A critical analysis of respondent quotes used as titles of qualitative research papers that are published in peer-reviewed journals

ABSTRACT

The use of respondent quotes to headline qualitative research papers is a popular literary device found in many academic journals. This practice has increased over the last four decades and now appears normalised within qualitative research writing. This article provides a critical analysis of this trend in academic writing and concomitant publishing. Content and framework analyses of 40 papers employing this literary device to summarise the respective studies identified (i) a lack of methodological rigor, (ii) incomplete analysis, (iii) an overall mis-representation of the wider qualitative dataset, and (iv) possible investigator bias associated with using respondent quotes as titles of qualitative research papers. This article questions the credibility of purposely selecting a single experience that reduces the wider collective experience into one deterministic statement. This article contends that such practice is antithetical to the principles of qualitative research. Recommendations are provided to better monitor this practice throughout the academy.

KEYWORDS:

• Qualitative research
• methodology and methods
• respondent quotes
• research titles

Introduction

In the Editorial of the inaugural issue of Qualitative Research, Atkinson, Coffey, and Delamont (2001) outlined a vision for that journal and its content. In that vision, the editorial team stated that, “As qualitative research methods achieve ever wider currency in the social and cultural disciplines, we need constantly to apply a critical and reflective gaze. We cannot afford to let qualitative research become a set of taken-for-granted precepts and procedures. Equally, we should not be so seduced by our collective success or by the radical chic of new strategies of social research as to neglect the need for methodological rigor.” (Atkinson, Coffey, and Delamont 2001, 5). This article seeks to contribute to those editors’ “debate about our canon” (ibid) in providing such a critical and reflexive gaze upon the seemingly widespread and undisputed use of quotes by research respondents as the titles of qualitative research papers.

In the two decades that have passed since the above editorial was written, respondent quotes used as titles have become a literary device that appears to have become firmly entrenched within the aforementioned “set of taken-for granted precepts and procedures” when attached to the production of qualitative research findings. Indeed, the use of respondent quotes as research titles has become so quotidian and commonplace that their current use potentially gives rise to accusations of “neglecting the need for methodological rigour” within qualitative research. Furthermore, to date (as far as the authors are aware), the selective use of respondent quotes that headline research papers has never been questioned within the literature. Accordingly, the authors aim to provide a methodological challenge to the use of this literary device within qualitative research and to simultaneously question those (be they authors, scholars, journal editors, peer reviewers) who choose to present, permit, accept and/or publish findings obtained from qualitative research in this unchecked manner. In doing so, the aim of this article is to contribute to the field of qualitative research by reminding those who neglect methodological rigor the qualities of qualitative research.

Respondent quotes used as a title: a fictional, illustrative, example

In order to proceed further, it is necessary to illustrate the relevant literary device, so that the reader may fully appreciate the topic in hand. Therefore, the following fictional “journal paper” provides an example of what this article seeks to review and critique.

Fictional title. “Because I’m a clean freak”: a qualitative study of why people who inject drugs avoid sharing needles and syringes. Written by Anon (n.d.).

In the above example, the first clause of the title (“Because I’m a clean freak”) consists of a quote (an interview extract) made by one individual (respondent) during an interview attached to a specific research project, involving a cohort of people who inject drugs. This is followed by a more descriptive and illustrative second clause after the colon (namely, 'a qualitative study of why people who inject drugs avoid sharing needles and syringes'). The second clause provides an indication of the study (and its research question and sample) as conducted by the anonymous author/s at an unknown date (n.d.). This sentence structure (namely, two independent (yet connected) clauses separated by a colon), is currently the standardised model of presenting respondent quotes within research titles. Headlining research papers in this way have become widespread, normalised, practice in the presentation of qualitative research within (English language) academic journals.

The proliferation of respondent quotes used as research titles in peer reviewed journals

This article has been inspired by the authors’ recognition that the use of respondent quotes as research titles appears to have become standard practice within a wide range of social science/health-focused journals that are sympathetic toward publishing qualitative research articles. Indeed, the number of articles featuring quotes as titles has steadily increased over the last four decades (since c.1980) to the extent that their inclusion in journals has now become widespread and seemingly “accepted” (quite literally) without question.

The following article therefore provides a critical and methodological reflection on this current trend in academic writing and concomitant publishing. This trend essentially involves the use of one research respondent’s view (within a given dataset) that somehow represents (or signals) the views of a total cohort involved in each study. As such, this paper seeks to question the credibility and integrity of purposely selecting a single experience that serves to decrease the wider collective experience into one reductionist and deterministic statement. In short, this critique seeks to challenge research practice that appears antithetical to the fundamental practice and principles of qualitative research.

The core elements of qualitative research

Whilst accepting that there is no single “correct” way of conducting qualitative research, the authors also concur with Ormston et al.’s (2014) view that such research should also be founded upon key elements that define qualitative research. These key elements essentially refer to the methodological design (see, for example, Cresswell 2013) attached to a study, and relate to features such as philosophical orientation (e.g., ontological and epistemological position of the researcher), the commitment to a given research tradition (such as ethnography, critical theory, realism, interpretivism, constructionism, symbolic interactionism and phenomenology), from which the researcher will draw upon relevant social theories to frame and situate their qualitative study.

A further methodological keystone of qualitative research is qualitative data analysis. As with the “methods” of qualitative research, data analysis is an equally broad and varied church. For example, this church embraces, among others, forms of analyses such as inductive/deductive analysis, thematic analysis, content analysis, narrative analysis, discourse analysis, framework analysis, visual analysis and interpretative phenomenological analysis as part of its varied congregation. Additionally, regardless of the form, the focus of qualitative data analysis tends to be either substantive or structural. Spencer et al. (2014, 272) define substantive approaches to data analysis as those which capture and interpret meanings within data and are concerned with what the texts “say” (e.g., critical analysis). Structural approaches to data analysis however are more concerned with what the text “does” (ibid) due to a focus upon the way in which language and social interaction are constructed and co-constructed by respondents to provide meaning to given issues (e.g., narrative analysis).

The craft of writing-up qualitative research

The final stage of a qualitative project – and arguably the topic of this article – involves the way in which qualitative researchers choose to report, present and disseminate findings obtained from their studies. As with all aspects of qualitative research described above, there appears to be an entire industry premised on the production of guidebooks and “how to” manuals dedicated to developing and improving the skills of “writing-up” qualitative research data. These texts (whether journal papers, textbooks or research monographs) typically provide useful and constructive advice (from experienced researchers) regarding the representation of the social world studied as well as how to re-present that social world in ways that should “remain grounded in the accounts of participants” (White et al. 2014). Accordingly, the use of respondent quotes within the corpus of a research report is a requirement for the representation and re-presentation process, as it is typically these data extracts (and associated interpretations) that will ultimately illuminate the experiences and social worlds to which the qualitative researcher has been given temporary and privileged access.

Interview extracts that are embedded in qualitative research reports are therefore important (and necessary) to the craft of writing-up qualitative research. However, despite this importance, there is a comparative lack of guidance in the literature that documents best practice use (regarding selection and application) of verbatim material for inclusion within research output. Exceptions do exist, however. For example, Weiss (2004) provides a didactic summary of “how” interview extracts may be selected and integrated within research output for descriptive purposes. Weiss also includes a short account of “investigator bias” and how this may influence the selection and use of interview data during the writing-up process. Namely, “an investigator who is determined, consciously or unconsciously, to have a particular theme emerge from his or her study can choose respondents whose interviews are likely to produce that picture … . and write a report that neglects whatever might disconfirm it” (Weiss 2004, 49). Elsewhere, Green (2013) provides a reflexive account regarding the rationale and decision-making processes that informed her selection and use of interview data in a specific manuscript. Here, Green emphasises how verbatim quotes were selected to contextualise and describe her respondents’ setting and to highlight the importance “of marshalling materials to progress an argument, and of writing with a particular readership in mind.” (Green 2013, 105). More recently, Lingard (2019), has developed these shared positions in providing guidance on the selection and integration of quotes into articles that are premised upon the “principles of authenticity and argument.” For Lingard, the authenticity principle is concerned with the way in which (i) the data illustrate an author’s finding, (ii) how succinct the quote may be and, (iii) most significantly, how the quote represents the patterns noted in the entire dataset and if the quote selected is “distributed across participants” (p362). The argument principle however relates to the way in which data extracts are presented grammatically and rhetorically – in which effective use of colons seek to emphasise the specific argument advocated throughout the entire text.

During 2003–2005, Corden and Sainsbury (2006) conducted a qualitative study commissioned by the Economic and Social Research Council (UK) to investigate the use of verbatim quotes by researchers in applied research reports. This unique study, in part, sought to explain and demystify the way in which “evidence” and “conclusions” are derived from qualitative studies in recognition that such data may be persuasive and convincing. The subsequent findings identified multiple constructs used by researchers for including verbatim quotes within their written outputs. Namely,: as a “matter of enquiry; as evidence; as explanation; as illustration; to deepen understanding; to give participants a voice, and to enhance readability” (Corden and Sainsbury 2006, 11).

The overall conclusion within the body of work summarised above is that the use of quotes is attached to an entire body of work associated with analyses. This equally infers that asolitary quote used as a research title is detached and separated from the main analyses.

Finally, titles of research papers/monographs are perhaps equally important constructs in the academic environment of research and inquiry. For example, research titles not only concisely summarise the content of an entire body of work written by an author, but they can also provide essential keywords for facilitating online literature reviews by other researchers. However, there appears to be little guidance within the current literature regarding the construction of this essential component of research output.

As part of this study, the authors were unable to identify any specific publication about constructing “titles” in qualitative research. However, a handout (by persons unknown) prepared by the University of Manitoba gives instructions for “writing a great title” for student assignments and essays. According to this handout, titles are considered important for the reader as they summarise the argument/content contained within the essay. In addition, they are important to the writer as “the title is a ‘privileged’ place in the text, because it draws the reader’s most complete attention. Take advantage of that fact. Use the title’s privileged position to express your ideas clearly and to articulate your argument” (Anon, n.d). The subsequent section on “what a great title looks like” includes the advice that this should consist of a main title and a subtitle, separated by a colon. In addition, the University of Manitoba suggest that a main title should consist of a “catchy phrase, quote or clever hook” (ibid) followed by a subtitle that provides an “informative phrase, including details about the content, organization and method(s) used” (ibid). In short, based on this information, it would appear that advice given to neophyte researchers within the academy is such that it actively advocates the use of quotes in the construction of “great titles.”

For this study, content analysis of research papers featuring respondent quotes as titles sought (i) to answer a series of questions connected to the methodological principles of qualitative research, and (ii) to determine if the use of the respondent quote as title reflected the methodological principles underlying the relevant author’s work. A purpose of this article is therefore to redress the advice given regarding the construction of “catchy titles” without first considering the methodological implications of such an approach.

Methods

The original aim of this study was to identify 40 research papers that featured respondent quotes as titles and which had been published in English-language, peer-reviewed health/social science journals during the period 1980–2019. The rationale for this sampling design was concerned with making critical assessments of the way in which authors and journal editors had incorporated respondent quotes into research titles and publications, respectively, throughout a forty-year period. This design also aimed to identify the way in which authors justified (or not) their use of respondent quotes as titles as part of health research that was informed by social science.

However, this study commenced in March 2020 just as “lockdown” procedures were introduced on a national scale in the UK as a result of the COVID-19 pandemic. “Lockdown” in London resulted in the authors’ host institution closing the university campus, and all staff were encouraged to work from home in attempts to reduce the spread of viral infection. Accordingly, the authors subsequently worked detached and remotely throughout the production of this article; each with limited access to the university library and its electronic resources (such as journal databases). For these latter reasons, the original sampling design was amended to become a convenience sample of journal papers which featured titles containing respondent quotes. In addition, these were papers that could be accessed from electronic resources while working remotely. Inclusion criteria were that the paper was published in the English language during the period 1980–2020 and was concerned with qualitative studies of health and grounded in any of the social science disciplines. Exclusion criteria were papers that used legitimate expressions in inverted commas (such as policy terms, technical expressions, etc.) that were evidently not verbatim quotes from respondents involved in the relevant study. Over 20 such papers were identified but were excluded from analysis based on these criteria.

Although over 40 papers were identified that matched the inclusion criteria for this study, the authors limited analysis to the initial figure stated in the original design. All 40 papers were then subject to content analysis and organized using the framework method. Content analysis (Downe-Wamboldt 1992, Parkin 2020; Elo and Kyngas 2008; Stemler 2001) is concerned with the systematic description and quantification of a given phenomenon in order to generate more generalised understanding of the issue under inquiry. The questions which guided the content analysis of all papers can be found in Table 1.

Table 1. Questions that informed content, framework and comparative analyses

1.Title of Paper (and “Is Second Clause of title suffice?”)

2.Year Published?

3.Full Journal Reference?

4.Impact Factor (date)?

5.Number of Authors?

6.Social Science Discipline?

7.Health Issue Addressed?

8.Qualitative Methods used?

9.Sample/Cohort Size?

10.Type of analysis?

11.Does quote feature as data extract in the paper?

12.Does author explain use of quote as title?

13.Does quote seem to represent all data presented?

14.Does the quote in title reflect methodology?

15.What does the quote infer by being a title?

Content analysis was further complemented by organising the responses to each question using the framework method of categorisation. Framework analysis involves the creation of a matrix (here in the form of an Excel workbook) consisting of rows and columns. The tabular organisation of such a matrix establishes “cells” into which information can be “charted” (Gale et al. 2013; Parkin 2020b, Parkinson et al. 2016) as data summaries. In this study, the individual papers formed 40 rows within the matrix, and the questions that guided content analysis (Table 1) formed the 16 columns. The intersections of each row/column therefore produced the cell into which the relevant “answer” to the question could be inserted. Accordingly, when the matrix was complete, the respective columns provided a wide range of responses and data summaries that were subject to further comparative analysis in order to identify similarities, differences and common themes throughout the dataset. It is the latter stage of analysis that typically informs the “results” of any research that follows the framework method. As such, the findings below represent those obtained from the three-levels of analyses (content, framework and comparative) attached to this research study.

Results: A descriptive overview of the 40 journal papers in this analysis

First clause: Second clause

As noted above, 40 journal papers were selected for inclusion in this study on the basis that the title of the paper contained a respondent quote as part of the title. More specifically, titles of papers were typically sentences constructed from two independent clauses separated by a colon (:), in which the first clause was an interview extract, and the second clause contextualized the quote by providing further details of the study/cohort/topic. For all 40 papers in the sample, the second clause provided enough information about the study and could have stood alone without the first clause of the title. That is, the quote in all papers was typically superfluous and was used as an illustrative device throughout the sample.

Authors

A total of 109 authors wrote/contributed to the 40 papers, with the average authorship consisting of 2.725 authors. 13 were single authored papers, 12 were written by two authors and 15 papers were written by 3 or more people (range 3–7 authors).

Year published

All 40 papers were published during the four-decade period 1980–2020. However, 32/40 were published between 2000–2020, and only 8 published during 1980–2000. Almost half (19/40) of the sample were published during the period 2011–2020, the most recent decade. Table 2 provides these details by decade.

Table 2. Publication Year

Decade published	Number of papers in sample	% of sample
1981–1990	4	10
1991–2000	4	10
2001–2010	13	32.5
2011–2020	19	47.5

Impact factor

The Impact Factor (IF) of a journal is a metric that is concerned with the number of times an average article within a given journal is cited over a specific period (typically every 2 years). In more simplistic terms, a higher IF suggests more academic prestige and journals with higher IF are those that offer opportunities of higher citation rates as a result. In the 40 papers of this study, the most recent (typically 2018) IF of all journals ranged from 1.09 − 27.6. Of these, 37 papers were published in journals holding an IF between 1.4–3.4. One journal had no information regarding its IF, one had an IF of 1.09, one more had an IF of 27.6.

Journals

The 40 papers in this study were sourced from a wide range of journals (from the well-established to less well-established). All journals were dedicated to physical/mental health and the social sciences. Table 3 provides a summary of the 18 journals from which the 40 papers were obtained.

Table 3. Journals in this review

Addiction Research and Theory British Medical Journal British Journal of Criminology Drug and Alcohol Dependence Family Practice Health and Social Care in the Community Health Expectations Health, Risk and Society International Journal of Older People Nursing International Journal of Qualitative Studies on Health and Well-being Journal of Addiction Medicine Journal of Advanced Nursing Journal of Asthma Journal of Public Health Journal of The Association for Persons with Severe Handicaps Qualitative Research Social Science and Medicine Sociology of Health and Illness

Health issue addressed

Papers subject to analysis were located mainly in the fields of public health (13/40) and psychology (6/40). Of the former, 7/13 were concerned with studies of substance use (and/or its management) and the latter addressed issues such as trauma, identity construction and stigma management. Other papers were located more broadly within primary care (5/40) and secondary care (5/40), and each addressed health issues in the relevant environments (e.g., community and hospital setting and tertiary care). Other papers addressed health issues from the perspective of specific groups. These included studies of lay participation in health studies (4/40), gender-specific health issues (3/40), disability (3/40) and congenital conditions (1/40) (See Table 4).

Table 4. Field of Health

Discipline/Field	Number of Papers Analysed
Public Health	13
Psychology	6
Primary Care	5
Secondary Care	5
Public Patient Involvement	4
Gendered Health	3
Disability	3
Congenital Conditions	1
Total	40

Qualitative methods used

35/40 papers involved research that had been informed by face-to-face interviews with respondents. Of these, 15/35 used semi-structured interviews as the main qualitative method, or in combination with other methods of research (such as ethnographic methods, focus groups and participatory research). 20/35 studies used other forms of interview techniques defined as structured/unstructured interviews, iterative interviews and/or in-depth interviews. Throughout the sample of 40 papers, a wider range of other qualitative methods were noted which also included audio/video recorded consultations, participant observation, and diary recordings. In some studies, qualitative research methods supplemented quantitative research as mixed methods studies of health issues (3/40).

Cohort/sample size

The cohort range (per paper) was 3 people (in a case study project) to 139 people. Most notably, 15 papers reported on cohorts/samples involving 21–40 respondents (474 respondents in total) and 10 papers had cohorts of 41–60 people (510 respondents in total). Two papers reported on research involving 61–100 people (totaling 138 respondents), and three papers involved cohorts of over 100 people (387 respondents in total) (see Table 5). Although some of these cohorts (especially those involving over 20 people) involved mixed groups of respondents (patients, professionals), it is noteworthy that the average sample/cohort across all research studies was approximately 42 people.

Table 5. Cohorts across 40 research projects

Cohort/Sample Size	Number of Papers	Total Respondents
0–10	4	30
11–20	6	112
21–40	15	474
41–60	10	510
61–100	2	138
100+	3	387
Total	40	1651

Type of qualitative analysis

In 18/40 papers, thematic analysis was reported as the primary form of data analysis. Grounded theory was the analytical approach applied in 8/40 studies. Several papers (7/40) made reference to a variety of combined forms that included inductive and deductive analyses that were used iteratively and sequentially. Other forms of analysis used in this manner included discourse analysis, phenomenological analysis and conversation analysis. One study also referenced the use of iterative categorisation and one other study conducted an analysis of responses to one key question asked during all research interviews.

7/40 papers did not report of any analytical framework attached to the relevant study.

Does the quote used in the title also feature as a data extract in the paper?

For 24/40 papers, the quote in the title also featured in the main text as part of a larger data extract that was typically located within the Findings section. In most of these papers (19/24), the relevant authors usually provided accompanying information (such as participant identification number, pseudonym, gender and/or age) that added contextual detail to the quote. The respondent quote used as a title was often presented as one of several findings reported throughout the paper. That is, the title of the paper did not typically illustrate all of the key findings reported in the paper.

Additionally, in 5/24 of these papers, the title quote was found to be an edited version of what was actually said when compared to the (unedited) data extract contained within the main text. In this regard, the relevant authors had paraphrased the verbatim text yet included those words in quotation marks to give the appearance they were in vivo statements (extracted from interview). Furthermore, no explanations were offered by the relevant authors why the “responses” had been edited in this way to establish the first clause of the relevant research titles. In one paper, the authors used a verbatim statement generated from another study to headline their work – and made no formal acknowledgment to the relevant researchers concerned (i.e., those who conducted the original study) nor provided any explanation of the decisions underlying the appropriation of this quote in their text.

In 16/40 papers, the quotes contained within the first clause of the title were completely absent from the main body of the paper. In this regard, no further information was provided about the title quote regarding its origins, attribution and its relation to the actual data that were analysed. Two exceptions to this finding were noted. One paper provided an explanatory footnote regarding the selection of song lyrics chosen as a title (which had the outward appearance of being a verbatim expression due to the presence of quotation marks) in which they also provided the bibliographic details of the relevant artist/recording. In another study, it became apparent that the “respondent quote” used as title was actually a paraphrase of a key question asked during the study. In this latter case, the quote appears to have been a “re-imagining” of how the researcher interpreted the relevant responses provided by the people interviewed.¹

Do authors explain the use of respondent quote as title?

The majority of papers (34/40) involved in this study did not include any explanation why the respective authors had used respondent’s quotes as a title for their research paper. In only 4/40 papers the authors explained the choice of the quote as the title of the paper, usually in the form of a footnote. In 2/40 papers, the relevant authors provided a partial explanation why the respondent quote was selected, but this was in a discursive form in which the explanation was inferred rather than made explicit.

Does title quote appear to represent all data presented in the study?

Only 4/40 papers appeared to use respondent quotes as a title in a way that seemed to be a fair representation of the relevant study and concomitant cohort group. In these studies, the respondent quote as title related to “case study” projects involving small groups of people per study (ranging from 3–7 participants).

Conversely, 36/40 papers used respondent quotes as titles that did not represent the full range of views/experiences of the wider cohort in all 36 studies. In many papers, the title quote did not represent the wider findings of the study. In short, all 36 title quotes typically reflected the view of one person (representing the relevant cohort) regarding an opinion of one facet of multiple findings described in the research.

In addition, the overall lack of representation (inferred by the title quote) noted throughout 36/40 papers took several forms. Namely, representation was either “individualised,” “selective,” “appropriated,” “disconnected,” or “unexplained.”

“Individualised representation” involved the use of specific words spoken by one respondent during an interview to summarise a specific health experience or phenomenon under investigation. In this regard, 14/40 research papers were titled with views that appeared to represent the more idiosyncratic, emotional and/or marginal experiences of one individual from a wider cohort. Accordingly, the collective views and experiences of the latter were diminished as they were subsumed by the more extreme experiences described in the words provided by one individual. The quote used as a title in these instances therefore brought to the fore the more marginal and less representative experiences of the wider views and experiences generated from the relevant studies. Indeed, in at least two papers that adopted individualised representation as research titles, the authors (somewhat paradoxically) also acknowledged the limitations of “generalising from a cohort of 30 people” and another stated that the research title was the result of one person’s response to a specific question.

“Selective representation” was noted in those qualitative studies that compared the views and experiences of heterogenous groups/populations affected by common health concerns/experiences. Papers of this nature included comparative studies of services from the perspective of structure (professional staff within institutions, organisations, services) and agency (service users, clients, patients). Selective representation therefore involved the use of a specific quote (as title of paper) from only one of these groups. As such, the selective use of a quote from one person, was located within a specific sub-group from a study involving mixed cohorts. Therefore, the quote selected as title not only (mis)represented the mixed cohort groups of the study, but also represented an individualised view as described above (as it was only one view, of one person situated within a specific sub-group of the wider study). Selective representation of respondent quotes was identified in 12/40 research papers.

“Appropriated representation” involved the use of quotes that had been “acquired” from other sources not directly connected to the primary research of the authors. For example, one paper used a song lyric, one used a popular English language expression (that was not accredited to any respondent in the study), one involved the paraphrasing of a respondent’s words to establish an “artificial” quote and one paper even used the words of a respondent from a differentstudy as the titles of the relevant research papers. In only one of these four papers did the relevant author explain and justify the use of “appropriated” representations in their work. However, the use of these “quotes” as titles gave the outward impression to the reader that they had been generated by the qualitative research reported within each article.

“Disconnected representation” involved the use of quotes as titles that did not appear connected to relevant qualitative dataset. In these papers (3/40), the relevant authors appear to have paraphrased a respondent quote that seems to have better reflected the qualitative methodology attached to the paper, rather than reflect the views and experiences of those they interviewed. As an example, one researcher, who had shared experience of the relevant health issue, paraphrased a respondent’s view in order to reflect the researcher’s own epistemological understanding of the same topic.

A further 3/40 used quotes as titles of research papers in an “unexplained” manner. That is, throughout each of these papers, the use of the quote was not contextualised nor explained in any part of the research report. It is therefore not possible to comment on the authenticity of these quotes as the relevant authors did not indicate their source.

Analysis of the quote used (indication of author motivation/subjectivity)

A final component of content analysis aimed to identify any motivation by the authors in selecting the relevant respondent quotes to headline their work. This aspect of analysis involved comparing the overall conclusion and/or recommendations attached to the research alongside the authors’ affiliation, role and, where possible, source of funding for the study. In the absence of any explanation for using the quote as a title (by the authors), this aspect of analysis aimed to identify why authors may have selected specific interview extracts as headline statements for their work. Four domains of author motivation/subjectivity emerged from this process. These four domains are here termed “possible investigator bias,” “shock appeal,” “political” and “unexplained.”

Possible investigator bias

The most frequent (13/40) motivation for using quotes as research titles appeared to be connected to “possible investigator bias.” This form of subjectivity appeared to use respondent statements as a literary device that typically reinforced the remit of the academic institution/department that conducted the study, the academic role and position of the author, and in some cases, the agenda of the organisation that funded the study. In each of these cases, the use of quotes as titles reflected these various epistemological orientations of the study in the overt presentation of a singularly positive, or singularly negative message. Various forms of possible investigator bias were noted, including selection bias, confirmation bias, and interpretation bias.

Selection bias involved emphasising a singular point of view that was overtly hyperbolic/rhetorical/tautological and possibly represented the most extreme experience within a given dataset. However, within the same studies, these extreme views were typically contradicted by other respondents who voiced less extreme opinions within the main text. Similarly, selection bias involved the use of specific experiences from within mixed cohort groups (e.g., male vs. female; child versus adult; youth vs. older people, etc.) or purposely selecting a singular homogenous experience from studies that involved heterogenous groups (relating to gender, professional background, nonprofessional background, etc.).

Confirmation and interpretation biases were each evident in various authors’ use of quotes as titles when considered alongside particular theories, paradigms or epistemological orientation of the relevant researchers. For example, some titles reflected and reinforced a point of view that succinctly crystallised aspects of social theory, a particular research method or even a way of living that was antithetical to the raison d’etre of the relevant research center. In these latter examples, the authors appeared to deliberately highlight a cause of concern, relevant to their research interests (such as public health), by highlighting (and confirming) negative experiences reported by respondents that reinforced the need for specific policies/agendas.

Indeed, many of the titles informed by possible investigator bias typically read as slogans such as those published by health promotion agencies. In addition, many titles in this category had a quasi-moralistic tone to the title, in which the researchers implied (through the words of a research respondent) what is (in)appropriate and (un)acceptable behaviour/lifestyle choices within the relevant field of health.

Shock appeal

The second most frequent (12/40) motivation for using verbatim quotes appeared to be one concerned with an overt intention to “shock” the audience using verbatim extracts from respondents. Such quotes included extreme profanity, or expressions that were deeply emotive and/or thought-provoking. In purposely selecting this type of challenging statement as the research headline, the relevant authors appear to have used these extracts mainly as emotional, confrontational, provocative and/or adversarial devices. In most cases, “shock appeal” also appeared to have been associated with research conclusions/recommendations that advocated a change/restructuring or introduction to certain health policies and practices. However, this aspect of subjectivity also had the (unintended) consequence of establishing titles that were more akin to the tone of tabloid journalism or similar forms of non-academic reportage.

Political

In 10/40 papers, the use of title quotes (of respondent) reflected a more political agenda (of researchers), in which the former supported/condemned government policy or supported the aims and objectives of an (overtly political) funding organization. These respondent statements as titles also included sentiments that reflected well-known, established failings within given health-care systems. Other title quotes in this domain also reflected the views of advocacy services and the remit/agenda of those organisations (for example, one paper included a staff member of such an organisation as a co-author).

Unexplained

A final category of subjectivity is here termed “unexplained,” in which analysis was unable to identify any obvious motivation attached to the use of a quote as a research title (noted in 5/40 papers).

Discussion

This critical analysis of respondent quotes used as research titles highlights a variety of issues relating to the (mis)conduct and (mis)management of qualitative research per se. Namely, in purposely selecting a singular quote from a single respondent to headline a research paper, authors appear to be making a (conscious/unconscious) decision that is antithetical to some of the core elements of qualitative research, whether as a paradigm or as a form of social science inquiry. In addition, the purposive selection of a single “soundbite” to highlight an entire corpus of research indicates (i) a lack of methodological rigor attached to the study, (ii) incomplete analysis, (iii) an overall mis-representation of the wider qualitative dataset, and (iv) possible investigator bias.

A lack of methodological rigor

As noted earlier, the methodological design of a qualitative study will normally seek to make intellectual and empirical links between any philosophical orientation of the researcher, the concomitant commitment to a given research tradition, and include thorough explications how the researcher made interpretive connections between data and theory. These are regarded as essential, fundamental and basic requirements of any research study that adopts of qualitative mode of inquiry.

However, over one-third (14/40) of studies in this analysis did not provide details regarding the methodological design of the relevant qualitative research. Instead, these studies typically provided a summary account of the qualitative “methods” employed during data generation. Nevertheless, none of these studies attempted to explicate the intellectual decision and/or methodological connection regarding the deliberate selection of a singular sentence provided by one individual to headline and signal the findings of a study involving many other participants.

As such, these flaws in acknowledging and applying methodological principles attached to qualitative research lead to basic errors in the interpretation and presentation of data chosen to headline qualitative research reports. To simplify this criticism, just as one would not select an individual response to an open question contained within a survey instrument completed by 1,000 respondents to summarise a quantitative study; then, one would similarly not expect to choose an individual statement from a study involving 40 participant to similarly summarise a qualitative study. This is due to the methodological issues of epistemology attached to each research paradigm (positivism in the former and interpretivism in the latter) in which the respective “route map” of each approach requires the researcher to amass a corpus of data from which conclusions may then be drawn from that collected whole. Accordingly, decisions relating to the deliberate selection of one part of that whole (regardless of epistemological paradigm) to headline the study immediately indicates a lack of methodological rigor attached to the relevant study.

Incomplete analysis

As noted above, qualitative data analysis is an important key element that is integral to the methodological design of a qualitative study. The form and foci of analytical frameworks are essential components upon which qualitative research reports founded and developed. However, in this analysis of 40 papers headlined by a quote, the account of “analysis” was superficial in many and cursory throughout. This was particularly noteworthy in the 26/40 studies that prioritised thematic analysis or grounded theory as the primary forms of data analyses. In each of these respective studies, the relevant authors did not provide any details regarding the way in which analyses were actually conducted. Instead, short sentences such as “data was organised into themes,” (sic) or “a grounded theory approach to data analysis was employed,” typified the depth of detail devoted to explanations of data analysis. More significantly, no details were provided regarding the performance of reflexive and critical engagement with voluminous datasets or how thematic analysis/grounded theory assisted with the conceptualisation of a coding framework (and vice versa), how the data/theme were interrogated to establish theory or how dis/connections were made between categorical themes to establish confirmation or refutation of a given issue. Somewhat bizarrely, 7/40 papers did not contain any account of the form of analysis applied to the relevant study.

Accordingly, the selection of singular quotes to headline research papers appears flawed on yet further methodological bases. On this occasion, this fault relates to the amplification of findings obtained from under-developed accounts of analyses, and associated coding processes and related frameworks. Such analytical oversight, especially with regard to title composition, reflects Rhodes et al.’s (2010, 443) concerns regarding the reduction of qualitative data to “no more than supplementary illustration, with qualitative analytical practices such as induction and iteration becoming opaque.” As such, title quotes obtained from thematic analysis/grounded theory possibly provide an indication of incomplete analysis attached to the study. This form of headlining, in turn, should instantly signal concerns regarding the integrity, trustworthiness and accuracy of the research concerned. As noted by Braun and Clarke (2014), thematic analysis is useful for applied purposes and for specific target audiences. However, they also caution that this form of analysis should be used “wisely” and, by inference, should not necessarily be the form of analysis considered appropriate for those journals that prioritise methodological engagement with academic research (such as those with higher Impact Factors).

A mis-representation of the wider qualitative dataset

Almost all (36/40) respondent quotes used as titles in this critical analysis did not represent the full range of views/experiences expressed by the wider cohort in each of the respective studies. In addition, this lack of representation took several forms, including “individualised representation,” “selective representation” and “appropriated representation” in which the authors purposely and deliberately selected a comment by one individual to compose a research title that summarised the collective views of all that were interviewed.

The interpretation and presentation of qualitative data in this manner is also flawed on methodological grounds. This is because the construction of a literary device consisting of one person’s thoughts/opinions has the effect of distorting and reducing the collective experiences generated from the entire cohort into one deterministic soundbite (subjectively chosen by the relevant researcher). In making such blasé decisions regarding the representation and re-presentation of qualitative data, researchers employing this tactic are failing to attend to “form” associated with research purposes, rationale, methods and questions asked. According to Sandelowski (1998, 382), attending to these forms should not be considered a frivolous pursuit, but instead as essential in maintaining the integrity of the qualitative enterprise.

Whilst recognising that individual experiences and accounts of phenomena are important (whether those are marginal or shared throughout a given cohort), it is equally necessary to acknowledge that “one narrative size does not fit all” (Tierney 1995, 389). Similarly, on the topic of representation, Kohler (2016) quite categorically states:

authors need to strike a very fine balance between presenting their data (i.e. showing) and interpreting that data (i.e. telling). Authors need to ensure that they present “parts” of their story or their data in a way that the “whole” of their story or findings also makes sense. This means that authors need to keep in mind the big picture that they want the reader to understand while they are providing evidence for the pieces that make up the big picture .

Accordingly, the decision to “show” a specific data extract as a title of a research paper immediately signals to the audience that an imbalanced interpretation of the complete dataset has taken place. Likewise, in purposely prioritising the views of one person, the author simultaneously diminishes any polyvocality (Richardson 1994, 521) associated with the wider cohort and contained within its collective experience. Failing to attend and recognise these forms attached to representation and re-presentation is a significant methodological error. Indeed, Sandelowski’s (ibid) warning regarding the representation of qualitative data could equally apply to the construction of titles using quotes, as she notes, “there is nothing more wasteful of both researchers' and participants’ efforts than having important findings lost in clichéd, overwrought, and disorderly prose.’

Investigator bias

The final issue relating to the (mis)conduct and (mis)management of qualitative research relates to the identification of various forms of (conscious and/or unconscious) bias and subjectivity made (implicit or explicit) in an author’s use of quotes as titles. This study identified various forms of subjectivity and bias that appeared to reinforce various political, institutional and epistemological agendas of researchers and their research settings (including funding agencies). In short, the use of quotes as titles appeared to provide a suitable and convenient soundbite that reinforced and legitimised the study, research positions and structural issues. For these reasons, it may be suggested that respondent quotes used in this way may be a tactic in which authors are able to subliminally project their own views upon a particular audience. This therefore poses a significant challenge to the presentation of qualitative data and possibly confirms Sandalowski’s (ibid.) view that some “readers may find it difficult to determine when participant or researcher is speaking.”

A valid counter-position to this criticism may be that all research is political, that subjectivity (and bias) are inevitable and that the relevant authors are adopting Lingard’s (2019) “argument principle” in the use of literary devices that are purposely rhetorical and political in tone. Such an approach, it may be suggested, serves to use data to advance and advocate a particular experience, to give voice to the voiceless and to highlight marginality. Whilst these may be noble defenses,² when used as an approach to headline research papers, they nevertheless sit slightly uncomfortably within the parameters of the core elements that define the qualitative research paradigm. This discomfort relates to the way in which prominence of a single utterance can obfuscate an entire dataset, distort polyvocality and misrepresent entirely “other” experiences located within the overall collective experience. Such obfuscation, distortion and misrepresentations, as also noted by others (Morrow 2005; Weiss 2004), pose a major threat to the overall validity of qualitative studies.

For example, the over-emphasis and over-interpretation of a single statement may cast further doubt within skeptical audiences about the origins of those interpretations and representations of data. These latter concerns, it is worth remembering, were central to Corden and Sainsbury (2006) study of researchers’ use of verbatim quotes within research reports. As they too noted, qualitative data may be persuasive and convincing to those familiar with qualitative methodology. However, they equally infer that such data may be mysterious and enigmatic to those less familiar with qualitative methodology and methods. As such, qualitative researchers need to be mindful of the ways and means in which data are presented, especially in terms of providing “evidence,” initiating conclusions and advocating recommendations from their studies. Unfortunately, the deliberate selection of rhetorical, hyperbolic and shocking soundbites as research titles do little to demystify the enigmatic qualities of interview data. Indeed, such practice may perpetuate misunderstandings of qualitative research, in turn, casting doubt upon the overall credibility, reliability, validity, veracity, and integrity attached to the wider conduct and management of qualitative research per se.

Conclusion and recommendations

To conclude, the authors of this article contend that the use of respondent quotes as titles of research papers is not good research practice and is antithetical to core elements of qualitative research. Indeed, the use of respondent quotes as a research title possibly speaks more of the agenda of the relevant researcher/s than it does of the study concerned. For these reasons, the authors here conclude that the practice of headlining research reports with verbatim quotes should be actively discouraged (and challenged) throughout all levels of the academy.

For example, at an undergraduate level, “Methods and Analysis” modules embedded within social science degree courses could easily incorporate into existing seminars/lectures the ways to effectively present qualitative research findings without resorting to devices more associated with trite tabloid journalism. Similarly, research supervisors of postgraduate (especially doctoral) students could likewise challenge the use of verbatim quotes as research titles via discussion/debate within supervisory sessions/tutorials (especially with those who may be tempted to use such devices). Post-doctoral students and Early Career Researchers should also be discouraged from this practice, especially in the preparation of manuscripts pertaining to their doctoral research or if leading on a multi-authored post-doctoral research report. In the latter situation, more experienced colleagues within the relevant academic department/authorship team may be able to discourage such practice and suggest alternative titles as part of co-authorship.

Further recommendations relate to the academic industry of manuscript submission and publication. Journals, their editors and peer-reviewers are each ideally placed to discourage a practice that undermines “good quality” qualitative research/ers. For example, peer-reviewers could proactively challenge the authors of those papers they are invited to review that are headlined by a single interview extract. The peer-review process may request the relevant author to explain, legitimise and validate the decisions underlying the selection of a specific quote to headline the manuscript. Similarly, editors of journals could adopt an equally proactive role in scrutinising and challenging those papers headlined by verbatim quotes. This could also take the form of requesting authors to explain the rationale for such practice at some point in the main body of the paper submitted (and not relegated to an obscure footnote). Editors could also introduce more stringent means of monitoring the ever-increasing use of quotes as research titles now appearing in academic journals. For example, the “Instructions for Authors” section of the relevant journal homepage could include a set of guidelines regarding the use of respondent quotes as research titles. Similar guidelines and protocols already exist for many journals regarding the content, format and style of qualitative research papers per se (e.g., see Kohler 2016; Neale, Miller, and West 2014; Neale and West 2015; Rhodes et al. 2010). Accordingly, journal preferences relating to titles of research papers could be easily incorporated into such guidelines.

Finally, many journals currently require authors’ manuscripts to adhere to various domains within a checklist in order to confirm comprehensive reporting of qualitative studies. These checklists have emerged in recognition that “poorly designed studies and inadequate reporting can lead to inappropriate application of qualitative research in decision-making, health care, health policy and future research” (Tong, Sainsbury, and Craig 2007, 349). One example of such a checklist is the “consolidated criteria for reporting qualitative research” (COREQ) (ibid) which has been adopted by several international journals for monitoring qualitative research submissions. COREQ provides a 32-point checklist for assuring quality (pre-peer review, pre-publication), with regard to the research team, research design, and analysis/findings. For example, COREQ Item 29 concerns the use of quotations contained within research reports and asks: “were participant quotations presented to illustrate the themes/findings? Was each quotation identified? e.g. participant number” (Tong, Sainsbury, and Craig 2007, 352). This type of checklist could therefore be adapted to include an item that involves the use of respondent quotes as titles of manuscripts (e.g., does the author explain why a respondent quote headlines this paper?). Perhaps more accurately, given the content and findings of this review, a checklist pertaining to title-writing guidelines reports should be made widely available throughout the academy. Indeed, this directive is most necessary as, to close the circle opened at the onset of this article, “we cannot afford to let qualitative research become a set of taken-for-granted precepts and procedures” (Atkinson, Coffey, and Delamont 2001).

Disclosure statement

No potential conflict of interest was reported by the authors. The views expressed in this article are those of the authors' only (and do not reflect those of any other body, institution or organisation associated with the authors).

Additional information

Funding

Andreas Kimergård, King’s College London, is supported by the National Institute for Health Research (NIHR) Applied Research Collaboration South London (NIHR ARC South London) at King’s College Hospital NHS Foundation Trust. The views expressed are those of the author and not necessarily those of the NIHR or the Department of Health and Social Care.

Notes

1. These two papers could equally have been categorized as “misleading,” due to the title “appearing” to include a respondent quote in the title. However, during analysis, it became apparent that they were not respondent quotes but were not excluded in order to maintain the sample – and reflect diversity in the use of “quotes” as titles.

2. And equally valid when applied to the use of data extracts within the main text of an article and framed in the appropriate epistemological terms