Quality in qualitative research: a relational process Open Access

Guba and Lincoln (1994) emphasize the primacy of paradigms. Qualitative methods can be utilized within different paradigms, and the chosen paradigm determines the approach to qualitative research: “Questions of method are secondary to questions of paradigm, which we define as the … worldview that guides the investigator … in ontologically and epistemologically fundamental ways” (Guba and Lincoln, 1994, p. 105). As this article is not a detailed guide to different ontologies within the qualitative research family, a useful simplification here is that if your ontology is (post)positivist, your quality criteria will bear greater resemblance to quality criteria of quantitative research. If your ontology is interpretivist or constructivist (i.e. based on the assumption that realities are subjective, multiple and socially constructed), your quality criteria will place more emphasis on reflexivity, authenticity and resonance. The critical realist position occupies the middle ground, asserting that there is one reality, but people have different perspectives on it which, in turn, shape reality (Archer, 1998). Although often seen as an ontological position underpinning mixed-methods research (Clark et al., 2007; Zachariadis et al., 2013), critical realism also guides stand-alone qualitative research (e.g. Fletcher, 2017) especially among researchers who seek practical relevance.

Critical, post-structuralist and post-modern perspectives challenge the idea that reality is shaped by social, political, cultural, economic, ethnic and gender-based forces into objective social structures. Instead, agents’ own subjective knowledge and understandings are privileged as capable of revealing socially and historically constituted values and power relations as well as other forms of being and doing. In critical approaches, researcher-participant relations including reciprocity are crucial and value-driven research for transformative effect is the goal (Spencer et al., 2020). Post-humanist perspectives reject humanist assumptions of an independent, disembodied and autonomous human subject as the basis for acting and knowing, positing instead human existence as part of a complex and adaptive co-existence with other-than-human entities including animals, plants and things that have coevolved with and shaped the human species. This approach challenges how we think about the human, drawing attention to the ethical consequences of our thinking (Wolfe, 2009). While relationality is centre stage in these critical approaches, our understanding of relationality encompasses all qualitative research paradigms, as explained further in the text.

Different stances on quality align with different paradigms and sometimes a particular stance on quality can be misaligned with a paradigm. Take for example triangulation (Moon et al., 2019) – the idea that two (or more) sources of information, types of data or interpretation, when combined or integrated, can yield a truer picture of a phenomenon of interest. Whether triangulation makes sense or not depends on your stance regarding (the existence and nature of) reality. From those perspectives where a singular social reality does not exist, the purpose of triangulation (drawing an accurate “map” of reality) is not meaningful. In this case, researchers can seek to arrive at a consensus over the meaning of data, but they will not approach this task as an attempt at unearthing the reality of the situation. A framework referred to as crystallization involves combining multiple research genres and forms of analysis that draw attention to researchers’ own positionality and how meaning is constructed (Ellingson, 2014). This boundary-spanning process allows divergent and multi-genre forms of knowledge to co-exist in support of interpretive and critical perspectives.

While triangulation therefore is not well aligned with constructivism and critical approaches, another widely mentioned “quality control” mechanism – member checking – is. Member checking refers to the practice of returning to one or more participants to check whether the interpretation of data by the researcher(s) holds resonance or rings true to participant(s) (Thomas, 2017). This makes sense for those who understand research findings as co-constructed in the interaction between researcher(s) and participant(s) in the research and for those who privilege situated and subjective perspectives. However, from a (post)positivist stance, this can be problematic: it is possible that the subjects of research cannot see the contours of “the map” (reality) clearly as they are too immersed in their subjective experience. In the (post)positivist perspective, it is the researcher, equipped with the tools of research (theories, concepts, measurements and so on) who is better positioned to see “the shape of things” and hence consulting participants about the veracity of findings might risk compromising rigor. Respondent validation is a term preferred by realists (see the discussion of Mays and Pope, 2000), where the aim is establishing alignment between the researcher’s analysis and the researched subjects’ situation.

Because paradigms act as guides to different quality criteria (Guba and Lincoln, 1994), it is not surprising that terminologies denoting elements of research quality vary. Realist quality criteria incorporate various indicators of rigor: internal validity (findings reflect reality); external validity (generalizability); reliability (stability of observations) and objectivity (researcher as distanced observer) (Healy and Perry, 2000). In constructivist quality criteria, trustworthiness parallels the realist notion of internal validity; transferability parallels the realist notion of external validity (Lincoln and Guba, 1985). Interpretivism, constructivism and post-structuralist approaches strive for dependability (instead of using the realist term reliability) and confirmability (instead of objectivity). They seek to achieve authenticity by conveying close understanding of the research topic in its context (Lincoln and Guba, 1985). While it is possible to identify such broadly corresponding terms, paradigms approach quality criteria from different ontological viewpoints.

When we have our bearings regarding different ontological approaches to quality, we can better understand the proponents of varying criteria. One widely used source is the Mays and Pope (2000) article succinctly titled “Assessing quality in qualitative research”. Mays and Pope (2000) have a health services research background (implicitly) grounded in (post)positivist or realist ontology. In keeping with this, the two key quality criteria that they focus on are validity and relevance. With reference to validity, they stress triangulation (comprehensive and convergent analysis), respondent validation and negative case analysis. These are strongly anchored in (post)positivist epistemology: the assumption that we can observe reality and draw an accurate “map” of it if we carefully select our instruments and use them with precision to examine reality from multiple perspectives. Reflexivity and clear, systematic reporting are also included as components of validity by Mays and Pope (2000) in their strong orientation to what they call avoidance of bias.

The starting point for reflections on relevance for Mays and Pope (2000) evidently was generalizability, which they translate into transferability in the case of qualitative research. In the quest for relevant (maximally generalizable, optimally transferable) research findings, they argue that “probability sampling … can have its place” in qualitative research (Mays and Pope, 2000, p. 52); this harks to the idea that reality is out there, and researchers seek to map it by applying methods that originate in quantitative research where random (and ideally, representative) samples are drawn with the view to generalizability. Mays and Pope (2000) list seven considerations intended to encapsulate the quality of qualitative research. However, arguably five of these are universalistic criteria that – with some minor modifications – are also applicable to quantitative research (value and relevance; clarity of research question; match between research question(s) and method(s); varied sample; systematic and replicable data collection and analysis procedures) and only two are of distinctive relevance to qualitative research: contextualization and reflexivity. Furthermore, even the contextualization requirement in Mays and Pope (2000) links to generalizability: contextualization is about sufficient clarity “so that the reader could relate the findings to other settings” (p. 52).

In contrast to the positivist-realist criteria of Mays and Pope (2000), Charmaz (2014), specifies the criteria of credibility (data enable incisive analysis), originality, resonance (concepts provide insight) and usefulness (revealing pervasive processes). Charmaz emphasizes reflexivity understood as transparency in documenting and analysing the researcher’s thought processes and biases. Following a review of different Grounded Theory schools’ quality criteria, Charmaz and Thornberg (2021) set out their broad checklist that incorporates: striving for methodological self-consciousness (i.e. how the ontological and epistemological position of the researcher fits with the approach taken, and how the methods fit with the research objectives and questions); having a critical understanding of the literature; gathering rich data sufficient to make meaningful comparisons between data; remaining open to the data and moving back and forth between data, asking progressively focused questions to expand key analytic concepts in the data; deciphering connections between key concepts and categories; providing a transparent record and justification for sampling, data collection and analysis; looking for as many theoretical connections as possible between data and comparing data with the literature and other sources to examine fit and how the findings extend or alter understanding of the phenomenon being investigated.

The proliferation and divergence of quality criteria by paradigm has not stopped the search for shared (cross-paradigm) criteria; we will next summarize and analyse two such criteria.

As we seek to establish and justify a shared approach to quality, we refer to existing cross-paradigm criteria. Our purpose here is threefold. First, we wish to separate the idea of necessary criteria from the idea of sufficient criteria and demonstrate that attention hitherto has been focused on the former, leading to a situation where the quality criteria might well be necessary but not sufficient. Related to this, we argue that universal criteria (shared by all research traditions across sciences and humanities) are not sufficient when seeking the crux of quality in qualitative research. Second, we argue that the multiplicity of criteria in the cross-paradigm definitions of quality does not adhere to the principle of parsimony (Long and Godfrey, 2004) which is necessary for promulgating pragmatic approaches to research quality. Third, we distil two criteria that are necessary and sufficient and parsimonious and shared across the different qualitative research paradigms.

A consensus process among qualitative researchers in Germany (Strübing et al., 2018) yielded a summary of quality standards published in the Zeitschrift für Soziologie. The first of these is adequacy (Gegenstandsangemessenheit) which refers to the fit between method and question. Flexibility and reflexivity are required when looking for this fit and “taking the empirical field seriously while maintaining distance from it by generating tension through theoretical reasoning” (p. 83). We argue that adequacy is not so much a hallmark of quality but rather its fundamental prerequisite. In other words, adequacy is a necessary basic requirement for all research, but not a sufficient sign of quality in qualitative research specifically.

The second and third criteria as outlined by Strübing et al. (2018) – empirical saturation and theoretical pervasiveness – appear to be strongly influenced by Grounded Theory principles. The emphasis on “the quality of the theoretical relations in which the research is articulated and, consequently, stimulated” (p. 83) and the absence of a strict separation between sampling and analysis, the primacy of data and connection to data are strongly reminiscent of Grounded Theory. The fourth criteria, textual performance or “strong authorship” – are arguably something that all researchers seek. While the challenges might differ – quantitative reporting must be on point and economical – all forms of (social) scientific writing demand strong writing skills and hence it is hard to see “strong authorship” as a special preserve of qualitative research.

While the meaning of originality is doubtless somewhat different for qualitative researchers, we have not yet encountered a quantitative researcher (of the ambitious kind) who would not want to produce original research, meaning that it is also hard to see originality as the special preserve of qualitative researchers. Quantitative researchers can also employ inductive analytical strategies such as cluster analysis and multiple correspondence analysis in search of original patterns for further, open-ended exploration. In short, there is not much to distinguish criteria for good quality qualitative research in Strübing et al.’ (2018) distillation; the criteria reflect necessary but not sufficient hallmarks of quality, and some lean strongly towards a specific qualitative research tradition (Grounded Theory).

In another synthesis of cross-paradigm (“big tent”) quality criteria, Tracy (2010) proposes eight key markers of quality: (1) worthy topic, (2) rich rigor, (3) sincerity, (4) credibility, (5) resonance, (6) significant contribution, (7) ethics and (8) meaningful coherence. When each marker with associated, more detailed bullet points is set out (Tracy, 2010, p. 840), these number 30 in total. This is not quite as detailed as the checklists that Tracy critiques in the article, but it is far from parsimonious. Further, several of these 30 items are also relevant for quantitative research (in fact, virtually all empirical research), for example the requirement to use appropriate theoretical constructs and adequate samples. Moreover, some of the detailed criteria are arguably highly subjective or at least subject to debate, such as the requirement that the topic should be “interesting”. Although intuitively appealing, some of the suggestions in Tracy (2010) are cryptic for a novice researcher who might wonder what exactly is meant by statements such as “a researcher with a head full of theories, and a case full of abundant data, is best prepared to see nuance and complexity” (p. 841).

We argue that quality criteria must be more parsimonious than the various expansive sets of criteria while also avoiding lapsing into formulaic checklists. In seeking to elucidate widely acceptable, sufficient (beyond necessary and basic) criteria that are specific to qualitative research, we will next explain what we mean by quality in qualitative research and the criteria that must be met to ensure quality. To accomplish this task, we address the fundamental issues of what is quality and how do I ensure quality in my research? We propose a conceptualization of quality that is new in its orientation to quality as a process that puts the researcher centre stage and calls for commitment to constructing and maintaining relations that are constitutive of quality. As outlined in the next section, we emphasize the applicability and utility of relationality across paradigms.

There is no single, widely-used definition of relationality (Cruz, 2024); hence we briefly summarise some uses of the term and distil common ground across perspectives on relationality.

Second wave feminist theory departed from the rational economic actor paradigm as the assumed form of relationality and depicted an alternative way of being in the world as a relational being (Gilligan, 1982). Positionalities, activities and discourses are seen through a relational modality in constructivist feminist and critical theorist research traditions; “feminist-infused participatory and action research is … built on the development of relationships” (Lykes and Coquillon, 2007, p. 315). Particularly strong examples of relationality can be found in Indigenous research (Bolton et al., 2023; Mbah et al., 2024) where relationalities of the self are embedded within networks of interdependent relationships knitted together through mutual obligation, common element being relational connections that are prioritised over the self. These approaches in turn can be seen as resonating with sociological theories that are deeply relational, such as Latour’s Actor Network Theory (Latour, 2005; see also Cruz, 2024), as well as new materialist and post-humanist ideation (e.g. Braidotti, 2013) that extend the scope of relationality to the more-than-human (Clark, 2023; Hamm et al., 2023).

While social scientists therefore associate relationality (correctly) with constructivist, critical, feminist and Indigenous perspectives, we consider it important to note that relationality features in many disciplines and fields of research. For example, we find cognitive science approaches illuminating, not in the least because they chime with the abovementioned research paradigms by connecting relationality to the higher cognitive processes of reasoning and categorisation, symbolic processes and the creation of novelty (Halford et al., 2010). Such understandings of relationality as structuring of meanings have significant parallels with constructivist, critical, feminist and Indigenous approaches which all view knowledge creation as a relational process, strive for novel and deep insights and understanding of (symbolic) processes. As parallels in conceptualisation of relationality span realist and post-humanist ontologies, relationality provides a framework for considerations of quality across qualitative research paradigms.

We argue that quality in qualitative research rests on a set of relations built (constructed) throughout the research process. Accordingly, we define quality in qualitative research as a relational process. The most important research instrument in all qualitative research is the researcher. “Quality control” is a process where researcher engagement with quality is continuous and relational. This process has a beginning, progresses through various stages and concludes. The researcher steers the process by asking the following questions that are anchored in the generalist conceptualization of relationality outlined in the previous section:

Quality in this understanding is a set of relationships between the researcher and the researched concern (most commonly: the people and communities implicated in the concern) and the research context and the data and the analysis of the data and the audience or readership of the research. Hence, quality is threaded through the entire research process from the point at which the research focus is sharpened to the point at which the research is communicated to others. Quality in this understanding arises from (1) formulating a research question that is relevant (for the studied concern, participants, researchers and practitioners); (2) sampling for (or accessing) data through a credible process that pertains to the relevant phenomenon/experience; (3) engaging analysis in a manner that is credible – true to the data and (3) doing justice to the data (drawing credible conclusions), while continuing to remain relevant to the studied concern. We argue that relationality is the vehicle through which these components are achieved.

Quality in qualitative research involves holding these relationships together, and this in turn requires paying attention, throughout the research process, to (1) credibility and (2) relevance. Of these two shared quality criteria, credibility pertains chiefly to the how (is this research (being) conducted?). Credibility is discernible throughout the research process. It is deeply relational and processual. Relevance pertaining to what (what does this research pertain to and seek to contribute?) is relational as it involves connecting to existing frames of reference with a view to contributing to knowledge and understanding and relaying these contributions to those who wish to apprehend (and utilize) them. Indeed, in projects taking a critical approach, the research findings ought to be usable towards transformative, emancipatory conditions, yet credibility is also at the heart of realist and (post)positivist approaches.

Relevance and credibility remain after we exclude from the multitude of quality criteria reviewed in the previous sections those that are either too generic or objectionable to researchers from different paradigms (e.g. validity does not feature in post-structuralism). Relevance and credibility are the shared hallmarks by which we know that a piece of qualitative research is of good quality, as endorsed (in different ways) by all paradigms. While we argue that credibility and relevance are the quality criteria that all qualitative researchers can agree on, arriving at these hallmarks of quality necessitates certain steps that engage the relational approach.

We certainly are not the first to emphasize credibility as a core constituent of quality in (qualitative) research (e.g. Barbour, 2003; Glaser and Strauss, 1966; Lincoln and Guba, 1985). All qualitative researchers seek to be plausible and want to engage in research that others (within and ideally also outside their own paradigm) consider credible. As outlined above, in the (post)positivist paradigm, credibility features in the form of exhortations to reliability and replicability. Constructivists employ the terms trustworthiness and dependability when referring to credibility. Credibility therefore constitutes shared ground, from different perspectives but underpinned by relational research practices.

The widely agreed-upon vehicles for demonstrating credibility in qualitative research are transparency, contextualization, reflexivity (positionality) and consistency. From a relational perspective where the qualitative researcher is central, transparency demands openness about methods, data and their limitations; linking analysis and data and openness about any data that might run counter to argument. Where mutuality and reciprocity inform the research approach, transparency entails (relational) openness and accountability with research participants about research purposes, procedures for meaning-making and potential application of findings. Contextualization is necessary because all qualitative researchers regardless of paradigm attest that experiences and perceptions are shaped by context and relations therein (Korstjens and Moser, 2017). Contextualizing the findings of any qualitative study is necessary so that they can be compared or related to findings from other studies conducted in similar and different contexts.

Reflexivity takes lighter and deeper forms, depending on the researcher’s ontology; a qualitative researcher whose ontological viewpoint aligns more with realism than constructivism might not elaborate on their reflexivity as extensively as a researcher who is deeply rooted in constructivism. In contrast, critical perspectives require researchers to interrogate the effects of their own positionality on their research and to anticipate power relations lest the researcher’s worldview risk overshadowing the worldviews of those being researched. Critical researchers are called on to recognise and challenge unexamined assumptions inherent in their own thinking within the inquiry so that they can privilege participants’ standpoints. Regardless of such differences, practising reflexivity entails a relational process of committed engagement with the researched concern directed towards disclosing researcher positionality and being mindful of power imbalances in the research process.

Consistency – regardless of paradigm – enhances credibility: this can be demonstrated by giving examples of analytical processes such as searching for consensus with another coder (a relational process); or checking both the data and findings with the population of concern and other stakeholders (again, a relational process). Where singular realities are assumed, consistency can be demonstrated by keeping from start to finish a detailed log of all research activities and study procedures should other researchers seek to replicate the study in a different context or consider their own investigation against the background of prior studies. In research where coding is part of the methodology, credibility requires credible coding and subsequently credible categorization or thematic grouping of the data (St. Pierre and Jackson, 2014). Different qualitative methodologies approach data analysis differently but in all cases, coding needs to be balanced – the researcher must not neglect relevant parts of data or pay disproportionate attention to some elements (although every dataset has parts that are richer than others). Themes or categories should be internally coherent, not overly complex or confusing; codes find a home within them. Themes or categories should be examined in relation to each other (Bazeley, 2009) to identify connections between them that can elucidate key processes and patterns across the data. All these tasks require researcher relationality vis-à-vis the data and analysis and, ultimately, accountability for how this was achieved.

We advise that, in pursuit of credibility, researchers should not just allow data to accumulate, but rather ought to think about and/or with their data during the process of generating (or accessing) the data. This involves paying close attention to what is going on in the data while seeking to draw out an abstract understanding of the studied concern. This level of attention necessitates reasonable degrees of probing, prompting and clarification, to understand and account for key incidents, behaviours, perceptions, experiences or other relevant concerns in the data. In paradigms emphasising mutuality between researcher and researched, a collaborative process of building meanings is iteratively achieved by researcher(s) sharing insights to deepen meaning as data generation progresses. Thinking about the data involves varying degrees of relationality through co-construction between the researcher(s) and the participants. Memoing, traditionally associated with Grounded Theory (Birks et al., 2008), can be highly conducive towards relating the data and analysis through documenting the research process for most qualitative researchers and can feature as part of the overall approach to establishing credibility in the form of transparency regardless of paradigm (McGrath, 2021).

We argue that all qualitative research as relational process should orient to relevance throughout the research, starting with topic and focus that are relevant – of concern (and of potential use) – for research participants, research communities and society. Critical approaches concerned with transformation and emancipation underpin the entire research process with relevance. For example, participatory action research approaches involve the constituency of interest in forging the parameters of the study. Relevance, like credibility, points to context. Context is relationally constituted by the researcher(s), participants, practitioners (where applicable) and/or sites and sources for data so that the research question(s) are both enabled and addressed (Riese, 2019). Qualitative research must always consider the natural contexts where groups and individuals of concern to the research exist in order to arrive at an understanding of their situatedness (Korstjens and Moser, 2017).

Relevance is not just an outcome of qualitative research. Rather, relevance begins at the design stage and features throughout data collection and analysis. This means that the researcher is actively engaged in (i.e. relates to) the generation (or selection) of data and understanding from start to finish (an outcome achieved with participants where mutuality is a concern). Relevance of the data and findings is manifest as resonance and potential utility for the people and communities that the research relates to. The researcher must ask: do the findings resonate with the participants in the study and the constituency and positionality they represent? However, this is not the same as generalizability or everyone agreeing what should be done about the matter of concern. Resonance is closely associated with transferability and naturalistic generalizability (Tracy, 2010), processes whereby those who read or hear about the research can relate to its relevance and possible (partial) applicability to another context familiar to them. For the (practitioner) research community, resonance is evidenced through useful abstractions that are endorsed as contribution to knowledge, although establishing this might take a long time and take various forms.

Persistence, sensitivity and flexibility in the search for understanding (Verstehen) run through research design, sampling, data collection and analysis in qualitative research, which collectively steer engagement with both scholarship and the empirical enquiry to achieve relevance (Hennink et al., 2020; Pultz, 2018). Relevance also entails considering the role of theory. All research relates to theory somehow but not all qualitative research is guided by a “theoretical framework” – it is important to be clear about the role of theory (Collins and Stockton, 2018) in your research. Theoretical sensitivity refers to the ability to know what data are important in the context of theory and why – in other words, what is both meaningful and significant in the data relative to existing or developing theory. Generalising to abstractions and/or at the level of theory involves explicating what patterns feature in the data, and elucidating these patterns in the data – when saturated for meaning and context – strengthens the relevance of the findings.

Relevance also relates to and arises from the political nature of research (Bandola-Gill, 2019). Both the researcher and participants are members of society and its power structures, and these structures influence the relational nature of research undertaken and how it is funded (Tarozzi, 2013). Finally, the relevance of research relates to its specification of future research. In many cases, qualitative research seeks to map out key parameters of the topic of concern before they are subject to further (qualitative and quantitative) investigation. Clear articulation of remaining research lacunae and new questions that have arisen in research publications (and other forms of communicating findings) are important markers of relevance and relationality.

Some practical steps are useful in seeking to ensure that quality is built into the process of qualitative research by remaining focused on relevance and credibility through attention to relationality. Qualitative research is often initiated with a topic or area of interest rather than a question in mind. Formulating the research question(s) can take a long time in a qualitative project, but this process is aided by attempts to formulate an actual question (a sentence followed by “?”). Being clear about the focus of a qualitative enquiry helps researchers to ensure that their research is credible and will also assist in establishing its relevance. When the researcher is clear about the focus, they can relate their study to the relevant literature, plan collecting relevant data, and make a credible case for contributing to understanding in relevant fields.

Relevance of the research question can be probed by asking whether the question matters to people, especially those whom the research concerns. An early indicator to the researcher in the field is how their attempts to recruit participants to the study are received. Research questions that resonate with the constituency of concern and succeed in having people come forward to participate indicates relevance. The data generated during early stages of a project are useful testing ground for the direction of the study.

Critical perspectives go further and seek potential for transformation and emancipation in considering what is relevant. Researchers concerned with quality should be attuned to these indicators of relevance and opportunities to fine tune studies for relevance from the outset of the study design and early data generation. In some areas of research such as health care where practitioner researchers often feature, the prevalent view now is that relevance means ensuring that the public whom the research relates to are actively involved in designing and overseeing the research in formal collaborator or co-investigator roles. Research questions formulated by the researcher without relational engagement with the public of concern can fall short in terms of relevance (impact and benefits) for the population of concern. Relevance cannot be established without relationality; again, this is particularly well-illustrated by Indigenous research where relevance and relationality may encompass culturally significant more-than-human elements such as the land, animals, ancestors and other actors and actants at the centre of the enquiry (Denzin and Salvo, 2020; Kerr and Adamov Ferguson, 2021).

In qualitative research, it is usually essential to approach data collection with an open mind, but also informed of prior theories to the extent that is suitable for the research paradigm and methodology. It is helpful to establish at least some understanding of concepts that are likely to be relevant before initiating the data collection process. Ensuring that the data in fact relate to and are relevant to the research question might necessitate revisiting the relevance of the research question and refining it. The flexibility of qualitative research allows the researcher to do so where necessary. Ensuring that the data give scope to say something interesting means that the data must “speak” (relate) to the research question in a nuanced manner. This in turn necessitates skill and care in generating data (or in accessing and understanding the background to the generation of secondary data). When the research approach is concerned with reciprocal research relations, the fit between data and research question can be iteratively explored during data generation. For example, as tentative insights emerge, these can be explored in subsequent data generation episodes to test their relevance and credibility with the researched community (Charmaz, 2014; Salzano et al., 2023).

Focusing on relevance and credibility through relationality can help to pre-empt some of the (often unfair or misinformed) questions from people with little or no familiarity with qualitative research – such as “what can you say on the basis of interviewing only 15 people” or “how do I know you are not just presenting selected data extracts to support your opinion”? These can be countered by saying, for example, that interviewing more people is unlikely to result in additional data that is relevant for answering the research question, and that the findings are credible for research participants. Detailing iterative processes of sampling and meaning making during the data generation process and emphasising the value of situated knowledge (as appropriate for the chosen ontology) can also counter such challenges. Misinformed comments often fall on the issue of generalizability because of a lack of understanding about transferability in qualitative research. It is not unusual for qualitative researchers who submit their work to journals that publish more quantitative than qualitative research in their field, to be asked to highlight in the limitations section of a manuscript that their work is limited because it does not generalize. In these situations, a focus on relevance and credibility, including weaving these criteria explicitly into the methods section of the work, is an effective platform from which to emphasize the quality and relational nature of the research. Finally, although disinclined to provide or recommend checklists, we suggest asking (and answering) the questions in Table 1 during the research process. These are the same questions that were posed above, and the resolutions are a short version of key steps recommended when working towards credibility and relevance by engaging the relational approach.

We define quality in qualitative research as a relational process that aims at, and delivers, credible and relevant contributions to knowledge and understanding. Qualitative research progresses towards and meets the shared quality criteria of credibility and relevance where researchers are able to give certain reassurances to themselves, others working with them, those they research (with) and ultimately those who read their research. All these reassurances are grounded in the idea of relationality where the researcher(s) acknowledge(s) and act(s) according to the precept that the researcher(s) is (are) responsible for credible and relevant accounts of the relationship of their enquiry to existing knowledge, the research context and participants, the data, the findings and the ideas, arguments and theorising that ensue from the research.

Quality in qualitative research necessitates that the researcher – academic, applied and practitioner alike – knows how, why and under what conditions the empirical material (data) have been generated, selected or accessed; this might sound obvious, but these elements are surprisingly often absent, even in projects led by senior investigators. A qualitative researcher committed to quality has a strong grasp of the data, has thought about and with the data in a theoretically engaged way, thereby forging a relationship between the data and analysis that the researcher can account for. The analysis and argumentation are in dialogue with – relate to – other research, concepts and theory. These are the hallmarks of qualitative research of high quality, achieved through relational processes focusing on credibility and relevance.

We contend that the root cause of common missteps in qualitative research often lies in lacking commitment towards research quality. If we are committed to and care about our research and participants, we will think about them more deeply and take care in approaching them: the best route to quality runs through care and commitment which are enabled and strengthened through the relational approach. The foundations of qualitative inquiry, across all paradigms, demand an engaged, committed researcher attending iteratively and with care – relationally – to every stage of the research process in pursuit of quality. Formalistic procedures for quality control in human inquiry are not a sufficient guarantee of quality. Concern with quality of qualitative inquiry goes hand in hand with commitment to relevance and credibility threaded through and evidenced at every stage of the relational research process.

Relational knowledge involves delving into substructures of meaning-making and signification which differs significantly from “box ticking” – the researcher is trying to access novel insights into how meaning is built and how shared understandings are arrived at. Qualitative research is never just about capturing responses; the qualitative researcher cannot be a passive collector of data, but rather needs to interact with (relate to) persons, fields of knowledge, ways of seeing and through these interactions they build understanding relationally. As an important addition (that needs to be fully teased out in another paper), we posit that relationality is also the key reason why Artificial Intelligence (AI) is not able to replace the qualitative researcher: AI is not able think relationally and cannot generate relational knowledge. Creation of insight through relational processes remains the domain of the human qualitative researcher.

Virpi Timonen thanks Professor Monika Florczak-Wątor and Professor Jolanta Perek-Białas, organisers, and students of the Interdisciplinary Doctoral Program “Society of the Future” at the Jagiellonian University, Cracow, Poland, for the opportunity to present an early version of the ideas in this article in May 2023. She also thanks Professor Blanca Deusdad for the invitation to present an earlier version of this paper to the anthropology and ethics seminar as part of the Programa de Doctorado en Antropología y Comunicación at the Universitat Rovira i Virgili in November 2023. Virpi Timonen also acknowledges the support of the European Union (LEGACIES, ERC, project number 101094124). Views and opinions expressed are those of the author(s) only and do not necessarily reflect those of the European Union or the European Research Council. Neither the European Union nor the granting authority can be held responsible for them.