C. L. (1996). Parents of children with
cancer: A longitudinal study of emotional distress, coping style, and marital
adjustment two and twenty months after diagnosis. Journal of Pediatric
Psychology, 21, 541-554.

Defares, P. B., Brandjes, M., Nass, C. H. T., & van der Ploeg, J. D. (1985). Coping styles,
social support, and sex differences. In I. G. Saraon & B. R. Sarason (Eds.), Social
support: Theory research and applications (pp. 7-35). Boston: Martinus Nijhoff
Publishers.

Farrell, M. P., & Barnes, G. M. (1993). Family systems and social support: A test of the
effects of cohesion and adaptability on the functioning of parents and adolescents.
Journal of Marriage and the Family, 55, 119-132.

Fiscella, K., Franks, P., Doescher, M. P., & Saver, B. G. (2002). Disparities in health care
by race, ethnicity, and language among the insured. Medical Care, 40, 52-59.

Fitzpatrick, J., Feng, D., & Crawford, D. (2003). A contextual model analysis of
women’s social competence, affective characteristics, and satisfaction in
premarital relationships. Journal of Family Communication, 3, 107-122.

Fragoso, J. M., & Kashubeck, S. (2000). Machismo, gender role conflict, and mental
health in Mexican American men. Psychology of Men and Masculinity, 1, 87-97.

Good, G. E., & Wood, P. K. (1995). Male gender role conflict, depression, and help
seeking: Do college men face double jeopardy? Journal of Counseling &
Development, 74, 70-75.

Greenberg, H. S., & Meadows, A. T. (1992). Psychological impact of cancer survival on
school-age children and their parents. Journal of Psychosocial Oncology, 9, 43-
56.

Hamovitch, M. B. (1964). The parent of the fatally ill child. Los Angeles: Delmar
Publishing Company, Inc.

83
Hoeskstra-Weebers, J. E. H. M., Jaspers, J. P. C., Kamps, W. A., & Klip, E. C. (1998).
Gender differences in psychological adaptation and coping in parents of pediatric
cancer patients. Psychooncology, 7, 26-36.

Hoekstra-Weebers, J. E. H. M., Jaspers, J. P. C., Kamps, W. A., & Klip, E. C. (1999).
Risk factors for psychological maladjustment of parents with children with
cancer. Journal of the American Academy of Child and Adolescent Psychiatry,
38, 1526-35.

Hoeskstra-Weebers, J. E. H. M., Jaspers, J. P. C., & Kamps, W. A. (2000). Factors
contributing to the psychological adjustment of parents of paediatric cancer
patients. In L. Baider, C. L. Cooper, & A. Kaplan De-Nour (Eds.), Cancer and the
family, 2
nd
Edition, (pp. 73-92). Chichester, UK: John Wiley & Sons.

Hoekstra-Weebers, J. E. H. M., Jaspers, J. P. C., Kamps, W. A., & Klip, E. C. (2001).
Psychological adaptation and social support of parents of pediatric cancer
patients: A prospective longitudinal study. Journal of Pediatric Psychology, 26,
225-235.

Hoekstra-Weebers, J. E. H. M., Josette, E. H. M., Heuvel, F., Jaspers, J. P. C., Kamps, W.
A., & Klip, E. C. (1998). Brief report: An intervention program for parents of
pediatric cancer patients: A randomized controlled trial. Journal of Pediatric
Psychology, 23, 207-214.

Jöreskog, K., & Sörbom, D. (1996). LISREL 8: User’s reference guide. Lincolnwood, IL:
Scientific Software International.

Kalnins, I. V., Churchill, M. P., & Terry, G. E. (1980). Concurrent stress in families with
a leukemic child. Journal of Pediatric Psychology, 5, 81-92.

Kazak, A. E. (1989). Families of chronically ill children: A systems and social-ecological
model of adaptation and challenge. Journal of Consulting and Clinical
Psychology, 57, 25-30.

Kazak, A. E., Barakat, L. P., Meeske, K., Christakis, D., Meadows, A., Casey, R., et al.
(1997). Post-traumatic stress, family functioning, and social support in survivors
of childhood leukemia and their mothers and fathers. Journal of Consulting and
Clinical Psychology, 14, 175-191.

Kenny, D. A., Kashy, D. A., & Cook, W. L. (2006). Dyadic data analysis. New York:
The Guilford Press.

Koesten, J. (2004). Family communication patterns, sex of subject, and communication
competence. Communication Monographs, 71, 226-244.

84
Kouneski, E. F., & Olson, D. H. (2004). A practical look at intimacy: ENRICH couple
typology. In D. J. Mashek, & A. Aron (Eds.), Handbook of closeness and
intimacy (pp. 117-133). Mahwah, NJ: Lawrence Erlbaum Associates.

Kunkel, A. W., & Burleson, B. R. (1998). Social support and the emotional lives of men
and women: An assessment of the different cultures perspective. In D. Canary, &
K. Dindia (Eds.), Sex differences and similarities in communication (pp. 101-
125). Mahwah, NJ: Erlbaum.

Kupst, M. J. (1993). Family coping: Supportive and obstructive factor. Cancer, 71, 3337-
3341.

Kupst, M. J., Natta, M. B., Richardson, C. C., & Schulman, J. L. (1995). Family coping
with pediatric leukemia: Ten years after treatment. Journal of Pediatric
Psychology, 20, 601-617

Kupst, M. J., & Schulman, J. L. (1988). Long term coping with pediatric leukemia: A six-
year follow-up study. Journal of Pediatric Psychology, 13, 7-22.

Lakey, B. & Cohen, S. (2000). Social support theory and measurement. In S. Cohen, L.
G. Underwood, & B. H. Gottlieb (Eds.), Social support measurement and
intervention: A guide for health and social scientists (pp. 29-52). Oxford:
University Press.

Lara, M., Gamboa, C., Kahramanian, M. I., Morales, L. S., & Bautista, D. E. H. (2005).
Acculturation and Latino health in the United States: A review of literature and its
sociopolitical context. Annual Review of Public Health, 26, 367-397.

Lavee, Y., & Mey-Dan, M. (2003). Patterns of change in marital relationships among
parents of children with cancer. Health & Social Work, 28, 255-263.

Lyons, R., Langille, L., & Duck, S. (2006). Difficult relationships and relationship
difficulties: Relationship adaptation and chronic health problems. In D. C.
Kirkpatrick, S. Duck, & M. K. Foley (Eds.), Relating difficulty: The process of
constructing and managing difficult interaction (pp. 203-224). Mahwah, NJ:
Lawrence Erlbaum Associates.

Marin, G., Sabogal, F., Marin, B. V., Otero-Sabogal, R., & Perez-Stable, E. (1987).
Development of a short acculturation scale for Hispanics. Hispanic Journal of
Behavioral Sciences, 9, 183-205.

Marteau, T. M., & Bekker, H. (1992). The development of a six-item short-form of the
state scale of the Spielberger State Anxiety Inventory (STAI). British Journal of
Clinical Psychology, 31, 301-306.

85
Matta, D. S., & Knudson-Martin, C. (2006). Father responsivity: Couple processes and
the coconstruction of fatherhood. Family Processes, 45, 19-37.

McNeil, T. (2007). Fathers of children with a chronic health condition: Beyond gender
stereotypes. Men and Masculinities, 9, 409-424.

Mintz, R. D., & Mahalik, J. R. (1996). Sex role ideology and gender role conflict as
predictors of family roles for men. Sex Roles, 34, 805-821.

Morrow, G. R., Carpenter, P. J., & Hoagland, A. C. (1984). The role of social support in
parental adjustment to pediatric cancer. Journal of Pediatric Psychology, 9, 317-
329.

Murphy, S. T. (1994). A mile away a world apart: The impact of independent and
interdependent views of the self on health care decision-making. In J. Power, &
T. Byrd (Eds.), Health care communication on the US/Mexico border (pp. 141-
170). Newbury Park, CA: Sage Publications.

Murphy, S. T., Palmer, J. M., Azen, S., Frank, G., Michel, V., & Blackhall, L. J. (1996).
Ethnicity and advance care directives. Journal of Law, Medicine, & Ethics, 24,
108-117.

National Cancer Institute (2004). Young people with cancer: A handbook for parents.
Retrieved on Sept. 22 from www.cancer.gov

O'Neil, J. M., Good, G. E., & Holmes, S. (1995). Fifteen years of theory and research on
men's gender role conflict: New paradigms for empirical research. In R. Levant &
W. Pollack (Eds.), A new psychology of men (pp. 164-206). New York: Basic
Books.

Ostroff, J., Ross, S., & Steinglss, P. (2000). Psychological adaptation following
treatment: A family systems perspective on childhood cancer survivorship. In L.
Baider, C. L. Cooper, & A. Kaplan De-Nour (Eds.), Cancer and the family, 2
nd

Edition, (pp. 155-173). Chichester, UK: John Wiley & Sons.

Reinhardt, J. P., Boerner, K., & Horowitz, A. (2006). Good to have but not to use:
Differential impact of perceived and received support on well-being. Journal of
Social and Personal Relationships, 23, 117-129.

Reis, H. T., Clark, M. S., & Holmes, J. G. (2004). Perceived partner responsiveness as an
organizing construct in the study of intimacy and closeness. In D. J. Mashek, &
A. Aron (Eds.), Handbook of closeness and intimacy (pp. 201-225). Mahwah, NJ:
Lawrence Erlbaum Associates.

86
Roberts, J. K. (2004). An introductory primer on multilevel and hierarchical linear
modeling. Learning Disabilities: A Contemporary Journal, 2, 30-38.

Robertson, J. M., & Fitzgerald, L. F. (1992). Overcoming the masculine mystique:
Preferences for alternative forms of assistance among men who avoid counseling.
Journal of Counseling Psychology, 39, 240-246.

Rubin, R. B., Palmgreen, P., & Sypher, H. E. (1994). Communication research measures:
A sourcebook. New York: The Guilford Press.

Sarason, B. R., Sarason, I. G., & Pierce, G. R. (1990). Traditional views of social support
and their impact on assessment. In B. R. Sarason, I. G. Sarason, & G. R. Pierce
(Eds.), Social support: An interaction view (pp. 319-66). New York: Wiley.

Saurer, M. K., & Eisler, R. M. (1990). The role of masculine gender role stress in
expressivity and social support network factors. Sex Roles, 23, 261-271.

Schulz, U., & Schwarzer, R. (2004). Long-term effects of spousal support on coping
with cancer after surgery. Journal of Social and Clinical Psychology, 23, 716-
732.

Schwarzer, R., & Leppin, A. (1989). Social support and health: A meta-analysis.
Psychology & Health, 3, 1-15.

Segrin, C., & Flora, J. (2005). Family Communication. Mahwah, NJ: Lawrence Erlbaum.

Seiffge-Krenke, I. (2002). ‘Come on, say something, dad!’: Communication and coping
in fathers of diabetic adolescents. Journal of Pediatric Psychology, 27, 439-
450.

Sharpe, M. J., Heppner, P. P., & Dixon, W. A. (1995). Gender role conflict,
instrumentality, expressiveness, and well-being in adult men. Sex Roles, 33, 1-18.

Sloper, P. (2000). Predictors of distress in parents of children with cancer: A prospective
study. Journal of Pediatric Psychology, 25, 79-91.

Speechly, K. N., & Noh, S. (1992). Surviving childhood cancer, social support, and
parents’ psychological adjustment. Journal of Pediatric Psychology, 17, 15-31.

Street, R. L. (2003). Communication in medical encounters: An ecological perspective. In
T. L. Thompson, A. M. Dorsey, K. I. Miller, & R. Parrott, Handbook of health
communication (pp. 63-89). Mahwah, NJ: Lawrence Erlbaum Associates.

87
Streisand, R., Braniecki, S., Tercyak, K. P., & Kazak, A. E. (2001). Childhood illness-
related parenting stress: The pediatric inventory for parents. Journal of Pediatric
Psychology, 26, 155-162.

Thoits, P. A. (1995). Stress, coping, and social support processes: Where are we? What
next? Journal of Health and Social Behavior, (extra issue), 53-79.

Thompson, E. H., Pleck, J. H., & Ferrera, D. L. (1992). Men and masculinities: Scales for
masculinity ideology and masculinity-related constructs. Sex Roles, 27, 573-607.

Torres, J. B., Solberg, S. H., & Carlstrom, M. S. (2002). The myths of sameness among
Latino men and their machismo. American Journal of Orthopsychiatry, 72, 163-
181.

Varni, J. W., Sahler, O. J., Katz, E. R., Mulhern, R. K., Copeland, D. R., Noll, R. B., et
al. (1999). Maternal problem-solving therapy in pediatric cancer. Journal of
Psychosocial Oncology, 16, 41-71.

Vega, W. A., & Amaro, H. (1994). Latino outlook: Good health, uncertain prognosis.
Annual Review of Public Health, 15, 39-67.

Walker, D. F., Tokar, D. M., & Fischer, A. R. (2000). What are the eight popular
masculinity-related instruments measuring? Underlying dimensions and their
relations to sociosexuality. Psychology of Men and Masculinity, 1, 98-108.

Wasserman, S., & Faust, K. (1994). Social network analysis: Methods and applications.
New York, NY: Cambridge University Press.

Watts, A. S., & Crimmins, E. M. (in press). At risk populations: Elderly. In
Encyclopedia of Aging. Elsevier.

Wiemann, J. M. (1977). Explication and test of a model of communicative competence.
Human Communication Research, 3, 195-213.

Wellman, B. (1992). Men in networks: Private communities, domestic friendships. In P.
M. Nardi (Ed.), Men’s friendships (pp. 75-114). Newbury Park: Sage
Publications.

Wellman, B., & Hiscott, R. (1985). From social support to social network. In I. G.
Saraon, & B. R. Sarason (Eds.), Social support: Theory research and applications
(pp. 36-58). Boston: Martinus Nijhoff Publishers.

88
Wills, T. A., & Shinar, O. (2000). Measuring perceived and received social support. In S.
Cohen, L. G. Underwood, & B. H. Gottlieb (Eds.), Social support measurement
and intervention: A guide for health and social scientists (pp. 86-112). Oxford:
University Press.

Wortman, C. B., & Lehman, D. R. (1985). Reactions to victims of life crises: Support
attempts that fail. In I. G. Saraon, & B. R. Sarason (Eds.), Social support: Theory
research and applications (pp. 123-156). Boston: Martinus Nijhoff Publishers.

Abstract (if available)

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Social media and health: social support and social capital on pregnancy-related social networking sites

PDF

Benefits of social networking in online social support groups

PDF

Relationship formation and information sharing to promote risky health behavior on social media

PDF

The social groups approach to quitting smoking: An examination of smoking cessation online and offline through the influence of social norms, social identification, social capital and social support

PDF

Building personal wellness communities: meaningful play in the everyday life of a network society

PDF

From conversation to conversion: children's efforts to translate their immigrant families' social networks into community connections

PDF

Social support in Cambodia: the role of peer educators in behavior change

PDF

Rediscovering the communication engine of neighborhood effects: how the interaction of residents and community institutions impacts health literacy and how it can be leveraged to improve health c...

PDF

The effect of Twitter social support on health outcomes and its mediators: a randomized controlled trial of a social support intervention on Twitter for patients affected by cancer

PDF

Social support, constraints, and protective buffering in prostate cancer patients and their partners

PDF

The formation and influence of online health social networks on social support, self-tracking behavior and weight loss outcomes

PDF

Creating community online: the effects of online social networking communities on college students' experiences

PDF

An examination of the association between spousal support and type 2 diabetes self-management

PDF

Social capital and community philanthropy: the impact of social trust and social networks on individual charitable behavior and community foundation development

PDF

Mobile self-tracking for health: validating predictors, effects, mediator, moderator, and social influence

PDF

The effects of a COVID-19 narrative message on beliefs and intentions among vaccine hesitant parents

PDF

The link between maternal depression and adolescent daughters' risk behavior: the mediating and moderating role of family

PDF

Children's maltreatment disclosures, recantations, and supporting evidence of maltreatment

PDF

Understanding normative influence of neighborhoods: a multilevel approach to promoting Latinas’ cervical cancer prevention behaviors in urban ethnic communities

PDF

Community integration of individuals with serious mental illness: a network perspective from India and United States

Asset Metadata

Creator Hall, Jeffrey Adam (author)

Core Title Communication and social support of parents of children treated at Childrens Hospital Los Angeles

School Annenberg School for Communication

Degree Doctor of Philosophy

Degree Program Communication

Publication Date 10/24/2007

Defense Date 10/04/2007

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag family communication,masculinity,OAI-PMH Harvest,pediatric oncology,social networks,social support

Place Name California (states), Los Angeles (counties), USA (countries)

Language English

Advisor Cody, Michael J. (committee chair), Messner, Michael A. (committee member), Murphy, Sheila (committee member)

Creator Email hallj@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-m885

Unique identifier UC1433108

Identifier etd-Hall-20071024 (filename),usctheses-m40 (legacy collection record id),usctheses-c127-557581 (legacy record id),usctheses-m885 (legacy record id)

Legacy Identifier etd-Hall-20071024.pdf

Dmrecord 557581

Document Type Dissertation

Rights Hall, Jeffrey Adam

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Repository Name Libraries, University of Southern California

Repository Location Los Angeles, California

Repository Email cisadmin@lib.usc.edu