Skip to main content

The role of analytic direction in qualitative research



The literature on qualitative data analysis mostly concerns analyses pertaining to an individual research question and the organization of data within that research question. Few authors have written about the entire qualitative dataset from which multiple and separate analyses could be conducted and reported. The concept of analytic direction is a strategy that can assist qualitative researchers in deciding which findings to highlight within a dataset. The objectives of this paper were to: 1) describe the importance of analytic direction in qualitative research, and 2) provide a working example of the concept of analytic direction.


A qualitative dataset from one of the author’s research programs was selected for review. Ten potential analytic directions were identified after the initial phenomenological analysis was conducted. Three analytic directions based on the same coding template but different content areas of the data were further developed using phenomenological analysis (n = 2) and qualitative description (n = 1) and are the focus of this paper. Development and selection of these three analytic directions was determined partially relying on methodological criteria to promote rigour including a comprehensive examination of the data, the use of multiple analysts, direct quotations to support claims, negative case analysis, and reflexivity.


The three analytic directions addressed topics within the scope of the overall research question. Each analytic direction had its own central point or story line and each highlighted a different perspective or voice. The use of an inductive and deductive approach to analysis and how the role of theory was integrated varied in each analytic direction.


The concept of analytic direction enables researchers to organize their qualitative datasets in order to tell different and unique “stories”. The concept relies upon, and promotes, the conduct of rigourous qualitative research.

Peer Review reports


Reports on data analysis in qualitative research are well documented. Procedural steps have been described [1,2,3,4,5,6,7] and authors have made distinctions between the concepts of coding, analysis, and interpretation [1, 2, 8, 9]. Authors have written about different researchers accessing different representations of a topic or phenomenon [2, 10] or multiple interpretations being applied to the same transcript [11]. The literature on data analysis mostly concerns analyses pertaining to an individual research question and the organization of data within that research question. Few authors have written about the entire qualitative dataset from which multiple and separate analyses could be conducted and reported.

The data collected by qualitative researchers can be voluminous and often surpass the data pertaining to objectives outlined in grant proposals. These data may be compelling but analyses of some data are often given lower priority if they do not align directly with the stated objectives.

There comes a point during data collection and analysis where qualitative researchers must choose “which story, of the many stories available to them in a data set, to tell” (p. 376) [12]. According to Arthur Frank, “[a] fter the methods, there has to be a story” (p. 431) [13]. “Stories” should have a central point or storyline [12]. The final report can be told from the perspective of different voices [12] and organized by time such as emphasizing key turning points and milestones in the sequence of events studied [12, 14] or by using other forms of representation such as metaphors [2, 12]. Theory can be central or more peripheral in the account [15]. The question remains, what “story”, or “stories”, do we tell?

The concept of analytic direction

The concept of analytic direction is a strategy that can assist qualitative researchers in deciding which “stories” to highlight within a dataset. Sandelowski reports that researchers account for their data and then determine the different “paths” [1] or “analytic paths” [16] they can pursue. Others have proposed that decision-making throughout analysis implies analytic ideas at every stage of the coding process [8] and that researchers define for themselves what analytic issues are to be explored and what ideas are important [8]. Charmaz [17] reports that grounded theory researchers pursue more than one analytic direction by focusing on certain ideas first and then returning to the data to address an unfinished analysis in another area later. While the concept of analytic direction has been referenced, or alluded to, by these and other authors [1, 8, 16, 18, 19], operationalization of this concept is not well articulated. In this paper, the term analytic direction refers to a message developed by the researchers about the data that may or may not require further substantiation. An analytic direction can be presented as a single message or theme, and can stand alone or be supported by multiple sub-messages or sub-themes. Analytic directions can be developed during the coding process, in later stages of analysis, or possibly during analyses of new datasets. Relying on strategies to promote rigour can assist with the development, substantiation, and selection of analytic directions. If substantiated, each analytic direction could be the focus of an individual publication. The objectives of this paper were to: 1) describe the importance of analytic direction in qualitative research; and 2) provide a working example of the concept of analytic direction.

Why analytic direction is important

The concept of analytic direction is important because it has implications for methodological rigour. We have an obligation to conduct methodological rigourous studies [20], especially when studies require primary data collection that involves a burden to participants [21]. The author proposes that methodological rigour is embedded within, and contributes to, the concept of analytic direction. Several strategies to promote rigour that are universal to many qualitative approaches, including phenomenology, are discussed. These strategies include, but are not limited to, a comprehensive examination of the data, the use of multiple analysts, direct quotations to support claims, negative case analysis, and reflexivity. It is important to support the quality of analytic directions so that researchers can then determine which analytic directions may or may not require further substantiation. The quality of the analytic direction will also assist in determining which directions may be selected for reporting.

The relationship between analytic direction and methodological rigour

This paper focuses on the stage where data collection is considered to be complete and does not directly address how data collection, and methodological rigour related to data collection, contributes to the concept of analytic direction. The assumption is that data collection and analysis were conducted iteratively [22, 23] and that the team decided when data collection was complete, perhaps relying upon one of the various conceptualizations of saturation discussed by Saunders and colleagues [24]. A decision about saturation would not necessarily apply to any, or all, analytic directions being developed.

The author proposes that several strategies for promoting rigour assist with the development and selection of analytic directions. One aspect of methodological rigour is that authors carry out a comprehensive examination of their data [5, 25]. By thinking about, and engaging in, analytic direction, researchers are encouraged to attend to all of their data rather than attending only to data that interests them initially.

The use of multiple analysts promotes a comprehensive examination of the data [2, 26] and thus, contributes to the concept of analytic direction. Different viewpoints lead to an enrichment of the analysis and can lead to a conceptual clarification of the interpretations [2]. Multiple viewpoints can be used at the level of coding but also at the level of the larger team as data collection and analysis proceeds. Discussions about the novelty, clinical significance, and relevance [27] of the analytic directions may occur at this time and continue through to the writing of the respective manuscripts. Analytic directions are relevant if they add knowledge, or increase the confidence with which existing knowledge is regarded [28]. According to Malterud [26], engaging multiple researchers in a qualitative study strengthens the design of the study, not for the purpose of consensus or identical readings of the data but to supplement and contest each others’ statements.

The use of direct quotations to support the claims made about the analytic directions (and/or themes within) is another strategy to promote rigour [29]. Not only do quotations illustrate and clarify the results but they also demonstrate whether there is substantive evidence to support the analytic directions being proposed. In contrast, data that do not support the analytic directions (and/or themes within) should be accounted for and their exclusion justified when promoting methodological rigour [30]. Authors may refer to this as attending to negative cases [28] or deviant case analysis [25, 31]. This strategy promotes that “deviant cases” or “outliers” are not forced into categories or ignored but used instead to aid understanding or theory development [25]. For example, these cases may explain why the patterns developed from the data or the more normative behaviours are not always found in the researchers’ interpretations [25, 31].

Reflexivity is an essential component of methodological rigour [26]. Reflexivity has been described as “an attitude of attending systematically to the context of knowledge construction, especially to the effect of the researcher, at every step of the research process” [26] (p. 484). Being reflexive means being aware of your own position in producing partial knowledge [32]. The qualitative researcher acknowledges his or her personal influence on what that partial knowledge is (for example, the data collected are dependent on the interviewer’s questions and prompts). According to Eakin and Gladstone [33], knowing one’s standpoint helps one to recognize the forces that might drive certain interpretations and stifle other conceptualizations of the data. Knowledge production is also partial because it is not possible to report all interpretations of the data and therefore, the research team has to decide what to report. Researchers engaging in the concept of analytic direction are more likely to be reflexive about what they are, and are not, reporting from their datasets.


Rationale for the chosen example

The dataset chosen for this example was from a study where the author and her team identified 10 potential analytic directions based on a compilation of the memos and team discussions pertaining to analysis and interpretation of the data. The publications developed from this dataset reflected the selection of three analytic directions that focused on different content areas [34,35,36]. The same coding template was the foundation for the three publications and the timing of the reporting was ordered based on the author’s interests. The author chose the dataset as an example primarily because it was not heavily theory-laden and therefore accessible to novice qualitative researchers. The resulting publications have practical implications for clinical and health services research and the process of developing these publications could inform graduate students who are embarking on a qualitative program of research for their thesis work.

Original research funded

The goal of the original research project was to reduce the burden of illness due to fracture through improved bone health investigation and treatment. Specifically, the aim was to examine what researchers could learn from members of a patient group. The study was approved by the Research Ethics Board at Unity Health Toronto (REB# 10–371). The study team consisted of scientists, clinicians, a policy maker, and a patient representative with expertise related to bone health. Informed by the Theory of Planned Behaviour [37, 38], the team set out to examine members of a patient group to ask them about their intentions and actions toward bone health diagnosis and treatment and their experiences with diagnostic tests and treatment recommendations. All individuals (n = 28) were 50+ years old and had sustained a fragility fracture. The overall project relied on a phenomenological approach conceptualized by Giorgi and Wertz [30, 39,40,41].

We developed a master coding template of 27 broad codes that were designed to organize the data with minimal reliance on theory. The coding template was revised four times as data collection and analysis proceeded. The codes were developed from a combination of inductive and deductive codes. More specifically, inductive codes were developed from topics discussed in the interviews. Other codes were pre-specified from the overall aim of the original funded study and from the domains of the Theory of Planned Behaviour.

Development of analytic directions from the dataset

Qualitative researchers can use several strategies to develop analytic directions. The author started the organization process early in order to think about how best to maximize the data collected. Coding began after the first couple of interviews had been conducted; this is conventional advice for analysis in qualitative research [1, 2, 23, 42]. As soon as the coding process began, a document specific to analysis was created. Miles and colleagues have referred to this as “analytic memoing” [6]. This document is different from other documents in which the team discusses design features, decisions, and interview logistics related to the study. Analytic ideas were added to this document after coding and discussing each transcript. The author engaged two individuals in the coding/analysis process, as multiple analysts promote a comprehensive examination of the data [2, 26]. The author met regularly with members of the team during the process of data collection and analysis to discuss the data, interpretations of them, and different lines of inquiry. These discussions were recorded in the analysis document. Table 1 outlines the potential analytic directions considered for this paper. The 10 analytic directions were developed prior to publication of analytic direction #1. Some of these directions were posed as questions that required further analysis and substantiation. Tables were then created to help us to visualize patterns during analysis. As an example, for analytic direction #2, a table was created in which each participant was assigned a row and perceived messages from the various health care providers (for example, primary care providers and specialists) were placed in columns. Perceived messages were presented as quotations from participants. We examined the columns to compare perceived messages across provider groups for each participant and then examined the columns to compare the perceived messages within each provider group. For analytic direction #3, a table was created with each participant assigned a row and the domains of the Theory of Planned Behaviour assigned to columns. The table was populated with data in the form of quotations from each participant that we believed corresponded to each of the domains. Strategies such as matrices [5, 6] or thematic maps [42] can also be used to visualize developing patterns when presenting or organizing data.

Table 1 Potential analytic directions considered

Selection of the three analytic directions

The number of analytic directions selected likely depends on circumstances including the quality of the data, the quality of the analysis, and available resources. The research team considered the multiple analytic directions, discussing their relevance [27], novelty, and clinical significance and also the interests of the team in order to incorporate the perspectives of the different stakeholders. It was important to the author that the content of each analytic direction was bounded in that it did not overlap with the content of the other analytic directions. For example, analytic directions #2 and #3 discuss the potential influence of others in participants’ lives. However, analytic direction #2 focused on health care providers while analytic direction #3 focused on family members, friends, and colleagues of participants and specifically excluded health care providers from the analysis based on the Theory of Planned Behaviour domain “subjective norm”. In narrowing down the list of analytic directions, the author ensured there were sufficient data (quotations) to support the claims. Cases that did not fit the general results were acknowledged in order to justify their exclusion or explain why they did not fit. For example, in analytic direction #3, we examined instances where the data did not appear to fit with the Theory of Planned Behaviour and explained what happened in these instances where the model did not appear to be predictive of intentions.

The master coding template was important as it assisted with the organization of evidence for each analytic direction. The master coding template also assisted the team with the creation of tables for each analytic direction discussed. Table 2 demonstrates the relationship between the master coding template and the three selected analytic directions.

Table 2 Master coding template and relationship to the three analytic directions selected (codes are not mutually exclusive)

The impetus for analytic direction #1 [34] was based on an assumption held by the author as she was working on the research proposal. Her expectation was that members of a patient group would be patient advocates who were experts in navigating for care. She was interested in what patients could learn from members of this patient group. The analytic direction for the paper came from surprise, and subsequent disappointment, that those assumptions were not supported by the data and that members of the patient group did not all appear to be advocates and experts in navigating for care. One commonality that defined the patient group was that members appeared to be in favour of taking prescribed medication.

Analytic direction #1 included elements of both inductive and deductive analysis in that codes were developed for the master coding template from the data (inductive) but the author’s expectations also influenced how those codes were combined and how the team interpreted the data (deductive). Drawing from the literature, the term “advocacy” was equated with the theoretical concept of “effective” or “activated” consumer [43, 44]. The code “effective consumer” did not exist in the original master template, partly because we preferred to not apply theoretical labels prematurely to the data. Based on the coding template, we drew from six codes to create a table about “effective consumer” behaviours (see Table 2). Participants were then coded along a continuum between what was referred to as “few effective consumer behaviours” (patients who followed orders with minimal involvement in their care and demonstrating the least amount of advocacy) to “many effective consumer behaviours” (individuals demonstrating significant involvement in their care, those who demanded diagnostic testing and requested specific medications).

Analytic direction #2 [35] was developed concurrently with analytic direction #1. The role of theory was minimal in analytic direction #2 and perhaps implicit in the methodology of phenomenology which focuses on individuals’ experiences [23, 39]. The impetus for analytic direction #2 was our proposal that messages from health care providers might determine individuals’ strategies or behaviours that were the focus of analytic direction #1. The analysis was more inductive than that of analytic direction #1 as the team had no pre-contemplated plan to examine how messages from health care providers might determine individuals’ behaviours. In conducting the analysis, the team wondered whether conflicts about what individuals did with the recommendations they received (their actions) appeared to be due to messages perceived across, and within, health care provider groups. Health care providers discussed in the interviews included clinic staff, primary care providers, specialists, nurses, physiotherapists, and chiropractors.

For analytic direction #2, we used seven of the codes in the master coding template (see Table 2). Five of these seven codes were also used in analytic direction #1 but for very different reasons and drawing from different data within these codes. We were interested in individuals’ understanding or interpretation of recommendations by health care providers, not how individuals interacted with health care providers or what they did with information received from health care providers. In other words, we were interested in the meaning of what health care providers reportedly said to participants and not what participants did with that information.

The publication for analytic direction #3 [36] was written 3 years after that for analytic direction #1. This was the author’s least preferred paper, despite the Theory of Planned Behaviour being the theoretical framework guiding the original funded research. Analytic direction #3 involved a primarily deductive analysis where the Theory of Planned Behaviour guided the coding and analysis. Because of the restrictions of forcing exploratory data from open-ended questions into pre-defined domains, the author selected a qualitative description approach for the research design.

Contrary to memos and reflexive notes documented by the author about the potential value of this analysis and whether the team had learned anything about the application of the Theory of Planned Behaviour in the context of our study, the pursuit of analytic direction #3 became an interesting methodological exercise for a number of reasons. We collected data on several behaviours including receiving diagnostic tests, taking supplements, exercising, attending falls prevention classes, and initiating medication. The author believed that one particular behaviour had to be selected for analysis which entailed examining the data for each of the behaviours in depth. The author chose to focus on medication initiation and/or medication use because of a longstanding interest in medication use. Also, there was sufficient data to substantiate the Theory of Planned Behaviour domains in relation to medication initiation and/or medication use. The Theory of Planned Behaviour did not appear to be particularly relevant to intentions to attend a bone mineral density test and there did not appear to be sufficient data to support any one of the non-pharmacological treatment strategies mentioned. The team also had to make decisions about what counted as “perceived behavioural control”, “subjective norms”, and “attitudes” which were the three domains of the Theory of Planned Behaviour [37, 38]. In particular, participants’ discussions about medication side effects were problematic to conceptualize in reference to these domains. The team decided to code “experiences with side effects” as “perceived behavioural control” but “anticipated side effects” as an “attitude”.

For analytic direction #3, the team drew from five codes, three of which were pre-specified prior to analyzing the interviews and meant to capture the domains of the Theory of Planned Behaviour. The code “attitude to BMD testing” and “attitude to bone health treatment” were existing codes based on the Theory of Planned Behaviour. The code “subjective norm” was not part of the coding template because the team believed it was too specific. We instead examined the code “social influence” which captured a broader array of information about peers such as family members and friends. Similarly, “perceived behavioural control” was not part of the coding template because we found it too specific. Information for this domain was taken from another code labelled “bone health treatment” which captured data pertaining to participants’ medications, including past behaviour with medication and how difficult it was, or not, to take the medication. The code “intentions” was an existing code.


The three selected analytic directions varied in how the team used an inductive and deductive approach to analysis [15, 45] and how the role of theory was integrated (“central” vs. more “peripheral”) [15]. Each publication was within the scope of the overall research goal or question. As proposed by Agee [46], this overall question offered the potential for more specific questions during analysis. Finally, each publication had its own central point [12] and highlighted a different perspective or voice [12].

The following is a summary of the three analytic directions labelled with the first few words of the titles of each publication (see Table 3).

Table 3 Three analytic directions selected

Analytic direction #1 (Strategies used by a patient group; inductive and deductive-driven)

In this publication, we examined the strategies described by three groups of individuals: individuals demonstrating few effective consumer behaviours, individuals demonstrating many effective consumer behaviours, and individuals demonstrating both types of behaviours. We discussed how the continuum was contrary to our expectations of what behaviours members of a patient group would exhibit. Having acknowledged this finding, we reported that more than half of the participants described effective consumer behaviours including making requests of health care providers for referral to specialists, bone mineral density tests, and prescription medications. Our overall message was that members of a patient group described a range of effective consumer behaviours that could be incorporated as skill sets in post-fracture interventions.

Analytic direction #2 (Perceived messages about bone health; inductive-driven)

In this publication, we described the perceived messages across the different provider groups and then the perceived messages within each provider group. We reported that participants perceived that specialists were more interested in their bone health than general practitioners and that very few messages about bone health were perceived from other health care providers. We also reported that perceived messages about one’s bone health and recommendations for management across provider groups were inconsistent (for example, with regard to medication initiation). The message for analytic direction #2 was that patients perceived inconsistent messages within, and across, various healthcare providers, suggesting a need to raise awareness of bone health management guidelines to providers.

Analytic direction #3 (Theory of Planned Behaviour explains intentions to use medication; deductive-driven)

In this publication, we described the data in each domain of the Theory of Planned Behaviour and the apparent relationship between these domains and participants’ intentions with regard to medication use. Our message was that the Theory of Planned Behaviour appeared to be predictive of intentions to take prescribed medication in approximately three-quarters of participants and when it was not predictive, a positive attitude to medication was the most important domain in determining participants’ intentions.


This working example of analytic direction resulted in three publications highlighting distinct "stories”. The publications differed in a number of ways. Each publication had its own central point or story line [12]. The role of theory [15] was minimal in analytic direction #2 but was more central in analytic directions #1 and #3 with the concept of “effective” or “activated” consumer and the Theory of Planned Behaviour dominating the analyses, respectively. Acknowledging that the authentic voices of participants may always be manufactured by the authorial account [32, 47], all papers were written from the perspective of “I” or “we”. However, we focused on participants at the forefront for analytic direction #1 and we focused on participants’ perceptions of their providers’ voices for analytic direction #2. For analytic direction #3, the voice of the research team dominated as we struggled with methodological decisions. It is proposed that the voice of the model (Theory of Planned Behaviour) also dominated in analytic direction #3.

One implication related to analytic direction is that the research team may need to modify elements of the original research design to better suit the analytic direction selected. If such a modification is made, the team should ensure theoretical consistency in how the methods and methodologies are integrated [48, 49]. For example, Crotty [49] proposes that theoretical consistency is needed between methods, methodology, theoretical perspective, and epistemology because these four elements inform one another. Similarly, Carter and Little [48] argue that consistency between methods, methodology, and epistemology contribute to the rigour of a qualitative study. Authors should demonstrate that elements of their theoretical perspectives and research design are compatible if they are applying another methodological approach to the data. Carter and Little [48] suggest that methodologies can be combined or altered if the researcher retains a coherent epistemological position and justifies the choices made. In the funded grant, a phenomenological program of research was proposed and the data were collected through in-depth interviews conducted from a phenomenological perspective. Analytic direction #3 was not purely consistent with a phenomenological approach because of the restriction to force exploratory data into domains of a theoretical framework and so we pursued this analytic direction with a different approach (qualitative description). As pointed out by Sandelowski [50], using phenomenology and qualitative description in this way is not to be confused with misuses of methods or techniques. Unlike quantitative research, qualitative research is not produced from any “pure” use of a method, but from the use of methods that are variously textured, toned, and hued [50]. According to Sandelowski [50], qualitative description can be used in conjunction with phenomenological research in a number of ways. For example, phenomenological analyses can be applied to qualitative descriptive studies [50]. However, the pursuit of other approaches to analysis, such as grounded theory or a participatory action approach, might lead to epistemological tensions if the original study design and data collection was guided by a phenomenological approach. Future discussion about the concept of analytic direction when considering theoretical and methodological positions that differ epistemologically from the original design and conduct of the study is needed.

There are a number of other implications related to the concept of analytic direction. Practically, it is advised that researchers start to think about analytic directions early so that they are aware of the potential analytic directions being developed as soon as data collection and analysis begin. By thinking about the “larger picture” at this early stage in the research, the team is better equipped to make the most of the data collected. Having said this, one will likely never use the entire dataset. As researchers, we rarely have sufficient funds or personnel to pursue all analytic directions. Data are often set aside because researchers are eager to analyze data collected for new projects or pressured to seek future funding opportunities. Analytic directions that are not pursued can be transferred to student projects. Alternatively, it is possible to draw on a sub-set of the transcripts/observations to carry out a secondary analysis. The author has developed subsequent analytic directions that span across studies and draw from a subset of transcripts for several secondary analyses [51,52,53]. Analytic directions can also contribute to ideas for new grant proposals that enable the researcher to generate more data on analytic directions that need further substantiation and further exploration.

This paper demonstrates some guidance about how to bound each analytic direction. Bounding the analytic direction is necessary so one does not re-use the data or produce multiple, yet quite similar, papers on the same topic. Researchers are encouraged to be open and transparent and acknowledge related publications so reviewers and other audiences reading the work are able to determine for themselves that the analyses are different.

There are ethical considerations in developing an analytic direction or framing the analytic direction in a way that might be different or supplementary to the original design. It is not always feasible to obtain subsequent consent from participants for use of the data if this use differs from that of the original goal of the study. As a result, analytic directions pursued should be within the scope of the approved research ethics application. One strategy is to keep the study goal or aim broad in the research ethics submission so that it encompasses many topics that might be discussed during data collection. Another consideration is to not prematurely close a research ethics application because researchers may be able to use the data for a secondary analysis at a later date.

This paper makes novel contributions to qualitative research methodology by demonstrating how the process of analytic direction works, by operationalizing the concept and providing an example, and by describing the connection between analytic direction and rigour. This paper further contributes to the advancement of rigour by demonstrating how the development and selection of analytic directions relies on several strategies to promote rigour, such as a comprehensive examination of the data, the use of multiple analysts, providing quotations to support claims made, checking for negative cases, and reflexivity.


In conclusion, the concept of analytic direction enables researchers to organize their qualitative datasets in order to tell different and unique “stories”. The concept relies upon, and promotes, the conduct of rigourous qualitative research. As with all elements of qualitative analysis, researchers are encouraged to think about the role of analytic direction as soon as data collection commences.

Availability of data and materials

The datasets generated and/or analysed during the current study are not publicly available due to participants not consenting to having their data deposited in a public dataset but are available from the corresponding author on reasonable request.


  1. Sandelowski M. Qualitative analysis: what it is and how to begin. ResNurs Health. 1995;18:371–5.

    CAS  Google Scholar 

  2. Kvale S. Interviews: an introduction to qualitative research interviewing. Thousand Oaks: Sage Publications; 1996.

    Google Scholar 

  3. Saldana J. The coding manual for qualitative researchers. Los Angeles: Sage Publications; 2009.

    Google Scholar 

  4. Miller WL, Crabtree BF. The dance of interpretation. Doing qualitative research. Newbury Park: Sage Publications; 1999.

    Google Scholar 

  5. Spencer L, Ritchie J, O'Connor W. Analysis: practices, principles and processes. Qualitative research practice: a guide for social science students and researchers. Los Angeles: Sage Publications; 2003. p. 199–217.

    Google Scholar 

  6. Miles MB, Huberman AM, Saldana J. Qualitative data analysis. 3rd ed. Los Angeles: Sage Publications; 2014.

    Google Scholar 

  7. Crabtree BF, Miller WL. A template approach to text analysis: developing and using codebooks. Doing qualitative research, vol. 3. Newbury Park: Sage Publications; 1992. p. 93–109.

    Google Scholar 

  8. Coffey A, Atkinson P. Making sense of qualitative data: complementary research strategies. Thousand Oaks: Sage Publications; 1996.

    Google Scholar 

  9. Kelly M. The role of theory in qualitative health research. Fam Pract. 2010;27:285–90.

    Article  PubMed  Google Scholar 

  10. Malterud K. Shared understanding of the qualitative research process. Guidelines for the medical researcher. Fam Pract. 1993;10(2):201–6.

    Article  CAS  PubMed  Google Scholar 

  11. Slaughter S, Dean Y, Knight H, Krieg B, Mor P, Nour V, et al. The inevitable pull of the river's current: interpretations derived from a single text using multiple research traditions. Qual Health Res. 2007;17(4):548–61.

    Article  PubMed  Google Scholar 

  12. Sandelowski M. Writing a good read: strategies for re-presenting qualitative data. Res Nurs Health. 1998;21:375–82.

    Article  CAS  PubMed  Google Scholar 

  13. Frank AW. After methods, the story: from incongruity to truth in qualitative research. Qual Health Res. 2004;14(3):430–40.

    Article  PubMed  Google Scholar 

  14. Sandelowski M. Time and qualitative research. [review] [31 refs]. Res Nurs Health. 1999;22(1):79–87.

  15. Sandelowski M. Theory unmasked: the uses and guises of theory in qualitative research. Res Nurs Health. 1993;16:213–8.

    Article  CAS  PubMed  Google Scholar 

  16. Sandelowski M. “To be of use”: enhancing the utility of qualitative research. Nurs Outlook. 1997;45:125–32.

    Article  CAS  PubMed  Google Scholar 

  17. Charmaz K. Constructing grounded theory: a practical guide through qualitative analysis. London: Sage Publications; 2006.

    Google Scholar 

  18. Thorne S. Metasynthetic madness: what kind of monster have we created? Qual Health Res. 2017;27(1):3–12.

    Article  PubMed  Google Scholar 

  19. Sharp EA, GD DC. What does rejection have to do with it? Toward an innovative, kinesthetic analysis of qualitative data. Forum Qualitative Sozialforschung/Forum. Qual Soc Res [On-line Journal]. 2013;14(2):1–12.

    Google Scholar 

  20. Streiner DL, Norman GR. PDQ epidemiology. 2nd ed. St. Louis, Missouri: Mosby; 1996.

    Google Scholar 

  21. Ulrich CM, Wallen GR, Feister A, Grady C. Respondent burden in clinical research: when are we asking too much of subjects? IRB. 2005;27(4):17–20.

    Article  PubMed  Google Scholar 

  22. Polkinghorne DE. Language and meaning: data collection in qualitative research. J Couns Psychol. 2005;52(2):137–45.

    Article  Google Scholar 

  23. Schwandt TA. Dictionary of qualitative inquiry. 2nd ed. Thousand Oaks: Sage Publications, Inc.; 2001.

    Google Scholar 

  24. Saunders B, Sim J, Kingstone T, Baker S, Waterfield J, Bartlam B, et al. Saturation in qualitative research: exploring its conceptualization and operationalization. Qual Quant. 2018;52:1893–907.

    Article  PubMed  Google Scholar 

  25. Lewis J, Ritchie J. Generalising from qualitative research. In: Ritchie J, Lewis J, editors. Qualitative research practice: a guide for social science students and researchers. London: Sage Publications; 2003. p. 263–86.

    Google Scholar 

  26. Malterud K. Qualitative research: standards, challenges, and guidelines. Lancet. 2001;358(9280):483–8.

    Article  CAS  PubMed  Google Scholar 

  27. Giacomini MK, Cook DJ, for the Evidence-Based Medicine Working Group. Users’ guides to the medical literature XXIII. Qualitative research in health care B. what are the results and how do they help me care for my patients? JAMA. 2000;284(4):478–82.

    Article  CAS  PubMed  Google Scholar 

  28. Mays N, Pope C. Quality in qualitative health research. In: Pope C, Mays N, editors. Qualitative research in health care. Malden: Blackwell Publishing; 2006. p. 82–101.

    Chapter  Google Scholar 

  29. Dixon-Woods M, Shaw RL, Agarwal S, Smith JA. The problem of appraising qualitative research. Qual Saf Health Care. 2004;13:223–5.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  30. Giorgi A. Concerning a serious misunderstanding of the essence of the phenomenological method in psychology. J Phenomenol Psychol. 2008;39:33–58.

    Article  Google Scholar 

  31. Silverman D. Qualitative research: issues of theory, method and practice. 3rd ed. London: Sage Publications Ltd; 2011.

    Google Scholar 

  32. Finlay L. Negotiating the swamp: the opportunity and challenge of reflexivity in research practice. Qual Res. 2002;2(3):209–30.

    Article  Google Scholar 

  33. Eakin JM, Gladstone B. “Value-adding” analysis: doing more with qualitative data. Int J Qual Methods. 2020;19:1–13.

    Article  Google Scholar 

  34. Sale JEM, Cameron C, Hawker G, Jaglal S, Funnell L, Jain R, et al. Strategies used by an osteoporosis patient group to navigate for bone health care after a fracture. Arch Orthop Trauma Surg. 2014;134:229–35.

    Article  PubMed  Google Scholar 

  35. Sale JEM, Hawker G, Cameron C, Bogoch E, Jain R, Beaton D, et al. Perceived messages about bone health after a fracture are not consistent across healthcare providers. Rheumatol Int. 2015;35:97–103.

    Article  PubMed  Google Scholar 

  36. Sale JEM, Cameron C, Thielke S, Meadows L, Senior K. The theory of planned behaviour explains intentions to use antiresorptive medication after a fragility fracture. Rheumatol Int. 2017;37:875–82.

    Article  PubMed  Google Scholar 

  37. Ajzen I, Fishbein M. Understanding attitudes and predicting social behavior. Englewood Cliffs: Prentice-Hall; 1980.

    Google Scholar 

  38. Fishbein M, Ajzen I. Belief, attitude, intention, and behavior: an introduction to theory and research. Reading: Addison-Wesley; 1975.

    Google Scholar 

  39. Giorgi A. The theory, practice, and evaluation of the phenomenological method as a qualitative research procedure. J Phenomenol Psychol. 1997;28:235–60.

    Article  Google Scholar 

  40. Giorgi A. The descriptive phenomenological method in psychology: a modified Husserlian approach, vol. 2009. Pittsburgh: Duquesne University Press; 2009.

    Google Scholar 

  41. Wertz FJ. Phenomenological research methods for counseling psychology. J Couns Psychol. 2005;52(2):167–77.

    Article  Google Scholar 

  42. Braun V, Clarke V. Using thematic analysis in psychology. Qual Res Psychol. 2006;3:77–101.

    Article  Google Scholar 

  43. Kristjansson E, Tugwell PS, Wilson AJ, Brooks PM, Driedger SM, Gallois C, et al. Development of the effective musculoskeletal consumer scale. J Rheumatol. 2007;34(6):1392–400.

    PubMed  Google Scholar 

  44. Hibbard JH, Stockard J, Mahoney ER, Tusler M. Development of the patient activation measure (PAM): conceptualizing and measuring activation in patients and consumers. Health Serv Res. 2004;39(4 Pt 1):1005–26.

    Article  PubMed  PubMed Central  Google Scholar 

  45. Sale JEM, Thielke S. Qualitative research is a fundamental scientific process. J Clin Epidemiol. 2018;102:129–33.

    Article  PubMed  Google Scholar 

  46. Agee J. Developing qualitative research questions: a reflective process. Int J Qual Stud Educ. 2009;22(4):431–47.

    Article  Google Scholar 

  47. Cooper N, Burnett S. Using discursive reflexivity to enhance the qualitative research process. Qual Soc Work. 2006;5(1):111–29.

    Article  Google Scholar 

  48. Carter SM, Little M. Justifying knowledge, justifying method, taking action: epistemologies, methodologies, and methods in qualitative research. Qual Health Res. 2007;17(10):1316–28.

    Article  PubMed  Google Scholar 

  49. Crotty M. The foundations of social research. Los Angeles: Sage Publications; 1998.

    Google Scholar 

  50. Sandelowski M. Whatever happened to qualitative description? Res Nurs Health. 2000;23:334–40.

    Article  CAS  PubMed  Google Scholar 

  51. Gheorghita A, Webster F, Thielke S, Sale JEM. Long-term experiences of pain after a fragility fracture. Osteoporos Int. 2018;29:1093–104.

    Article  CAS  PubMed  Google Scholar 

  52. Sale JEM, Ashe MC, Beaton D, Bogoch E, Frankel L. Men’s health-seeking behaviours regarding bone health after a fragility fracture: a secondary analysis of qualitative data. Osteoporos Int. 2016;27(10):3113–9.

    Article  CAS  PubMed  Google Scholar 

  53. Sale JEM, Frankel L, Paiva J, Saini J, Hui S, McKinlay J, et al. Having caregiving responsibilities affects management of fragility fractures and bone health. Osteoporos Int. 2020;31:1565–72.

    Article  CAS  PubMed  Google Scholar 

Download references


Not applicable.


Funding for the work described in this paper was provided by the Canadian Institutes of Health Research (Funding Reference Number: CBO-109629). The Canadian Institutes of Health Research had no involvement in the design of the study and collection, analysis, and interpretation of the data and in the writing the manuscript.

Author information

Authors and Affiliations



Joanna Sale made substantial contributions to conception and design and analysis and interpretation of the data, drafted and revised the manuscript critically for important intellectual content, approved the final version of the manuscript submitted, and agreed to be accountable for all aspects of the work.

Author’s information

JEMS is a Scientist and Associate Professor who has been teaching qualitative research courses and lectures at the introductory and intermediate level at the University of Toronto since 2007.

Corresponding author

Correspondence to Joanna E. M. Sale.

Ethics declarations

Ethics approval and consent to participate

The study and protocol upon which this manuscript is based was approved by the Research Ethics Board at Unity Health Toronto (REB# 10–371). All methods were carried out in accordance with the Declaration of Helsinki and the relevant guidelines and regulations set by the Research Ethics Board at Unity Health Toronto. Informed consent was obtained from all participants.

Consent for publication

Not applicable.

Competing interests

The author declares that she has no competing interests.

Additional information

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit The Creative Commons Public Domain Dedication waiver ( applies to the data made available in this article, unless otherwise stated in a credit line to the data.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Sale, J.E.M. The role of analytic direction in qualitative research. BMC Med Res Methodol 22, 66 (2022).

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: