grounded theory.
Providing an alternative to positivism (Bryant & Charmaz, 2007), feminist research and constructivist grounded theory are evolving and
organic, emerging from the data (Wuest, 1995). Both are interactive processes involving the researcher and the participants, wherein data are co-created (Bryant & Charmaz, 2007; Charmaz, 2006; King, 1994). Diverse perspectives are explored and integrated, not dismissed or explained away (Wuest, 1995). Both approaches recognize multiple realities and focus on contextual
influences (Bryant & Charmaz, 2007; Wuest, 1995). Thus, the researchers and the participants are recognized as mutually embedded in the social context of the research (Bryant & Charmaz, 2007; Charmaz, 2006).
Feminists ask questions pertinent to women’s lives, of interest to
women, grounded in women’s experience, and born of socio-political struggle (King, 1994). Constructivist grounded theory and feminist research both emphasize the emergence of theory from the data and allow for the
exploration of women’s experiences without the constraints or guidance of a predefined framework (Bryant & Charmaz, 2007). Both approaches allow for the analysis of complex questions and for the voices of participants to be heard
as they tell their stories (Keddy et al., 1996; Wuest, 1995). Using constructivist grounded theory methods, feminist researchers can explore salient problems in people’s lives and how they respond to them from the participants’ point of view (Keddy et al., 1996).
In both constructivist grounded theory and feminist research, power imbalances between the researcher and the participants are recognized and addressed as researchers strive to share power and assume a non-hierarchal, non-oppressive stance (Bryant & Charmaz, 2007; Charmaz, 2006; King, 1994; Wuest, 1995). Participants are recognized as experts on their own experiences and their subjective voices are privileged as legitimate sources of knowledge (King, 1994; Wuest, 1995). Reflexivity, fundamental to feminist research and to constructivist grounded theory, encourages researchers to reflect upon,
examine critically, and explore analytically their impact on the research, their own and the participants’ social locations, and the contextual influences of the research and the participants’ lives (Bryant & Charmaz, 2007; Charmaz, 2006; King, 1994; Wuest, 1995).
Recognizing that theory building is a process that flows from the data (Wuest, 1995), feminist and constructivist grounded theory researchers and participants co-construct knowledge (Bryant & Charmaz, 2007; Charmaz, 2006). As such, both approaches employ member checking to ensure that the researcher’s interpretation of the data and theory building is consistent with
the participants’ perspectives (Wuest, 1995). The resulting theories, written to be accessible to the public, policy-makers, and academics, are judged on whether they fit with participants’ experience and attributed meanings, and whether the theories can be applied in the social world (Walker & Myrick, 2006; Wuest, 1995). In both feminist research and constructivist grounded theory, the aim is to transform practices and social processes and, in doing so, to contribute to a better world (Charmaz, 2006).
Of particular pertinence to the current research, Allen (2010) posited the value of a feminist perspective in concert with constructivist grounded theory methods for giving voice to women’s experience of violence. She suggested that feminism and constructivist grounded theory blend well to facilitate alternate understandings of women’s experiences. Those experiences, Allen (2010) contends, can inform professional interventions through the
development of the theoretical analysis of diverse experiences. 4.3 Retrospective Design
Given that the current research addresses disclosure of intra-familial child sexual abuse, what happens after disclosure, and how abuse ends, it is logical to employ a retrospective design. That said, the choice of a
retrospective design was guided by ethical concerns as well as an awareness of the challenges and risks concerns associated with research with sexually
powerlessness, children’s ability to provide informed consent, the dynamics and relationships between victims and abusers, the proximal consequences of abuse, and the impact of discussions of abuse with victims too soon after the abuse has ended (Kinard, 1985, 1994). The decision also was driven by
concerns about the potential reactions of the relevant Research Ethics Boards to interviewing children, and the inherent time and resource implications were there to have been barriers presented to undertaking research with children. 4.4 Sampling and Recruitment
The research design incorporates theoretical sampling which, like most qualitative methods, begins with purposive, non-probability initial sampling, involving a variety of recruitment methods and strategies. “Initial sampling provides a point of departure, not of theoretical elaboration and refinement” (Charmaz, 2006, p. 100). The constant comparative method, an evolving cyclical process of data gathering and analysis, prompts the researcher to gather more data from existing and new sources to fill gaps and saturate categories. Data gathering stops once theoretical saturation is reached. (This process is described in greater detail earlier in the chapter).
In designing and completing the current research, I strove to engage in theoretical sampling, recognizing the confines imposed on completing a PhD thesis (e.g., learning curve, time, education leave, financial resources). Two visits were planned to Saskatoon to gather data: the first in June/July when I
interviewed 10 of the 16 women; the second in September when I interviewed the remaining six of the 16 participants. I deliberately and consciously left ample time between each of the interviews to engage in memo-writing and preliminary analysis, and between the two visits to engage in data analysis, in order to assess the need for more data. In analyzing the data gathered during the first visit, patterns were beginning to emerge. The second visit provided the opportunity to interview six more women. That said, the decision to stop gathering data was based more on time-lines and the number of women who came forward to be interviewed than on actually reaching theoretical
saturation, as not all of the interviews had been analyzed when the data gathering stopped.
That said, the sample did provide a range of experiences, some of which contained aspects that were unique to individual participants and some of which overlapped with the experiences of other participants. In reviewing the extant literature and comparing it to the findings of the current research, the experiences that are lacking, or at least limited, in the current research involve positive experiences of disclosing, other’s reactions to disclosures, and ending the abuse. This lack of positive experiences may be a result of not achieving theoretical saturation or it may be an accurate reflection of the experiences of the women who met the criteria to be interviewed.
theoretical sampling methods. Very few researchers using grounded theory methods actually practice theoretical sampling. Most begin with purposive sampling; few reach theoretical saturation using the constant comparative method (Bryant & Charmaz, 2007; Charmaz, 2006; Wertz et al., 2011).
To provide clearer expectations and to facilitate completing the thesis, a decision was made during the proposal stage about the parameters of an acceptable sample size, based on achieving variability among participants plus depth, breadth, and richness of data. A potential sample size of 12 to 20 was deemed large enough to be relevant and diverse, but not overwhelming in size or scope (Carey, 2009; Guest, Bunce, & Johnson, 2006; Luborsky & Rubinstein, 1995; Thomson, 2011). Twelve to 20 participants was regarded as adequate to gather a range of perspectives and experiences by virtue of the potential participants’ varied social locations and experiences of abuse and disclosing. Upon completing 16 interviews, which involved interviewing all of the women who had volunteered for the research by the end of the second visit to
Saskatoon, it was determined that I had interviewed a sufficient number of women to fulfill the need for variability among participants plus depth, breadth, and richness of data.
In order to participate in the research, women, before reaching 18 years of age, must have: (a) experienced contact or non-contact sexual abuse by one or more family members; and (b) disclosed the abuse (indirectly or directly;
accidentally or purposefully; spontaneously or when elicited or prompted). Due to the differing experiences of sexual abuse and disclosing based on the sex and gender of the victims and their relationship to their abusers (i.e., intra- versus extra-familial abuse) (Dorais, 2002; Haugaard & Emery, 1989; Kinard, 1994; O'Leary & Barber, 2008), the scope of the sample was limited to girls sexually abused by one or more family members.
The sample was drawn from Saskatoon and surrounding area, in part because I had lived and worked there for many years and had established enduring relationships with professionals in a variety of health and human services agencies, including those addressing child sexual abuse. These connections were beneficial in recruiting participants and in ensuring that support was available to participants through local agencies, including the Saskatoon Health Region and the Saskatoon Sexual Assault and Information Centre. A further benefit of drawing the sample from Saskatoon is the region’s long history of offering support and therapy to adult female survivors of child sexual abuse. As a result, child sexual abuse is part of the community
discourse and a local community of women have addressed the abuse and undertaken a healing journey.
Participants initially were recruited through the Saskatoon Health Region, community agencies, sexual abuse programs, and personal and professional contacts. Letters to Agencies (Appendix C) and Advertisements for
Participants (Appendix B) were emailed to agencies likely to provide services to adult female survivors of child sexual abuse. The email requested that
recipients post the information and share it with colleagues and potential participants. Advertisements for Participants (Appendix B) also were posted on agency, university campus, and other public bulletin boards, as well as distributed via personal and professional contacts.
Snowball sampling, which is especially beneficial with hard to reach groups or participants who are unlikely to volunteer without a link to the research or researcher (Bryant & Charmaz, 2007; Carey, 2009), was used in the initial purposeful sampling. Women were invited to share information with others who fit the sample criteria. A number of participants invited other women to participate and some requested permission to post the information on their social media pages.
Women interested in participating in the research contacted me via email, text, or telephone with questions and to schedule an interview. Those who expressed interest in the research were screened to ensure they fit the sample criteria of being sexually abused by at least one family member and disclosing that abuse before age 18. Participants also were asked to indicate: (a) that they were willing and able to discuss their childhood experiences of intra- familial sexual abuse, focusing on disclosure, what happened after disclosure, and how the abuse ended; and (b) that they had access to formal or informal
support to deal with any surfacing issues. 4.4.1 Potential sources of sampling bias.
Selection and social competency bias may have affected who is included in the sample and who volunteered to be interviewed. Criteria requiring
participants to have access to support for surfacing issues may have eliminated potential participants or biased the sample by recruiting participants further along in their healing journey or experiencing greater stability in their lives (Kinard, 1994; Teram et al., 2005). Regardless of the potential for bias, it is unethical to increase risk to participants by interviewing women unprepared, unable, or lacking support to cope with surfacing issues (Brzuzy, Ault, & Segal, 1997; Teram et al., 2005).
Women of a particular age, class, ethnicity, or other socio-demographic identity may have been more prone to volunteer to participate in the current research, although no obvious bias is evident in reviewing the participants’ demographics. Widely circulating information about the current research and relying on personal and professional connections to encourage diverse
participation was intended to mitigate any selection bias. Advertisements for Participants (Appendix B) and Letters to Agencies (Appendix C) were widely disseminated. The information was emailed to agencies serving women with diverse cultural, religious, and sexual identities, and shared with personal and professional contacts with connections in a variety of communities.
Data gathered from retrospective self-reports may present challenges to reliability, validity, and trustworthiness. The passage of time may impede accurate and complete recall of past events or participants may be selective in responses and memories (Berliner & Conte, 1995; Dorais, 2002; London et al., 2005; Tang et al., 2008). Participants also may have been susceptible to re- authoring their subjective truths, altering their perceptions of the meanings attached to various experiences, or reinterpreting past events through present knowledge and understanding (Briere, 1992b; Dorais, 2002; Tang et al., 2008). Despite the potential limitations, retrospective designs are widely accepted in child sexual abuse research.
4.5 The Interviews
Prior to my arrival in Saskatoon, arrangements were made for
temporary office space at the University of Regina’s Faculty of Social Work, Saskatoon campus. On-site office space was also negotiated for interviewing women referred by therapists at the Saskatoon Health Region Adult
Community Mental Health Clinic.
Each in-person interview was scheduled for up to two hours, at a time and place convenient and comfortable for both the participant and the
researcher. Time was allotted between interviews to begin to analyze the data through memo writing. Interviews were structured to begin with building rapport through introductions, discussing the current research process,
reviewing the ethical guidelines, and building or renewing acquaintance. Following the grounded theory methods outlined by Charmaz (2006), an open-ended interview guide (Appendix E) loosely shaped the exploration of the participants’ experiences of intra-familial child sexual abuse, disclosing, and how the abuse ended. Included in the interview guide were six open- ended, semi-structured, focused questions inviting participants to describe their experiences of growing up in their family and community, the sexual abuse, disclosing, what happened after they disclosed, how the abuse ended, and what they thought could have been different. Participants also were invited to ask any questions and to provide clarification around anything they had discussed.