Citation
Family Environment Types and Their Association with Family Support Satisfaction among African American and Caucasian American Women with Breast Cancer

Material Information

Title:
Family Environment Types and Their Association with Family Support Satisfaction among African American and Caucasian American Women with Breast Cancer
Creator:
MIRSU-PAUN, ANCA
Copyright Date:
2008

Subjects

Subjects / Keywords:
African Americans ( jstor )
Breast cancer ( jstor )
Emotional support ( jstor )
Ethnic groups ( jstor )
Families ( jstor )
Family members ( jstor )
Questionnaires ( jstor )
Research studies ( jstor )
Women ( jstor )
Womens studies ( jstor )
City of Gainesville ( local )

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Copyright Anca Mirsu-Paun. Permission granted to the University of Florida to digitize, archive and distribute this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Embargo Date:
2/28/2005
Resource Identifier:
436098577 ( OCLC )

Downloads

This item is only available as the following downloads:


Full Text

PAGE 1

FAMILY ENVIRONMENT TY PES AND THEIR ASSOCI ATION WITH FAMILY SUPPORT SATISFACTION AMONG AF RICAN AMERICAN AND CAUCASIAN AMERICAN WOMEN WITH BREAST CANCER By ANCA MIRSU-PAUN A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE UNIVERSITY OF FLORIDA 2004

PAGE 2

Copyright 2004 by Anca Mirsu-Paun

PAGE 3

ACKNOWLEDGMENTS Special thanks go to my advisor, Dr. Carolyn M. Tucker, who plays a major role in my development as a scholar. She represents a model of excellence that inspires and motivates me along the graduate school pathway. She always had an ear for me and supported me along the way. iii

PAGE 4

TABLE OF CONTENTS page ACKNOWLEDGMENTS.................................................................................................iii LIST OF TABLES.............................................................................................................vi ABSTRACT......................................................................................................................vii CHAPTER 1 INTRODUCTION........................................................................................................1 2 REVIEW OF THE LITERATURE..............................................................................8 Cultural Disparities and Breast Cancer.........................................................................8 Family Support and Adjustment to Breast Cancer.....................................................11 Family Support Perceived Versus Received/Needed Among Women with Breast Cancer...................................................................................................13 Family Support Types and Levels, and Support Satisfaction Among Women with Breast Cancer...........................................................................................14 Helpful Versus Unhelpful Support for Women with Breast Cancer...................17 Family Environment and Family Support Among Women with Breast Cancer........18 Family Systems Theories for Understanding Family Environment and Family Support...................................................................................................................20 Family Stress Related to Coping with Breast Cancer.................................................24 Culture and Family Environment...............................................................................26 Purpose of the Proposed Study...................................................................................29 3 METHODS.................................................................................................................32 Participants.................................................................................................................32 Measures.....................................................................................................................35 Procedures...................................................................................................................39 4 RESULTS...................................................................................................................44 First Research Question..............................................................................................44 Second Research Question.........................................................................................48 Third Research Question............................................................................................52 Fourth Research Question...........................................................................................58 iv

PAGE 5

5 DISCUSSION.............................................................................................................61 Findings for Each Research Question.........................................................................61 Limitations of the Research........................................................................................68 Implications for Future Research................................................................................70 Implications for the Field of Counseling Psychology................................................72 Conclusions.................................................................................................................73 APPENDIX A RESEARCH QUESTIONNAIRE..............................................................................77 B INVITATION LETTER FOR PACKETS..................................................................85 C INFORMED CONSENT FORM................................................................................87 D SCRIPT FOR CLINIC RECRUITMENT..................................................................89 E SCRIPT FOR SUPPORT GROUP RECRUITMENT...............................................90 F INVITATION LETTER FOR WEB..........................................................................91 G INFORMED CONSENT FORM FOR WEB.............................................................93 LIST OF REFERENCES...................................................................................................95 BIOGRAPHICAL SKETCH...........................................................................................101 v

PAGE 6

LIST OF TABLES Table page 1 Demographic Description of the African American and Caucasian American Participants...............................................................................................................34 2 Descriptive Data for the Subscales of the Family Relationship Inventory (FRI) by Ethnic Group.......................................................................................................45 3 Distribution of Cases for the Three-Cluster and Four-Cluster Solutions from the Cluster Analysis Applied to the FRI Data................................................................46 4 Cohesion, Expressiveness, and Conflict Levels Characteristic of the Three Family Environment Types......................................................................................47 5 Correlation Coefficients for the Associations Among Perceived Informational, Tangible, and Emotional Support by Ethnic Group.................................................54 6 Correlation Coefficients for the Associations Among Satisfaction with Informational, Tangible, and Emotional Support by Ethnic Group.........................54 7 Correlations Between Demographic Variables and the Perceived Support Variables and Self-Reported Support Satisfaction Variables for African American Women....................................................................................................55 8 Correlations Between Demographic Variables and the Perceived Support Variables and Self-Reported Support Satisfaction Variables for Caucasian American Women....................................................................................................56 9 Hierarchical Regression Predicting Support Satisfaction from Family Cohesion, Expressiveness, and Conflict for African American Women Participants..............59 10 Hierarchical Regression Predicting Support Satisfaction from Family Cohesion, Expressiveness, and Conflict for Caucasian American Women Participants..........60 vi

PAGE 7

Abstract of Thesis Presen ted to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Science FAMILY ENVIRONMENT TY PES AND THEIR ASSOCI ATION WITH FAMILY SUPPORT SATISFACTION AMONG AF RICAN AMERICAN AND CAUCASIAN AMERICAN WOMEN WITH BREAST CANCER By Anca Mirsu-Paun August 2004 Chair: Carolyn M. Tucker Major Department: Psychology This study primarily explored types of fa mily environments and the support they offer to African American and Caucasian Amer ican women with breast cancer. Overall, findings from this study suppor t the view that family envi ronment types of women with breast cancer are different, and that there may be an association between family environment type and satisfaction with support or with perceived s upport. Three types of family environments were identified among African American and Caucasian American women with breast cancer: Cohesive-E xpressive, Conflictual-Expressive, and Conflictual-Nonexpressive. Findings from this study indicate that overall the women in Cohesive-Expressive family environments ha d higher education and were more likely to have a spouse/partner present in the hous e compared with women in ConflictualExpressive and Conflictual-None xpressive family environments. These findings suggests that perhaps women with lower educati on levels and women who do not have a

PAGE 8

spouse/partner should be assessed more closely with regard to their family relationships and support needs from their family members. Findings from this study also indicated that perceptions of emotional support were associated with family environment type among African American women, but satisfaction with emotional and tangible support was associated with family environment types for Caucasian American women. Satisfaction with overall family support was predicted by family conflict among African American women participants. The present study provides evidence for the fact that ethnically diverse families should not be reduced to strict categories or types, especially not on the basis of ethnicity of their members. Instead, ethnically diverse families should be understood through the lens of the embededness theory. Specifically, individual families should be viewed in the context of their social environment with careful consideration given to the individual differences among family members. viii

PAGE 9

CHAPTER 1 INTRODUCTION Breast cancer is the leading form of cancer among women in the United States (Holland, 1998). The American Cancer Society reported that in 2002 approximately 203,500 women were diagnosed with invasive breast cancer and 39,600 women died from this disease. However, these incidence and mortality rates vary significantly among different ethnic groups. Caucasian women have the highest incidence of breast cancer, followed by African-American, Asian American, Hispanic American, and American Indian women, respectively (American Cancer Society Facts and Figures, 2002). Concomitantly, African American women have the highest mortality rates, followed by Caucasian, Hispanic American, American Indian, and Asian American women, respectively. Despite the fact that the incidence of breast cancer in the general population declined significantly in recent years and that the breast cancer survival rate is higher than it has ever been in the past, breast cancer is nonetheless considered a life-threatening disease that dramatically affects women’s lives on a long-term basis. Confronting breast cancer involves a continuous series of events that include a main treatment option, adjuvant medical interventions, and periodic medical examinations (Bloom, 2000). In addition, subsequent to the initial emotional shock of receiving a diagnosis of breast cancer, the patient usually experiences long-term uncertainty and fear due to the possible recurrence of cancer once it is successfully treated (Holland, 1998). These persistent emotional sequels of a breast cancer diagnosis, as well as the required long-term medical 1

PAGE 10

2 monitoring that is part of cancer treatment, bring significant changes into the lives of cancer survivors and affect their families and support systems (Bloom, 2000). For example, when a woman undergoes treatment for breast cancer, family member roles within her family usually change, with family members having additional responsibilities as a result of her interruption of the work-related activities. The cancer survivor research literature emphasizes that there are both similarities and differences among women with breast cancer regarding their experiences with the disease and the changes it brings in their lives. The similarities involve difficulties with the treatment for breast cancer and with the associated lifestyle changes, while differences between women with breast cancer typically involve women’s attitudes and beliefs about breast cancer, screening behaviors, and treatment availability and adherence. Overall, women’s experience with breast cancer is personal and unique (Holland, 1998). According to Bloom (1998), the personal and unique experiences of women with breast cancer can be understood only by considering both the composition of their families and their cultural background. Specifically, family relations and cultural background, and their convergent action impact women’s adjustment to breast cancer and its treatment (Spinetta, 1984; Wellisch et al., 1999). According to Hill, Amir, Muers, Connolly, and Round (2003), it is the perceptions that families of women with breast cancer have of this disease that influence women’s interpretation of their experience with breast cancer and its treatment. Moreover, there are cultural group differences in how stressful situations (e.g., the hospitalization of a spouse with breast cancer) are perceived and interpreted (Hill, Amir, Muers, Connolly, & Round, 2003). Being diagnosed with a severe illness such as cancer is generally considered an

PAGE 11

3 extremely stressful event (Hagerdoon et al., 2000); however, there are also culture-specific differences in the perception of breast cancer and its treatment. For example, African American women as compared with Hispanic and Caucasian American women reported more fears and worries as barriers to mammography screening (Friedman et al., 1995). Consequently, healthcare professionals need to carefully consider cultural variations and family involvement in the health care process (Kagawa-Singer & Nguyen, 2000). Families of women with breast cancer are important not only because of their capacity to influence women’s attitudes and belief systems, but also because they are a major source of support for women coping with the challenges that come with a cancer diagnosis. The level of support that family members can provide to a woman diagnosed with breast cancer is associated with their ability to adjust to this diagnosis and ability to cope with its associated stressors, and with the family members’ use of the family support resources that are available. Level of family support perceived has been linked to one’s psychological state and well-being during times of hardship (Rigazio-DiGillio, 2002), and to one’s ability to adjust and cope with the experience of breast cancer (Spinetta, 1984; Wellisch et al., 1999). Specifically, family support for women with breast cancer has been associated with their physical functioning, social adjustment, anxiety and depression levels, and levels of well being (Primomo et al., 1990). Family members are also often tremendously impacted emotionally by a family member’s breast cancer diagnosis (Northhouse, Templin, Mood, & Oberst, 1990; Primono, Yates & Woods, 1990). Thus, a family systems framework seems necessary for

PAGE 12

4 understanding and promoting the support that breast cancer survivors need to successfully cope with the experience of having breast cancer. The family support offered to women confronting breast cancer typically involves complex psychosocial mechanisms. Consequently, the relationship between the actual support offered by family members and the woman’s level of satisfaction with the perceived family support is not always a linear one. Even when family caregivers are well intentioned, their genuine attempts to be helpful might not be perceived as such. This misperception may occur when women experiencing breast cancer need some other type of support than the type being provided, or when these women’s level of distress is so high that it distorts their perception of the support received. Moreover, when women with breast cancer have the impression that family members are not providing the needed or desired support, these women may display negative reactions toward their family members. These negative behavioral responses might trigger an elevated level of distress in family caregivers, who might then believe that their helping efforts are not recognized or valued. As a consequence, the supportive efforts of family members might decrease. It is also the case that family members have their own perceptions regarding the support they should offer and the support they are actually able to offer to a woman family member confronting breast cancer. Furthermore, family members, as expected sources of support, may experience pressure to cope with a new set of obligations (Bloom, 2000), and they may have their own fears and concerns regarding cancer (Krishnasamy, 1996) which reduce their emotional resources for helping their family member with breast cancer with the detection, diagnosis, and treatment of this disease (Bloom, 2000).

PAGE 13

5 In sum, the family systems framework takes into account the family’s environment, as well as the mutual influence of family members’ emotional reactions and psychological states. The notion of “family environment” refers to the overall social-environmental characteristics of a family. Three relationship dimensions are usually employed to describe a family environment: cohesion level, expressiveness level, and conflict level. Family cohesion refers to the degree of help, commitment, and support that family members provide to one another. Family expressiveness is the degree to which family members are encouraged to act openly and to directly express their feelings, either positively or negatively. Family conflict is the amount of openly expressed anger, aggression, and conflict among family members. The present research primarily focuses on “family environment type,” which is a term that the primary investigator is using to describe the particular patterns of family communication, expressiveness, and conflict levels that characterize individual families of women with breast cancer. These family environment types will be determined through cluster analytic procedures. Another important aspect of the present research is its inclusion of minority women and majority women. Traditionally, studies on family variables and cancer survivorship have included participants from a single ethnic group, and most often that group has been White/Caucasian Americans. Indeed, there are increasing calls for including patients from different racial/ethnic backgrounds in studies of people with cancer (e.g., Kagawa-Singer & Nguyen, 2000). According to Andersen and Sabatelli (1999), understanding ethnicity is important and must include both an understanding of the central tendencies of each ethnic group, as well as an understanding of group variations within an ethnic group. An implication of this assertion is that research on culturally different family systems

PAGE 14

6 also involves the recognition of family themes and communication patterns that are characteristic of an ethnic group as well as recognition of the family themes and communication patterns that are characteristic of particular families within ethnic groups, independent of ethnic group membership. Culture and family are usually included in research studies independently from one another (Szapocznik & Kurtines, 1993). However, in reality there seems to be an integration, overlap, and interaction between individuality of family members, family systems, and cultural environment. Evidence for this interaction is provided by the embeddedness theory (Szapocznik & Kurtines, 1993). The embeddedness theory advocates an understanding of individuals in the context of their families, and an understanding of families in the context of their cultural environment (Szapocznik & Kurtines, 1980). Two major assertions of the embeddedness theory are as follows: (a) that different contexts (i.e., individual, family, and culture) are embedded within each other, and that (b) the cultural context is heterogenous and pluralistic and should not be regarded as a uniform variable. The embeddedness theory promotes a flexible view of families and the avoidance of preconceived culture based family models. In contrast, preconceived cultural assumptions of family environments cannot describe the complex reality and the richness of the unique experiences of diverse women with breast cancer. In accordance with the principles of the embeddedness model, the present study assumes the existence of independent types of family environments in a group of African American and Caucasian women. Even though the existence of a possible correlation between ethnicity and family environment will be studied, it will not be assumed that certain family environments are

PAGE 15

7 characteristic of specific ethnic groups. Specifically, the present study will use an exploratory approach for identifying different family environments and examining their links to ethnicity and to family support satisfaction of African American and Caucasian American women with breast cancer. The exploratory nature of the study derives from the fact that it uses empirical data to determine family environment categories (versus a deductive approach).

PAGE 16

CHAPTER 2 REVIEW OF THE LITERATURE Cultural Disparities and Breast Cancer Each year, breast cancer affects a large number of women in the United States. Despite the declining incidence and increased survival rates of breast cancer in the general population, there are still particular groups of women who are disproportionately affected by breast cancer. More specifically, lower survival rates have been reported in association with lower economic resources, older age, and ethnic minority group status (Walker, Figgs, & Zahm, 1995). Ethnic minority status is associated with the nature of healthcare services received and implicitly with cancer survival rates. Evidence of this assertion is the finding that physicians are less likely to recommend screening mammography to minority women than to Caucasian women (Friedman, Webb, Weinberg, Lane, Cooper, & Woodruff, 1995). The ethnic minority group the most affected by cancer are African Americans. This group of Americans has an age-adjusted mortality due to cancer that is 27% higher compared with that of general population (Kang & Bloom, 1995). These statistics suggest a need for research studies to understand African American women’s experience of breast cancer and to understand the support that these women need to successfully cope with this experience. Socio-economic variables are also associated with breast cancer survival rates. For example, family income is associated with the availability of health care services and with the nature of these services (Friedman et al., 1995). Healthcare availability includes 8

PAGE 17

9 but is not limited to transportation to and from one’s medical clinic, or available resources to pay for medical insurance that would cover necessary medical procedures. Availability and appropriateness of health care services usually influence the early detection of breast cancer and women’s adherence with the required medical treatment for this disease. Late detection and reduced treatment adherence in turn interfere with treatment effectiveness and reduce survival time (Holland, 1998). Late detection and poor treatment adherence are also associated with the belief systems and health behaviors of culturally diverse women with breast cancer (Long, 1993). More specifically, research findings provide evidence for the existence of ethnic group differences in perspectives and belief systems associated with breast cancer. For example, among African American and Hispanic women there seems to be a strong “cancer phobia” (Bloom et al., 1987; Perez-Stable, 1991), despite the facts that breast cancer is an increasingly common experience among women in the United States (Epping-Jordan et al., 1999) and the prognosis for this disease is generally favorable (American Cancer Society statistics). Furthermore, research based evidence lead Bloom and Kessler (1994) to affirm that 65% of African American and 30% of Hispanic American women may believe that getting breast cancer is like receiving a death sentence. Although Bloom and Kessler caution that further research is necessary to validate these very high percents, they still assert that levels of information and knowledge about cancer and associated treatments are still disproportionately lower among African American and Hispanic American women in comparison with Caucasian women.

PAGE 18

10 Cultural values, attitudes, and practices affect women’s knowledge and awareness of cancer, their perceptions of vulnerability to developing breast cancer, their causal attributions and perceived personal responsibility regarding having breast cancer as well as their health-seeking behaviors and expectations for cure (Long, 1993). Ethnic group differences regarding breast cancer belief systems and related practices have been found to be associated with women’s readiness to seek treatment immediately after the detection of breast abnormalities (Friedman et al., 1995) and with their willingness to openly talk about their experience of breast cancer (Kang & Bloom, 1993). In a sample of Caucasian American women, openness to talk about one’s experience with breast cancer influenced one’s ability to ask for and receive adequate and helpful support from one’s spouse, family members, and/ others in her/his larger social environment (Manne & Schnoll, 2001). Increasing social support in turn has been shown to be an effective way of increasing the use of healthcare services for breast cancer in a sample of older women with breast cancer (Kang & Bloom, 1993). Clearly, it appears that the overall experience of breast cancer and the difficulties in coping with this disease are different among culturally diverse women. Consequently, it is possible that the needs for support and the actual support received vary among culturally diverse women with breast cancer (see Bloom, 1998). Culturally sensitive research is needed to address this matter. It is important to note that culturally sensitive research requires more than simply recruiting an ethnically diverse sample or simply comparing outcomes and performances across ethnic/racial groups (Dumas et al., 1999; Nagayama Hall, 2001). Cole and Stewart (2001) asserted that comparative research can be made less discriminatory and more culturally sensitive by (a) studying ethnic/racial

PAGE 19

11 categories as lived experiences rather than self evident realities; (b) sampling a wide range of settings and experiences; (c) choosing appropriate control or comparison groups that have been living in similar cultural and environmental conditions, and (d) describing within-group differences (Cole & Stewart, 2001, p. 302). The present study used the culturally sensitive “difference model” research approach (Oyemade & Rosser, 1980) to test the association between satisfaction with different types of perceived family support of women with breast cancer and family environment types among women with breast cancer. The “difference model” approach emphasizes the importance of recognizing cultural differences when investigating academic, cognitive, or social behavior of diverse cultural groups. Specifically, it suggests performing data analyses separately for participants grouped according to their major cultural background (especially ethnicity and socioeconomic status), versus collapsing data from culturally different participants for collective data analyses and statistically examining racial and socioeconomic differences. According to Oyemade and Rosser, this type of research is in contrast to typical traditional “deficit model” research that uses one statistical model to compare groups that are different (e.g., different by race) and then explains group differences in performance using the performance of Caucasian American middle class as the performance standard. In “deficit model” research lower or different performance by minority groups is viewed as deficit performance that often further reinforces negative stereotypes of minority groups. Family Support and Adjustment to Breast Cancer Given the multitude of challenges that come with the diagnosis of breast cancer, the networks of social and family support play important roles in women’s adjustment to, and coping with this disease. Social support has been defined (Hobfall, 1982, p.121) as “those

PAGE 20

12 social interactions or relationships that provide individuals with actual assistance or that embed individuals with a social system believed to provide love, caring, or sense of attachment to a valued social group or dyad”. Some research findings suggest that family support for women with breast cancer represents a specific domain of social support, independent from the global support received from other available sources. The difference between the family and the overall social environment as sources of support for women with breast cancer is related to the nature of the relationships and interactions that take place between the support recipient and other family members. Family support is the focus of the present study due to its importance for women with breast cancer. One common method used by individuals to adapt to cancer is to obtain support from their families (Manne & Schnoll, 2001). The family (and especially one’s spouse) represents the main source of support for women with a chronic illness (Primomo, Yates, & Woods, 1990), and in a study by Primomo et al. (1990) women participants perceived their families to be the main provider of emotional, tangible, and informational support, in comparison with friends and other people. Moreover, findings from Primomo et al. Yates, and Woods (1990) evidenced that family members of women with a chronic illness (such as breast cancer) provided the most effective support in comparison with friends and others. The reactions of family members have been shown to have an overall significant impact on how women cope with and adapt to their breast cancer diagnosis and treatment (Manne at al., 1997), and the presence versus absence of a partner has been identified as predictive of the emotional distress often experienced by women with breast cancer (Ford et al., 1995).

PAGE 21

13 Although a unitary concept, family support includes multiple dimensions. Accordingly, some assessment instruments measure perceived support, while others measure actual support being offered (Manne & Schnoll, 2001), and still some others focus on support types or support levels. The present study investigated the level of need for versus the perceived occurrence of three common types of family support: emotional support, tangible support, and informational support. Family Support Perceived Versus Received/Needed Among Women with Breast Cancer The distinction between perceived support versus received support is an important one, since the perception that women with breast cancer have of the support they receive does not always fully coincide with the actual support received. Support received represents the amount of supportive and helpful actions that are performed in reality by family members or members of the social network, while perceived support has been defined as a general belief that particular types of support are available and/or would be available if one would need them (Bloom & Kessler, 1994). Research findings indicate that, for women with breast cancer, the association between perception of support and satisfaction with support is stronger in comparison with the association between actual support being offered and satisfaction with support (Krishnasamy, 1996). Satisfaction with support is an important variable, as it is associated with psychological well being and positive health outcomes. For example, satisfaction with family support during hospitalization and three month after surgery has been found to be significantly associated with levels of anxiety and depression in a sample of women with breast cancer (Neuling & Winefield, 1998).

PAGE 22

14 The above research findings provide evidence for the differentiation between actual and perceived support for women with breast cancer. More importantly, this research suggests that satisfaction with family support among women with breast cancer is an important research topic. Thus, family support satisfaction and factors that influence it was a major focus of the research being presented. Family Support Types and Levels, and Support Satisfaction Among Women with Breast Cancer Support levels refer to the quantity and frequency of supportive actions. For women with breast cancer, their perceptions of levels of family support are associated with their level of satisfaction with support received (see Gurowka & Lightman, 1995). The stronger the correspondence between the level of support perceived and the level of support needed, the higher is the level of satisfaction with support (Hagedoorn et al., 2000). Moreover, some authors agree that there is no linear association between levels of family support and satisfaction with received support of women with breast cancer. Specifically, whereas limited support may not be enough to meet the needs of women with breast cancer, too much support (or overprotection) might undermine women’s self-efficacy in dealing with breast cancer (Coyne et al., 1990). Consequently, both too low and too high levels of support have been found to be associated with decreased levels of women’s satisfaction with support (Hagedoorn et al., 2000). However, it is important to note that women confronting breast cancer might not define in a similar manner what “too much” or “too little” support means for them. Their needs for support might indeed be different, as well as their perceptions of support received. For example, according to Shankar (1997), the perception of support received from family members is different among women with breast cancer from different ethnic backgrounds. Ethnic differences

PAGE 23

15 in perception of support availability have also been found among Caucasian and African American women with breast cancer (Bourjolly & Hirschman, 2001). Specifically, African American women, in comparison with Caucasian women, perceived their husbands as being less supportive but their friends as being more supportive. Support types refer to the different types of actions performed with the goal of helping. An exploratory factor analysis performed by Manne and Schnoll (2001) revealed the following types of support: emotional, instrumental, cognitive information and guidance, encouraging distancing and self-restraint, and criticism and withdrawal. It is noteworthy that one could be satisfied with one type of support (e.g., tangible support) received from their family members, but not be fully satisfied with another type of support (e.g., emotional support) received from their family members. The present study examined the distinct levels of satisfaction with informational, tangible, and emotional support in a sample of women with breast cancer. Emotional, informational, and tangible types of support are the most cited in the research literature. Emotional support. This type of support seems to be the most commonly studied form of social support for women confronting breast cancer. Research findings indicate that emotional support is especially helpful when it is offered by family members or close friends (Neuling & Winefield, 1988). Expression of positive affect and provision of information that one is cared for, loved, esteemed, or understood (Gurowka & Lightman, 1995) have been consistently evaluated as helpful by women with breast cancer. Moreover, expressing agreement with, acknowledging a person’s feelings, and encouraging the open expression of beliefs and feelings have been found to be useful types of family emotional support for women with breast cancer (Gurowka & Lightman,

PAGE 24

16 1995). Feelings of closeness, intimate interactions, esteem building, comfort, and encouragement (Stokes & Wilson, 1984), as well as listening and reflecting understanding (Krishnasamy, 1996) have also been found to be particularly helpful for women with breast cancer. Greater levels of emotional support from family members have been associated with positive outcomes such as enhanced well-being (Bloom & Spiegel, 1984) and decreased anxiety and depression levels (Funch & Mettlin, 1982). Lower levels of depression in women who receive more emotional support from their partners and families have also been found by Primomo, Yates, and Woods (1990). Emotional support tends to be offered by family members to a greater extent in comparison with other sources of social support. However, in a study by Neuling and Winefield (1998) women with breast cancer expressed their need for more emotional/empathic support from their family members, which suggests that sometimes the emotional support provided by family members to women with breast cancer is not sufficient to meet their needs. Guidance or informational support. This type of support refers to the provision of advice, information, and feedback (Stokes & Wilson, 1984), and problem-solving skills (House, 1981). Although sometimes considered to be auxiliary to emotional support, guidance or informational support from family members generated the greatest amount of satisfaction in a sample of women with breast cancer at three months following surgery (Neuling & Winefield, 1998). Tangible support/practical assistance. Tangible support, or instrumental support, is usually understood as the provision of material aid, money, goods, or services (Primomo, Yates, & Woods, 1990). Examples of tangible support include transporting

PAGE 25

17 the patient to or from the hospital, performing household tasks (cleaning, cooking, etc.), or offering useful and tangible goods and gifts. According to Gurowka and Lightman (1995), the provision of practical help or assistance is the most helpful behavior within the category of problem-solving behaviors. Moreover, research findings indicate that higher levels of tangible support (or material assistance) from family members are related to improved physical recovery (Funch & Mettin, 1982) and increased self-esteem (Dunkel-Schetter, Feinstein, & Taylor, 1992) in women with breast cancer. Helpful Versus Unhelpful Support for Women with Breast Cancer Research findings suggest that women with breast cancer perceive specific supportive behaviors as more helpful than others. For example, among behaviors of family members generally evaluated by women with breast cancer as supportive are expression of concern, love, and understanding, practical assistance (Gurowka & Lightman, 1995), and being available to help when necessary (Bloom & Kessler, 1994). Some behaviors evaluated as unsupportive are avoidance, repeatedly inquiring about one’s history of cancer, not providing the right amount and type of support (Gurowka & Lightman, 1995), giving unsolicited advice, encouraging recovery, and “positive thinking”, minimization and forced cheerfulness, and naming another person’s feelings (Bloom & Kessler, 1994). Women who confront breast cancer have particular needs for support from family members versus friends, colleagues, or professionals (e.g., their doctor, priest, nurse, counselor, etc.). However, the assumption that women with breast cancer need all the support that can be offered to them and that they particularly need a specific type of support may be sometimes erroneous and may foster feelings of dependency and inadequacy among these women (Dunkel-Schetter, 1992). Not all dimensions of social

PAGE 26

18 support are perceived as helpful to the same extent, and not always does more support automatically imply more help to the woman confronting breast cancer (Gurowka & Lightman, 1995). The same supportive action may be perceived as supportive or unsupportive, depending upon the source it is coming from and upon the specificity of the situation (Gurowka & Lightman, 1995; Krishnasamy, 1996). Moreover, the needs for support of women with breast cancer may vary according to the source of supportive actions, timing along the course of the illness, or ethnic background. For example, Caucasian women with breast cancer evaluated as more helpful the information and advice received from healthcare professionals than information and advice from family members or friends (Dunkel-Schetter, 1984). In a study by Neuling and Winefield (1988), women with breast cancer expressed their need for different types of support before, immediately after, and three month after the surgery. Women’s needs for empathic support and reassurance from family members decreased at three months after surgery, while their needs for informational and tangible support from family members slightly increased in the first month after surgery but ultimately followed a descending slope. Family Environment and Family Support Among Women with Breast Cancer The term family environment usually refers to the different modalities of interaction between family members, and more specifically to levels of family communication, expressiveness, and conflict that exist among family members. The present study focused on family environment types among women with breast cancer. Family environment type is being defined as the particular pattern of family cohesion, expressiveness, and conflict levels reported by research participants as revealed via cluster analytic techniques.

PAGE 27

19 Family cohesion refers to the degree of commitment, help, and support family members provide to one another (Moos & Moos, 1986). Increased family cohesion is associated with increased support provided to women with breast cancer by family members (Stokes & Grimard, 1984). Moreover, when confronted with stressful situations, more cohesive families display increased functional levels in comparison with less cohesive families. Women with breast cancer who perceived their families to be highly cohesive reported lower levels of psychological distress than women from less cohesive families (Bloom, 1982). It is noteworthy that there are ethnic group differences in family cohesion (Anderson & Sabatelli, 1999). Consequently, support types and levels offered to ethnically diverse women with breast cancer might be different and these differences might be associated with family cohesion. A second dimension of family environment, family expressiveness, represents the extent to which family members are encouraged to act openly and to express their positive or negative feelings directly (Moos & Moos, 1986). In contrast, the dimension of family conflict is described by Moos and Moos (1986) as openly fighting and becoming angry with one’s family members. More specifically, family conflict represents the amount of overt anger and aggression among family members. A cluster analytic study of family environment types among women with breast cancer that was conducted by Kissane et al. (1988) provides evidence for the existence of patterns of higher conflict in sullen and hostile families. The types of family environments described by these researchers suggest an association between the dimensions of family conflict, family cohesion, and family expressiveness. More specifically, these researchers found that sullen families were characterized by both

PAGE 28

20 moderate expressiveness and cohesion as well as moderate conflict and hostility; additionally families characterized by low levels of cohesion and expressiveness between members also evidenced high levels of conflict. It has also been found that among the families of women with breast cancer, levels of family conflict are associated with lower levels of support offered by family members to the women with breast cancer (Kissane et al., 1988). An important conceptual difference should be noted between the notion of family environment with its three dimensions (family cohesiveness, expressiveness, and conflict), and the notion of family identity. While family environment refers to behavioral aspects of the socio-relational climate of a family, a family’s identity emerges from the spoken and unspoken rules of relationship that guide family members in how they relate to each other (Patterson & Garwick, 1994). The rules governing the family life refer to the definition of family boundaries and membership, the assignment of roles for accomplishing family tasks, and rules and norms for family interaction (Patterson & Garwick, 1994). Family Systems Theories for Understanding Family Environment and Family Support Family systems theories emerged from general systems paradigms formulated in the 60’s. The fundamental notion of family theories is that there is a relationship and interdependency among subsystem components (Bertalanffy, 1968). The systems perspective proved very useful when applied to families (Patterson & Garwick, 1994) and has been adopted by the fields of psychology and psychotherapy. According to the family systems theory set forth by Bertalanffy (1968), the individual, dyadic, and family levels are simultaneously distinct and interdependent, and they are related in a dynamic and

PAGE 29

21 mutual manner (Bertalanffy, 1968). As Pinsof (1992, p. 436) points out, a family system is “a set of people related by blood or intention who mutually affect each other.” Each of the family subsystems (i.e., individual family members or subgroups of family members) maintains boundaries that are permeable to some extent to other subsystems and to the larger social environment. Boundary ambiguity becomes a major problem for families confronting a chronic disease such as breast cancer (Patterson & Garwick, 1994), especially because functional families have clearly defined boundaries (Minuchin, 1974). More precisely, families function on a continuum from rigid boundaries on one end, to clear boundaries in the middle, and to diffuse boundaries at the other end (Minuchin, 1974). Members of families with diffuse boundaries are often disengaged and isolated from one another. Thus, these families are minimally cohesive and the communication between their members is minimal. Members of these families may rely on external sources of support outside of the family (Becvar & Becvar, 1996, in Navarre, 1998). In contrast, members of families with rigid boundaries may experience a loss of independence and autonomy, and they may be overly involved and responsive under stressful situations. Moreover, members of families with rigid boundaries may insist on preserving the family’s structure, which might hinder a flexible adaptation of the family and a positive usage of family’s resources (Minuchin, 1974). Consequently, families with rigid boundaries might be overly supportive and they might offer support that in fact is not needed. Another assumption of family systems theories is that each system is composed of smaller component subsystems. The family system can be composed, for example, by individual, marital relationship, parent-child, and sibling subsystems. Each of these

PAGE 30

22 subsystems in the family system has some elements that are unique and some that are common with other family subsystems (Bertalanffy, 1968). Family systems theories focus on the relationship between family members and, most specifically, on patterns of interaction that occur with regularity between members of a particular family. Understanding the way a family system operates can provide a clearer understanding of the way individual members think, feel, and behave under specific circumstances (Anderson & Sabatelli, 1999). An implication of family systems theories is that different families will react differently to and will make use in different ways of the resources they possess when confronted with stressful situations. There are several ways in which family systems theories are relevant to the issues of family coping and support in response to breast cancer. Specifically, it has been suggested that in response to a chronic illness, the family system experiences boundary ambiguity (Navarre, 1998), which in turn is associated with disengagement of the family members and with support seeking from external sources. Additionally, the family tends to reorganize itself and mobilize resources in response to a crisis situation such as breast cancer (Patterson & Garwick, 1994). Reorganizing the family structure results in new roles for the members of the family system. For example, the interruptions of work-related/professional activities that often comes with a diagnosis of breast cancer may result in family members having new requirements and roles to accommodate this interruption and to meet the demands that come with the diagnosis of breast cancer (Bloom & Kessler, 1994). According to family stress theories, this added set of demands that comes with the diagnosis of breast cancer is defined as “pileup demands” (Patterson & Garwick, 1994). The notion of “pileup” often involves an increased appeal to family

PAGE 31

23 resources to meet the demands of the current situation. In well functioning families, family cohesion is an important resource for dealing with such stressful situations (Patterson & Garwick, 1994). In accordance with family systems theories, findings from Bloom & Kessler (1994) support the conclusion that interaction patterns between women confronting breast cancer and their family members may influence the level of support that these women receive from their families. Using an inferential research design, Bloom and Kessler found that levels of emotional support received by women with breast cancer from family members are the result of their interaction with these family members. For women with breast cancer, improvements in their interactions and social involvement with family members and in their role functioning in daily activities have been related to an increase in their perceived emotional support (Bloom & Kessler, 1994). The notion of “fit” between family resources and external demands (Hobfoll & Vaux, 2003) explains, from a family systems perspective, how families usually adjust to breast cancer. According to Hobfoll and Vaux (2003), individual members of a family adapt (fit) their resources to meet the demands they encounter. It is not the family’s resources per se, but the family members’ ability to draw on these resources that promotes coping with stressful events. Consequently, individuals from more adaptable families will be able to “mold” their resources to meet the challenges that come with a breast cancer diagnosis. More importantly, families who are able to successfully adjust to stressful life events are often described in the literature as “cohesive, supportive, and able to positively communicate” (Hobfoll & Vaux, 2003).

PAGE 32

24 Family Stress Related to Coping with Breast Cancer Family stress has been defined as a state of tension that arises in the family functioning from an actual or perceived imbalance between demands and the family’s capabilities for meeting or coping with these demands (McCubbin, 1988). According to Olson et al. (1983), an important component of a family’s response to stressful life events is the family’s perception and appraisal of the event, which is usually influenced by social and cultural norms (McCubbin, 1988). A stressor such as a chronic illness is accompanied by the employment of specific coping strategies and/or substantial changes in the family system (Patterson & Garwick, 1994). Changes within the family system become possible due to the fact that, in general, families are capable of managing their available resources. Moreover, families are better able to cope with stressful situations as a result of structural re-arrangements and reorganization within the family and use of the external social environment (Carey et al., 1991). In general, families function better in their attempt to cope with stress when their members have common goals and tasks, share a sense of family history, experience emotional bonding with one another, develop strategies for meeting the needs of both individual members and the global family system, and maintain flexible but firm boundaries (Anderson & Sabatelli, 1999). As previously indicated, one assumption of family systems theory is that different systems within the family affect each other mutually and this mutual interaction becomes most salient in times of hardship, crisis, or stressful situations (Pinsof, 1994). Consequently, the level of distress in women with breast cancer and their partners or family members may clearly be related not only to the disease itself, but also to the family interactions that take place in the context of their illness (Baider et al., 1998). The importance of family interactions in the context of breast cancer is emphasized by

PAGE 33

25 research findings indicating that poor family dynamics are related to high levels of distress, decreased well being, and avoidance coping by survivors of breast cancer (Clark, 1993; deRuiter et al., 1993; Manne et al., 1997; Tylor, 1996). Furthermore, a high level of distress experienced by women with breast cancer may impact not only their psychological state (e.g., their level of depression, or their well-being) but also such health-related factors as treatment compliance, survival rate, and survival time (Hagedoon et al., 2000; Kissane et al., 1998). As a consequence of understanding the importance of the family in the context of breast cancer, increasing numbers of studies shifted from identifying problems in families dealing with a chronic illness to identifying factors associated with positive outcomes. For example, according to the Family Adjustment and Adaptation Response (FAAR) Model (Patterson, 1988), family resources used to cope with the demands of a chronic illness are manifested in family organization, members’ ability to communicate effectively, flexibility of family interactions, and cohesion between family members. According to Patterson and Garwick (1994), it is easier for an individual with cancer to ask for and receive needed support from family members when he or she perceives the family environment as cohesive (Holahan & Moos, 1981). Spiegel, Bloom, and Gottheil (1983) indicated that communication and mutual understanding between family members are related to a family’s capacity to provide support and that openness between family members can increase women’s easiness of expressing their feelings, which leads to a sense of family cohesiveness. A more expressive and less conflict-prone family environment has been found to be predictive of psychological adjustment of women with metastatic breast cancer (Spiegel, Bloom, & Gottheil, 1983).

PAGE 34

26 Culture and Family Environment The term “culture” has been defined as “an integrated pattern of human behavior that includes thoughts, communication, actions, customs, beliefs, values, and institutions of a racial, ethnic, religious, or social group” (Cross, Bazron, & Isaacs, 1989, in Sue et al., 1998). These patterns may be explicit or implicit and are transmitted through a socialization process. According to Rigazio-DiGillio (2003, p. 400), “culture permeates the way we come to see ourselves and our relationships, and the family is a central mediator of how we construct these meanings”. Consequently, ethnic, racial, or subcultural identities may influence a family’s system (Anderson & Sabatelli, 1999). The relation between family environments and cultural background of the members is recognized and accepted as a fact. The diversity among families is expressed through differences in the number of family members (Baider, 2000), the relationships between family members, determinations of who is a member of the family, and/or the patterns of interaction between family members (Spinetta, 1984). Moreover, socio-economic status influences in a significant manner the family interactions and the family environment and should be taken into consideration when studying the diversity of family systems (Anderson & Sabatelli , 1999). Generally, people from diverse cultural/racial/ethnic groups consider family to be more inclusive than do individuals in the dominant cultural group in the United States (Rigazio-DiGillio, 2003). For example, extended families and a broader network of relationships seem to be common among African American, Hispanic, and Native American groups (Rigazzio-DiGillio, 2003). Holland (1998) described differences between Caucasian, Hispanic American, and Asian American families with regard to models of family network, support provided by family members to one another, and

PAGE 35

27 perceived autonomy of family members. The Caucasian model of family network generally defines family in terms of autonomy, individual responsibility, and a limited number of members. In contrast, Hispanic American and Asian American families are usually larger and more encouraging of interdependence of family members in comparison to Caucasian families. Moreover, Holland (1998) identified a tendency of members of Hispanic American and Asian American families to be more involved with a family member’s health care and health decision-making than members of Caucasian families. There are, unquestionably, similarities in how members of the same cultural environment define the meaning of a family and their family life. However, differences can also be found in family functioning, even in the same cultural context. Individual families may filter, modify, accept or reject certain influences of the cultural background of their members (Sciarra, 1996). Illustrative of this assertion is the fact that, across cultures, there is an increased frequency of single parenthood, divorce, remarriage, and “blended” families. Moreover, within different cultures the family is not necessarily defined in objective terms (as blood ties), but in a subjective and individual manner (Gotay, 2000). The “family culture” or the “culture” of a particular family is related not only to its members’ ethnicity, race, religious affiliation, age, socio-economic status, national origin, ability or disability, and sexual orientation, but also to the history of that family (similar to the way the process of self-identity development is related to but not determined by an individual’s cultural background). The embededdness model (Szapocznik et al., 1988) proposes an empirically-driven description of the relationship that exists in reality between individuals, their families,

PAGE 36

28 and their cultural background. More specifically, the embeddedness model expands the contextualist paradigm that views individuals in the context of their families and in the context of their cultural environments, separately. The disjunction between the family and the wider cultural context raises the concern of missing elements of the microcontext (i.e., the family) when the influence of culture is studied, or missing elements of the macrocontext (i.e., cultural environment) when the influence of family variables is studied (Szapocznik & Kurtines, 1993). To address this difficulty, the embededdness model views individuals in the context of their families, and views families in the context of a culturally pluralistic environment. The notion of embeddedness of contexts has been promoted by studies conducted by Jose Szapocznik and his collaborators, which have sought to identify patterns of family interaction that impact the development of conduct disorders among Hispanic youth. However, the implication of these research findings can be extended beyond this specific population to all families that are currently confronted with an increasingly complex and multicultural environment (Szapocznik & Kurtines, 1993). The embededdness model provides support for the view that different families may develop their own patterns of communication, and thus the use of preconceived culture based family models can prove to be too rigid (Steele, 1990). For example, despite the fact that the literature almost unanimously describes the Hispanic family as a key source of support for its members during times of hardship, there is some evidence contradictory to this perspective (e.g., Alferi, Carver, Antoni, Weiss, & Duran, 2001). Spinetta (1984) also cautions upon the use of preconceived cultural models, based on the notable differences he recorded between Spanish-speaking families of cancer

PAGE 37

29 patients from Central America and those from South-America, and also between Vietnamese and other families from neighboring countries. Based on his findings, Spinetta (1984) concludes that research involving families of people with cancer should take into account the particular background of a family instead of assuming particular characteristics of family functioning based on general descriptions of the broad cultural group to which the family belongs. Cultural background may influence the support that women with breast cancer need, the way they express their needs, their perception of the support received, as well as the support they actually receive from different sources. For example, Wellish and his colleagues (1999) found that there are differences in expression of needs for support between Caucasian and Asian American women with breast cancer, but no differences were found in terms of wanting more support (Wellish et al., 1999). Moreover, despite the fact that the needs for support were found to be the same for Caucasian and Asian American women, Caucasian and Asian American women reported receiving social support to different degrees. Purpose of the Proposed Study The proposed exploratory study will examine the associations of family environment and demographic and disease-related characteristics of women with breast cancer with (a) the type and level of support that these women desire and actually receive from their families, and (b) the level of satisfaction that these women experience in relation to the family support they perceive receiving. Multiple descriptors of family support (i.e., support perceived, support desired, and satisfaction with support) are used because measuring only support received as perceived by the patient does not still measure how satisfied the patient is. In order to have an accurate measure of how helpful

PAGE 38

30 family support really is, it is necessary to take into consideration support desired versus support perceived. Little, if any, research has specifically explored such discrepancies between desired and perceived family support. The major goals of this exploratory study are: (1) to determine the family environment characteristics (levels of cohesion, expressiveness, and conflict as indicated by scores on the Family Relationships Index) experienced by African American women and Caucasian women with breast cancer; (2) to explore the associations between family environment types (i.e., patterns of family cohesion, expressiveness, and conflict level as revealed by the results of cluster analyses applied to the family cohesion, expressiveness, and conflict data) experienced by African American women and Caucasian women with breast cancer and certain demographic (i.e., ethnicity, age, number of family members, and income) and disease-related characteristics (i.e., stage at diagnosis) of these women; (3) to determine if, among African American and Caucasian women with breast cancer separately, the characteristics of their family environments (levels of cohesion, expressiveness, and conflict) are associated with their levels of perceived family support, desired family support, and satisfaction with perceived family support (as indicated by the discrepancy between family support received and family support desired); and (4) to determine if a particular type of family environment (i.e., a particular pattern of cohesion, expressiveness, and conflict levels) can significantly predict the level of satisfaction with family support perceived among African American women with breast cancer as a group, and Caucasian women with breast cancer as a group. The following research questions were addressed:

PAGE 39

31 1. What are the types of family environments (i.e., patterns of cohesion, expressiveness, and conflict levels) that are experienced by African American and Caucasian women with breast cancer? 2. Among African American and Caucasian American women with breast cancer, are there differences in family environment types in association with their ethnicity, age, education, income level, and presence of spouse/partner? 3. The third research question is as follows: Among African American women with breast cancer as a group, and Caucasian American women with breast cancer as a group, are there differences in (a) levels of perceived emotional, tangible, or informational support, and in (b) levels of satisfaction with emotional, tangible, or informational support in association with type of family environment (i.e., type of cluster)? 4. Among African American women with breast cancer as a group and Caucasian American women with breast cancer as a group, is level of overall satisfaction with family support significantly associated with the dimensions of family environment (i.e., cohesion, expressiveness, and conflict)?

PAGE 40

CHAPTER 3 METHODS Participants Participants were recruited through (a) a physician oncologist who has a private practice in Gainesville, Florida, (b) two coordinators of support groups for women with breast cancer from Alachua County, Florida, and (c) two breast cancer survivors associated with the Gainesville section of the American Cancer Society. The criteria for inclusion in this study were (a) being 18 years of age or older, (b) being able to give informed consent, and (c) having a breast cancer diagnosis. Ninety-seven (97) out of 200 Research Questionnaires were returned to the principal investigator, resulting in a return rate of 48%. The final sample was composed of 97 women who have received a diagnosis of breast cancer and whose ages ranged from 32 to 82 years old (mean age = 57.41, sd = 11.81). Eighteen (19.8%) of the women were African American, 2 (2.0%) were Asian American, 73 (74.4%) were Caucasian American, and 3 (3.0%) were Hispanic American. One woman did not specify her ethnicity, and thus her data could not be considered. For the purpose of this study, only African American data and Caucasian American data were considered for statistical analyses. The final analyses included data from 91 women participants. Among the African American women, more than half (55.6%) did not work while 22.2% worked full time. The levels of schooling completed by African American women participants in this study were as follows: elementary school (5.6%), high school 32

PAGE 41

33 (27.8%), some college (33.3%), college (5.6%), and graduate/professional school (16.7%). Half (50.0%) of these African American women had family income levels below $20,000. Fifty percent of these women participants reported that their husband/partner was present in the home, and 72% of these women reported having one or more children. The percent of these African American women who rated the importance of family support in each of the specified family support importance rating categories is as follows: not important (5.6%), somewhat important (0%), important (33.3%), very important (16.7%), or extremely important (22.2%). The mean number of extended family members reported by these African American women was 15. Slightly more than half (55.6%) of these women had a history of cancer in their families, and 33.3% of them participated in a support group. Among Caucasian American women, 52.1% did not work, while 41.1% worked full time. The levels of schooling completed by the Caucasian American women participants in this study were as follows: elementary school (1.4%), high school (21.9%), some college (28.8%), college (12.3%), and graduate/professional school (34.2%). The family income level of almost half (47.9%) of the Caucasian American women was higher than $40,000, while 75.3% had a family income higher than $20,000 and only 16.4% had a family income lower than $20,000. Of these participating women, 72.6% reported that the husband/partner was present in the home and 75.3% reported having one or more children. The percent of these Caucasian American women participants who rated the importance of family support in each of the specified family support importance rating categories is as follows: not important (0%), somewhat important (2.7%), important (19.2%), very important (34.2%), and extremely important

PAGE 42

34 (41.1%). The mean number of extended family members reported by these Caucasian American women participants was 11.7. Over sixty-one percent (61.6%) of these women had a history of cancer in their families, and 20.5% of them participated in a support group. Table 1. Demographic Description of the African American and Caucasian American Participants Variable African American Caucasian American Age Range Mean 39-82 60.2 32-80 57.5 Presence of Spouse/Partner: Yes No N 9 6 % 50 33 N 51 21 % 70 29 Employment Status: Work Full Time Work Part Time Do Not Work N 4 1 10 % 22 7 56 N 30 4 38 % 41 5 52 Income Level Below 10,000 10-20,000 20-30,000 30-40,000 Above 40,000 N 5 4 3 0 3 % 28 22 17 0 17 N 6 6 6 14 35 % 8 8 8 19 48 Education Level Elementary/Secondary High School Some College/Technical College Graduate/Professional N 1 5 6 1 3 % 6 28 33 6 17 N 1 16 21 9 25 % 1 22 29 12 34 Extended Family Members: Range Mean number 2 42 15 1 200 12 Currently Under Treatment Yes No N 6 9 % 33 50 N 52 19 % 71 26 Time Since Diagnosis (Month) Mean Standard Deviation Range 51 44 10 144 24 29 .01 123 Note: The percent columns under each variable do not add up to 100% because of missing data (i.e., no responses to some items on the Demographic Data Questionnaire)

PAGE 43

35 Table 1. Continued Variable African American Caucasian American Type of Treatment Surgery Only Chemotherapy Only Surgery + Other Other N 3 1 9 0 % 17 6 50 0 N 6 2 56 6 % 8 3 77 8 Support Group Participation Yes No N 6 9 % 33 50 N 15 55 % 20 75 Note: The percent columns under each variable do not add up to 100% because of missing data (i.e., no responses to some items on the Demographic Data Questionnaire) Measures The Research Questionnaire that participants were asked to complete consisted of the following specific questionnaires: Family Relationships Index (FRI) (Holahan & Moos, 1974). The FRI is a self-report index of the quality of a family’s social relationships. The FRI consists of the three subscales which are Cohesion, Expressiveness, and Conflict. Each of these subscales consists of nine true-false items. The Cohesion subscale assesses the degree of commitment, help, and support family members provide to one another (e.g., ‘Family members really back each other up’). The Expressiveness subscale assesses the extent to which family members are encouraged to act openly and to express their feelings directly (e.g., ‘We tell each other about our personal problems’). The Conflict subscale assesses the amount of openly expressed anger, aggression, and conflict among family members (e.g., ‘Family members often criticize each other’). The items for all three subscales can be rated as 0 or 1, and the total score for each subscale represents the sum of points for all scale items. For the Expressiveness and Cohesion subscales, higher to lower scores indicate higher to lower quality of family functioning, respectively. The scores for the

PAGE 44

36 Conflict subscale have a reversed meaning, that is higher scores indicate more conflict in the family and lower quality of family functioning. The three subscales of the FRI represent the Communication dimension of the larger Family Environment Scale (Moos & Moos, 1986) for which the normative sample data was collected for families from all areas of the U.S.A. These families included single-parent and multigenerational families, families from different ethnic groups (Hispanic and African American), as well as families of various age groups. The FRI has moderate to high internal consistencies ranging from .61 to .78. The two-month test-retest reliabilities are all within the acceptable range with .86 for Cohesion, .73 for Expressiveness, and .85 for Conflict (Moos & Moos, 1986). The FRI was also found to significantly correlate with other measures of social support and outcome measures, thus suggesting that it has good construct validity. The FRI has often been used in research involving women with breast cancer. The findings generally indicate that supportive family environments, characterized by high cohesion and expressiveness and low conflict, are associated with family members’ better adjustment when coping with personal physical illness. Women with breast cancer from less cohesive families were more likely to use coping responses that resulted in poor health outcomes (Bloom, 1982). Also, a better adjustment during the year after breast surgery was predicted by more expression and less conflict in the family (Spiegel, Bloom, & Gottheil, 1983; Baider & De-Nour, 1984). Inventory of Socially Supportive Behavior (ISSB) (Barrera, Sandler & Ramsay, 1981). The ISSB is a 40-item self-report measure designed to assess how often individuals received various forms of assistance during the preceding month. Three

PAGE 45

37 studies have examined the factor structure of the ISSB (Barrera & Ainlay, 1983; Caldwell & Reinhart, unpublished; Stokes & Wilson, 1984) with considerable agreement across the findings. The present study will use the factors proposed by Barrera and Ainlay. The three scales of social support that were used in this study are: Informational Support (e.g., “Taught you how to do something”), Emotional Support (e.g., “Expressed interest and concern in your well-being”), and Tangible support (e.g., “Gave you transportation”). The Emotional Support scale used in this study represents a combination of the Emotional Support and Social Interaction scales of the structure proposed by Barrera and Ainlay. There was considerable overlap between the items of these two subscales, and the Social Interaction scale was only comprised of five items. Respondents were asked to rate the frequency of each item (support behavior) on a 5-point Likert scale (1 = not at all to 5 = about every day). The 5-point ratings of each item were averaged to form a total frequency score for each of the three subscales and an average score is then calculated for each subscale. For the purpose of this study, each item was answered and scored on both a Support Perceived and a Support Desired Form. The Support Perceived Form referred to the perceived frequency with which specific supportive behaviors occurred, and the Support Desired Form inquires about the desired frequency of the same supportive behaviors. The test instructions were modified to refer specifically to family support (and not to support from other sources). The authors of the test have given their permission for usage of ISSB in a modified version in order to meet the needs of different research projects. The internal consistency reliability for the ISSB has been consistently above .90 (Barrera, 1981; Barrera, Sandler, & Ramsay, 1981; Cohen & Hoberman, 1983; Cohen et

PAGE 46

38 al., 1984; Stokes & Wilson, 1984). Test-retest reliabilities over a 1-month interval were .80 and .63 for samples of undergraduate students (Barrera & Ainlay, 1984) and female graduate students, respectively (Valdenegro & Barrera, 1983). ISSB total scores have shown correlations between 0.24 and 0.42 with measures of network size (Barrera, 1981; Sandler & Barrera, 1984; Valdenegro & Barrera, 1983). The ISSB also correlates (.359) with the Cohesion subscale of Moos' Family Environment Scale (Barrera et al., 1981). A short version of the Patient Satisfaction Questionnaire (PSQ) (Marshall & Hays, 1994). The PSQ-18 is a general measure of the satisfaction with one’s medical healthcare. The questionnaire consists of 18 items divided in seven scales (General Satisfaction, Technical Quality, Interpersonal Manner, Communication, Financial Aspects, Time with the Doctor, and Accessibility and Convenience), but a total score can be obtained as well. Respondents are asked to rate how much they agree with different statements regarding their satisfaction with medical care received, on a scale from 1 (Strongly Agree) to 5 (Strongly Disagree). Some items are worded so that agreement reflects satisfaction with medical care, whereas other items are worded so that agreement reflects dissatisfaction with medical care. The total satisfaction score is computed as the average for all items in the scale that were answered. The data from this questionnaire was obtained for another related study to the present study and thus was not analyzed as part of the present study. Demographic and Medical Data Questionnaire (DMQ). The DMQ was constructed by the primary researcher and has three parts: (a) questions that solicit demographic data (age, ethnicity, employment status, level of education, and income

PAGE 47

39 level), (b) questions about participants’ families and the family support received (number of family members and kinship type, number of family caregivers, principal caregiver within the family, family’s ethnic background, and the presence of a history of cancer in the family), and (c) questions that solicit one’s history of breast cancer and its treatment (age at diagnosis, if the participant is currently under treatment or not, time since treatment was received, type of treatment, current and prior counseling/ support group participation). Each participant received an Invitation Letter and two copies of the Informed Consent Form as well (see Appendices B and C, respectively). Procedures Two different methods for collecting data were used. In the first method all women with a breast cancer diagnosis who came to a community-based clinic of a private oncologist were invited to participate in this study by the clinic’s receptionist. In the second method participants were recruited by (a) two arbitrarily selected breast cancer survivors who are advocates for breast cancer prevention education, and (b) two leaders of the local cancer survivor support groups. In the clinic recruitment method, as a first step the oncologist at the participating private oncology clinic was contacted. The oncologist was provided information regarding the following: (a) the general goals of the study (i.e., to identify and describe types of family environments among women with breast cancer, and to determine the association of these family environment types with women’s satisfaction with their perceived family support); (b) the research procedures; (c) the measurement instruments, (d) the potential benefits of participating in the study (i.e., the opportunity to attend a publicly announced presentation of the results from the study), (e) the procedures for

PAGE 48

40 ensuring participants’ confidentiality, (f) the length of time that participation in this study would involve (i.e., approximately 20 minutes), and (g) anticipated duration of the data collection process (i.e., three months). As a second step, the oncologist at the participating private oncology clinic was asked for permission to recruit patient participants from among patients who are treated at his clinic. Patient participant recruitment at the clinic involved having the clinic receptionist extend a research participation invitation to all women with a breast cancer diagnosis who came in the clinic for their medical appointments. The invitation message used by this clinic receptionist is provided in Appendix D. This clinic receptionist gave each patient who came to the clinic a packet containing an Invitation Letter, two copies of the Informed Consent Form, and the Research Questionnaire. Women who decided to participate were asked via the invitation letter to read and sign one of the Informed Consent Forms prior to completing the Research Questionnaire and to return the signed Informed Consent Form in the designated confidential drop box made available at the clinic. These women participants were also instructed to return their completed Research Questionnaire in a designated confidential drop box that had also been made available at the clinic. This box for research questionnaires was different from the box for Informed Consent Forms. Participants were encouraged via the invitation letter to answer the Research Questionnaire while they were waiting for their medical appointment or while receiving chemotherapy. However, participants had the option of completing the Research Questionnaire at home in case they preferred to do so or if they did not finish completing it while at the clinic. Participants who chose to take the Research Questionnaire home

PAGE 49

41 were asked to return it by mail within one week, using a stamped envelope available at the clinic for their convenience. The return rate for this method of recruitment was 67% (around 80 Research Questionnaires were returned out of the 120 that were distributed). Because the first method of participant recruitment rendered almost exclusively Caucasian American participants, the support group leader and community member method of recruitment (i.e., the second method of participant recruitment) was implemented with the goal of increasing the number of African American women participants. The first step for this recruitment method was to contact two African American leaders of culturally diverse support groups for women with breast cancer and two African American breast cancer survivors well known for their advocacy for breast cancer prevention education in a local or nearby African American community (i.e., in Gainesville Florida, or Trenton, Florida). After being given an overview of the present study, the two contacted support group leaders and the two contacted community member breast cancer prevention advocates agreed to help with research participant recruitment. In separate meetings with the support group leaders and community members, the primary researcher presented these participant recruiters with more detailed information about the following: (a) the general goal of the study, (b) the research procedures, (c) the Research Questionnaire, (d) the benefits of participating in this study (i.e., the opportunity to attend a publicly announced presentation on the results from this study and $10 from the person who invited them to participate as soon as the participant reported having mailed in her questionnaire), and (e) the procedures used to ensure participants’ confidentiality. The $10 monetary incentive was used in the second but not in the first method of participant recruitment. This is because the first recruitment method, which did

PAGE 50

42 not have a $10 monetary incentive, had not been successful in recruiting the African American breast cancer survivors who constituted approximately 15-20% of the women with breast cancer who received treatment at the community-based oncology clinic participating in this study. Support group coordinators and the community members individually and informally presented to potential participants a brief overview of the goals and procedures of the present research study and then invited these potential participants to participate (see Appendix E for the script). Support group leaders and community members handed women who expressed interest in participating in this study a packet containing a Cover Letter, two copies of the Informed Consent Form, and the Research Questionnaire together with a stamped, self-addressed envelope. Women who agreed to participate were asked to return these materials no later than one week. The Invitation Letter mentioned a phone number to be called to communicate questions about the study or requests to have the Research Questionnaire read to the participants by a research assistant. No such calls were received. The return rate for this second recruitment method was 38% (15 out of approximately 40 Research Questionnaires disseminated were returned). The overall duration for the data recruitment was 5 months. It is noteworthy that women recruited at the private clinic by the receptionist of the participating oncologist were offered the option of completing the Research Questionnaire on line. However, none of these women selected this on-line completion option. Thus, this option was not offered to subsequent potential research participants (i.e., those recruited by support groups and those recruited by the community member breast cancer prevention advocates). The Invitation Letter and Informed Consent Form

PAGE 51

43 prepared for the online administration of the Research Questionnaire are attached in Appendix F and Appendix G respectively.

PAGE 52

CHAPTER 4 RESULTS This chapter is organized in four parts. Each of these parts is dedicated to presenting the results of the main statistical analyses that were conducted to address each of the four research questions set forth in this study. The results of preliminary data analyses to determine variable that should be covariates on the main statistical analyses are also presented. Additionally, descriptive data on the variables of interest in this study are presented. All statistical analyses were performed using the Statistical Package for Social Science (SPSS). First Research Question The first research question is as follows: What are the types of family environments (i.e., patterns of cohesion, expressiveness, and conflict levels) that are experienced by African American and Caucasian American women with breast cancer? In order to address this research question, a Ward’s cluster analysis was performed on data from the Cohesion, Expressiveness, and Conflict subscales of the Family Relations Index administered to all of the women with breast cancer in this study. To address this research question, data from African American and Caucasian American women participants were included into the same cluster analysis to avoid imposing any categories on the exploration of family environment types that exist among women with breast cancer. The association between family environment types and ethnicity was further addressed. A clustering analysis allows us to merge scores into groups in such a way that cases within a group exhibit certain similarities and, at the same time, are different from cases 44

PAGE 53

45 in other groups. This methodology has the advantage that preconceived notions are not influencing the grouping of scores obtained for the studied variables. Within this research study, the groups of scores formed through the cluster analytical procedure indicate different types of family environments that are characterized by different scores on the three subscales of the FRI (e.g., one type might be low in cohesion, average in expressiveness, and high in conflict, whereas another family environment type might be characterized by average scores in all three FRI subscales). The means, standard deviations, and actual score ranges for each of the three FRI subscales are presented in Table 2. Table 2. Descriptive Data for the Subscales of the Family Relationship Inventory (FRI) by Ethnic Group African American Caucasian American Cohesion Mean Score Range Standard Deviation .89 .44 1.00 .14 .90 .22 1.00 .12 Expressiveness Mean Score Range Standard Deviation .64 .11 .89 .14 .67 .00 1.00 .23 Conflict Mean Score Range Standard Deviation .16 .00 .56 .16 .19 .00 .78 .19 Note: FRI mean subscale scores were computed using valid items. Ward’s cluster analysis was initially used to test multiple cluster solutions (i.e., two, three, four, and five cluster solutions). A range of solutions (versus a definite number of clusters) was proposed to identify the best fit for the data in this study. A pre

PAGE 54

46 established number of clusters might have presented the risk of forcing the existent data into a fixed pattern, which raises the risk of errors. All four clustering solutions obtained using the hierarchical clustering option of SPSS were analyzed using the agglomerative schedule and the dendogram to determine the best cluster solution. Standardized scores were used for the purpose of cluster analysis. Both the agglomerative schedule and the dendogram suggested that either a three-cluster or four-cluster solution is the most appropriate. The twocluster and five-cluster solutions were excluded on the basis of their weaker distribution of cases. Specifically, the two-cluster distribution of scores was considered to be too general. The five-cluster solution was considered to be redundant, since it did not add a significant new group of scores. The three and four-cluster solutions were compared next to determine their fit. Table 3. Distribution of Cases for the Three-Cluster and Four-Cluster Solutions from the Cluster Analysis Applied to the FRI Data Cluster 1 Cluster 2 Cluster 3 Cluster 4 Three-cluster solution Frequency Percent 63 69.2% 11 12.1% 17 18.7% N/A Four-cluster solution Frequency Percent 63 69.2% 3 3.3% 17 18.7% 8 8.8% Note: Frequencies and percents are computed based upon the total number of participants. As Table 3 indicates, the three-cluster and four-cluster solutions had in common two groups of scores. Specifically, these groups of scores or clusters included scores from the same individual cases (participants). Moreover, the three-cluster solution combines the two remaining groups from the four-cluster solution into a single one. The agglomerative schedule and the cluster membership chart also indicate that the four-cluster solution did not add a significant new cluster. Consequently, the three-cluster

PAGE 55

47 solution was determined to be the best solution. The three-cluster solution maximized the differences between and minimized the differences within each cluster to the greatest extent, compared with the two, four, and five cluster solutions. For the purpose of further analyses, a new categorical variable named “family environment” was created based upon the cluster membership of each individual case. The new variable named “family environment” indicates patterns of family cohesion, expressiveness, and conflict. The first family environment type was the most common among women participants in this study and it comprised 63 cases (69%), while the second and third family environment types were less common, comprising 11 (12%) and 17 (19%) cases respectively. Each family environment type is described in Table 4. Table 4. Cohesion, Expressiveness, and Conflict Levels Characteristic of the Three Family Environment Types Family Environment 1 Family Environment 2 Family Environment 3 Cohesion Mean Range .94 .67 – 1 .75 .22 1 .86 .72 1 Expressiveness Mean Range .77 .44 1 .61 .44 .89 .29 0 .56 Conflict Mean Range .12 0 .44 .55 .22 .78 .19 0 .44 Overall characteristics Highest cohesion Highest expressiveness Lowest conflict Lowest cohesion Moderate expressiveness Highest conflict Moderate cohesion Lowest expressiveness Moderate conflict Proposed Label Cohesive-Expressive Conflictual-Expressive CohesiveNonexpressive

PAGE 56

48 In order to validate the final three-cluster solution chosen, the three clusters were compared to test for significant differences between them. Specifically, a one-way ANOVA and Tuckey post-hocs were used to compare mean Cohesion, Expressiveness, and Conflict scores of each cluster with the mean scores of the other clusters. Significant differences in Cohesion scores (F(2,90) = 14.99, p < .0001), Expressiveness scores (F(2,90) = 82.04, p < .0001), and Conflict scores (F(2,90) = 48.97, p < .0001) were obtained. Consequently, all three family environment types significantly differed from each other on all three family dimensions (i.e., cohesion, expressiveness, and conflict). 00.20.40.60.81Cohesion Expressiveness Conflict Family DimensionsFamily Environment CohesiveFamilies ConflictualFamilies Non-expressiveFamilies Figure 1. Family Environment Types Among African American and Caucasian American Women with Breast Cancer Second Research Question The second research question is as follows: Among women with breast cancer, are there differences in family environment types in association with women’s ethnicity, age, education, income level, and presence of spouse/partner? Family environment type, ethnicity, education, income level, and presence of spouse/partner are categorical variables, and age is a continuous variable. To address this research question, several analyses were performed. First, preliminary Kendall correlations between family environment type, ethnicity, education, income level, presence of spouse/partner status, and age were conducted. Second, chi

PAGE 57

49 square analyses were conducted to test for the independence between the demographic variables (i.e., ethnicity, education level, income level, and presence of spouse/partner status) and types of family environments (i.e., Cohesive-Expressive, Conflictual-Expressive, and Conflictual-Nonexpressive) among women with breast cancer. These analyses were performed using the data from African American and Caucasian American women combined and then with the data from the African American women as a group and with the data from the Caucasian American women as a group. The latter separate group analyses were performed as indicated by the tenants of the “difference model” research approach (Oymenade & Rossier, 1980). When the combined data was analyzed, ethnicity was included in the chi square analyses as a categorical variable. Kendall correlations using the combined data indicated that family environment type was significantly correlated with education level (r(88) = -.240, p < .05) and with presence of a spouse/partner status(r(87) = -.243, p < .05). Chi-square analyses of independence on the combined data indicated that education and presence of spouse/partner status were not independent from family environment type. The chi-square coefficients were (88) = 20.78, p < .01, df = 8 for education, and (87) = 8.51, p < .05, df = 2 for presence of spouse/partner status. These findings suggest that differences may exist between family environment types in association with education levels and presence of spouse status. Standardized residuals were used to test the interactions between family environment type and education level, and between family environment type and presence of spouse/partner. Standardized residuals represent measures of the difference between expected and observed frequencies. The higher the value of the standardized

PAGE 58

50 residuals, the more likely it is that they are contributors to a significant chi-square coefficient. The standardized residuals for the education by family environment type chi square indicated that a larger than expected number of women in Cohesive-Expressive family environments had a graduate/professional education (standardized residual = 1.1); a larger than expected number of women in Conflictual-Expressive families had a high school education (standardized residual = 2.1); and a larger than expected number of women in Conflictual-Nonexpressive families had a high school and college education (standardized residuals = 1.0 and 1.5 respectively). At the same time, a lower than expected number of women in Cohesive-Expressive families had a high school education (standardized residual = -1.4), a lower than expected number of women in Conflictual-Expressive families had some college/technical school education and college education (standardized residuals = -1.3 and .1 respectively), and a lower than expected number of women in Conflictual-Nonexpressive families had a graduate school education (standardized residual = -2.3). The standardized residuals for presence of spouse/partner status by family environment type indicated that a lower than expected number of women in Cohesive-Expressive families did not have a spouse/partner (standardized residual = -1.2), a larger than expected number of women in Conflictual-Expressive families did not have a spouse/partner present (standardized residual = 1.9), and a lower than expected number of women in Conflictual-Nonexpressive families had a spouse/partner (standardized residual = -1.3). These findings suggest that women with breast cancer in the Cohesive-Expressive families differed from women with breast cancer in the Conflictual-Expressive and Conflictual-Nonexpressive families with regard to their level of education and presence

PAGE 59

51 of a spouse/partner status. Specifically, women with breast cancer in Cohesive-Expressive families were more likely to have achieved higher education levels compared with women with breast cancer from Conflictual-Expressive and Conflictual-Nonexpressive families. At the same time, a spouse/partner was more likely to be present in Cohesive-Expressive families than in Conflictual-Expressive and Conflictual-Nonexpressive families of women with breast cancer who participated in this study. For African American women as a group, none of the Kendall correlation coefficients were significant. Moreover, for African American women, none of the chi-square coefficients were significant, indicating that family environment types and education levels, income levels, and presence of spouse/partner status were independent from each other. For Caucasian American women, Kendall correlations indicated that family environment type was significantly correlated with education levels (r(72) = -.26, p < .05) and with presence of spouse status (r(72) = -.25, p < .05). Significant chi square coefficients on data from Caucasian American women indicated that family environment type was not independent from income levels and presence of a spouse/partner. Chi square coefficients were (67) = 15.58, p < .05, df = 8 for income levels, and (72) = 8.51, p < .05, df = 2 for presence of spouse/partner status. The standardized residuals for the income level by family environment type indicated that (a) a lower than expected number of women in Cohesive-Expressive families had an income of $10,000 – $20,000 (standardized residual = -1.5), (b) a larger than expected number of women in Conflictual-Expressive families had an income lower than $10,000 (standardized residual = 1.3) and an income between $10,000 – $20,000 (standardized residual = 2.4), (c) a

PAGE 60

52 lower than expected number of women in Conflictual-Expressive families had an income lower than $40,000 (standardized residual = -1.7), and (d) a lower than expected number of women on Conflictual-Nonexpressive families had an income lower than $10,000 (standardized residual = -1.1). Additionally, a lower than expected number of women in Cohesive-Expressive families did not have a spouse/partner present (standardized residual = -1.1), while a higher than expected number of women in Conflictual-Expressive families did not have a spouse/partner (standardized residual = 1.5). These findings suggest that Caucasian American women with breast cancer in the Cohesive-Expressive families, Conflictual-Expressive families, and Conflictual-Nonexpressive families differed from each other with regard to their family incomes. Specifically, Caucasian American women with breast cancer in Cohesive-Expressive and Conflictual-Nonexpressive families were less likely to have lower family incomes; and Caucasian American women with breast cancer in Conflictual-Expressive families were more likely to have lower family incomes and they were less likely to have higher family incomes. At the same time, a spouse/partner was less likely to be absent from Cohesive-Expressive families, while a spouse/partner was less likely to be present in Conflictual-Expressive families. Third Research Question The third research question is as follows: Among African American women with breast cancer as a group, and Caucasian American women with breast cancer as a group, are there differences in (a) levels of perceived emotional, tangible, or informational support, and in (b) levels of satisfaction with emotional, tangible, or informational support in association with type of family environment (i.e., type of cluster or levels of cohesion, expressiveness, and conflict in the family)?

PAGE 61

53 The satisfaction with support received from family members was computed as the difference between the Desired and Real (Perceived) ratings for each of the ISSB items. A higher absolute value of this difference indicated a bigger discrepancy between the support desired and the support received by women with breast cancer from their family members. Thus, greater absolute values of this difference were interpreted as less satisfaction with the support received. The positive versus negative direction of the difference indicated whether the needs for support of women with breast cancer are not fully met or are exceeded. More specifically, the support received was not enough to meet a woman’s expectations when Desired minus Real had a positive value, and the support received exceeded women’s needs when the Desired minus Real had a negative value. For the purpose of this study, satisfaction with support was computed as the absolute value of the difference between support desired and support perceived. To determine whether to use ANOVAs or MANOVAs to address the third research question, preliminary Pearson correlation analyses were performed by ethnic group to determine if the dependent variables (i.e., perceived emotional, tangible, and informational support) were significantly associated with each other. For both ethnic groups of women with breast cancer, analyses indicated (a) significant correlations among perceived emotional, tangible, and informational support (see Table 5), and (b) significant correlations among self-reported satisfaction with emotional support, tangible support, and informational support (see Table 6). Specifically, all three forms of support perceived were significantly correlated with each other, and all three forms of self-reported support satisfaction were significantly correlated with each other. These findings

PAGE 62

54 provided support for using multivariate analyses (versus univariate analyses) to test the third research question. Table 5. Correlation Coefficients for the Associations Among Perceived Informational, Tangible, and Emotional Support by Ethnic Group Informational Support Perceived Tangible Support Perceived Emotional Support Perceived Informational Support Perceived African American Caucasian American 1.00 1.00 .72** .67** .78** .77** Tangible Support Perceived African American Caucasian American .72** .67** 1.00 1.00 .56* .67** Emotional Support Perceived African American Caucasian American .78** .77** .56* .67** 1.00 1.00 Note: *p < .05, **p < .01 Table 6. Correlation Coefficients for the Associations Among Satisfaction with Informational, Tangible, and Emotional Support by Ethnic Group Satisfaction with Informational Support Satisfaction with Tangible Support Satisfaction with Emotional Support Satisfaction with Informational African American Caucasian American 1.00 1.00 .97** .57** .74* .82** Satisfaction with Tangible African American Caucasian American .97** .57** 1.00 1.00 .82** .51** Satisfaction with Emotional African American Caucasian American .74* .82** .82** .51** 1.00 1.00 Note: *p < .05, **p < .01 Preliminary Pearson correlation analyses were also performed to determine whether any of the demographic variables of interest in this study should be covariates in the analyses to address research question three. Specifically, Pearson correlation analyses

PAGE 63

55 were performed separately by ethnic group to determine if there are any associations between both the perceived support variables and support satisfaction variables, and the following demographic variables: education levels, presence of a spouse/partner status, treatment type, and support group participation status. It is noteworthy that the correlation coefficients and their significance were different for the African American women group versus the Caucasian American women group. The correlation coefficients for African American women appear in Table 7, and for Caucasian American women appear in Table 8. These correlation coefficients indicate that some demographic variables must be controlled for in multivariate analyses to address research question three. Table 7. Correlations Between Demographic Variables and the Perceived Support Variables and Self-Reported Support Satisfaction Variables for African American Women Education Level Presence of Spouse/ Partner Treatment Type Support Group Participation Perceived Informational Support .212 .797** .435 -.079 Perceived Tangible Support .123 .497 .043 -.065 Perceived Emotional Support .353 .589* .511 -.348 Satisfaction with Informational Support -.440 -.435 -.964** .494 Satisfaction with Tangible Support -.372 -.330 -.966** .621 Satisfaction with Emotional Support -.405 -.440 -.638* .632* Note: ** p < .01; * p < .05 In sum, Table 7 indicates that for the African American women participants, presence of spouse/partner was significantly correlated with perceived informational support and perceived emotional support; treatment type was significantly correlated with satisfaction

PAGE 64

56 with informational support, satisfaction with tangible support, and satisfaction with emotional support; and support group participation status was significantly correlated with satisfaction with emotional support. Table 8. Correlations Between Demographic Variables and the Perceived Support Variables and Self-Reported Support Satisfaction Variables for Caucasian American Women Education Level Presence of Spouse/ Partner Treatment Type Support Group Participation Perceived Informational Support -.158 -.075 .048 .121 Perceived Tangible Support .029 .123 .116 .284* Perceived Emotional Support .012 .177 -.057 .139 Satisfaction with Informational Support -.324* -.269* .100 .172 Satisfaction with Tangible Support -.178 -.285* -.090 .069 Satisfaction with Emotional Support -.343** -.297* .084 .086 Note: ** p < .01; * p < .05 In sum, Table 8 indicates that for the Caucasian American women participants, education level was significantly correlated with satisfaction with informational support and satisfaction with emotional support; presence of spouse/partner was significantly correlated with satisfaction with informational support, satisfaction with tangible support, and satisfaction with emotional support; and support group participation was significantly correlated with perceived tangible support. Two MANCOVAs were performed separately by ethnic group, as the main analyses to address research question three. In the first MANCOVA using data from African American participants, perceived informational, tangible, and emotional support were the dependent variables; family environment type was the independent variable; and presence of spouse/partner status was a covariate. No significant multivariate effects were obtained. This finding indicates that African American women with breast cancer who

PAGE 65

57 participated in this study did not differ in their reported levels of perceived informational, tangible, and emotional support in association with their family environment type. In the second MANCOVA performed using data from African American women participants, satisfaction with informational support, satisfaction with tangible support, and satisfaction with emotional support were the dependent variables, family environment types was the independent variable, and type of treatment and support group participation status were covariates. The multivariate test was significant for family environment type, Wilks’ lambda F(6,8) = 5.95, p < .05. Follow-up ANOVA analyses indicated a significant univariate effect of family environment type on satisfaction with emotional support (F(2,6) = 5.51, p < .05). Post-hoc Tukey’s tests could not be performed because one of the three family environment types contained only one case. However, a t-test revealed that the African American women in Cohesive-Expressive families as compared to the African American women in Conflictual-Expressive families are significantly more satisfied with the emotional support they receive from family members (mean difference = -1.38, p < .01). In the first MANCOVA using data from the Caucasian American women participants, informational, tangible, and emotional support perceived were the dependent variables, family environment type was the independent variable, and support group participation status was a covariate. A significant multivariate effect was found for family environment type, Wilks’ lambda F(6,106) = 2.75, p < .05. Follow-up ANOVA analyses indicated significant univariate effects of family environment type on tangible support perceived (F(2,55) = 3.31, p < .05) and on emotional support perceived (F(2,55) = 5.29, p < .01). Post-hoc Tukey’s tests revealed that Caucasian American women participants in

PAGE 66

58 Conflictual-Expressive families perceived that they received significantly more tangible support than women in Conflictual-Nonexpressive families (mean difference = 1.13, p < .05). Additionally, the Caucasian American women participants in Cohesive-Expressive families perceived that they received significantly more emotional support from their family members compared with women in Conflictual-Nonexpressive families (mean difference = .99, p < .01), and the Caucasian American women participants in Conflictual-Expressive families perceived that they received more emotional support than the women participants in Conflictual-Nonexpressive families (mean difference = 1.12, p < .05). In the second MANCOVA performed on the data from the Caucasian American women participants, satisfaction with informational, tangible, and emotional support were the dependent variables, family environment type was the independent variable, and education level and presence of spouse status were covariates. No significant multivariate effects were found, suggesting that the Caucasian American women participants in the present study did not differ in their reported satisfaction with informational, tangible, and emotional support from their family members in association with their family environment type. Fourth Research Question The fourth research question is as follows: Among African American women as a group and Caucasian American women as a group, is level of overall satisfaction with family support significantly associated with the dimensions of family environment (i.e., cohesion, expressiveness, and conflict)? To address the fourth research question, preliminary Pearson correlation analyses were performed separately by ethnic group on the data for the following variables: (a) overall satisfaction with family support, (b) the dimensions of family environment (i.e.,

PAGE 67

59 cohesion, expressiveness, and conflict), and (c) the demographic variables of interest which may need to be covariates in the main analyses to address research question four. Results revealed that for African American women, type of treatment received for breast cancer was significantly correlated with satisfaction with overall support, r(11) = -.91, p < .0001. Moreover, a significant moderate correlation was found between family cohesion and family expressiveness, r(18) = .51, p < .05). Results of the hierarchical regression analysis revealed that African American women’s satisfaction with their overall perceived family support was predicted by family conflict (F(4,10) = 48.21, MSe= 9.86, adjusted r 2 = .95 , p < .0001. This finding indicates that variables such as family conflict increase as satisfaction with overall perceived family support decreases. The standardized and unstandardized beta coefficients and the probability values for the regression analysis on data from African American women appear in Table 9. Because of the small sample of African American women participants (n = 18), the performed hierarchical regression analysis was performed as an exploratory analysis for which findings are viewed as suspect. Table 9. Hierarchical Regression Predicting Support Satisfaction from Family Cohesion, Expressiveness, and Conflict for African American Women Participants Unstandardized Standardized Variable Beta Beta t Sig. Model 1 Constant Treatment type 5.749 -1.066 -.91 8.506 -6.824 .000 .000 Model 2 Constant Treatment type Cohesion Expressiveness Conflict 5.258 -1.112 -1.617 2.082 3.395 -.956 -.139 .189 .317 5.492 -11.509 -1.401 1.823 4.025 .002 .000 .211 .118 .007 Note: Positive beta coefficients indicate an inverse relationship and negative beta coefficients indicate a direct relationship.

PAGE 68

60 Results of the Pearson correlation applied to the data from the Caucasian American women participants revealed that education level and presence of a spouse/partner status were significantly correlated with satisfaction with overall family support, r(57) = -.32, p < .05 and r(57) = .-33, p < .05, respectively. Additionally, a low significant correlation was found between education level and family conflict (r(72) = -.24, p < .05). Other findings were that family cohesion was significantly correlated with family expressiveness (r(73) = .32, p < .01) and with family conflict (r(73) = -.32, p < .01); and family conflict was significantly correlated with overall satisfaction with family support (r(58) = .32, p < .05). In the hierarchical regression analysis applied to the data from Caucasian American women participants, satisfaction with overall family support was the dependent variable; family cohesion, family expressiveness, and family conflict were the independent variables; and education level and presence of spouse/partner status were the controlled variables. The results of this hierarchical regression analysis revealed that none of the beta coefficients for family cohesion, expressiveness, or conflict scores was significant (see Table 10). Table 10. Hierarchical Regression Predicting Support Satisfaction from Family Cohesion, Expressiveness, and Conflict for Caucasian American Women Participants Unstandardized Standardized Variable Beta Beta t Sig. Model 1 Constant Education Level Spouse/Partner 3.016 -.332 -.937 -.260 -.268 4.811 -2.040 -2.108 .000 .046 .040 Model 2 Constant Education Level Spouse/Partner Cohesion Expressiveness Conflict 2.888 -.254 -.811 -.396 -.275 1.454 -.199 -.232 -.034 -.040 .192 1.833 -1.478 -1.791 -.238 -.294 1.397 .073 .146 .079 .813 .770 .168 Note: Positive beta coefficients indicate an inverse relationship and negative beta coefficients indicate a direct relationship.

PAGE 69

CHAPTER 5 DISCUSSION This chapter provides a summary and interpretation of the results of this study and is organized in four parts. First, a discussion of the results from each research question and their significance are presented. Second, limitations of this study are discussed. Third, implications for future research and for the field of counseling psychology are presented. Fourth, conclusions are presented. Findings for Each Research Question The first research question asked whether there are different types of family environments among African American and Caucasian American women with breast cancer. The application of a cluster analytic technique to these women’s scores on family cohesion, expressiveness, and conflict resulted in the emergence of three different types of family environments. The three family environment types were significantly different from each other with regard to their patterns of cohesion, expressiveness, and conflict levels. One family environment type was labeled “Cohesive-Expressive” because it had high cohesion, high expressiveness, and low conflict scores. This pattern of scores suggests that this family type would be highly functional while dealing with stressful situations and that women with breast cancer who are members of this family type would perceive more support from their family members and would be more satisfied with the family support they receive in comparison with women in families having other family 61

PAGE 70

62 environment types. The majority of women participants’ families were Cohesive-Expressive (almost 70%). A second family environment was called “Conflictual-Expressive”. High conflict, moderate expressiveness, and low cohesion scores characterized this family environment type. High conflict seems to be a prevalent dimension of these families; furthermore, members of these families may be disconnected from each other as a way of dealing with the conflict between them. It is likely that members of Conflictual-Expressive families have difficulties offering support of any type since cohesion scores are low and conflict scores are high for these families. A low percent (12%) of the participants in this study belonged to this family environment type (12%). The third family environment identified among women participants in this study was named “Cohesive-Nonexpressive” and was characterized by moderate cohesion and conflict scores, and low expressiveness scores. Members of these families likely avoid expressing their feelings to each other, and thus they might not consider emotional support as a valuable type of support. Low expressive scores, as well as moderate conflict scores that characterize Cohesive-Nonexpressive families suggest that family members might not provide enough emotional support. This family environment type included 18.7% of the participants in this study. The family environment types identified and described in the present study are similar to the Cohesive-Authoritative, Conflictive-Authoritarian, and Defensive-Neglectful family types that Mandara & Murray (2002) identified among African American teenagers. The Cohesive-Authoritative family type, characterized by high levels of cohesion and expressiveness and low levels of conflict is similar to the

PAGE 71

63 Cohesive-Expressive family environment type described in the present study. High levels of internal conflict and disorganization of family relationships characterized the Conflictive-Authoritarian family type described by Mandara & Murray, analogous to the Conflictual-Expressive family environment type described in the present study. Finally, the Defensive-Neglectful family type associated with low family cohesion and expressiveness and high conflict would correspond to the Conflictual-Nonexpressive family environment type in the present study. The identification and description of family environment types can be useful for counselors working with women with breast cancer and their families. Specifically, the three proposed family environment types promote a conceptual understanding of the family relationships and dynamics that exist among the family members of women with breast cancer. Identification and understanding of the family environment type experienced by women with breast cancer can be helpful in developing family interventions to increase satisfaction with family support among women with breast cancer. The second research question asked whether these identified family environments are associated with women’s ethnicity, age, education, income level, and presence of a spouse/partner. Results using the combined data from African American and Caucasian American participants indicated significant differences between family environment types in association with education levels and presence versus absence of a spouse/partner. Specifically, women with breast cancer in Cohesive-Expressive families were more likely to have a graduate/professional education, women in Conflictual-Expressive families

PAGE 72

64 were more likely to have a high school education, and women in Conflictual-Nonexpressive families were more likely to have high school or some college/technical school education. Significant differences were also found regarding the presence/absence of a spouse or partner. Specifically, women in the Cohesive-Expressive families were more likely to have a spouse/partner present in the house, while women in the Conflictual-Expressive and Conflictual-Nonexpressive families were more likely to not have a spouse/partner present. The finding that there were no differences in the family environment types in association with age and ethnicity suggests that these family environment types may not be culture bound with regard to ethnicity and age. However, when the analyses for research question two were conducted separately by ethnic group different findings were obtained for the African American participants group versus the Caucasian American participants group. Results for the African American women participants group revealed no significant differences between family environment types in association with the demographic variables tested to answer research question two. For the Caucasian American women participants group, family environment types differed in association with family income level and spouse/partner presence status. Specifically, Caucasian American women with breast cancer in Cohesive-Expressive and Conflictual-Nonexpressive families were less likely to have lower family incomes, and Caucasian American women with breast cancer in Conflictual-Expressive families were more likely to have lower family incomes and they were also less likely to have higher family incomes. Additionally, a spouse/partner was less likely to be absent from Cohesive-Expressive families, while a spouse/partner was less likely to be present in Conflictual-Expressive families.

PAGE 73

65 In sum, significant differences were obtained between family environments in association with spouse presence status and education level (for all women participants) and in association with spouse presence and income level for Caucasian American women only. These differences among types of family environments provide empirically valid information about families of women with breast cancer. This empirical information suggests that special attention should perhaps be paid to these demographic variables when determining which women with breast cancer should be assessed more closely with regard to their family relationships, the impact of these relationships, and their family support needs. This assessment of family environments of women with breast cancer might also allow timely identification of those in need of family counseling and/or support group participation. The assessment of family environments might also facilitate identification of counseling and support interventions that are appropriate/helpful for these women. The third research question asked whether levels of perceived support and self-reported support satisfaction among African American women and Caucasian American women with breast cancer differ in association with family environment type (i.e., Cohesive-Expressive, Conflictual-Expressive, and Conflictual-Nonexpressive). For the African American women participants it was found that those from Cohesive-Expressive families were significantly more satisfied with the emotional support received from family members compared with women from Conflictual-Expressive families. Interestingly, no significant differences were found between the African American women participants from Cohesive-Expressive and Conflictual-Nonexpressive families regarding their satisfaction with emotional or any other type of support. For the

PAGE 74

66 Caucasian American women participants, the findings for research question three were as follows: (a) women from Conflictual-Expressive families perceived that they received significantly more tangible support from their family members compared with women from Conflictual-Nonexpressive families, (b) women from Cohesive-Expressive families perceived that they received more emotional support compared with women from Conflictual-Nonexpressive families, and (c) women from Conflictual-Expressive families perceived that they received significantly more emotional support than women from Conflictual-Nonexpressive families. Overall, the findings for research question three suggest that (a) African American women with breast cancer differ in their satisfaction with emotional support in association with their family environment type (i.e., Cohesive-Expressive versus Conflictual-Expressive), and (b) Caucasian American women with breast cancer differ in their perceptions of tangible and emotional family support in association with the type of their family environments (i.e., Conflictual-Expressive versus Conflictual-Nonexpressive for perceived tangible support, and Conflictual-Nonexpressive versus Cohesive-Expresive and Coflictual-Expressive for perceived emotional support). It is noteworthy to mention that the analysis of differences in family support perceived and in self-reported family support satisfaction in association with family environment types that was done separately by ethnicity revealed different findings by ethnicity. Specifically, it was found that levels of satisfaction with emotional support, but not perceptions of emotional or other examined support types were associated with differences in family environment types among African American women. In contrast, for Caucasian American women participants, perceptions of emotional support and

PAGE 75

67 tangible support were associated with family environment type. Furthermore, the Caucasian American women participants’ satisfaction with informational, tangible, or emotional support was not associated with family environment type. The meaning of these findings is not clear, but may help form directions for future research. It is noteworthy that these findings provide support for the use of the “difference model” research approach in studies investigating family support and family environments among ethnically diverse women with breast cancer. Overall, findings for research question three tend to validate the initial description of the three family environments obtained as a result of cluster analytic techniques. Specifically, women from Cohesive-Expressive families appear to perceive more support or be more satisfied with the support received from their family members compared with women from Conflictual-Nonexpressive or Conflictual-Expressive families. Additionally, emotional support appears to be the most significant form of family support in that satisfaction with emotional support and emotional support perceived were associated with family environment types among African American and Caucasian American women participants, respectively. The fourth research question tested if levels of satisfaction with family support can be significantly predicted from family cohesion, expressiveness, or conflict scores among African American women with breast cancer and Caucasian American women with breast cancer. Although data from African American women with breast cancer were very limited, significant results were obtained. These results indicated that among these women family conflict is a significant predictor of the satisfaction with overall support from family members. Specifically, African American women with breast cancer from

PAGE 76

68 more conflictual families were less satisfied with their overall perceived family support, and as family conflict increased by one unit, women’s satisfaction decreased by .31. However, because of the small sample size of the African American women participants in this study, these results should be viewed with caution. Limitations of the Research One limitation of this study was that the number of African American participants was disproportionately low, especially when compared with the number of Caucasian American participants. Special efforts were made to increase the participation of African American women with breast cancer, such as through the involvement of community member research participant recruiters. However, only 18 women in this ethnic category agreed to participate (versus 72 Caucasian American women). This low participation rate might be related to the low trust of researchers and research among African Americans (Mouton, Harris, Rovi, Solorzano, & Johnson, 1997). The view of breast cancer as a taboo topic of conversation among African Americans (Mouton et al., 1997) might also have contributed to their low participation in the present study. Moreover, the sample size of this study was relatively small and restricted to a particular geographical area (data from 97 participants were ultimately included in the statistical analyses and these participants were from Gainesville or Trenton, Florida). Given these characteristics of the sample, the present results have limited generalizability and should be viewed with caution. However, the sample size in this study is comparable to that of other exploratory studies involving women with breast cancer. Moreover, no results indicating specific differences in support needs of women with breast cancer by U.S. geographic region have been reported and the location of the participants in this study was not expected to influence the findings.

PAGE 77

69 This study used solely self-report measures of family relationships and family support, which might represent another limitation. However, this study’s focus solely on the perceptions of women with breast cancer was not accidental. Previous research findings indicate that subjective perceptions of the received support are better predictors of support satisfaction and health care outcomes compared with objective measures of received support. Since different women may assess differently similar levels of support or similar ways their family members interact with each other, subjective measures and self-reports are appropriate for the purpose of this study. Another possible limitation consists of the inclusion in this study of women with different treatment options, with different amounts of time since diagnosis, and who participated or not in a support group. As previous research findings suggest, these variables can influence the type and amount of support women need from their family members. The inclusion of women who were different on these variables was necessary in order to ensure the inclusion of an adequate number of participants in this study for performing the statistical analyses to investigate the research questions set forth. The potential influence of treatment type and support group participation status (i.e., participant versus non-participant) was controlled for in the statistical analyses. Another limitation of this study is related to the Index of Socially Supportive Behaviors (ISSB) as a measure of perceived and desired family support. Because ISSB was originally a measure of support obtained from various social sources, certain ISSB items were not applicable for some participants in this study (for example, “Loaned you under $25” as an item of the Tangible Support subscale). However, only a few participants reported that between 2 and 4 of the ISSB items did not reflect their

PAGE 78

70 experiences. Moreover, the method used to control for missing/unanswered items was computing scale scores as arithmetic means for all the answered items. Future research can prevent this limitation through the development of a measure of family support (versus social support) based on reports of ethnically diverse women with breast cancer as to what constitutes family support for them. Implications for Future Research Findings from this study have several implications in terms of directions for future research. First, the limitations of the present research need to be addressed in future investigations. Specifically, similar future research studies with larger samples and with a more representative sample of African American women with breast cancer are needed. Such research studies would allow a more adequate assessment of the impact that family variables have on the support perceptions and satisfaction among women confronting breast cancer in general, and among African American women with breast cancer in particular. Although past research explored family environment types among African Americans, no research until now explored family environment types among ethnically diverse women with breast cancer. The present paper argues not only for the inclusion of culture-related variables into further studies investigating the needs for family support of women with breast cancer, but it also argues for the use of the “difference model” research approach in future studies similar to the present study. The “difference model” research approach used in this paper led to the finding that similar family environment types that exist among African American and Caucasian American women with breast cancer may be differentially associated with perceived family support and family support satisfaction among these two groups of women. Specifically, family environment type

PAGE 79

71 was significantly associated with (a) support satisfaction levels versus perceptions of support among African American and Caucasian American women respectively, and (b) distinct dimensions of family support for each ethnic group separately. Moreover, preliminary findings from this study indicated that family conflict was significantly associated with satisfaction with overall family support among African American women with breast cancer but not among Caucasian American women with breast cancer. Another implication for future research is that it is necessary to take into consideration the influence of demographic variables when assessing family environment type experienced by women with breast cancer. For Caucasian American women with breast cancer in particular, income levels and presence of spouse/partner status need to be considered in association with family environment types. The present study proposes an exploratory research model of different types of family environments. Future research is necessary to validate the family environment types obtained in this study. Such research should include a larger number of participants in order to determine which family dimensions are significant predictors of the satisfaction levels experienced by ethnically diverse women with breast cancer. Furthermore, if the family types identified in the present study are validated by future studies, an assessment instrument can be developed to assist clinicians in their effort to identify women with breast cancer who might not be fully satisfied with their received support. Finally, it is necessary that future research further assess any links between the support needs of women with breast cancer and their health-related outcomes. The present study provided evidence for differences in support needs among African

PAGE 80

72 American and Caucasian American women with breast cancer, and yet it did not test the association between these support needs and health related outcomes such as distress levels, treatment adherence, distress or anxiety levels, level of well being, or survival time. Implications for the Field of Counseling Psychology One immediate implication of the findings of this study is that health care providers and counselors need to be aware of the differences in family environment types that exist among women with breast cancer. Specifically, the present study proposed a classification of family environments composed of three different types (i.e., Cohesive-Expressive, Conflictual-Expressive, and Conflictual-Nonexpressive). This family environment type taxonomy promotes an accessible conceptualization by practitioners of the communication patterns characteristic of families of women with breast cancer. Moreover, practitioners can easily identify which one of the three family environment types is characteristic to families they work with by using the FRI (Family Relationship Index), which is a popular and easy to use assessment instrument. Awareness regarding the existence of different family environment types is also important due to the association between these family environment types and (a) satisfaction with emotional support among African American women with breast cancer, and (b) perceived tangible and perceived emotional support among Caucasian American women with breast cancer. Specifically, based on findings from this study practitioners might assess more carefully the family support needs of African American women in Conflictual-Expressive families, since participants in this study from Conflictual-Expressive families were less satisfied with the emotional support perceived compared with women participants in Cohesive-Expressive families. Practitioners might also need

PAGE 81

73 to carefully assess family support needs of Caucasian American women with breast cancer in Conflictual-Nonexpressive families, since participants in this study from Conflictual-Nonexpressive families reported lower levels of perceived tangible and perceived informational support compared with women participants in Cohesive-Expressive families. Another implication for the field of counseling psychology is related to the finding that while African American women differed in their satisfaction with family support, Caucasian American women differed in their perceptions of support. Consequently, it may be important that interventions targeting family support for women with breast cancer focus more on level of satisfaction that African American women with breast cancer experience, and on perception of family support among Caucasian American women with breast cancer. Although of an exploratory nature due to the limited number of African American participants, results from analyses applied to the data from African American women with breast cancer who participated in this study are promising and relevant for the development of family intervention programs. Specifically, these results suggest that family intervention programs should be sensitive to the cultural background and support needs of women with breast cancer. In particular, family-counseling interventions could be aimed at helping teach families how to modify the communication between their members in order to decrease the conflict between family members and improve members’ ability to provide support to women with breast cancer. Conclusions This study primarily explored types of family environments and the support they offer to African American and Caucasian American women with breast cancer. Overall,

PAGE 82

74 findings from this study support the view that family environment types of women with breast cancer are different, and that there may be an association between family environment type and satisfaction with support or with perceived support. Moreover, the findings in this study indicate that overall, the women in Cohesive-Expressive family environments had higher education and were more likely to have a spouse/partner present in the house, and that women in Conflictual-Expressive and Conflictual-Nonexpressive family environments had lower education levels and were less likely to have a spouse/partner present in the house. Findings also indicated that Caucasian American women in Cohesive-Expressive families were more likely to have higher family incomes and to have a spouse/partner present, while women in Conflictual-Expressive families were less likely to have higher income levels and to have a spouse/partner present in the house and women in Conflictual-Nonexpressive families were less likely to have a spouse/partner in the house. These findings suggests that perhaps women with lower education levels and women who do not have a spouse/partner should be assessed more closely with regard to their family relationships and support needs from their family members. Also, Caucasian American women with breast cancer with lower income levels should be assessed more carefully. It is noteworthy that although there were no ethnic group differences associated with family environment type, findings from this study support the view that African American and Caucasian American women with breast cancer who report similar family environment types may have different family support needs and family support satisfaction. For example, presence of spouse/partner was associated with African American women’s perceptions of family support and with Caucasian American

PAGE 83

75 women’s satisfaction with family support. Moreover, perceptions of emotional support were associated with family environment type among African American women, but satisfaction with emotional and tangible support was associated with family environment types for Caucasian American women. A possible explanation of these findings is that African American women with breast cancer and Caucasian American women with breast cancer have a different set of expectations regarding the support their family members should offer to them. Clearly, however, support is provided for future culturally sensitive research with larger and more representative samples of women with breast cancer that examine (a) the existence of the family environment types found in this study, and (b) the association of these family environment types with the perceived family support and family support satisfaction experienced by African American women and Caucasian American women with breast cancer. Finally, findings in the present study provide support for the “difference model” research approach. When regression analyses were conducted for each ethnic group separately to test the associations between family environment dimensions (i.e., family cohesion, family expressiveness, and family conflict) and satisfaction with overall support, it was found that family conflict predicted overall satisfaction only for African American women with breast cancer but not for Caucasian American women with breast cancer. It is important to note that ethnically diverse families should not be reduced to strict categories or types, especially not on the basis of ethnicity of their members. Instead, ethnically diverse families should be understood through the lens of the embededness theory (Szapocznik & Kurtines, 1980). Specifically, individual families should be viewed

PAGE 84

76 in the context of their social environment with careful consideration given to the individual differences among family members.

PAGE 85

APPENDIX A RESEARCH QUESTIONNAIRE

PAGE 86

78 Continued on the next page

PAGE 87

79 Continued on the next page

PAGE 88

80 Continued on the next page

PAGE 89

81 Continued on the next page

PAGE 90

82 Continued on the next page

PAGE 91

83 Continued on the next page

PAGE 92

84

PAGE 93

APPENDIX B INVITATION LETTER FOR PACKETS

PAGE 94

June 2003 Dear Patient: We are inviting you to participate in our study of women’s experiences with breast cancer. This study will involve asking you about your needed family support, the way members of your family interact with each other, and your health care satisfaction. The findings from this study will provide important information that can be used in the future to better understand what family members can do to help women live well with breast cancer. Furthermore, results from this study may help other women who experience breast cancer receive more of the support they need from their families. If you decide to participate in our study, your participation will involve the following: Reading and signing an Informed Consent Form Completing the attached Research Questionnaire (It will take approximately 25 minutes to complete the Research Questionnaire.) Placing the signed Informed Consent Form and the completed Research Questionnaire in the boxes provided at your clinic. (You will be asked to place the Informed Consent Form in one box and the Research Questionnaire in the other box so that your confidentiality can be protected.) If you prefer to complete the Informed Consent Form and the Research Questionnaire on line, you may do so at the following address: http://grove.ufl.edu/~ancamp/ We hope that you will find this study interesting and valuable. We look forward to your participation. If you have any questions or if you would like further information, please contact Anca Mirsu-Paun or Dr. Carolyn Tucker at 392-0601 Ext. 260. Sincerely, Carolyn M. Tucker, Ph.D. Distinguished Alumni Professor Robert R. Carroll, M.D., P.A. Breast Oncology Specialist Anca Mirsu-Paun Graduate Student University of Florida Deborah Crom, R.N. Research Coordinator 86

PAGE 95

APPENDIX C INFORMED CONSENT FORM

PAGE 96

Dear Patient: I would like to invite you to participate in a research study on how satisfied women experiencing breast cancer are with the support being provided to them by members of their families. This research study is being conducted by Anca Mirsu-Paun, a graduate student in Counseling Psychology at the University of Florida, under the supervision of Dr. Carolyn M. Tucker. If you agree to participate, your participation will involve completing the Research Questionnaire attached to this Informed Consent Form. The Research Questionnaire contains questions about the following: (a) the support that you need and the support that you currently receive from your family, (b) the way your family members interact with each other, and (c) your overall satisfaction with the medical care you receive (or received) for cancer treatment at your health clinic. Completing the Research Questionnaire will take approximately 25 minutes. Although it is anticipated that the Research Questionnaire items will not be disturbing to you in any way, you do not have to answer any questions you do not wish to answer. All information that you provide will be anonymous and will be kept completely confidential. You will not have to write your name on the Research Questionnaire, and your signed Informed Consent will be collected in a separate box from the measures you complete. Your participation in this study is completely voluntary and you may decide to discontinue your participation at any time without consequence. Your decision to participate or not to participate in this research will not influence in any way your relationship with the health care providers at your health clinic. There are no anticipated risks nor any compensation or other direct benefits to you as a participant in this research study. However, it is expected that the results of this study can be used to help other women experiencing breast cancer receive more of the support they need from their families. If you have any questions about this research study, please contact Anca Mirsu-Paun, B.S., or Dr. Carolyn M. Tucker at (352) 392-0601 Ext. 260. Questions or concerns about your rights as a research participant may be directed to the UFIRB office, University of Florida, Box 112250, Gainesville, FL 32611 or at the phone number (352) 392-0433. Please sign below if you would like to participate in this study. I, _______________________________, voluntarily agree to participate in this research. (Print name) I have read the procedure described above and I have received a copy of this description. I voluntarily agree to participate in this research study. _______________________________________ ___________ (Signature of participant) Date 88

PAGE 97

APPENDIX D SCRIPT FOR CLINIC RECRUITMENT Dr. Carolyn Tucker and Ms. Mirsu-Paun, from the University of Florida, in collaboration with Dr. Carroll’s office, are conducting a study in order to identify what type of support women with breast cancer need from their family members and how much help they would like to receive. We hope that you will be willing to participate; if you would like to do so, please take a copy of the Research Questionnaire from this pile [indicating the pile of Research Questionnaires]. Please note that you can call the number specified in the Invitation Letter if you would prefer to have the questionnaire read to you, or if you have any questions. Please note that your participation will be anonymous or confidential, and your name will not be placed on the Research Questionnaire. In case you do not have enough time to finish answering all the questions today, please take the Research Questionnaire home and return it no later than [date will be specified after consulting with somebody at the clinic]. Thank you! 89

PAGE 98

APPENDIX E SCRIPT FOR SUPPORT GROUP RECRUITMENT Dr. Carolyn Tucker and Ms. Mirsu-Paun, from the University of Florida, are conducting a study in order to identify what type of support women with breast cancer need from their family members and how much help they would like to receive. We hope you will be willing to participate; if you would like to do so, you are invited to answer the Research Questionnaire on-line, but a few hard copies are available as well. The Internet address where the Research Questionnaire is posted is specified in the Invitation Letter that I will give you in a moment. Hard copies of the Research Questionnaire are placed here, in this pile. If you would like to get a hard copy, please take one with you at the end of our meeting and complete it at your earliest convenience in the following week. You can return the completed Research Questionnaire using the stamped envelope, which has been provided. Please note that your participation will be anonymous or confidential, and your name will not be placed on the Research Questionnaire. The time frame for answering the Research Questionnaire (via the Internet or hard copy) is one week. Please note that you can call the number specified in the Invitation Letter if you would prefer to have the questionnaire read to you, or if you have any questions. Thank you! 90

PAGE 99

APPENDIX F INVITATION LETTER FOR WEB

PAGE 100

June 2003 Dear Support Group Participant: We are inviting you to participate in our study of women’s experiences with breast cancer. The study will involve asking you about your needed family support, the way members of your family interact with each other, and your health care satisfaction. The findings from this study will provide important information that can be used in the future to better understand what family members can do to help women live well with breast cancer. Furthermore, results from this study may help other women who experience breast cancer receive more of the support that they need from their family members. If you decide to participate in our study, your participation will involve the following: Logging on the Internet to the following address: http://grove.ufl.edu/~ancamp/ Reading the Informed Consent Form and checking the appropriate box to indicate your intention to participate in this study (Please print a copy of the Informed Consent Form and keep it for your personal records.) Completing the Research Questionnaire (It will take approximately 25 minutes to complete the Research Questionnaire.) If you prefer to complete printed copies of the Informed Consent Form and of the Research Questionnaire, you can obtain copies to complete from your support group coordinator. In this case you can return your completed Informed Consent Form and the Research Questionnaire to us in the stamped pre-addressed envelope attached to the Research Questionnaire. We hope that you find this study interesting and valuable. We look forward to your participation. If you have any questions or if you would like further information, please contact Anca Mirsu-Paun or Dr. Carolyn Tucker at 392-0601 Ext. 260. Sincerely, Carolyn M. Tucker, Ph.D. Distinguished Alumni Professor University of Florida Anca Mirsu-Paun Graduate Student University of Florida 92

PAGE 101

APPENDIX G INFORMED CONSENT FORM FOR WEB Dear Patient: I would like to invite you to participate in a research study on how satisfied women experiencing breast cancer are with the support being provided to them by members of their families. This research study is being conducted by Anca Mirsu-Paun, a graduate student in Counseling Psychology at the University of Florida, under the supervision of Dr. Carolyn M. Tucker. If you agree to participate, your participation will involve completing the Research Questionnaire on this Web page. The Research Questionnaire contains questions about the following: (a) the support that you need and the support that you currently receive from your family, (b) the way your family members interact with each other, and (c) your overall satisfaction with the medical care you receive (or received) for cancer treatment at your health clinic. Completing the Research Questionnaire will take approximately 25 minutes. Although it is anticipated that the Research Questionnaire items will not be disturbing to you in any way, you do not have to answer any questions you do not wish to answer. All information that you provide will be anonymous and will be kept completely confidential. You are not required to provide your name anywhere on this web page. Your participation in this study is completely voluntary and you may decide to discontinue your participation at any time without consequence. There are no anticipated risks nor any compensation or other direct benefits to you as a participant in this research study. However, it is expected that the results of this study can be used to help other women experiencing breast cancer receive more of the support they need from their families. If you have any questions about this research study, please contact Anca Mirsu-Paun, B.S., or Dr. Carolyn M. Tucker at (352) 392-0601 Ext. 260. Questions or concerns about your rights as a research participant rights may be directed to the UFIRB office, University of Florida, Box 112250, Gainesville, FL 32611 or at the phone number (352) 392-0433. Please print this page and keep it with your records for future reference. 93

PAGE 102

94

PAGE 103

LIST OF REFERENCES Alferi, S.M., Carver, C.S., Antoni, M.H., Weiss, S., & Duran, R.E. (2001). An exploratory study of social support, distress, and life disruption among low-income Hispanic women under treatment for early stage breast cancer. Health Psychology, Vol. 20, No. 1, 41. American Cancer Society Cancer Facts and Figures 2002. Web page http://www.cancer.org/downloads/STT/Cancer%20Facts%20&%20Figures%202002%20RevCovP6P7TmPWSecured.pdf. Last accessed on 01/14/2004. Andersen, S.A. & Sabatelli, R.M. (1999). Family interaction: A multigenerational developmental perspective (2 nd edition). Boston, MA: Allyn & Bacon. Bloom, J. (2000). The role of family support in cancer control. In Baider, L., Cooper, C.L. & Kaplan De-Nour, A. (Eds.), Cancer and the family, 2 nd edition (pp. 55-72). West Sussex, England: John Wiley & Sons, Ltd. Bloom, J.R., & Kessler, L. (1994). Emotional support following cancer: A test of the stigma and social activity hypotheses. Journal of Health and Social Behavior, 35(2), 118-133. Bloom, J.R., & Spiegel, D. (1984). The relationship of two dimensions of social support to the psychological well-being and social functioning of women with advanced breast cancer. Social Science and Medicine, 19(8), 831-837. Boss, P. & Mulligan, C. (2003). Family stress. Classic and contemporary readings. Thousand Oaks, CA: Sage Publications, Inc. Bourjolly, J.N. & Hirschman, K.B. (2001). Similarities in coping strategies but differences in sources of support among African American and White women coping with breast cancer. Journal of Psychosocial Oncology, Vol. 19, No. 2, 17–37. Carey, P.J., Oberst, M.T., McCubbin, M.A., & Hughes, S.H. (1991). Appraisal and caregiving burden in family members caring for patients receiving chemotherapy. Oncology Nursing Forum, 18(8), 1341-1348. Carlson, L., Bultz, B.D., Speca, M. & Pierre, M.St. (2000). Partners of cancer patients. Part I. Impact, adjustment, and coping across the illness trajectory. Journal of Psychosocial Oncology, Vol. 18(2), 39. 95

PAGE 104

96 Carlson, L., Bultz, B.D., Speca, M. & Pierre, M.St. (2000). Partners of cancer patients. Part II. Current psychosocial interventions and suggestions for improvement. Journal of Psychosocial Oncology, Vol. 18(3), 33. Cheatham, H., D’Andrea, M., Ivey, A.E., Bradford Ivey, M., Pedersen, P., Rigazio-DiGillio, S., Simek-Morgan, L., & Sue, D.W. (2002). Multicultural Counseling and Therapy I: Methateory – Taking Theory into Practice. In: Ivey, A.E., D’Andrea, M., Bradford Ivey, M., & Simek-Morgan, L. (Eds.), Theories of Counseling and Psychotherapy. A Multicultural Perspective. 5 th edition. (pp. 291-325). Boston, MA: Allyn & Bacon Inc. Conoley, J.C.& Buterick Werth, E. (1995). Family assessment. Lincoln, NE: Buros Institute for Mental Measurements. 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 interactional view (pp. 129-140). New York: Wiley. Cutrona, C. E. (2000). Social support principles for strengthening families. In J. Canavan, P. Dolan, & J. Pinkerton (Eds.), Family support in disadvantaged families (pp. 103). Dublin, Ireland: Routledge. Cutrona, C.E. (1996). Social support in couples. London: Sage Publications, Inc. Dakof, G.A. & Taylor, S.E. (1990). Victims’ perceptions of social support: What is helpful from whom? Journal of Personality and Social Psychology, Vol. 58, No. 1, 80. Davis, M.C., Tucker, C.M. & Fennell, R.S. (1996). Family behavior, adaptation, and treatment adherence of pediatric nephrology patients. Pediatric Nephrology, 10, 160. Davis, M.H., Morris, M.M., & Kraus, L.A. (1998). Relationship-specific versus global perceptions of social support: associations with well-being and attachment. Journal of Personality and Social Psychology, 74(2), 468-481. Davis-Ali, S.H., Chesler, M.A. & Chesney, B.K. (1993). Recognizing cancer as a family disease: Worries and support reported by patients and spouses. Social Work in Health Care, Vol. 19(2), 44. Dunkel-Schetter, C., Feinstein, L.G., & Taylor, S.E. (1992). Patterns of coping with cancer. Health Psychology, 11(2), 79-87. Dunn, C.W. & Tucker, C.M. (1993). Black children’s adaptive functioning and maladaptive behavior associated with quality of family support. Journal of Multicultural Counseling and Development, Vol. 21, 79.

PAGE 105

97 Friedman, L.C., Webb, J.A., Weinberg, A.D., Lane, M., Cooper, H.P., & Woodruff, A. (1995). Breast cancer screening: Racial/ethnic differences in behaviors and beliefs. Journal of Cancer Education, 10(4), 213-216. Funch, D.P., & Mettlin, C. (1982). The role of support in relation to recovery from breast surgery. Social Science and Medicine, 16(1), 91-98. Gotay, C.C. (2000). Culture, cancer, and the family. In: Baider, L., Cooper, C.L. & Kaplan De-Nour, A., Cancer and the family, 2 nd edition (pp. 95-110). West Sussex, England: John Wiley and Sons, Ltd. Grotevant, H.D. & Carlson, C.I. (1989). Family assessment: A guide to methods and measures. New York: Guilford Press, Gurowka, K.J., & Lightman, E.S. (1995). Supportive and unsupportive interactions as perceived by cancer patients. Social Work in Health Care, 21(4), 71-88. Hagerdoon, M., Kuijer, R.G., Buunk, B.P., DeJong, G.M., Wobbes, T., & Sanderman, R. (2000). Marital satisfaction in patients with cancer: does support from intimate partners benefit those who need it the most? Health Psychology, Vol. 19, No. 3, 274. Hill, K.M., Amir, Z., Muers, M.F., Connolly, C.K., & Round, C.E. (2003). Do newly diagnosed lung cancer patients feel their concerns are being met? European Journal of Cancer Care, 12(1), pp. 35-45. Hobfoll, S.E., & Vaux, A. (2003). Social support and social context. In L. Goldberger & S. Breznitz (Eds.), Handbook of stress: Theoretical and clinical aspects, 2 nd edition (pp. 685-705). New York: Wiley. Holahan, Ch.J. & Moos, R. (1981). Social support and psychological distress: A longitudinal analysis. Journal of Abnormal Psychology, Vol. 90, No. 4, 365. Holland, J.C. (1998). Psycho-Oncology. New York: Oxford University Press, Inc. Kagawa-Singer, M. & Nguyen, T. (2000). A cross-cultural comparison of social support among Asian-American and Euro-American women following breast cancer. In: Baider, L., Cooper, C.L. & Kaplan De-Nour, A. (2000). Cancer and the family, 2 nd edition (pp. 73-94). West Sussex, England: John Wiley and Sons, Ltd. Kissane, D.W., Bloch, S., McKenzie, M., Mcdowall, A.C., & Nitzan, R. (1998). Family grief therapy: a preliminary account of a new model to promote healthy family functioning during palliative care and bereavement. Psycho-Oncology, No. 7(1), 14. Krishnasamy, M. (1996). Social support and the patient with cancer: a consideration of the literature. Journal of Advanced Nursing, 23, 757-762.

PAGE 106

98 Maccormack, T., Simonian, J., Lim, J., Remond, L., Roets, D., Dunn, S., & Butow, P. (2001). ‘Someone who cares:’ a qualitative investigation of cancer patients’ experiences of psychotherapy. Psycho-Oncology, No. 10(1), 52. Mandara, J. & Murray, C.B. (2002). Development of an empirical typology of African American family functioning. Journal of Family Psychology, 16(3), 318-337. Manne, S. & Glassman, M. (2000). Perceived control, coping efficacy, and avoidance coping as mediators between spouses’ unsupportive behaviors and cancer patients’ psychological distress. Health Psychology, Vol. 19, No. 2, 155. Manne, S. & Schnoll, R. (2001). Measuring cancer patients’ psychological distress and well-being: A factor analytic assessment of the Mental Health Inventory. Psychological Assessment, 13(1), 99-109. Manne, S.L., Taylor, K.L., Dougherty, J., & Kemeny, N. (1997). Supportive and negative responses in the partner relationship: their association with psychological adjustment among individuals with cancer. Journal of Behavioral Medicine, Vol. 20, No. 2, 101. McCubbin, M.A. (1988). Response to “Measures of stress and related constructs”. Scholarly inquiry for nursing practice, 2(1), 71-75. Minuchin, S. (1974). Families and family therapy. Cambridge, MA: Harvard University Press. Moos, R., & Moos, B. (1986). Family Environment Scale manual, 2 nd edition. Palo Alto, CA: Consulting Psychologists Press. Mouton, C.P., Harris, S., Rovi, S., Solorzano, P., & Johnson, M.S. (1997). Barriers to black women’s participation in cancer clinical trials. Journal of the National Medical Association, 89(11), 721-727. Navarre, S.E. (1998). Salvador Minuchin’s structural family therapy and its application to multicultural family systems. Issues in Mental Health Nursing, 19(6), 557-570. Neuling, S.J. & Winefield, H.R. (1988). Social support and recovery after surgery for breast cancer: Frequency and correlates of supportive behaviors by family, friends, and surgeon. Social Science and Medicine, 27(4), 385-392. Northhouse, L.L., Templin, T., Mood, D. & Oberst, M. (1998). Couples’ adjustment to breast cancer and benign disease: a longitudinal analysis. Psycho-Oncology 7, 37–48. Oyemade, U., & Rosser, P. (1980). Development in Black children. Advances in Behavioral Pediatrics, 1, 153-179.

PAGE 107

99 Patterson, J.M. & Garwick, A.W. (1994). The impact of chronic illness on families: a family systems perspective. Annals of Behavioral Medicine 16(2), 131. Pinsof, W.M. (1992). Toward a scientific paradigm for family psychology: The integrative process systems perspective. Journal of Family Psychology, 5(3), 432-47. Primomo, J., Yates, B.C., & Woods, N.F. (1990). Social support for women during chronic illness: The relationship among sources and types to adjustment. Research in Nursing and Health, 13, 153-161. Ptacek, J.T., Pierce, G.R., Dodge, K.L. & Ptacek, J.J. (1997). Social support in spouses of cancer patients: What do they get and to what end? Personal relationships, 4, 431–449. Rigazzio-DiGillio, S. (2002). Family counseling and therapy: Theoretical foundations and issue of practice. In H. Cheatham, M. D’Andrea, A.E. Ivey, I.M. Bradford, P. Pedersen, S. Rigazio-DiGillio, L. Simek-Morgan, & D.W. Sue (Eds.), Theories of counseling and psychotherapy. A multicultural perspective (5 th edition) (pp. 390-428). Boston, MA: Allyn & Bacon. Schaefer, C., Coyne, J.C. & Lazarus, R.S. (1981). The health-related functions of social support. Journal of Behavioral Medicine, Vol. 4, No. 4, 381. Sciarra, D.T. (1996). Assessment of diverse family systems. In: Suzuki, L.A., Meller, P.J. & Ponterotto, J.G. (Eds.), Handbook of multicultural assessment. Clinical, psychological, and educational implications (pp. 135-168). San Francisco: Jossey-Bass Publishers. Shankar, A.D. (1997). Pilot study: The influence of cultural components on breast cancer patients’ perception of their support system. The Howard Journal of Communications, No. 8, 101. Sormanti, M. & Kayser, K. (2000). Partner support and changes in relationships during life-threatening illness: Womens’ perspectives. Journal of Psychosocial Oncology, Vol. 18, No. 3, 45-66. Spiegel, D., Bloom, J.R. & Gottheil, E. (1983). Family environment as a predictor of adjustment to metastatic breast carcinoma. Journal of Psychosocial Oncology, Vol. 1(1), 33. Spinetta, J. (1984). Measurement of family function, communication, and cultural effects. Cancer, Vol. 53, 2330. Stokes, J.P., & Wilson, G.D. (1984). The Inventory of Socially Supportive Behaviors: Dimensionality, prediction, and gender differences. American Journal of Community Psychology, 12(1), 53-69).

PAGE 108

100 Szapocznik, J. & Kurtines, W. M. (1993). Family psychology and cultural diversity: Opportunities for theory, research, and application. American Psychologist, 48(4), 400-407. Szapocznik, J., Perez-Vidal, A., Brickman, A. L., Foote, F. H., Santisteban, D., & Hervis, O. (1988). Engaging adolescent drug abusers and their families in treatment: A strategic structural systems approach. Journal of Consulting and Clinical Psychology, 56(4), 552-557. Thomas, C., Morris, S.M. & Harman, J.C. (2002). Companions through cancer: the care given by informal carers in cancer contexts. Social Science and Medicine, 54, 529 –544. Tracy, E.M. & Whittaker, J.K. (1990). The social network map: assessing social support in clinical practice. Families in Society: The Journal of Contemporary Human Services, 461. Walter, B., Figgs, L.W., & Zahm, S.H. (1995). Differences in cancer incidence, mortality, and survival between African Americans and Whites. Environmental Health Perspectives, 103(8), 275-281. Weihs, K. & Reiss, D. (2000). Family reorganization in response to cancer: a developmental perspective. In: Baider, L., Cooper, C.L. & Kaplan De-Nour, A. (2000). Cancer and the family (2 nd edition). West Sussex, England: John Wiley and Sons, Ltd. Wellisch, D., Kagawa-Singer, M., Louise Reid, S., Lin, Y., Nishikawa-Lee, S., & Wellisch, M. (1999). An exploratory study of social support: a cross-cultural comparison of Chinese-, Japanese-, and Anglo-American breast cancer patients. Psycho-Oncology 8: 207.

PAGE 109

BIOGRAPHICAL SKETCH Anca Mirsu-Paun was born in Romania, where she lived until the age of 23 years. She graduated with a B.S. from the University of Bucharest, with a major in psychology. She entered a master’s program in psychotherapy at the University of Bucharest, and she concomitantly worked as a research assistant for the Institute for Research of Youth Issues, affiliated with the Romanian Ministry of Youth and Sports. She then entered the doctoral program in counseling psychology at the University of Florida. Anca is currently doing research under the supervision of Dr. Carolyn M. Tucker, her advisor, for a project on culturally sensitive healthcare. She anticipates receiving her Ph.D. in 2006. Anca is the recipient of the Outstanding International Student Award from the College of Liberal Arts and Sciences of the University of Florida. 101