1 AFRICANAMERICAN WOMENS BELIEFS ABOUT THE INTERNET AS A SOURCE OF BREAST CANCER PREVENTION INFORMATION By HEATHER M. EDWARDS A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILL MENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2009
2 2009 Heather M. Edwards
3 ACKNOWLEDGMENTS I wish to thank my dissertation committee members collectively for those characteristics which they have in common as well as their distinct gifts and attributes that enhanced my studies at the University of Florida. All four of these women impressed me with their abilities to see relevance of research questions across disciplines and to envision how all of my academic studies could culminate in this project. I count myself fortunate to have learned under these women whose abilities to discuss their areas of expertise in such engaging ways set a standard that I one day hope to reach. I thank Dr. Kim Walsh Childers, my thesis and dissertation chair and mentor, for recruiting me into this program and offering me a great opportunity to work on her research grant. Her willingness to listen to my ideas and offer direction without being dismissive was inte gral to my ability to think in depth about the complexities of communications and society. Dr. Debbie Treise prompted me to make decisions and set realistic expectations about the dissertations projects parameters. She graciously allowed me to work as he r graduate assistant during the second half of my doctoral program, teaching me about the considerations and processes of seeing a project from start to end. Dr. Youjin Choi taught my public communications campaigns, a class that left such an impression I hope to pursue a career in this area. She has provided careful feedback that always required me to rethink my conceptualization of the studys constructs and has made this project much more comprehensive. I thank Dr. Barbara Curbow, my MPH program advisor, whose public health communication class and encouragement focused my interest on melding mass communication and public health perspectives into an area of study.
4 I am very grateful to my grandmother for her encouragement She was an extraordinary help wi th several projects during my doctoral program, including helping collect back issues of magazines and identifying resources for my dissertation. My parents have lent unquantifiable amounts of love, support and encouragement over the years I thank them for their sacrifices for my education, their examples of perseverance, and their efforts to infuse my most stressful days with humor and levity. I wish to thank all of the people who helped identify recruitment channels for my dissertation. Sincerest thanks go to the women who participated in interviews and surveys. Their willingness to be interviewed and surveyed about this personal health subject was the foundation of this projects success.
5 TABLE OF CONTENTS page ACKNOWLEDGMENTS ...............................................................................................................3 LIST OF TABLES ...........................................................................................................................7 LIST OF FIGURES .........................................................................................................................9 ABSTRACT ...................................................................................................................................10 CHAPTER 1 INTRODUCTION ..................................................................................................................12 Breast Cancer & Prevention ...................................................................................................14 Breast Can cer Risk Factors .............................................................................................16 History of the Cause .....................................................................................................18 Cultural Sensitivity .................................................................................................................25 The Internet & Technology ....................................................................................................27 New Media Communication Theories ....................................................................................31 2 LITERATURE REVIEW .......................................................................................................38 The Breast Cancer Experience in African American Women ...............................................38 Health Beliefs and Beliefs about Breast Cancer in African American Culture .....................41 Health Literacy .......................................................................................................................45 Ecological Studies of InformationSeeking ............................................................................49 Breast Cancer Information Seeking .......................................................................................53 Obstacles to Use ..............................................................................................................59 Minority Use of the Internet ............................................................................................60 E ffects of Internet Use on Breast Cancer Information Seekers .......................................61 Summary and Research Questions .........................................................................................61 3 QUALITATIVE METHODS .................................................................................................65 Qualitative Research ...............................................................................................................65 Symbolic Interactionism ..................................................................................................67 Appropriateness of Qualitative Design to the Study Topic .............................................68 Grounded Theory .............................................................................................................70 Mixed Methods .......................................................................................................................74 Research Design .....................................................................................................................77 Interviews ........................................................................................................................78 Theoretical Sensitivity .....................................................................................................81 Interview Method ............................................................................................................82 4 QUALITATIVE RESULTS ...................................................................................................84
6 Summary of Interviews ...........................................................................................................84 Interview Themes ...................................................................................................................85 Community Beliefs and Interpersonal Communication about Breast Cancer .................99 Prevention Informati on and Internet Beliefs .................................................................108 5 SURVEY METHODS ..........................................................................................................128 Role of the Survey in the Study ............................................................................................128 Survey Development ............................................................................................................131 Questionnaire Format and Pretest .........................................................................................134 Survey Questions ..................................................................................................................135 General Questions .........................................................................................................138 Theme Questions ...........................................................................................................142 Survey Sampling ...................................................................................................................150 Survey Analysis ....................................................................................................................154 6 SURVEY RESULTS ............................................................................................................160 Survey Results ......................................................................................................................160 Integration of Interview and Survey Results ........................................................................175 7 DISCUSSION .......................................................................................................................180 Summary of Major Themes ..................................................................................................184 Themes Unrelated to Research Questions ............................................................................193 Practical Implications of the Study .......................................................................................204 Study Limitations ..................................................................................................................205 Future Research ....................................................................................................................211 APPENDIX A INTERVIEW PROTOCOL SUBMITTED TO THE INSTITUTIONAL REVIEW B OARD ................................................................................................................................217 B INTERVIEW GUIDE ...........................................................................................................219 C SURVEY PROTOCOL SUBMITTED TO THE INSTITUTIONAL REVIEW BOARD ..221 D SURVEY INSTRUMENT ....................................................................................................223 LIST OF REFERENCES .............................................................................................................230 BIOGRAPHICAL SKETCH .......................................................................................................248
7 LIST OF TABLES Table page 41 Women interviewed and their characteristics ....................................................................84 61 Survey respondents ag e, education level, income, and relationship status ....................160 62 Frequency of perceived risk factors for breast cancer in total sample and by family history ..............................................................................................................................161 63 Descriptive statistics for beliefs about perception of personal risk and beliefs about breast cancer prevention ..................................................................................................162 64 Correlations between age, education, and income, and beliefs about personal risk of breast cancer and beliefs about preventi on ( rs ) ...............................................................163 65 Problems facing respondents communities, by number and percent of women who cited the p roblems as top three priorities .........................................................................164 66 Use of the Internet for breast cancer information ............................................................165 67 Womens sources of health information ..........................................................................165 68 Number and percent of women who said they found out about health topics from each source of information, by family history of breast cancer .......................................166 69 Amount of time women reported to spend on the Internet in an average week ..............167 610 Reasons given for not using the Internet by women who reported spending no time using it in a typical week .................................................................................................168 611 Number and percent of women who use the Internet for each activity ...........................169 612 De scriptive statistics for belief items about using the Internet for breast cancer information .......................................................................................................................170 613 Correlations of age, education, and income with Internet beliefs ( rs ) .............................170 614 Percentages of women reporting specific reasons they would use the Internet to find breast cancer prevention information ...............................................................................172 615 Signif icant relationships between reasons for potential Internet use for prevention information and demographic variabl 2 ) ....................................................................172 616 Number and percent of women who indicated they would want to learn about each breast cancer topic, by women with and without a family history of the disease ............173
8 617 Reasons why women would not use the Internet to find breast cancer prevention information .......................................................................................................................174 71 Comparisons between the HBM and t he studys findings ...............................................191 72 Potential measures in future research about Internet health informationseeking and health prevention ..............................................................................................................201
9 LIST OF FIGURES Figur e page 71 Antecedents for online breast cancer prevention information seeking ............................186
10 Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy AFRICANAMERICAN WOMENS BELIEFS ABOUT THE INTERNET AS A SOURCE OF BREAST CANCER PREVENTION INFORMATION By Heather M. Edwards Aug ust 2009 Chair: Kim Walsh Childers Major: Mass Communication Although African American women are less likely to be diagnosed with breast cancer than Caucasian women, the mortality rate for African American women with breast cancer is higher. Part of the effort to reduce this disparity includes education about prevention and early detection. The Internet has become a popular source of information for people who seek to know more about health topics. The medium can be a vehicle to disseminate general preve ntion information to the public and tailored information to subgroups. However, the literature reveals little about if, why, and how AfricanAmerican women who do not have breast cancer seek out prevention information. This project is a mixed methods stu dy of African American womens beliefs about breast cancer prevention and attitudes about looking for prevention information on the Internet Specifically, the research sought to discover if beliefs about the disease or beliefs about the medium related to a womans online prevention informationseeking. The study included indepth interviews with nine African American women without personal histories of breast cancer and 160 written surveys developed based on the interview results. V ariables of particular i nterest were family history of breast cancer and Internet use.
11 Participants included a mix of women with and without family histories of breast cancer, as well as low and high users of the Internet Inferences about the relationship s between these variabl es were based on analyses of both qualitative and quantitative data. Internet nonuse seemed to be more related to physical access to the medium than to beliefs about the medium. Seeking behavior among women with Internet access was related to beliefs abou t personal risk of the disease. Results indicate that family history is overwhelmingly thought to put a woman at risk for developing breast cancer. Women who do not have this family history may be unlikely to seek information about preventing the disease. The implications of this finding are that breast cancer primary prevention interventions targeting women without family histories of breast cancer should aim to increase womens knowledge about the diseases risk factors and emphasize the womens need to be proactive in prevention behaviors
12 CHAPTER 1 INTRODUCTION Healthy People 2010 ( United States Department n .d.), the federal governments written plan for improving the nations health and setting goals to achieve by the year 2010, recognizes that the h ealth of citizens and residents of the United States has been improving or maintaining for some subgroups in the population while others lag behind in health status. The document points out the need to reduce disparities that exist in the health of differe nt demographic groups. Breast cancer among African American women represents one health disparity. Despite the fact that, in general, breast cancer mortality rates are relatively low compared to those of other cancers, there are race differences in death r ates from the disease. African Americans are less likely to be diagnosed with breast cancer, but those with the disease are more likely to die from it than their white peers. The Healthy People 2010 objective is to reduce age adjusted mortality due to brea st cancer in African American women from 35.1 (per 100,000) in the baseline year of 1999 to 21.3 by 2010 ( Centers for Disease Control 2007). In 2005, the half way mark of Healthy People 2010 and the most recent year for which data was found, mortality rat es for African American women with breast cancer were slightly lower at around 33 (per 100,000) than the baseline but not close to the targeted rate (CDC, 2009a ) While science has not yielded conclusive answers as to what causes breast cancer in its many forms, there are some identified risk factors that women should know, including inherited genetic and behavioral characteristics. Some of these factors are characteristics over which individuals exercise little control. However, some research suggests tha t overall healthy behaviors may have a protective effect, including not smoking, limiting alcohol consumption, exercising regularly and eating a nutritious diet (McPherson, Steel & Dixon, 2000).
13 A growing body of research now demonstrates that the majorit y of American adults use the Internet (Fox, 2006), and one of the most common online activities is seeking information about health, including chronic diseases such as breast cancer. At the time of this papers writing, the medical community c ould not tell women what women can do to prevent breast cancer; there are no clear answers. However, sciences lack of definitive answers likely will not forestall peoples efforts to seek information on ways to prevent breast cancer. Even though the present study s purpose is not to investigate factors believed to cause breast cancer, including those related to unhealthy behaviors (obesity and overweight, tobacco consumption, etc.), it does suggest that having more people engage in positive behaviors will reduce the ef fect of breast cancer (Miller, 1995). In addition, although screenings for breast cancer, including mammography, are not considered primary prevention methods, mammography and other secondary prevention methods are often discussed as prevention options for women with no personal histories of breast cancer The purpose of this study, therefore, will be to add to our understanding of the social, cultural, and even individual level determinants of whether or not a woman uses the Internet to seek breast cancer information. An ultimate goal of this line of research would be to find out more about the most likely users of an online breast cancer information resource so that Web sites and other interventions can be developed to serve these women most effectively. Using in depth interviews and surveys the project explore d womens beliefs about breast cancer, their intentional seeking of breast cancer prevention information from any source, their attitudes about and use of the Internet in general, and, linking the se concepts, the womens attitudes about and use of the Internet for breast cancer prevention information. The phrase beliefs about the Internet will be used throughout this paper to refer to womens beliefs about the value of the Internet as a health info rmation source as well as beliefs about their ability to use the Internet
14 successfully to obtain information. The latter is a form of self efficacy, which Bass and colleagues refer to as confidence in being able to complete a task ( 2006, p. 221). Primary and secondary breast cancer prevention are both considered in this studys literature. Other screening methods such as genetic testing could be included as factors in secondary prevention because some participants had family histories of breast cancer ( or as pr imary prevention, if women opt for prophylactic mastectomies based on genetic tendencies toward the disease) but for this study mammography, as a very common and emphasized screening for women older than age 40 ( American Cancer Society, 2007d ) w as the main representation of secondary prevention. The remainder of this chapter will discuss the epidemiology of breast cancer, the different levels of breast cancer prevention, and how breast cancer has become a cause, therefore gaining a great deal o f media attention over the years. Chapter 1 also introduces Newhagen and Bucys (2004) new media content access model which organizes this studys exploration and explains the studys research questions. In Chapter 2, prior research is presented about the health beliefs and potential barriers to enacting healthful behaviors in some African American cultures and communities. Relevant literature about health and breast cancer informationseeking, as well as Internet use in general, is reviewed. Chapter 3 describes the methods used for analyzing the relationship between African American womens individual characteristics, their use of the Internet in general, and their use of the Internet for breast cancer prevention information. The findings of the in depth interviews and surveys used for data collection are discussed in Chapter 4, and Chapter 5 reports on and discusses the study conclusions. Breast Cancer & Prevention Of all diseases that affect women, breast cancer is most feared, even though the disease c laims fewer womens lives than does heart disease (Wider, 2005). Breast cancer is not even the
15 most deadly cancer among women; that distinction belongs to lung cancer ( ACS, 2007a ). However, a Google search for lung cancer produces 2.3 million results, compared to the 7 million Web pages found in a search for breast cancer Perhaps this is a reaction to demand: searches for breast cancer information are more common than searches for lung and prostate cancer information (Metz, Devine, et al., 2003). Media, gr ass roots movements, and the fact that breast cancer affects women in an intimate way also may contribute to the diseases inflated attention. As breast cancer continues to occupy a prominent place in medical research and funding advocacy, it will remain a popular topic for Web sites. The American Cancer Societys forecast ranked breast cancer as the number one site of new cancer cases in women in 2007 and the second most deadly ( ACS, 2007a) In African American women specifically, breast cancer was predicted to account for 27% of new cancer cases in 2007 and 19 % of deaths from cancer ( ACS, 2007b) Although deaths from breast cancer fell slightly among white women from 1975 2003, for African American women, mortality rates from the disease were a few percen tage points higher in 2003 than in 1975. Breast cancer has one of the highest fiveyear survivorship rates of all cancers, likely the result of awareness efforts and early detection initiatives. The United States showed a 33.8 (per 100,000 people) death r ate of African American women from breast cancer during the 2000 2004 period ( ACS 2008); among African American women in Florida the rate was slightly better at 30.7 ( ACS 2007b). In comparison, the death rate nationwide (per 100,000 people) among White women for the same period was 25 ( ACS 2008). Although the term breast cancer is commonly used in the singular, in reality it represents different types of cancer that affect the breast. The anatomical site is the same for these diseases, but their etiol ogies, manifestations, and treatments vary. Breast cancers develop when cells in the
16 breast begin to grow abnormally. Usually cells in the human body go through a cycle of dying and being replaced by new cells that know how to function properly in their pl aces. However, cells that grow incorrectly form tumors, and malignant tumors that begin to spread, or metastasize to surrounding tissues threaten to interfere with vital body functions. This is, in essence, breast cancer ( National Cancer Institute 2005b) Breast cancer is much more common in women than in men ( ACS 2008), which is why women have rallied to support research to eradicate the physical and emotional suffering caused by the disease. Breast Cancer Risk Factors There are many identified risk factors for developing breast cancer, some of which are related to controllable behaviors and lifestyle choices and others that are associated with inherited genes. As Teolis points out, however, being female and aging are the most predictive of risk factors for breast cancer (2004). Experts believe that smoking, excessive alcohol consumption and lack of exercise may increase a persons risk of developing breast cancer. Family histories of breast cancer and such personal factors as when a woman goes through m enarche and reaches menopause have been shown to relate to a higher chance of breast cancer. Additionally, a womans risk of breast cancer goes up as she ages ( NCI 2005b). Scientists have identified several genes that are related to breast cancer risk. Th e BRCA 1 and BRCA 2 genes are risk factors for breast cancer, although scientists caution that a low percentage of breast cancer cases occur among women with these genes. The concept of prevention typically is divided across three levels in public health practice, each with a different focus and goal. Primary prevention efforts aim to promote healthful behaviors and services in communities so that certain health problems do not develop. Turnock describes primary prevention as strategies that seek to preve nt the occurrence of disease or injury, generally through reducing exposure or risk factor levels. These strategies can reduce or
17 eliminate causative risk factors (risk reduction) ( 2004, p. 398). To reduce the risk of breast cancer, at the primary level, women are encouraged to eat healthful foods, to abstain from smoking and excessive alcohol use, and to get adequate physical activity. The hope is that such actions will reduce the number of breast cancer cases that occur. Targeted groups in an interventi on are those which are considered susceptible to the health problem but do not currently have it, so that incidence reduction is the ultimate goal. This includes women who may be helped by taking chemoprevention therapies, such as tamoxofin, or undergoing prophylactic mastectomies (Chen, Rosner & Colditz, 2007; Cummings, 2007). Naturally, primary prevention is the desired level of public health interventions. They are the most cost effective in terms of outcomes (Turnock, 2004), and they maximize quality of life in the targeted individuals. Secondary prevention programs focus on identifying health problems in early stages through the use of screenings and other tools. Detecting illnesses at this point, it is believed, will make them easier to treat and wil l reduce symptoms and mortality, protecting some measure of patients quality of life (Turnock, 2004). Prevention at the secondary level calls for diligent screening and testing so that breast cancer cases will be detected early, at stages enabling the mos t effective treatment. The mammogram is the most common medical diagnostic tool for detecting breast cancer, and it is generally recommended that most women have one annually after they reach age 40. Before that age, clinical exams should be included in ye arly gynecological appointments for women ( ACS 2007d). Yabroff found that mammography is associated with improved survival, lower annual costs of care, and lower continuing care costs than later detection (2000, p. 208). Evidence suggests mammograms can detect breast tumors at earlier stages than usually identified in clinical or self examinations (Elmore, et al., 2005).
18 The final level is tertiary prevention. This one involves treating individuals with later stages of a disease. It usually requires specialist medical teams and more invasive techniques, with the goal being preventing death and morbidity Strategies for breast cancer at this level include various treatments like radiation, chemotherapy, hormone therapy, mastectomy and lumpectomy. These tre atments, which differ by degree of invasiveness, may be used in combinations as part of an aggressive treatment effort. According to the American Cancer Society (ACS) the five year survival rate for breast tumors detected and treated at stages 0 and 1 is 100% .1History of the Cause Later stage tumors that may be larger and have spread to other parts of the body have lower five year survival rates, from 93% for stage IIA to 20 % for tumors detected at stage IV ( ACS 2007c ). The proliferation of Web sit es about breast cancer is likely due, at least in part to increases in the number of awareness campaigns and the intense media focus on the issue. Braun (2003) credits an emphasis on breast health and more media attention to breast cancer victims stories in the early 1980s with raising awareness of breast cancer as a critical public health problem. The media effectively plac[e] and maintain health issues on the publ ic agenda and provide straightforward information (Downie, et al., 1996, p. 32). Whether the media focus first raised consciousness of the disease or the awareness campaigns captured media attention, the point is that breast cancer awareness in all its forms (research for a cure, prevention education, and patient support) has become a recogni zable cause. 1 The National Cancer Institute describes breast cancer stages 0 IV. Very generally, stage 0 is breast carcinoma in situ where abnormal cells are in the breast lining but they have not spread beyond that location. On the other end of the breast cancer spectrum, patients at stage IV have cancers that have spread to other body organs. S tages in between those two extremes reflect increasing tumor size, cancerous activities in the lymph nodes, and metastasis to other body parts ( NCI 2009a ).
19 October is national Breast Cancer Awareness Month in the United States. During this month, and indeed year round, many individuals wear or display pink ribbons or pink ribbon symbols, which are meant to raise awareness of the disease. Organizations like the Susan G. Komen f or the Cure foundation and the American Cancer Society have educational programs and event fundraisers designed to honor those who have died from breast cancer, to support breast cancer patients and survivors, and to raise money for research and curerelated initiatives. The pink ribbon has become a widely recognized symbol of the collective movement to develop treatments and eventually find a cure for breast cancer. No studies could be found showing the extent to which the public associates the pink ribbons with breast cancer causes. However, the sheer number of commercial enterprises seeking to align themselves with the breast cancer awareness and cure cause indicates that this associative relationship is very strong (Orenst ein, 2003; Jackson, 2006). Some companies offer to donate portions of purchase proceeds to a breast cancer charity in exchange for the positive association with the cause (Till & Nowak, 2000). Advertisements in the media show that alignment with the cause crosses industries, with companies like Ford Motor Company, Oreck Corporation (vacuum cleaners), and the Avon cosmetics company forming partnerships with breast cancer organizations and sometimes establishing their own fundraising foundations.2 One positive effect of the media attention devoted to breast cancer cau ses may be increased awareness of the disease. Women who have information about breast cancer Another indication of the successful breast cancer awareness month initiative is the increase in the number of mammograms performed during October each year ( Effect of breast c ancer, 2005). 2 See Fords Warriors in Pink campaign ( http://www.fordvehicles.com/), Orecks Clean for the Cure ( http://www.oreck.com ), and Avons foundation ( http://walk.avonfoundation.org/ ).
20 prevention methods including good diet and exercise at the primary prevention level and compliance with recommended screenings at the secondary prevention level, may be able to make changes in their lives to lessen their risks of developing breast cancer or detecting the disease at a late stage. Not only is the idea of reducing cancer risk advantageous to the individual women and their families who would be spared the pain and emotional effects of the disease, these types of prevention are economically ideal for the nation. Populationbased and primary prevention efforts are better investments in public health than are tertiary care plans. This logic supports the idea that the Internet should serve as a medium for providing women with information they can use to reduce their risks of developing breast cancer.3 3 Clearly this is a simplistic line of reasoning that in no way assumes that information receipt will translate into behavior change. As this paper suggests, the literature on how certain subpopulations make use of this online prevention content has no t been established. How that content impacts behavior is an additional step in the research that cannot be made until the immediate questions are answered. It may especially serve as a resource for women, including many African American women, who face disad vantages in regular access to the health service system (Gustafson, et al., 2005). However, there appears to have been no published research specifically exploring whether and how African American women who have not been diagnosed with breast cancer use the Internet to find information about preventing the disease. Several studies have examined how African American women with breast cancer (or with personal histories of breast cancer) use the medium (e.g. Fogel, Albert, et al., 2002; Talosig Garcia & Davis, 2005) T here are also studies about how African American women construct beliefs about breast cancer and prevention tools such as mammography (Beck j ord & Klassan, 2008; Russell, et al., 2007). Findings from these and other related areas of scholarship about breast cancer will be discussed in the following sections of this paper. They fall into categories of health and Web literacy, access to the Internet and perceived benefit from online breast cancer information. Such studies allow us to explore aspects of the
21 online breast cancer experience in depth and then assemble them into a larger picture. However, it remains to be seen whether this reductionist approach to understanding how women use the Internet for breast cancer information really captures that experience in a way that can be used to improve how organizations present and deliver breast cancer information through the medium. Furthermore, it is unclear if and how the prevailing public health paradigm of multiple determinants has been married to a mo dern theory of new media use in a way that provides an agenda for future research in this area. Through the literature revi ew and study methods, this project seeks to raise several issues for discussion in the health communications research forum in addi tion to answering specific research questions that will be presented later in the document. The contentions, starting from a micro research perspective and extending to the macro, are that use of the Internet for breast cancer information by a woman who ha s been diagnosed with the disease will likely be different from the mediums use by women who have no personal history of breast cancer. If this is true, then the plethora of findings about how African American breast cancer patients or survivors use the I nternet offers limited insight in to prevention informationseeking. The second point is that ecological studies about how the Internet has diffused across race, gender, education level and income groups tell us only about physical access to the technology, not how the users construct meaning from the content they find.4 4 Of course this is not the stated purpose of such studies. In other words, even if two people are provided the same Web site about breast cancer prevention, they likely will interpret the content differently because they employ unique cognitive abilities, life experiences, socially constructed meanings about disease, and other filtering variables. Neither of these ideas is new. Theories such as the Health Belief Model (Janz, Champion & Strecher 2002) include variables such as perceived
22 susceptib ility to a health problem as influences on how people attend to health messages and adopt behaviors. A breast cancer diagnosis naturally increases someones perceived vulnerability to breast cancer. Likewise, theories suggest that individuals understandin g and behavior are mediated by social, cultural and community influences.5The ultimate idea upon which this study s thesis rest ed wa s that there may be something unique about African American womens perceptions of breast cancer prevention and intention to seek information about it; and there may be something unique about African American womens use of the Internet a nd the ways in which they interpret online content. The answer to the question of how and why African American women use and create meaning from Internet information about breast cancer prevention may not be the sum of the two parts. Results from studies of how African American women conceptualize breast cancer prevent ion and studies of how African American women use the Internet may not tell us how African American women use the Internet for breast cancer prevention information. Where the existing literature fails to offer non assumptive questions is at the intersection of these influencers on both Internet use and conception of personal health at a mea ning level. In other words, we cannot know how African American women without breast cancer use the Internet for breast cancer information based on the literature that has focused on women with the disease. Also, we cannot know if African American womens use (or nonuse) of the Internet for such educational material is related to their beliefs about the Internet being a good or bad source for breast cancer information or their beliefs about preventing breast cancer. 5 For example, Heaney and Israel (2002) explain the role of soci al networks on peoples health. Baranowski, Perry and Parcel (2002) describe how Banduras Social Cognitive Theory relates to interpersonal influences on individuals health.
23 Although published studies about how breast cancer patients and survivors use the Internet for information about (and for coping with) the disease are numerous (e.g. Rees & Bath, 2001; Williamson, 2005; Dickerson, et al., 2006; Ziebland, Chapple, et al., 2004; Eysenbach, 2003), the body of literature exploring how women without personal histories of breast cancer use the Internet for prevention information is scarce. A rather obvious explanation for this is that perhaps such women do not seek breast cancer prevention information, so such studies would be unnecessary. If women do not believe that they are at high risk for breast cancer, they may not be prompted to enact the health behavior of prevention informationseeking. As stated before, however, clearly a breast cancer diagnosis should increase a womans perceived personal susceptibility to the disease and change her behavior, including informationseeking and use of the Internet for other breast cancer purposes.6Using the Web to find out about treatments or to join an online support group are two examples of how women with breast cancer utilize the technology. Such activities may contribute to a womans psychological, emotional and physical survival during and after treatment, thus warranting their placement on a list of sec ondary or tertiary prevention tools. As described earlier, s econdary prevention in public health includes efforts aimed at detecting health problems like breast tumors early so that they can be treated in less invasive ways. Mammography and other screening s fall into this category. Tertiary prevention involves efforts to red uce mortality and morbidity from the disease the latter of which will be discussed later These are important in the planning of breast cancer mortality and morbidity prevention strateg ies. However, it is a goal of primary public health interventions to reduce the incidence of 6 For example, Satterlund, McCaul and Sandgren (2003) found that for women who ha d been diagnosed with breast cancer, patterns of Internet use for information about the disease changed as they progressed through stages of treatment and recovery. From this we might surmise that a similar difference in health information seeking might be found before a breast cancer diagnosis.
24 breast cancer in the population. Although there is no known guaranteed way to avoid getting breast cancer, taking suggested steps to reduce ones chance of develop ing the disease, such as not smoking, may do more than just prevent death. Women with advanced stages of breast tumors experience higher rates of depression, anxiety and other psychiatric issues as they deal with the disease and its consequences ( Grabsch, et al., 2006). But women at other stages of breast cancer also may face psychological and physical side effects from the ordeal, treatment, and living with possibility of recurrence (Wilmoth & Sanders, 2001) affecting their quality of life. These types of morbidity issues are associated with tertiary prevention. Compared with women who have not had breast cancer, survivors of the disease report lower quality of general health (Paskett, 2008) and a reduced quality of life from their pre cancer states (Micha el, et al., 2000). In addition to general health concerns, women with breast cancer may suffer fatigue and experience difficulty concentrating as side effects of treatments (Ashing Giwa, et al., 2004). Other physical side effects of breast cancer treatment s are hair loss from chemotherapy and arm and chest swelling (lymphedema) resulting from disruption to the lymph system from breast surgery and radiation ( ACS 2009a ). Decreased sexual desire has also been reported in post treatment women (Ashing Giwa, et al., 2004). Physical effects of breast cancer treatment and survival may impact womens psychological and emotional well being, too. Mastectomy, which involves removing the breast, may result in specific concerns for African American women who are more prone to forming keloid (cell overgrowth) scars and may worry about finding prosthetics that match their skin tone (Wilmoth & Sanders, 2001). Women, especially young survivors, may experience body image issues, possibly believing they are not as attractive as they once were (Ashing Giwa, et al., 2004; Fobair, et al., 2005). And women from ethnic and racial minorities may report embarrassment for
25 having the disease and concerns about discrimination in the medical system affecting their care (AshingGiwa, et al. 2004). For these reason s it is important to stress prevention for women who have not been diagnosed with breast cancer, and that aim is one of the key foundations of this project. It is also important for researchers, medical professionals and others involved in looking into the future of cancer prevention to better understand the tools and opportunities for public health education available through new media. Cultural Sensitivity Providing culturally sensitive health interventions and services may help reduce some of the health disparities that affect minority residents in the U nited S tates ( Office of Minority Health 2005). Developing cancer prevention information that is more culturally relevant to subpopulations may increase the perceived personal re levance of that information to individuals in those groups because ideas about health well being vary across cultural groups (Friedman & Hoffman Goetz, 2006; Guidry, Fagan, & Walker, 1998) For example, African Americans may prefer to have their race and c ulture considered in the cancer information they receive, including cancer statistics and risk calculations (Brown & Williams, 1994). Cultural sensitivity is an often illdefined term that refers to incorporating understanding of, and elements from, a targ et populations culture or experience, including customs, language, and values, into health interventions ( Office of Minority Health 2005). According to one explication of the term, cultural sensitivity includes both congruence between the health message content in terms of language and appearance and the message receivers culture, and the understanding that the receivers culture affects her health perception and behavior (Friedman & Hoffman Goetz 2006).
26 Culturally sensitive interventions are based on the idea that audience members may have distinct culturally formed and maintained expectations about health behaviors that should be addressed in order for interventions to be effective. Research suggests that encouraging health behavior adoption, such as getting mammograms, in African American women requires understanding the concept of culture, a patterned way of life that has special meaning to the individual and his or her social group (Bailey, Erwin, & Belin, 2000, p. 136). The cultural sensiti vity literature suggests that health information must take into account audiences beliefs (including knowledge, attitudes, and values), barriers and enablers to enacting change, and the mediation of beliefs and activities by individuals peer and social spheres (Kline, 2007). Culturally sensitive materials and interventions may identify things that appear to be cultural barriers to good health behavior but that can be turned into cultural assets (Bailey, Erwin, & Belin, 2000). For example, peer perpetuat ed myths about diseases and disease prevention may be an obstacle to individuals adopting healthful behaviors. However, the importance placed on personal testimonies and relationships that allowed the myth to flourish can be an asset if peer educators are trained to raise community awareness about the health condition. It is important to note the obvious fact that the self classification term African American represents a range of education levels, incomes, religions, beliefs and histories. Cultural sensi tivity is not to be confused with stereotypes. Instead, researchers acknowledge that shared experiences and histories may be more unique within groups of African American women than between African American women and women of another race or ethnicity ( Gui dry, Matthews Juarez, & Copeland, 2003).
27 In their review of printed cancer education materials, Guidry, Fagan and Walker (1998) found that 45% of breast cancer materials for African Americans were not culturally sensitive. In the researchers words, Many materials included pictures of African Americans on the exterior, but failed to address African Americans either visually or verbally. Messages were placed in pamphlets that are not relevant to African Americans. This also may cause the reader to lose interest in the topic. Similarly, many printed materials did not include pictures, and researchers should be aware that African Americans are a visual culture. In addition, the format of the materials was less than conducive to the visual culture (p. 168). Som e of the approaches developers of culturally sensitive cancer prevention materials should use include addressing cancer stigma and fatalism (Bailey, Erwin, & Belin, 2000), providing risk information tailored to the audience, describing alternative and comp lementary treatments (Friedman & Hoffman Goetz, 2006), and managing myths about how people get cancer (Guidry, Matthews Juarez, & Copeland, 2003). Other issues that may act as barriers to people of that racial or ethnic group seeking health services such as distrust of the medical system, should also be considered (Guidry, Matthews Juarez, & Copeland, 2003). Friedman and Hoffman Goetz suggest that cancer material should be assessed for cultural sensitivity based on whether or not the racial or ethnic group is mentioned in the content and described as the target audience for that information; if risk perceptions and alternatives to traditional medical treatment are mentioned; if the information is written in a way that the audience can understand it; if cues to the desired action are included in the content; if the producing organizations contact listed on the materials is of the same racial or ethnic group; and if the sources for the message would be trusted by readers or audience members (Friedman & HoffmanGoetz, 2006, p. 442443). The Internet & Technology For all the advantages offered by the Internet as a form of communication, its very nature results in some challenges for researchers and content providers. In fact, for every desirable
28 feature of Inter net use, one might identify a potential difficulty for those individuals who study the technology. For instance, anonymity may be an attractive feature for Web users wishing to find out about a sensitive subject like a personal health condition; however, c ontent developers might benefit from having basic information about the people who use their sites. Information seekers may like the fact that they can follow hyperlinks from site to site, thus locating information that specifically pertains to their inter ests. On the other hand, researchers know that this vast information opportunity for users severely complicates researchers abilities to summarize or measure information seekers online experiences. Street and Rimal describe the Internet as having two cha racteristics: interactivity and modular components.7 7 A detailed account of the Internet s history and mechanisms is not necessary for reading this paper. Although the term encapsulates many data sharing forms, including email and the World Wide Web, its use in this study refers more generally to the act of using a network connected computer or other device to find information about breast cancer through whatever online mechanism desired. The latter term refers to the ability of a technological program to link many sources of information, thus creating a nonlinear information experience; the former term refers to the user control and responsiveness provided to users (1997, pp. 23). Taken together in the context of an online search for health information, these characteristics produce an online experience wherein a user can control what content she finds by making keystroke and mouse click decisions while the online program responds by producing a result based on those choices. As she navigates through Web pages, typically she will have many additional choices on each site opportunities to select links to pages with more and different information. By offering numerous links on every page and not forcing a user to pick a particular link, the Web site extends more control over the content selection to the user. It also requires of the Internet user
29 more personal processing of the messag es. Street and Rimal (1997) explain how this relates to health information seeking and processing by saying that effectiveness of health promotion interventions depends, in large part, on how users process health messagesand whether this experience produ ced the desired results. How message processing unfolds will depend on the interaction of user, media, and message characteristics (p. 14) Compared to a print, radio and television user the Internet user has more opportunities to construct a unique colle ction of information based on her selections. She employs her mental faculties, personal and cultural preferences, and computer skills to gather a subset of Internet information and applies those same elements to interpreting what she finds. This is an overview of why the very features that attract users to new media may hinder researchers efforts to explore the phenomenon in a robust manner. As with all mass media studies, those that examine Internet communication either apply scientific rigor to focus on a very small aspect of use or they attempt to synthesize the experience, often placing it within the parameters of the larger social and cultural trends. On its face this study do es neither. The expected outcome of the study is not a list of answers to t ested hypotheses, nor is it a grand theory about the use of and effect of the Internet in peoples lives. Rather the anticipated end to the project is, instead, a merger of both perspectives. It takes the knowledge already gained through other researchers studies of the subject of Web use for health information and synthesize s it with research on African American womens perspectives on breast cancer prevention. However, before those specifics are presented, a discussion of what the Internet is and how res earchers conceptualize it as a form of communication is necessary. The Internet is more than the sum of its parts. Many of the things the Internet allows users to do send messages to friends, research health topics, buy products are activities people h ave done for a very long time. However, the Internet offers a different way to do these things and the
30 fact that these activities can be done in a different manner may change the effect of the activities. For example, a person may maintain larger connecti ons of acquaintances with people in different geographic regions than she would without the Internet because email, Internet phone, social networking, and other interpersonal communication technologies facilitate exchanges more easily than postal service mail and long distance telephone call coordination across time zones. A person who might not have been inclined to research a health topic in a public library a few decades ago might find that the convenience of fast access, home Internet service and the w ealth of resources on the Web removed some or all of his barriers to health information seeking. Some of the features that make the Internet a useful health information tool are user anonymity, extensive and varied information, accessibility from across the world, the mix of media (such as text, audio, and video) offered, and interactivity. The first four characteristics are commonly understood, straightforward concepts but interactivity is more complicated. One oftcited definition of interactivity is th e extent to which users can participate in modifying the form and content of a mediated environment in real time (Steur, 1992, p. 84). This definition implies that the users action whether keystrokes, mouse clicks, or other decisions gives the system input and the system responds, without delay, by presenting information based on that input. In other words, both parties in the interaction receive feedback from the information exchange (Street & Rimal, 1997). Each persons Internet experience will be d ifferent because the experience is based on the systems available offerings and what the person seeks from it. The degree of interactivity in any given Internet tool, including individual Web sites, varies according to how responsive the source is to the users prompts and how much control the user actual ly has in shaping the content that results from his commands or selections (Street & Rimal, 1997).
31 Stout, Villegas and Kim (2001) suggested that interactivity is composed of many dimensions, including cont ent and physical accessibility; multimedia forms of message delivery; navigation; data entry and user personalization; linkages between individuals; and time required to use the Internet among other dimensions that are less relevant to this study. Advant ages for health information seekers are that the Internet content can be accessed virtually at any time (Rimal & Flora, 1997) ; in addition, it offers social support options for people with diseases and health issues, provides opportunities for interaction with physicians and other health professionals (Street & Rimal, 1997), and the potential for messages to be tailored to meet the users preferences (Rimal & Flora, 1997). The se Internet features may hold importance for health behavior change and health pro motion The features allow for active user involvement in interactive media that offer opportunities for individuals to learn vicariously (Jaffe, 1997) and to practice or model skills (Stout, Villegas & Kim, 2001) When the user successfully accomplishes his or her task, be it answering quiz questions correctly or finding information that meets his or her needs (Rimal & Flora, 1997), these interactive activity successes may lead to greater self efficacy for the user to do the promoted behavior in real life (Street & Rimal, 1997). The converse also may be true; the inability to navigate the Internet may result in reduced self efficacy (Jaffe, 1997). New Media Communication Theories Theories about the roles of message deliverers and receivers have changed as both the media and the field of communication scholarship have progressed, moving from a view that the media have very powerful and direct effects on message recipients to one in which receiver processes are integrated more fully into the effects equation (Severin & Tankard, 2001). As media channels increase and new types are added, existing theories of communication and interaction with media must be tested, expanded, and have their utilities and application
32 parameters defined. Newhagen and Bucy (2004) off er one model for understanding the myriad variables that impact Internet use and messagereceiver communication in new media. Before their ideas are presented, core concepts must be described. In support of their belief that new media require scholars to reexamine theories of the communication process, Newhagen and Bucy (2004) define several keywords that help conceptualize the players and roles in the Internet phenomenon (Also see Bucy, 2004) First, they contend that information consists of data that are or ganized to have meaning that have been derived through the use of an abstract symbol system such as language; access to that information is contingent on an understanding of that language (p. 17). The word organized is important to their point becau se they imply a difference between random or haphazard pieces of sense data and content. The latter is a collection of words and images structured into narrative units called content (Newhagen & Bucy, 2004, p. 5). These narratives may be constructed and influenced through various avenues, including personal cognitive abilities and social and cultural inputs. The fact that narrative creation is not formulaic leads to the inevitability that slightly different versions of a narrative may exist. Contents pur pose is to convey meaning to an audience, yet the variables that allow people to understand and interpret that meaning also may determine whether or not they have access to the content in the first place and whether or not they attend to the content messages. Newhagen and Bucy (2004) suggest that access is a multidimensional concept that exists on personal, social and system levels. First, there is physical access to technology that might be explained in terms of certain people possessing a personal compute r with an Internet connection (or having a place where they can use a computer). Some determinants of physical access to the Internet include income and education level, which are cofounding variables with race and ethnicity in the United
33 States at the time of this writing (Brodie, et al., 2000).8Second, there is potential for differ ential content access on the personal cognitive level because the words, texts, and sounds must be decoded and interpreted using real neurological processes and skills that vary among individuals. There are also content access differences at a social level wherein social communities place varying levels of importance on being online as opposed to other activities. The latter suggests that ones social environment helps form a persons schema, and that schema helps him process and interpret content. Many efforts have been made to increase physical access to computers and the Internet among people identified as being have nots in the Digital Divide that separates the digital rich from the digital poor. B ased on Bucy and Newhagens thoughts, one can see how physical access is both a personal access barrier and an aggregate access barrier when multiple people within some segment of the population face the same barrier. If the se contentions possess truth, the tenets, to some extent, would be observed in communication in the era before the Internet As in the case with print and other electronic media, communicators could not exercise total control over how receivers decoded and interpreted the meanings in information they read or observed. However, the Internet is different because content creation is not centralized and produced by a limited number of newspapers or radio or television stations owned by media conglomerates actin g as information gatekeepers. Through the use of Weblogs, social networking sites, and other emerging platforms for digital conversations, alternative narratives can be introduced into the communication landscape. A persons prior schema may determine whet her or not she chooses to use the Internet at all and what information she chooses to attend to online. The nonlinear nature of the Web may result in 8 The income variable is not found to be significant in all studies (e.g. Kind et al., 2005).
34 Person A having a completely different Internet information vig nette than Person B because the two chose t o use the technology in different ways. These concepts intersect with the present project by suggesting that getting a Web user to attend to a certain message is a complicated effort that includes understanding the system, social, and personal level charac teristics that may affect her decisions to 1) use the Internet for information about breast cancer, and 2) consider the Web page information consistent with her own constructed narrative. The ideas will be explored in the next chapter by identifying who pr esently uses the Internet to find health information and how and why they seek such information. Various concepts related to the factors that shape views on what constitutes a patients role in health management as well as her identification of good health information sources also will be discussed. It is hoped that this review later will be used by others to identify a version of health information tailoring/targeting which has been explored in other media (e.g. Kreuter & Wray, 2003; Kreuter, Skinner, et al., 2004; Noar, Benac, & Harris, 2007; Paisley & Skrzypczyk, 2005; Strolla, Gans, & Risica, 2006) that specifically meets the needs of African American women using the Web for information about primary and secondary prevention of breast cancer. One of the reasons this perspective of an expanded diversity of inputs into narrative construction is interesting is how it largely mirrors the variables public health scholars believe affect a person and, by extension, the aggregate publics health. It is theori zed that a persons health is a product of individual level factors such as genetics and health behaviors, social and interpersonal level variables that include cultural views of health, and system level constructs like history, racism, and politics (Kapla n, Everson & Lynch, 2000). This multilevel determinant model of health also allows for psychological factors to exert impact on personal health. For
35 example, studies of how individuals perceive racism are distinctly different concepts than identified systemic racism, based on measurement strategies ; however they are clearly related. A racial minority persons health may be negatively affected by his or her perception of being discriminated against (Williams, Neighbors & Jackson, 2003), so whether or not a ctual discrimination occurred is of secondary importance. This example is pertinent because it is a nonphysical variable that results from narratives people and groups create and it culminates in physical effects on health. The literature is replete with s tudies documenting that some African Americans existing distrust of the health care system is based on historical racism and medical experimentation. That is to say, the lens through which some subgroups in the population see health and health care seems to be a product of a narrative based both in fact and in perception. In such a case, whether or not actual racism exists in the individuals dealings with health care practitioners may be secondary in importance to the individuals perception that it exists. At first glance this may seem entirely different from the Internet use narrative, but in essence, both suggest that the effects of culture and history on the formation of a persons narrative about health and/or the Internet as a source of health infor mation may be as important as education, income level, and age are as use indicators. These may be critically o verlooked areas of scholarship. Street and Manning (1997) proposed a similar model to the one examined here: the effectiveness of health message s designed to inform or persuade is contingent on the juxtaposition of two contexts of experience. One is the mediated information environment within which the message is presented. The other is the health behavior environment that consists of other social cultural, and economic factors that also will influence health behaviors and outcomes of interests. (p. 122). Numerous studies and online interventions have explored evolving options for communicating cancer information to the public. These add a great deal to the body of knowledge about online health promotion but do not always adequately explain within a single
36 project the idea explored herein, namely how the intention and act of looking online for breast cancer prevention information is influenced by both health beliefs (about breast cancer) and beliefs about the Internet as a tool for accessing such information. Intuitively we can expect individual, social and cultural, and macro factors to influence womens schemas or meanings about breast cancer and about the Internet ; however, this study is designed to explore the intersection of those meanings in order ultimately to hypothesize about whether or not women without cancer diagnoses would or would not utilize online information about breast cancer prima ry and secondary prevention. If it can be determined that their beliefs about and meanings of breast cancer influence their decisions not to seek information, then the point of intervening to increase their information seeking behavior would be different than the strategy for increasing searches in women who describe technology barriers (physical access, distrust of the medium, Internet skills, etc.) as their hindrance. Th is study did not answer this important abstract question, but the approaches employe d provide d good initial insight into the meanings women make of primary breast cancer prevention and the use of the Internet for information about the topic. It should allow researchers to probe the raised questions using other measures to determine ho w we ll the themes discovered in the study represent African American womens beliefs as well as to employ additional qualitative methods to explore the topic in more depth. Th e study specifically focused on African American women because they represent an ide ntifiable population subgroup about which there is research that reveals certain cultural views about health behavior and technology use as well as unique breast cancer patterns. Also, the African American population has lower income and education than whi te Americans, thus limiting their health resources. Gustafson et al., partnering with the National Cancer Institutes
37 Extraordinary Opportunity in Cancer Communications program, suggested that a mechanism for reducing cancer disparities is increasing the knowledge about, tools for, access to, and use of cancer communications by diverse populations (2005, p. 174). Understanding how to design, implement, and disseminate materials that are tailored to diverse populations needs is vital to this goal. The st udy is not based on assumptions of consistency among African American wom e ns health and Internet behaviors and personal beliefs about these things; however, it does utilize the extensive literature that tells us something about the social and cultural inf luences on health and technology use.9 9 Another important note is that men can and do develop breast cancer, but because the rates are much lower than those in women, this study focused only on women. Logically, mens perceived risks of breast cancer a nd their intention to look for information also should be lower and that is why including men in this study would be inappropriate to the studys aims. Within this proposal are reviews of the literature about breast cancer in women and specifically in African American women, Internet use by African Americans and other Americans, online health information seeking by the general public and by those with breast cancer, and how these may or may not support the idea that the act and process of looking online for breast cancer prevention information may be affected by beliefs about breast cancer as well as beliefs about the Internet
38 CHAPTER 2 LITERATURE REVIEW The Breast Cancer Experience in African American Women African American women have lower rates of breast cancer diagnoses than white females, but they are more likely to die from the disease ( ACS 2007b). There a re numerous theories about why this is the case. Like most ecological health models, these theories exist at the personal level in the form of such factors as genetic differences in breast cancer types (Amend, Hicks & Ambrosone, 2006), social and cultural levels including different culturally held views of health, breast cancer and prevention (Lannin, et al., 1998), and at the system level in terms of access to health care and racism or perceived racism within the medical community (Williams, Abbott & Tay lor, 1997). Some scientists suggest that a different, more aggressive variant of the disease may be present disproportionately among African American women (Amend, Hicks & Ambrosone, 2006). One curious piece of evidence that may support a biological expla nation is the fact that breast cancer is diagnosed more often in much younger black women than among younger white women. Another proposed reason for the discrepancy is that tumors are found at a more progressed stage in African American women. This hypothesis can be interpreted in several ways. It might mean that by the time the disease is discovered in a black woman, it has progressed further than is the case for the average white woman with breast cancer. It can also be interpreted that theres a longer delay between diagnosis and treatment for African American women (Gorin, Heck, Chang & Smith, 2006). The distinction may be small, but it is important to discuss because it may identify any system level problems that prevent early detection and/or treatme nt of breast cancer in African American women.
39 Mammograms, the tests included in both the American Cancer Society and National Cancer Institutes guidelines for breast cancer screenings in women older than 40, are administered as a form of secondary preven tion for breast cancer. The goal of mammography in breast cancer control is to detect tumors at their earliest, most treatable stage, thus preventing death from the disease and spread to other parts of the body. There have been conflicting findings about t he rates of mammography screenings in African American women possibly reflecting fluctuating rates per year Tessaro, Eng and Smith (1994) found lower utilization of mammography in black women compared to white women, but in a later study, Jones and colle agues (2003) found no significant difference in screening rates by race. Recently, Williams, Lindquist and colleagues (2008) found an association between lower income and less screening in older women. Although that study did not focus on race, the findin gs raised questions about the connection between money (of which African Americans, on average, have less than white peers) and screening behaviors. Access to healthcare is another factor that may influence breast cancer prevention (Wolff, et al., 2003) diagnoses, and outcomes in African Americans. Private health insurance rates for black citizens are lower than for white Americans. They are more likely than their white peers not to have a regular care doctor and more likely to utilize emergency rooms and county health departments for health services. Even though many of these differences disappear when controlling for income and similar socioeconomic factors, suggesting that the relationship may be a result of financial circumstances and not race it remains that such negative socioeconomic factors disproportionately apply to African Americans in the United States; therefore understanding the effect of health care utilization is important to identifying the factors that result in health disparities.
40 Distrus t of the medical system also may play a role in African Americans health care decisions (Wolff, et al., 2003). A history of unethical medical practices on black people in the United States dates back hundreds of years but occurred in its most concrete a nd perhaps best known form in the Tuskegee syphilis experiments. Beginning in 1932, the U.S. Public Health Service agency withheld information about known treatments for syphilis from a group of male farmers in Tuskegee, Alabama, in order to learn more a bout the natural progression of the disease ( Centers for Disease Control 2008). As a result, the men enrolled in the study and their families suffered unnecessarily from the illnesss effects. Although formal recognition of this unethical practice has sin ce been made, and some financial restitution to the families was provided, the incident is often cited in research on African Americans perspectives on the medical system as a reason for distrust. Also, use of culturally insensitive or ineffective prevent ion strategies targeting African American communities may have a nonoptimal effect on health (Wolff, et al., 2003). Kreps (2006) suggested that prevention information campaigns should take into account the trusted communication channels and interpersonal networks for the targeted health messages. Poverty as an overarching and related factor may play a major role in the burden of breast cancer disease in African American women. Slightly more than 23 % of this racial minority lived in poverty in 2006, compare d to about 10% of white Americans (DeNavas Walt, Proctor & Smith, 2007). Poverty has been associated with many negative health indicators (LaVeist, 2005). Moving backward from screening and mortality to primary prevention, it has been suggested that a lar ge percentage of cancers are preventable through behavior modification ( Institute of Medicine 2003). Tobacco use, low levels of physical activity and obesity are three behavioral components believed to increase breast cancer risk. While no agencies have produced
41 estimates of how many breast cancer diagnoses might be prevented if the obesity problem in African Americans was assuaged, recent research has yielded some evidence that excessive overweight may be negatively associated with breast cancer survival (Hede, 2008). Estimates are that as many as 80 % of adult female African Americans are overweight or obese ( Minority Womens Health 2007). Phillips and Williams Brown (2005) suggest that community resources necessary to sustain behavior change in areas l ike tobacco use are lacking in ethnic minority communities. Similarly, these communities may lack the social capital to examine links between environmental conditions and disease development. Health Beliefs and Beliefs about Breast Cancer in African Ameri can Culture Views of breast cancer and general health perspectives in some African American cultures may present challeng es for public health experts working with communities to encourage prevention. The same idea of distinct cultural views of health issue s would be found in many other subcultures, be they categorized by race/ethnicity, age, or lifestyle groups. However, this paper is concerned with the culturespecific qualities identified in the literature as possibly interfering with healthful breast can cer prevention behaviors. Studies have found that breast cancer was (and may still be) perceived as a private topic among African Americans, such that mothers often did not share experiences or information about the disease with their daughters (Matthews, et al., 2002; Moy, et al., 2006). In communities where this is the case, women may not talk openly about the risks of breast cancer or encourage information seeking. In families that have experienced breast cancer first hand, the situation may actually hav e the effect of discouraging information seeking about the disease. Matthews et al. (2002) suggested that a family history of cancer was also cited as a major emotional barrier to cancer information seeking and active involvement in treatment (p. 212) because cancers in
42 African Americans are often found at late stages, when treatments are more invasive and survival rates are lower. Conflicting reports about African American womens perception of risk have come from the literature. Some have found that perceived risk of cancer is lower in African Americans than among white women (Honda & Neugut, 2004; Haas, et al., 2005). Another study revealed that African American women had lower perceptions of mammography benefits than white women but that the black participants perceptions about mammograms efficacy was actually more realistic than the white participants reports (Haggstrom & Schapira, 2006). In other words, white women overestimated the efficacy of mammography screenings while the black womens esti mates were more accurate. Myths about breast cancer and treatment may prevent some women from seeking treatment. One major issue is the belief that breast cancer is a death sentence, which may result from the high mortality rates in African American breast cancer patients (Moy, et al, 2006). For example, Haggstrom and Schapira found that black women were more pessimistic about their survival chances compared to white women, but a greater proportion of black women accurately perceived breast cancer survival than white women (2006, p. 373). In other words, black women are more likely than white women to believe they will die if they get breast cancer, and based on the mortality rates previously cited, there may be ecological support (as opposed to individual case prognoses) for that pessimism. And Matthews et al. (2002) found that fear about deaths from a cancer diagnosis might prevent African American breast cancer patients from seeking health information and from complying with their recommended treatment s following a cancer diagnosis, if they do not think cancer is survivable Another belief that may be prevalent in some African American communities is that cancer is spread by cutting into it (Matthews, et
43 al., 2002) and that breast cancer is related to breast trauma, which is sometimes caused by domestic physical abuse Simon found that some women might not be screened for breast cancer due to the fear that it could result in unwanted revelation of the abuse (2006, p. 39). Although one recent study by R etsky and colleagues (2007) suggested a possible link between surgery for some tumors and increased mortality, according to the National Cancer Institute (2005a ) there is no scientific support for this association and it remains a cancer myth Trust issu es also play a role in breast cancer health decisions. Mistrust of medical doctors (Matthews, et al., 2002) as well as the medical test technologies (Powell, et al., 2005) and the previously mentioned pessimism about mammography benefits compared to white peers (Haggstrom & Schapira, 2006) have been suggested as reasons why African American women may not adopt preventive and screening behaviors. For this reason, Moy et al. (2006) emphasize the importance of good relationships between primary care physicians and staffs and their patients, and Glanz and colleagues (2003) raise the point that understanding the communication differences between providers cultures and those of the patients is vital. Another way trust plays a role in health beliefs and behaviors is in the area of faith and religion. The church is an important part of many African American communities. While researchers point out the complexity of the relationship between religion and personal health perspectives (Bowie, Curbow, et al., 2001), it r emains that the theme of God being in control of ones health and life may affect attitudes and beliefs about breast cancer screenings and treatments. Simon found evidence that participants believed only God can cure cancer and thus they questioned the purpose of breast cancer screening (2006, p. 39). A participant in Matthews and colleagues qualitative study said, They say that God is going to take care of me, he is
44 going to cure me, and there is [such] a spiritual belief in some individuals that th ey dont take the extr a step [to care for] themselves (2002, p. 213). Some of the other barriers to breast cancer prevention behaviors are logistical in nature. African Americans often face many competing health issues, such as social problems in their co mmunities, so that breast cancer prevention may be perceived as less of a priority (Wolff, et al., 2003). Similarly, the prevalence of other diseases like HIV in the community may be considered more pressing needs (Moy, et al, 2006). Lack of time for mammography screenings has been identified as a barrier (Powell, et al., 2005), as has the fear of losing ones hair due to chemotherapy if a tumor is found and treated (Simon, 2006). One study that examined African American breast cancer patients and survivor s barriers to losing weight (a positive health behavior change thought to prevent recurrence) found that the women wanted to adopt more healthful eating and exercise habits but faced barriers similar to those faced by the general public, as well as some s pecific to their cultures (Stolley, et al., 2006). For example, women c ited having too many nutrition recommendations by experts, not liking the taste of healthy foods, and finding it difficult to prepare a dish just for themselves as factors that negative ly impacted their dietary change perceptions. A culturespecific (be it a racial/ ethnic or Western culture commonality) finding was that food served as a centerpiece of family occasions (p. 619). All of the barriers mentioned so far relate to personal o r social health beliefs and behaviors, but not all public health advocates accept the idea that changing health behavior is an effective way of decreasing health problems in disadvantaged populations (Marks, et al., 2006). In other words, changing behavior may only do so much if the people face obstacles like environmental, education or economic hardships. Low income and low education African American women
45 have a 50% higher risk of being diagnosed with advanced stage breast cancer than more educated women with higher incomes (Gustafson, et al., 2005, p. 174). And Consedine and colleagues found that, (i)n many cases, race and income have independent or additive predictive value in terms of cancer and screening behaviors (2007, p. 551). Health Literacy Fed eral health care officials have referred to public health literacy as a major challenge facing the nation, suggesting that reducing the countrys health disparities depends on increasing health literacy (Carmona, 2006; United States Department of Health 2000). According to a 2003 survey by NALS, 75 million American adults score in the basic and below basic he alth literacy categories (Kutner, 2006), with the highest number of low literates being U .S .born white citizens, but higher ingroup percentages bein g found in minorities (Sentell & Halpin, 2006), the elderly, the poor, the homeless, prisoners, and military recruits (Weiss, 2005). In addition to reading skills, people need other skills to understand health issues and make choices about their personal health to be health literate. Health literacy is a multidimensional measurement of abilities to function in a health care environment. Healthy People 2010 defines health literacy as, "The degree to which individuals have the capacity to obtain, process, and understand basic health information and services needed to make appropriate health decisions" ( United States Department of Health n.d.). It has also been defined as the skills and abilities used to navigate through health messages, information, and the health system (Bernhardt, Brownfield, & Parker, 2005). Although health literacy is related to peoples basic reading and numeracy abilities (Ad Hoc, 1999) and education level is highly correlated, it is not always predictive of health literacy (Conlin & Schumann, 2002). Health literacy goes beyond recognition of words and numbers; health literate people also understand this information, are able to use it to make good decisions about their health (Weiss, 2005), and can adapt to changes
46 in the continuously developing health science (Kickbusch & Ratzan, 2001) and communication technology fields with evolving skills and competencies (Zarcadoolas, Pleasant & Greer, 2003, p. 119). Health literacy is context specific meaning that a person might possess good s kills when negotiating communication with his doctor but poor skills when searching the Web for health information. People need skills and knowledge to read medical prescription bottles, complete paperwork at doctors offices, and search for and understand health information in the mass media (Bernhardt, Brownfield, & Parker, 2005) in order to make good health decisions for themselves and their families. Low health literacy results in an estimated $73 billion wasted annually due to poor information sharing and exchange (Zarcadoolas, et., al, 2002, p. 320; Carmona, 2003). Improving health literacy not only is good for the countrys bottom line and general health, it also serves as an empowerment tool (Zarcadoolas, et al., 2002), giving people skills to make their own health decisions. Understanding health has been linked to a more comprehensive view of health opportunities. Guerra, Dominguez and Shea found that functional health literacy was associated with awareness of colorectal cancer and tests to detect it (2005). This type of health literacy might be better classified as health awareness or health knowledge (Baker, 2006). Although scholars regularly cite a lack of a clear definition of health literacy (Zarcadoolas, Pleasant & Greer, 2003; Baker, 2006), it appears that health literacy is a multidimensional concept not unlike the model of Internet access and the ecological model of health described in this paper in terms of its antecedents and effects H ealth literacy may require possessing fundamenta l/numeracy skills (such as reading, vocabulary, and math skills), science and technology literacy and community/civic, and cultural literacy (Zarcadoolas, Pleasant & Greer, 2003) Researchers have connected income, education, and language skills (Pawlak, 2005;
47 Nutbeam & Kickbusch, 2001) and even individual beliefs and attitudes with health literacy although the association between functional health literacy and beliefs and attitudes is inconclusive (Guerra, Dominguez, & Shea, 2005). As an example of health literacy scholarship, a large body of literature focuses on how patients health literacy levels correspond with their understanding of medical directions and prescription labeling (e.g. Conlin & Schumann, 2002; Ross, Potter & Armstrong, 2004; Davis, W olf, Bass, et al., 2006), with implications that patients adherence to medical practitioners advice may be linked to the ability to read the instructions (Wolf, Bennett, et al., 2005). Another line of health literacy research involves assessing the publi cs understanding of health and science, which studies indicate is low among the United States population (Payne & Schulte, 2003). Health literacy researchers in the communications field also study the readability of health information produced for patient s and the general public (e.g. Daghio, Fattori & Ciardullo 2006). Overall, there is evidence of a gap between the levels at which health materials are written and the reading levels of the general public (Rutherford, Holman, et al., 2006). And a fourth that which is most relevant to this study is the area of research focused on health information provision and informationseeking behavior on the Internet (e.g. John, 2005; Bremner, et al., 2006; Hong, 2006). Included in this category are findings that Internet sites for breast cancer may be written at 11thto 12thgrade and college reading levels (Friedman, Hoffman Goetz, & Arocha, 2006), much higher than the level at which the NALS study indicates many Americans read, and that the average online health information consumer may not be able to distinguish between high and low quality information (Bates, Romina, Ahmed, & Hopson, 2006).These findings are significant to cancer communication because low health literacy contributes to
48 individuals difficulties understanding cancer information, their knowledge of screening tests, and their beliefs about cancer prevention (Davis, Williams, Marin, Parker, & Glass, 2002). A major tenet of health literacy understanding is that it is contextual, meaning that a p erson with high health literacy in one area (e.g. understanding television news stories about a health issue) may have low health literacy in another situation (e.g. finding good sources of health information on the Web). The gap between the levels at which health materials are written and the reading levels of the general public parallels the gap in computer skills. When older, poorer, and less educated women use the Internet people in these subgroups have a harder time locating the information they need quickly, most likely due to a lack of familiarity with the technology (Hargittai, 2002). Digital divide issues continue to concern communication researchers studying e health information. Although the Pew study suggests that Internet use has diffused through much of America, there are still questions about who has access to the technology (Fox, 2006; Baur, 2005). Beyond access, there are major barriers to the Internet s use. Specialized technical skills are required to browse the Web using toolbars, links, popups and radio buttons. The highly textual Web platform presents difficulties for people with low reading and comprehension skills. These may be insurmountable obstacles for people with visual, mental and some movement disabilities. A final issue relat ed to health literacy on the Web is information seeking skills. Research into the search strategies employed, page selection, and decision making by users is ongoing. Still, Baur (2005) suggested that the use of the Internet in health information provi sion may be too new to understand its part in the health literacy problem.
49 Ecological Studies of Information Seeking Large scale surveys have yielded very useful insights into the scope and trends of Internet access and health informationseeking. The Pew Inte rnet & American Life Project regularly conducts surveys to determine what segments of the countrys population are using the Internet for what purposes they use it, and by what means they access the Web. Similarly, the Health Information National Trends S urvey (HINTS) is administered by the National Cancer Institute and produces extensive data useful to many studies of health communication. Based on these data sources, we know that use of the Internet for health information increased from 2003 to 2005 ( Ru tten, et al. 2007). Eighty percent of Americans with Internet access have searched for health information online, with most of them looking for information about diseases and medical problems (Fox, 2006). Rates of Internet health informationseeking vary by age but are relatively high across groups, ranging from a low of 68% among users older than age 65 to 84% among 30to 49year olds (Fox, 2006). Women search for health information on the Internet at a slightly higher rate (82%) than do men (77%); however a significant portion of both sexes engage the medium for this purpose (Fox, 2006). Health information seekers on the Web do not only look for information for themselves. Almost an equal number of users searched for personal medical information (58.4%) as said they sought this type of information for someone they knew (59.5%) (Rutten, et al., 2007, p. 16). The Internet as a health information source is gradually diffusing across demographic groups, but some personal characteristics have been associated w ith higher rat es of online health information seeking. For example, health status is associated with use, as people in poor health are more likely to seek information than healthy people (Houston & Allison, 2002). Individuals with chronic illnesses are amo ng the highest users of the technology for this health
50 information seeking purpose, though other factors such as health insurance status, education, and age, to some extent, affect the usage rates (Bundorf, Wagner, Singer & Baker, 2006; Helft, Eckles, Jo hnsonCalley & Daugherty, 2005). Women are more likely to use the Internet for health information than men (Fox, 2006), but older, low income, and low education women use it at lower rates than other women (Mancini, et al., 2006; Ramanadhan & Viswanath, 2006; Hall, Dunkelberger & Wheat, 2005). Health information on the Internet is a much studied topic. While the Internet is a broad medium that encompasses interpersonal activities such as e mailing, message board postings, and other forms of twoway communication, the World Wide Web component dealing strictly with the provision of health information on Web sites has received a great deal of scholarly attention. Four frequently studi ed areas are health information seeking as a process activity (e.g. Bader & Theofanos, 2003; Morahan Martin, 2003; Coiera & Vickland, 2008), the quality of health information on Web sites (e.g. Meric, et al., 2002; The Internet s democratic nature allows anyone the opportunity to develop his or her own Web site devoted to a disease like breast cancer. Quality and accuracy of this informatio n is mostly unverified and not subject to a set of medical standards. Physicians report some concerns that their patients are not accessing good, evidence based health sources on the Web. They also Schmidt & Ernst, 2004; Walsh Childers & Edwards, 2007), the reading level and amount of health literacy require d to comprehend Web page information (e.g. Kaphingst, Zanfini & Emmons, 2006; Friedman, HoffmanGoetz, & Arocha, 2006; Berland, et al., 2001), and the effect of finding online health information on disease treatment and health maintenance (e.g. Sillence, Briggs, Harris & Fishwick, 2007; Powell, Low, Griffiths & Thorogood, 2005).
51 express worry that patients with low literacy skills and t hose who speak Spanish are not able to find credible information (Perocchia, Rapkin, et al., 2005). Poor quality or inaccuracy in health information Web sites is one area of communication researchers concern (Eysenbach, Powell, Kuss, & Sa, 2002). People s reliance on generic information search engines such as Google that use link popularity to generate page recommendations may result in their selection of misleading or medically inaccurate sites. For example, Ekman and colleagues (2005) found that sites offering cancer gene test services very commonly included fearful messages that ones family history might make individuals more susceptible to cancer. They concluded that instead of providing good prevention advice, (t)he abundance of these family histor y questionnaires could give a skewed idea of the major cancer risk factors and the magnitude of the risk they convey irrespective of the type of cancer studied (2005, p. 771). Christensen and Griffiths (2000) found that depression Web sites included infor mation that was of poor quality and inconsistent with the consensus among the mainstream medical community. A recent analysis of popular breast cancer Web sites found that the type of information provided on them was very different from the information br east cancer experts thought patients should know about the disease (WalshChilders & Edwards, 2007). Source credibility is a major concern for researchers. Previous research has suggested some criteria people use to determine the quality of health informa tion. The inclusion of authoritative sources, scientific references on the page, and a professional look to the site are a few elements researchers cite (Eysenbauch & Koher, 2002). Factors like peoples knowledge, in addition to their perception of a Web s ites trustworthiness, source expertise, and depth have been found to influence whether or not the Internet users return to that site in the future (Hong, 2006). However, in experiments, participants looking at health Web sites did not pay enough
52 attention to the authority or source of the information to recall it when prompted later (Eysenbauch & Koher, 2002). The distinction between these contradictory findings is that people cite one set of criteria, but experiments suggest that the reality may be differ ent. When considered together, these findings raise the question of whether or not people are influenced by bad health information both in the short term and long term by not fully evaluating a Web sites authority. If people use trustworthiness as a reason to return to a site, it is important that they associate trustworthiness with credible sources like the W eb sites of the National Cancer Institute or the American Cancer Society Research suggests that this association is not happening because people may not consider the Web site author and other factors, such as references to scientifically accepted sources of information about cancer, when they select a health information Web site (Bates, Romina, Ahmed, & Hopson, 2006). It is possible that people may ch oose Web sites because the material is presented in a certain appealing way. Personal stories by individuals who have experienced cancer may be more desirable to seekers than comprehensive and scientific sites that are harder to understand (Wathen & Harris 2007). Due to the current state of Web information and the possibility of finding incorrect advice, it is important to improve users Web skills (Hall, Dunkelberger, & Wheat, 2005). There is a disconnect between where people want to get cancer informatio n and from which source they actually seek it, as demonstrated by the fact that close to half (47.6%) of HINTS respondents said they had turned to the Internet the last time they needed cancer information compared to the 23.5% who had asked their health ca re providers (Rutten, et al., 2007, p. 2830). While most Americans favored their personal health care provider (55%) as the source of cancer information, nearly 28 % cited the Internet as their preferred source, a slight
53 decrease from when the same questio n was asked in 2003 ( Rutten, et al., 2007, p. 1920). The order of source preferences was the same for white and black Americans, even though the percentages differed some. A recent survey suggests that people are unlikely to talk to their doctors about what they learn from Internet health sites (Fox, 2006). This is important because more than half of Internet health information seekers report that what they find affects their disease treatment decisions (Fox, 2006). Breast Cancer Information Seeking Women account for more than half of Internet cancer information seekers ( Rutten, et al. 2007, p. 25). Search rates for cancer information vary across age and ethnic groups, ranging from 42% to 55% in women aged 18 through 79 years ( Rutten, et al. 2007, p. 25). P eople identifying themselves as Hispanic use the Internet for cancer information at a much lower rate (25%) than members of other race and ethnicity groups. Approximately 47% of black Americans use the medium for this type of information, and almost 55% of white Americans are online cancer information seekers. Education attainment is positively associated with rates of Internet cancer information seeking, as is household income ( Rutten, et al., 2007). It is difficult to estimate how many breast cancer pa tients use the Web to find information or services related to their disease. A metaanalysis suggests that about 39 % of cancer patients do so (Ramanadhan & Viswanath, 2006). However, isolated studies suggest that it is used by anywhere between 10 % and 43% of breast cancer patients (Satterlund, McCaul, & Sandgren, 2003; unpublished data cited in Weissenberger, et al., 2004). Also hard to determine are specific usage differences among demographic groups of women. Similar to overall Internet trends, breast can cer patients who use the Internet for health information are generally younger, better educated, and have higher incomes than nonusers (Fogel, Albert, Schnabel, F., et al., 2002).
54 Minority women with breast cancer use the Internet to find information about their disease at very low rates (Talosig Garcia & Davis, 2005). Access to the technology is certainly an issue that influences use. When that barrier is removed, such as when Gustafson, McTavish, and fellow researchers (2005) studied women who were prov ided with computers as part of the Comprehensive Health Enhancement Support System program, lower income women actually used the Internet more often than affluent participants. Uses also differed by race as it more often served a communication purpose for Caucasian women, whereas African Americans used it for information. For some women with cancer, the Internet is an important resource because of the mediums characteristics, namely on demand access and user privacy (Gustafson, McTavish, et al., 2005; Ziebland, Chapple, et al., 2004). Christensen and Griffiths (2000) offered that the Internet s capacity to be interactive, to be accessed from across the world, to disseminate information quickly, and to provide a nearly infinite amount of information make it a natural and exciting potential t ool for increasing health literacy. T he Internet s appeal as a health resource cannot be denied. It is a multi use tool that offers opportunities for patients to look for health information, to participate in online suppor t groups and message boards dedicated to cancer issues, and to communicate via e mails exchanged with their doctors and with friends family members and other cancer patients and survivors All of these activities may serve as coping mechanisms for the pat ients (Eysenbach, 2003). Being well informed on matters related to the disease and treatment also allows women to be active participants in their breast cancer management (Ramanadhan & Viswanath, 2006). Women with breast cancer consider information about their disease, such as the likelihood of its recurrence after treatment and family members risk of developing it, to be highly
55 important (Raupach & Hiller, 2002). Patients receive more information from their medical care providers in the early months afte r diagnosis and during treatment but come to rely on other sources as time passes (Raupach & Hiller, 2002). Often cancer patients have questions that are not fully answered by their doctors (Engelman, Perpich, et al., 2005). Despite the high rates of peop le seeking cancer information on the Internet and from other sources, the information they find often lacks usefulness. Reported trust in the Internet as a source of health information appears to be inconsistent with rates of use, as only 18.9% of people place a lot of trust in online health information (Rutten, et al., 2007, p. 22). Arora and colleagues (2007) suggest that looking for cancer information is a frustrating experience for many Americans. While women would like to receive answers to their brea st cancer questions from their physicians, they also seek out additional sources of information (Johnson & Meischke, 1991). Health information seekers consider their personal doctors online information to be their most trusted source, with medical univers ities and the government following in second and third place (Dutta Bergman, 2003). In the mix of additional information resources, 13 % of patients consulted the Internet first in a study published in 2005 (TalosigGarcia & Davis). Patients also report co nsulting the Internet before talking to their medical practitioners in order to be knowledgeable about the condition (Sillence, Briggs, Harris & Fishwick, 2007). Satisfaction with the Internet as a breast cancer information provider is quite high among women, certainly higher than their satisfaction with other medias delivery of such information (Raupach & Hiller, 2002). Beyond looking for answers to breast cancer questions, women use the Internet to become active participants in their disease management. Ziebland and colleagues suggested that Internet
56 use is transform[ing] roles and chang[ing] involvement in health care because it allows patients to question medically prescribed treatment plans and become knowledgeable and empowered instead of merely di sease victims (2004, p. 328). They found that the anonymity of online searching let s patients maintain nonconfrontational relationships with their physicians while secretly checking the professionals advice as well as the medical facilities quality. Not surprisingly, patients use of the Internet for health information may affect the doctor patient relationship when patients find information about alternative treatments or facts that conflict with what their doctors have told them (Bass, et al., 2006; Ce rully, Klein, & McCaul, 2006). If this furthers dialogue between the medical team and patient, the result may be positive. However, it may have negative implications for patient treatment compliance (Bass, et al., 2006). Oncologists opinions regarding th e Internet s impact on their patients are mixed, with most of these specialists saying that the information patients find generally raises hopes (in addition to anxiety) but offering conflicting views of whether this hope is good or not. At any rate, onco logists report that discussing information patients find on the Web has noticeably increased appointment lengths (Helft, Hlubocky & Daugherty, 2003). More than half of cancer patients caregivers who search for information about the disease online say tha t they ask the physicians about the results during consultations (Kirschning, Kardorff & Merai, 2007). Women would like to receive Web site recommendations from their health care provider to help the patients avoid pages filled with complex medical jargo n (Mancini, et al., 2006). It remains unclear how doctors interpret this new information source. In one study broaching the topic, more than 90% of rural West Virginian patients said their doctors never suggested any Internet source for reference (Chandra, Rutsohn, & Carlisle, 2004).
57 What information do they seek? Research conducted when the Internet was gaining popularity as a source of health information found that women used the medium to search for more information about their disease than their physicians had provided, as well as information about other treatment options (Pereira, Koski, et al., 2000). Among breast cancer patients, specific information needs vary by age, as younger women are more likely to want information about the effect of treatment s on childbearing and such than are older patients (Williamson, 2005). Also, the types of information breast cancer patients want changes as they move through stages of their disease experience (Leydon, Boulton, & et al., 2000), but usage remains consistent (Satterlund, McCaul, & Sandgren, 2003). Newly diagnosed patients want to know about disease prognosis, causes, risk factors, treatments, and side effects (Talosig Garcia & Davis, 2005). The type of breast cancer information patients seek online may be di fferent from what they get from physicians and other health professionals. Personal testimonies by cancer survivors have been found to provide new patients with information for managing their own disease (Leydon, et al., 2000). Needs also depend on whether or not the Internet user seeks the information for herself or someone else. Research shows that rural patients seek cancer information for others, presumably friends and relatives, at high rates (Engelman, Perpich, et al., 2005). Cancer patients loved ones report seeking information about the disease for their own knowledge first, but also to share with the patient (Kirschning, Kardorff & Merai, 2007). Research has suggested that women may consciously seek or avoid information about breast cancer follow ing their diagnoses with the disease. One study found that information seekers found the action to be a way to cope with cancer, to gain a perception of control and security in the face of uncertainty, and to help them make health care decisions. The women in
58 the study who avoided cancer information did so because it depressed, scared or worried them (Rees & Bath, 2001). Conversely, use of the Internet by cancer patients may be an important coping strategy that is associated with improved quality of life (O wen, et al., 2005). Whether or not this Internet coping link exists across the board for all cancer types is disputed. Fogel (2004) found that the Internet did not meet a coping need for breast cancer patients using the medium independently. He suggested t hat this contrast with previous scholars findings of a coping effect for patients might be explained by the earlier studies use of a medical librarian to help patients locate good information instead of the patients seeking the information themselves. T his discrepancy raises important questions about whether the benefits of accessing a wealth of cancer informat ion on the Web are offset by womens frustrations in attempting to locate good material and the overwhelming task of surfing through all of the av ailable sites. Support for cancer patients is important for coping and managing their health. However, patients who live a great distance from available support groups and programs may not be able to travel to meetings or functions that offer this benefit. Studies of Internet support group involvement have indicated that the medium does provide welcome psychological and physical coping strategies for its users (Owen, et al., 2005). It may be a useful resource for breast cancer patients even after the successful treatment of their disease (Satterlund, McCaul, & Sandgren, 2003). Questions of how and why women use the Internet for cancer information have been explored using a variety of methods, including indepth interviews (e.g. Leydon, et al., 2000). From them we have learned the aforementioned fact that people tend to start health searches with generic search engines. Choice of search engine varies, with some reasons being the search engine format, it appearing on peoples home pages, and finding an engine useful after seeing
59 someone else using it (Peterson, Aslani & Williams, 2003). Searchers spend a median of 37 seconds on a page and usually click only on links presented on the first page of search engine results (Eysenbach & Khler 2002, p. 575). Alth ough one might presume that gaining information about breast cancer would be associated with knowledge increases, this educational process is more complicated. Variables such as the way online information is presented (e.g. if it is in a two way communicat ion format or a straight text presentation) may lead to different learning outcomes for women accessing breast cancer information on the Internet (Shaw, Han, et al., 2007). Ability to comprehend the Web content also may affect both what women learn about breast cancer and which sites they select as sources. Research analyzing cancer Web sites reading levels reveals that many use text consistent with upper high school and college level reading skills (Friedman, HoffmanGoetz & Arocha, 2004), although breas t cancer sites appear to be written at a lower reading level than those devoted to other cancer types (Friedman, HoffmanGoetz & Arocha, 2006). Obstacles to Use The biggest and most obvious barrier to womens use of the Web to find breast cancer information is lack of computer access (e.g. Ziebland, Chapple, et al., 2004). An Australian study of breast cancer patients interest in using the Web for information about the disease revealed that deficiencies in skills, lack of access to the Internet and a lac k of interest were key reasons why women did not use the Internet for this purpose. However, 41% of survey respondents indicated an interest in learning more about using the Internet to find health information, with no significant difference between rural and urban womens interest (Winefield, et al., 2003, p. 31). In response to this perspective, programs and research attempting to reduce this barrier have provided study participants with computers to examine how the women would use the technology (e.g. Shaw, et al., 2006). Those studies have shown that women provided with
60 computers may use them to find breast cancer information, and these women reported more self efficacy in making decisions about their medical treatment (Gustafson, et al., 2002). The HIN TS survey ask s respondents about their cancer informationseeking experiences. Rutten and colleagues (2007) analysis of the 2003 and 2005 HINTS data found that nearly 40 % of respondents said it took a lot of effort to find, about one quarter said it w as frustrating to obtain, almost half said they were concerned about the informations quality, and over 20% said it was too hard to understand (p. 24). Although the rates of people reporting agreement with these statements in 2005 is high enough to cause health communicators some concern, all have dropped considerably from when the survey was administered in 2003, when almost 50% reported great effort to find information, about 40% feeling frustrated, nearly 60 % expressing concerns over information qual ity, and almost 40% finding the information hard to understand ( Rutten, et al. 2007) Minority U se of the Internet There have been studies aimed at defining and describing the Internet experience for racial and ethnic minority groups. Like other surveys of the Internet use landscape, many of them have gauged access to the technology in order to see if and where a digital divide exists. In their study of urban African Americans and Internet use for information about diabetes, Jackson and colleagues (2005 ) found that 46% of their participants knew how to use computers (based on self reports), 40 % owned their own computers, and the majority of people who did not have computer skills were willing to learn how to use the technology (p. 578). Although home c omputer ownership was limited to 58% of participants in a study of urban African American caregivers of pediatric outpatients, 93% agreed that there is useful information about health on the Internet and almost the same percent thought that online medic al information should be discussed with a doctor. Fifty two percent reported having
61 searched for health information online, and 82% of them said the information impacted their health decisions (Kind, et al., 2005, p. 118). The authors of that study put the findings in context of physician patient communication and pointed out that doctors should not assume that minority and low income patients and their caregivers have not used or been influenced by the Internet Effects of Internet Use on Breast Cancer Inf ormation Seekers While a great deal of research has investigated the use of the Internet by breast cancer patients, this likely differs substantially from the experience of someone who has not received a breast cancer diagnosis and seeks information about primary or secondary prevention of the disease. Unfortunately, extensive studies about what prompts African American women without cancer histories to seek or avoid prevention information on the Web have not been published in the literature. Therefore much of what is known about African American womens use of the Internet for breast cancer information comes from research with subjects identified as cancer patients or survivors. The extent to which this circumstance affects the studies relevance to this cu rrent project is unknown, but it is a probable issue. There likely is something profound about a breast cancer diagnosis that causes women to change their informationseeking habits. Fogel and colleagues (2003) found that for African American women who us ed the Internet during and after their breast cancer experiences, the technology provided them more perceived social support than the white women in the study reported. Minority participants in that study also saw the Internet as providing tangible benefi ts of possible concrete advice or directives relating to breast cancer (p. 114). Summary and R esearch Q uestions African American women have higher rates of mortality from breast cancer than do white women. Detecting tumors at later stages reduces survival rates for breast cancer patients and, based on some evidence, black women find and/ or begin treatment for their breast cancers at
62 later stages than white women. There are many hypothesized reasons for this disparity and likely it is the result of many fa cto rs, some of which have been described Promoting healthy lifestyles that may help prevent some breast cancer cases and encouraging African American women to utilize secondary prevention methods, such as mammograms, may reduce this mortality burden. The Internet has become a source of consumer breast cancer information for many people, and the technology may offer opportunities to deliver tailored and targeted health education interventions that more effectively reach women in different demographic groups, including ethnic/racial groups, women with disabilities, and women facing other barriers to enacting healthy behaviors. However, in order to maximize effectiveness of interventions and breast cancer information Web sites, more needs to be learned about who among these groups seeks breast cancer information, why they seek it, what they want from a Web site, and what they do with this information. This study acknowledges that beliefs about breast cancer and beliefs about the Internet both are likely to inf luence a womans decision to search for breast cancer prevention information. Therefore, the project sought to discover more about how African American womens perceived personal risk of breast cancer, beliefs about the disease and its prevention, beliefs about the Internet and beliefs about the Internet as a source of prevention information interact. This study does not seek to present a priori research questions to be tested by the data. Instead, as explained before, it proposes to develop a theory about African American womens use of the Internet for breast cancer information based on womens experiences. Extant literature does little to inform this topic, certainly stopping short of providing hypotheses about the relationships under study or even clear indications of what components Internet use, individual,
63 psychosocial influences on breast cancer beliefs, and perceived susceptibility to the disease, to name a few should be anal yzed in concert. Therefore, it was hoped that hypotheses about these re lationships would emerge from the developed theory and will direct future research in the area. That said, the following serves as the initial and overarching question: Among members of these two groups African American women with breast cancer family h istories and those without them are their reasons for using or not using the Internet more related to their beliefs about breast cancer or to their beliefs about the Internet ? The studys research questions are as follows: RQ1: How do the women describe their beliefs about breast cancer and breast cancer prevention? RQ2: What individual, interpersonal and cultural factors influence African American womens beliefs about breast cancer prevention? What do they think breast cancer means to their communities ? How do members of their social sphere and community talk about breast cancer? RQ3: What emphasis do women place on personal, proactive breas t cancer prevention information seeking? RQ4: What do they think about the Internet ? How do they describe the Int ernet s role or importance in their social setting or community? RQ5: What do African American women think about the Internet as a source of breast cancer information? For women who use the Internet for breast cancer information, how do they describe their experience? RQ6: What features or characteristics of the Internet enhance or deter womens experiences seeking breast cancer information?
64 RQ7: Among members of these two groups African American women with breast cancer family histories and those without them are their reasons for using or not using the Internet more related to their beliefs about breast cancer or to their beliefs about the Internet ?
65 CHAPTER 3 QUALITATIVE METHODS Th is project was a sequential, exploratory mixed methods design with qualitative paradigm dominance. The study consisted of an initial qualitative exploration of African American womens beliefs about the Internet as a source of breast cancer prevention information. This phase involved in depth interviews with women between t he ages of 35 60 years. Based on analysis of this qualitative data, a survey was created and administered to African American women in the same age group to provide quantitative complimentary exploration of the theory developed from the interview research This section will describe the qualitative paradigm, mixed methods approach to research, survey research methods, and the proposed study design. Qualitative Research The qualitative research paradigm comes from a viewpoint that reality is subjective and multiple. There are many truths as opposed to the one objective Truth that is more common in the quantitative research paradigm. This study assumes that the women participating in the research have views of, ideas about, experiences with, and meanings for the subjects that are part of their realities. Creswell (2007) cites the five philosophical assumptions about qualitative research as including ontological, epistemological, axiological, rhetorical and methodological. Ontology is the nature of the thing s we seek to know (Littlejohn, 2002; p. 28), and in the area of social sciences includes debate about actional and nonactional human behavior. This study reflects an actional perspective, assuming that individuals create meanings, have intentions, and ma ke real choices as opposed to being ruled strictly by biological and natural forces (Littl e john, 2002, p. 29). The women in this study we re assumed to be part of an active social system, wherein they
66 make conscious decisions and have interactions with other people who are also making decisions. As mentioned, part of the qualitative method ontological assumption is a subjective reality. The practical application of this assumption is that qualitative researchers present the participants actual words to show their perspectives. Participants words may be grouped by similarity into themes but the themes are descriptive and not absolutes. These themes are not intended to represent a larger populations reality In this study, rich descriptions of the data from transcripts were presented. This was done to ensure that the womens words we re interpreted as consistent with their meanings. Also, it allows readers to understand, dispute, or agree with the emergent thematic groups. Next in the list of qualitative method assumptions is epistemology, which covers questions concerning the nature, scope, and sources of knowledge (DeRose, 2005, 1). It considers the possible ways humans come to knowledge, either intrinsically or through their senses. Epistemological ass umptions in qualitative research concern the relationship between the researcher and what is being researched. Creswell (2007) says that the researcher takes efforts to reduce distance between him or her self and the research by interacting with the participants in the field. In this case, the researcher personally conduct ed the interviews with the women. Axiology, the philosophical study concerned with values, is the next important assumption and involves the scientists observational or participatory role in research (Littlejohn, 2002). Quantitative purists may believe that subjecting one s values to research or subjecting research to the researchers values taints it, while others believe that the initial selection of any study subject is essentially a value laden action (Littlejohn, 2002). The axiological assumption in qualitative methods requires the researcher to identify and present biases in the research by, for
67 example, reporting in the final paper both his interpretations and the participants interpretation s (Creswell, 2007). The section on study quality will discuss these matters more fully Rhetorical assumption asks about the language of the research. In qualitative research, the investigator may use qualitative terms and an informal style, including first person pronouns. This reflects quantitative research s basis in a science that seek s an objective Truth. Methodological assumption is about the process of research. Qualitative research employs inductive logic that emerges from the study itself (Creswell, 2007). The grounded theory used in this study fostered this emergence. Symbolic Interactionism The underlying theoretical perspective of this study wa s symbolic interactionism. Lindlof and Taylor (2002) describe symbolic interactionism as the study of how the self and the social environment shape each other through communication (p.41). It reflects the idea that meaning ascribed to communication or a symbo l comes from the context of the symbols use. The meaning a person ascribes to symbol s (concepts, words, etc.) is influenced by his or her experience with it, both personally and socially. This allows for and predicts different meanings in people and social groups, as well as a sense of dynamic, ongoing interpretation as new interactions a re added, as is consistent with the qualitative research view of reality. Associated with the work of George Herbert Mead (Charon, 1998) and explicated in Herbert Blumers 1969 publication, s ymbolic interactionism is based on the notions that individuals act toward objects on the basis of the meanings that these objects have for them, these meanings arise out of social interaction with ones fellows, and these meanings are modified through an interpr etive process by the individual ( Rogers, 1994, p. 170). As opposed to the idea that peoples sense of self evolves naturally and instinctively, Mead saw personal conceptions as being influenced by social processes ( Rogers, 1994, p. 168).
68 Symbols are important because they represent shared meanings. Humans in terpret symbols based on individually and socially determined ideas about the items. Religious art i facts are rather obvious examples of symbols because it is the ascribed meaning of the crucifix, for example, and not the material from which it was made tha t conveys significance to individuals. Language is another symbol a very important one in the study of communication. Charon offers some central ideas about symbolic interactionism. First is the notion that social activities among people are dynamic (p. 27). T he dictionary definition of the word dynamic includes ideas of change, power, influence, and group roles and processes. Charons next four ideas reflect many of these definitions. He goes on to say that it is not only interpersonal interactions but intrapersonal actions that cause human actions (p. 27) and that people define their situation as they go along in their action, implying a continuous or cumulative process (p. 27). Next, Charon says it is the present circumstances and interactions with t hings in the present that cause people to take an active part in causing their actions (p. 28). Symbolic interactionism has been the theoretical principle behind several studies of womens perceptions of breast cancer treatments and risks (e.g. Halkett, Kristjanson, & Lobb, 2008; Halstead, 2001; HoffmanGoetz & Mills, 1997). In the prese nt study, the perspective allowed the researcher to view participants as active players in making meaning about t he concepts in the study health, disease prevention, med ia, the Internet beliefs about life, and any number of potential symbols that arose from the study. The women make active choices toward these things based on individually and socially developed meanings about the represented items. Appropriateness of Qu alitative Design to the Study Topic This study sought to answer questions about a topic that has not been well studied in its target population. More specifically, these questions have not been studied under the proposed
69 combination of a new media content access model and public health behavior determinant parameters. When there is little or no prior knowledge about a topic, qualitative inquiry allows researchers to gather data from participants in order to generate a theory about the concepts under stud y ( Morgan, 1988). Morgan explained that qualitative methods are useful for exploration and discovery, context and dept h and interpretation ( 1998b, p. 12). In this particular case, the literature review show ed that the questions of why and how African Am erican women without breast cancer histories would seek primary breast cancer prevention information from the Internet have not been fully answered. Consistent with Morgans explanation of qualitative methods use, the study explore d beliefs about seeking breast cancer prevention information in the context of a new media content access model. The research examine d that question using in depth interviews to gain insight from women in the target population about why they do or do not use the Internet for that purpose followed by a quantitative approach to further explore the question. The qualitative design allow ed the researcher to delve into three related areas of inquiry: 1. The meaning s women make about primary breast c ancer prevention and mammograms 2. The meanings women make about the Internet as a communication technology 3. The meanings women make from online information about breast cancer prevention While these three lines of inquiry are somewhat ambitious for a single study, the purpose of this project wa s to understand two important and oft studied subjects in light of one another. As this has not yet been done in the recent published literature, it serve d as a test of whether or not such medium theory meets research application studies can be conducted. Fi rst and foremost, however, it allow ed the researcher to generate theory grounded in the data and possibly to use these findings to inform further explorations in this scholarship area.
70 Grounded Theory Although this study recognizes and mentions some of t he Health Belief Models components, the investigation does not use that model as its theoretical underpinning because the researcher believed that the use of or intention to use the Internet for breast cancer prevention information might be influenced by more variables than are described by that model. Instead, the study develop ed a theory from the information collected in these focus groups, producing a theory grounded in the data. Glaser and Strauss (1967) described grounded theory as a way of generating theories of social phenomena from emerging data observed or collected through social interactions. This qualitative method seeks to collect data, analyze it for themes or core concepts, and then collect more data to analyze in this light and integrate int o the framework. The cycle repeats and eventually produces a theory based in the collected data. As Richards and Morse say suggest a qualitative method such as this is useful for developing theory based on reality instead of based on the researchers bias es or prior research that does not explicitly match the questions in the current study (2007). The approach and interpreted meaning of grounded theory as a means of generating theory in research splintered after the early description by Glaser and Strauss Since then, various versions of grounded theory research have emerged One of those interpretations was the systematic process by Strauss and Corbin ( 1990). The grounded theory process begins by theoretical sampling choosing participants who will help a theory develop and then follows what Creswell (2007) describes as a zigzag pattern of collecting data from the field, analyzing the data, returning to the field for more data, and so on. New data are constantly being compared to existing data and emerg ing concepts; thus the term constant comparison is used to describe the process. Due to this cycle, the theory is based on inductive and deductive research approaches
71 unveiling concepts from the data and then testing the fit of those concepts with new da ta (Green & Thorogood, 2004). Coding Steps Grounded theory processes begin with coding the data. The first coding phase is open coding, a step that applies the question What is happening here? to each line of the first pieces of collected data (transcr ipt portions, in thi s case) and helps identify the core phenomenon being observed ( Creswell 2007). Strauss and Corbin ( 1990) refer to open coding as the part of analysis that pertains specifically to the naming and categorizing of phenomena through close examination of data, stressing that coding requires comparison making and asking questions (p. 62). Open coding deconstructs the data so they can be examined in smaller units. Each unit of data (words, sentences, or paragraphs) should be assigned codes or labels based on what it represents (Strauss & Corbin, 1990). Green and Thorogood explain that such an approach generates many potential conceptual labels that ultimately may fit or not fit the data, depending on what emerges from the rest of the collected data (2004). Once some codes in the early data are established, the researcher should constantly compare new data units to see if they represent the existing codes or new ones. After the codes are established, the theorist begins to group related codes into provisional categories (Strauss and Corbin, 1990) that are given abstract names to represent them in order to be memorable. The categories have attributes, which Strauss and Corbin ( 1990) refer to as properties and those properties occur along a cont inuum identified as the propertys dimensions that help researchers understand relationships between categories and subcategories Attributes of the categories are referred to as subcategories. Subcategories include causal conditions, strategies and inte rvening conditions, and consequences of using the strategies ( Creswell 2007), all of which relate to and qualify the core category.
72 Axial Coding. In the next phase, axial coding, the researcher reconstructs the data in new ways by making connections bet ween a category and its subcategories (Strauss & Corbin, 1990, p. 97). Relationships between labels are considered and these relationships help the researcher understand some of the bigger picture concepts at play. For this to happen, the researcher may explore which codes are conditions and consequences of other codes, therefore mapping out a relationship between the codes context. After these relationships are provisionally made, the data must be consulted again to verify the appropriateness of the rel ationships. Selective Coding In the final coding phase, selective coding, researchers see core categori es arise from the analyzed data, and they can start forming more abstract theories (Green & Thorogood, 2004) and develop hypotheses ( Creswell 2007). St rauss and Corbin describe selective coding as consisting of five steps: explicating the core story of the phenomenon in the form of a story line; explaining the relationship of the other categories to the core category; doing so according to conditions, c ontext, strategies, and consequences (p124); and validating relationships (p. 128) the researcher made by checking it against the data; and finally to make refinements to fill in category gaps (p. 141). Research Pro c ess & Interpretation Integrity In qu alitative methods, the researcher plays a key role in the research by his or her direct involvement with study participants by interviewing them, moderating focus groups of participants, and conducting deep ethnographic studies. The researchers presence l ikely influences whatever phenomenon is being studied, as her own biases will possibly affect her interpretation of the data. There are several concepts in qualitative methods that are employed to ensure high quality results and reduce researcher interference with the integrity of the study.
73 Researchers have suggested many qualitative specific approaches to research quality that are similar to the aims of quantitative researchs reliability and validity ( Creswell 2007). For example, Lincoln and Guba (1985) used credibility to refer to the researchs value as it reflects the presented reality, transferability to represent the findings application to other contexts, dependability to mean consistency, and confirmability as a reflection of neutrality. Lindlof and Taylor describe validity as a researchers desire to inspire confidence in readers (and themselves) that they have achieved a right interpretation (2002, p. 240) of observations. Johnson and Turner use more general terms to represent validity, sayin g that it means high quality research (2003, p. 299). Some of the terminology development by qualitative researchers may be interpreted by others as efforts to make qualitative research palatable to members of quantitative dominant research fields ; thus it may be discounted by qualitative purists. Creswell (2007) does not attempt to rename concepts ; rather he reinterprets them in the form of methods qualitative researchers can use to ensure a level of research quality that rivals that of quantitative meth ods. He says validation is an effort to assess the findings accuracy Qualitative methods can achieve accuracy by spending a great deal of time in the field, using thick descriptions in reports, and by establishing relationships with participants. Cresw ell says qualitative research can be enhanced through triangulating methods, peer review of methods and interpretations, clarifying biases at the front end, and using ex ternal audits and member checks. Keeping good records of these things may not only make results seem more valid to other researchers but it reminds the studys investigator to orient herself in the study as a strategic researcher. Reliability, according to Creswell (2007) is increased by the quality of field notes and transcribing and by using multiple coders for the data analysis and a codebook, and finding stable agreement on the codes.
74 Mixed Methods The qualitative research paradigm was described first because it undergird ed the purpose and approaches used in the study. However, this s tudy included a mix of qualitative and quantitative methods to explore the research questions more fully The use of mixed methods in social science research is based on the notion of multiple ways to ask questions about social phenomenon (Greene, 2007). M ixed methods use a combination of qualitative and quantitative research methods, as opposed to multi method studies that may use more than one type of qualitative method or quantitative method in one study (Tashakkori & Teddlie 2003). Greene says, A mixe d methods way of thinking is thus generative and open, seeking richer, deeper, better understanding of important facets of our infinitely complex social world. A mixed methods way of thinking generates questions, alongside possible answers; it generates re sults that are both smooth and jagged, full of relative certainties alongside possibilities and even surprises, offering some stories not yet told ( 2007, p. 20). In other words, using mixed methods allows for multiple pathways to understand complex ideas and to reveal unanticipated aspects of the studied phenomenon. This study type may also reveal connections between social processes and generate new theories about phenomena (Greene, 2007). Central to the justification of mixed methods research are the idea s that these different research angles may provide better understanding and that differences revealed by the methods are to be appreciated as opportunities for embrac[ing] dissonance, instead of opportunities to delegitimize one method (Greene, p. 23). M orse suggests both quality and efficiency justifications for mixed methods, saying that use of multiple methods gives us a more complete picture of human behavior and experience, allowing us to reach research goals for understanding the phenomenon faster than we might with only one method (2003, p. 189). Two different methods may also help overcome the obstacles or disadvantages inherent in each single method and answer questions that one method could not answer (Creswell, et al., 2003;
75 Tashakkori & Teddl ie 2003). Whereas mixed methods have comprehensiveness as a strength, their weakness may be a presumed lack of rigor. Use of clear data collection and analysis descriptions helps overcome others skepticism about the studys rigor (Morse, 2003). In fact, the qualitative paradigms techniques for ensuring high quality studies should reduce or eliminate concerns about rigor. One of the important aspects of conducting mixed methods research is understanding that the differences in qualitative and quantitative methods extend beyond study design into philosophical schools of thought about how knowledge is generated (i.e. deductively versus inductively) and the construction of the social world. Quantitative pursuits usually involve a priori constructs and hypothe sis testing whereas qualitative research seeks to allow the reality of human behavior to emerge. These stances influence the selection of inquiry methods. However, Greene (2007) points out several studies that suggest these kinds of epistemological stances are not always used in method choices, with the context of the problem being studied often influencing what methods are used. When using mixed methods, researchers may identify with one stance among many along the continuum of incorporating qualitative and quantitative philosophical assumptions. Each studys purpose will determine the underlying theoretical stance. For example, studies that seek to confirm relationships will be direct ed by the quantitative research tradition and deductive reasoning (Morse, 2003). Research decisions are guided by research purposes, which is different from the research questions. The research purpose requires the designer to think about the reasons for the study (Newman, et al., 2003). In this dissertation project, the resear ch questions African American womens beliefs about breast cancer prevention information on the Internet are different from the underlying purpose of asking those questions, which is that African American women with
76 breast cancer face worse mortality o dds than white women; therefore interventions to encourage black women to take steps to prevent breast cancer may reduce a health disparity. There are many reasons to mix methods, including triangulation, complementarity, development, initiation, and expa nsion (Greene, 2007, pgs. 100103). In this study, the mix of methods will be used for complementarity and, to a lesser extent, development purposes. Greene describes complementarity as appropriate when the study seeks broader, deeper, and more comprehens ive social understandings by using methods that tap into different facets or dimensions of the same complex phenomenon (2007, p. 100). With this purpose, one method offers insights that further and enhance the inferences from the other method. Development purposes are rather self explanatory; they involving using a method to help design another research instrument. In the first phase of research, themes and statements emerge. They are used in the next phase to help the researcher identify questionnaires in the literature that are appropriate for use in the study or to develop new questionnaire items, although the difficulty of translating qualitative data analysis into quantitative question items is said to be one of the disadvantages of this approach (Cres well, et al., 2003). Development is describe d by this researcher as a secondary purpose to the study because the development concept implies that the first research phase may only be conducted in order to develop the instrument, placing more importance on the latter phase. Mixed methods studies may be concurrent, with data collection through qualitative and quantitative methods occurring at the same time, or sequential, where one phase is completed before the other begins. Studies may begin with qualitativ e research first and then quantitative data collection or vice versa. Each phase of data collection and analysis is completed separately
77 and then findings and interpretations are integrated at the end of the study (Creswell, et al., 2003). R esearchers usi ng mixed methods can decide how to incorporate the philosophical stances of qualit ative and quantitative research treating the two methods as distinct but equal studies, each subject to its own paradigmatic assumptions, or putting more emphasis on one typ e of research for the studys interpretative lens. In studies w he re one method is dominant, the first phases findings serve as the main theoretical base and the other method supplements the first (Morse, 2003). This study use d a dominant less dominant sta nce in which the qualitative portion wa s considered more dominant than the quantitative ( Creswell 1994). Designs may also be explanatory or exploratory in purpose. This study will use a sequential exploratory design, with the qualitative portion of the r esearch being conducted first, followed by the quantitative method, with the purpose of exploring, not explaining, the phenomenon (Creswell, et al., 2003). This type of exploratory design is useful for exploring emerging theories from the qualitative portion by using a different research angle in the second (Morgan, 1998a ). The symbols QUAL quan signifies the direction of the research as well as the emphasis placed on each part (Creswell, et al., 2003). Sequential mixed methods designs beginning with qual itative phases first identify groups of similar people in the research and then use quantitative research to make comparisons between groups (Tashakkori & Teddlie, 1998). The quantitative portions of the study are used to add to the qualitative part (Steckler, McLeroy, Goodman, Bird, & McCormick, 1992). Research Design The research design involved recruiting African American women for interviews, seeking a mix of women who ha d and did not have family histories of breast cancer and a mix of Internet users an d nonusers. F amily history of breast cancer was an important distinction because
78 women with such histories tend to overestimate personal risks of the disease and therefore may have different health informationseeking patterns and different general approa ches to prevention (Haas, et al., 2005). Internet use status wa s o f obvious import; as the study wa s ultimately interested in the use of Web based b reast cancer interventions, it was necessary to understand why women do use the Internet and why they do not use it. Interviews Lindlof and Taylor (2002) say that interviews are particularly well suited to understanding the social actors experience or perspective (p. 173). Frey and Fontana (2003) distinguish between two extremes of interviewing approaches in structured and unstructured interviews by saying that in the first case, interviews ask the same questions in the same order to all the participants, but in the second kind, no a priori categorization that may limit the field of inquiry is imposed (p. 75). In other words, developing a preset interview guide would violate the spirit of unstructured interviews. A third, middle type of interview approach is the semi structured interview. In these data collection sessions, the researcher has a preset interview guide based on questions sought to be answered by the participants, but the researcher can insert other probing questions to further explore a participants comments. The researcher may ask questions out of order if the participants answers to a pr evious question lead into a lat er one. Lindlof and Taylor (2002) informally distinguish between structured and semi structured interviews by describing benefits and uses of associated tools. Interview schedules are structured sets of questions that are asked of each participant in a specific order, so that the answers can be compared from one participant to the next. In this study, a less rigid interview guide was used. This tool represents a rough grouping of study questions and topics but does not require that the questions be asked in a particular order or even in the exact words on the page. Instead, the conversation dictates the question flow and can be adapted to the verbal style of the participant
79 ( Lindlof & Taylor, 2002, p. 195). Semi structured intervi ews with an interview guide were selected to allow for the interview to bring up insights and topics that the researcher did not anticipate when conceiving the design. This approach is more consistent with the proposed grounded theory method, which advises researchers to reduce personal and professional biases when conducting the research in order for new theories about the phenomenon to emerge. The purposive sampling approach used in this study guides researchers to select participants for traits, experien ces, or characteristics that will yield insight into the phenomenons meaning, as opposed to using random selection techniques in hopes of achieving a representative sample. Lindlof and Taylor (2002) explain that it is appropriate to select individuals bas ed on their experiential knowledge, including membership in a social category or status they are [presumed to share] life conditions or challenges (p. 173). In this projects case, race and age we re two social categories and breast cancer family history and Internet use were two status variables for which the researcher purposively sought out participants for inclusion in the study. Questions and prompts on the interview guides varied slightly according to which session was being conducted. For instance, all participants were asked about their reasons for using or not using the Internet for health information purposes, but the self reported Internet users were asked questions specific to their experiences using the Internet for health information. The thre e categories of questions in the interviews were beliefs about breast cancer and prevention, beliefs about and use of the Internet as a general tool, and beliefs about the Internet as a source of breast cancer prevention information. Most questions were open ended and explore d how the women make meaning out of the concepts of breast cancer prevention informa tion seeking and Internet use, with the hope that these resulting data would help the researcher begin to understand how
80 the social, cultural, interpers onal, and macro level factors might influence womens beliefs about seeking breast cancer prevention information on the Internet The research questions posed were explored using interviews of African American women older than 35 who had not been diagnosed with breast cancer. This group was chosen because it include d women who are near the age (age 40 and older) at which annual mammograms are prescribed as a course of secondary prevention, so that women who are approaching this age or have reached this age could benefit from primary prevention information now and screenings in the future. Another reason why the lower boundary for age was chosen is that the women would be the next age group of women to benefit from Web sites about breast health. Of all the literature reviewed for this study, the most similar qualitative research design found was in a study conducted by Unruh, et al. (2004). They sought to understand what online breast cancer risk information women without personal breast cancer histories prefe rred, ultimately finding that women preferred sites from credible sources that offered personalized information and answers to specific questions. The research was conducted with focus groups divided by users and non users of the Internet Within the study researchers asked what barriers prevented women from using the information. Three of these elements women without personal cancer histories, online breast cancer risk information and the division of groups by Internet use status were common to this s tudy and yield ed helpful direction in the methods section of this paper. However, the two studies diverge d in four important ways: 1) P articipants in this study were also divided by absence or presence of a family history of breast cancer; 2) African Ameri cans exclusively participated in the interviews; 3) barrier questions concerned health beliefs and beliefs about the Internet ; and 4) this study s methods were in depth interviews
81 and a subsequent survey design based on the results of the interviews, inste ad of the cited studys focus group method. Theoretical Sensitivity Strauss and Corbin ( 1990) say that theoretical sensitivity describes the researchers personal qualities, expertise, and experiences that may be factors in her interpretation of meaning in the research. They say it is the attribute of having insight, the ability to give meaning to data, the capacity to understand, and capability to separate the pertinent from that which isnt (p. 42). Everything from expectations for the study outcome bas ed on those of prior research to the researchers own professional and personal experiences may influence how she perceives meaning from the qualitative data. Strauss and Corbin ( 1990) say researcher biases can be checked by consciously throughout the stud y asking oneself if the data are really supporting what the researcher believes she is seeing in the data, by remaining skeptical, and by adhering to rigorous methods for data collection and analysis (pp. 4445). The author recognized the biases researcher s can bring to qualitative studies, both based on her personal presence in the study and her expectations. In order to remain aware of the potential for biases and to try to limit their influences on the research, before gathering data, she described her r elationship to the topic. First in the demographic category, unlike the interview participants in the study she is younger and identifies as white. Similar to the interview participants, she has never been diagnosed with breast cancer. Second, related to the Internet use variable, the researcher is a heavy user of the Internet and regularly searches for health information as part of her job. She could not recall having ever sought information specifically about breast cancer prevention for personal reasons The answer to this question was slightly surprising to the author, as she considered herself a proactive health maintainer. However, she recognized that she had been immersed in cancer prevention information since her mid20s when the researcher first
82 began working with cancer organizations and on cancer related projects. B ecause she had learn ed about cancer prevention as part of her job, the researcher did not need to search for that information for her own health purposes. However, she does see her self as relatively proactive in terms of seeking health information and taking actions to reduce health risks. Interview Method The study recruit ed self identified African American women (Matthews, Sellergren, Manfredi, & Williams, 2002) from a large Southern city to participate in interviews. Local churches businesses and community organizations serve d as the primary recruiting venues. Because this project was concerned with women who have no personal histories of breast cancer, identifying women through clin ics or other health centers, as is often done in studies about breast cancer patients experiences using in the Internet for informationseeking, wa s not necessary. A short screening guide was used to determine if a woman me t the criteria for the study. W omen were excluded from the study if they ha d had a personal diagnosis of breast cancer or if they d id not meet the age criteria. Potential recruits with professional careers in medicine or public health (i.e. doctors, nurses, or similar professions as opposed to merely working in a hospital) also were excluded because the se individuals are expected to have specialized knowledge about health that probably would influence their attitudes and beliefs about the subject. A short questionnaire was administered b efore each interview to gather demographic information about the participants. Questions included yielded information about the womens age, education, Internet use and breast cancer family history. Although mental or cognitive functioning w as not measured, women had to be able to demonstrate understanding of their rights as a participant as well as show a willingness to take part in the study. A consent form approved by the Institutional Review Board (IRB) and signed by each interview participant demonstr ated her agreement to participate. The protocol submitted to IRB state d the parameters
83 of the study, explained how collected data w ould be used, and assure d that the participants identities would remain confidential ( see Appendix A with investigators co ntact information removed ) The interview sessions were held in safe, quiet and comfortable locations of the interviewees choosing, including a church, restaurant, and offic e building In accordance with traditional sentiment in qualitative research, the number of interviews conducted depend ed on when saturation was reached and the data being collected no longer offer ed new insights on the subject; rather it repeat ed what ha d been provided in previous sessions. No new themes emerged after the first four in terviews T he women repeated different versions of the same themes. For example, some women said they did not think they were at risk for breast cancer because they did not have a family history of the disease, while other women said they were at risk beca use they did have this history. In both cases, family historys associat ion with perception of risk was the theme. T he researcher continued to interview five more women in an effort to include many views and see if any more themes would emerge. However, no new themes emerged in subsequent interviews. An interview guide for the sessions is provided in Appendix B It outlines key topical areas that were covered in the interviews. Some of the questions were asked in every session, but others were presented di fferently or not at all with certain women For example, questions about breast cancer prevention beliefs we re relevant to every category of women, but questions about beliefs about the Internet were posed only to self reported Internet users, inquiring wh y and for what purposes they use the Internet Internet nonusers were asked a variation of those questions to understand why they do not use the Internet
84 CHAPTER 4 QUALITATIVE RESULTS Summary of Interviews Nine women were interviewed in depth about the ir Internet and breast cancer prevention beliefs. These women represented age, education level, marital status, and vocation diversity. Table 41 shows the nicknames and key characteristics of the women interviewed. Those women with a mother, sister, or gr andmother who had breast cancer are indicated by italics. Interviews were conducted in places the participants suggested would be convenient and comfortable for them, including workplaces, a fast food restaurant, and a church. The interviews took between 40 minutes and a little more than an hour each. The sessions were recorded with a digital audio recorder and the audio files were later transferred to a computer and transcribed and analyzed in NVivo7 by the researcher. Table 41. Women interviewed and the ir characteristics Nickname Age Relationship Education Level Ada 42 Married High School Betty 56 Married Some College Joan 43 Single College Joyce 46 Divorced Some College Linda 43 Single Some College Mandy 38 Single High School Melissa 32 1 Married Graduate School Rachel 38 Married Some College Tisha 38 Single College Although the interview participants represented a range of ages, education levels, Internet use, and experiences with breast cancer, major themes were repeated with su ch frequency that after four interviews very little new information was presented in additional sessions. Subsequent 1 One woman was found to be a few years younger than the initial minimum age level set but her intervi ew was included because her education and perspectives represented characteristics that enhanced the study.
85 interviewees answers fit within existing open coded themes. However, the researcher continued to interview women who had different demogra phic and other attributes that might offer unique insights into beliefs about the Internet and breast cancer prevention, or allow the researcher to make stronger inferences about themes. After nine interviews, the researcher determined that the no new them es were emerging The next section of this chapter presents major themes and subthemes from the interviews. While not every theme addresses a research question, many of the themes do provide insights into the concepts asked in these questions. A summary of each research question and its answer from the interview follows themes and sets of themes. Interview Themes B reast cancer b elief s This subsection of findings about breast cancer and prevention beliefs provides the qualitative data that support the summa rized findings. In addition to womens knowledge related beliefs about breast cancer, such as their beliefs that the biggest risk factor for breast cancer is family history major themes found include proximity to the disease affecting beliefs, a focus on secondary prevention, and changes in social views and science of breast cancer. A final theme that permeates the other themes is a lack of knowledge about the disease and its prevention. Beliefs about breast cancer related to the interviewed womens impressions of the diseases impact on other people, as well as beliefs about their own risks of developing breast cancer. For women whose families were not affected by breast cancer, the disease was a less concrete fear, meaning that the women thought breast c ancer was scary and often associated it with bad outcomes, but they spoke from a position of an unaffected observer or commentator. For women who had mothers, sisters, or grandmothers who had survived or died of breast cancer, these personal experiences served as a basis for the womens beliefs about the disease.
86 These experiences carried over to preventive behaviors, too. T he women who had experienced breast cancer in their families were more attentive to prevention strategies, such as screenings. Breast cancer prevention beliefs were also associated with the womens beliefs about breast cancer risk factors. Although some controllable risk factors, such as smoking and diet were cited, the women mostly expressed the belief that breast cancer is an inherited disease; a woman s inherited genes determined whether or not she would get breast cancer. With this belief that breast cancer may or may not be a womans genetic fate came a view of breast cancer prevention as almost entirely a secondary prevention domai n function. Inherent in the interview participants beliefs about breast cancer was a perceived (and likely real) lack of knowledge about breast cancer, its risks, and its prevention. At times, this theme was represented by the womens own admissions that they did not know much about breast cancer; at other points the lack of knowledge was apparent in that the women espous ed myths about breast cancer causes that leading cancer organizations dismiss as scientifically inaccurate. Proximity to breast cancer influenced womens beliefs about breast cancer. When first asked what breast cancer meant to the interviewed women, differences in perspectives based on experience were immediately evident Women who did not have family histories of the disease answered fr om a knowledge or fact based angle, such as Joyce who said breast cancer meant like a lump or a mass in the breast, Linda who initially answered with a definition, Having, uh, a cancer of the breast. Being checked for it, and Betty, who associated breast cancer with its impact on women more than men. Ada similarly thought of breast cancer as the diagnostic process and common treatments and outcomes, saying, I think about you having your
87 mammogram and they find a knot or whatever, and you get it tested and if its cancer, then they tell you what you need to dochemo. Having your breasts taken off. These definitional or process answers given by women who had not experienced the diseases effects from a close vantage point reflected perspectives of looki ng from the outside in on what breast cancer would be like. Contrasting this perspective was that of women who had been on the inside of their family members breast cancer battles and used that experience to anchor their responses. For example, Mandy said, Scary. Its been in my family. My mother, my auntand others. So its kind of a scary thing. Even though several women without breast cancer family histories cited fear of negative outcomes and side effects, such as Rachel who thought of her hair falli ng out and, dying, um, its just, its just something I cant explain. Um, its. its not a good feeling, women who had not experienced breast cancer in their relatives spoke more abstractly about what they had heard or read about the disease, rathe r than what they had witnessed up close. Proximity was a key theme in womens beliefs about their personal risks of breast cancer. This concept related to the womens perceived personal risks and the reasons they sought prevention. Overwhelmingly, the inte rviewed women framed risk of breast cancer in terms of family history of the disease. The women who had close relatives with breast cancer indicated a higher degree of perceived personal risk, whereas women without this family history mostly characterized their risk as low, citing the belief that breast cancer is largely inherited. Ada, whose only family members with breast cancer were in laws, said : Um, I never its not in my family, my immediate family, so I really I feel like youre kind of safe if y our mother havent had it or your grandmother, or you know, sister. I feel like I dont worry about it, you know. I mean, Ive had mammograms but I dont worry about it. I guess I figure if its in the family, youll get it, but if its not, you wont ge t it. Im not saying you cant, but I wasnt concerned because its not in my immediate family, so
88 Melissa, also from a family without breast cancer experiences, relat ed this history to her own risk. She said, So I think I have a low risk simply because no one in my family has it or no one has been diagnosed with it just yet, and I know that there are certain things that are hereditary. And Im assuming that breast cancer may be one of them. While most of the women who had not experienced breast cancer in a loved one said they had not really thought about their risk, or they thought the risk was likely low, the women who had this family history expressed that they had thought about the diseases impact and potential risk for them. Joan, whose family inc luded several breast cancer diagnoses, said she based her risk perception on research or conversations with physicians. She said, I guess I have to agree with the doctors and statisticians. I guess I have to say the genetics. She said that family history likely increased her risks but that a combination of a healthful lifestyle and diligence in terms of checking ones breasts and getting mammograms did not make a breast cancer diagnosis an inevitable event for her, explaining, My own risk? I believe it co uld be high because of the genetics and because its so strong in my family. Like I say, on my mothers side. But I guess Im not really worried about it because Im trying to take, you know, precautionary measures, trying to get my mammogram, eat healthie r, and exercise more. And I do self examinations every month. So Im trying to stay on top of things. Women with family histories incorporated their family members ex periences into many of their answers about breast cancer beliefs including their assessm ents of breast cancer risk factors. When Mandy, the daughter of a survivor, was asked what she thought put women at risk for breast cancer, she discussed her mothers health and the absence of things she thought would increase risk : I dont know because it s like when my mother was diagnosed with it, she didnt smoke and I had heard when I think about cancer in the first place, you know, I think about smoking. Its just a, its just a gene or something, that [breast cancer is] all in my family.
89 Mandys co mment about smoking being a risk factor for breast cancer but not one her mother had when she was diagnosed shows how women with close proximity to the disease filtered their beliefs about risk and prevention through their existing schema developed through their loved ones experiences. Mandys own personal perception of breast cancer risk was kind of high due to an awareness that she has a sensitivity to breast cancer because of my mom however, she also expressed religious faith that her fate was not determined only by genetics, saying, But I do think that just because my mother, my aunt, and my cousins had it doesnt necessarily mean that Ill get it, too. Im a believer in God and I dont have to have breast cancer. Also consistent with the the me of proximity to breast cancer affecting ones beliefs about the disease was an example of a participants friend developing breast cancer despite not having any of the risk factors family history and smoking the participant associated with the disease. Bettys schema regarding breast cancer risk may have been violated by this friends experience. Although Betty maintained a belief that breast cancer was hereditary, the story of her friend, who did not have the cancer in her family, made Betty think, it could happen to us. The effect of this experience on Bettys perception of risk is complicated by the fact that she had had some suspicious mammograms in the past and thus seemed to have thought about her personal breast cancer risk more than the other women who did not have family histories of the disease. Both the friends experience and Bettys own breast cancer scares put her in closer proximity to the disease. Of her own risk, Betty said: I think its a possibility. Like I said, Ive had a couple of personal scares. Ive had a mammogram that came back suspicious twice, actually. And had to have a biopsy, but it was fibroid tumors and sonothing, no cancer was found. But I think that the possibility is I wouldnt count it out. I wouldnt say that I have a great possibility, but I wouldnt count it out.
90 Perception of risk seemed to affect womens breast cancer prevention behaviors. Women who did not have family histories of breast cancer reported that they had had at least one mammogram in the pas t,2Tishas mother had breast cancer. She described her mothers ordeal of surgery, medic ation side effects, and getting a prosthetic bra, and the effect of that experience on her personally : but they described the screening behavior in more casual, routine, and less urgent terms than the women whose female relatives had had breast cancer diagnoses. For me, it was scary. I was hurt like it was me. You see what Im saying. So that threw me into checking myself every month. Being paranoid, I overly che cked. I mean, I checked every day, you know. But you would have thought it happened to me, you know. Tishas experience and view of screening importance was echoed by the other women with breast cancer family histories, who often referred to their relative s tumor detection and their perceived personal risk as prompting them to seek screenings. The women who did not have the same proximity to the disease stated very different reasons for getting mammograms and doing other breast cancer checks. Simply put, these latter women had had mammograms because the tests were conve nient (and sometimes free) and/ or because getting mammograms is something a woman is supposed to do when she reaches age 40. For example, Joyce said of her personal physician: She this is my regular doctor that Ive had since probably I was like in my 20s or something. And I would go every year to get a checkup and when I turned 40, she was like, Okay, its time for you to go have your mammogram done. And, okay, Ive had one mammogram don e. When asked why she had the mammogram when she did not think she was at risk for breast cancer, Linda said she did it because the doctor sent me over there: 2 Each womans physician should have instructed her about how often she needs to have mammograms. However, if the general guidelines of every one to two years were applied to the women over forty in the study, each should have had at least two screenings by the time the interviews occurred.
91 I had it done because they say at age 40 you need to have, um, a mammogram, and I just went and had a mammogram, and they said check your breasts and I checked my breasts and said, No lumps, soI just dont do it. Which I need to. And Im at fault for that. Ill own up to that part. In the last few sentences of that quote, Linda brought up an i nteresting point that low perceived risk plus an initial mammogram that detected no tumors may have actually reduced her perceived risk further by confirming that she did not have a problem (at the time). T wo women, Linda 43, and Joyce, 46, who both had had mammograms upon turning 40, did not have regular mammograms after that age. When prompted about why they did not have more frequent screenings, the women said they did not know how often mammograms were advised and believed that the doctors would likely have told them if the y needed more tests. Again, this notion related back to the womens beliefs that risk was tied to family history. Focus on secondary prevention. The interview participants were asked what things they believed could prevent women from developing breast cancer. Even though the questions terminology implied primary level prevention ways women could avoid ever getting the disease womens answers always included secondary prevention such as screenings and tests. In fact, none of the women answered the question about preventing breast cancer with only primary level steps; if a woman mentioned healthful eating or another primary behavior, she quickly moved on to describe secondary prevention actions and usually spent more time describing prevention behaviors on this level. Support for this idea comes from analyzing the womens answers and the amount of time devoted to each type of prevention in their comments. Tisha whose mother had breast cancer, is an example of this (for word count purposes, frequent ums were removed from this answer). She said a person might help prevent breast cancer by, Eating healthy. Of course making sure that you check your breasts. Preventionwise, I guess it would be probably just getting checked,
92 you know? Two of 23 words related to primary prevention and the rest were about secondary detection of breast cancer. Echoing the inference that actually getting breast cancer might be fated by genes, the participant wondered how much a person could really do to prevent the disease if she had genes predispos ing her to developing breast cancer. Tisha said, I guess to treat myself healthy and make sure I get regular checkups, umum, you know, as far as that goes, I mean I cant say that even if you did it, genetica lly if it was to be, it would be. Answers like this supported the connection between inherited genes and risk. If one believes her risk is determined by her DNA, she may believe the diseases development cannot be prevented. In that belief scenario, the l evel of prevention shifts from thwarting tumor development to detecting it early enough to save the womans life. Joan, another participant with a family history of breast cancer, also emphasized awareness, eating healthfully, and getting tested for cancer She devoted 16 words each to primary and secondary level preventions, saying, I think diet and nutrition is part of it. And education is a part of it. And, like I say, mammograms. And other whatever other means of early detection thats out there. J oyce pointed out some primary level prevention behaviors but emphasiz ed secondary actions, including checking for signs of cancer and being screened: I think like eating healthy, you know, watching your diet, you know, taking in the information of what you need to do, whether that is with your doctor or checking on the Internet what symptoms, you know, like doing your checkups, breast checkups, like they tell you to do at home, like when you go to the doctors office, they would do it, or either, you know, doing your mammogram. Betty came the closest to a pure primary level prevention answer to the question about preventing breast cancer by emphasizing diet and exercise. However, she also mentioned noting changes in your body and acknowledged that this conce pt was a secondary level of prevention, one that relates to the cancer already being in the body.
93 You know, as far as with diet and exercise. Things like that. And uh, just knowing your body and knowing when theres something that has changed in it that ne eds toand I know thats still kind of saying, Okay, its there but as far as prevention, its been more like diet and exercising, so forth. And different foods that you can eat that can help you prevent cancer. Other answers were much more overtly focus ed on secondary prevention. Clearly expressing this view of prevention as secondary was Lindas answer: Either they can prevent it longer, prolong it, prolong their life or they can eventually die from it. And Melissa reiterated the idea that preventing breast cancer means finding problems early: I think of a mammogram when I think of breast cancer prevention. It is important to note here that all but one of the interviewed women reported that they had had a mammogram at least once in their lives, a fac tor that might affect their perceptions of breast cancer prevention. The one woman who had not had this screening test was younger than 40, the age at which guidelines recommend most women begin regular screenings ( NCI, 2002). Even women who said they did not think they were at risk for breast cancer said they had received a mammogram, either because their regular doctor said they needed to have one at that age or because a free mammogram service was offered. The quotes below illustrate the similarities amo ng descriptions of prevention expressed by five of the women: Linda: Self check every month and before your cycle and after your cycle. They say to check in the mirror. And still do your annual breastwhenever you have to do your mammogram, kind of keep y our appointments. Betty: Um, the self examinations, um, regular, I guess at least once a year at a certain age for mammograms and make certain that they do that on a regular basis. And just be as knowledgeable as possible about what to look for. Mandy: Mammograms, um, self breast exams, go to the doctor. Basically just get the mammograms when youre supposed to and go to the doctor.Exercising and I just, basically, most of the women that I heard they do the tests, just the proper things to prevent it like the mammogram and the self breast exam. So Im a real believer in the mammograms and the self examining your breasts.
94 Joyce: Um, eating healthy, um, get your mammogram done to check and make sure that youre not, you know, forming breast cancer, and so forth, like doing the self checks at home. You know. I forgot how they told me to do it at home because I dont really do it but, you know, Ive thought about the self checks at home. Joan: What do I think about? Self examination, early detection, for myself, because it was a while before I even, um, (started) to get mammograms because they was asking my history and finally, I think, somebody finally noticed that how strong it was on the maternal, on my maternal side. So they started me at age 30 how ol d was I? Thirty four? Or maybe younger, between 33 and 34, because I know theres a certain age but because of my family history, they didnt want to wait. They started me immediately. So I get I have a mammogram every year. Changes and advancements in c ancer treatment. At the beginning of this research questions findings section, the researcher asserted that proximity meant something more than geographic or genetic closeness to breast cancer; it also was associated with prevailing and distinct social no rms and scientific paradigms concerning the disease. M any of the women in the age group interviewed were part of generations that had seen notable changes in how society and science treats breast cancer. Some of the women made clear distinctions between th e taboo ascribed to breast cancer by older generations and the more open advocacy efforts in younger generations. M ost discussion of this theme will be presented later in the chapter However, to the extent that interviewed womens beliefs about breast can cer were associated with improved technology, awareness, and trust in screenings to increase survival, the theme will be described here, too. Betty offered a good example of beliefs associated with social and scientific changes in breast cancer perspective s. She said, I think theres a lot of research available. I think that there are a lot of things in place now that have more and more every day the technology is increasing to the point that I think, you know, that there are better chances to prevent it and um, I think that doctors are more knowledgeable and I think overall that people are more knowledgeable because its in front of them. People are constantly youve got commercials, youve got Race f or the Cure, um, youve got those types of things in place, too, as a constant reminder of the importance of self examination and just preventing it if you can. And if not be able to prevent it to catch it as soon as possible so that your chances of survival are far better.
95 Although women believed that br east cancer might be unavoidable due to genetics, they placed some hope in technology and medicine to make the disease survivable. In other words, as Betty put it, If [secondary prevention] doesnt prevent you from actually getting cancer, it prevents you from being a statistic a death statistic. T here was a sense of women being able to look back in history to associate brea s t cancer with death, while also looking at the present and futures advancements in the ability to reduce breast cancer mortality. Two womens comments illustrate this point. First, looking back, Joyce said that when she thought about breast cancer, she thought of j ust so many people having cancer and ending up dying from it and stuff like that. Similarly, Melissa said, Most peopl e I have heard of who have had breast cancer usually die before it is treated. Despite their answers to the associative, firstthing that comes to mind inquiries, these and other women explained that they thought improvements had been made in cancer treat ment so that diagnosed individuals might not die from it. Within their explanations of this change were stories of friends and family members who had survi ved different types of cancer. Ada described two inlaws who had breast cancer, saying, The mother had to have her breast taken off, but the aunt, they was able to get it out and save her breast, so I know theres different ways that you can deal with it. You know, um, the one that they was able to get, get it out, she had chemo and the other one did, t oo, but theyre both doing fine. So I dont think of death, you know, now. I used to, but now that I know somebody thats been you know thats dealt with it for years, then I know theres a possibility that you can live longer. Similarly, Betty said, Ju st the word cancer, when you heard it usually you associated it with death. That when the person was diagnosed with cancer, no matter what form it was, they were going to die of it. However, her husband had survived a form of cancer and she cited him a s an example that cancer used to be a death sentence but she had this example of overcoming the disease with todays medical treatments.
96 The words used to demonstrate the womens belief that cancer experiences are not what they were many years ago. Eve n Mandy, who lost her husband to colon cancer, spoke of changes in treatment offerings in terms of her experience with him. When I think of cancer patients, I think of I used to think, until my husband died of colon cancer I used to think of peoples hair falling out. But now they have some things with cancer that make your hair grow. You know, treatment for cancer that make your hair grow even more so. Several other s interviewed said that technology and scientific understanding of cancer has changed o ver the years, perhaps leading to different outcomes for breast cancer patients today versus patients who were diagnosed in previous generations. Tisha reflected on her mothers breast cancer diagnosis 15 to 20 years earlier, noting that in her mothers ca se, the doctors had conducted the biopsy and mastectomy in a single surgery. Who knows if that nodule was, uh, that could have been caught before then. I think theres better technology now, Tisha said. She compared the decision to remove her mothers br east to a less invasive approach she believed might be more common today: Even back then, I dont remember people talking about radiation then, because they do that a lot, now before they actually say give somebody a mastectomy. In an earlier quote, Bett y explained that she thought womens chances of surviving breast cancer were better now than they were in the past. She cited advancements in many aspects of cancer control as well as awareness events as contributing to better outcomes. Joan described chan ges in technology that were not available when a family member was diagnosed with the disease. It seems like there is more technology today than there was what seven, eight years ago. And Im glad I mean, its bittersweet because you know how the mind i s, youre like, Okay, it wasnt here around when she needed it, but Im glad somebody is benefiting from it today.
97 Lack of knowledge about breast cancer and prevention. As some of the earlier parts of this section hint at, the womens beliefs about bre ast cancer often contained misinformation or incomplete information about breast cancer. For example, several women did not know how often mammograms should be performed. Women said they did not know how often they needed to have the test done or even if t hey needed to have subsequent mammograms after the first one detected no tumors. For instance, Linda said: To be honest with you, I dont know if mammograms are done every year, every two years ; I just know they get them at age 35 or 40, something like that. I had did them two times. But I hadnt had it again and I just turned 43 this year. So I know I hadnt had one in about two years. Lack of knowledge is a tricky belief theme because the phrase implies a comparison between one persons truths and an obj ective standard. However, this subtheme is appropriate in a discussion about beliefs for two reasons: one, the womens answers to questions about breast cancer beliefs sometimes included expressions of doubt or lack of knowledge, such as Ada, who said, I dont really know what causes breast cancer, and Linda, who said, Really all I know is that its just breast cancer. I dont know what causes it. Two, science does offer some clues as to breast cancer etiology. At the time of this writing, many question s remain about the diseases risk factors and causes, but extensive research has offered women some preliminary information about risks and prevention. Many of the women interviewed in this study did not report to have basic information about breast cancer For example, none of the interviewed women suggested a connection between growing older and developing breast cancer, even though age is the second biggest risk factor besides being female. Also, by consistently citing genetics, they seemed to emphasize the association between family history and breast cancer diagnosis. Even though a few women Betty, Mandy, and Tisha -mentioned smoking as risk factors for cancer, and Rachel mentioned overweight, the women almost always reported that they
98 thought fami ly history determined risk or that they did not know what increased risk of the disease. Rachel was the only woman who reported common myths, including wearing bras and men grabb ing on your breasts. The lack of knowledge about breast cancer risk and cau ses carries over to a lack of knowledge about prevention. This makes sense, as someone who does not know what causes a disease would not know how to prevent it. Rachel said she did not understand breast cancer causes but would be pleased with outcomes and positive emotions associated with prevention, saying, Prevention, it sounds good because theyre it makes me feel like you know the women have hope. But it just seems like, I dont know how you get breast cancer but prevention does sound good. Simila rly, Ada seemed familiar with the idea of prevention, even though she did not understand what it was or if it was really possible. She said, Ive been at health fairs, at [place of employment] and at different places and I would pick up a brochure and look at it. But I couldnt tell you anything that somebodys told me how to prevent breast cancer because Ive never known that there was something that just caused it. I just thought there was something in your body or was hereditary. I didnt know there was something you could do to get breast cancer. This last sentence was the major theme in the study. The womens association of risk with hereditary genetics something that is, at the time of this writing, unable to be altered moves the level of prevention from primary to secondary except in women with genetic risk factors who elect to have preventive mastectomies without being diagnosed with breast cancer Throughout the interviews, the emphasis in womens answers to questions about breast cancer prevent ion was not about avoiding the disease, rather about avoiding death from it. Answers to Research Question 1. These themes shed light on Research Question 1, which asked how the women describe their beliefs about breast cance r and breast cancer prevention. The women described beliefs both in terms of what they knew about the disease itself, such as breast cancer being a lump in ones breast and the disease affecting more women
99 than men, and its impact on women, especially women they knew personally. They de scribed risk as relating to uncontrollable genetic factors. Insomuch, women with breast cancer in their families seemed to feel more susceptible to the disease than women without this risk. Due to this focus on genetic predisposition to breast cancer, the women interpreted prevention as early detection screenings that do not prevent developing cancer; rather the tests reduce mortality by finding cancers when they are more treatable. Community Beliefs and Interpersonal Communication about Breast Cancer Experiences with breast cancer influence communication about and the communities views of the disease. The main influence of breast cancer beliefs in the interviewed women was experiences with the disease, including having loved ones who had faced it or having had a mammogram that spotted something suspicious. Breast cancer was seen as less of a problem in the womens communities compared to other diseases that were more prevalent. Although many of the women who did not have family histories of breast cancer sa id they did not talk about the disease, their reasons were that it had not affected them or their friends ; therefore the topic had not come up in conversations. While the women described some negative connotations of breast cancer held by some in their com munities that had once made the topic taboo, the women also recognized a shift in the way breast cancer is viewed and talked about due to awareness efforts that may de stigmatize the disease and medical advancement in treatment and detection that increase survival rates Witnessing changes in medical technology, various cancer outcomes in people they know, and seeing more emphasis on awareness messages may have developed (or prompted shifts in) womens beliefs about breast cancer prevention. As the first fi nding that womens perspectives of breast cancer were influenced by their family members experience with the disease has been described in answer to Research Question 1, the findings in this section primarily address the other social and community the mes
100 revealed by the research. Those themes are 1) the tendency to measure the importance or priority of diseases based on how many people in the community are affected by them; 2) the spectru m of breast cancer perspectives from ascribed taboo, embarrassment, and moribund qualities on one end, to awareness, increased survival, and advocacy on the other; and 3) the notion that interpersonal conversation about breast cancer occurs in the context of someone the women know having the disease. Disease priority relates to quantity of observations. The women seemed to determine problem priority based on pervasiveness of the condition as opposed to severity of effect on peoples lives. Mandy, who personal ly believed breast cancer was an important issue because it w as pervasive in her family, still said it was less common in her community than other diseases, like diabetes, high blood pressure, because I know a lot of people that have diabetes and high blood pressure. From the interviews, it was evident that other diseases and problems were more common in the womens communities, and the women suggested that the more common a problem is, the more it was talked about and the more serious it is High blood pressure, heart problems, and diabetes were mentioned by almos t every woman interviewed. Mandys explanatory phrase because I know a lot of people indicates an association between ranking particular diseases as priority problems and observing how many people are affected by the conditions. Tisha reinforced this i dea that the importance of a disease relates to prevalence in the community. Because breast cancer wa s less prevalent in her community, it received less emphasis. She said: Because like right now, a lot of people are just more concerned with heart conditi ons and stuff like that. High blood pressure. Thats the more thing that people are talking about as far as being more out open than they would breast cancer. Nobody would.nobody sees [breast cancer], its like I said, unless it happens to somebody or som ething like that. Melissa also brought up the idea that problem severity is associated with personal exposure to
101 the problem in the community, High blood pressure and strokesbecause weve had people in our community who have suffered from high blood pres sure. A lot of family a lot of my family members have problems with high blood pressure. And weve had people who have either um attended our church or are attending our church now who have had strokes. And I also have a grandfather who died of diabetes.I would say [breast cancer is] mainly last on the list. Betty said something very similar: As far as like hypertension and heart disease and strokes and things of that that natureamongst the people that I know, I know more people who have high blood pr essure, so to speak, certainly than Ive known who had breast cancerSo I know that breast cancer is important, but I personally know more people are hypertensive and so forth than I know who have been stricken with breast cancer. Spectrum of breast cancer perspectives based on advocacy and scientific progress A previously mentioned finding suggested the influence of individual proximity to breast cancer on a womans beliefs about the disease. Similarly, generations of people may have beliefs influenced by their proximity to changing social norms about breast cancer and scientific advancements in cancer treatment and detection. Even though the interviews resulted in a range of interpretations about how communities approached cancer topics, the answers refle cted past versus present and old versus modern views about the disease. The past themes involved breast cancer mortality and embarrassment about having the disease. On the other end of the spectrum were the present themes, including a belief that eliminating the problem of breast cancer means making people more aware of the disease and supporting and encouraging breast cancer patients and survivors. By past and present, it does not mean that the association s of death and shame with breast cancer beliefs ar e extinct. The interviewed women however, drew distinctions between two disparate views and suggested that the advocacy, openness and awareness theme correlated with social and scientific changes younger generations associated with the disease. Of note, t he women did not always clearly identify their own places along the spectrum of community beliefs.
102 Past views. Women said the low survival odds for breast cancer in the past had shaped community members association of the disease with a death sentence or very negative side effects of treatments. This association had several effects One was that some people thought only bad people got the disease. The second, and likely somewhat related issue, was embarrassment by those people who had the disease and fear of lost social support. The third mentioned outcome, also related to the first two, was that people did not talk about breast cancer. Fear of breast cancer is still very common today ( Wider, 2005) but the womens assessments of past views of breast cancer included a strong sense that lack of knowledge about the disease and myths about the disease striking people of bad character likely enhanced the pasts fears of cancer. The difference seemed to be that whereas women now may fear getting cancer, in previ ous generations, there may have been a stigma of the complicit victim; that the women did something to get breast cancer Betty said: especially the generation thats just before me and the generations before that, for some reason or another, cancer was like something that only happened to bad people. So there was ait was hush hush. It was like you had the plague, but it was something bad you had done and I dont know where that came from. But I know that I have experienced it with some of my older rel atives, that they were that it was just hush hush because [whispers] T heyve got cancer, the C word. And it was like it was a curse instead of something that could happen to anybody. Betty and Ada both used the term hush hush when describing how pe ople once viewed breast cancer. Adas comments highlighted the possibility that the short and longterm side effects of cancer treatment might be another aspect people did not want to think about or discuss: I just remember people it was kind of quiet, you know. You know such and suchs wife has cancer? [ Ada whisper ed ]. And it was kind of like you really didnt talk about it. You didnt say any kind of cancer.
103 Whether the stigma about cancer stemmed from beliefs that a person who got it was inevitabl y going to die from it or that the side effects of treatment were embarrassing, this view of cancer as a taboo subject affected how people talked about the disease. Tisha shared that she thought fear of cancer, a lack of knowledge, and peoples association of the disease with death likely affected womens views of bre ast cancer prevention. She said : I think theyre scared. They feel like some people now might think it could be the end of your life, per se, as far as like, if it happened to you (at a) youn ger age, they might think you cant get married, you know, that you cant do this or that. Rachel expressed a similar opinion about peoples views of the outcomes of cancer: A lot of black people, when you say cancer, they just stop and act like youre g oing to die. The mortality concern was one of several reasons cited for why women would not be inclined to talk about breast cancer. Others were fears of losing social support during the treatment, embarrassment about side effects, and the desire to prot ect ones health privacy. Representing the social fears about breast cancer was Joyces comment about womens fears that they will lose social support systems if they get cancer: Sometimes family membersyou know a lot of people are scared of the word c ancer. And a lot of people dont you know, they tend to shy away from you so you dont have that support if you actually did end up and have breast cancer or something. Ada thought that people were reluctant to talk about breast and prostate cancer becaus e sexual side effects for men and mastectomies for women were common and embarrassing outcomes of treatment. And Joyce felt that women with breast cancer wanted to keep the condition private because they would probably think that people will treat them di fferent, or talk about them like, Oh you know, they done lost this weight and they had to have this done. Linda actually shared an experience that supports Joyces suggestion. Linda learned how important it was for a friend with breast cancer to contain and control who knew about her health.
104 And since [getting cancer], my friend had to have her breast taken off. And what happened, I thought I had done something bad, I told her ex husbands mother that she had a mastectomy and she called me and blessed me out, said, You shouldnt have told them I had one of my breasts removed. She was embarrassed. Speaking further about this experience with her friend, Linda explained that the incident had ended their friendship. She wanted to keep it quiet she didnt want nobody to know. Like it was a secret, she said. Joyce also suggested that women who did not talk about breast cancer were probably embarrassed or ashamed. Advocacy and Openness. Interviewed w omen often referred to progressive views about breast cancer, even if the older view s about death and embarrassment persisted in their communities Comments about increasing awareness, participation in cancer charity events, and interpersonal communications about breast cancer were much more common in the women who had family histories of breast cancer. That subgroup of womens conversations will be discussed later. However, it should not be assumed that lack of conversation or awareness of breast cancer is related to beliefs that the disease is taboo. Many of the women interviewed responded to a question about how women in their communities talked about breast cancer with a simple reply like Lindas answer: We do not, Melissas answer, We dont discuss it really at all, or very much, or Joyces, We don t. We never have. These women all had the same answer to a follow up question about why they di d not talk about breast cancer: the disease has not affected them or their friends or family members. As Linda explained, You talk more when it happens to you. You get more involved if its you or somebody close to you. Similarly, Tisha said she and her friends are open to discussing breast health. I have talked to lots of people my age in the community about fibrocystic di sease, not breast cancer itself, Tis ha said. Fibrocystic disease is very common in
105 her peer group, whereas with breast cancer, she did not know a lot of people who have been affected by that besides her mother who is deceased. Within the subgroup of interviewed women who did have family histories of breast cancer, the theme was consistent with the idea that people think and talk about conditions affecting themselves and their loved ones. It is open and told, said Tisha, whose mother had breast cancer, as did Mandy and Joans mothers. These latter two women shared similar views. Mandy said: I ts not a sensitive subject. After my mom was diagnosed, her first born niece was diagnosed with breast cancer... Its kind of like, its not a delicate subject. We can talk. Its not a sensitive subject for us. Its okay. We can talk about it. She spoke of a breast cancer survivor aunts advocacy within the family. The aunt, who is still living, um she has educated us a lot about breast cancer and breast cancer awareness and precautions and such. She reported that one of the reasons female family members talk about breast cancer frequently is to encourage each other to get mammograms. Joan also attributed conservations about breast cancer to the fact that her family has experienced it first hand. We communicate a lot anyway and because of the fact that it has occurred so much in our family So it comes up. I mean, we dont talk about it much because weve discussed it for so long. I mean, were comfortable with it. You know, theres no tension, theres no problem talking about it or discussing it. Even women who did not have family histories of breast cancer serving as context for talking to their relatives about the disease had noticed changes in society and their communities views of breast can cer. Ada thought the changes in her local communitys approach to breast cancer communication resulted from having people in the church congregation affected by the disease. My church is pretty open with it, she said, and described the churchs participa tion in several health fairs and breast cancer awareness events. In the larger society, Ada attributed some
106 breakdown of the taboo to increased survival and to increased attention to cancer among famous people. She said, I think a lot of things thatsyou know, people are living longer, you hear good things; the cancer walks. Um, people talk about it celebrities I mean, people are really talking about it. Theyre more open about it. Um, I think you know, its not whispered. And it could be because of th e years. You know, but its not whispered anymore. They just talk about it like any other sickness now. Betty, who earlier had mentioned the older generations take on cancer, also spoke of the openness about cancer due to experiences with it by people in her community. These individuals, according to Betty, are using their experiences to help educate others about awareness. Betty was involved in a church health ministry that included breast cancer education events She said: I know that it is very importan t to get the word out and to talk to people that you need to do whatever you can do to prevent. And thats just with any woman, you know, any time the subject comes up or anyone has the suspicion, get yourself to the doctor, have a mammogram, have yourself examined. Be sure. Dont think about it. Similar to this awareness of breast cancer and the move from discussing it in whispered tones to more open dialogue about the disease, Tisha, whose mother had breast cancer, mentioned her friends efforts to proac tively raise the profile of breast cancer prevention in their community. Im in a sorority and we do functions and we do walks together for that, Tisha revealed. She also referred to a breast cancer education group her friends had formed to pass out pamp hlets and materials at malls and health fairs. Several other women recalled breast cancer events and breast cancer awareness efforts, including Susan G. Komen events, pink ribbon magnets on cars, and a local television news anchor who survived breast cancer and tries to raise awareness in the community. Answers to Research Question 2. The second research question asked about the individual, interpersonal and cultural factors that influence African American womens beliefs about breast cancer prevention. It also asked w hat the women thought breast cancer means to
107 their communities and h ow members of their social sphere and community talk about breast cancer The themes that emerged from the interviews did not answer the first sub question exhaustively becaus e it seemed that despite some differences in the womens backgrounds, most held the view that breast cancer was closely tied to genes and that genes were uncontrollable. The other factor that may have influenced views about prevention, at least in terms of being diligent to adopt secondary prevention behaviors, was having a family history of breast cancer. Answers to the sub questions regarding the communities treatment of breast cancer varied. Some women cited old views of previous generations that bel ieved cancer was to be dreaded, it was almost always a death sentence, and that people with the disease may have done something to cause them to develop it. More modern views recognized that technology and medicine had advanced to the point that cancer was often survivable. Women did not say that their communities held one or the other view exclusively. Instead, like many of the women interviewed expressed of their own views, the fear of cancer may be coupled with recognition of improved treatment opportuni ties. Despite views that cancer is associated with negative outcomes, most of the women cited other health problems as being more dominant in their community. These women referred to perceptions of that high blood pressure, diabetes, and other diseases aff ected more people in the community, therefore these diseases may be higher priorities or at least more commonly talked about in the community. Most women without family histories of the disease said they did not talk about breast cancer much in their social circles, mostly because they said they did not talk about things that were not affecting them personally. Women with family histories reported that they did talk about the disease with family members.
108 Prevention Information and Internet B eliefs Womens p erceived lack of need for breast cancer prevention information The women mostly had not looked for primary breast cancer prevention information. This seemed to stem from the lack of perceived personal risk of the disease due to family histories free of br east cancer and the related belief that the disease is a genetic fate that cannot be prevented, only detected early through secondary prevention. Of the women with family histories of the disease, only two said they had sought information recently and neit her had looked for primary prevention information.3Given the rare event of women seeking out prevention information, the interviewed women were asked more general questions about health informationseeking. The majority opinion was that women would be unlikely to seek information about a health topic unless it was affecting the women personally or they had family members with the condition. This proximity to the disease concept answer, either in the form of a loved ones diagnosis or a personal exper ience with it, was common to both breast cancer information and general health topic and diseas e information seeking prompts. All of the women who had this family history expressed some beliefs that they remembered the information they had received around the time their loved one was being treated for breast cancer, and thu s they did not feel the need to actively search for additional information. Among the women who did not have a family history of breast cancer, only one said she had proactively sought breast cancer information. She did so after a personal scare and said s he did not look for prevention information at that point. Just as it was uncommon among the interviewed women to seek information about preventing breast cancer, the women thought other w omen in their community also did not look 3 One wanted information about new technologies to detect tumors. The other needed some information for a class she was taking.
109 for such information. As Tisha put it, I dont know anybody who has looked up information for that particular disease, myself. Mandy said, Its not a common thing. The women in my community, they arent going to get out and look for anything. If it comes up, itll be from them going to the doctor and the doctor, you know, telling them what to do. The two reasons women gave for why they might look for information about breast cancer prevention were a family membe r getting the disease or having a personal scare. Linda said she did not search for this information because it never crossed my mind. I didnt have it and no one close to me has it to make me want to be curious. Ada said she would not look for informati on on a topic like breast cancer unless it was something dealing with me. And Joyce said, Well, I just.I guess I dont. If nobody is affected with it or if its not affecting me directly with it, or somebody I know, I just never thought to just look for it. From the perspective of a woman whose family members were affected by the disease, Mandy surmised that most women do not seek this information because because I dont think they know the seriousness of breast cancer. I dont think theyve done the research on it unless its happened to them. I think thats the reason why. Women were consistent with the theme of proximity to breast cancer when the question was rearranged to ask what would make them or other women want to find information about breas t cancer. Whereas women said they had not looked for breast cancer or prevention information because the disease had not affected them, they replied to the question about what would make them or another women seek information by saying that a family member being diagnosed or the women themselves finding a lump or having a suspicious mammogram would cause them to learn more about breast cancer.
110 Tisha said a personal scare or a family member having breast cancer would cause a woman to look for more informati on about breast cancer. She said: Maybe some people might just go on there just generally for information but that probably will be small, a small number. I would think people just pretty much go on there to learn about something thats happened to someone they care about or know about. Joyce also said women would not look for this information unless it is concerning them, referring to the women or their families. But like if its pertaining to my family, I will [go on the Internet ] so I can try to get a better understanding of it. Ada also said that she would only look for this information if she experienced someone else someone close to me getting it. Like a mother, sister, close family. Similarly, Rachel said women would likely only look for brea st cancer information if somebody in their family has had it. She said, Maybe because its hereditary, you knowin black peoples communities, uh, they get more interested in stuff if they feel like like I told you before like their aunt or their mother or their grandmother had it. Thats when they get more concerned. This theme of seeking disease information if a family member is diagnosed seem ed to have two dimensions. One is that this information would help the women know how to help their loved ones. For example, Ada described how she sought information for her husband after he was diagnosed with cancer in order to find the best treatments available. Joyce said she would look for information about a condition affecting a family member in order to offe r that person support and to be prepared for the treatment process, Because I would want to know. I would want to know what other people are facing, you know. What I could do to help that person. You know, to talk to that person, to comfort that person, l ike, Oh, its going to be okay. You know, whether theyre taking chemo, if its going to make them sick, throw up, you know, lose their hair. Whatever. What I could do to help that person. The other reason related to family members diagnoses wa s the pos sibility that the womens own assessment of personal risk might increase when their relatives are stricken with the disease. Tisha put it like this:
111 Say a family member has had one of the diseases and they go a lot of people I know might have been a p arent, see, that has had the disease or whatever, and then they go and do extensive research themselves because they want to find out ways to prevent it. Like if that parent had it and a couple of sisters or brothers had it, then we want to find out how to prevent it. B reast cancer scare prompting information seeking The women mostly thought they would look for breast cancer information if (meaning after) they received that diagnosis. When asked why women might look for information about breast cancer, Linda said, Yeah, because they have it. Because they got it. Rachel thought she would only look for breast cancer information if she had tests that indicated she might have the disease. First of all, like the only reason why I would look for that informati on is if I had like a positive mammogram or something. Ada said she would not look for breast cancer information unless it affected her directly. In fact, Ada said that she was not prompted to learn more about the disease even when women close to her were diagnosed with it. Um, unless I had a scare, I mean, unle even when my friends and my mother in law had it, I didnt try to go out and find any information. I didnt go and try to find outno, I didnt. She said her mother in laws experi ence did not pr ompt information seeking because the woman was not a blood relative, referring to the inherited gene association with risk. Betty also surmised that a personal scare would make her look for breast cancer information, but unlike Ada, she would also look for information about diseases that affected other members of her family. She said: You know, if somebody if I went to the doctor and they said, You have a nodule about this size I would probably seek it out and you know go after all the information I could about it as it pertains to me and what is said about me or certainly my h usband or somebody that I know.
112 Joan, whose family had been affected by breast cancer, thought other women would not be interested in learning more about the disease unless one of the two conditions of personal experience or a family members diagnosis made the women curious: Somebody Im going to be honest with you either they have to be diagnosed or somebody they know or a family has to be diagnosed because a lot of times we dont really want to search out or find things until it actually happens or we know somebody that it happens to. Another woman, Joan, who had a family history, said she sought information because she was just curious, but not unlike the theme of looking for information after a family members diagnosis, Joan said this search for knowledge was to see what detection technologies were available now versus when her sister had breast cancer seven or eight years earlier. Three women associated the lack of per ceived ri sk with the lack of informationseeking by tying their opinions back to my projects interviews. Most generally, Rachel said she thought a lack of knowledge about breast cancer would prevent women from seeking information. She said, I think one re ason why black women most black women dont know anything about breast cancer is because of denial and lack of information. Um, its just uh, um, its a lot of things in our community that we just dont know anything about. I dont know some of us, a lot of us do want to know about breast cancer and some of us just dont care, but I do want to know about breast cancer because I want to know how to prevent it and uh how can I help someone else prevent it. This seemingly paradoxical statement about lack of knowledge as a barrier to seeking information may be explained by community and cultural views of breast cancer, as well as the belief among the interviewed women that breast cancer risk relates to family history and therefore cannot be prevented. Ada and Betty made statements that may support this idea that a lack of knowledge about breast cancer its causes, risk factors, prevention, and survival rate may make women not want to know about breast cancer or not feel vulnerable enough to the disease, based on lack of
113 family history, to look for information. Ada suggested that participating in the interview about breast cancer was leading her to think more about the disease and her risk. These thoughts might make her want to learn more about the disease and prevention so she could avoid it. I would try to find out what I need to do so I wouldnt be in thatcounted in that theory. I would want, you know, to do better. Cause like I said, I didnt know that it was something you can do to prevent it; I jus t thought it was hereditary so that would be inform a lot of information for me because then Id find out that its not. Betty predicted that other women interviewed might be prompted to learn more about breast cancer prevention after speaking with me abo ut it. I mean, some people you might find that you interview that as you begin to ask questions about it, it will kind of plant questions in their minds to make them want to go and you know just simply ask like I said, I use Yahoo to say Okay, Im an African American women, 56 years old, with da da da da da da da da, certain weight, what I eat, da da da da daWhat are my chances? You know, What are the risk factors that I should be looking for and um, what do I need to do to best take care of mys elf? She said that even women who had not been interviewed might sometimes seek information after viewing a billboard about breast cancer awareness or a TV program about the disease. Melissa also said women might seek information if they heard something on the news. However, the fact that some interviewed women said they probably had seen news stories about breast cancer but had not actively sought information after that suggests Melissa and Bettys theory did not reflect the other women in the qualitative study. Answers to Research Question 3. This research question asked w hat emphasis women placed on personal, proactive breast cancer prevention informationseeking The answer was that most women had not searched for this information; therefore little e mphasis seemed to be placed on proactive breast cancer prevention informationseeking. Internet non use related to physical access Those interviewed women who were not regular Internet users gave logistics centered reasons for their nonuse, including lac k of physical access and time and low Internet skills, as opposed to beliefs about the usefulness or
114 trustworthiness of the tools information. The women who used the Internet had positive views about the medium. They described multiple anticipated tasks f or which they use the Internet such as for e mail communication and information seeking. This subgroup of interviewed women was very good at describing these specific functions they used the Internet for but only a few women interpreted the questions about the mediums role outside of concrete ex amples of it use. The women spoke of the Internet allowing people to do some of the same tasks they have always done (pay bills, renew library books, find movie times, or research medications), but quicker, easier and more privately. These womens view of the Internet s role in their communities was that it generally mimicked the role in the interviewees lives. They thought their friends and neighbors likely used the Internet for purposes similar to their own. Th e women described the Internet in very simple terms. For the women who used the Internet a great deal, the medium was a tool to do many things at work and at home. They appreciated its benefits more than they were frustrated by its outcomes. In fact, most of the statements by these women were very positive. O ne of the women waxed eloquent about the Internet s impact on her life by describing how things were done before the Internet came along: It makes information a whole lot more accessible to people beca use um, Im just at a unique age where I can remember life without computers and Internet and where you had to dig a lot deeper and a lot harder in order to find. The other Internet users described what they used the medium for and what they liked about it; the latter issue will be described in detail in the discussion of subsequent research questions In no case did the women give abstract answers about the Internet s role in a larger social context, such as acknowledging the impact of a digital divide on solidifying class distinctions.
115 For the women who did not use the Internet or did not use it often, the primary barrier s were time and access. These women did not seem to have negative opinions about the Internet ; rather their needs and life priorities did not include the medium. However, even one woman who was not very familiar with the Internet had a computer with access at her home because she viewed it as integral to her childrens futures. This woman, Linda, explained that her employer at her housekeeping job (one of three jobs she had) had given her a computer and she paid for Internet access to help her children do schoolwork without having to go to the library. So Id never let my computer Internet go out. It cost more, a lot of money, but I just keep it going for the kids. Theyre in school, she said. The least frequent Internet users gave answers indicating that their non use was not related to suspicions of the Internet or beliefs that it was a bad tool. Mandy said, simply, I dont have one. I dont have access to the Internet I dont even own a computer. After several probes, this woman volunteered little about why she did not use the Internet saying only, Its not that I dont like it. I just dont have time. I be busy all the time. I ju st be busy all the time. But this woman did say she had asked other friends and family members to look for information on the Internet for her, indicating that she saw some value in the medium. Linda, the woman who wanted the Internet for her children, sa id she used it sometimes. I dont know much about it. No more than [laughs] playing games, suggesting a lack of familiarity with the medium. She also cited the three jobs she worked and the lack of time for things like the Internet People do use the In ternet more. Like I said, you have some who are on it every day, all day. But Im not. It depends on your job. When asked what role the Internet plays in their lives, the womens answers were very simple and to the point. Its an everyday function, Joan said. Melissa and Tisha both said the
116 Internet plays a major role in their lives, as they use the medium every day and for multiple communication and information purposes. Betty and Ada also described a range of tasks for which they use the medium. Bett y laughed, called it the information highway, and said it had become part of her life. Ada summed up the Internet s purpose as a user driven tool as allowing one to find out different things that you want to find out. Joyce and Linda both described the Internet as being something their children were learning or had learned about in school and the influence of their childrens Internet use on the womens interpretation of the mediums value. Linda said, Cause you got information right at hand. You can key it in. Once you get the use of it, cause my kids download and burn everything, set up the computer. I dont know how to do none of that. I let my 16year old do all of that. Because theyre taking it at school. So thats how. My 12 year old, she knows because she take it at school. She goes to [hi gh school name]. So they know. And Joyce associated her own use of the Internet as opposed to its nonuse by other women in her community as bei ng related to having children. I dont think, well, with me being the age that I am, well, a lot of people, they dont have kids growing up, and you know, using the Internet or computers and stuff; I dont think they really just go on the computer or Internet. I just have two grown kids so I made sure I had a computer in my house when my kids were growing up, so Im used to being on a computer. A lot of people dont work with computers, you know. If they work in a coal mine or plants or something like that, they dont have a computer that theyre working on with their jobs and like doing Internet stuff so I dont really think theyre into it. Joyces view of why some women do not use the Internet segues into the womens answers to questions about the Internet s role in their community. Again, none of the women answered i n analytical terms, such as the impact of the Digital Divide, the need for Internet skills for success in some careers, or cultural beliefs about the Internet s utility. Instead the women spoke more concretely about how people in their social circles used the tool. Melissa summarized a common answer that she thought her communitys use of the Internet was similar to her own. I think, for
117 the most part, my community uses the Internet, um for the same purposes that I use it for checking email and just fin ding information, for just news, current events. With the exception of Mandy, who reported to know only one other person in her community who used the Internet most of the women said they thought use was very common among their friends, family members an d neighbors. Some womens conjectures indicated that older people and those without children living at home might be less likely to use the Internet ; however, womens answers to the questions about the Internet s role in their communities indicated that th ey believed the medium had largely diffused to reach the older and childless, too. Tisha said, Um, a lot of older people on the Internet Um, you know, young, old. Uh, middle aged. Um, everybodys on it. Theyre looking for a recipe, theyre looking for or trying to find information about how to buy tires, or something like that. They look on the Internet pretty frequently now, a lot of people in m y neighborhood, old and young. Betty made similar comments, suggesting that the myriad features of the Interne t made it a useful tool for a range of people. Joan echoed this sentiment, responding to the question about what the Internet means to her community by saying: I mean, it plays a broad role because theres so much you can do with it. I mean, there is so mu ch that you can do on the Internet And I think it plays a very big role for those who are willing to go out there and access it. Because you can find any information, anything you want to know is out there. Answers to Research Question 4. This research q uestion asked w hat women thought about the Internet and how they described the Internet s role or importance in their social setting or community The women who used the Internet said they thought it was a useful tool for a variety of tasks. Very few of the women cited negative aspects of Internet use. Even the women who did not use the Internet much did not express negative beliefs about the Internet. Those women merely cited lack of time, access, and skills as reasons why they did not use the medium. Alth ough the women did not speak abstractly about the positive and negative impacts of the
118 Internet on their community, such as the advantages it might offer users in the workforce or the disadvantages associated with the Digital Divide, the women suggested th at other people in their communities likely used the Internet in ways similar to the womens own online habits and purposes for use. A belief that the Internet would be a good source of info about breast cancer, if a woman wanted that information Most of the interviewed women had not looked for breast cancer information in the past. Those who did were prompted by personal or family events, rather than simple curiosity or an interest in primary prevention. The women thought that the Internet would be a good source of breast cancer information if a woman had a reason to look for it. In other words, if a woman wanted to learn more about the disease, the Internet would be a good tool for her search. However, the interviewees beliefs that women are unlikely to seek out information about a condition that is not affecting them revealed that the Internet would not be a good standalone primary prevention intervention because women without family histories of the dis ease probably would not use it. Interviewed women said the Internet would be a good tool for finding information about breast cancer, but a woman would be unlikely to look for it unless it was affecting her. Joyce said, The best way to find out is through the doctors or the Internet And if you dont have [breast cancer] you wouldnt just, Oh, let me just go to the doctor and let him tell me about breast cancer. Betty said she knew of some women who had used the Internet for breast cancer research and said they found some very good information conce rning exactly what they needed for themselves, so she would be likely to use the medium if she needed information about the disease. Ada also spoke hypothetically that if she wanted to learn more about breast cancer, she would turn to the Internet I mea n, I think its good. The information, I mean, she said, and
119 went on suggest that she imagined the Susan G. Komen Web site probably had some useful information about the disease. In summary, the women thought, based on their experiences looking for other information online that the Internet would probably be a good source of information if a woman wanted to look for information but might not be a good method to share prevention information with women who were not actively trying to learn about the diseas e. Tisha explained, you can pass out pamphlets in the mall and [women] might look at them. They might throw them away. But you just pass it to each woman. But if its on the Internet more than likely somebodys going to actually go and take an interest and look it up themselves, because the Internet doesnt just pop up. Like a pamphlet, somebody just gives it to you. You really have to go resear ch and you look it up yourself [on the Internet]. Another theme was a preference for doctors advice. Mandy said, The women in my community, they arent going to get out and look for anything. If it comes up, itll be from them going to the doctor and the doctor, you know, telling them what to do. Linda also said that she might listen to what a friend said about a disease, or consider the information from a Web site, but normally she would go to the doctor to learn about a health condition. The things Melissa said she might use the Internet to find out about are health questions that she did not feel comfortable asking her doctor about. Tisha said she might look on the Internet but the majority of the time, if its something I want to ask, Id wait until I go to the doctor and Id ask them, per se, and just go from there. She continued to describe some concern over the quality of information on the Internet and how well that information applied to her situation and her ideals, [the Internet ] has some good information on it but if itslike if I had a rare cancer or something like that if I looked it up on the Internet a lot of stuff that might pop up later might say, You need to you might eat this certain type of grass that might help you. Or over here might say, You take this certain type of drug. But of course, you know you got the natural healing and the medicine healing, you know. And depending on what type of uh, I guess Id say what kind of background you come from, what you do, which one are you going to go with? So, I would ask the doctor instead.
120 Answers to Research Question 5 This research que stioned asked w hat African American women think about the Internet as a source of breast cancer information. It also asked how women who use the Internet for breast cancer information describe their experience. Few of the women had actively searched for br east cancer information in the past. However, most thought they would use the Internet for that information, if they perceived a need for breast cancer information in the future. The women who had searched for breast cancer information described successful attempts at getting answers to their questions. Health information seeking on the Internet is a common activity Active b reast cancer information seeking using any medium was uncommon among the women interviewed, but the women did report use of the Inte rnet for other health information seeking. Women who did not have family histories of the disease indicated their noninformation seeking behavior was related to factors other than the medium, such as low knowledge of breast cancer and low perceived risk of the disease, rather than feelings about the Internet as a breast cancer information source. Women who did have a family history of breast cancer said they had looked for information on the disease when their relatives were diagnosed and felt confident in their knowledge of the disease and prevention. Therefore, a simple answer to the question is that no features or characteristics of the Internet were cited as enhancing or deterring these wo mens breast cancer information seeking. This is not to say that certain features or characteristics would not be facilitators or barriers to seeking information; rather the question cannot be answered with the available interview data. Among the women who did not use the Internet for health information seeking, the pr imary obstacles were lack of time, lack of access, or lack of computer skills, so it is technically possible that associated characteristics such as cost, infrastructure, and complexity of the Internet acted as an obstacle to breast cancer information seek ing. However,
121 the women in this category did not express that they had in the past or currently wished to use the medium for breast cancer informationseeking but were deterred by these characteristics. Given this finding, the interview questions about breast cancer information seeking were re conceptualized to describe more generally womens impressions of the Internet s features for finding health information. Also, women were asked a hypothetical question about what Web site cha racteristics they woul d prefer if they sought breast cancer information. The three main features and characteristics women cited as enhancing their health information seeking experiences were convenience, the wealth of available information, and the simplicity of searches. Con venience of searches. For women with computers and Internet connections at home, the convenience of using the medium at any time was very appealing. Ada explained the convenience by comparing the tools use to the methods people would have used before the Internet : Instead of going to the library and you have to go through all these books, um, you can just put in one thing and get, you know, all kinds of different information. Even Linda, who said she did not know much about the Internet said it was impo rtant to her to have a computer with Internet access at home to keep her children from having to go back and forth to the library. Cause you got information right at hand. You can key it in, Linda said. Its a good feeling because you dont have to go to the public library for one. And you dont have to go check out a book, and thats my main source why I got the computer, because I didnt realize that my kids was going to the library just to use the computer. Once we got it, we saved a lot of time. Th e library replacement notion came up with Tis ha and Melissa, too. Tisha said: You can find a lot of information quick. You dont have to leave your home. Say, you dont have to go to um, go to the library and pull up a book and then go through the different topics to find exactly what youre looking for. The Internet just kind of labels topics and then just say, okay, boom, and it gives you just kind of a brief summary under it so you can just say, yeah, thats what I want. And if its not what you wa nt, you can
122 just come out of there and look for something else. So its better than say going to the library and they say you cant get more than five books a t a time, something like that. And Melissa said: Its easily accessible, more so than the library because you dont have to drive down to the library. And its just more time consuming to go so somewhere else versus using the Internet because all you have to do is turn the computer on and, you know, youve got informatio n at your hands, in an instant. The issue of privacy, which is often cited in the literature as a reason why people like to do health searches at home, only came up in those specific terms in the interview with Tis ha. She said: Some people do it in the privacy of their home. You know, t hey wont feel like, S omebody sees me going out here, if its something personal going on with them. I know people that are really like, I dont really want nobody to know about that, you know. Some people are, some people dont care. But its good if they can have the access in their home, and you know, if they couldnt get out of the house they could look up stuff and see how could they take care of themselves. This issue may go back to the previously discussed idea that health and especially breast cancer are very private matters for many people, and they may be more likely to look for information about health concerns when they believe they can do so without anyone finding out about it. Simplicity of searches and the wealth of information These two themes are discussed together because so many of the womens quotes regarding what they liked about the Internet mentioned these two concepts in one sentence. For example, Betty said about the medium, Its pretty easy its usually pretty easy, Ill say to find what youre looking for and it gives you a lot of choices and a lot of different areas that you can look into, uh in order to find that answer. Similarly, Joan noted that positive characteristics of the Internet include the access and the ease of finding the information that youre looking for and the broad scope of information thats out there.
123 The simplicity of searching for online health information involved two related concepts: the search engine technology was easy to use, requiring just a few keywords and mouse clicks, and this method resulted in efficient, even intuitive, answers to the womens questions. The distinction is largely a matter of steps in the process. Tisha described the first aspect of the simple search the ability to fin d information without having to t ype in a specific Web address or uniform resource locator ( URL ) She said: I like Google, I mean, because its easy to do. Because pretty much if you type in the topic under Google, it comes up with whatever it is youre looking for. If you type in www dot something in the World Wide Web, its gonna take forever, and youre going to have a bunch of nonchalant uh, no nonsense, you know, its not going to be direct. It seems like if you go to Google, its direct. Rachel sai d the information just pops up after she enters keywords for a topic. I would just go into address bar and just type in lymphedema, or newest research on lymphedema and thats it, she said, describing how she would search for information on that condition and what her expectations would be for that process. Joan called the Google search process user friendly. The second aspect is that the keyword entry method is effective; it locates the information the women seek and presents the information in a way that enables them to find answers to their questions easily, either on the search engine page hits list or on the actual Web pages generated by the search. Joan said, I mean, theres so much there are so many sources, especially when you Google stuff. When you put in a word and you have all of these sources that come up and you know you get to see it from different aspects or different views. Betty referred to this presentation and organization of Web page content as the useful way the medium cate gorizes things so that a person can move from topic to topic. She said, just follow the line. Follow the yellow brick road through it and see where it leads in terms of topics and information that can be learned.
124 Tisha also explained this concept tha t of being able to quickly see if a Web page has the information she wanted: The Internet just kind of labels topics and then just say, okay, boom, and it gives you just kind of a brief summary under it so you can just say, Yeah, thats what I want. A nd if its not what you want, you can just come out of ther e and look for something else. Many of the women liked being able to scan a Web page quickly and see, due to the way information is presented in headings and subheadings, if the information wil l be useful to them. Joyce said: It tells you everything. It tells you what I mean, if youre looking for just like you said, breast cancer it tells you what causes it, what it is, whether its a mass, what to look for, what procedures. It breaks it dow n, stepby step. If you have to have surgery, what type of surgery it is, what they do in the surgery. It tells you about the chemo. It just breaks everything down, step by step. They saw this as being a major advantage over more traditional forms of print material research. Presumably, it is an advantage over television and radio information sources that do not allow for user control, although neither of these media was mentioned. Two of the women did say that the Internet s health information was more us eful to them than brochures and information in doctors offices when the women wanted answers t o specific questions. Joan said: Because its easier to access, its easier to find. For me it is. I mean, because you cant just walk into a health facility, I mean, they dont have pamphlets on everything. And especially their technology its brief, and I guess I wanted a l ittle more in depth knowledge. When Joyce described a recent online health search, I asked her why she had chosen the Internet for her sea rch and she said, There just wasnt a lot of information in the doctors office about the rare condition she wanted to learn more about. Another key theme was the wealth of information about health topics that one can find online. For instance, Betty sai d, And then, I cannot emphasize enough the information thats available on it, you know, and how easy it is to attain that information. Joan referred to the
125 broad scope of available information. She said, You can find any information; anything you wan t to know is out there. Answer to Research Question 6. This question asked about the f eatures or characteristics of the Internet that enhance or deter womens experiences seeking breast cancer information Due to the low reported cases of breast cancer in formation seeking, the specific question could not be answered. Overall, women who used the Internet for information about other health topics cited the ease of searches, the way Web sites organize content to make answers to questions easy to find, the wea lth of available topics, and the convenience of searching at home or other place at any time as things they like. Limitations or drawbacks included content quality and accuracy issues and a preference for communicating with doctors about health questions. Summary of womens beliefs about online breast cancer information seeking Reported reasons for not using the Internet in general were related to lack of physical access, Internet skills, and time for online activities, not beliefs about the medium. Reaso ns for not using the Internet to find breast cancer information were related to perceived personal risk of the disease or family experiences with breast cancer, not beliefs about the Internet as a source of that information. In fact, the women who had low perceived risks reported that they had not actively sought information about the disease from any source. Among women without breast cancer histories, the interview participants beliefs that they were not at personal risk for breast cancer seemed more res ponsible for their lack of informationseeking behavior than any belief about the Internet as a source of this information. This association has already been discussed. With the exception of the low Internet user s the women in this particular category rep orted to have sought online information about other health conditions, indicating that they believed the medium was a useful source. In addition, some of
126 the themes presented earlier in this chapter show that women believed the Internet would be a good source of information if they wanted that information. However, with their lack of perceived personal risk, the women did not think about looking for this information. The women who did have family histories of breast cancer said they believed they were at r isk for the disease, but only one said she had looked up breast cancer information recently (the secondary prevention screening information sought by Joan). The women did not report a belief that the Internet was a bad source of information about breast ca ncer. In both groups women with and without family histories the interviewed women had in common a belief that they did not need more information about breast cancer. Women with histories believed they already had the information they needed; women wit hout histories did not express the vulnerability to the disease that might prompt them to want information. Answers to Research Question 7. The final research question asked if among members of these two groups African American women with breast cancer f amily his tories and those without them their reasons for using or not using the Internet were more related to their beliefs about breast cancer or to their beliefs about the Internet. The answer seems to be that beliefs about breast cancer influence thei r beliefs about the need to seek prevention information. However, this answer should not imply that women who believed themselves to be at risk would automatically be prompted by this belief to seek prevention information. While women who did not have a f amily history of breast cancer said they would not be likely to search for this information unless they had a personal scare or a family member was diagnosed with the cancer, women with the family history might not search for information if they believed t hey already knew everything they needed to know about breast cancer. Another clue that beliefs about breast cancer influence decisions to seek more than beliefs about the Internet is the womens comments
127 that if they wanted more information about breast ca ncer and prevention, many if not most of the women with physical access to the Internet would use the medium for this purpose.
128 CHAPTER 5 SURVEY METHODS Chapter 5 will describe the role of the survey in this study, how the survey questions were developed, the sampling method used, and data analyses performed The chapter will also show how the literature and need for further exploration of research questions contributed to specific questions on the instrument. Role of the Survey in the Study After the in terviews we re an alyzed, they contribute d to the development of a questionnaire that further explore d emergent themes. Chapter 3 explained, in greater detail, the general advantages of using two different methods within one study. The role of the second met hod in this study was to complement the findings from the interviews. In this case, the word complement refers to following up on the specific themes that developed in the interviews as well as examining some variables and relationships between variables that the interviews were not equipped to examine An example of the latter is exploring the role of age, income, and education level on beliefs about breast cancer prevention. Statistical analyses of these reported variables in relation to answers to ques tions about beliefs and Internet use allowed the researcher to see if correlations exist. Advantages and disadvantages of survey methods Surveys traditional purpose is to find out something about a large population by using statistics and sampling techni ques to extrapolate findings from a small, representative group of people. The interest is in seeing how certain traits are distributed in the sample and, by extrapolation, in the larger population the sample is thought to represent (Babbie, 1990). This em phasis on counting cases and using statistics to make interpretations about the sample and projections about the population it represents places survey methods within the quantit ative paradigm. However, as Chapter 3s
129 section about mixed methods explained, surveys can still be used to complement qualitative research by further exploring the themes that emerged from interviews. The difference is one of interpretation; this present study ma de no attempt to claim that the findings from either phase of t he rese arch we re representative of the larger population. While some surveys purpose s are to explain relationships, this survey only describe d and explore d the topic in more detail. Surveys can be conducted in a variety of ways using many formats ; however they all consist of sampling, question design, and data collection (Fowler, 2002). Quantitative research designs usually use probability sampling to increase likelihood that the randomly selected sample represents the population under study and statistically ac counts for chance. This sampling technique and an appropriate sample size allow the data to be explored with advanced statistical analyses that can point to significant and sometimes causal relationships between variables. While quantitative only studies use large, probability sampling, qualitative focused projects may use smaller, purposive ly selected sampl es so as to select only those cases that might best illuminate and test the hypothesis of the research team (Kemper, et al., 2003, p. 279). Although s urveys offered this study the advantage of being able to explore themes emerging from the qualitative research phase using well crafted questions, survey methods ha ve some drawbacks. Most of the disadvantages reflect survey researchs roots in the quantita tive objective paradigm as they strive for some clinical, uniform conditions and control over the study design. Included are selection biases in respondents, low response rates, responders misinterpretation of questions, costs associated with administerin g the surveys, and the possibility of people providing socially desirable answers to questions as opposed to answers that reflect their real opinions (Fowler, 2002, p. 98). Some of these disadvantages, such as selection
130 bias, are difficult to overcome wh i le others such as the tendency of respondents to give socially desirable answers can be mitigated through good instrument design. The person (respondent) is the unit of analysis in this study. The population in this study was African American women aged 3565, without breast cancer histories. These race, gender, age and disease history variables represent the parameters. The survey used in this study was written, self administered and structured, meaning that the questionnaire had preset answer options a s opposed to blanks for people to write in their own answers to every question The survey was paper based, as opposed to online, bec ause the research questions sought feedback from some wom en who do not use the Internet, and an instrument the participants read and completed themselves. Self administered surveys have the advantages of allowing people to answer sensitive questions without having to share that information directly with an interviewer and letting the researcher ask many related questions about themes (Fowler, 2002). However, the researcher recognized that vision and literacy issues might have preclude d some participants with useful thoughts to share from completing the written form. Disadvantages specific to self administered surveys include t he lack of control over who answers the questions and the quality of those answers, the possibility that literacy issues may limit responses, and the limited practicality of including open ended questions (Fowler, 2002). The survey was administered in some group settings. Although group settings may not be appropriate for selecting probabilistic samples, they offer logistical advantages to the researcher, including opportunities for her or a representative to explain the study to the group, lower costs, and good cooperation rates (Fowler, 2002; Johnson & Turner, 2003). Groupadministered surveys were appropriate for the present study because of the need to reach a minority within a
131 minority, meaning the eligible participants represent a smaller percent of pe ople who are already part of a racial minority. By identifying existing social groups of African American women in the studys age range, such as religious groups and community organizations, the researcher increased the number of survey respondents fitting the studys parameters. Survey Development The survey w as developed after analysis of the interviews. Themes and subthemes that emerged in th e interview stage inspire d the surveys questions and design. Not every theme or subtheme was addressed in the su rvey, as some would require asking many questions to explore the concepts. Themes that were incorporated included a womans proximity to breast cancer influencing her beliefs about the disease, which was probed in the survey by asking women if they had fir st degree relatives with breast cancer and then comparing answers to other questions according to whether the respondents had a family history The theme of women emphasizing secondary prevention prompted the researcher to include items about mammography where she originally sought to focus on primary prevention. The theme about womens lack of knowledge about breast cancer was partially explored with a question asking women what they thought were risk factors for the disease. An interview finding that wom en tended to talk about breast cancer more if they had a family member who had experienced the disease prompted a question asking women if they had talked to another woman about breast cancer. To explore the conditional part of the theme that women with family members who had breast cancer might be more likely to talk about the disease the question asking women if a female relative had had breast cancer was used with the question about breast cancer conversations. The theme of community disease priorit y relating to observing how common the problem is in the community was not explored in those exact terms, because that would have been
132 complicated in this survey; rather the researcher tried to see if the same other diseases mentioned in the interviews wer e cited by the surveyed women as community priorities. The theme that Internet nonuse related to lack of physical access to the Internet was incorporated into several survey questions, most notably in a question asking women who did not use the Internet r easons for their nonuse, with a list of choices that included physical access as well as belief based reasons. The theme about womens perceived lack of need for breast cancer prevention information was addressed in the survey by asking women if they had looked for prevention information, results of which might imply need (although it is fully understood that perceived need does not always translate to behavior). A theme that women were prompted to look for breast cancer information after experiencing a p ositive mammogram or having some other personal scare was addressed by asking women who had sought breast cancer information for the reasons why they searched for this information. Answer options included a personal scare. A final theme that the Inter net would be a good source of information about breast cancer, if women wanted that information was incorporated into the survey with questions that asked women what Internet features and characteristics would prompt them to, or not to, use the Internet for breast cancer information, if they needed that information. Also, a belief question asked women how likely they would be to use the Internet to seek breast cancer information. Surveys were anonymous and confidential. Information asked include d basic de mographics such as age, education level, relationship status, number of children, family history of breast cancer, and income level (see Appendix D). Other questions included beliefs about risk factors for breast cancer, the time a woman spends online in a typical week, and her beliefs about breast cancer prevention. Every persons questionnaire had the same questions but
133 the instrument instructed some participants not to complete all the questions. For example, the questionnaire ha d items asking if a woma n might look for information about breast cancer on the Internet. If she answered that she would do this, another question asked her what Internet features might make her want to use this source. Women who had answered no to the question about using the Internet for prevention information were directed to check reasons why they would not do so. The survey protocol and questionnaire form were submitted to the University of Floridas Institutional Review Board for approval. The interviews, the literature d escribed in Chapter 2, and the research questions contributed to the survey instruments development. In its role as complement to the interviews, the survey explored themes and subthemes that emerged from the interviews. In addition to this interpretation of complementing the interviews, the surveys were able to pose questions in a different, close dended format that helped answer research questions the interviews could not explore very thoroughly. A prime example of the latter is the research question tha t asked what individual factors might influence womens breas t cancer prevention informationseeking. Although the interviews suggested some factors, the surveys allowed the researcher to compare wo mens self reported informationseeking with their demogra phic information, family history, and other variables. One very important point that will be discussed in the limitations section of Chapter 7 is that the desire to measure many different themes and concepts in a single survey yet not make the survey so long that it discouraged participation was an important consideration. Women who reviewed the survey before it was finalized suggested that they would be unlikely to complete the questionnaire if any more questions were added to the instrument. Because this study is exploratory, the researcher opted for a cursory overview of many major themes instead of an in -
134 depth examination of only a few themes. As a result, many themes were explored with only one or two questions. Another important point to note is tha t not every qualitative theme was explored. The reason for this is that the survey had limited space to explore questions and some themes were considered too difficult to examine with only a few questions or they were deemed less relevant to the research questions. In some cases, a subtheme of that theme was explored on the survey. One example of such a case was the interview theme of openness and advocacy about breast cancer. While this was an important finding, it would have been difficult to assess with a few questions. Ultimately, the subtheme about womens discussions of breast cancer relating to personal or family experiences with the disease was included because it was more directly related to the differences in beliefs and behavior by women who do a nd do not have family histories of the disease. This discussion of question development will not describe the survey questions in their chronological order because that order was influenced by question format (questions were grouped by answer format in an effort to simplify the instrument for respondents) ; rather it will describe themes and the questions that relate to each theme. Questionnaire Format and Pretest Format and scale decisions, such as whether the questionnaire use d Likert scale or agreedisagr ee items, were made based on existing instruments in the literature and the format the researcher thought would provide the best design for each question. Survey literature provided some general guidelines for question design that inform ed the instrument d evelopment. Fowler (2002) advises survey researchers to use little variation in answer option formats to limit confusion. In other words, the survey avoided excessive switching back and forth between different category scales, agree disagree questions, and multiple choice answer items. The
135 instrument made efforts to use common, well defined, familiar words instead of terms that might be too ambiguous and to avoid asking multiple questions within one question. The questionnaire was reviewed by ten white and African American women in the studys age range b efore it wa s widely administered, in order to identify and reduce potential for problems. The reviewers were recruited from among key informants and other contacts the researcher used to recruit women for i nterviews. The reviewers were not provided an incentive to participate. They examined the instrument and made comments, when necessary, about question wording and questionnaire length. Comments were used to revise the instrument to its final form. Survey Q uestions The survey included general questions that were useful for comparing womens cha racteristics with their beliefs. The instrument also included questions that were associated with the themes and subthemes from the interviews. This subsection describes those questions and the reasons they were used in the survey. A final category of questions was summary belief questions assessed with Likert scale items. Those Likert items will be discussed first as potential scales together and then as individual que stions falling under themes. The survey included nine Likert scale belief questions that fell into the general categories of Prevention Benefits Beliefs Personal Risk Beliefs, Fear Beliefs, and Internet Efficacy Beliefs, with the latter referring to a wom ans belief about how well she could find information about breast cancer on the Internet if she wanted that information. These four categories are not actual themes from the interviews ; rather they related to the categories of an existing scale instrument that was found to have relevance to the interview results. Individual survey questions came from Champion and colleagues (2008) scales for measuring mammography and breast cancer beliefs in African American w omen based on the Health Belief Model After the theory emerged from the qualitative portion of the study and revealed similarities to components in the Health Belief
136 Model, the researcher decided to include relevant scale items in the present survey. In this way, the interview themes themselves are not reflected by these categories ; rather the overall theory emerging from the interviews prompted the items inclusion in the survey. Champion and colleagues scales included questions categorized under susceptibility to breast cancer, benefits of mammogr aphy, barriers to mammograms, self efficacy to get mammograms, and fear of cancer. This paper presents slightly different category terms to reflect the language used in the current study: perceived risk beliefs correspond to the scales susceptibility and was changed because interviewed women used the word risk; prevention beliefs replaced the scales benefits category to reflect that prevention in this study was not limited to mammography and because the interviews revealed few primary prevention behavio rs for the women to assess benefit for; and fear be liefs corresponded to the scales barriers category. The current projects emphasis on primary prevention limited the utility of many of these questions about mammography. Due to relevancy issues and the general exploratory as opposed to explanatory focus of this project as well as the inability to cover the scales more than 40 items about mammography beliefs in addition to questions about health and breast cancer information seeking in a single surve y design only a few of the scale items for susceptibility, barriers, fears, and benefits were included. The items about self efficacy to get a mammogram were rephrased with terms reflecting womens perceived ability to find breast cancer information on th e Internet, which was more appropriate to this study. First, for the perceived risk beliefs, the following two mammography scale items were modified: How likely is it that I will get b reast cancer during my lifetime ? and Compared to other women my ag e, would you say your chances of getting breast cancer are higher or lower? These questions
137 became the following Likert scale agreement questions, with strongly agree, agree, disagree, strongly disagree, and no opinion as answer options : I will get brea st cancer during my lifetime and Compared to other women my age, my chances of getting breast cancer are higher .1The prevention beliefs items from Champion and colleagues (2008) scales used in the study were the following: If breast cancer was found early, how like ly is it that the cancer could be successfully treated ? and Having a mamm ogram will decrease my chances of dying from breast cancer These items were reworded into statements with which the women would agree or disagree. In order to include a n items related to primary prevention, one item not in the scales was included. It said, I can do something to keep from getting breast cancer. The existing scales included a fear category, but those items were not explored here. Possibly creat ing confusion, the present study used two barrier category items that related to fear and recategorized these two items as fear beliefs. One survey item was derived from the scale item, How likely is it that The treatment I would get for breast cancer wo uld be worse than the cancer itself? Another survey item rephrased the scale item, How li kely is it that Being afraid of finding a breast lump would keep me from having a mammogram ? into a prevention information seeking behavior, worded in the follow ing way: Being afraid of breast cancer would keep me from looking for information about preventing breast cancer. The scale items for s elf efficacy involved perceived ability to perform tasks related to secondary prevention. Due to the studys focus on prevention information seeking behaviors instead of prevention screening behaviors, no scale items were relevant. Nevertheless, the language of one scale item formed the foundation of one survey item. That scale item was I can 1 The no opinion answers were not used in ordinal measurements in this survey because they did not represent a measurable degree of agreement.
138 find a place to get a mammogr am (or know where to go to get a mammogram) and the resulting survey item was I believe I could find information about breast cancer prevention on the Internet easily. No other similar items were found among the scale items. In effort to find another i tem related to using the Internet for breast cancer prevention information, the researcher created the following question: If I wanted to find out what I could do to prevent breast cancer, I would use the Internet for that. These scale items use in this study was not to validate the items. In fact, the items formats and measurements differ in the original scale and the survey. The purpose was to take phrasing from items in the literature that reflected scholars testing, consideration for theory, and ta iloring to African American women. Given the exploratory nature of this survey, the use of single questions in this manner was appropriate. The researcher first attempt ed to use the nine items to create four scales within the categories of prevention belie fs, personal r isk beliefs, fear beliefs, and I nternet efficacy beliefs however each scale included two to three items each, resulting in a small number of items. Reliability analyses revealed poor internal consistency for all of these scales, with Cronbac h alpha coefficients of .55 for Internet efficacy, .5 5 for risk, .49 for prevention, and .27 for fe ar. The mean inter item correlation values were also too low for all four scales to be considered useful for measuring the belief concepts together. Therefor e, the questions intended for use in constructing scales were used separately, for exploratory purposes, to compare with other variables including survey respondents Internet use and breast cancer family history General Q uestions Demographic questions as well as some questions that provided helpful background information f e ll into the general questions category. The former should be clear from the description below. However, the latter the post analysis questions require more explanation.
139 Health infor mation seeking was not a theme or subtheme; it was an overarching behavior guiding this research. Therefore, questions related to womens health informationseeking behavior were included. Similarly, women in the interviews and in the literature cited in C hapter 2 referred to some beliefs they had about the Internet, such as that it was easy to use or that it might not have trustworthy information. Those remarks were not full themes but they were important details for understanding womens beliefs about the Internet; thus they were included in the survey. Demographic q uestions Useful surveys about health and Internet use require a significant number of items about respondent demographics. In an effort to understand potential factors involved in Internet us e, researchers include questions about income, education level, and age, because these variables are, to some extent, influencers of Internet use. This studys survey included these demographic questions, plus marital or relationship status, and whether or not the woman had medical insurance. Age was also used to assess if the woman fit the studys age criteria being between ages 35 and 65. A question about whether or not the woman had ever had breast cancer also served this purpose of eliminating women outside of the studys parameters. Another question asked women if they had ever had any form of cancer ; however that question was not used for analyses because only three women marked that they had had cancer in the past. One of the major questions asked w as about the womens family history of breast cancer. Survey respondents were asked to check if they had a mother, grandmother, sister, aunt or cousin, or close friend with breast cancer. The close friend selections were not analyzed because they were not found in the interviews to affect womens personal perceived risks of breast cancer. Aunts and cousins also were not considered to reflect a family history of breast cancer
140 because the medical literature included no information about the relationship bet ween womens risk of breast cancer and these extended family members diagnoses with the disease. The two were included because several women in the interviews referred to experiences with friends or extended family going through breast cancer experiences. The researcher anticipated that some women might write in these individuals because the women wanted to acknowledge the impact of these loved ones ordeals on their views of breast cancer. Internet use questions Interviewed women who used the Internet sa id they used it for many purposes, including communication, entertainment (games and music), and information seeking. The women suggested that the Internet replaced many of the tasks and chores they once had to do differently, such as going to the library for research, talking to a doctor about a health condition, or opening a paper map for directions. Question 9 on the survey was based on this theme of multiple uses for the Internet. It asked women What do you do on the Internet? A list of items was a vailable for women to select from, if they used the Internet for those purposes. Items on the list included those that interviewed women cited information, maps, playing games, listening to music, helping children with homework, looking for jobs, sending and receiv ing email, and read ing news that are also included on Pew Internet polls about peoples online activities (Pew Internet, 2009). Question 6 also related to Internet use and served as a stratification variable. This item asked women to indicate how much time they spent online in a typical week. The options included no time, under three hours, between three and seven hours, and more than seven hours per week. Although other ways to approach this measurement would be to ask women to write in a dis crete time or estimate how much time they spent on the Internet in the last week, the
141 researcher decided that the limitation of atypical online activity in the preceding week would outweigh the precision of the other two approaches. Another question that i s common on online activity polls is where people use the Internet. In the interviews, the theme of convenience emerged as a reason why women used the Internet. Many of the women said they had the Internet in their home and liked being able to use it whene ver they wanted to. Question 7, Where do you use the Internet? asked women to select from a list including home, work, library, and somewhere else (such as a friends house or community center). Women could also check that they did not use the Internet. Internet n on use q uestions The interviews revealed that for women who did not use the Internet, or did not use it much, time and lack of physical access were the primary reasons for nonuse. Therefore, Question 8 asked women, If you dont use the Interne t, why dont you? Listed as possible options were both logistical and financial aspects of Internet use and beliefs about the Internet and Internet skills and self efficacy. A lack of time and not owning a computer or affording Internet service were optio ns related to physical access. A skill and self efficacy explanation was not knowing how to use the Internet and, to an extent, having somebody else look on the Internet for the respondent. The latter is also commonly asked in online informationseeking p olls such as those from the Pew Internet and American Life Project (Fox, 2006). Beliefs about the Internet Beliefs about the Internet itself were represented by three statements: I dont trust the information, t he Internet is slow, and t heres too much i nformation. Efficacy for seek ing breast cancer prevention information on the Internet The question resulted from post interview analyses during which the researcher realized that interviewed womens answers to questions about their Internet beliefs did not include references to self efficacy. It was decided that womens beliefs about the Internet as a source of good
142 information did not necessarily equate to womens beliefs that they could find good information on the Internet. This item was used to measure what women believed about their abilities to find information about breast cancer prevention on the Internet easily This measure explored impressions of the Internet beyond what women think it is (such as a reservoir of cancer information, which even wom en who do not use the medium might believe), but also the womens views of their own abilities to use the Internet to find the information they need or want. Preference for tailored or general breast cancer Web sites Consistent with the notion of surveys complementing interviews is the opportunity to use surveys to explore concepts that were not clear in the interviews. Preference for tailored or general breast cancer Web sites was not an interview theme because there was no agreement with either preferen ce. Some women said they would rather have a Web site for all women, whereas others thought they might prefer one specifically for African American women. This question sought to examine what the surveyed women preferred. Theme Q uestions Theme questions ar e those that arose directly from major themes or subthemes in the interviews. The survey questions do not match the wording of the interview questions, due to the different formats, but they we re designed to explore themes and relationships that emerged fr om conversations with the women. Health information seeking on the Internet is a common activity The study was primarily about women seeking online information about preventing breast cancer, so the survey required questions specifically about that topic. However, the infrequency of those searches by interviewed women prompted a broader look at health information. The questions conceptual
143 progression was this: 1) Are women looking for breast cancer prevention information on the Internet? If they are not l ooking for this information on the Internet, are they looking for 2) prevention inform ation about other diseases and/ or 3) information about breast cancer in general? Perhaps, as the interviewed women suggested, women do not look for information about preventi ng conditions they do not presently have. Based on this line of thought, the survey included one general question asking if the women have ever looked for information about a health problem or a disease. This question was not specific to the Interne t because some women do not use the Internet but may be health information seekers. Question 2 asked women if they ha d ever looked for information about ways to keep thems el ves from gettin g a health problem or a disease. And the next question asked women w hich sources they use to find out about health topics. This question is differently formatted from the others yes and no answer options, and includes a list of possible media, interpersonal, and health care sources from which women might have sought infor mation Moving into online informationseeking explorations, Question 4 asked women if they have ever looked up health information on the Internet and Question 5 asked if they had ever asked someone to look up this information on the Internet for them. From there, questions became more specif ic to breast cancer information seeking on the Internet, including whether or not the women had ever looked for breast cancer information on the medium and if they had ever looked up information about preventing (neve r getting) breast cancer? Prompts for breast cancer information seeking related to family members experiences or personal scares. This theme of women looking for health information in general, and breast cancer information specifically emerged from the i nterviews. The survey question relating to that theme asked women If you looked for breast cancer information on the Internet, what made
144 you want to find it? The respondents had answers that included the two theme scenarios family members developing breast cancer or the women personally having a scare, such as finding a breast lump or a suspicious mammogram as well as prompts of curiosity, research for a school project, media coverage of breast cancer, and a breast cancer awareness event prompting interest. A belief that the Internet would be a good source of info rmation about b reast cancer, if a woman wanted that info rmation None of the women interviewed indicated that the Internet would be a bad source of information about breast cancer, although s ome suggested that opportunities for misinformation persisted across the medium, due to the ease of Web site development by nonauthorities on breast cancer. Therefore, two questions originally sought to explore whether women believed they would look for breast cancer prevention information on the Internet, if they wanted that information, and an Internet skills self efficacy measure. These latter questions had Likert scale response formats, with Strongly Agree, Agree, Disagree, Strongly Disagree, and No Opinion, as answers. However, it was later determined that the two questions were measuring two different concepts: perceived utility of the Inte rnet for prevention information seeking and self efficacy to use the Internet for that purpose. Therefore, the se cond item is described separately. Simplicity of searches and the wealth of information Interviewed women like d the Internet for health information because it has lots of information, it breaks topics down into subtopics, and information is easy to find. Also, they like d generic search engines because of this convenience factor. Interviewed womens preference for generic search engines was consistent with previously cited research (Fox, 2006) This preference also has important implications for developing tailored and targeted Web sites for African American women if they use Google,
145 Yahoo and similar search engines that generate search results based on popularity formulas, and the keywords they use are simply breast cancer or similar terms, the women may not encountered such Web sites in the first few pages of search results. For this reason, the women were asked to check the Web sites or search engines they had used for health information seeking based on those supplied by interviewed women Google, Yah oo, WebMD Medline Mayo Clinic a doctors Web site or another site not listed. The interviews and to some extent, the literature discussed in Chapter 3 about Internet features provided some reasons why women who wanted to find out about preventing brea st cancer might turn to the Internet. Most of these reasons related to the reasons the interviewed women said they used the Internet for other health informationseeking. The list included that the Internet has a lot of information; t he Internet is fast; the Internet makes it ea sy to find answers to questions; t he Internet lets me read about ot her women who had breast cancer; t he Internet gives me information I can share with my doctor or ask him/her about; t he Internet lets me talk about breast cancer with othe r women through email and chats; t he Internet has information that is easier to und erstand what my doctor tells me; and that the Internet can be used from home, at any time of the day. Conversely, women were asked if they would not use the Internet f or breast cancer prevention information, what their reasons would be. The answer options were, like those related to overall Internet use, logistic, economic, self efficacy, and belief based. The list included the following: I dont have a computer with Internet access; I dont know how to use the Internet; I dont trust the information on the Internet; I cant find the answers to my questions on the Internet; the Internet has too much information; the language on web pages is confusing; another reason not listed.
146 Fear of cancer. Interviewed women frequently cited fear of cancer in general, and breast cancer specifically. Sometimes the fear was mentioned in the context of what the women thought other women might experience with cancer. This theme wa s consist ent with Champion and colleagues (2008) scales about mammography beliefs that they adapted for African American women. From Champion and colleagues (2008) mammography scales came th is studys questionnaire item 25: The treatment I would get for breast ca ncer would be worse than the cancer itself This was a Likert scale formatted statement to which women indicated their level of agreement. It was included in the survey because interviewed women suggested that, for some cancers, concern about treatment sid e effects such as breast removal and sexual dysfunction were some reasons women did not talk about cancer prevention or did not seek prevention. Another Likert style item created and added to the survey was item 26: Being afraid of breast cancer would ke ep me from looking for information about preventing breast cancer. This item was another attempt to gauge womens beliefs about actively seeking prevention. It was largely based in the literature that suggest s that some people blunters intentionally a void seeking information about health risks, while others monitors actively seek or scan available sources for risks (Case, et al., 2005). The interviews also revealed that women thought people viewed breast cancer as something to be feared. This surv ey question was used to assess whether or not fear of breast cancer might be a barrier to women trying to learn more about the disease. These two questions both involved fear beliefs that might affect a womans behavior regarding breast cancer prevention a nd prevention informationseeking. However, as Chapter 7 s limitation section discusses, the underlying concepts of fear of treatment and fears as barriers to information seeking may be fundamentally different
147 Proximity to breast cancer. The women held the b elief that risk is mostly associated with family history. The interviewed women implied that they thought they were at low risk for breast cancer if they did not have a family history of the disease. In addition, many women suggested that they did not know what else besides smoking and possibly obesity might increase a womans risk of breast cancer. Question 11 asked women to check from a list of items or to write in an alternative item they believed puts people at risk for breast cancer. Most of the ite ms came from the interviews including the myth about breast trauma one woman believed, but others were added from the literature (e.g. ACS, 2009b) to present more options and to determine whether the womens beliefs were consistent with the interviews and with some of the science related to breast cancer. The list was not exhaustive and did not include some items that would have been more difficult to explain in a short answer, such as those related to womens birthing patterns However, other items listed as things that might increase womens risks of breast cancer included having a family member who has or had breast cancer ; the foods they eat (diet) ; b eing overweight ; s moking ; having a breast hit or pulled; g rowing older ; a lcohol overuse ; s omething in th e environment and nothing i t just happens to some people. As mentioned, i nterviewed women focused on family history as being the primary risk factor for breast cancer. Many of them viewed breast cancer risk as uncontrollable or unable to be changed with prevention. Women without family histories viewed themselves as unlikely to get breast cancer. Women also interpreted prevention as a secondary function in that women can detect tumors early, as opposed to preventing their development. Because of this, se veral survey items attempted to measure womens perceptions of personal risk. Items 20, 21, and 23 asked the following: I can do something to keep from getting breast cancer I believe I will get breast cancer during my life, and C ompared to other wom en my age, my chances of getting breast
148 cancer are higher. These were Likert scale items that were compared with respondents reported family experience with breast cancer. The intention was to see if women reporting family histories of breast cancer had higher perceived risks of the disease than women without this history. The three items were compared with the family history variable. Frequency of secondary prevention behaviors. Almost all of the women interviewed had had at least one mammogram. The wome ns reasons for this screening varied, with some taking advantage of free tests even though they expressed low perception o f risk and other women having a higher perceived risk and getting the mammograms as a precaution. Two questions were included with ye s and no answer options to determine how many women in the survey had had mammograms or engaged in monthly self examinations. This measure was, in part, a demographic type assessment to see how common secondary prevention behavior was Even though this beh avior is not part of primary prevention, active prevention of breast cancer in any form may indicate general orientations toward prevention. Those two items, questions 39 and 40, asked, H ave you ever had a mammogram or another test for breast cancer? and D o you do monthly self exams for breast cancer? A belief in secondary prevention. Some of the interviewees indicated a belief that breast cancer is no longer a death sentence; that it is treatable. Champion and colleagues (2008) Likert scale items abou t mammogram beliefs included two about the efficacy of mammograms. These items were included in part to see if surveyed women believed in a benefit from any prevention at all, or if they believed breast cancer is a death sentence, as some interviewed wom en thought older generations believed. The two items were (item 22) which asked, I believe if breast cancer is found early, it can be treated successfully and (item 24), H aving a mammogram will decrease my chances of dying from breast cancer.
149 Import ance of breast cancer relative to o ther diseases and problems in the community. This theme occurred in the literature (e.g. Moy, et al, 2006) and in the interviews. Women perceived breast cancer to be less common in their communities, and therefore less o f a priority. In order to see if the women in the survey had similar perceptions, they were asked to check the three problems that were the biggest concerns for their community. The women were not asked to rank these items, only to check three of them from the list. Options included an other and opportunity to write in a problem that was not listed, in addition to some of the diseases and conditions that came from the literature and interviews. HIV/AIDS; high blood pressure; diabetes; abuse; sickle cell anemia; other cancers; and breast cancer were some of the answer options. Infrequency of interpersonal discussions about breast cancer Question 10 asked surveyed women if they had ever talked to friends and fam ily members about breast cancer. This yes or n o question was included because women in the interviews reported that they would be unlikely to talk about breast cancer unless they or someone close to them developed it. The questions purpose was to compare its responses to womens self reported breast cancer family history to see if more women with this history talked about the disease than women without the history. No reported preference for tailored breast cancer Web sites. Many of t he interviewed women expressed that they did not care if a breast c ancer Web site specifically addressed African American women. The researcher included Question 18, which asked, If you wanted information about preventing breast cancer, would you want a web site that talked about breast cancer in African American women or in women in general? in an effort to see if this theme carried over to survey respondents. She did not know if the interview responses were influenced
150 by the women being interviewed by a Caucasian woman so she posed this question under survey conditions were women could answer anonymously. The three answer options for this question were women in general, African American women, and it wouldnt matter to me. Survey Sampling The survey method used in this study was purposive, nonprobability sampling (Ba bbie, 1990), allowing the researcher to better select survey settings that include d women meeting the study parameters. The main reason why this study used purposive sampling wa s that the study participant parameters we re somewhat narrow, using age, race, and gender as conditions of the study. African American females accounted for only about 6.5% of the U S population in 2000, with an even smaller percentage being between the ages of 35 and 65 years ( McKinnon & Bennett 2005) and another slight reduction in the studys conceivable study population based on some portion of these women having survived breast cancer, thus not being appropriate for the study. Random sampling techniques such as random digit dialing would have result ed in a high e xpenditure of t ime and effort in order to yield significant numbers of participants meeting the studys criteria. Despite this low percent of their population representation, African Americans are disproportionately low income, representing nearly one quarter of American s living below the poverty line in 2000 ( McKinnon & Bennett 2005) This income variable may affect the likelihood of women meeting the st udys focus parameters having telephones or long term addresses so random selecti on for phone or mail surveys also wo uld have been problematic. Even before developing the survey instrument, the researcher expected to be interested in receiving data from women with and without breast cancer in their families and from women who do and do not use the Internet. Focusing sam pling efforts in a purposive way so that most or all survey responde nts fit the initial gender, race, age, and no breast cancer diagnosis parameters allow ed the researcher to get more usable and useful data for inferences. Kemper echoes this idea
151 by explai ning that purposive, nonprobability sampling allows researchers to select specific cases that will provide the most information for the questions under study, (2003, p. 279). Th e sampling technique was opportunistic and snowball sampling. Absence of probability in sampling design prevent s researcher s from making inferences about the survey findings distribution in the larger population, but it allow s researcher s to get more cases to explore the themes. Although these types of purposive nonrandom sampl ing reduce researcher s ability to make generalizable inferences and to make the kind of strong statements about the datas relationships as are often seen in purely quantitative pursuits, this method is not inconsistent with the qualitative paradigm of se lecting cases based on qualities that will inform the understanding of a phenomenon. Even though the survey itself is quantitative, due to its close ended questions and opportunities to count cases, its purpose and design remain exploratory. Additionally, it was necessary for this younger, white researcher to seek help from informants in the community who are African American women within the studys age ranges and who could assist with identifying cases for the study. Kemper says this type of sampling us e[s] insider knowledge to maximize the chan c e that the units included in the final sample are strong (highly appropriate) cases to include in the study (2003, p. 283). Snow balling begins with one or two participants matching the studys description and as ks them to identify more participants to be part of the research. Less defined is opportunistic sampling which Kemper describes as taking opportunities as they come along and following up on leads as they arise within fieldwork (2003, p. 283). Sample s ize in quantitative studies should be estimated based on the need for data from subgroups being studied (Fowler, 2002). In describing quantitative studies, Fowler says, Most sample size decisions do not focus on estimates for the total population; rather, they are
152 concentrated on the minimum sample sizes that can be tolerated for the smallest subgroups of importance (2002, p. 36). Estimating sample size for this study wa s more complicated. T he survey phase sought responses from women of various ages withi n the studys range who ha d different Internet use patterns and a mix of family breast cancer history statuses. As implied in the mixed methods section of this paper, the researcher had to make many decisions about the studys purpose and design. The fact that the study is sequential as opposed to concurrent; exploratory and not explanatory; qualitative inquiry followed by the quantitative method; qualitative perspective dominant as opposed to equally weighted between qualitative and quantitative; etc., gi ves it a unique design DNA. Each of these study attributes has its own underlying perspective that would be violated with haphazard decisions such as in the case of sampling approaches. As the qualitative paradigm does not seek generalizability of findi ngs the statistical formulas often used to determine sample sizes in quantitative research were not appropriate for this exploratory study with a nonpurposive sample. Without a sampling formula, the researcher consulted literature for guidance in determ ining an appropriate sample size. As previously described, e very mixed method study may become a distinctly different design from another mixed method study by choosing a different research stance. The literature includes examples of mixed method studies using interviews and surveys but because existing studies did not have the same attributes of this particular study, they offer ed little direction for sampling size. Morse (2003) points out that when emphasis is placed on one phase instead of treating the qualitative and quantitative portions as equal, the less dominant research method portion is not considered a complete and separate study. Findings from the less dominant data collection phase (the quantitative portion of this study) do not need complete v erification and saturation is not required (p. 196). The main anticipated determinant
153 of sample size for this study was the desire to explor e many themes and subgroups. Sampling decisions reflect ed this interest in seeing rich data from people with multip le backgrounds and experiences (e.g. Internet experience, family history, and perhaps sociodemographic indicators). It was decided before the survey data collection phase that approximately 150 surveys would provide that information. Surveys were administe red to 259 women in one large Southern city with an MSA population of more than one million and in one small Southern city with a population of about 10,000 people, in order to get a variety of responses. Church pastors and individuals affiliated with loca l health care services and cancer non profit organizations served as initial contacts for identifying women, and groups of women, who might complete the survey. Some of these contacts suggested group meetings and events where the researcher distributed and collected surveys on site. When no convenient meeting or venue was available, the researcher left copies of the survey for any interested woman to complete at her leisure and return to the researchers post office box in a confidential, pre stamped, addre ssed envelope. Eighty four surveys were distributed with mailing instruc tions, of which 19 were returned. The response rate was 22.6% for that method. Surveys were administered in person as well. For the onsite method, it is estimated that only two of the 175 surveys were taken but not returned. Thirty of the surveys collected through both methods were eliminated because the respondents answered that they were outside the studys age parameters or the respondents were breast cancer survivors, leaving 160 s urveys in the sample. The surveys were analyzed at the end of the collection process. Then the survey findings were presented separately from the qualitative results in Chapter 6
154 Survey Analysis To remain consistent with the qualitative paradigm and the i nterpretation of nonprobablistic sampling, statistical inferences about the strength of relationships between variables were limited However, descriptive statistics and other analyses were used to examine the relationships between variables in order to further explore the themes (Johnson & Turner, 1993). Variables explored included answers to questions about Internet use and nonuse, health in formation seeking, breast cancer informationseeking beliefs about breast cancer risk factors, respondents perce ptions about personal risk of breast cancer, efficacy of prevention, family history of breast cancer, and other demographic data, such as education and income levels. Analyses varied by variable level. All of the demographic questions, including the one ab out womens breast cancer family history were first analyzed with descriptive statistics to show frequencies. Ordinal variables, including education and income level, and nominal variables such as relationship status and breast cancer family history, were described by number and percentages of cases they represented within the sample. Age, the only continuous variable, was also assessed with measures of central tendency. Further analyses depended on the appropriateness of measures to each variable and res earch question. The three analyses and accompanying tests of significance and association used in the paper were Chi square with cross tabulated nominal and nominal or ordinal variables, independent samples t tests to compare means of two groups of nominal variables, and Spearmans rank correlation coefficient to examine ordinal variables. The purpose of the Chi square test is to examine whether apparent relationships between variables are significant or are likely due to chance based on the distribution of actual survey responses compared to expected distribution if no relationship existed (Babbie, Halley, & Zaino, 2003). The t tests (two tailed in all cases within this survey) allow researchers to examine means of responses between two
155 different groups, s uch as male and female or someone with a breast cancer family history and someone without a breast cancer family history, and see if any observed differences in means are likely significant or the result of the sampling used. The p value indicates whether a statistically significant association between measured variables exists. Spearmans statistic for correlations allows researchers to assess if two ordinal variables are associated, the direction of the relationship, the strength of the relationship, and the significance level. The rest of this analysis section will describe which tests were used to answer each research question based on the survey data. The first research question about womens beliefs about breast cancer was assessed by running frequenc ies for the risk factors women cited in survey question 11, in order to see what women believed put a person at risk for breast cancer. Frequencies of cited risk factors were also examined and presented separately for women with and without family historie s of breast cancer to see if differences existed in what these women thought put women at risk of breast cancer. Cross tabulation with the Chi square statistic was performe d to see if any statistically significant relationships existed between family histo ry of breast cancer and the factors women believed put one at risk for the disease. Another analysis used to examine factors influencing beliefs was a correlation using Spearmans rank correlation coefficient statistic ( rs) to explore how income, educati on, and age related to the f ive Likert scale questions about a womans beliefs about her perceived risk of developing breast cancer and beliefs about prevention. The prevention belief questions were : I can do something to keep from getting breast cancer I believe if breast cancer is found early, it can be treated successfully and H aving a mammogram will decrease my chances of dying from breast cancer Beliefs about perceived personal risk of cancer included the following: I
156 believe I will get breas t cancer during my life and Compared to other women my age, my chances of getting breast cancer are highe r Descriptive statistics for the answers were presented. T he Spearmans rank correlation coefficient was useful to see if agreement with the belie fs was correlated with higher levels of age, education and income. The second research question asked about how communities treated or viewed breast cancer. The surveys could not fully answer that question, as it was clearly written from a qualitative pers pective; however, the surveys offered an opportunity to see which diseases and health problems women thought were most common in their communities. Frequencies were run for the problems women checked as the top three priorities. Then, to see if women with breast cancer family histories might perceive breast cancer to be a bigger priority tha n women without this history, a cross tabulation with the chi square statistic was used with the family history variable. Further exploration of these relationships was not appropriate due to the small subsample (n=37) of women with family histories of breast cancer. Also f or Research Question 2, the results of the survey question about whether women talk ed to friends or family members about breast cancer was described w ith frequencies and then cross tabulated with family history of breast cancer to see if, as was the case in the interviews, the women who had this history were more likely to talk about the disease due to this experience with loved ones. Both of these vari ables were nominal so the Pearsons C hi square statistic was used to test for statistical significance between talking about breast cancer and having a family history of the disease. To answer Research Question 3 about breast cancer prevention information seeking, frequencies were run for the informationseeking questions (about health informationseeking, prevention informationseeking, breast cancer information seeking, and breast cancer prevention
157 information seeking) and then cross tabulated with the f amily history variable to see if differences existed according to this history using the chi square statistic Frequencies were also examined on the health information sources mentioned to see where the women were finding out about health topics and if these sources differed by breast cancer family history. Chi square values revealed where there were significant differences in sources by family history. Research Question 4 sought to discover more about how the women used the Internet. First, womens responses to time spent online were examined with frequencies. Of particular interest to the researcher were the reasons women who do not use the Internet give for their non use. The cases of women who said they spent no time on the Internet in a given week wer e probed with the question about why they do not use the medium by running frequencies for the answers. Frequencies also were used to show where the women used the Internet. In an effort to see if there was a correlation between survey respondents ages and the time they spent online, Spearman s rank correlation coefficient was used to examine relationships between these two variables. Similarly, correlations were examined to determine if there was a relationship between education or income and time spent o nline Frequencies were run for the online activities surveyed women said they did. A c hi s quare test indicated whether or not the age, income, and education level variables were significantly related to the most relevant activity looking up information. Research Question 5 concerned breast cancer information seeking. An earlier research question (RQ3), showed frequencies for breast cancer online informationseeking For RQ5 the prompts for these breast cancer information searches were explored by selecting only the cases of women who had looked for the information and running frequencies for the prompts. The reason for this filter is that some women marked that they had not looked for breast cancer
158 information on the Internet but did mark some reasons w hy they looked for breast cancer information. This selection filter allowed the researcher to examine only the cases where the answers were consistent for both questions. Three of the Likert style scale questions that related to Internet use for breast can cer i nformation were examined using S pearmans rank correlation coefficient to note any significant correlations between these beliefs and age, education, and income. For Research Question 6, f requencies explored the most common Web sites and search engine s women used to find health information. Then frequencies analyzed reasons why women reported they would use the Internet for breast cancer prevention information, if they wanted that information. Chi square statistics looked for significance between those reasons and womens age, education, income, and family history. Frequencies were also used to explore reasons women reported for why they would not use the Internet to find breast cancer prevention information, even if they wanted that information. The r easons to or not to use the Internet were computed into new variables so that if a woman cited any reason she would use the Internet for breast cancer information, the new variable reflected a case of women who have a reason to use the Internet, where 1 represented this state and 0 represented cases where women did not mention a reason. The same computation was used for the question asking women for reasons they would not use the Internet for this information. Then the new variables were cross tabulated wit h time spent online and family history of breast cancer Frequencies were also run for the topics of breast cancer information women would want to find on the Internet. Only frequencies were run here because the answers were not exclusive women could sel ect many of the topics from the list instead of choosing only one or ranking their
159 selections in order of importance However, the topics selected were also cross tabulated with breast cancer family history to see if that variable was statistically related to desires for different information about the disease. Research Question 7 was explored by comparing responses to the individual scale questions by the family history variable to look for differences in beliefs held by women with and without breast canc er family histories. Family history was treated as a dichotomous variable, with women who had a family history of breast cancer and those who did not T tests were used to compare the mean responses of women without this history to women with the history
160 CHAPTER 6 SURVEY RESULTS Survey Results One hundred and sixty surveys were included in the final dataset. All of the surveys represented self reported female African Americans. The breakdown of respondents reported annual household incomes, highest educ ation level completed, age, and relationship status appear in Table 6 1. The average age was 54.8 and the median age was 49. Table 61. Survey respondents age, education level, income, and relationship status Variable n % Yearly household income Less th an $20,000 Between $2050,000 Between $5075,000 More than $75,000 42 66 24 12 29.2 45.8 16.7 8.3 Highest education level completed Middle school High school College Graduate school 7 90 32 19 4.7 60.8 21.6 12.8 Age 3539 4044 4549 5054 5559 6065 Relationship status Single, never married Married Living with a partner Divorced Widowed 24 31 36 28 24 7 32 89 3 24 12 16.0 20.7 24.0 18.7 16.0 4.7 20.0 55.6 1.9 15.0 7.5
161 The highest percentage of women surveyed (45.8%, n = 66) r eported a yearly income of between $20,000 and $50,000. A majority (60.8%, n = 90) of surveyed women had high school degrees but no further education. More than half (55.6%, n = 89) said they were currently married. RQ 1: How do the women describe their beliefs about breast cancer and breast cancer prevention? The survey results reflect ed the belief that family history puts one at risk for breast cancer. Approximately 71% ( n = 113) of the surveyed women indicated they thought family history is a risk factor Family history was the most commonly marked risk factor, followed by smoking ( 50.6%, n = 81), diet (39.4%, n = 63), and a tie between being overweight and nothing, it just happens to some people (28.8%, n = 46). Growing older, a major factor in breas t cancer risk, was marked by only 28 women (17.5%) Table 62 shows the risk factors in order of their frequency Table 62. Frequency of perceived risk factors for breast cancer in total sample and by family history Risk Factor Total Sample Family History No Family History n % n % n % Family history 113 70.6 33 89.2 80 65.0 Smoking 81 50.6 16 43.2 65 52.8 Diet 63 39.4 17 45.9 46 37.4 Overweight 46 28.8 12 32.4 34 27.6 Nothing 46 28.8 14 37.8 32 26.0 Environ ment 32 20.0 6 1 6.2 26 21.1 Growing o lder 28 17.5 7 18.9 21 17.1 Breast h it /pulled 17 10.6 5 13.5 12 9.8 Alcohol 15 9.4 0 0.0 15 12.2 The three most common risk factors indicated are the same for women with and without family histories. Howe ver, the order and percentages of two risk factors were different, with women who have breast cancer family histories ranking smoking above overweight, and women without a family history marking overweight above smoking as a risk factor. Also important is
162 the higher percentage ( 89.2%) of women with relatives who have had breast cancer indicating that they thought family history was a risk factor, compared to women without this personal family history (6 5%). Despite these different percentages for women wit h and without family histories of breast cancer only two of the differences were statistically significant. F amily history of breast cancer status had a statistically significant relationship with the belief s that family history is a risk factor for the d isease ( 2 = 7.995, p < .005) and that alcohol use is a risk factor ( 2 Five Likert scale questions assessed women s beliefs about their perceived risks of developing breast cancer and beliefs about prevention. Table 63 shows the means, st andard deviations and other measures for these five questions. = 4.979, p < .03). Table 63. Descriptive statistics for beliefs about perception of personal risk and beliefs about breast cancer prevention Survey item n Mean SE SD Variance Compared to other women my age, my chances of getting breast cancer are higher 116 2.43 .08 .86 .73 I can do s omething to keep from getting breast cancer 127 2.09 .0 8 .89 .80 If breast cancer is found early, it can be treated successfully 154 1.49 .05 .62 .38 I believe I w ill get br east cancer in my life. 96 3.17 .08 .79 .63 Having a mammogram will decrease my chances of dying from breast cancer 140 2.06 .07 .82 .67 The treatment I would get for breast cancer would be worse than the cancer itself 111 2.96 .07 .78 .61 The rese archer wanted to see if any correlations existed between these beliefs and the variables of age, income and education level (see Table 6 4) Only one item revealed statistically significant results. There was a small correlation between income and belief that one is at higher risk of breast cancer ( rs = .205; p < .05). Even though this statistic suggests a positive relationship, the sign is an artifact of the measurement; in all of the beliefs scales, strongly agree was coded as 1 and strongly disagree was coded as 5. Therefore, higher income is correlated with lower beliefs that one is at risk for breast cancer.
163 Table 64. Correlations between age, education, and income, and beliefs about personal risk of breast cancer and beliefs about prevention ( rs Survey item ) Age Education Income Higher chances .1 1 .0 7 .2 1 Do something .0 2 .12 .0 6 Early detection .08 .05 .05 I will get breast cancer .04 .12 .17 Mammogram .0 7 .1 0 .11 *p < .05 R Q 2: What individual, int erpersonal and cultural factors influence African American womens beliefs about breast cancer prevention? What do they think breast cancer means to their communities? How do members of their social sphere and community talk about breast cancer? Surveys re vealed that other diseases were more predominant or bigger priorities in participants communities. Women were asked to check the three problems that we re the bigg est concerns for their community from a short list of common health issues that came from the literature and interviews, as well as the option to check other if the list was not inclusive enough. The three most common health problems checked on the surveys were high blood pressure (n = 136, 85%); diabetes (n = 118, 73%), and HIV/AIDS (n = 79, 49%). Table 65 has a complete list of cited problems By comparison, breast cancer was only listed as a topthree disease priority by 29 women (18%). Even among women who had a family history, breast cancer was only marked by 22% (n = 8). Women with a fami ly history of breast cancer were not significantly more likely to select breast cancer as a priority problem in their communities ( 2 = .397, p > .528).
164 Table 65. Problems facing respondents communities, by number and percent of women who cited the problems as top three priorities Problem n % H igh blood pressure 136 28.1 Diabetes 118 24.4 HIV /AIDS 79 16.3 Heart d is ease 55 11.4 Other c ancer s 34 7.0 0 B reast cancer 29 6.0 0 Other p roblem s 15 3.1 0 Abuse 10 2.1 0 Sickle cell anemia 8 1.7 0 Among the survey responses, 78% ( n = 126) of surveyed women indicated that they had talked to friends or family members about brea st cancer. Among women with a family history of breast cancer, 95% ( n = 35) said they had talked about the disease; 74% ( n = 90) without this history said that they had talked about breast cancer. Chi square tests revealed that this difference may be signi ficant ( 2RQ 3: What emphasis do women place on personal, proactive breast cancer prevention information seeking ? The survey did not ask women about breast cancer information seeking from any source; rather it focused on the Internet for that information. Less than one third ( 30.3%, n = 37) of survey participants who did not have a family history of breast cancer indicated that they had actively sought breast cancer information on the Internet whereas slightly more ( 38.9%, n = 14) of the women who had a mother, sister, or grandmother with the disease had search ed (Table 6 6). Chi square tests revealed that the relationship between family history of breast cancer and breast cancer informationseeking on the Internet was not statistically significant. = 7.221, p < .01) This questionnaire item was a simple yes or no question, allowing the researcher to infer no more details or insights about the way women talk about breast cancer, what prompts the conversation, how regularly it is discussed, what is said in conversations about the disease, and how the survey results compare to the interviews
165 Table 6 6. Use of the Internet for breast cancer information Breast cancer information seeking status Women with no family history Women with a family history Total Has never sought breast cancer information on the Internet 85 (69.7%) 22 (61.1%) 107 (67.7%) Has sought breast cancer information on the Internet 37 (30.3%) 14 (38.9%) 51 (32.3%) Total 122 (100%) 36 (100%) 158 (100%) Three most commonly mentioned sources from which women find out about health topics are their doctors (75%, n = 120), television (61%, n = 97), and from friends and family members (60%, n = 96; see Table 6 7). While women may actively seek out health information from these sources, the most common sources accommodate passive information attainment, as opposed to actively seeking information in books or on the Internet Women may actively seek health information from their friends and doctors but they also may receive unsolicited advi ce about health when they go to medical offices for annual check ups or when they engage in conversation with family members and friends. Table 67. Womens sources of health information Source N % Doctor 120 75.0 Television 97 60.6 Friends & f am ily 96 60.0 Books 88 55.0 Internet 80 50.0 Work 53 33.1 Health f airs 48 30.0 Hospital 43 26.9 Radio 40 25.0 Church 31 19.4 Sources of information about health topics did not vary much according to womens family history of breast cancer (see Table 6 8) None of the relationships between specific sources of information and family history were statistically significant.
166 Table 68. Number and percent of women who said they found out about health topics from each source of info rmation, by family history of breast cancer Sources of health information Women with no family history Women with a family history Doctor or hospital 91 (75.2%) 27 (73.0%) Television 74 (60.2%) 23 (62.2%) Radio 28 (22.8%) 12 (32.4%) Friends 71 (58. 7%) 24 (64.9%) Health fairs 35 (28.9%) 12 (32.4%) Church 22 (18.2%) 7 (18.9%) Internet 59 (48.8%) 21 (56.8%) Work 39 (32.2%) 12 (32.4%) Books 68 (56.2%) 19 (51.4%) Despite this low breast cancer information seeking, 87.5% of surveyed women s aid they had searched for health information in general ( n = 140), and 83.1% ( n = 133) had looked for prevention information from some source about some disease or condition. However, only 34.4% ( n = 55) of the women said they had ever sought out informati on about preventing breast cancer. Only 50% of the survey respondents ( n = 80) said they find out about health topics on the Internet, but 65% ( n = 104) said they had looked up health information on the Internet, a discrepancy that may be explained by semantics: the question asking women if they find out about health topics implies passive information acquisition whereas the wording in the question asking women if they looked up health information on the Internet was decidedly more active. Regardless, at 32 % (n = 51), a lower percentage of women had looked up information about breast cancer on the Internet While a higher percent of women with breast cancer in their families (39%, n = 14) looked for breast cancer information online than women without this hi story (31%, n = 37), an earlier finding showed that the difference in rates was not statistically significant RQ 4: What do they think about the Internet? How do they describe the Internets role or importance in their social setting or community? The sur vey respondents were asked
167 how much time they spent on the Internet each week and the most common answer was less than three hours (37.5%, n = 60), followed by no time at 34% ( n = 55). Forty five women (28%) said they spent more than three hours each we ek on the Internet (Table 69) The percent age of women who answered that, during an average week, they spend no time on the Internet (34 % n = 55) is higher than the percent age of women who answered another question saying they do not use the Internet at all (23% n = 36) ; however the discrepancy could be explained by the two questions wording. Given the utility to this study that comes from comparing other variables according to the multiple categories of time spent online, the reported weekly Internet use statistic will be used throughout the results section. Table 69. Amount of time women reported to spend on the Internet in an average week Time N % No t ime 55 34.4 Less than 3 hours 60 37.5 Between 3 and 7 hours 32 20.0 More than 7 hours 13 8.1 Total 160 100.0 Age, education and income were all significantly correlated with time spent online with m oderate correlations between each Age was negatively correlated ( rs = 31, p < .01), suggesting that as age increased in the sample, time online decreased. Education ( rsp < .01) and income ( r = .44, sAmong the 55 women who said they spend no time on the Internet in a typical week, 20 women did not mark a reason for this nonuse from the survey list. N one said that their reasons for not using the Internet stemmed from a lack of trust of the medium. The two most common reasons given for nonuse were not having physical access to the Internet (42%, n = 23) and not knowing how to use the Internet (20%, n = 11). The only other explanation marked by more than = .38, p < .01) were both positively correlated with time spent online, indicating that higher degrees of education and income were related to more Internet use.
168 two people was I have somebody else [use the Internet] fo r me, with about 15% ( n = 8) of the women indicating that situation. It is important to note that these categories were not mutually exclusive on the survey instrument, meaning that women may have selected more than one reason to describe why they do not use the Internet (Table 610 ) Table 610. Reasons given for not using the Internet by women who reported spending no time using it in a typical week Reason n % I dont have access 23 41.8 I dont know how to use it 11 20.0 Someone else does it for me 8 14.5 I dont have time 1 1.8 I dont trust the Internet 0 0.0 The Internet is too slow 0 0.0 The Internet has too much information. 0 0.0 Most of the women who said they did spend time on the Internet in a typical week indicated that they used it at home (88%, n = 92), and 48% ( n = 50) said they used the Internet at work. Less popular places to use the Internet were at the library (16%, n = 17) and some other place, such as a friends house or a community center (9%, n = 9). These categories were not presented as mutually exclusive either; some women used the Internet in more than one of the places. The most common activities on the Web were looking for information and using email (see Table 611) Looking up information, the activity relevant to this study, was related to age ( 2 = 10.77), education ( 2 = 26.44, p < .001), and income level ( 2 = 14.72, p < .01), although the relationships were only statistically significant for the latter two.
169 Table 611. Number and percent of women who use the Internet for each activity Online Activity n % Look up information 89 84.8 Email 80 76.2 Read news 68 64.8 Find maps and directions 60 57.1 Help children with work 46 43.8 Buy things 46 43.8 Pl ay games 32 30.5 Listen to music 26 24.8 Look for jobs 24 22.9 Time spent online and survey respondents age as a recoded categorical variable with approximately five years representing each category, were negatively correlated ( rsn = 151, p < 0.001) meaning that older age was correlated with less time online ; and respondents education level was correlated with time spent online ( r = 0.31, s = 0.4 4, n = 148, p < 0.001) indicating a medium positive correlation between the variables so that more educated women tended to spend more time online Also, income was positively correlated with time spent online ( rsRQ 5: What do African American women think about the Internet as a source of breast cancer information? For women who use the Internet for breast cancer information, how do they describe their experience? Of the 51 surveyed women who said they had sought online information about breast cancer, the three most common prompts to seek that information were having a family member diagnosed with the disease (75%, n = 38), having a personal breast cancer scare (39%, n = 20), and seeing something in a pamphlet that made them want to find more information (39%, n = 20). The questions instructions encouraged women to check all items that applied to them, accounting for the percentages adding up to more than 100%. = 0.38, n = 144, p < 0.001) with higher incomes correlated with more time using the Internet None of these correlations were v ery strong. Three of the remaining Likert items explo red womens beliefs about their potential use of the Internet for breast cancer information. Table 6 12 shows the descriptive for these questions.
170 Table 612. Descriptive statistics for belief items about using the Internet for breast cancer information Su rvey item n Mean Std. Dev Being afraid of breast cancer would keep me from looking for information about preventing breast cancer 139 3.11 .81 If I wanted to find out what I could do to prevent breast cancer, I would use the Internet to find that. 135 2.03 .65 I believe I could find information about breast cancer prevention on the Internet easily 139 1.86 .65 T ests revealed statistically significant correlations between education level and two of these Internet beliefs (Table 613) Both demogr aphic variables were negatively and weakly correlated ( rsTable 613. Correlations of age, education, and income with Internet beliefs ( r = .19 for both, p < .05). Again, the coding scheme for beliefs assigned higher numbers to disagreement, so the negative direction suggest that higher income and education levels we re correlated w ith less disagreement or more agreement that women would use the Internet to find prevention information, if they wanted that information, and more agreement that they would find the information they wanted easily. This implies that higher education is cor related with higher Internet self efficacy. s Survey item ) Age Education Income Being afraid .07 .0 3 .04 Would u se the Internet .15 .19* .0 8 Could find informati on easily .04 .19 .1 1 p < .05 RQ 6: What features or characteristics of the Internet enhance or deter womens experiences seeking breast cancer information? This question was not fully answered in either the interview or the survey, largely be cause the interviews, as sources of survey question development, indicated that breast cancer information seeking might be an uncommon activity by women who have never had the disease. Therefore, survey questions helped describe how women search for health information on the Internet and reasons why they might use the Internet for breast cancer prevention information.
171 Among the 104 women who said they had looked on the Internet for health information, the most popular sites and search engines on the survey instrument were Google (70%, n = 73), Yahoo (67%, n = 70), and WebMD (42%, n = 44). Women were asked questions about the reasons they would or would not use the Internet to find breast cancer prevention information. Although the researchers presumption wa s that women would either choose reasons why they would use the Internet or reasons why they would not use it, some women checked reasons from both categories. Eighty six percent ( n = 137) of all women surveyed indicated one or more reasons why they would use the Internet, and 44% ( n = 71) marked at least one reason why they would not use the Internet to find breast cancer prevention information. Instead of attempting to interpret reasons for this overlap in this results section, these results represent th e number of times each reason was mentioned Women who indicated that they would use the Internet to find information about preventing breast cancer were asked to select all applicable choices from eight reasons or write their own reason why the medium would be an attractive source of information for this topic. At the top of the list of the reasons why they would use the Internet were that the medium has a lot of information ( n = 51), the Internet makes it easy to find answers ( n = 49), and the Internet le ts them read about other women who have had breast cancer ( n = 40); however, the next three most common reasons (as Table 614 shows) were indicated by 3538 women, too.
172 Table 614. Percentages of women reporting specific r easons they would use the Intern et to find breast cancer prevention information Reason n % The Internet has a lot of information 51 37.2 The Internet makes it easy to find answers to questions 49 35.8 The Internet lets me read women's stories 40 29.2 The Internet gives me informat ion I can share with my doctor 38 27.7 The Internet can be used from home, at any time of the day 38 27.7 The Internet has information that is easier to understand 35 25.5 The Internet is fast 27 19.7 The Internet lets me talk about breast cancer w ith other women 20 14.6 Other 1 0.7 Chi square tests revealed very few statistically significant relationships between family history status and reasons why women would use the Internet for prevention information. One significant relationship was b etween family history and reading about stories of women who experienced breast cancer ( 2Six statistically significant relationships were found between the income, age, and education variables and the characteristics about the Internet that would make the women use that medium for prevention informationseeking. These significant relationships are shown in Table 615. = 8.138, p < .005). More than 62% (n = 23) of women with family histories of breast cancer reported that they would be interested in this information, compared with only 36% (n = 44) of women without the family history. Tabl e 615. Significant relationships between reasons for potential Internet use for prevention information and demographic variables 2 Reason ) Age Education Income The Internet has a lot of information 23.25 **** The Internet gives me information I can share with my doctor 18.88**** The Internet can be used from home, at any time of the day 8.48* 10.63 The Internet h as information that is easier to understand 12.77 *** The Internet is fast 11.28** *p<.05, **p<.01, *** p<.005, ****p<.001
173 The surveyed women were asked to check which of 12 questions about breast cancer they would want to have answered by a Web sit e about the disease. Anecdotally, the researcher observed that many women checked nearly every listed topic; in fact, the item chosen least by the women information about hair falling out was still marked by 40% of respondents (n = 64). Table 616 show s the order of items by percent of women who said they wanted questions answered about each topic. Although differences appear in the frequencies, none of the differences were statistically significant for women with and without family histories of breast cancer. Table 616. Number and percent of women who indicated they would want to learn about each breast cancer topic, by women with and without a family history of the disease Breast cancer q uestion or topic No family history n (%) Family history n (%) What causes it? 103 (85.1%) 32 (86.5%) Warning signs of breast cancer 96 (79.3%) 30 (81.1%) Does it affect African American women differently? 85 (70.2%) 25 (67.6%) Treatment options 79 (65.3%) 28 (75.7%) Treatment side effects 75 (62.0%) 22 (59.5%) Diet/nutrition to reduce risks 73 (60.3%) 24 (64.9%) Personal stories from people with cancer 67 (55.4%) 19 (51.4%) Survival 63 (52.1%) 18 (48.6%) Mammography process 61 (50.4%) 18 (48.6%) How to find support groups 61 (50.4 %) 15 (40.5%) How to get a mammogram if you cant afford one 53 (43.8%) 20 (54.1%) Hair falling out 52 (43.0%) 11 (29.7%) The two most common reasons women marked for why they would not use the Internet for breast cancer prevention information w ere because they did not have physical access to the Internet ( n = 29) and they do not know how to use the Internet ( n = 18). Table 617 lists all of the reasons given in the order of popularity.
174 Table 617. Reasons why women would not use the Internet to find breast cancer prevention information Reason n % I dont have a computer with Internet access 29 40.8 I dont know how to use the Internet 18 25.4 I cant find the answers to my questions on the Internet 7 9.9 The Internet has too much info rmation 3 4.2 I dont trust the information on the Internet 2 2.8 The language on web pages is confusing 2 2.8 Another reason not listed 0 0.0 When frequencies were run only on the 71 women who mentioned a reason why they would not use the Internet to find breast cancer information, 61% ( n = 43) reported spending no time online each week. An additional 30% ( n = 21) said they spend fewer than three hours online weekly suggesting that these women who would not use the Internet for finding breast cancer prevention information do not use the Internet much for any purpose RQ 7: Among members of these two groups African American women with breast cancer family histories and those without them are their reasons for using or not using t he Internet more related to their beliefs about breast cancer or to their beliefs about the Internet? T he question about influence remains difficult to answer because of its inherent formulaic phrasing. For one thing, the question implies that behavior would be influenced by one set of beliefs or the other, when a combination is likely involved. To what extent knowledge, socioeconomic factors, health status, social norms, faith, and other variables interplay to mold a womans beliefs about the disease s prevention is unknown and undiscoverable with the present survey sample size Identifying influencing factors was difficult because most interviewed women answered not in terms of preventing the incidence of breast cancer but in reference to early detection efforts, in hopes of saving the patients life. In fact, the emphasis on inherited risk factors across the interviews suggested that some of the demographic differences in the women, such as age,
175 marital status, and education level, likely played very little to no role in how they perceived breast cancer prevention. The researcher hoped the survey portion of the study would provide more clues about differences in breast cancer and prevention beliefs among women in various demographic categories. However, t he results failed to shed light on potential measurable influences because the scales were unreliable, likely due to the low number of items in each proposed scale and possibly some face validity issues. Instead of comparing breast cancer family histories to the scales, responses to each Likert scale question were compared with family history. Family history was treated as a dichotomous variable and t tests were performed. There was no significant relationship between these beliefs and family history of bre ast cancer except for the belief statement that one is at higher risk for developing breast cancer than other women. For this belief the mean for women without a family history of the disease was 2.54 ( SD = .832, n = 87) and 2.10 for women with fam ily his tories ( SD = .859, n = 29) ; t (114) = 2.43, p < 0.018) indicating that women who had a family history of breast cancer perceived their own risk as higher. This result should be viewed with some caution because the total response ra te for that question was only 73 %. Integration of Interview and Survey Results The interviews and surveys examined the studys research questions from very different perspectives. The interviews included open ended explorations of womens beliefs whereas surveys presented part icipants with limited number of options from which to choose. Interpretation of answers to research questions reflects this difference. For example, the interviews allowed the researcher to ask women the equivalent of what do you believe about breast canc er? The surveys basically asked participations of these listed items, what do you believe about breast cancer? As such interview and survey results are not direct comparisons; however, together they provided a more comprehensive look at the research questions.
176 RQ1: How do the women describe their beliefs about breast cancer and breast cancer prevention? Study participants in the interview and survey portions both emphasized family history as the predominant risk factor for breast cancer. Family history was mentioned by all of the interviewed women and cited as a risk factor by more than 70% of surveyed women. The nature of the interviews did not allow the researcher to distinguish certain factors that might have influenced the womens beliefs about breas t cancer. However, in general, the women interpreted prevention to be secondary level prevention methods. The belief about family history being the biggest risk factor may have led to this interpretation that breast cancer cannot be prevented, only detecte d early and treated. Two survey questions answers point to more ambiguity than clarity on this issue about prevention levels. One item, I can do something to keep from getting breast cancer, implied a primary prevention level. The mean for that answer w as 2.09, representing a point between agreement and disagreement with that statement. Although response rates were higher, a mean of 1.49 for the statement If breast cancer is found early, it can be treated successfully, points to a stronger belief in se condary prevention. Most of these beliefs were not significantly correlated with age, income, and education for the survey participants. RQ2: What individual, interpersonal and cultural factors influence African American womens beliefs about breast cancer prevention? What do they think breast cancer means to their communities? How do members of their social sphere and community talk about breast cancer? The interviews revealed that the belief that breast cancer was related to family history was rather com mon across participants, including those with and without family histories of the disease. Surveys echoed this finding that women perceived family history to be the biggest risk factor for breast cancer. On an individual demographic level,
177 income was the o nly variable that was significantly related to beliefs about ones chan c es of getting breast cancer. Lower income women tended to believe more strongly that they were at risk for the disease. Both the interviews and surveys revealed that other diseases and problems were more prevalent or bigger priorities in the womens communities. Specifically, high blood pressure and diabetes were often mentioned in the interviews and cited on the surveys. The prevalence of diseases seemed to influence womens perception of those health conditions as community concerns. The interviewed women without breast cancer history suggested that they would be unlikely to talk about the disease unless it affected them or a loved one. A higher percentage of women with family historie s in the survey sample reported that they had had conversations about the disease with other women. RQ3: What emphasis do women place on personal, proactive breast cancer prevention information seeking ? Health information seeking on the Internet was common among interviewed and surveyed women. However, breast cancer informationseeking, and particularly prevention informationseeking was less common. Few women interviewed said they had actively sought this information online but over one third of surveyed w omen had done so RQ4: What do they think about the Internet? How do they describe the Internets role or importance in their social setting or community? Internet use was commonly reported in the interviews and nearly twothirds of surveyed women said they used the Internet during an average week. Surveys revealed that age, education and income levels were all correlated with the amount of time spent Internet. Interviewed women described a range of activities and
178 purposes for which they used the Internet, including communication, looking for information, and entertainment. Survey results showed that a variety of activities were well represented. Over 40% of respondents cited using the Internet for informationseeking, emailing, news reading, directions, as sisting children with school work, and performing job searches. RQ5: What do African American women think about the Internet as a source of breast cancer information? For women who use the Internet for breast cancer information, how do they describe their experience? Although interviewed women reported low rates of breast cancer information seeking, they said they frequently use the Internet for other health searches. When asked what would prompt the women to seek breast cancer information, the two answers were a family members experience with the disease and a personal scare, such as a suspicious mammogram or a lump found in the womans breast. These two prompts were the most commonly indicated reasons why the surveyed women who said they had looked for breast cancer information online initiated that search. RQ6: What features or characteristics of the Internet enhance or deter womens experiences seeking breast cancer information? Again, few i nterviewed said they had searched for this information, so the survey was developed to include general questions about health information searches. Interviewed and surveyed women reported preferences for generic search engines, such as Yahoo and Google, for their health searches. Even though interviewed women were no t asked to list every reason why they would use the Internet for a breast cancer information search, some of the Internet characteristics they mentioned liking were the convenience of using the Internet at home and at any time, the wealth of information available, and ease at which the information could be accessed. These answers were common in the survey responses, too. Reasons why interviewed and surveyed women said they would not use the
179 Internet for breast cancer information seeking, even if they wanted information about breast cancer, cited physical access limitations. Surveyed women also reported that not knowing how to use the Internet would prevent them from using it for this purpose. Although interviewed women were not asked to estimate or quantify their perceived risk of breast cancer, compared to other women their age, the women who did not have a family history shared that they did not consider themselves especially at risk for the disease. RQ7: Among members of these two groups African American women with breast cancer family histories and those without them are their reasons for using or not using the Internet more related to their beliefs about breast cancer or to their beliefs about the Internet? Beliefs about the Internet did not seem to influence womens searches for breast cancer prevention information. Prevention informationseeking did not seem to be related to trust issues about the Internet. Women interviewed did not search for this information because they did not think they needed it. Either the women believed they had all the information they needed about the disease or, not contradictory, they did not need to know about a disease that was unlikely to involve them. This view seemed to be associated with womens perception of breas t cancer as uncontrollable and most likely to affect women with family histories of the disease.
180 CHAPTER 7 DISCUSSION This chapter will put the studys findings into context. Included in this portion of the paper are several subsections: some main points and findings revealed by the study, important themes and answers to research questions, the model of variables that emerged from the study comparison of results to those of previous research, study limitations, and suggestions for future research. Some l eeching of future research into important themes was unavoidable due to relevance to those themes. Several findings that were extraneous to the research questions but relevant to the studys overall purpose of finding ways to reduce breast cancer incidence and mortality will also be discussed as part of the general themes The study examined two important yet constantly evolving subjects. Breast cancer, which many interviewed women said was once considered a death sentence, has been the focus of research an d awareness over the years and has become much more survivable. The Internet is an innovation that is ever increasing its functions and range of users. When studied together, breast cancer and the Internet represent fertile ground for study but also presen t difficulties keeping pace with advancements in the diseases science and the technologys progression. However, by focusing on beliefs about the disease and about the Internet as opposed to measures of accurate knowledge about the subjects, this study p ut the point of emphasis on the body of literature and health behavior theories that recognize that peoples beliefs about health and taking health actions are important components of people actually adopting (or not adopting) a behavior. Four main conclusions were reached. They are summarized here and then explained more throughout the rest of this chapter: family history was often cited as the main risk factor for breast cancer; women thought about breast cancer prevention in terms of secondary prevention and screenings; online health informationseeking was a common behavior among the women
181 but breast cancer information seeking and prevention informationseeking behaviors were less common; and the study did not fully answer the question about the role of beliefs about breast cancer prevention and beliefs about using the Internet for breast cancer prevention information but it identified methods that might be used to answer those questions in the future. First, family history of breast cancer was the most commonly cited risk factor for breast cancer among survey takers and interviewed women. Due to the open ended question format, more about the relationship between beliefs about breast cancer risk factors and the womens perceptions of personal risk could be inferred from interviews than from surveys. Interviewed women, for the most part, seemed to believe they were at higher risk for breast cancer if they had a family history of breast cancer. Proximity to breast cancer experiences was an important concept in the interviews. The difference in how the two groups expressed beliefs about breast cancer might be explained with a weather analogy: residents of Kansas know that hurricanes can be scary and devastating, but they do not feel the susceptibility or have first hand experiences with the storms impact that coastal residents have. Every woman in the interview believed that breast cancer was a fear inducing disease, but the degree to which each woman thought she had reason to fear getting breast cancer varied, with higher risks expressed by the women who had family histories of the disease. Consistent with the weather analogy was the idea of breast cancer beliefs being influenced by proximity to the disease, either by having had a personal scare involving a benign lump or suspicious mammogram, or experiencing the disease vicariously through close female relatives. Likely influencing the womens personal beliefs about breast cancer were the social environments where they engage and the generations to which the y belong. This component also had a proximity element; instead of involving proximity to the disease in ones immediate
182 family, this latter type involved proximity to changes and trends in the way breast cancer is treated in society and scientific advancem ents in diagnoses and treatments. Specifically, descriptions of breast cancer spanned a spectrum from thinking death was the primary outcome at one end, to strong beliefs in early detection and survivability on the other end, with the middle of the spectru m representing the pivot point where women said they had associated breast cancer with death but now recognize awareness efforts and new medical techniques that can extend womens lives after diagnosis. Although previously discussed research does point to a connection between family history and breast cancer risk, the ACSs data (2009b) show that women with family histories represent a relatively small percent of total breast cancer cases These statistics represent women in general, not African American wo men specifically. More research is needed to parse the influence of family history of breast cancer on personal risk of diagnosis in women of different ethnic groups. However, one study (Simon, et al., 2006) suggested that white women with family histories of breast cancer may have a higher risk of developing the disease themselves than do African American women with this family history. Referring to the literature, the authors suggested that factors such as birthing patterns where African American women on average give birth at younger ages than Caucasian women, may play a role in the smaller risks for African American women with this history. Second, discussion of prevention focused on secondary screenings for breast cancer. The NCIs Web site still desc ribes prevention in terms of avoiding getting breast cancer ( NCI, 2009b) Screening is the word used for mammograms, X rays, MRIs, clinical breast exams, and other forms of detecting breast tumors. Both the NCI and the ACS frame primary prevention of breas t cancer in terms of avoiding known risk factors. Those factors include tobacco use, obesity,
183 hormone replacement therapy (HRT), late childbearing, radiation exposure, and other potential risks suggested by research. The extent to which many of these and other factors may be avoidable varies and will be rather specific to individuals. For example, despite physical and financial obstacles, most women may have some control over their food and exercise choices, alcohol and tobacco use, and use of HRT A woma n s ability to become pregnant and the years when she begins and ceases menstrual cycles are less controllable. The lack of clear guidelines about what women can do if anything, to prevent breast cancer, as opposed to things they can avoid to prevent the di sease such as not smoking, may impact their views about prevention and efficacy of prevention. Third, use of the Internet for health informationseeking was rather common. However, the use of the Internet to look for breast cancer information especially prevention information was less common among surveyed and interviewed women. Interviewed women revealed that they are more likely to seek information about conditions that are affecting them personally or affecting a family member. The perception of low risk among women without family histories of breast cancer may reduce their interest in such information searches. From this, it is clear that the womens beliefs about their risks for breast cancer, as opposed to their beliefs about the Internet, influence their decisions to search or not to search for prevention information. Fourth, the studys methods were insufficient for examining the kinds of content access issues presented in the papers earlier chapters. Although not a listed research question, the concept of womens content access to the Internet and online information about breast cancer prompted and even undergirded much of the research. In other words, the study sought to look beyond the dichotomous states of physical access to the Internet w here the women either have or do not have access to computers with the Internet that seem to be common measures of
184 Digital Divide issues, to explore the actual experiences of women who do and do not have the Internet However, examining beliefs about the Internet in the surveys and interviews did not answer the larger questions about content access the factors beyond having physical access to an Internet connected computer that influence how women understand and make use of online content, and experience. Discussion presented later in this chapter addresses the limitations of the methods to answer content access and even health literacy questions. Summary of Major Themes One them e was that a womans proximity to breast cancer influenced her beliefs about the disease, with interviewed women who had close female relatives with past breast cancer diagnoses filtering their beliefs through memories of that experience and interviewed and surveyed women perceiving their personal risks based on family history. Women without the family history sometimes mentioned friends or extended family with breast cancer, but the womens descriptions and beliefs were based on more abstract or second hand knowledge of the disease. Fear and death associated with breast cancer was sometimes cited, too. Another theme was the tendency for women to think of breast cancer prevention in terms of secondary prevention and early detection. Most of the women reported having had at least one mammogram. Related was t he theme of womens lack o f knowledge about breast cancer. Interviewed and surveyed women overwhelming cited genetic family history as the number one risk factor for breast cancer, when age is a bigger factor. The interviews revealed a theme that women did not perceive a need for breas t cancer prevention information. This was largely because the women without breast cancer histories in their families did not think they were at risk for breast cancer or that anything could cause or prevent cancer, except in some cases, smoking and b eing overweight.
185 The theme of women tending to talk about breast cancer more if they had a family member who had experienced the disease came up in the interviews but was not found in the surveys. This chapter suggests future studies should revise survey wording for this question to see if and with whom women talk about breast cancer. Both the interviews and the surveys revealed t he theme of other diseases being more common, therefore higher priorities in the womens communities. In the interviews, disease priority related to observing how common the problem wa s in the community, not necessarily on the extent of outcomes from each disease. An important theme from both study methods was that Internet non use related to lack of physical access to the Interne t rather than beliefs about the Internet being a good or bad source of information. Online searches for breast cancer and breast cancer prevention information seeking may have been somewhat uncommon, yet most women who used the Internet said the medium wou ld be a good source of that information, if the women had questions about the disease. Among those who had sought breast cancer information, prompts included having some other personal sc are or a family member with the disease. M odel for breast cancer prev ention information seeking Relationships between the studys major themes came rather clearly from the qualitative portion of the study but were more ambiguous in the surveys, a result that may be an artifact of the closedended questionnaire format that allowed no room to explore the meaning of womens answers. Given this, and the ability of the qualitative portion to explore these meanings, priority will be assigned to the interview findings. Figure 71 shows a concise theory of characteristics and beliefs that the interviews point to as antecedents to using the Internet to find breast cancer prevention information for oneself
186 These points, which are randomly ordered, are based on womens comments about why they looked for some health topics online, why they did not look for breast cancer prevention information, and what they believed about breast cancer risk and prevention. Each concept will be discussed in further detail later. Figure 7 1. Antecedents for online breast cancer prevention informations eeking Physical access to the Internet & perceived Internet skills The most obvious condition for women to use the Internet for breast cancer prevention informationseeking is having physical access to the Internet None of the women interviewed expressed any concerns about the medium or trust issues that they said would keep them from using the Internet for breast cancer information seeking, although some women expressed a preference for physicians advice. Instead, nonuse of the Internet was related to lack of familiarity with the medium or not having a computer with an Internet connection. One of the studys research questions asked if womens beliefs about the Internet or beliefs about breast cancer were related to use of the medium for breast cancer prevention
187 information seeking. I n the results chapters, the researcher somewhat dismissed the notion that not seeking prevention information was related to beliefs about the Internet as a source of breast cancer information For women who used the Intern et for health information purposes this description may be adequate. The relationship between beliefs and prevention information seeking is less clear in the women who did not use the Internet for health information. T he interviewed women who did not use the Internet attributed their nonuse to logistical issues, such as lack of familiarity and lack of physical access, instead of trust or quality concerns about online content. However, lack of familiarity and access are actually beliefs about the Internet At very least these are beliefs about the utility or efficacy of the tool and womens abilities to use it. Although time and money are integral to the decision to use the Internet the basic concept is one of cost benefit analysis in economics. Even for a time or budget strapped individual, her view of the Internet as being a useful tool for her health, education, or childrens future might incline her to make the technology investment at the expense of other items or activities. We saw this in Lindas c ase the housekeeper who admitted to a limited income but thought the Internet was worthwhile for her childrens sake. However, for simplicitys sake, this box in the model refers to physical access instead of beliefs about the utility of having physical access. The other interpretation will be further addressed in the future research portion of this discussion section. A belief that breast cancer can not be prevented This theme was implied by the womens opposite reaction to questions about breast cancer risk and prevention. The participants thought that breast cancer risk was related to family history. Several women said they did not know breast cancer could be prevented, in a primary prevention sense of the word. In fact, most of the prevention answers w ere related to secondary prevention prolonging life by detecting
188 tumors early through mammograms and other screenings. To some extent, breast cancer may not be avoidable for many women, but others may adopt healthy lifestyles and screenings to decrease t he i r chances of developing, suffering morbidity related effects, or dying from the disease. However, i f a person does not believe that a disease can be avoided, she likely will not seek information about things she can do to prevent it or reduce her risk Preventing breast cancer was a concept associated with secondary prevention, a view that might be explained by the very commonly expressed thought that breast cancer risk is associated with family histories of the disease. Presumably this association is p erceived as an inherited genetic risk. Despite ongoing research in the area of genetic engineering and modification, no studies were found to date that allow scientists to turn on or off identified genes such as BRCA1 and BRCA 2 that may lead to breast cancer. However, it is also important to reinforce the estimates that only 20 30% of women who develop breast cancer have a family history of the disease ( ACS, 2009b) As inherited genes are seen as uncontrollable at the time of this research, women with the family history may think there is nothing they can do to prevent getting the disease if that is their genetic fate. Women who do not have the disease may perceive themselves as somewhat immune from breast cancer. The other potential interpretation follows the out of sight, out of mind axiom, where women who do not have experiences with family members with breast cancer just have not thought about the disease, much like people who live in the United States may not think about their susceptibility to diseas es that affect people in other parts of the world. The corollary to this explanation would be that for women who have experienced breast cancer in their families, breast cancer is more present and impacting on their lives. A belief that she is at risk for breast cancer. This theme crosses over the one about prevention beliefs. Women who thought breast cancer is completely (or mostly) based on family
189 history and who have no such family history may think they have low risks of the disease; whereas women with a mother, sister, or grandmother who developed breast cancer described the possibility of them getting the disease as likely or higher than normal. However, having had a breast cancer scare, such as a benign lump or an abnormal mammogram, may increase a wo mans sense of proximity to breast cancer hence her perceived risk regardless of her family history. A cue to seek information or a belief that she lacks information about breast cancer prevention The first part of this theme, a cue to seek, repres ents an umbrella concept, compris ing several possible cues. The second part, a belief that she lacks information, is also important and may or may not be conditional upon a major cue. This distinction may be unclear until examples of the various circumst ances are described. First, cues to seek were mostly described in the interviews as personal scares, such as finding a benign breast lump or having a family member diagnosed with breast cancer. When a family member is diagnosed, the woman may seek general breast cancer information to help that relative with treatment decisions or support needs. She may also find her own risk schema is affected by the relatives diagnosis. If the woman has not felt at risk before, due to lack of family history of breast canc er, a close, maternal relatives experience could alter the womans perceived susceptibility. If the relative is farther genetically removed an in law or a cousin, perhaps and this diagnosed person also did not have a family history of breast cancer, t he womans association of risk with genetics may also falter as she sees that women without the history can get the disease. The latter condition in the figure s box, lacks information, describes the condition that was not found in the interviewed women with breast cancer family histories. These three women believed they already had the information they needed to know; therefore these women might not
190 become breast cancer information seekers unless they decided, at some point, they lacked some information about the disease. They experienced the disease firsthand with their mothers and other relatives and reportedly absorbed information at the time. These women may feel that they already know what they need to know about breast cancer. This condition is important because a woman could meet all of the other ones have Internet access, believe that breast cancer risk can be reduced through certain behaviors, and feel susceptible to the disease due to family history ; however, if she believes that the information she received from doctors and pamphlets when her relative had the disease was sufficient, she may not believe she needs to actively seek more information. Two of the women in the interviews implied this much; they thought they knew what they needed to know about breast cancer. One of the women expressed a different view when she described a recent search for information about new technology for detecting breast cancer. Even though this example is one of secondary prevention, it showcases an important vari able at play with women who have family histories time since the relatives cancer experience. Medical science changes rapidly. Women whose mothers or other family members had breast cancer decades ago might benefit from learning about advancements and n ew knowledge about breast cancer prevention. Similarity to the Health Belief Model The theory grounded in the qualitative data about breast cancer prevention informationseeking in women without the disease is a good fit to the Health Belief Models (HBM ) concepts. The HBM is a framework for thinking about why people do or do not take actions to prevent diseases, based on their beliefs about the disease, their perceived susceptibility to the disease, and the weighing of the costs and benefits of the preve ntion (Janz, Champion, & Strecher, 2002). Even though the present study was not a theory testing exploration, many of the HBMs tenets were present in the data, including perceived
191 severity, perceived susceptibility and cues to action (See Table 7 1). One component of the HBM not specifically addressed in the model is perceived severity. The study did not reveal enough about how perceived severity might affect seeking. The study suggest ed that women do not rate breast cancer as a major concern in the commun ity and that women may not seek breast cancer prevention information at high rates. However we do not know if the reverse would be true: if women do perceive the problem to be severe, would they seek information about the problem? Based on this lack of cla rity the research er thought it premature to stress the condition in the model. Table 71. Comparisons between the HBM and the studys findings Health Belief Model t enet Studys f indings Perceived susceptibility Based mostly on the absence or presence of breast cancer in a womans family history; to a lesser extent, based on personal scares, such as a benign breast lump or a suspicious mammogram Perceived severity Priority in community is seemi ngly based on how many people a disease affects, rather than the impact the disease has on the patients lives; deserving more attention are beliefs about what happens to women who develop breast cancer because, even though participants sometimes associated breast cancer with death, they still did not express many r isk avoidance behaviors Perceived benefits Women ascribed positive outcomes to early detection, such as self checks and mammograms. However, the view that breast cancer was related to family history of the disease may affect views about the benefits of pr imary prevention and seeking information about breast cancer prevention on the Internet Perceived barriers Myths, lack of knowledge about breast cancer prevention; barriers to seeking information online include physical access issues Cues to action The woman having a personal scare or a family member developing breast cancer; some implied that news stories that stressed severity of a health problem or risk factors she had might make her seek out more information about the condition Self efficacy No goo d measure about how confident women were to look for and find breast cancer prevention information
192 Despite some of the similarities in the HBM and the studys findings, one important distinction should not be glossed over: seeking breast cancer preventi on information is not actually a form of prevention ; it is an intermediary step for prevention s such as diet changes, alcohol consumption reduction, and smoking cessation to occur. The prevention goals are those healthful lifestyle suggestions, and the beh aviors are adopting those lifestyles, not attaining knowledge about them, which the Internet provides. The Internet may be a helpful facilitator of those behaviors, providing support from other women in online communities or providing tools for nutrition a nd exercise, but finding breast cancer prevention information is merely a step to the goal of adopting the behaviors. On the flip side, the rise in health care consumerism ( Calabretta, 2002; Keckley & Eselius 2009) and empowered patients requires us to think more analytically about the role of information along the disease prevention behavior spectrum. While Internet information seeking may not be a true health behavior in the vein of nutritional attainment or toxin avoidance, it may be becoming a bigger i nfluence on many of the well vetted health behavior theories. Where once information may have played a role in giving people knowledge about health topics and the media may have shaped views through coverage, the Internet as a form of mass media may do bot h of these things as well as serve as a support system, self efficacy enhancer, and a maintenance tool. For example, another widely known health behavior model, the Precaution Adoption Process Model (PAPM) posits that people move through seven stages when approaching a health behavior (Weinstein & Sandman, 2002). In the first two stages, a person moves from not being aware of the health issue to being aware but unengaged by it. Proceeding from that point are the stages wherein people decide if they will ac t, how they will act, actually acting, and maintaining that action. Clearly the Internet has a potential role in changing
193 unawareness about breast cancer prevention into awareness through facts and explanations of the disease. Through many multimedia and i nterpersonal communication tools a woman may also become engaged with the issue. Interactive features on the Internet may also serve as decision aids. For example, the National Cancer Institute has an online breast cancer risk assessment tool to help a woman estimate her own risk (NCI, n.d. ) And a plethora of Web sites offer services to plan and track healthy lifestyle goals such as diet and exercise, which could be considered maintenance functions in disease prevention. When electronic health records an d personalization features become commonplace, more patient specific online tools may give women opportunities to integrate multiple health functions into their health management. For example, the Department of Veterans Affairs (VA) already offers its heal th system patients the online portal, MyHealth e Vet, to perform general health information searches, access lab and test results, and manage prescription medications. Beyond the VA, more integration of personal health history with personal risk of diseases is a likely outcome of the available technology and trends in patient empowerment (U nited S tates Department of Veterans Affairs, 2005) This paper offers no conclusions about the place of online breast cancer prevention information seeking in contemporary health behavior theories and certainly does not argue that the two discussed here are the most appropriate models. More accurately, the study suggests that additional research is needed to understand the place of online prevention informationseeking in each theory, including those with strong social belief components. T hemes Unrelated to Research Questions This study is intended as an academic exploration of the relationship between two theoretical models the ecological model of public health and a new media model of access. However, the ultimate purpose of this type of research is to contribute to the reduction or
194 elimination of breast cancer disease burden. During interviews, women brought up topics that were unrelated to the projects research questio ns but that were, nonetheless, interesting and relevant to the overall effort of preventing breast cancer. For that reason, a few of the topics are presented here. The first topic concerns knowledge transfer and communication between primary care physician s, patients and mammogram service providers about breast cancer screenings. Women in the interview without family histories of breast cancer reported low knowledge of breast cancer and their own risks, even though nearly all of them had received a mammogra m in the past. Two women said they did not know how often mammograms should be scheduled, with one saying she thought her physician would tell her if her personal risk was high enough to require another mammogram. Although women reported a trust in their physicians and sometimes a dependence on the doctors to tell the women what health tests they needed and diseases they might be at risk for, this dependence may present obstacles to educating women about breast health. Joyce described changes she saw in generations approaches to health care seeking, saying: Y ou know, its so different now. I was, the way I was raised, you go to the doctor every year for a checkup. When I was raising my kids, every year they had a scheduled time, okay, Im going to take yo u now to get a checkup. Now that theyre grown, you know, my oldest son is 28, my youngest son is 24. T hey might go to the doctor if theyre sick but other than that theyre not going to go. So its just the way youre raised different, I guess. You know, I was raised up like that They dont think to go to the doctor to get a checkup every year or get their cholesterol checked, blood pressure checked. They just dont think like that. While generational issues involving patients views of sick and preventive care may exist it is important to note that the medical community continues to debate the value of preventive medical care for otherwise healthy people. One particular study that caught the mainstream medias attention (e.g. Blue, 2008) was based on Mehrotra, Zaslavsky, and Ayanians (2007)
195 study of preventive health examinations that showed routine, preventive medical office visits accounted for $7.8 billion annually and a significant portion of physicians time Those who highlight drawbacks of annual physical exams as part of preventive health care suggest that the financial costs to patients and the health care system outweigh the benefits, suggesting that such routine care has not been shown to dramatically improve outcomes (Gordon, Senf & Campos Ou tcalt, 1999) However, other researchers have found positive benefits of preventive health exams, including alleviating patient worry (Boulware, et al., 2007). In the case of mammography, Mehrotra and colleagues (2007) found that these preventive screening s and education were taking place largely outside of regular physical exams. Th e debate about periodic physical exams efficacy is significant to this paper only in that if other women like the few interviewed here rely on interactions with physicians during periodic exams to prompt their breast cancer screening behavior, and that health care venue may be assigned diminished value in the future, other interventions must reach women with breast cancer screening education. Also, i f women perceiving themselves at low risk for breast cancer wait for physicians to initiate mammography discussions, where does that leave women who only go to the doctor when they are ill? This area remains fertile for more research into the prompts that may lead women to adopt screen ings The survey instrument intentionally lacked the precision needed to shed light on mammogram frequency because, again, secondary prevention was not the main focus of this study. However, future research into general breast cancer prevention behaviors s hould endeavor to flesh out issues of perceived adequate prevention (getting one mammogram at age 40 but no more after that age) versus medically recommended prevention (including regular mammograms,
196 clinical breast exams, and/ or whatever physicians recomm end for each woman based on her risk factors for breast cancer). Another common theme among the women who had had past mammograms was the importance of free screenings as part of health fairs or workplace initiatives. When asked why they had received mammo grams when they did not think they were at risk for breast cancer, many women said the reason was that the mammograms were free. One woman who was concerned about promoting health in her community said that she wished more mobile mammograms and free, conve nient screenings were made available to women in poor neighborhoods. States where the interviews and surveys were conducted offered free or low cost mammograms through the National Breast and Cervical Cancer Early Detection Program to women meeting certai n eligibility standards, including income, insurance status, and age (CDC, 2009b) Details of the offers vary somewhat by state. For example, one of the states in the study area offered the screenings for women older than 40 whereas the other state set the minimum age at 45, unless the woman had symptoms. This program means that women who were financially unable to afford mammograms may be able to get one for free, reducing costs as a barrier to the screenings. In addition, positive test results and treatme nts necessitated may also be covered through the program. Among the most useful exercises the researcher engaged in post data analysis was one of semantic dissection of the studys objectives and data collection tools words. Several discussion points emer ged from considering the studys findings and thinking critically about the nuanced meanings in them. Posed as questions, some of the points included the following:
197 What do we consider breas t cancer prevention information seeking? Does it require actually wanting to know what prevents breast cancer and framing Internet searches in those terms? Or can a persons mere search for topics that might prevent breast cancer outside of that context constitute a search for prevention information? The distinction is s o subtle that these words likely do not convey the difference and an example is required. In a case of the former, a woman might go to the Internet with questions about preventing breast cancer using those search terms During her search, the woman might discover that scientists encourage women to eat healthfully in order to prevent many different diseases, including breast cancer. However, in the second case, a woman could look for information about other topics such as diet, exercise, smoking cessation assistance and alcohol consumption reduction. A woman may look for inf ormation on those topics for reasons other than cancer prevention and, by enacting certain behaviors based on the found material, may reduce her breast cancer risk without that having be en her purpose The distinction is one of intent. By looking for and adopting healthy lifestyle behaviors in effort to be physically fit, a woman may find advice about including a variety of fruits and vegetables in ones diet and being physically active. These are the kinds of diet and exercise advice experts recommend for reducing cancer risk. Yet the information she finds may not make any connections to cancer reduction benefits. While this example seems like a minor semantic issue, it had bearing in the studys interviews. Several women said they used to the Internet for healthy recipes, weight management, and to find out the caloric and nutrient contents of foods. Joan, the woman who had breast cancer in her family and was determined to stay healthy her self, mentioned using the online diet plan site, Spark People, for nutrition management. The question arises about how
198 controllable risk factors for breast cancer prevention, including diet and exercise, not smoking, and drinking alcohol in moderation, should be emphasized and branded as breast cancer prevention behaviors or cobranded as good health lifestyles that reduce the risks of a wide range of diseases As previously mentioned, science offers no guaranteed ways to prevent breast cancer except proph ylactic surgery in some cases, (especially since two of the biggest risk factors being female and aging are uncontrollable ); however, these healthy lifestyle behaviors may be of some value in breast cancer prevention efforts. How important are beliefs about online health information seeking? How do beliefs about this relate or translate to health outcomes? This line of questions emerged from listening to women describe their health information seeking processes, including how they decide if a source is good. Constituting one of the extraneous study themes alluded to earlier was this idea of Web site evaluation approaches. In addition to evaluating based on source credibility, which was uncommon, and based on consensus of information across multiple Web s ites, one method employed by the interviews was evaluation based on how well information fits the situation. All of the interviewed women turned to generic search engines such as Google and Yahoo to search for health information, and the surveys pointed to similar reliance upon these online tools. Because the interviewed women reported only searching for health topics that were affecting them or their family members, they approached searches with specific questions in mind, such as: What are the side effects to this medication the doctor prescribed for me? and What treatments are available for this condition affecting my husband? Not surprisingly, the women considered a Web site valuable to them if it answered their questions or helped them understand a topic better by using easy to understand terminology. Betty and Rachel offer explanations of a major evaluation criterion good informational fit to ones information needs.
199 Betty said: Cause usually if I have time Ill sit there and kind of speed read through it and say Nah, thats not the type of information Im looking for, and Ill go back to the list of things and click another one and look at it. Yeah, thats more what Im looking for Similarly, Rachel said she enters her keywords into a search engine, generates hits, and Um, a lot of times I just go through each one. Or if um the exact thing came up, I would just click on that. When asked what she meant by the exact thing, Rachel said it was the answer to the question she had or what she wanted to find. The interviewed women who searched for online health information reported general confidence in their ability to find answers to their questions. However, their confidence stemmed from finding answers to their questions, not necessarily fr om finding accurate answers to their questions or from accurately interpreting the health information on the Web pages. The sites that satisfy their curiosity and answer their questions may not be medically accurate and could even cause the women to make d ecisions that are not in their best interests. Returning to the concept of patient empowerment, women who find inaccurate information about breast cancer may feel a sense of empowerment but lacking true facts and explanations, they are not fully able to m ake informed health decisions. The new media model described at the beginning of this paper posits that ones ability to benefit from online health information is a multicomponent determination, meaning here that a person with low health literacy may feel just as confident as a person with high health literacy that she can find, understand and can implement in her life the information on Web pages. However the two peoples actual benefits may differ greatly. Beliefs about the Internet and self efficacy of Internet skills have some importance in health information seeking likely as a type of initiator of the behavior ; however future research must focus on other measures such as those described in a non exhaustive list in Table 72.
200 An earlier chapter posed the possibility of using a new media model of access ( Newhagen & Bucy 2004) along with the ecological model of health to conceptualize a research agenda for health information seeking. Even though this study did not provide clear direction for this, or even speak to whether or not such a conceptualizing would be appropriate, the possibility remains. With a clearer picture of the appropriate measurements (e.g. beliefs, attitudes, and behaviors about breast cancer and the Internet ; physical access, user demographics, and health literacy), it may be useful to understand how factors influencing Internet use and benefit complement or are distinct from factors influencing health prevention behaviors. Thinking in these terms may help identify behavioral interv entions and opportunities to better utilize technology to promote disease prevention in different populations. If this discussion seems too abstract, consider one shared domain of health behavior and Internet use: self efficacy. What if a woman has low sel f efficacy for enacting preventive health behaviors but high self efficacy for Internet use? Could this womans high self efficacy for using the Internet with its interactivity and multiple opportunities for social modeling interventions increase her sel f efficacy to adopt prevention behaviors? This study can only be used to speculate that there might be opportunities to explore relationships between health status and behaviors with Internet use and benefits of use in order to maximize effectiveness of co nveying health information to the public and subpopulations. Doing this may yield a more interesting and more comprehensive body of literature about Internet use than simple analyses of demographic information as it relates to health information seeking. I t should extend beyond describing how often and for what purposes people of each age category, race, income level, and other demographic variables use the Internet.
201 Table 72. Potential measures in future research about Internet health information seeking and health prevention Health Measures Internet Measures Individual characteristics, including health status, socioeconomic indicators, demographics, family history of the disease, insurance coverage, health literacy Individual beliefs, att itudes and behaviors about a health condition, including self efficacy to prevent a disease Social influences of beliefs, attitudes and behaviors, including interpersonal networks and group membership and the importance placed on health behaviors by a persons social circle Individual characteristics, including socioeconomic indicators, demographics, computer literacy, literacy, physical and cognitive abilities Individual beliefs attitudes and behaviors about the Internet including self efficacy to use the Internet Social influences of beliefs, attitudes and behaviors, including interpersonal networks and group membership, and the importance placed on Internet use by a persons social circle System level measures, including physical access to health services; system racism System level measures, including physical access to the Internet by demographics Ecological studies, including e pidemiolog ical explorations about disease incidence and prevalence, and prevention behavior, and statistics about health system access and use Ecological studies, including surveys of Internet use and health information seeking behavior Health literacy as it relates to the ability to successfully understand condition information and the extent to which this literacy relates to successful outcomes of preventive behavior adoption Health literacy as it relates to the ability to successfully understand condition information and the extent to which this literacy relates to successful outcomes of preventive behavior adoption This future research needs clear definitions of what the ultimate measure of information seeking is and which other measures are important or superfluous to the goal of the Internet serving as a tool for disease prevention. Perhaps the ultimate measure is translating online information into prevention outcomes Included among needed future research are studies comparing perceived understanding of online health content and accuracy of that understanding. The concept of this gap between perceived and real understanding is alr eady considered in clinical settings, where providers are encouraged to explain health instructions to patients and have the patients repeat those instructions in their own words, to ensure the patient interpreted directions as the provider intended (Williams, Davis, Parker & Weiss, 2002)
202 Do womens perceptions of family history equating to breast cancer risk necessarily refer to inherited genes? Only during final analysis did the research er consider that her interpretation of the womens belief that family history is the same thing as hereditary genetics might be shortsighted. Although this interpretation seems most likely, based on context in the interviews, it is possible that the women were referring to families living together in the same environment breathing common air, drinking water from the same source, and eating the same foods After the interview ended, one woman pondered aloud if there might be something in the environment that was causing her family to be disproportionately affected by breas t cancer. The researcher had prior knowledge that the womans community was home to a Superfund site and that locals sometimes expressed concern over what they perceived to be clusters of several health conditions in the area. No other interviewed women me ntioned this environmental possibility. Nevertheless, s everal survey questions were designed to further explore beliefs about breast cancer origins, including separately listed family history and environment items. As reported in the results section of this paper, family history was much more frequently marked than environment, implying that the women saw these as distinct risk factors. Comparisons of study findings to previous research The studys attempt to incorporate and build upon the rich literat ure about African American womens beliefs about breast cancer necessitates a review of what the research revealed compared to previous scholars research. Where Moy and colleagues (2006) had found that African American women do not talk much about breast cancer, this study s survey found that 78% of the respondents said they had spoken to another woman about breast cancer.
203 Another finding compared with Moy and colleagues research was the priority placed on other diseases in the community over breast canc er. Both the surveys and the interviews revealed that problems such as high blood pressure and diabetes were more common in the womens communities; therefore those conditions were considered bigger problems by the women. The interviewed women said that pr oximity to a health condition influences how they talk and seek information about it. While the study did not specifically explore barriers to mammogram use, the interview participants frequent mention of worksite programs and mobile mammography services is interesting in light of Powell and colleagues (2005) finding that lack of time for mammograms was a major barrier to having the screening. For most of the interview ed women it appears that time was not a barrier because they were offered convenient ma mmograms as part of workplace and community wellness initiatives. Several findings relate to studies in the literature about information seeking and Web site selection. First, Kind and colleagues (2005) found that many people who did not have computers a nd were not familiar with the medium were still interested in learning to use them. The current study did not explore this concept but it is complemented by the earlier research that implied peoples nonuse of computers was due more to physical access an d skill deficiency than to disinterest in using the technology or beliefs about the medium. Second, many of the Web site credibility evaluation concerns raised by other researchers (e.g. Bates, Romina, Ahmed & Hopson, 2006; Eysenbauch & Khler 2002) were raised by this study. Even though the study did not focus on processes by which women judge Web site credibility, the i nterviews suggested that women had not ha d training i n evaluating sources of online information.
204 Third, the effect of family members ex periences with breast cancer on prevention seeking was not clear from the study. Matthews and colleagues (2002) had found that emotional issues experienced during relatives cancer experiences sometimes prevented women from searching for cancer information None of the interviewed women cited this as an obstacle. The women did not say it was not an obstacle; instead they just attributed nonseeking to other factors, such as believing they already had enough information about the disease. However, the study did not attempt to measure relatives outcomes from breast cancer.2Practical Implications of the Study Perhaps women whose mothers, sisters, or grandmothers survived the disease would have different beliefs about seeking breast cancer information than women whose relatives died from the cancer. One of the purposes of studies like this one is to apply what is learned from the research to health interventions. The mortality rate disparity between African American and Caucasian women suggests that interventions for breast cancer prevention should remain a priority. Based on this studys research, health organizations seeking to encourage primary or secondary prevention among African American women should stress that women are at risk for breast cancer even if they do not have a family history of the disease. Also, the study suggests a need for more education about mammogram frequency. If a public service announcement or other communication campaign were designed to address these issues, that message might incorporate information about how to find mammography services, including availability of low or nocost mammograms in each state. 2 This would have been a difficult measure in interviews and certainly in surveys. The interview s revealed that the womens self reports about their relatives causes of deaths may not be reliable measures of actual causes. For example, one woman said her mother did not die of breast cancer; rather she died when the cancer spread to another organ. Mo re than likely the cause of death would have reflected the original site, the breast. Despite this inconsistency, studies of womens beliefs prompting or discouraging information seeking could use perception of cause of death as an appropriate measure.
205 Intervention planners seeking to use the Internet for breast cancer prevention education and promotion should anticipate tha t women are online health information seekers but may be unlikely to seek out breast cancer prevention information unless they perceive themselves at risk for the disease. An effective communication campaign incorporating the risk message described above m ay be a cue for women to seek breast cancer prevention information on the Internet. However, the intervention should put effort into incorporating a campaign Web site into promotion messages so the women viewing or hearing the campaign will know where to g o for more information Unless the women are directed to the campaign URL, they may find Web sites through generic searches that have low quality, inaccurate information about breast cancer prevention. Intervention planners should think of ways to increase the likelihood that women will remember the site URL. A short or memorable site name is one possible way. Incorporating the Web site on promotional materials given to women at health fairs is another way. Study Limitations As already stated, this project emphasized the qualitative portion of the study, which describes meaning as resting with the participants instead of the participants meaning representing other women. The survey in this study, although a typically quantitative method of exploration, was ruled by the same interpretation of meaning that applied to the qualitative stage; namely that the survey results were not intended to represent other African American women in the general public. Even though the results included some statistics appropriat e to larger sample sizes, significance and other measures discussed in the findings section are not to be treated as generalizable to the population. The results attempted to approximate beliefs and behaviors of only the women in the group of 160 participa nts constitut ing a nonrandom sample with inherent self selection biases. For that reason, the survey could only be used to enhance inferences made from the qualitative portion and to generate questions for future research.
206 This study is not limited by the qualitative methods employed. Rather, our ability to generalize beyond the participants is inconsistent with the method. Our ability to see how well the results represent other women and what significance the findings have beyond the study is restrained by the method. Therefore, future research might seek to replicate this study with the existing methods as well as complement the study with other methodolog ical explorations. The first study limitation from the interviews relates to the womens ages. The results of interviews may have been different if older women had been included. Despite eligibility extending to women in their early to mid sixties, the oldest woman interviewed was 56. The studys age parameters were designed to cover two things: women w ho were within the ages of mammography screening recommendations and women who might be targets of new media interventions for breast cancer prevention. There are two ways to look at this study ; first for its potential to add to the body of research on the topic and second, for its practical importance. To address the first aim, stratification by age categories would have been necessary and results from interviews within each grouping might have yielded interesting insights into similarities and differences in Internet use by older and younger women. However, t he latter inter pretation seeks to answer questions of how best to communicate messages to women through the Internet. F rom a practicality standpoint, any interventions directly resulting from this stud y would take several years to implement, making women in their fifties and younger more relevant research subjects In other words, if women closer to age 65 had been included, they would be over the age by the time any resulting project began. Although th e studys parameters included women up to age 65, the purpose of this range in the qualitative study was not to look for similarities and differences between the ages, only to set an upper bound for interviewed womens ages not to exceed. It was thought that women older than age 65 might not have had as many natural opportunities to use
207 the Internet in their jobs and life, as the Internet diffused somewhat recently. Future studies should specifically examine age cohorts and Internet use, asking more sophist icated questions about how women in each age group use the Internet and how each group might be targeted with breast cancer prevention interventions For example, opportunities may exist for social marketing and viral marketing of breast cancer prevention messages for younger women but those methods may not effectively reach older women. Another limitation that might be overcome with additional study is the recall bias in asking women to remember details of the last time they used the Internet for differen t purposes. While the Internet users were able to give examples, such as looking up information about a prescription medication, they often had difficulty remembering their search strategies or the Web sites they used. O ne interpretation of the latter migh t be that the women chose sites that popped up near the top of the search engine results, as is consistent with the literature ( Eysenbach & Khler, 2002), but it is also possible that they used a more complex method for selecting Web sites. Several study d esigns might compensate for this recall issue Future research should include studies using a t hinka loud protocol or another form of auto driving wherein the participants are asked to do specific tasks or recreate a past search and explain to the resear cher why they made online decisions at each step. Another possible study design to address this recall issue might be incorporation of a log or diary for women to record details of future online health information searches or using a keystroke capture program to record such searches, with womens permission. However, these methods and even the question about how women might search for breast cancer prevention information are not directly related to the present study, which sought to learn first if women even used the Internet for prevention information, and if they did seek this information, why and how they did it. As the first condition was rarely
208 satisfied in the interviews, questions about search details were hypothetical and of proportionately less value to the study. Other limitations to the survey include the in ability to glean much from yes or no questions about topics like health insurance and mammogram use. Having health insurance does not mean a woman is fully insured. In fact, health insurance studies often refer to a middle class of underinsured individuals ( Kaiser Commission, 2002; Monheit, 1994) The survey did not fully assess a womans access to health care, limiting the ability to make analyses based on that variable. Adding more questions about previous health care system experiences, the type of health insurance coverage held, and the frequency of mammograms might have enhanced the study and should be included in future studies However, analyzing mammography and health care access w as not the ori ginal goal of this study, nor w ere those topics specifically related to research questions. More indepth explorations of mammography beliefs were inconsistent with this projects aims but certainly should received continued research focus given how few women described topics related to primary prevention. Another limitation identified during the survey analys is stage was the low consistency of the scales adapted from Champion and colleagues (2008) mammography belief scales. Mammography beliefs and beliefs about breast cancer prevention and informationseeking are clearly different concepts, although they may be quite related. In addition to measuring different concepts, the question format and the categorization of the original items were modif ied in this study These alterations may account for the low reliability found in the present studys scale questions. T his study was exploratory and intended to measure many beliefs and factors related to breast cancer prevention beliefs and Internet beli efs, not to develop a testable scale for beliefs about using the Internet for breast cancer prevention information. However, future studies should
209 endeavor to create more scales directly examining how beliefs about breast cancer influence breast cancer inf ormation seeking. The survey and interviews both were intended to explore if and how African American women talked about breast cancer and its prevention. One limitation of the methodologies that was revealed by the relatively high number of women reportin g that they had talked to someone about breast cancer was that the nature and frequency of those conversations are unknown and not measureable in the present survey. T he surveyed women may have construed the questionnaire questions meaning to include comm ents with medical professionals about the disease as opposed to strictly dialogue with friends and family members. The interviews impl ied this interpretation as several women said the topic of breast cancer does not come up in their social circles, yet the women mentioned receiving mammograms and having interactions with medical doctors about breast health that would technically qualify as talking about breast cancer Future studies using surveys to assess how women talk about breast cancer with other women should address this limitation by describing contexts of conversations. For example, women might be asked if they had every talked to a female friend about having a mammogram, or if they had discussed primary prevention of breast cancer with their physici ans. Surveyed women might also be asked to select from a list of breast cancer topics ( such as nutrition as a form of primary prevention) they discussed with family members and friends and medical professionals. Distinctions such as these would help resear chers understand whether women are having casual, interpersonal conversations about breast cancer or if dialogue is confined to medical offices. Mammogram use is another question with limited inferential opportunities. As the interviewed women implied, knowledge about the proper use and frequency of mammograms may be low. Surveyed women may have received one mammogram in their lives but not regular
210 screenings every one to two years as guidelines recommend ( NCI, 2002 ). Future surveys could ask women how ma ny mammograms they had had since they turned 40. These answers could be compared to the womens self reported ages to estimate frequency of mammograms in the sample. Although the survey was not primarily about breast cancer and prevention knowledge, a ques tion asking women how often they thought women over 40 needed a mammogram might reveal interesting findings, especially when considered alongside answers about the womens mammogram frequencies. If a 50year old woman believes one mammogram at age 40 is su fficient screening for breast cancer, and she reportedly has had only one of these tests in her lifetime then her perception and behavior are seemingly consistent. However, if another 50year old woman believes mammograms should be performed every one to two years but she has only had one of the screenings, then other factors may be influencing her prevention behaviors. Barrier s such as lack of insurance, inconvenient mammography centers, and other obstacles might be identified. Another limitation involved the relationship between women perceiving problems other than breast cancer to be more widespread in their social group and perceiving themselves to be at risk for those conditions. The study did not address whether perception of other diseases as more co mmon or higher priorities is an actual barrier to breast cancer prevention. It also did not determine whether womens perceptions of personal risk of diabetes, heart disease, high blood pressure, and the other more common diseases reflected the priority ranking they assigned to each condition. This issue might be addressed in future research by using the list of problems found in this survey and including scale beliefs about susceptibility to or risk for each of the health conditions. For example, the Lik ert scale item asking women to think about their personal risk of developing breast cancer compared to risks faced by other women their age could be
211 reworded to ask women about their personal risk of developing high blood pressure, HIV/AIDS, diabetes, and other diseases. Analyzing results of the womens rankings of priority health problems with the womens perceived risk of those conditions might reveal more about how the women see their own health risks as similar to or different from those facing their co mmunities. Future Research Among the many questions that should be addressed in future research is, What is an appropriate cue to action for a woman to seek breast cancer prevention information? In the absence of family histories of breast cancer and per sonal scares with the disease, what mass media, clinical intervention, or interpersonal cues might prompt women to reassess their risks of breast cancer and their needs for information about the disease ? The cue question is important because the study foun d that women who do want health information are relatively self directed in their searches for answers to their questions. In the area of breast cancer prevention, we must determine what the potential cues are if we want to trigger this information seekin g. Several of the women mentioned having seen something about breast cancer in the media but they could not recall the stories and said they did not seek out additional information about breast cancer after seeing the stories. At the same time, most of th e women in the interviews who did not have family histories of breast cancer perceived their risks for the disease to be low. The interviewed and surveyed women focused on tangible risk factors namely heredity, smoking, and diet that could be assessed by looking no further than ones immediate family and ones own behaviors, as opposed to environmental and other factors. This implied that prevention education must include risk education that extends farther than these three factors to show that all wom en are potential ly susceptible to breast cancer ; therefore the woman being educated is at risk. Given the interviewed womens reliance on their physicians for answers and
212 direction about their health, primary care offices may be good opportunities for prevention education. However, many potential types and sources of cues should be developed and tested with a variety of research methods, including experiments and more qualitative studies. Somewhat related to officebased intervention research is the opportunity for fu ture exploration of how primary care doctors and radiologists/breast cancer doctors communicate and how they see each others role in the process of patients breast cancer education. Do primary care physicians expect they can promote mammograms to their patients at age 40, even scheduling appointments for their patients, but expect the mammogram clinic to take over from there and remind women to get future screenings? What happens, then, to women whose mammograms are conducted in more general radiology settings rather than those specializing in mammography? Another line of research should focus more on comparing womens perceived beliefs about the Internet with t he i r Internet skills. A key tenet of the new media model of access that inspired part of this study is that all access is not equal access. Two women with the same computer, Internet connections, reported confidence in online skills, and health questions in mind may have very different experiences in their health searches, depending on how well they actually navigate through Web sites, evaluate source credibility, and understand the often complex language on health Web pages. This concept of content access has at least two implications and interpretations for future studies. The first inte rpretation remains on the personal belief and attitudinal level, with studies of womens perceived Internet skills, perceived utility of the medium for health searches, and related concepts. The second, more interesting line would focus more on outcomes than perception. This area would fall under the general health literacy heading which spanned both
213 Internet and health measure columns in Table 72 For example, for women who want to find out more about taking primary prevention steps to avoid breast cance r such as eating healthfully, limiting alcohol consumption, and exercising how well are they able to find scientific consensus information about those topics versus finding Web advice they believe is good but which lacks scientific merit? When women want to find out how they can get a low cost mammogram, can they find that information? Basically, is the information, if acted upon, likely to help the womans efforts to prevent or detect breast cancer early or is it merely giv ing her a sense of assurance? Although the present study focuses on beliefs, the Internet as a health communication source must either incorporate a behavior change function, such as a woman acting on advice about tobacco use cessation or adopting screening behaviors after she reads a bout prevention on a Web site, or be integrated into a larger health education strategy that ultimately is concerned with healthy behavior adoption or changes. Other Web specific future research may include experiments to find out what African American wom en want in a Web site about breast cancer prevention and what presentation most effectively leads to learning. Also, explorations into how Web sites may be designed for women of many levels of Internet expertise may be useful. Perhaps most interesting woul d be research about how to incorporate Web sites about breast cancer prevention into clinic visits such as in the form of an information prescription. Ritterband and colleagues (2005) said information prescriptions are opportunities for clinicians to dire ct patients and their families to Web sites that contain information that is of high quality, appropriate to their condition, and consistent with their clinicians approach to treatment, in hopes that patients will benefit from the Internet s advantages (e.g. frequently updated information) without choosing and relying upon inaccurate or poor quality information (p. e643). The interviewed women reported that they trust their
214 physicians and prefer information from that source. Integrating breast cancer education at that site might help guide seekers to credible sources of information and serve as a cue for the women to think about prevention when ordinarily they would not. An important reason for studies such as this current one is to account for and try to correct disparities in health outcomes. Inherent in the research questions is the speculation that women of diverse racial and ethnic groups that experience different breast cancer outcomes have unique and multiple obstacles to getting the health care and education needed to prevent or detect breast cancer early. Some of these barriers may exist on the individual level, cultural, and system level. Therefore future research examining Internet and breast cancer beliefs should continue to look at disparities, considering women in general but also women who may have unique issues with technology and the health system. For example, women with low English language skills may face content access issues if Web pages are not available in the womens native language or are written at high reading levels and these women may also face health system access problems and culturally insensitive experiences with health care. This point about cultural sensitivity recalls the literature about tailored health messages discuss ed in Chapter 2. In the studys survey, 64 women said they would prefer Web sites about breast cancer that were specifically for African American women whereas 110 said they would prefer sites for women in general and 36 said it would not matter to them. As the statistics suggest, this survey item was of limited use due to women selecting more than one option (for a total of 210 answers representing 160 respondents), and several explanations exist for the observed greater preference for general audience Web sites. Two possible explanations include respondents perceived social acceptability of a raceneutral answer and a distrust of race -
215 specific health information. Yet race specific information was a very popular area of interest women noted on the survey. Several interviewed women responded to questions about whether they would want breast cancer information Web pages with content tailored to African American women by saying no, a breast is a breast. However, these women and those surveyed, also said th ey might be interested in reading whether or not African American women were differently affected by breast cancer than white women. This area of discussion about preferences regarding health information tailoring is interesting in light of research sugges ting that tailored information is more effective. The surveys wording limited the utility of these questions about race specific breast cancer information tailoring, and the sample size was too small to infer much from the statistics. However, these limitations do not preclude conjecture about the findings and future research. Among questions that arise include the extent to which women are consciously aware of information tailoring in health pamphlets and other communication materials they receive and how cultural tailoring fits into the Internet information seeking experience. The latter, more ambiguously expressed question relates to the studys findings (and confirmation of previous research) that women rely on generic search engines for health informat ion searches and that they typically select from the first Web sites the search yields. African American women conducting a simple search with the keywords breast cancer are likely to find general audience breast cancer resources among the initial hits. This likelihood is due to search engines using formulas that rank site popularity and produce the most popular links among the first sites in search results. Given that African American women are a minority, it is logical that Web sites
216 focusing only on Af rican American women and breast cancer would not be among the most popular sites for general breast cancer keyword searches. The Internet allows us to streamline the information we want RSS blogs, subscriptions to newsletters, and such so that even though it offers opportunities for exposure to many different topics, we still have many options to avoid the information we do not want or do not feel is relevant to us. Joyces description of an online pharmacy information subscription is a good example of this. She subscribes to an electronic newsletter Even thought it offers nutrition, exercise, and other prevention tips, Joyce specifically mentioned that she received information about her medications. More information may have been available. She may eve n have read it. But she remembered that the information was about her medication. The issues of proximity (currently experiencing a condition) and perceived risk likely influence whether or not a woman is attuned to these other pieces of information. Futur e research must incorporate what women believe about disease, how they use the Internet and what opportunities might exist to use the Internet to activate potential health information seekers, not just anticipate seekers visits to health Web pages.
217 APP ENDIX A INTERVIEW PROTOCOL S UBMITTED TO THE INST ITUTIONAL REVIEW BOARD UFIRB 02 Social & Behavioral Research Protocol Submission Title of Protocol: African American womens beliefs about the Internet as a source of breast cancer prevention information Principal Investigator: Heather M. Edwards UFID #: Degree / Title: PhD candidate Department: College of Journalism & Communications Mailing Address: Email Address & Telephone Number: Co Investigator(s): UFID#: Supervisor: Kim Walsh Childer s, PhD UFID#: Degree / Title: Professor of Journalism Department: College of Journalism & Communications Mailing Address: Email Address & Telephone Number: Date of Proposed Research: August 2008 August 2009 Source of Funding (A copy of the grant proposal must be submitted with this protocol if funding is involved): N/A Scientific Purpose of the Study: The study seeks to understand why African American women use or do not use the Internet for breast cancer information. The projects goal is to bridge gaps in the literature about the breast cancer experience in African Americans and Internet use by African Americans. A more important, applied purpose of the study is to better identify potential opportunities and obstacles for using the Internet as a breast cancer prevention promotion tool. Describe the Research Methodology in Non Technical Language: ( Explain what will be done with or to the research participant. ) Women who respond to the flyers will be asked a few screening questions to deter mine if they qualify for the study. The questions include age, self identification as an African American woman, personal history of breast cancer, a mother, sister or daughter with a breast cancer history, and Internet use. Eligible women will be intervie wed for 6090 minutes. They will answer questions and talk about their opinions and beliefs about breast cancer, breast cancer prevention, the Internet, and using the Internet to find information about breast cancer prevention. The researcher will pose q uestions from the interview guide (attached). The sessions will be audio recorded for data analysis, with the womens permission and with the assurance that the researcher will not use the audio file in a way that compromises confidentiality. After the pri ncipal investigators dissertation based on the study is complete, she will destroy the recordings.
218 Participants will be asked to complete a short questionnaire before the interview to provide the researcher with descriptive data about the participants and to facilitate conversation about the topic. Included will be questions about their education level, income, church attendance, Internet use, media use, self described health status, and questions consistent with the interview guides topics. The women wi ll be told that they dont have to answer any question they dont want to answer. The qualitative nature of this study does not allow the researcher to set a participant sample size, but it is expected that no more than 100 women will participate in this phase of the study. Describe Potential Benefits and Anticipated Risks: ( If risk of physical, psychological or economic harm may be involved, describe the steps taken to protect participant.) Minimal physical benefits or risks are anticipated. The partic ipants will not be brea st cancer patients or survivors; therefore they wont be asked to disclose personal medical histories. H owever, they may have close friends and family members who have experie nced the disease. Talking about breast cancer may be emoti onally difficult for some women, both because of mortality issues and the personal nature of the disease. Women will be told upfront what the general purpose of the study is and will be able to opt out of the focus groups and interviews before and at any t ime during the sessions. Local counseling resource information will be available for women who may express emotional distress, such as pain associated with experiencing the illness or death of a loved one. Describe How Participant(s) Will Be Recruited, th e Number and AGE of the Participants, and Proposed Compensation: Participants for the interviews will be recruited in several ways. First, members of community organizations, such as churches and healthrelated interest groups will be contacted for help r ecruiting participants. These individuals will be asked to place fliers advertising the study in places where potential participants might encounter them and also to personally tell women who might be interested in the studys subject about the opportunity The ad may also be emailed to key informants. Second, the researcher will seek to attend meetings of organizations with potential participants and announce the study at these meetings The researchers phone number and email address will be provided for women to contact her and sign up for the study. It is expected that fewer than 1 00 women will participate, and all of them will be older than 18 years of age. Participants will be compensated with $15 for their time, and they will be given printed information about breast cancer prevention. Describe the Informed Consent Process. Include a Copy of the Informed Consent Document: The consent process will consist of providing each participant with a hard copy of the informed consent document, which she will be required to sign stating that she agrees to participate. The womens confidentiality will be ensured by not associating the womens names or other identifying information with the transcribed conversations. Lists including the womens contact informati on (used for organizing the sessions) will be destroyed by shredder or locked in a professors office after the interviews are complete. Women will be given the opportunity to withdraw at a ny time and for any reason. The women will be provided the names, email addresses and phone numbers of the researcher and her supervisor The researcher will answer any questions the participants may have after reading the informed consent document. A copy of the interview consent form is attached. The following contact information will be provided to all participants: Questions/Contact Information: If you have any questions or comments regarding the study please contact UF PhD student Heather Edwards or her supervisor Dr. Kim Walsh Childers. Questions or concerns regar ding study participants rights may be directed to the University of Florida Institutional Review Board (UFIRB) at (352) 3920433.
219 APPENDIX B INTERVIEW GUIDE 1. When I say breast cancer, what do you think about? 2. When I say breast cancer prevention, wh at comes to mind? 3. What kinds of things can women do, if anything, to lower their chance of having breast cancer? 4. What does breast cancer prevention mean to you personally? a. Importance of mammography, BSE b. Lifestyle changes? 5. What does breast cancer mean t o the African American community? a. How do you, your family, and friends talk about breast cancer? b. Compared to other concerns, how important do you think breast cancer is to African American women? 6. What are some of the things youve heard other African Am erican women say about breast cancer prevention? a. Why do African American women think about breast cancer prevention the way they do? b. Tell me about any programs or events about breast cancer that youve seen or heard about in your community. 7. Have you see n or heard information maybe a brochure or on TV, radio, or the Internet that tells you how to help prevent breast cancer? Tell me about what you saw or heard. a. How common is it for women in your community to look for information about preventing breas t cancer? b. Tell me some reasons why you and your friends or family would look for information about preventing breast cancer? 8. Lets shift subjects for a few minutes. Tell me a little about how you use the Internet a. What role does the Internet play in your life? b. What role does the Internet play in your community? c. Tell me about the last time you used the Internet i. What did you use the Internet for? ii. How do you usually use the Internet ? d. What are some features that you like about the Internet as an information source? e. What do you NOT like about using it for information? f. Have any of you ever used the Internet to get health information, and if so, what for? g. Tell me about a time when you looked for health information on the Internet h. What do you think about the Internet as a way to get breast cancer information? i. How likely are women in your community to use the Internet for breast cancer information?
220 j. What about you yourself? How often, if ever, have you used the Internet for breast cancer information? k. (For wom en who use the Internet for breast cancer information) Id like to hear about your experiences looking for breast cancer on the Internet Tell me about what you wanted to learn about breast cancer when you looked for information on the Internet i. What are some reasons why you used the Internet to find information about breast cancer? ii. What sources did you use? iii. How well did the information you find answer your questions about breast cancer? iv. How did finding or not finding the information make you feel? v. Did you use it to talk with health care providers? vi. Did you make any changes in your life after learning about the prevention info rmation? l. [For women who dont use the Internet for breast cancer information], what are some reasons you havent used the Internet for breast cancer information?
221 APPENDIX C SURVEY PROTOCOL SUBM ITTED TO THE INSTITU TIONAL REVIEW BOARD UFIRB 02 Social & Behavioral Research Protocol Submission Title of Protocol: Survey to assess African American women's beliefs about the Internet as a source of br east cancer prevention information Principal Investigator: Heather Edwards UFID #: Degree / Title: MAMC, MPH, PhD candidate Department: College of Journalism & Mass Communications Mailing Address: Email Address & Telephone Number: Su pervisor: Kim Walsh Childers UFID#: Degree / Title: PhD, Professor of Journalism Department: College of Journalism & Mass Communications Mailing Address: Email Address & Telephone Number: Date of Proposed Research: April 2009 Source of Funding (A copy of the grant proposal must be submitted with this protocol if funding is involved): n/a Scientific Purpose of the Study: This studys purpose is to explore African American womens beliefs and opinions about using the Internet for breast cancer prevention information. The ultimate goal of this research is to better understand how to use the Internet and other media to increase awareness of breast cancer prevention in order to reduce the mortality rates from breast cancer in African American women. Describe the Research Methodology in Non Technical Language: ( Explain what will be done with or to the research participant. ) This study will be a quantitative follow up to the UFIRB #2008U 887 qualitative portion of my dissertation. The survey research will incorporate the findings from prior indepth interviews into question items on a written questionnaire. The interview guide submitted to IRB02 in the #2008U 887 study shows the general themes explored in the total project. Participants will b e asked to anonymously answer questions exploring what they think about breast cancer and breast cancer prevention as well as what they think about the Internet The women will also provide answers to demographic questions, such as age, income, whether the y have medical insurance, Internet use, education level, and other statuses that are found in the literature to relate to womens views of health and their health behavior. No details of protected health information will be collected. The women will be ask ed if they or their mother, sister, grandmother, or other relative has ever been diagnosed with breast cancer. Even though women with personal histories of breast cancer will not be included in the study, they will not be prevented from taking and completi ng a survey because some of the surveys will be administered in a group setting. Under those public circumstances, the women would effectively reveal their histories of breast cancer to the group if they did not take
222 a survey. Therefore, all women within t he studys age and race parameters may complete the survey, but they will mark on the questionnaire whether or not they have ever had breast cancer. Describe Potential Benefits and Anticipated Risks: ( If risk of physical, psychological or economic harm may be involved, describe the steps taken to protect participant.) The potential benefit to participants is the possibility of reducing breast cancer incidence and mortality in African American women by increasing public health researchers and social scien tists understanding of how to best encourage cancer prevention. Also, the last page of the survey will contain phone numbers and Web addresses of organizations that provide public education about breast cancer so the participants can learn more about the disease and reducing their own risks. Anticipated risks include potential emotional issues resulting from women with painful personal or familial experiences with cancer being asked to answer cancer related questions. In addition to the breast cancer info rmation resources on the final page of the survey, numbers for counseling services will be listed. Also, the women may be concerned about employers or peers learning about their family histories of cancer or their income, for example. Both of these anticipated risks will be mitigated by stressing the voluntary nature of participation. The women do not have to complete the surveys, and if they choose to do so, they can skip questions they do not wish to answer. However, the women will be encouraged to complete the survey to its fullest extent to provide the most robust information for the study, and they will be assured of their anonymity and the confidentiality of their answers. No names will be asked and no specific information that might identify a woman w ill be cited in the study results. Confidentiality will be protected according to the method by which the surveys are administered. When it is possible for the surveys to be given in groups, the researcher will be present in the group to collect surveys i n an envelope, and she will seal the envelope after all of the questionnaires are returned. The envelope will remain in her possession at all times while she is at the survey facility. In some cases where women fitting the study parameters may not be reached in groups, surveys may be handed out to participants by proxies, such as public librarians or church ministers. When the questionnaires are administered in this way, they will be accompanied by a preaddressed, stamped envelope with the researchers PO Box address on it so the participants can mail it directly to the researcher. Even though the proxies will not be responsible for collecting the surveys, the researcher will explain to them the need to ensure confidentiality and anonymity in the study. D escribe How Participant(s) Will Be Recruited, the Number and AGE of the Participants, and Proposed Compensation: Participants will be women between the ages of 3565 who identify as African American. A maximum of 1000 surveys will be administered. The part icipants will be recruited by word of mouth and in groups. For the latter, groups of women fitting the study parameters, such as churches, hobby groups, libraries, and workplaces, will be identified and the researcher will ask group attendees to complete t he survey. No financial compensation will be provided to the women. Describe the Informed Consent Process. Include a Copy of the Informed Consent Document: The first page of the survey will describe the purpose of the study and that it is being conducted by a graduate student at the University of Florida. It will tell the potential participants that their participation is completely voluntary but that if they choose to complete the survey, the researcher will treat the surveys with strict confidentiality and will not ask for their names. Consent is implied by their reading of the consent document and then completing a questionnaire.
223 APPENDIX D SURVEY INSTRUMENT Breast Cancer Survey Please mark the best answer or answers that describe you. 1. Have you ever looked for information about a health problem or a disease? Yes No 2. Have you ever looked for information about ways to keep yourself from getting a health problem or a disease? Yes No 3. Where do you find out about health topics? (You may choose m ore than one. Check all that apply to you.) Doctor or hospital TV Radio Friends or family members Health fairs Church Internet Work Books 4. Have you ever looked up health information on the Internet ? Yes No 5. Have you ever asked someone to look up he alth information on the Internet for you? Yes No 6. How much time do you spend on the Internet each week? No time Less than 3 hours Between 3 hours and 7 hours More than 7 hours
224 7. Where do you use the Internet ? (Check all of the places where you use t he Internet ) I dont use the Internet I use it at work. I use it at home. I use it at the library. I use it at a friends house, a community center, or somewhere else. 8. If you dont use the Internet why dont you? I dont have time. I dont have acces s (dont have computer, cost of service). I dont know how to use the Internet I have somebody else do it for me. I dont trust the information. The Internet is slow. Theres too much information. 9. What do you do on the Internet ? (Check all of the th ings you do on it) Play games Listen to music Read news Email Buy things Find maps and directions Look up information Look for jobs Help children with school work 10. Have you ever talked to friends and family members about breast cancer? Yes No 11. W hat do you think puts people at risk for breast cancer? (Check all that apply) A family member has or had breast cancer The foods they eat (diet) Being overweight Smoking Getting a breast hit or pulled Growing older Alcohol overuse Something in the environ ment Nothing. It just happens to some people. Other _________________________________ (write in other risks)
225 12. Have you ever looked up information about breast cancer on the Internet ? Yes No 13. If you looked for breast cancer information on the Inter net what made you want to find it? (Check all that apply) Personal scare Family member or friend was diagnosed with breast cancer Doctor mentioned it Saw a pamphlet about breast cancer and wanted to learn more Participated in a cancer Walk or Relay event School project Just curious Another reason _________________________________ 14. Have you ever looked up information about preventing (never getting) breast cancer? Yes No 15. If you would use the Internet to find out about preventing breast cancer, w hat are the reasons youd use it? (Check all the reasons that apply to you.) The Internet has a lot of information. The Internet is fast. The Internet makes it easy to find answers to questions. The Internet lets me read about other women who had breast c ancer. The Internet gives me information I can share with my doctor or ask him/her about. The Internet lets me talk about breast cancer with other women through email and chats. The Internet has information that is easier to understand than what my doctor tells me. The Internet can be used from home, at any time of the day. Another reason not listed _______________________________________________________________ 16. If you would not use the Internet to find out about preventing breast cancer, what are the reasons you would not use it? Check all that apply. I dont have a computer with Internet access. I dont know how to use the Internet I dont trust the information on the Internet I cant find the answers to my questions on the Internet The Internet has too much information. The language on web pages is confusing. Another reason not listed ______________________________________________________________
226 17. If you have used any of these web sites to find health information, which ones did you use? (Che ck all that apply) Yahoo Google WebMD Medline Mayo Clinic A doctors Web site Other __________________________________________ 18. If you wanted information about preventing breast cancer, would you want a web site that talked about breast cancer in Afri can American women or in women in general? African American women Women in general It wouldnt matter to me. 19. What types of questions would you want answered about breast cancer? (Check all that apply) What causes it? Does it affect African American wo men differently than women of other races? Mammography process How to get a mammogram if you cant afford it Warning signs or symptoms Treatments options, such as chemo and radiation Treatment effects How can I find support groups for people with cancer? How can I talk to other people with cancer? Diet/nutrition to reduce my risks Survival Hair falling out Personal stories from people who had it Please read each sentence. Think about how much you agree or disagree with each sentence. Circle whether you strongly disagree, dis agree agree or strongly agree with the sentence. If you dont have an opinion about the sentence, circle No opinion. 20. I can do something to keep from getting breast cancer. Strongly agree Agree Disagree Strongly dis agree No opinion 21. I believe I will get breast cancer during my life. Strongly agree Agree Disagree Strongly disagree No opinion 22. I believe if breast cancer is found early, it can be treated successfully.
227 Strongly agree Agree Disagree Strongly disagree No opinion 23. Compared to other women my age, my chances of getting breast cancer are higher. Strongly agree Agree Disagree Strongly disagree No opinion 24. Having a mammogram will decrease my chances of dying from breast cancer. Strongly agree Agree Disagree Strongly disagree No opinion 25. The treatment I would get for breast cancer would be worse than the cancer itself. Strongly agree Agree Disagree Strongly disagree No opinion 26. Being afraid of breast cancer would keep me from looking for information about preventing breast cancer. Strongly agree Agree Disagree Strongly disagree No opinion 27. If I wanted to find out what I could do to prevent breast cancer, I would use the Internet to find that. Strongly agree Agree Disagree Strongly disagree No opinion 28. I believe I could find information about breast cancer prevention on the Internet easily. Strongly agree Agree Disagree Strongly disagree No opinion Check one answer for each question: 29. Which describes you best? Single, never married Married Live with a partner Divorced Widowed 30. Do you have children? Yes No
228 31. Highest e ducation finished: Middle school (6th8th High school graduate grade) College graduate Graduate school Other 32. How much money do you and the people living in your house make each year combined? Less than $20,000 Between $20,000 and $50,000 Between $ 50,000 and $75,000 More than $75,000 33. Do you have medical insurance? Yes No 34. Do you have a regular doctor that you see? Yes No 35. Check the 3 problems that are the biggest concerns for your community. HIV/AIDS High blood pressure Diabetes Heart d isease Abuse Breast cancer Sickle cell anemia Other cancers besides breast cancer Other problems 36. Have you ever had breast cancer? Yes No 37. Have you ever had any kind of cancer? Yes No 38. Have any of these people in your life ever had breast cance r? Yes Mother Yes Grandmother Yes Sister Yes Aunt or cousin
229 Yes Close friend None of these people 39. Have you ever had a mammogram or another test for breast cancer? Yes No 40. Do you do monthly self exams for breast cancer? Yes No Please wri te your age in years (for example, 52): _____________
230 LIST OF REFERENCES Ad Hoc Committee on Health Literacy for the Council on Scientific Affairs, American Medical Association. (1999). Health literacy: Report of the C ouncil on S cientific A ffairs. Journal of the American Medical Association, Amend, K., Hicks, D. & Ambrosone, C. B. (2006). Breast cancer in AfricanAmerican women: Differences in tumor biology from EuropeanAmerican women. Cancer Research 66, 83278330. 28, 552557. American Cancer Society. (2009a). Lymphedema: What every woman with breast cancer should know. Retrieved April 21, 2009, from www.cancer.org American Cancer Society. (2009 b). What are the risk factors for breast cancer? Retrieved June 8, 2009, from www.cancer.org American Cancer Society. (2008). Breast cancer facts & figures 20072008. Retrieved June 6, 2008, from www.cancer.org American Cancer Society. (2007 a ). Cancer facts & figures 2007. Retrieved November 20, 2007, from www.cancer.org American Cancer Society. (2007 b ). Cancer facts & figures for African Americans 2007 2008. Retrieved November 20, 2007, from www.cancer.org American Cancer Society. (2007c). Detailed guide: Breast cancer: How i s breast cancer staged? Retrieved June 5, 2008, from www.cancer.org American Cancer Society. (2007 d ). Mammograms and other breast imaging procedures. Retrieved May 23, 2008, from www.cancer.org Arora, N. K., Hesse, B. W., Rimer, B. K. Viswanath, K., Clayman, M. L. & Croyle, R. T. (2007). Frustra ted and confused: The American public rates its cancer related information seeking experiences. Journal of General Internal Medicine, 23(3), 223228. Ashing Giwa, K. T., Padilla, G., Tejero, J., Kraemer, J., Wright, K., Coscarelli, A., et al. (2004). Under standing the breast cancer experience of women: A qualitative study of African American, Asian American, Latina and Caucasian cancer survivors. Psycho Oncology, 13(6), 408428. Babbie, E. (1990). Survey research methods ( 2nd ed.). Belmont, California: Wadsworth Publishing Company. Babbie, E., Halley, F. & Zaino, J. (2003). Adventures in social research (5th ed.). Thousand Oaks, California: Sage Publications.
231 Bader, J. L. & Theofanos, M. F. (2003). Searching for cancer information on the Internet : Analyz ing natural language search queries. Bailey, E. J., Erwin, D. O., & Belin, P. (2000). Using cultural beliefs and patterns to improve mammography utilization among African American women: The Witness Proje ct. Journal of the National Medical Association, 92, 136142. Journal of Medical Internet Research, 5 (4), e31. Baker, D.W. (2006). The meaning and the measure of health literacy. Journal of General Internal Medicine, 21, 878883. Baranowski, T., Perry, C. L. & Parcel, G. S. (2002). How individuals, envi ronments, and health behavior interact (pp. 165184). In K. Glanz, B. K. Rimer & F. M. Lewis (Eds.), Health behavior and health education: Theory, research and practice San Francisco: Jossey Bass. Bass, S. B., Ruzek, S. B., Gordon, T. H.; Fleisher, L.; M cKeown Conn, N. & Moore, D. (2006). Relationship of Internet health information use with patient behavior and self efficacy: Experiences of newly diagnosed cancer patients who contact the National Cancer Institutes Cancer Information Service. Journal of H ealth Communication, 11, 219236. Bates, B. R.; Romina, S.; Ahmed, R.; & Hopson, D. (2006). The effect of source credibility on consumers perceptions of the quality of health information on the Internet Medical Informatics and the Internet in Medicine, 31(1), 45 52. Baur, C. E. (2005). Using the Internet to move beyond the brochure and improve health literacy. In Understanding Health Literacy: Implications for medicine and public health. Eds. J. G. Schwartzbert, J. B. VanGeest, & C. G. Wang. USA: AMA Pr ess. Beckjord, E. B., Klassen, A. C. (2008). Cultural values and secondary prevention of breast cancer in African American women. Cancer Control, 15(1), 6371. Berland, G K, Elliott, M N, Morales, L S, Algazy, J I, Kravitz, R L, Broder, M S, Kanouse, D E, et al. (2001). Health information on the Internet : accessibility, quality, and readability in English and Spanish. Journal of the American Medical Association, Bernhardt, J.M., Brownfield, E.D. & Parker, R.M. (2005). Understanding he alth literacy. In Understanding Health Literacy: Implications for medicine and public health. Eds. J.G. Schwartzbert, J.B. VanGeest, & C.G. Wang. USA: AMA Press. 285(20), 26122621. Blue, L. (2008, June 4). Is an annual physical really necessary? Time.com Retrieved June 10, 2009, from www.time.com Blumer, H. (1969). Symbolic interactionism: Perspective and method. Englewood Cliffs, New Jersey: Prentice Hall.
232 Boulware, L. E., Marinopoulos, S., Phillips, K. A., Hwang, C. W., Maynor, K., Mer enstein, D., et al. (2007). Systematic review: The value of the periodic health evaluation. Annals of Internal Medicine, 146 (4), 289300. Bowie, J.; Curbow, B.; LaVeist, T.; Fitzgerald, S. & Pargament, K. (2001). The relationship between religious copi ng style and anxiety over breast cancer in African American women. Braun, S. (2003). The history of breast cancer advocacy. The Breast Journal, 9(Suppl. 2), 101 103. Journal of Religion and Health, 40(4), 411422. Bremner, J. D., Quinn, J., Quinn, W., & Veledar, E. (2006). Surfing the Net for medical information about psychological trauma: An empirical study of the quality and accuracy of trauma related Websites. Medical Informatics and the Internet in Medicine, 31(3), 227 236. Brodie, M.; Flournoy, R E.; Altman, D. E.; Blendon, R. J.; Benson, J. M.; & Rosenbaum, M. D. (2000). Health information, the Internet and the Digital Divide. Health Affairs, 19(6), 255265. Brown, L. W. & Williams, R. D. (1994). Culturally sensitive breast screening programs f or older black women. Nurse Practitioner, 19(3), 21, 2526, 31. Bucy, E. P. (2004). The interactivity paradox: Closer to the news but confused (pp. 4772). In E. P. Bucy & J. E. Newhagen (Eds.), Media access: Social and psychological dimensions of new tech nology use Mahwah, NJ: Lawrence Erlbaum. Bundorf, M. K., Wagner, T. D., Singer, S. J. & Baker, L. C. (2006). Who searches the Internet for health information? Health Services Research, 41 (3), 819836. Calabretta, N. (2002). Consumer driven, patient center ed health care in the age of electronic information. Journal of the Medical Library Association, 90(1), 3237. Carmona, R. H. (2003). Health literacy in America: The role of health care professionals. Remarks to the American Medical Association House of D elegates Meeting on June 14, 2003. Retrieved May 23, 2008, from www.surgeongeneral.gov Carmona, R. H. (2006). Health literacy: A national priority. Journal of General Internal Medicine, 21, 803. Case, D.O.; A ndrews, J.E.; J ohnson, J. D .; Al lard S.L. (2005). Avoiding versus seeking: the relationship of information seeking to avoidance, blunting, coping, dissonance, and related concepts Journal of the Medical Library Association, 93(3) 353362. Centers for D isease Control & Prevention. (2009a). Comparing breast cancer by race and ethnicity. Retrieved June 21, 2009, from www.cdc.gov
233 Centers for Disease Control & Prevention. (2009b). National Breast and Cervical Cancer Early Detection Program: About the Program. Retrieved June 7, 2009, from www.cdc.gov/cancer/nbccedp/ Centers for Disease Control & Prevention. (2008). US Public Health Service Syphilis Study at Tuskegee : The T uskegee Timeline. Retrieved March 23, 2008, from www.cdc.gov Centers for Disease Control and Prevention. (2007). Data 2010: The Healthy People 2010 database. Retrieved March 24, 2008, from www.cdc.gov Cerully, J. L., Klein, W. M. P. & McCaul, K. D. (2006). Lack of acknowledgment of fruit and vegetable recommendations among nonadherent individuals: Associations with information processing and cancer cognitions. Journal of Health Communication, 11, 103115. Champion, V. L., Monahan, P. O., Springston, J. K., Russell, K., Zollinger, T. W., Saywell, R. M., JR, et al. (2008). Measuring mammography and breast cancer beliefs in African American women. Journal of Health Psychology, 13(6), 827837. Chandra, A., Rutsohn, P. & Carlisle, M. B. (2004). Utilization of the Internet by rural West Virginia consumers. Journal of Consumer Health on the Internet 8(2), 45 59. Charon, J. M. (1998). Symbolic interactionism (6th ed.). Upper Saddle River, NJ: Prentice Hall. Chen, W. Y., Rosner, B. & Colditz, G. A. (2007). Moving forward with breast cancer preve ntion. Cancer, 109(12), 23872391. Christensen, H. & Griffiths, K. (2000). The Internet and mental health literacy. Australian and New Zealand Journal of Psychiatry, 34, 975979. Coiera, E. W. & Vickland, V. (2008). Is relevance relevant? User relevance ratings may not predict the impact of Internet search on decision outcomes. Conlin, K. K. & Schumann, L. (2002). Literacy in the health care system: A study on open heart surgery patients. Journal of Journal of the American Medical Informatics Association, published ahead of print. the Consedine, N. S.; Horton, D.; Magai, C. & Kukafka, R. (2007). Breast screening in response to gain, loss, and empowerment framed messages among diverse, low income women. Journal of Health Care for the Poor and Underserved, 18.3, 550566. American Academy of Nurse Practitioners, 14(1), 3842. Creswell J. W. (1994). Research design: Qualitative & quantitative approaches Thousand Oaks, California: Sage Publications. Creswell, J. W. (2007). Qualitative inquiry & research design: Choosing among five approaches. Thousand Oaks, Calif.: Sage.
234 Creswell, J. W., Plano Clark, V. L., Gutmann, M. L., & Hanson, W. E. (2003). Advanc ed mixed methods research designs. In A. Tashakkori, & C. Teddlie (Eds.), Handbook of mixed methods in social & behavioral research (pp. 209240). Thousand Oaks, California: Sage Publications. Cummings, S. R. (2007). Primary prevention of breast cancer: N ew approaches. Maturitas, 57, 3941. Daghio, M. M., Fattori, G., & Ciardullo, A. V. (2006). Assessment of readability and learning of easy to read educational health materials designed and written with the help of citizens by means of two nonalternative m ethods. Advances in Health Sciences Education, 11, 123132. Davis, T. C., Williams, M. V., Marin, E., Parker, R. M., & Glass, J. (2002). Health literacy and cancer communication. CA: A Cancer Journal for Clinicians, 52(3), 134149. Davis, T. C.; Wolf, M. S .; Bass III, P. F.; Middlebrooks, M.; Kennen, E.; Baker, D. W.; Bennett, C. L.; DurazoArvizu, R.; Bocchini, A.; Savory, S.; Parker, R. M. (2006). Low literacy impairs comprehension of prescription drug warning labels. Journal of General Internal Medicine, 21, 847851. DeNavas Walt, C., Proctor, B. D. & Smith, J. (2007). Income, Poverty, and Health Insurance Coverage in the United States: 2006. U.S. Census Bureau. Retrieved June 3, 2008, from www.census.gov DeRose, K. (2005). What is epistemology? A brief introduction to the topic. Retrieved June 4, 2008, from http://pantheon.yale.edu Dickerson, S. S., Boehmke, M., Ogle, C. & Brown, J. K. (2006). Seeking and manag ing hope: Patients experiences using the Internet for cancer care. Oncology Nursing Forum, 33(1), E8 E17. Downie, R. S., Tannahill, C. & Tannahill, A. (1996). Health promotion models and values New York: Oxford University Press. Dutta Bergman, M. (2003). Trusted online sources of health information: Differences in demographics, health beliefs, and healthinformation orientation. Journal of Medical Internet Research, 5 (3), e21. Effect of breast cancer awareness month on mammography use. (2005, October 19). Journal of the National Cancer Institute, 97(20), 1493. Ekman, A., Hall, P. & Litton, J. (2005). Can we trust cancer information on the Internet ? A comparison of interactive cancer risk sites Cancer Causes and Control, 16(6), 765772. Elmore, J. G., Arms trong, K., Lehman, C. D. & Fletcher S. W. (2005). Screening for breast cancer. Journal of the American Medical Association, 293(10), 12451256.
235 Engelman, K. K., Perpich, D. L., Peterson, S. L., Hall, M. A., Ellerbeck, E. F., & Stanton, A.L. (2005). Cancer information needs in rural areas. Journal of Health Communication, 10, 199208. Eysenbach, G. & Khler, C. (2002). How do consumers search for and appraise health information on the World Wide Web? Qualitative study using focus groups, usability tests, an d indepth interviews. BMJ, 324, 573577. Eysenbach, G. (2003). The impact of the Internet on cancer outcomes. CA: A Cancer Journal for Clinicians, 53, 356371. Eysenbach, G., Powell, J., Kuss, O., & Sa, E. R. (2002). Empirical studies assessing the qualit y of health information for consumers on the World Wide Web: A systematic review. Journal of the American Medical Association, 287(20), 26912700. Fobair, P., Stewart, S. L., Chang, S., D'Onofrio, C., Banks, P. J., & Bloom, J. R. (2006). Body image and sex ual problems in young women with breast cancer. Psycho Oncology, 15(7), 579594. Fogel, J. (2004). Internet breast health information use and coping among women with breast cancer. CyberPsychology & Behavior, 7(1), 5963. Fogel, J.; Albert, S. M.; Schnabe l, F.; Ditkoff, B. A. & Neugut, A. I. (2002). Use of the Internet by women with breast cancer. Journal of Medical Internet Research, 4(2), e9. Fowler, F. J. J. (2002). Survey research methods ( 3rd ed.). Thousand Oaks, California: Sage Publications. Fox, S (2006). Online health search 2006. Pew Internet & the American Life Project. Retrieved November 20, 2007, from www.pewi nternet .org Frey, J. and Fontana, A. (1993). The group interview in social research, (pp 2034). In D.L. Morgan, (Ed.) Successful focus groups: Advancing the state of the art. Newbury Park, CA: Sage Publications. Friedman, D. B., HoffmanGoetz, L., & Arocha, J. F. (2006). Health literacy and the World Wide Web: Comparing the readability of leading incident cancers on the Internet Medical Informatics and the Internet in Medicine, 31(1), 67 87. Friedman, D. B., HoffmanGoetz, L., & Arocha, J.F. (2004). Readability of cancer information on the Internet Journal of Cancer Education, 19(2), 1 17122. Friedman, D., & HoffmanGoetz, L. (2006b). Assessment of cultural sensitivity of cancer information in ethnic print media. Journal of Health Communication, 11(4), 425447. Glanz, K., Croyle, R. T., Chollette, V.Y & Pinn, V. W. (2003). Cancer rel ated health disparities in women. American Journal of Public Health, 93(2), 292298.
236 Glaser, B. & Strauss, A. (1967). The discovery of grounded theory. Chicago: Aldine Publishing. Gordon, P. R., Senf, J., & Campos Outcalt, D. (1999). Is the annual complete physical examination necessary? Archives of Internal Medicine, 159(9), 909910. Gorin, S. S., Heck, J. E., Cheng, B. & Smith, S.J. (2006). Delays in breast cancer diagnosis and treatment by racial/ethnic group. Archives of Internal Medicine, 166, 22442252. Grabsch, B., Clarke, D. M., Love, A., Mckenzie, D. P., Snyder, R. D., Bloch, S., et al. (2006). Psychological morbidity and quality of life in women with advanced breast cancer: A cross sectional survey. Palliative and Supportive Care, 4, 4756. Green J. & Thorogood, N. (2004). Qualitative methods for health research. Thousand Oaks, CA: Sage. Greene, J. C. (2007). Mixed methods in social inquiry San Francisco, California: Jossey Bass. Guerra, C. E., Dominguez, F., & Shea, J. A. (2005). Literacy and knowledge, attitudes, and behavior about colorectal cancer screening. Journal of Health Communication, 10, 651 663. Guidry, J. J., Fagan, P., & Walker, V. (1998). Cultural sensitivity and readability of breast and prostate printed cancer education material s targeting African Americans. Journal of the National Medical Association, 90(3), 165169. Guidry, J. J., Matthews Juarez, P., & Copeland, V. A. (2003). Barriers to breast cancer control for African American women. Cancer, 97(S1), 318323. Gustafson, D. H., McTavish, F. M., Stengle, W., Ballard, D., Jones, E., Julesberg, K., et al. (2005). Reducing the digital divide for low income women with breast cancer: A feasibility study of a population based intervention. Journal of Health Communication, 10(7) 173193. Gustafson, D. H.; Hawkins, R. P.; Boberg, E. W.; McTavish, F.; Owens, B.; Wise, M.; Berhe, H. & Pingree, S. (2002). CHESS: 10 years of research and development in consumer health informatics for broad populations, including the underserved. Interna tional Journal of Medical Informatics, 65(3), 169177. Haas, J.; Kaplan, C.; Des Jarlais, G.; Gildengoin, V; Perez Stable, E.; & Kerlikowske, K. (2005). Perceived risk of breast cancer among women at average and increased risk. Journal of Womens Health, 14(9), 845851. Haggstrom, D. A. & Schapira, M. M. (2006). Blackwhite differences in risk perceptions of breast cancer survival and screening mammography benefit. Journal of General Internal Medicine, 21(4), 371377.
237 Halkett, G. K. B., Kristjanson, L. J., & Lobb, E. A. (2008). 'If we get too close to your bones they'll go brittle': Women's initial fears about radiotherapy for early breast cancer. Psycho Oncology, 17(9), 877 884. Hall, L., Dunkelberger, J. & Wheat, J. (2005). Medical connections use of the Internet and traditional sources of health information by rural Alabama households. Journal of Agromedicine, 10(2), 17 26. Halstead, M. T., & Hull, M. (2001). Struggling with paradoxes: The process of spiritual development in women with cancer. Oncology Nursing Forum, 28(10), 1534. Hargittai, E. (2002). Secondlevel digital divide: Differences in peoples online skills. First Monday, 7(4). Retrieved September 18, 2007, from www.firstmonday.org Heaney, C. A. & Is rael, B. A. (2002). Social networks and social support (pp. 185209). In K. Glanz, B. K. Rimer & F. M. Lewis (Eds.), Health behavior and health education: Theory, research and practice San Francisco: Jossey Bass. Hede, H. (2008). Fat may fuel breast canc er growth. Journal of the National Cancer Institute, 100(5), 298299. Helft, P. R., Eckles, R. E., JohnsonCalley, C. S. & Daugherty, C. K. (2005). Use of the Internet to obtain cancer information among cancer patients at an urban county hospital. Journal of Clinical Oncology 23(22), 49544962. Helft, P. R., Hlubocky, F. & Daugherty, C. K. (2003). American oncologists views of Internet use by cancer patients: A mail survey of American Society of Clinical Oncology members. Journal of Clinical Oncology Hoffman Goetz, L., & Mills, S. L. (1997). Cultural barriers to cancer screening among African American women: A critical review of the qualitative literature Womens Health, 3(3 4), 183201. 21(5), 942947. Honda, K. & Neugut, A. (2004). Associations between perceived cancer risk and established risk factors in a national community sample. Cancer Detection and Prevention, 28(1), 1 7. Hong, T. (2006). Contributing factors to the use of healthrelated Websites. Journal of Health Communicati on, 11, 149165. Houston, T. K. & Allison, J. J. (2002). Users of Internet health information: Differences by health status. Journal of Medical Internet Research, 4(2) Institute of Medicine (2003). Fulfilling the potential of cancer pre vention and early detection. Washington, DC: National Academy Press. e7.
238 Jackson, C. L.; Batts Turner, M. L.; Falb, M. D.; Yeh, H.; Brancati, F. L. & Gary, T. L. (2005). Computer and Internet use among urban African Americans with type 2 diabetes. Journal of U rban Health, 82(4), 575583. Jaffe, J. M. (1997). Media interactivity and self Efficacy: An examination of hypermedia first aid instruction. Journal of Health Communication, 2(4), 235251. Janz, N. K., Champion, V. L., & Strecher, V.J. (2002). The health belief model. In K. Glanz, B.K. Rimer & F.M. Lewis (Eds.), Health behavior and health education: Theory, research, and practice (Third ed., pp. 4566). San Francisco, California: Jossey Bass. John, A.K. (2005). A critical appraisal of Internet resources on colorectal cancer. C olorectal Disease, 8, 217223. Johnson, B., & Turner, L. A. (2003). Data collection strategies in mixed methods research. In A. Tashakkori, & C. Teddlie (Eds.), Handbook of mixed methods in social & behavioral research (pp. 297 320). Thousand Oaks, California: Sage Publications. Johnson, J. C. & Meischke, H. (1991). Womens preferences for cancer information from specific communication channels. American Behavioral Scientist, 34(6),742755. Jones, B. A., Patterson, E. and Calvocoressi, L. (2003). Mammography s creening in African American w omen: Evaluating the r esearch. Cancer Kaiser Commission on Medicaid and the Uninsured. (2002). Underinsured in America: Is health coverage adequate? Retrieved June 10, 2009, from 97(S1), 258272. www.kff.org Kaphingst, K. A., Zanfini, C. J. & Emmons, K. M. (2006). Accessibility of web sites containing colorectal cancer information to adults with limited literacy (United States). Kaplan, G. A., Everson, S. A. & Lynch, J. W. (2000). Contribution of social and behavioral research to an understanding of the distribution of disease: A multilevel approach (pp. 37 80). In B. D. Smedley. & S. L. Syme (Eds.). Promoting Health: Inter vention strategies from social and behavioral research. Washington, D.C.: National Academy Press. Cancer Causes Control, 17(2), 147151. Keckley, P.H. & Eselius, L.L. (2009). 2009 Survey of health care consumers: Key findings, strategic implications Deloitte Center for Health Solutions. Retrie ved June 8, 2009, from www.deloitte.com Kemper, E. A., Stringfield, S., & Teddlie, C. (2003). Mixed methods sampling strategies in social science research. In A. Tashakkori, & C. Teddlie (Eds.), Handbook of mixed methods in social & behavioral research (pp. 273286). Thousand Oaks, California: Sage Publications. Kickbusch, I. & Ratzan, S.C (2001). Health literacy: Making a difference in the USA. Journal of Health Communication, 6, 8788.
239 Kind, T.; Huang, Z. J.; Farr, D. & Pomerantz, K. L. (2005). Internet and computer access and use for health information in an underserved community. Ambulatory Pediatrics, 5(2), 117121. Kirschning, S., Kardorff, E., & Merai, K. (2007). Internet use by the families of cancer pati ents -help for disease management? Journal of Public Health, 15(1), 2328. Kline, K. K. (2007). Cultural sensitivity and health promotion: Assessing breast cancer education pamphlets designed for African American Women. Health Communication, 21(1), 85 96. Kreps, G. (2006). Communication and racial inequities in health care. American Behavioral Scientist, 49, 760774. Kreuter, M. W. & Wray, R.J. (2003). Tailored and targeted health communication: Strategies for enhancing information relevance. American Jour nal of Health Behavior, 27(Suppl 3), S227S232. Kreuter, M. W., Skinner, C. S., et al. (2004). Responses to behaviorally vs culturally tailored cancer communication among African American women. American Journal of Health Behavior, 28(3), 195 207. Kutner, M.; Greenberg, E.; Jin, Y.; Paulsen, C.; & White, S. (2006). The health literacy of Americas adults: Results from the 2003 National Assessment of Adult Literacy. U.S. Department of Education Retrieved October 12, 2006, from http://nces.ed.gov Lannin, D. R.; Mathews, H. F.; Mitchell, J.; Swanson, M. S.; Swanson, F. H. & Edwards, M. S. (1998). Influence of socioeconomic and cultural factors on racial differences in late stage presentation of breast cancer. Journal of the American Medical Association, 279, 18011807. LaVeist, T. (2005). Minority populations and health: An introduction to health disparities in the United States San Francisco : Jossey Bass. Leydon, G. M., Boulton, M., Moynihan, C. et al. (2000).Cancer patients information needs and information seeking behaviour: In depth interview study. BMJ, 320, 909913. Lincoln, Y., & Guba, E. (1985). Establishing trustworthiness (Chp. 11). Naturalistic inquiry Beverly Hills, CA: Sage Publications. Lindlof, T. R., & Tayl or, B. C. (2002). Qualitative communication research methods (Second ed.). Thousand Oaks, California: Sage Publications. Littlejohn, S. W. (2002). Theories of human communication (7th ed.). Belmont, CA: Wadsworth/Thomson Learning.
240 Mancini, J., Nogus, C., Adenis, C., Berthet, P., Bonadona, V., Chompret, A., et al. (2006). Patients characteristics and rate of Internet use to obtain cancer information. Journal of Public Health, 28(3), 235237. Marks, D. F.; Murray, M.; Evans, B.; Willig, C.; Woodall, C. & Sykes, C. M. (2006). Health psychology: Theory, research & practice London: Sage. Matthews, A. K., Sellergren, S. A., Manfredi, C. & Williams, M. (2002). Factors influencing medical information seeking among African American cancer patients. Journal of H ealth Communication, 7, 205219. McKinnon, J.D. & Bennett, C.E. (2005). We the people: Blacks in the United States. U.S. Census Bureau. Retrieved June 4, 2008, from www.census.gov McPherson, K., Steel, C. M. & Dixon, J. M. (2000). ABC of breast diseases. Breast cancer epidemiology, risk factors, and genetics. Mehrotra, A., Zaslavsky, A. M., & Ayanian, J. Z. (2007). Preventive health examinations and preventive gynecological examinations in t he United States. Archives of Internal Medicine, 167(17), 18761883. BMJ, 321(7261), 624628. Meric, F., Bernstam, E. V., Mirza, N. Q., Hunt, K. K., Ames, F. C., Ross, M. I., et al. (2002). Breast cancer on the World Wide Web: Cross sectional survey of quality of information and popularity of websites. BMJ, 324(7337), 577 581. Metz, J. M., Devine, P., DeNittish, A., et al. (2003). A multi institutional study of Internet utilization by radiation oncology patients. International Journal of Radiation Oncology, Biology, Physics, 56, 12011205. Michael, Y. L., Kawachi, I., Berkman, L. F., Holmes, M. D., & Colditz, G. A. (2000). The persistent impact of breast carcinoma on functional health status. Cancer, 89(11), 21762186. Miller, A. B. (1995). Primary prevention of cancer: Needs and opportunities for research. Environmental Health Perspectives, 103(Suppl. 8), 313317. Minority Womens Health, U.S. Department of Health and Human Services. (2007). Overweight & obesity. Retrieved March 20, 2008, from w ww.4woman.gov Monheit, A. C. (1994). Underinsured Americans: A review. Annual Review of Public Health, 15(1), 461. Morahan Martin, J. M (2004). How Internet users find, evaluate, and use online health information: a cross cultural review. Cyberpsychol ogy & Behavior, 7(5), 497510. Morgan, D. L. (1998a). Practical strategies for combining qualitative and quantitative methods: Applications to health research. Qual Health Res, 8(3), 362376.
241 Morgan, D.L. (1998b). The focus group guidebook: Focus group kit 1. Thousand Oaks, CA: Sage. Morgan, D.L. (1988). Focus groups as qualitative research. Newbury Park, CA: Sage. Morse, J. M. (2003). Principles of mixed methods and multimethod research design. In A. Tashakkori, & C. Teddlie (Eds.), Handbook of mix ed methods in social & behavioral research (pp. 189 208). Thousand Oaks, California: Sage Publications. Moy, B.; Park, E.; Feibelmann, S.; Chiang, S.; & Weissman, J. (2006). Barriers to repeat mammography: Cultural perspectives of African American, Asian, and Hispanic women. Psycho oncology, 15, 623634. National Cancer Institute. (2009a ). Stages of breast cancer. Retrieved June 5, 2009, from www.cancer.gov National Cancer Institute. (2009b). What is prevention? Retrieved June 7, 2009, from www.cancer.gov National Cancer Institute. (2005a ). Cancer: Questions and answers. Retrieved June 21, 2009, from www.cancer.gov National Cancer Institute. (2005b). What you need to know about breast cancer. Retrieved February 4, 2008, from www.cancer.gov National Cancer Institute. (2002). NCI statement on mammography screening Retrieved June 9, 2009, from www.cancer.gov National Cancer Institute. (n.d.). The Breast Cancer Risk Assessment Tool Retrieved June 7, 2009, from www.cancer.gov Newhagen, J. E. & Bucy, E. P. (2004). Routes to media access (pp. 3 23). In E. P. Bucy & J. E. Newhagen (Eds.), Media access: Social and psychological dimensions of new technology use Mahwah, NJ: Lawrence Erlbaum. Newman, I., Ridenour, C. S., Newman, C., & DeMarco Jr., George Mario Paul. (2003) A typology of research purposes and its relationship to mixed methods. In A. Tashakkori, & C. Teddlie (Eds.), Handbook of mixed methods in social & behavioral research (pp. 167188). Thousand Oaks, California: Sage Publications. Noar, S. M., Benac, C. & Harris, M. S. (2007). Does tailoring matter: Meta analytic review of tailored print health behavior change intervention. Psychological Bulletin, 133(4), 673693. Nutbeam, D. (2000). Health literacy as a public health goal: A challenge for contemporary he alth education and communication strategies into the 21st century. Health Promotion International. 15(3), 259267.
242 Office of Minor ity Health, Department of Health & Human Services. (2005) What is cultural competency? Retrieved April 21, 2009, from www.omhrc.gov Orenstein, S. (2003, February 1). The selling of breast cancer: Is corporate America's love affair with a disease that kills 40,000 women a year good marketing--or bad medicine? CNN.com Retrieved June 3, 2006, from http://money.cnn.com Owen, J. E., Klapow, J. C., Roth, D. L., Shuster, J. L, Bellis, J., Meredith, R., et al. (2005). Randomized pilot of a se lf guided Internet coping group for women with early stage breast cancer. Annals of Behavioral Medicine, 30 (1), 5464. Paisley, J. & Skrzypczyk, S. (2005). Qualitative investigation of differences in benefits and challenges of eating fruits versus vegetabl es as perceived by Canadian women. Journal of Nutrition Education and Behavior, 37(2), 77 82. Paskett, E. D., Alfano, C. M., Davidson, M. A., Andersen, B. L., Naughton, M. J., Sherman, A., et al. (2008). Breast cancer survivors' healthrelated quality of l ife. Cancer, 113(11), 32223230. Pawlak, R. (2005). Economic considerations of health literacy. Nursing Economics, 23(4), 173180. Payne J. G. & Schulte, S. K. (2003). Mass media, public health, and achieving health literacy. Journal of Health Communicati on, 8:124125. Pereira, J. L., Koski, S., Hanson, J., Bruera, E. D. & Mackey, J. R. (2000). Internet usage among women with breast cancer: An exploratory study. Clinical Breast Cancer, 1(2), 148153. Perocchia, R. S., Rapkin, B., Hodorowski, J. K., Davis, N. L., McFarlane, A. R. & Carpenter, R. (2005). Raising awareness of online cancer information: Helping providers empower patients. Journal of Health Communication, 10(7), 157172. Peterson, G., Aslani, P. & Williams, K A. (2003). How do consumers search for and appraise information on medicines on the Internet ? A qualitative study using focus groups. Journal of Medical Internet Research, 5 (4), e33. P ew Internet & American Life Project. (2009). Internet activities. Retrieved June 20, 2009, from www.pewinternet.org Phillips, J. M. & Williams Brown, S. (2005). Cancer prevention among racial ethnic minorities. Seminars in Oncology Nursing, 21(4), 278285. Powell, J. A., Low, P., Griffiths, F. E., & Thorogood, J. (2005). A critical analysis of the literature on the Internet and consumer health information. Journal of Telemedicine and Telecare, 11 (Supplement 1), 4143.
243 Powell, M. E.; Carter, V.; Bonsi, E.; Johnson, G.; Williams, L.; Taylor Smith, L.; et al. (2005). Inc reasing mammography sc reening among African American w omen in rural areas. Journal of Health Care for the Poor and Underserved, 16, 1121. Ramanadhan, S. & Viswanath, K. (2006). Health and the information nonseeker: A profile. Health Communication, 20(2), 131139. Raupach, J. C. A. & Hiller, J. E. (2002). Information and support for women following the primary treatment of breast cancer. Health Expectations, 5(4), 289301. Rees, C. E. & Bath, P. A. (2001). Informationseeking behaviors of women with breas t cancer. Oncology Nursing Forum, 28(5), 899907. Retsky, M. W., Demicheli, R., Gukas, I. D., & Hrushesky, W. J. M. (2007). Enhanced surgery induced angiogenesis among premenopausal women might partially explain excess breast cancer mortality of blacks com pared to whites: An hypothesis. International Journal of Surgery, 5(5), 300 304. Richards, L. & Morse, J. M. (2007). Readme first for a users guide to qualitative methods. Thousand Oaks, Calif.: Sage. Rimal R. N. and Flora J. A. ( 1997 Rogers, E. M. (1994). A history of communication study: A biographical approach. New York: Free Press. ) Interactive technology attributes in health promotion: practical and theoretical issues. In Street, R. L., Jr, Gold, W. R. and Manning, T. (eds), Health Promotion and Interactive Technology. Lawrence Erlbaum Associates, Mahwah, NJ, pp. 1938. Ross, B. S., Potter, L. S., & Armstrong, K. A. (2004). Improving patient educational literature: An understandable patient package insert for the pill. Journal of Obstetric, Gy necologic, and Neonatal Nursing, 33, 198208. Russell, K. M., Monahan, P., Wagle, A. & Champion, V. (2007). Differences in health and cultural beliefs by stage of mammography screening adoption in African American women. Cancer, 109(2 Suppl) 38695. Rutherford, J.; Holman, R.; MacDonald, J.; Taylor, A.; Jarrett, D. & Bigrigg, A. (2006). Low literacy: A hidden problem in family planning clinics. Journal of Family Planning & Reproductive Health Care 32(4), 235240. Rutten, L. F., Moser, R. P., Beckjord, E. B., Hesse, B. W. & Croyle, R. T. (2007). Cancer communication: Health Information National Trends Survey 2003 & 2005. Washington, D.C.: National Cancer Institute. Satterlund, M. J., McCaul, K. D. & Sandgren, A. K. (2003). Information gathering over time by breast cancer patients Journal of Medical Internet Research, 5 (3), e15.
244 Schmidt, K., & Ernst, E. (2004). Assessing websites on complementary and alternative medicine for cancer. Annals of Oncology, 15(5), 733 742. Sentell, T. L., & Halpin, H. A. (2006). Importance of adult literacy in understanding health disparities. Journal of General Internal Medicine, 21, 862866. Severin, W. J. & Tankard, J. W., Jr. (2001). Communication theories: Origins, methods, and uses in the mass media (5thShaw, B. R., Hawkins, R., McTavish, F., Pingree, S. & Gustafson, D. H. (2006). Effects of insightful disclosure within computer mediated support groups on women with breast cancer. Health Communication, 19(2), 133142. ed. ). USA: Addison Wesley Longman. S illence, E., Briggs, P., Harris, P. R., & Fishwick, L. (2007). How do patients evaluate and make use of online health information? Social Science & Medicine, 64(9), 1853 1862. Simon, C. E. (2006). Breast cancer screening: Cultural beliefs and diverse populations. Health & Social Work, 31(1), 36 43. Simon, M. S., Korczak, J. F., Yee, C. L., Malone, K. E., Ursin, G., Bernstein, L., et al. (2006). Breast cancer risk estimates for relatives of white and African American women with breast cancer in the women's contraceptive and reproductive experiences study. Journal of Clinical Oncology, 24(16), 24982504. Steckler, A., McLeroy, K. R., Goodman, R. M., Bird, S. T., & McCormick, L. (1992). Toward integrating qualitative and quantitative methods: An introduction. Health Education Quarterly, 19(1), 1 8. Steuer, J. (1992), Defining virtual reality: Dimen sions determining telepresence, Journal of Communication Stolley, M. R.; Sharp, L. K.; Wells, A. M.; Simon, N.; & Schiffer, L. (2006). Health behavi ors and breast cancer: Experiences of urban African American Women. Health Education & Behavior, 33(5), 604 624. 42, 7393. Stout, P. A., Villegas, J., & Kim, H. (2001). Enhancing learning through use of interactive tools on healthrelated websites. Health Education Research, 16(6), 721733. Street, R. L., Jr. & Rimal, R. N. (1997). Health promotion and interactive technology: A conceptual foundation (pp. 118). In R. L. Street, W. R. Gold, & T. Manning (Eds.). Health Promotion and Interactive Technology Hillsdale, N J: LEA. Strolla, L. O., Gans, K. M., & Risica, P. M. (2006). Using qualitative and quantitative formative research to develop tailored nutrition intervention materials for a diverse low income audience. Health Education Research, 21(4), 465476. Talosig Ga rcia, M., & Davis, S. W. (2005). Informationseeking behavior of minority breast cancer patients: An exploratory study. Journal of Health Communication, 10, 5364.
245 Tashakkori, A., & Teddlie, C. (1998). Mixed methodology: Combining qualitative and quantitative approaches Thousand Oaks, California: Sage Publications. Teddlie, C., & Tashakkori, A. (2003). Major issues and controversies in the use of mixed methods in the social and behavioral sciences. In A. Tashakkori, & C. Teddlie (Eds.), Handbook of mixed methods in social & behavioral research (pp. 350). Thousand Oaks, California: Sage Publications. Teolis, M. (2004). Breast cancer Internet information. Journal of Consumer Health on the Internet 8(3), 69 79. Tessaro, I., Eng, E. & Smith, J. (1994). Bre ast cancer screening in older African American women: Qualitative research findings. American Journal of Health Promotion, 8(4), 286293. Till, B. D. & Nowak, L.I. (2000). Toward effective use of cause related marketing alliances. Journal of Product & Brand Management, 9(7), 472 484. Turnock, B. J. (2004). Public health: What it is and how it works Sudbury, Mass.: Jones and Bartlett. United States Department of Health and Human Services (2000). Healthy People 2010: Understanding and improving health. Ret rieved April 5, 2007, from www.healthypeople.gov United States Department of Health and Human Services. (n.d.). Healthy People 2010. Retrieved March 24, 2008, from ww w.healthypeople.gov United States Department of Veterans Affairs. (2005). About My Healthe Vet. Retrieved June 7, 2009, from www.myhealth.va.gov Unruh, H. K., Bowen, D. J., Meischke, H. Bush, N., Wooldridg e, J .A (2004). Women's approaches to the use of new technology for cancer risk information. Women Health, 40(1), 5978. Walsh Childers, K. & Edwards, H. M. (2007). Finding answers online: Comprehensiveness and accuracy in online information about breast cancer. Paper presented at the American Public Health Association 2007 Annual Meeting, Washington, DC. Wathen, C. N., & Harris, R. M. (2007). "I try to take care of it myself." How rural women search for health information. Qualitative Health Research, 17(5), 639651. Weinstein, N. D., & Sandman, P. M. (2002). The precaution adoption process model. In K. Glanz, B. K. Rimer & F. M. Lewis (Eds.), Health behavior and health education: Theory, research, and practice ( 3rd ed., pp. 121143). San Francisco, Cali fornia: Jossey Bass.
246 Weiss, B. D.; Francis, L.; Senf, J. H.; Heist, K.; & Hargraves, R. (2006). Literacy education as treatment for depression in patients with limited literacy and depression: A randomized controlled trial. Journal of General Internal Medi cine, 21,823828. Weissenberger, C., Jonassen, S., BeranekChiu, J., Neumann, M., Mller, D., Bartelt, S., et al. (2004). Breast cancer: Patient information needs reflected in English and German web sites. British Journal of Cancer, 91(8), 14821487. Wide r, J. (2005). Women most fear breast cancer, but heart disease is the top killer. Society for Womens Health Research. Retrieved September 20, 2006, from www.womenshealthresearch.org Williams, B. A.; Lindquist, K.; Sudore, R. L.; Covinsky, K. E.; & Walter, L. C. (2008).Screening mammography in older women: Effect of wealth and prognosis Archives of Internal Medicine, 168(5), 514520. Williams, D. R., Neighbors, H. W. & Jackson, J. S. (2003). Raci al/ethnic discrimination and health: Findings from community studies. American Journal of Public Health, 93, 200208. Williams, G. A., Abbott, R. R. & Taylor, D. K. (1997). Using focus group methodology to develop breast cancer screening programs that recr uit African American women. Journal of Community Health, 22, 4556. Williams, M. V.; Davis, T.; Parker, R. M. & Weiss, B. D. (2002). The role of health literacy in patient physician communication. Family Medicine Journal, 34(5), 383389. Williamson, K. (2 005). Where one size does not fit all: Understanding the needs of potential users of a portal to breast cancer knowledge online. Journal of Health Communication, 10, 567580. Wilmoth, M. C., & Sanders, L. D. (2001). Accept me for myself: African American women's issues after breast cancer. Oncology Nursing Forum, 28(5), 875879. Winefield, H., Coventry, B., Pradhan, M., Harvey, E., & Lambert, V. (2003). A comparison of women with breast cancer who do and do not seek support from the Internet Australian J ournal of Psychology, 55(1), 3034. Wolf, M. S., Bennett, C. L.; Davis, T. C.; Marin, E.; & Arnold, C. (2005). A qualitative study of literacy and patient: Response to HIV medication adherence questionnaires. Journal of Health Communication, 10, 509517. Wolff, M.; Bates, T.; Beck, B.; Young, S.; Ahmed, S.M.; & Maurana, C. (2003). Cancer prevention in underserved African American communities: Barriers and effective strategies A review of the literature. Wisconsin Medical Journal, 102(5), 36 40.
247 Yabroff, K. R. (2000). Breast cancer epidemiology, prevention and costs of care: Implications for disease management programmes. Disease Management and Health Outcomes, 8(4), 197210. Zarcadoolas, C., Pleasant, A., Greer, D.S. (2003). Elaborating a definition of heal th literacy: A commentary. Journal of Health Communication, 8, 119120. Zarcadoolas, C.; Blanco, M.; Boyer, J.F.; & Pleasant, A. (2002). Unweaving the Web: An exploratory study of low literate adults navigation skills on the World Wide Web. Journal of Health Communication, 7, 309324. Ziebland S., Chapple A., Dumelow C., Evans J., Prinjha S. & Rozmovits L. (2004) How the Internet affects patients experience of cancer: A qualitative study. British Medical Journal, 328, 564569.
248 BIOGRAPHICAL SKETCH Heath er Edwards graduated from Samford University in 2001 with a Bachelor of Arts degree in political science and later earned a Master of Arts in Mass Communication and a Master of Public Health from the University of Florida. She began the University of Flori das doctoral program in mass communication in 2006. During the program, Ms. Edwards studied many facets of health communication and social and behavior sciences, including social marketing, doctor patient communication, media coverage of health stories, a nd new media and health. While at the University of Florida, Ms. Edwards was a graduate research assistant to Drs. Kim Walsh Childers and Debbie Treise, and a research assistant at the Department of Veterans Affairs Rehabilitation Outcomes Research Center in Gainesville. Prior to begi nning the doctoral program at the University of Florida Edwards worked at the American Cancer Society.