Quotations in Qualitative Studies: Reflections on Constituents,Custom,and Purpose

Abstract

Qualitative studies are often found to be accompanied by quotations from interviews or similar data sources. As with any methodological tradition, it is essential to critically explore the general principle of including quotations in scientific papers: what is the purpose and justification for including quotations? Are there standards and, in that case, what are they and what are their scientific positioning? This paper presents an overview of the somewhat diverse guidance found in the literature in reference to the representation of quotations. Yet, both students and scholars use a variety of approaches to quote from their data, ranging from presenting numerous, extensive, and/or comprehensive quotations throughout the results section to the reporting of a few particular quotations to illustrate certain aspects of the findings only. While their function may be described, scientific reasoning for using quotations is scarce. Along with an overview of the scientific background and options for including quotations in qualitative studies, we discuss the consequences of the different epistemological and methodological aspects found in the literature. In conclusion, we argue that there are incentives for promoting a more deliberate employment of references from data; deriving from the human sciences tradition, a corresponding epistemological stance would suggest that quotations preferably apply for illustrating the analysis process and/or findings, while the idea that quotations can be employed to validate findings has limited support. Further critical examination of the application of and justification for including quotations in the reporting of qualitative studies is needed among researchers, journal editors and reviewers.

Introduction

As researchers and as reviewers for scientific journals as well as teachers in undergraduate and postgraduate healthcare training, we often encounter quotations and questions vis-à-vis quoting in the reporting of qualitative studies. In our experience, qualitative findings presented in scientific papers and reports seem to be rather habitually associated with quotations. Yet, it is rare to find an account that justifies the inclusion of quotations or that defines the purpose of including them. This paper is our venture to decipher the scientific basis for enrolling quotes in qualitative reporting, discuss the purpose and role of quotations, and delineate some theoretically just routes, while also addressing the potential liabilities in enrolling quotations in academic writing practices.

Background

What Are Quotations and Quotes?

While the terms “quotations” and “quotes” are sometimes used as synonyms, they originally represent a noun and a verb, respectively. Further, both terms have several connotations, but, as applied in qualitative research, the term “quotation” generally signifies “passages reproduced or repeated,” whereas it can also indicate the act of quoting—that is, “to quote,” which would be to reproduce or repeat a passage, including a person’s spoken statement (Oxford English Dictionary, 2020).

When Do Quotations Apply and Why—A Scope of the Literature

Presenting authentic citations of what informants have uttered has become the “gold standard” in qualitative studies, and most people who read such papers and reports expect that there will be a number of such quotations; they may even expect a particular type of quotation (Brown, 2010).

Quotations can bring content to life (White et al., 2014), and a sound application of research ethics prevents the (mis)use of catchy phrases; anecdotal use of striking examples provides neither in-depth focus nor validation (Silverman, 2014). Rather, it is vital that quotations are applied in a way that shows respect to participants and that they are presented in a reliable manner, including that they maintain participants’ confidentiality (Polit & Beck, 2016).

Creswell (2012) describes three types of quotations:

Discrete quotations, which are easy to read and take up little space. These types of quotations are described as signifying different perspectives.

Embedded quotations, consisting of briefly quoted phrases. These quotations can be used to prepare the reader for a shift in emphasis or to move on.

Longer quotations, which are used to exemplify more complex understandings. These are scarcer in scientific papers, where the word count is often limited.

Any quotation or extract used should support the interpretations and explanations presented, according to Yin (2011). The use of brief quotations interspersed within selected paragraphs can subtly signify richness of data. Yet, a lengthier extract, illustrating a condensed finding, over multiple paragraphs, will better represent extensive data and the informants’ experiences (Yin, 2011).

Corden and Sainsbury (2006) suggest that researchers quote for providing explanations, and to elucidate processes and informants’ perspectives. Further, quotations are also sometimes used as illustration; for example, a need to describe how something has affected an informant is illustrated by means of presenting his or her own words. Corden and Sainsbury (2006) even suggest that quoting deepens the understanding by means of enrolling the informant’s voice. All in all, quotations are supposed to enhance readability, providing vividness to the text. Thus, quotations aid communication. According to Corden and Sainsbury (2006), quotations also serve as evidence: original data help to serve the reader’s assessment of the accuracy of the analysis, thereby strengthening the findings.

The Scientific Claims Vis-à-Vis Quotations

As the idea of presenting quotations as evidence in qualitative research appears every now and then, we suggest looking a bit further into the basics of promoting scientific rigor, as in validity and reliability. First and foremost, many scholars argue that a better fit with a qualitative worldview is to talk about credibility (e.g., Noble & Smith, 2015). Of significance for the development of qualitative research was the work of Lincoln and Guba (1985), who presented four constructivist criteria for trustworthiness, parallel to but differing from the criteria for validity and reliability:

Credibility (similar to internal validity) is about providing assurance and fit between the informants’ reality as presented and the researchers’ re-construction and representation of this presented reality (Guba & Lincoln, 1994). One technique to achieve credibility is triangulation: using different or multiple sources, methods, researchers, and theories to present evidence.

Dependability (similar to reliability) focuses on the process of conducting the study, making it possible to follow its audit trail. The product/data should also be examined, in addition to discerning whether the records are accurate, providing an acceptable baseline.

Confirmability (similar to objectivity) is also about the audit trail, particularly in being able to follow the whole process—including data collection and analysis. Being able to reach an overall decision about the confirmability of the study is vital; data and interpretations are grounded in events and not in the researcher’s assumptions or point of view. Confirmability includes judgement about the appropriateness of the labelling of themes or categories as well as the quality of the analysis and how the findings fit the data.

Transferability (similar to external validity) is about generalization from case to case or group to group, addressing whether the findings apply to another context (Guba & Lincoln, 1994).

Credibility, or trustworthiness, is supposed to be reached by using these four criteria, making it possible and simple to follow the audit trail (Guba & Lincoln, 1994).

More recently, Morse (2015) has suggested the term “rigor” instead of trustworthiness, replacing the standards of dependability, credibility, and transferability with the more contemporary and scientifically well-known reliability, validity, and generalizability. Both reliability and validity are indicative characteristics when aiming to conduct rigorous qualitative research. Supposedly, rigor can be ensured by using verification strategies throughout the whole research process (Morse et al., 2002). The researcher is responsible for checking and adjusting the research processes to make certain that the results are robust (Spiers et al., 2018). Neither of these references specifically argue for or against quoting or quotations, but we will revisit these propositions in the discussion.

Although, historically, science was associated with a range of subjects, or liberal arts, including, for example, rhetoric, a later development favored the natural sciences (Kuhn, 1970). As scientists primarily occupied themselves with classification, quantification, and measurement, it comes as no surprise that quantitative methods were described preceding those of the qualitative traditions. While qualitative researchers long struggled for their methods to be acknowledged as science, both traditions are nowadays chiefly recognized by scholars and in society. In addition, most researchers are familiar with the different ontological, epistemological, and methodological standpoints of both quantitative and qualitative practices. However, while the understanding of science has evolved over time, today, many scholars would further recognize an alternative sequence, with the postpositivist, constructivist, transformative/critical theory and pragmatic considered the main paradigms (Creswell, 2012; Denzin & Lincoln, 1998). As such, these understandings are normative, offering scientific laws grounded in ontology and epistemology, shared by the scientific society as assumptions, values, methods, and expectations (Kuhn, 1970).

For the positivist/postpositivist paradigm, validity, reliability, and objectivity are well grounded in the assumptions of its ontological and epistemological framework (Krippendorf, 2013; Lincoln & Guba, 1985), while both internal and external validity also transpire in the transformative/critical theory and pragmatism paradigms. Within the latter tradition, internal validity is represented by a sound match between the researcher’s data collection and the findings presented; external validity, on the other hand, is about the degree of generalizability of the findings (Lecompte & Goets, 1982). This means that methodologies such as grounded theory, content analysis, and ethnography could well use the concepts of validity and reliability—or rather, internal and external validity—even though grounded theory proposes its own criteria, suggesting that findings are conceptual and expected to speak for themselves (Glaser & Strauss, 1967; Glaser, 1978). This positioning is similar to the ideas of phenomenology (Giorgi, 2005) and hermeneutics, mapped in the constructivism paradigm, in that the findings of an qualitative inquiry correspond to a meaning (Gadamer, 1960/2013), which reinvents itself as the subject encounters the text, as parts and as a whole (Ricoeur, 1989).

Consequently, the essential literature of the qualitative traditions, such as constructivism, pragmatism, or the more recently recognized realism (Pawson & Tilley, 1997), provides limited evidence as to the justification for quoting data extracts in qualitative studies. Even so, guidance provided by, for example, COREQ (Tong et al., 2007) does constitute criteria regarding the effective application of quotations. While this guidance is widely accepted as the norm in reporting qualitative research, we propose instead that further discussion regarding the practice of enrolling quotations is needed.

Discussion

The origin of including quotations in qualitative research remains uncertain, and, so far, limited guidance in the scientific literature concerning quotations is noted. Hence, it comes as no surprise that many scholars, regardless of their level of experience and expertise, have a somewhat vague idea as to why, how, or when to quote. Rather, in everyday practice, a variety of reasons and routes are suggested for quotations in qualitative research, most often including that it is “expected.”

Patton (2002) suggests that, by using quotations, the informants’ views are captured—representing these in their own words; the main purpose of including quotations and extracts in research would be to demonstrate how the findings and interpretations have arisen from the data. Further, Sandelowski (1994) suggests that quotations are used to support a researcher’s claims, illustrate experience, evoke emotion, and sometimes to provoke a response. Descriptions and direct quotations provide the basis for qualitative reporting and allow the reader to enter into the situation presented (Patton, 2002). Thus, entering quotations could be understood as a validation process (Matt, 2004). Even so, if the quotations are presented without an interpretive commentary, the reader is forced to take on the task of the analyst, although with a very small and limited data set (Brinkman & Kvale, 2015; White et al., 2014). Polit and Beck (2016) claim that, because there are no set guidelines as to how many quotations should be used or how long quotations or extracts should be, the researcher(s) must decide how to use them and where they are appropriate, and how to guarantee that the quotations fit the context. An overuse of quotations in the results can blur the findings and make them difficult to read (Elo et al., 2014). A somewhat contrary recommendation is to include “rich and thick verbatim descriptions of participants’ accounts to support findings” (Noble & Smith, 2015, p. 35). According to Patton (2002), verbatim quotations are synonymous with direct quotations, which are a basic source of raw data in qualitative research that serve to reveal the informants’ emotions and experiences.

As the purpose and validity of quotes relies profoundly on the appraisal and skills of individual researchers and/or teams, they must decide whether the quotations inserted into a paper or report are presented verbatim or are edited or condensed. Verbatim quotations can indeed convey a particular experience or statement, but unedited citations can also lead to a sense of exposure for an individual—especially if the number of participants in a study is limited. Other, more practical, considerations include whether quotations can be—or ought to be—shortened. Some suggest that condensation is acceptable if it is obvious or declared that the quotation(s) are edited, and if they are indicated by a consistent truncation; for instance, a text including either a double slash or brackets signifies that part of a verbatim utterance has been left out (Denzin & Lincoln, 1998). In group or focus group interviews, for example, quotations are employed to illustrate the associations between the participants. To highlight this purpose, longer passages from transcripts may be relevant, although they may also be shortened to indicate a particular relation or structure (Flick, 2018). Quotations should not exclude the provision of enough and reliable data which should be sufficient to sustain the analysis and findings (Wolcott, 1990). Furthermore, qualitative research should be presented with convincing arguments, including enough representative data to support the findings and any alternative explanations (Morse, 1994).

Understanding social encounters, which is often the case in qualitative research, calls for an intellectual effort requiring both rigor and diagnostic transparency (Bennett, 2010). Well-founded motives for the use of quotations would allow for academic development. It would also sustain future reviews, enabling both increased transparency and methodological advancement. Supplementary verification of quotations would bridge the common ambiguity about the rationale for using them. The literature that we have addressed that calls for a more discerning, rather than common, practice could aid in the further establishment of well-founded principles.

Yet, a 2007 review of checklists used to assess qualitative studies revealed that the most common provisos related to data collection and the validation of findings, that is to say, the description of how themes were developed and the inclusion of supporting quotations (Tong et al., 2007). Subsequently, this review article became the foundation of the COREQ (COnsolidated criteria for REporting Qualitative research) guidelines, a 32-item checklist used by numerous journals. Out of the current 32 items, two are about quotations:

“Were participant quotations presented to illustrate the themes/findings? Was each quotation identified? e.g. participant number,” and

“Was there consistency between the data presented and the findings?”

These expectations are consistent with suggestions such as, if sufficient data are presented to support the findings, the researcher(s) should include quotations from different participants to add transparency and trustworthiness to their findings and interpretations of the data (Cote & Turgeon, 2005). However, from an ontological and epistemological perspective, we have found limited support for this viewpoint in the contemporary qualitative research literature. Rather, a potentially superficial approach may thereby have been taken, neglecting the “if,” “why,” “how” and “when” queries relating to the use of quotations in qualitative research. Quoting can have a certain purpose and quotations can operate well, although there is not one but several ways to employ them. If quotations are applied, this should be deliberate, and transparent. As a result, qualitative papers should present and illustrate how and why quotations are used.

We propose that two strategies are sustained for the use of quotations, each occupying different epistemological positionings for their use in qualitative studies. Both are rooted in what is originally suggested as the human sciences, which later advanced into, for example, the constructivist paradigm (Creswell, 2012; Denzin & Lincoln, 1998). While the ontology of paradigms communicates a scientific worldview, we suggest that both routes correspond to a philosophical position that reflects the multiple perspectives of human experience. Thus, while the routes discard any proposition of the relevance of employing quotations as evidence, we suggest that there is ontological and epistemological support for particular use of quotations in qualitative studies:

Quotations might be used to illustrate the data analysis process, that is to say, the method applied. In this case, the use of quotations is positioned within the epistemological assumptions of a descriptive representation. The quotations can be presented in the section describing the analysis but can also be presented throughout the findings. In order to provide for transparency and a potential to judge the trustworthiness of the process, quotations to demonstrate the analysis should be sequential, and each phase of the analysis should be accompanied by quotations. In addition, to affirm the analysis process, the quotations are likely to be more detailed than when the purpose is to illustrate the various features of a study, as described below. Yet, the same ethical perspectives apply and, as a result, the use of quotations should signify a cautious approach in which the confidentiality of study participants is not jeopardized. Furthermore, in order to affirm the process and findings, the quotations should represent a careful selection of the entire dataset with only limited modifications expected; to sustain a descriptive analysis, the lifeworld of the study participants should remain intact.

Quotations might be used to illustrate findings, with the main purpose being to bring the text to life—or bring life to the text. Using quotations to illustrate the results would stem from the epistemological assumptions of an interpretative approach. Thus, the quotations can be used to highlight particular features of the data, while also making the text more appealing to the reader. However, researchers should seek to find only quotations that illustrate the findings, rather than cherry-picking controversial remarks or events. Such “outliers” are generally of limited value for an enhanced understanding of the qualitative analysis process and/or the findings, which rather resembles a shared, conceptual understanding of meaning. When there is an intention for quotations to illustrate the results of an interpretative analysis, these can be placed randomly throughout a section of findings. Modifications can be made to exclude superfluous text yet should fortify the content and meaning of the interpretation. Furthermore, to put flesh on the bones, the quotations may represent a selection, accurately reproduced, rather than the entire dataset.

Two examples of the above strategies, that is, to illustrate the analysis process and to illustrate the findings, respectively, could illustrate how researchers adopt either route. First, in stating that “every quote was condensed, and a code reflecting the content was given to the quotes,” Søreide et al. (2019, p. 891) explicate how quotations were used in a phenomenological-hermeneutical analysis, including a sample of the analysis process in the methods, along with a variety of quotations in the results. Second, van der Zijpp et al. (2016) propose that their qualitative results include “themes illustrated with the narratives from across countries” (p. 27), signifying a further visualization of their findings.

The above suggestions challenge a fairly common initiative in that different quotations from several informants should be presented to support each category/theme in particular. For example, while suggesting that their study was situated within the socialist-constructivist paradigm, Heery et al. (2019) propose that verbatim quotes enhanced the confirmability of the findings, whereas Roman et al. (2020) indicate that, in their phenomenological study, “each of the main themes was defined and illustrated with representative quotes from the participants” (p. 405). McPeake et al. (2020), on the other hand, conducted a qualitative inquiry analyzed with thematic content analysis and inserted “extracted quotations to support the thematic analysis” (p. 3). While we recognize that, although the above, random examples, and many others, did not equally propose a rationale for presenting the quotations, it may well be a result of a rule of thumb or guideline adherence, for example, to COREQ.

However, some of the random examples do indicate that researchers may employ quotations to validate the results of a qualitative analysis. Given the literature cited in this paper, and thus considering the ontological and epistemological foundations of qualitative research, there is limited support for such claims, that is, that quotations can substantiate a set of findings. Rather, several of the ontological traditions, such as grounded theory and phenomenological hermeneutics, stress that the outcomes of qualitative research constitute a comprehensive understanding of meaning and behavior presented as concepts (Glaser, 1978); the life-world of people is not to be divided or fragmented but captured as parts constituting a whole (Ricoeur, 1992). Although earlier qualitative guidance suggests that quotations should support and enrich accumulated findings, validating them is instead accomplished by adopting strategies such as triangulation and cross-checks (Malterud, 2001). Again, any qualitative analyses should be rigorous and transparent enough to provide trustworthiness (Graneheim et al., 2017; Guba & Lincoln, 1994). While certain procedures, such as member checking or triangulation, have been proposed, a purposeful application of quotations may further illustrate and thus accredit a qualitative analysis and/or illuminate the findings in qualitative studies. To settle on an appropriate route, the consideration of certain issues helps, such as ensuring congruence as to why one might quote and applying a coherent location for and type of quotations.

As we suggest above, from a research ethics perspective, it is important to protect confidentiality when quotations are used, particularly when they are presented verbatim. Consistent use of identifiers that do not reveal identities can aid this practice and will also inform the reader as to whether the data set is abundant; a deliberate sample including quotations from most or all of the interviews indicates ample data to sustain the findings (Denzin & Lincoln, 1998), although this would not exclude a further consideration, vis-à-vis, for example, saturation in qualitative research (Saunders et al., 2018).

Conclusions

Today, a variety of ways to use quotations is observed in the practices of reporting qualitative research. Quotations are seemingly applied for several purposes: as evidence, explanation, illustration, impression, representation, and/or to enhance readability. However, we suggest that the use and purpose of quotations needs to be more purposeful and transparently described than is commonly seen today. Thus, the objectives of the use of quotations should be rigorously upheld, with the scientific basis for quoting reported. Scholars should therefore, accordingly, consider the research design of their study and paradigm in which it is positioned, the methods applied for collecting and analyzing data, and the research ethics of extracting, phrasing, and organizing quotations in research papers and reports.

Fueled by the literature, we propose that drawing on quotations from data within qualitative research could be used to illustrate the analysis process and/or findings. However, the choice of how many quotations and the extent of the participants’ accounts is still a decision that must be made by the researchers, while considering the confidentiality of study informants. Although we expect both the purpose of quoting informants and/or the role of the quotations are increasingly being articulated in qualitative analyses and further reported to readers of such papers and theses, we also propose that these aspects are considered by editors and reviewers. This would support qualitative studies becoming more transparent and compelling to their audiences. Further, with such details at hand, future reviews could inform an understanding of what is appropriate quoting, or not. Although this may procure for further guidance for scholars, such efforts preferably move beyond a checklist approach but offer guidelines accompanied by appropriate epistemological underpinnings.