Introduction
In 2021, Croatian Sociological Review published an open call for a special issue, seeking papers that, from a qualitative perspective, examined sociological topics relevant to the context of Croatia, the Western Balkans region and, more generally, Central and Eastern Europe. The primary goal of this invitation was to expand the publishing spaces for qualitative studies, still underrepresented in Croatian sociology, but also to encourage and highlight good practices of reporting on qualitative research – primarily, theoretical-methodological transparency and reflexivity.
In our call, we recognised the diversity of approaches and understandings of qualitative methods and analyses. Still, for the sake of transparency and reflexivity, I must acknowledge that, in my role as editor-in-chief, I evaluated the suitability of submissions in light of my expectations and judgements of what makes a good qualitative paper. Accordingly, the editorial comments and suggestions reflected these expectations and judgements. Therefore, again for the sake of transparency and reflexivity, in this introduction to the special issue, I report on the main editorial requests posed to the authors – from writing about their methods to presenting their results – and reflect on why I considered such requests important. In the process, I also highlight how the articles published in this special issue are representative of good practices of transparency and reflexivity in qualitative reporting.
Why Transparency and Reflexivity?
Transparency – providing sufficient information for the reader to fully understand the research and the analytical process – is central to evaluating the validity of any research, and qualitative research in particular. Why is that the case? Because qualitative research is typically characterised by a “back-loaded” (Martin, 2017) or an emergent research design. In this type of research design, the bulk of the research and analytical decision-making happens after the initial conceptualisation, in a non-linear and non-deterministic manner, in the process of constant adaptation to the developments and growing understandings in the field (Maxwell, 2012;Rubin, 2021). Therefore, flexibility and adaptability are integral features – and strengths – of qualitative research.
This also means that qualitative research decisions can be properly evaluated only when transparently situated – and, even more importantly, justified – within the context of a particular study and a particular qualitative approach (Clarke and Braun, 2013;Small and McCrory Calarco, 2022). As Maxwell emphasises in his classic textbook Qualitative Research Design, “validity is a property of inference rather than methods, and … is also relative: It has to be assessed in relationship to the purposes and circumstances of the research, rather than being a context-independent property of methods or conclusions.” (2012: 21).
Reflexivity, similarly to transparency, also has an important role in assessing the validity of the decision-making and the conclusions of a qualitative study. Statements on the positionality of the researcher, their epistemological assumptions, relationships with the participants, and similar reflections on the role of the researcher in their study have long been standard requirements for qualitative reporting. Yet, the purpose of such statements of reflexivity is sometimes described as reporting on the sources of “bias”. Such an interpretation does a disservice to qualitative research (cf. Clarke, 2021) because it downplays “bias” in research in which one cannot clearly see the researcher or analyst. This plays into the illusion of “objectivity”, which supposedly characterises quantitative in contrast to qualitative research. But, it is a mistake to claim that qualitative research is subjective (which is often meant to say “biased”) while quantitative is not just because the qualitative researcher lays their presence bare, while quantitative researcher does not – as evidenced most recently by the open science movement, pushed into existence by the misdeeds of quantitative analysts.
Furthermore, the mistake of attributing some inherent “bias” to qualitative research is usually also caused by mixing up “subjectiveness” (i.e. making decisions and conclusions based on personal impressions, feelings and opinions) with “subjectivity”, which is a researcher's “sense of themselves” (Clarke and Braun, 2013: 337) – and “subjectivity” is then mistaken for “bias”. But, as the qualitative researcher is a research instrument in their study, their subjectivity cannot and should not be eliminated. Instead, the researcher's subjectivity should explicitly become part of the research process and of the contextualised analysis (Clarke and Braun, 2013). Indeed, Small and McCrory Calarco (2022) list “self-awareness” as one of the indicators that a qualitative study has been well executed: “self-awareness about one's identity, one's relation to others, and how both affect the actions of those interviewed and observed can lead to dramatically better data and more accurate conclusions about the social world” (121). Qualitative researcher both shaping and adapting to the needs and developing understandings of their study is one of the major strengths of qualitative research, but these shapings and adaptations cannot be unexamined or unconsidered – this is the purpose reflexivity serves in assessing the validity of qualitative reports.
Qualitative Explorations of Gendered Processes and Relationships
Croatian Sociological Review's call for submissions for the special issue was open in terms of theme, specifying only that we were looking for “sociological topics from a qualitative perspective”. Still, it is not surprising that most of the submitted papers as well as the four empirical studies brought together in this issue navigated to topics that examined processes and relations or spotlighted marginalised voices and perspectives. This is, indeed, the added value of qualitative approaches whose in-depth examinations are able to contextualise and better reflect the messiness and the complexity of social worlds and experiences and then use these analyses to go beyond the “usual suspects” of much of the mainstream approaches in many disciplinary fields.
Therefore, despite the very different topics that the first three papers in this special issue explored (migration, homelessness, social workers' handling of domestic violence cases), the studies by Poleti Ćosić, Greiner and Kujundžić shared an approach spotlighting gendered processes. In particular, their contributions explored the implications and consequences of taking a gender-neutral approach to a social problem that cannot be properly understood or managed without understanding the gendered mechanisms at work.
In the first paper, Dunja Poleti Ćosić takes advantage of a qualitative approach to study migration in its procesuality, based on biographical interviews with women migrating from Serbia to France and Germany. In this study, Poleti Ćosić shows how qualitative analysis of women's migration trajectories reveals their agency and also outlines women's migration strategies, some of which are not easily observable from the mainstream, mainly quantitative, studies. Throughout her analysis, the author also highlights how women's migration strategies are gendered and interconnected with the family sphere.
In the second paper, Paula Greiner builds on her earlier ethnographic observations with in-depth interviews with women affected by street homelessness in the Croatian capital. In this work, Greiner underlines in particular how women's experiences of homelessness differ from men's and how these differences reflect broader gender inequalities. This paper also contributes a perspective that received insufficient attention in previous studies of homelessness in Croatia. This is a perspective that is very much needed, although still conspicuously absent, in the shaping of public policy on homelessness.
In the third paper, Jana Kujundžić explores the gendered dimension of social work in Croatia in the case of domestic violence, based on expert (or elite) interviews with individuals with expertise on domestic violence in the Croatian legal and welfare system. In this study, Kujundžić illustrates the structural problems of the system in which the underfunded, understaffed and often unprotected social workers are charged with almost the sole responsibility of handling domestic violence cases without proper means to do so. Importantly, Kujundžić's analysis brings attention to the missing “proper means”, also including a deeper understanding of how social workers, although largely women, often also reinforce and normalise patriarchal attitudes towards victims of domestic violence, thus reflecting the structural problems of the larger society.
The final paper in this special issue, a study on modern dating among emerging adults in Croatia, by Lucija Šutić, Margareta Jelić and Ana Krnić, differs from others in that it does not focus on gendered experiences and is also not based on the individual but on the group method of data collection (focus groups). This study, however, spotlights another dimension that other types of research designs are less able to capture: relationships. This includes not only the identification of types of dating relationships (dating scripts) described by the participants in this study, young Croats aged 18–25, but also the role of relationships as a data collection tool. Specifically, the study authors' choice to examine young people's experiences and opinions on how they and others like them date through focus groups is valuable because the method of focus groups is crucially characterised by the role of group dynamics in expressed opinions and shared experiences (Cyr, 2019). Rather than viewing this as a limitation of a study, it is better viewed as a strength when the study's purpose, as in this case, is to identify societal scripts, which are always closely related to norms and societal expectations. In other words, the identification of the types of dating relationships described by Šutić, Jelić and Krnić is valuable precisely because the focus groups' responses on dating scripts are more than summaries of individual responses. Rather, these are the dating scripts agreed upon by groups of young people, who were bringing their expectations and perceived norms into their joint, interactional evaluation of today’s dating practices in Croatia.
Good Practices of Transparency and Reflexivity in Qualitative Reporting
While the studies described above – and others considered for this special issue, including some that will be published in later issues of Croatian Sociological Review – differed in their topics and approaches, I soon realised that I was repeating similar questions and suggestions to the authors. These mostly related to expectations concerning transparency and reflexivity when it comes to selecting and recruiting the participants, identifying ethical issues specific to the research and detailing the analytical procedures, although some comments also related to the manner of data presentation.
Some editorial comments were more negotiable than others. For example, I required information about the selection of participants, but I was only gently suggesting to the authors how the quotes might be presented differently; the final decision remained theirs. I also did not insist on all the issues on which I may have strong opinions when these opinions were not widely shared or I had just recently come to hold them so strongly. One of them is the opinion that qualitative researchers should not use the term “anonymity” since only “confidentiality” is possible in qualitative research (see the discussion on the epistemological difference between anonymity and confidentiality in chapter 5 inTolich and Tumilty, 2021) and, consequently, that qualitative researchers can talk only about de-identification and not about anonymisation. As a result, readers may find all these terms in the published texts, as well as some others that may not be congruent with my constructivist perspective on qualitative research but fit in with different epistemological positions.
Still, whether other qualitative practitioners agree with my requirements for a good qualitative study or not (and I may also shift my thinking in the future, as I work on these issues more), I believe that the questions I asked are useful to consider. For this reason, I present them in the text that follows, while also highlighting some examples from the studies published in this special issue that I see as examples of good practices.
Selection and Recruitment of Participants
Transparency and reflexivity about the selection and recruitment of participants are central to good qualitative reporting. This is so because the logic of selection and the pragmatics of recruitment are key in evaluating a study's findings and conclusions, including their trustworthiness and reach. While the selection and recruitment of participants are usually bundled together, I find it more useful to consider them as two separate processes, described and considered separately.
The first question, therefore, is about the selection of participants and/or cases. As is often repeated in the literature on sampling in qualitative research, participants are typically not meant to be representative of a population but to be “information-rich” and cases, likewise, should also be selected strategically (Patton, 1990). Therefore, it is beside the point to discuss whether a qualitative sample is non-probabilistic – this is so by design, as a qualitative sample is supposed to be purposive. Indeed, most qualitative textbooks today (see, for example,Patton, 1990 and its later editions) discuss various logics or purposes of selection, so I would also recommend using specific and appropriate terminology when discussing qualitative sampling. But, at minimum, I would ask about the logic or purpose of the selection – what is the strategic thinking behind the selection of particular participants or particular case(s)? Furthermore, what are the implications of this logic of selection for making conclusions based on these participants and case(s)? In particular, can you make an argument for theoretical generalisibility or transferability or some other similar relevant concept (for a discussion of different conceptualizations of generalization in qualitative research, seeGobo, 2008)?
While the question of selection is the question of logic or purpose, the question of recruitment is often a question of feasibility. Once the researcher has decided on what type of participants they need for their study, how did they reach them and how did they motivate them to participate? What is the role of the researcher in the process, both for gaining access or trust and for (not) reaching particular participants? For instance, the questions such as self(selection) of participants, the researcher's positioning and presentation to potential participants, or power relations in the field, are all possibly relevant here, as they might have implications for the reach and depth of the study's findings and conclusions.
Poleti Ćosić's paper on migration among Serbian women is a good example of how to transparently present the logic of case selection (France and Germany as destination countries for Serbian migrants) and justify the selection of a particular group of participants. In the latter case, Poleti Ćosić not only outlines two eligibility criteria required for participation in the study, but also justifies the selection of these criteria and then goes on to explain other criteria she was considering in forming the socio-demographic structure of the sample. However, going into the field and recruiting desired participants can be challenging and sometimes these challenges provide further insight into the social problem under investigation. This, indeed, is what happened in Poleti Ćosić's case. Poleti Ćosić transparently reports on the issues she encountered during recruitment and thoughtfully considers the reasons behind them – and how these developments led her to change the structure of her sample. Finally, Poleti Ćosić also reflects on her presentation to the participants and how this might have influenced both their agreement to participate and their rapport in the interviews. All this contextual information adds both richness and a better understanding of the data Poleti Ćosić collected and analysed, thus demonstrating the benefits of not trying to bracket the researcher's subjectivity, but making it a part of the interpretative framework of the study.
Ethical Issues
Since 2021, Croatian Sociological Review has required authors to submit an ethical approval statement when submitting their manuscript – this then becomes part of the published text, together with a Declaration of conflicting interests, Funding statement and Data access and transparency statement. This practice has revealed two things. First, the requirements for having a study with participants institutionally approved by an ethics committee differ widely among (and even within) institutions and countries in the region. Second, in the absence of the required institutional ethical approval, authors sometimes assume that the issue does not apply to their study. However, institutional approval signals solely that minimum ethical requirements have been considered; by itself, it is not sufficient for an evaluation of the study's ethical practices, regardless of whether the institutional evaluation was required or not. Transparency and reflexivity in reporting on ethical issues require an awareness of ethical issues specific to the study's research design and identification of its specific management strategies and solutions.
The difference between considering minimum ethical requirements and situating the study's ethical issues in relation to its specific design is best demonstrated in the discussions of confidentiality, which is often the only ethical aspect reported by many authors in the initial versions of their papers. This discussion is typically limited to reports on having promised the participants confidentiality (sometimes presented as “anonymity”) and using codes or pseudonyms in the presentation of findings. However, some qualitative research designs can raise additional issues that limit confidentiality, including, in particular, the issue of deductive disclosure or “internal confidentiality” (Tolich and Tumilty, 2021). In such cases, for example, members of participants' communities or even their acquaintances can guess participants' identities based on specific additional details presented in the text.
However, while detail sometimes has a very important place in a qualitative text (see the discussion of “palpability” later), oftentimes it also happens that some specific additional information does not have any heuristic or methodological added value, but compromises participants' confidentiality. For example, there is no reason to keep the name of a specific shelter for the homeless, even if mentioned by the participant in the included quote, if naming is not necessary to understand the described experiences – and naming it increases the chances that a person working in that shelter would recognise the participant. Likewise, is it really important to say that a participant was 27 years old (instead of, say, using aggregate information such as “in their twenties”) and “a nurse” if such specifics are not relevant for understanding the data and the patterns?
In light of these considerations, I generally discourage the practice of presenting all the participants, their socio-demographics and similar information in the tables. Relevant information is frequently easily incorporated in an aggregate form in the body of the text and/or selectively presented in relation to particular quotes or vignettes, without making all the specifics and their connections directly related to each participant, thus increasing their recognisability. Therefore, my recommendation to the authors in such cases is to privilege better protection of participants' confidentiality. I would even go as far as to suggest misrepresenting some unimportant details in order to confuse potential recognitions (but also transparently admitting to this strategy for protecting confidentiality).
There are also other ethical issues to consider in qualitative reporting. Some intertwine with the issue of reflexivity. For example, I consider the questions of access and establishing the relationship of trust and rapport with the participants, especially in the circumstances of a power differential, to be quintessential ethical questions. But, other than that, it is very difficult to give general advice regarding ethical issues since they ultimately depend on the specifics of each research. So, the best I can do here is to offer the following questions to the authors: What are the ethical issues that are specific to your study? Concerning this specifically, what did you do to protect your participants?
Kujundžić addresses these questions well in her study of domestic violence cases in the social welfare system. Kujundžić positions herself as a feminist and LGBTIQ activist, but she also considers how her positionality necessitated a strategic presentation of her research to some participants and a careful negotiation of what she can and cannot say of her own political and personal worldviews so as not to provoke a priori antagonistic reactions. Likewise, Kujundžić considers the different relationships she established with different types of participants and she reflects on diverse power imbalances in such situations. Finally, Kujundžić also reports on entering the field with a plan of action in case the difficult topic of her research causes distressing emotions among her participants. These reports and reflections signal an ethical and thoughtful researcher who understands the role her self-awareness must play in collecting, analysing and understanding the data.
Analytical Procedures
A transparent description of analytical procedures provides readers with a clear sense of the process leading from raw data to the patterns that form the findings of a qualitative study. In reporting on this process, I encourage authors to take responsibility for their analytical decisions. In the simplest form, this sometimes means not saying “thematic analysis was conducted”, but identifying who conducted the analysis (thematic or any other type), and what that type of analysis consisted of, specifically. For example, how did the analyst(s) identify the patterns (inductively, deductively, or by a combination thereof)? Did the analyst(s) consider how the specifics of their sample related to the patterns they identified and how did they do that? How did the analyst(s) check for the validity of their conclusions? For example, did they look for negative cases or discrepant evidence or did they use some other validity checks (see, for example, inMaxwell, 2012), if these were appropriate for their study (for a word of caution on using validity checks mechanically, seeVarpio et al., 2017)?
While I understand that many of these demands conflict with the word limits, the authors can take advantage of Croatian Sociological Review's data and analytical material sharing policies to present additional detail on their analytical procedures This is exactly what Šutić, Jelić and Krnić did, thus becoming the first authors in the history of the journal who shared some of their qualitative analytical materials when initially submitting their manuscript. These materials (Focus groups' discussion guide and the Coding matrix) were made available to the reviewers, and they are also now published as an Online Supplement to Šutić, Jelić and Krnić's article. Furthermore, in their text, Šutić, Jelić and Krnić identify they are operating within the post-positivist paradigm and they report on the analytical procedures that are considered good practices within such an approach to qualitative analysis: using independent coders as a check on the validity of the identified categories and themes, and a consensual construction of the coding matrix.
Poleti Ćosić provides another good example of transparent reporting on analytical procedures within a different epistemological paradigm. Situating her analysis within a constructivist framework, Poleti Ćosić justifies the appropriateness of collecting data through problem-oriented interviews and details the inductive process of identifying themes based on two specific types of coding. Poleti Ćosić also explains why she did not use another coder in this process as a validity check on the identified patterns, although it is worth highlighting here that, within a constructivist paradigm, using several coders in the analysis is not considered necessary, nor is it always appropriate as an indicator of the validity of a study's conclusions (seeBraun and Clarke, 2022). This little note of defensiveness regardless, Poleti Ćosić's description of analytical procedures provides sufficient detail and justification of her analytical decisions for the reader to build confidence in her conclusions.
Conclusion: Adding Palpability to Transparency and Reflexivity
Qualitative research is non-linear by design, often requiring adaptability in the field and back-and-forward movement in rethinking the design and collecting and analysing the data (Maxwell, 2012;Clarke and Braun, 2013;Rubin, 2021). Such a design is incompatible with too rigid and pre-determined research plans that treat uncertainty and flexibility as threats to “reliable” knowledge rather than as the first steps into a better and deeper understanding of social phenomena. Likewise, such a non-linear, emergent research design also means that every analytical step must be situated into its specific context and should be justified and evaluated accordingly (Maxwell, 2012;Small and McCrory Calarco, 2022). This makes universal criteria for evaluating qualitative research difficult to construct, as seen in responses Small and McCrory Calarco received when asking a great variety of researchers and evaluators what criteria they would use to distinguish between poor and good qualitative research: “just about everyone who answered our question expressed some uncertainty, and the single most common answer was some version of the phrase, ‘I'm not sure.’” (2022: x).
Still, qualitative research is regularly evaluated, and even though some of these evaluations rely on criteria that are sometimes mechanically taken over from quantitative research or that reflect only one qualitative research paradigm (e.g. see criticism byClarke, 2021), qualitative practitioners have constructed a variety of standards they use to judge qualitative studies within their epistemological framework (e.g.Seale, 1999;Maxwell, 2012;Small and McCrory Calarco, 2022). In my view, most of these standards rely on the general principles of transparency and reflexivity that could also be described as whatever you do, document it painstakingly, present it clearly and justify it appropriately. These criteria were therefore reflected in the vision I had for this special issue and in my evaluations and comments described above.
However, I would like to end the introduction to this special issue with a note on a more specific criterion for qualitative studies Small and McCrory Calarco recently introduced, in their excellent book Qualitative Literacy, as “palpability”, or “the extent to which the reported findings are presented concretely rather than abstractly” (2022: 80). In their view, a good “palpable” qualitative report uses rich data (specific detailed explanations, concrete contextualised interactions or examples…) rather than (authors' or participants') generalisations as evidence for their claims.
While at the point of a final editorial evaluation, it is much too late to ask the authors to present their evidence “more palpably”, I did sometimes ask the authors to “flesh out” their participants more and/or to contextualise their words and experiences. For example, if the participants were referred to by codes (such as M13 or F01), I would suggest to the authors to use pseudonyms instead so that the participants do not seem like variables, but more like the real people they were. Likewise, I sometimes asked the authors to attempt to provide a deeper sense of who the participants were and why they were saying what they were saying. I also occasionally suggested linking authors' interpretations more directly with quotes, so that the quotes do not feel just as general examples of authors' arguments – interchangeable with one another and possibly presented as big chunks of skippable text – but that quotes, instead, constituted part of the argument itself.
I expect that the authors were frequently frustrated with such comments. They hardly seem actionable, and I also often felt as if I were pushing my personal preference on the presentation of the data without articulating clearly what I meant by “fleshing out” or what precisely the authors needed to do when presenting the participants and their words or actions. The concept of “palpability” now finally gives me the conceptual tools and vocabulary to express that what I wanted to see was data being used as evidence instead of as an illustration. While “palpability”, of course, is a much more complex endeavour than just presenting participants with pseudonyms and more context, the latter is, in my view, at least a small step towards taking advantage of the depth that qualitative data can provide in understanding the specifics of our participants' social worlds, experiences and perspectives. In light of this, I am concluding this introduction on good practices of reporting on qualitative research with a plea for three principles of qualitative reporting in practice: transparency, reflexivity and palpability.