A qualitative grounded theory study design was chosen for the research project because it was considered most appropriate for exploring the research question, “How do paediatric osteopaths promote effective breastfeeding in mother and baby dyads with breastfeeding difficulties?” It also met the aim of developing a substantive theory to explain the processes involved; however, like all research, it has particular strengths and weaknesses. Acknowledging and uncovering these aspects of the study is important when evaluating results and their contribution to knowledge.
14.2.1 Strengths of the study
In the first instance, attention to aligning and making explicit the study’s constructionist epistemology, theoretical perspective of interpretivism and symbolic interactionism, and Corbin and Strauss’s (2008) version of grounded theory methodology and methods is strength of the study. From this point, determining a grounded theory study’s strengths tend to relate to issues concerning its overall ‘quality’. Quality relates to issues of academic rigour and trustworthiness, which have been examined in Chapter Four214. According to Charmaz (2006, p. 18), quality starts with the depth and scope of the data, acquired by entering research participants’ worlds and understanding their lives and perspectives. This study generated rich data with a range of views and experiences, drawn from clinical observations and interviews with two specific participant groups and sampling continued until saturation. From a clinical perspective, a more pragmatic approach to ascertaining quality in a grounded theory study arises as a result of the need to deliver effective and evidence-informed patient care. In this situation, quality relates to a consideration of how the research methods were used to produce findings that give a meaningful and accurate account of the phenomenon of interest, in the clinical setting.
One of the proposed strengths of a grounded theory study is that it provides a package of research methods that build quality into the research process (Elliott & Lazenbatt, 2005).
Implicit in this idea is the notion of methodological consistency (Morse et al., 2002). It is
214 See Chapter 4.9.
273
for this reason that the researcher faithfully employed the techniques and procedures of Corbin and Strauss’s (2008) grounded theory methods and provided a robust audit trail to enable the reader to follow the analytic process in order to have confidence in the results.
Attention was paid to visibility of data collection and analysis procedures. As analysis proceeded, theoretical sampling and memo writing became particularly important to add conceptual density to categories, test ideas, and gain insights into some of the more tacit processes of clinical practice; processes that paediatric osteopaths use but which can be difficult to articulate.
Overall, the researcher spent 12 months in the field gathering and analysing data and interacting with it over an extended period of time (four years) to develop categories with high levels of abstraction and dimensional range. Regular meetings with research supervisors and presentation of research material to colleagues provided opportunity for discussion, critique, and reflection, all of which assisted with auditing the research process. A common idea concerning the quality of a grounded theory study is that it is best judged by those who read and respond to it; by following the researcher’s data and analysis to make their own conclusions (Chiovitti & Piran, 2003; Cooney, 2011). When attempting to define a ‘quality study’, Corbin (2008, p. 302) places more emphasis on how the methods are used as tools to produce research that has depth, is creative, stimulating, and interesting. Overall a plausible and coherent theory of paediatric osteopathic practice in the situation of assisting dyads with breastfeeding difficulties was generated. The final substantive theory, however, has a number of limitations.
14.2.2 Limitations of the study
The first limitation relates to the general design of a qualitative study in that the results arise from a prescribed context; in this instance, paediatric osteopathic practice in metropolitan Melbourne and in the situation of treating dyads with breastfeeding difficulties. Findings pertain to the experiences, views, and actions of paediatric osteopaths and mothers and babies, which are interpreted by the researcher whose assumptions and expectations will have some influence on the final analysis215. The value of a substantive grounded theory is its ability to address specific issues relating to the clinical setting from which data was derived and, therefore, it cannot be assumed to be generalised (Petty, Thomson, & Stew, 2012). There is, however, some contention regarding the notion of generalisability, which traditionally represents a canon of quantitative research (Snowden & Martin, 2010). Rather than generalisability, findings from qualitative studies are discussed as being potentially transferable to other similar
215 See Chapter 4.4 for an account of the researcher’s assumptions and expectations.
274
situations and are evaluated in terms of trustworthiness216. They are also recognised and valued more for the purpose of gaining insight into clinical practice, rather than description and measurement (Morse & Field, 1995). This is particularly relevant when very limited literature is available in the area of interest; as is the case in the current study. Strauss and Corbin (1998) contend, however, that raising the level of abstraction of a grounded theory study through systematic and widespread theoretical sampling will elevate its generalisability in terms of explanatory power and precision, and that it can be generalisable in the sense that it might have meaning for other populations in similar circumstances and thus have “predictive ability” (p. 267).
The second limitation relates to the two participant groups, osteopaths and dyads, and some restrictions posed by the inclusion criteria. While both groups demonstrated variety, they tended to be homogenous in some respects. Mother participants shared a common socio-demographic and educational background217 and a trend toward older maternal age.
This means that the breastfeeding experiences of younger women and women from lower socio-economic or non-English speaking backgrounds were not represented. To avoid any sense of pressure on new mothers, who are generally recognised as a vulnerable social group, recruitment occurred primarily through indirect means by displaying Research Information Pamphlets218 in osteopaths’ clinics. This might have meant that only highly motivated breastfeeding mothers, interested in the research topic, chose to participate. It is also likely that this group of mothers might be particularly receptive to osteopathy as a new treatment approach. Additionally, mothers generally reported positive experiences of osteopathic treatment. It is possible that mothers who had negative experiences of osteopathy might have chosen not to return for further treatment and were thus less likely to participate in the study, and their accounts were missing from the data. Overall, limitations concerning diversity of mother participants on the basis of age, culture, education, and social circumstance could have been addressed by using more expansive recruitment methods. For example, this might involve providing study information pamphlets to women who attend larger public health care institutions for pre- natal and obstetric care or specialty outpatient breastfeeding clinics. Reducing the financial costs of osteopathic treatment for some families might be another strategy to encourage participation of a wider social group of breastfeeding women. In a similar way, expanding participation criteria to include not only healthy term babies, but also premature babies or babies with special needs could add another perspective to the data. Osteopath participants were informally classified by their peers as paediatric practitioners because of
216 See Chapter 4.9 for a detailed account of the concept of Trustworthiness.
217 See Chapter 5.2.
218 See Appendix F.
275
their interest and experience in this patient-base. Hence findings cannot be generalised to osteopathic practice as a whole. While these limitations might have narrowed the breadth of human experiences and interaction associated with the research topic, nevertheless they tend to more accurately reflect the everyday circumstances of paediatric osteopathic practice in the prescribed situation, and did not interfere with the collection of rich and diverse data.
The final limitation relates to practical issues and time constraints associated with completing a doctoral thesis. In particular, concurrent data collection and analysis took place over a 12-month timeframe and involved time spent travelling to participants’ homes for interviews and to 12 different geographically located clinics to deliver documents, gain access, and conduct observations and interviews. Recruitment of dyads took time and osteopaths had busy treatment schedules. One interview was conducted with each participant and while follow-up interviews might have added a longitudinal perspective, they were not conducted for two reasons. First, the limitations of time and second, rich sources of diverse data were made available from observations and interviews such that data saturation was achieved and categories were able to be conceptually well developed.
The data did not indicate that a series of observations or interviews was needed.
However, in a larger scale study, collecting data at the beginning and end of a course of treatment might be a worthwhile consideration.