Guest Editorial: A qualitative quandary: alternative representations and meta‐synthesis

Over the last decade or so, there have been two potentially conflicting trends having an impact upon the reporting of qualitative research for application to clinical nursing. These trends can cause a quandary for researchers. The first trend is still emerging: postmodernist encouragement for alternative representations of qualitative research results (Richardson 2000) and the increasing use of alternative representations like fictionalized stories (Gilgun 2004), poetry (Brady 2004), visual art (Sava & Nuutinen 2003), or performance such as dance (Bagley & Cancienne 2001) or a play (Ajwang & Edmondson 2003). Numerous researchers are experimenting with how best to represent their research product (Gilgun 2005). The second trend, growing in strength, is the notion of evidence-based practice for which the results of similarly focused studies are systematically reviewed via ‘meta-analysis’ (Hawker et al. 2002). However, for qualitative research studies, usually it is a ‘meta-synthesis’ that is conducted. This is more than, and differs from, a systematic review as the process produces a ‘new and integrative interpretation of findings that is more substantive than those resulting from individual investigations’ (Fingeld 2003, p. 894). An example is a meta-synthesis of the hidden suffering among breast cancer victims conducted by Arman and Rehnsfeldt (2003). Meta-synthesis seeks to identify key findings of similarly focused qualitative studies, then compare, contrast and collate identifiable common themes and subthemes to synthesize overall a description of the phenomenon (Evans 2003). Attention is also given to rigour/trustworthiness and an aim is to review data ‘generated from multiple populations, settings and circumstances’ (Evans 2003, p. 26). However, a dilemma for a researcher arises when it is hoped that the results will inform clinical practice but it is believed that, optimally, the result of their qualitative research should be constructed as an alternative representation. Providing an alternative representation where the research-derived themes, categories of concepts, or theory components directed the content (and possibly the style) but are not explicitly presented, will render the results ineligible for meta-synthesis. So, what are the choices and associated issues? A solution is to present, simultaneously, the elements of the results and the alternative representation. Therefore, elements for meta-synthesis are accessible and also a vicariously evocative representation is available for clinicians, but there is a constraint. If intending for the results to be published in a peer-reviewed journal, word-limits (the bane anyhow of researchers seeking to publish an adequate report of a qualitative study) can restrict acceptability of this option. A poem might be fitted into the manuscript submitted (e.g. Poindexter 2002), but even a ‘short’ story can flip content well over the word-limit, and this option is not viable for multiple stories, a novel or other forms of artistic representation. However, the benefit of a peer-reviewed journal is that electronic databases record the publication; the study can be easily found with the elements of results and explication of method accessible for meta-synthesis. Other examples of simultaneous presentation of the two forms of results are: a non-published, comprehensive research report (which may be a thesis); a monograph, the return of which is championed by Morse (2003); a novel, a tome of poetry, or a book of short stories that present method and elements of results in a concluding (or introductory) chapter (e.g. Koch et al. 1999a,b); or a song, dance or play either printed or performed with accompanying notes about the research method and elements of results from which the work is derived. When planning a meta-synthesis, the protocol lists how appropriate studies will be sought and found. Although some of these simultaneous presentation options may be readily identifiable and accessible as ‘grey literature’ (Conn et al. 2003), in this era, many will remain hidden except to the audience for which they are initially targeted, who could be patients, local nurses and health professionals, or the public. Another solution is to disseminate the two forms of results separately. The researcher may seek to publish a traditional representation of results embedded in a peer-reviewed journal article but also seek a suitable forum, stage or mode of dissemination for the alternative representation, with or without mutual reference to the other report or presentation. This could, possibly, involve two published articles, although needing to guard against ‘salami slicing’– multiple publication of results from one study that fracture gained knowledge and offer only partial understanding of the phenomenon researched. For the purpose of facilitating evidence-based practice, this solution is worthy as the traditional representation is likely to be accessible for meta-synthesis. There may be reservations by some researchers that within this solution the alternative representation, unless also appearing in a peer-reviewed journal, may be less disseminated when it is actually this representation that may have the greatest power directly to influence clinical practice via the reader, perhaps ‘in order to make others feel… the fundamental mover’ (Harrington 2003, p. 103) and to provide ‘grab’ (Gilgun 2005). This latter point is linked to why some qualitative researchers may find neither of the two posited solutions acceptable. Just as many qualitative researchers only seek to present results in a traditional form, a researcher may only want to present the alternative representation believing that the alternative representation is the ideal (or only) form within which the fullness or extent of the results can be embedded for the gaining of necessary understanding by a clinician. Presenting elements of results may be viewed as destructively splintering an ‘essence’ or ‘core category’. Otherwise, the researcher may view qualitative research as offering localized, contextual knowledge or understandings that primarily benefit clinical practice choices in that localized context; hence rejecting the ideal of meta-synthesis for evidence-based practice in a wider area or scope of practice. Therefore, representation style would be influenced by other criteria, which could vary between: comprehensive reporting for evaluation of trustworthiness/rigour; maximization of an evocative, vicarious response; acceptability to, and accessibility by, the target audience; or combinations thereof. Additionally, there may be cynicism regarding the ability of meta-synthesis to appraise and synthesize successfully results of qualitative research for worthy recommendations regarding clinical practice. This stance would also reject the need of a dissemination form and method that caters for meta-synthesis. Therefore, it will be interesting to discern in the next decade or so whether other ways of gathering and comparing qualitative research findings, like conducting a ‘qualitative meta-summary’ about a target domain as suggested by Sandelowski and Barroso (2003), will include and enable alternative forms of representation to inform evidence-based practice. A possibility is that qualitative researchers familiar not only with their own work but also of each other's research in a substantive area may jointly meta-sythesize the understandings offered collectively, as exampled by McCormick et al. (2003), and that this might incorporate alternative representations in the synthesis. In the meantime, the contemporary qualitative researcher is wise to ponder carefully how to optimally present and disseminate results with consideration of desired audience and how best to inform clinical practice. I recommend that this be deliberated in context to personally held beliefs about the philosophy of inquiry, how or whether to ensure trustworthiness/rigour, and the perceived usefulness of meta-synthesis.