Conducting Qualitative Research in Veterinary Settings

By Dr. Zubair Khalid, DVM, MS, PhD · 2026-08-11

Key Takeaways

• Qualitative research in veterinary settings addresses "how" and "why" questions regarding professional decision-making, client perceptions, and social influences, complementing quantitative data by exploring the reasoning behind behaviors such as antimicrobial prescribing patterns.
• Common qualitative methods include semi-structured interviews (typically 10-30 participants) and focus groups (4-7 participants per group), with data analyzed primarily through thematic analysis, content analysis, or framework analysis to identify patterns of meaning.
• Rigour in qualitative veterinary research is established through trustworthiness criteria (credibility, transferability, dependability, confirmability) and techniques like member checking, triangulation, reflexivity, and maintaining an audit trail, with reporting guided by standards like COREQ.
• Key failure modes include design drift (research question shifting post-data collection), premature saturation (claiming saturation without sufficient documentation), and analytic bias, which can be mitigated through rigorous protocol adherence and transparent documentation of analytic decisions.
• Salience of themes in qualitative analysis is determined by emotional weight and unprompted discussion, not mere frequency, and findings should be reported as range and variation, not quantitative prevalence estimates, unless the study design specifically supports it.
• Ethical considerations in veterinary qualitative research necessitate institutional ethics approval for identifiable data and adherence to guidelines like ARRIVE for animal procedures, with consultation of methodologists recommended for sensitive topics or integration with quantitative data.

Qualitative research answers questions that quantitative methods cannot address. It examines how veterinarians make clinical decisions, how clients perceive treatment recommendations, and how professional networks shape prescribing behavior. This article explains the design, conduct, and analysis of qualitative studies in veterinary settings. It serves veterinary researchers who need a procedural reference for interview studies, focus groups, and thematic analysis, and it addresses the question of how to generate trustworthy evidence from non-numerical data.

The veterinary literature increasingly uses qualitative approaches to study antimicrobial stewardship, One Health collaboration, and human-animal relationships. Studies of antimicrobial prescribing in UK small animal practice have used semi-structured interviews with 21 veterinarians to identify factors influencing drug selection, including colleague influence, client pressure, and awareness of guidelines. Similar work in dairy cattle has been synthesised through systematic review, showing that farmers' and veterinarians' beliefs about antimicrobial resistance vary and sometimes conflict. These examples demonstrate that qualitative methods produce insights into professional behavior that surveys cannot capture.

This article covers the scientific rationale for qualitative inquiry, study design decisions, data collection techniques, sampling strategies, and analytic procedures. It also addresses quality criteria and reporting standards. The companion article on reporting qualitative research in veterinary science provides detailed guidance on publication requirements.

At a Glance

Scientific Rationale for Qualitative Inquiry

Veterinary medicine involves complex social systems. Clinical decisions occur within a network of client expectations, economic constraints, practice culture, and professional norms. Quantitative methods measure the frequency of behaviors but cannot explain why those behaviors occur. Qualitative methods examine the reasoning, values, and social pressures that shape practice.

The study of antimicrobial use illustrates this distinction. A prescribing audit can show that a practice prescribes amoxicillin-clavulanate more often than guidelines recommend. It cannot explain whether this reflects client demand, habit, fear of treatment failure, or uncertainty about diagnosis. Semi-structured interviews can explore these factors directly. In the UK small animal study, thematic analysis of interview transcripts identified distinct influences on prescribing, including the use of hypothetical clinical scenarios to probe decision-making. The scenarios covered conditions such as deep pyoderma in a Shar-Pei and feline lower urinary tract disease, allowing researchers to compare stated principles with case-specific responses.

Qualitative methods also serve One Health research. A study of integrated zoonotic disease surveillance in Australia used ten semi-structured interviews with academic experts to identify barriers to interdisciplinary collaboration. Thematic analysis revealed that the absence of a shared definition of One Health hindered cooperation and that sectoral silos restricted communication. These findings have direct implications for surveillance system design.

Philosophical Foundations and Study Design Logic

Qualitative research rests on interpretivist assumptions. It holds that reality is socially constructed and that meaning emerges from context. This contrasts with the positivist assumption that a single objective reality can be measured. The researcher is not a neutral observer but an instrument of data collection and analysis. This position must be acknowledged in the study protocol and in the final report.

The choice of qualitative design follows from the research question. Exploratory questions about understudied phenomena suit grounded theory approaches. Questions about experiences and perceptions suit phenomenology. Questions about professional behavior in context suit ethnography or case study design. Questions about patterns across a defined body of literature suit qualitative systematic review. The design must be specified before data collection begins, although qualitative research allows iterative refinement of questions and sampling as understanding develops.

Defining the Research Question

A qualitative research question should be open, focused, and answerable. It should not presuppose a hypothesis. The question "What factors influence antimicrobial prescribing decisions in small animal practice?" is appropriate. The question "Does guideline adherence improve after an educational intervention?" is not, because it assumes a measurable outcome and a pre-specified intervention.

The question should also delimit the setting and population. A study of dairy farmer and veterinarian behavior regarding antimicrobial use must specify the production system, the region, and the professional roles included. The systematic review of dairy cattle antimicrobial use identified 35 studies from 349 articles, illustrating the breadth of existing work and the need for clear inclusion criteria.

Study Design and Participant Selection

Qualitative studies in veterinary settings typically use purposive sampling. Participants are selected because they hold relevant knowledge or experience, not because they represent a population statistically. Maximum variation sampling seeks a range of perspectives across practice types, regions, or professional roles. Snowball sampling asks participants to nominate colleagues, which is useful for reaching hidden populations such as veterinarians working in regulatory roles or traditional health practitioners.

Sample size in qualitative research is determined by saturation. Saturation occurs when new data no longer generate new themes or insights. In interview studies, saturation is often reached between 10 and 30 participants, depending on the homogeneity of the population and the breadth of the research question. Focus groups typically include 4 to 7 participants per group, as used in a Portuguese study of community pharmacists' attitudes to antibiotic dispensing. That study conducted groups across five districts to capture regional variation.

Interview Design

Semi-structured interviews use a topic guide instead of a fixed questionnaire. The guide lists themes to cover but allows the interviewer to follow participant responses. The UK antimicrobial study used a topic guide covering criteria for antimicrobial selection, influences by colleagues and clients, pet characteriztics, sources of knowledge, awareness of guidelines, and practice protocols. This structure ensures consistency across interviews while permitting exploration of unexpected topics.

Interview questions should be open-ended and neutral. Questions that begin with "why" can provoke defensive responses. Questions that begin with "tell me about" or "describe" encourage narrative accounts. Hypothetical scenarios can probe decision-making without requiring participants to admit to undesirable practices. The UK study used four scenarios, including vomiting in a Yorkshire Terrier due to dietary indiscretion and neutering of a six-month-old dog, to assess the appropriateness of antimicrobial use in realistic contexts.

Data Collection Methods

Interviews are the most common data collection method in veterinary qualitative research. They can be conducted face to face, by telephone, or by video call. Audio recording and professional transcription are standard. The interviewer should transcribe or verify transcripts to maintain familiarity with the data.

Focus groups generate data through interaction among participants. They are efficient for exploring shared norms and disagreements within a professional group. The Portuguese pharmacist study used focus groups with a moderator and a topic guide, with sessions audio-recorded and transcribed. Group dynamics can reveal consensus and conflict that individual interviews miss, but they require skilled moderation to ensure all participants contribute.

Observation and document analysis are less common but valuable. Participant observation allows researchers to study behavior in natural settings, such as consultation rooms or farm visits. Document analysis examines written protocols, guidelines, or clinical records. These methods can be combined with interviews in a mixed methods design.

Data Management and Transcription

Qualitative data management begins at the point of collection. Audio recordings must be transferred to a secure storage location immediately after each interview or focus group, with participant identifiers replaced by codes. A master list linking codes to identities should be stored separately from the data files, ideally in an encrypted file with restricted access. This separation protects participant confidentiality and simplifies later anonymisation during reporting.

Transcription converts spoken interaction into analysable text. Verbatim transcription is the standard for thematic analysis because it preserves the participant's own words, including hesitations, repetitions, and emphasis that may carry meaning. Some researchers use transcription software with manual correction, while others transcribe by hand. The choice depends on budget, time, and the need for accuracy. Mateus and colleagues transcribed interviews by hand in their study of antimicrobial prescribing decisions in UK small animal practice, a process that also familiarised the interviewer with the data before formal analysis began.

Transcription conventions should be established before analysis starts. Decide whether to include non-verbal cues such as laughter, pauses, and overlapping speech. For most veterinary research questions, a clean verbatim transcript that omits fillers such as "um" and "you know" is sufficient, but if the research question concerns communication patterns or interaction dynamics, a more detailed notation system is required. The decision should be recorded in the study protocol.

Coding and Theme Development

Thematic analysis is the most commonly used analytic approach in veterinary qualitative research. It involves identifying, organizing, and interpreting patterns of meaning across the dataset. The process is iterative: codes are generated from the data, grouped into candidate themes, reviewed against the data, and refined until they accurately represent participant experiences.

Coding begins with close reading of the transcripts. A code is a short label that captures the essence of a segment of text. Codes may be descriptive, summarizing what was said, or interpretive, capturing the underlying meaning. In the Australian One Health surveillance study, thematic analysis of ten semistructured interviews with academic experts identified barriers to integrated zoonotic disease surveillance, including the absence of a shared definition of One Health and the siloed structure of human, animal, and ecological health sectors. These themes emerged from systematic coding of participant accounts instead of from a pre-existing framework.

Two broad approaches to coding exist. Inductive coding derives codes entirely from the data, allowing themes to emerge without imposing prior assumptions. Deductive coding applies a pre-existing framework or theory to the data, testing whether the framework explains participant experiences. Many veterinary studies use a hybrid approach, beginning with a few deductive codes derived from the research question and allowing inductive codes to supplement them.

The coding process itself requires decisions about granularity. Fine-grained coding produces many codes that capture subtle distinctions, while coarse-grained coding produces fewer, broader codes. The appropriate level depends on the research question and the intended depth of analysis. A study exploring antimicrobial prescribing decisions might code at a fine grain to distinguish between client pressure, colleague influence, and practice protocols as separate factors, as demonstrated in the UK small animal study.

After initial coding, the researcher reviews the codes and groups them into candidate themes. A theme is a coherent pattern that captures something significant about the data in relation to the research question. Themes are not simply summaries of codes, they represent an interpretive level of analysis that explains what the patterns mean. The systematic review of dairy farmer and veterinarian antimicrobial use behaviors noted that findings across studies were varied and conflicting, which underscores the importance of rigorous theme development that stays grounded in the data instead of forcing coherence.

Theme refinement involves checking that each theme is internally consistent and distinct from other themes. The researcher returns to the transcripts to verify that the themes accurately represent the data and that no important content has been overlooked. This recursive process continues until the themes are stable and the researcher is confident that the analysis is complete.

Ensuring Rigour and Trustworthiness

Qualitative research is assessed by criteria of trustworthiness instead of statistical validity. Four criteria are widely used: credibility, transferability, dependability, and confirmability. Credibility concerns whether the findings accurately represent participant perspectives. Transferability concerns whether the findings apply to other contexts. Dependability concerns whether the research process is consistent and auditable. Confirmability concerns whether the findings are grounded in the data instead of researcher bias.

Several techniques support these criteria. Member checking involves returning the findings to participants for comment, allowing them to confirm or correct the researcher's interpretations. Triangulation uses multiple data sources, researchers, or methods to cross-check findings. Reflexivity requires the researcher to document their own assumptions, background, and potential influence on the data collection and analysis. An audit trail records all analytic decisions, providing transparency and supporting dependability.

Reporting standards also support rigour. The Consolidated Criteria for Reporting Qualitative Research (COREQ) provides a 32-item checklist covering research team and reflexivity, study design, and analysis and findings. The EQUATOR Network maintains a comprehensive library of reporting guidelines, including COREQ and other qualitative reporting standards, which researchers should consult before submitting their work for publication. Transparent reporting allows readers to assess the quality of the study and the credibility of its findings.

Focus Groups in Veterinary Research

Focus groups are a distinct data collection method that generates data through group interaction. A moderator guides a discussion among four to seven participants, using a topic guide to ensure coverage of the research questions while allowing participants to respond to and build on each other's contributions. The Portuguese study of community pharmacists' attitudes to antibiotic dispensing used focus groups of four to seven pharmacists, with sessions audio-recorded and transcribed for analysis.

Focus groups are particularly suited to exploring shared norms, professional cultures, and areas of disagreement. They can reveal how veterinarians discuss clinical decisions with colleagues, how practice protocols are negotiated, and how professional identities are constructed. They are less suited to sensitive topics where participants may be reluctant to speak openly in a group, and they require careful moderation to ensure that dominant voices do not suppress quieter participants.

The decision between interviews and focus groups depends on the research question and the participant population. Interviews provide depth and privacy, allowing exploration of individual experiences and sensitive topics. Focus groups provide breadth and interaction, allowing observation of how opinions are formed and modified through discussion. Some studies combine both methods, using interviews for detailed individual accounts and focus groups for broader exploration of shared perspectives.

The repeat group discussion design used in the South African study of traditional health practitioners offers another variation. Four repeat group discussions with nine practitioners allowed the research team to build rapport, explore topics in increasing depth, and observe how accounts developed over time. This design is resource-intensive but can generate richer data than a single session.

Documentation and Reporting

Documentation throughout the research process supports both analysis and reporting. A study protocol should record the research question, design, sampling strategy, data collection methods, and analytic plan. Field notes taken during and after data collection capture contextual information that may not appear in transcripts, such as the setting, participant demeanour, and interactions between participants. An audit trail records coding decisions, theme development, and the rationale for analytic choices.

The final report should describe the research process in sufficient detail that readers can assess its quality. This includes the sampling strategy and participant characteriztics, the data collection methods and setting, the analytic approach, and the steps taken to ensure rigour. The report should present themes with supporting quotations from participants, allowing readers to judge the fit between the data and the interpretation. The COREQ checklist provides a structured framework for this reporting.

Table 1 summarizes the key decisions in qualitative data management and analysis.

The choice of analytic approach should be made before data collection begins and recorded in the protocol. Changes to the analytic plan during the study are acceptable if documented and justified, but post hoc decisions that appear to be driven by the findings should be avoided.

Recognized Complications and Failure Modes

Qualitative studies in veterinary settings fail most often from design drift, not from data collection errors. Design drift occurs when the research question shifts after data collection begins, usually because early interviews reveal unexpected themes. The researcher then continues collecting data against the original question while analyzing against an unstated new one. Detect this early by returning to the written research question before each analysis session and asking whether the coding frame still answers it. If it does not, either revise the question formally in the protocol or stop data collection and re-scope.

A second failure mode is premature thematic saturation. Researchers declare saturation when they have run out of time or when interviews become repetitive, not when new data cease to yield new codes. The discriminating check is to document saturation explicitly: log each new code by interview number and show that the last three to five interviews added no substantive codes. Without this audit trail, reviewers will question the claim. The Consolidated Criteria for Reporting Qualitative Research (COREQ) framework provides a structured way to report sampling, data collection, and analysis decisions so that reviewers can assess whether saturation claims are credible.

A third failure mode is analytic bias introduced during transcription and coding. When the same person who conducted the interviews also transcribes and codes, they may unconsciously privilege passages that confirm their impressions from the field. Mitigate this by having a second coder independently code a subset of transcripts and by calculating simple inter-coder agreement on those passages. Disagreements should be resolved through discussion and documented, not by averaging.

Common Errors and Corrective Actions

Less experienced researchers frequently confuse frequency with salience. A theme mentioned by every participant is not necessarily the most important theme, it may simply be the most socially acceptable thing to say. Salience is better judged by the emotional weight participants attach to a topic, the length of time they spend on it, and whether they raise it unprompted. Correct this error by coding both the presence of a theme and the intensity markers around it.

A second common error is asking leading questions during interviews and then treating the resulting answers as participant-generated themes. For example, asking "Do you think clients pressure you to prescribe antibiotics?" invites agreement. The corrective action is to pilot the interview guide and have a colleague flag any question that contains an embedded assumption. The qualitative study of antimicrobial usage in UK small animal practices used hypothetical clinical scenarios to elicit decision-making without leading participants toward a preferred answer, a technique that transfers well to other clinical topics.

A third error is treating qualitative findings as if they were quantitative prevalence estimates. A thematic analysis of ten interviews cannot tell you what proportion of veterinarians hold a given belief. It can tell you that a belief exists and how it functions in clinical reasoning. Correct this by writing findings in the language of range and variation, not frequency, unless the study design genuinely supports counting.

Limitations of the Current Evidence

The veterinary qualitative evidence base is uneven. Antimicrobial use and prescribing behavior have received sustained attention, including systematic reviews of farmers' and veterinarians' behavior regarding antimicrobial use in dairy cattle, but other clinical areas remain thinly studied. Studies often rely on small samples from single regions, which limits transferability. The Australian One Health surveillance study illustrates a common limitation: findings from expert interviews describe perceived barriers but cannot establish whether those barriers actually operate in practice.

Expert opinion still differs on several points. One contested area is whether focus groups or individual interviews produce more candid data in veterinary settings. Some researchers argue that focus groups generate richer discussion because participants build on each other's comments, as seen in the Portuguese community pharmacist focus group study. Others contend that individual interviews are necessary because veterinarians are reluctant to admit uncertainty in front of colleagues. No comparative study has resolved this question for veterinary populations.

A second contested area is the role of quantitative counts within qualitative analysis. Some methodologists permit simple counts of how often a code appears as a transparency measure. Others reject counting entirely, arguing that it implies a generalizability the sampling strategy cannot support. The companion dog ownership study during Covid-19 demonstrates that large-sample qualitative work is possible with online data collection, but it also shows that such samples are self-selected and cannot be treated as representative.

Referral, Consultation, and Reporting Thresholds

Most qualitative veterinary studies do not require regulatory reporting. However, two circumstances do. First, if the research involves client or practitioner data that could identify individuals, institutional ethics approval is required before data collection begins. The AVMA practice resources provide guidance on professional obligations regarding client confidentiality and data handling. Second, if the study involves animal procedures, the ARRIVE guidelines specify the minimum reporting requirements for animal research publications, and compliance is expected by most veterinary journals.

Consult a qualitative methodologist early if your study involves vulnerable participants, such as clients discussing euthanasia decisions or practitioners discussing professional errors. These topics require particular skill in interview technique and in managing participant distress. Similarly, consult a statistician or epidemiologist if you plan to integrate qualitative findings with quantitative data, because the integration logic must be specified before data collection, not after.

Laboratory involvement is rarely needed for qualitative studies, but it becomes relevant when interview or focus group data reference diagnostic testing. If participants describe laboratory results as influencing their decisions, verify those descriptions against actual laboratory records where possible. This triangulation strengthens the analysis and guards against recall error.

Frequently Asked Questions

How Many Participants Are Enough for a Qualitative Veterinary Study?

Sample size depends on the research question, design, and the point of thematic saturation. For interview-based studies in veterinary settings, 10 to 30 participants commonly suffices, but saturation should be demonstrated instead of assumed. A study of antimicrobial prescribing across seven UK small animal practices used 21 veterinarians and reached sufficient thematic depth, which suggests that smaller, purposive samples often outperform larger convenience samples in qualitative work. For focus groups, plan four to seven participants per group, as used in a Portuguese study of community pharmacists. If resources are constrained, recruit in waves and stop when new data no longer generate new codes or themes. Document the saturation decision explicitly in your methods.

What Can I Do When Audio Recording Is Not Permitted or Feasible?

When recording is impossible, take detailed field notes during and immediately after each interview or focus group. Write down verbatim phrases where possible, because participant wording carries analytic weight. Expand abbreviations and clarify ambiguous statements while the interaction is fresh. Consider having a second researcher present to take notes independently, then compare and reconcile the two accounts. For focus groups, a note-taker is essential because the moderator cannot capture group dynamics alone. If you rely on notes instead of transcripts, acknowledge this limitation in your report and describe how the absence of full transcription may have affected the analysis. Some ethics committees require a data management plan that specifies how notes will be stored and anonymised.

How Do I Adapt Qualitative Methods for Studies Involving Production Animals?

Farm animal research often requires interviewing farmers, farm workers, and veterinarians across dispersed geographic areas, which makes in-person interviews costly. Telephone or video interviews are acceptable alternatives and may increase participation. A systematic review of antimicrobial use behavior in dairy cattle found that studies varied widely in how they sampled farmers and veterinarians, and that transparency of reporting was often poor. When visiting farms, schedule interviews around milking times and calving seasons. Use photo-voice methods, where participants photograph relevant practices, to prompt discussion about routines that have become habitual. For studies involving livestock keepers in low and middle income settings, community walks and repeat group discussions have been used successfully to explore health-seeking behavior, as demonstrated in research with traditional health practitioners in rural South Africa.

What Are the Minimum Resources Needed to Conduct a Defensible Qualitative Study?

A defensible study requires a clear research question, a purposive sampling strategy, a data collection guide, a method for capturing data, and a systematic analysis plan. You do not need expensive software. Word processors and spreadsheets can support coding, although dedicated qualitative data analysis software helps manage large datasets. Transcription is often the largest cost, budget for professional transcription or allocate substantial researcher time. Ethics approval is mandatory and may require proof of researcher training in qualitative methods. If you lack formal training, consult reporting standards such as those catalogued by the EQUATOR Network, which include the Consolidated Criteria for Reporting Qualitative Research, before you begin data collection. Pilot test your interview guide with two or three colleagues to refine questions and practice probing techniques.

How Should I Store and Anonymise Qualitative Data in a Veterinary Practice Setting?

Store audio files, transcripts, and field notes on encrypted, password-protected institutional drives instead of personal devices. Remove names, practice identifiers, and distinctive clinical details during transcription. Replace participant names with codes and keep a separate linking document under restricted access. Be aware that small veterinary communities may recognize individuals from context even after anonymisation, so describe participants in general terms and consider omitting identifying practice characteriztics. Data sharing agreements should specify who can access raw data and for what purposes. If you collect data in a practice where you also work clinically, clarify your dual role with participants and consider whether a colleague should conduct the interviews to reduce coercion concerns.

How Do I Explain Qualitative Research to a Client or Practice Owner Who Expects Numbers?

Explain that qualitative research answers questions about why and how, not how many. For example, understanding why veterinarians choose certain antimicrobials requires exploring clinical reasoning, client pressure, and practice culture, which a survey cannot capture. A UK study of small animal practitioners used hypothetical clinical scenarios to reveal how decision-making varies by case presentation, and this kind of insight supports guideline development and practice improvement. Emphasize that findings inform protocols, training, and policy. If the practice owner is concerned about time, offer a brief written summary of the research aims and a commitment to share findings. Reassure them that participation is confidential and that the goal is to improve care, not to audit individual performance.

Related Clinical & Scientific Guides

• Conducting Systematic Reviews of Veterinary Diagnostic Test Accuracy
• Bias in Veterinary Research: Types, Sources, and Mitigation
• Cluster Randomized Trials in Veterinary Research: Design and Analysis

References and Further Reading

• Qualitative study of factors associated with antimicrobial usage in seven small animal veterinary practices in the UK.. 2014.
• The challenges of implementing an integrated One Health surveillance system in Australia.. 2018.
• "I Couldn't Have Asked for a Better Quarantine Partner!": Experiences with Companion Dogs during Covid-19.. 2021.
• Understanding farmers' and veterinarians' behavior in relation to antimicrobial use and resistance in dairy cattle: A systematic review.. 2021.
• Attitudes of community pharmacists to antibiotic dispensing and microbial resistance: a qualitative study in Portugal.. 2013.
• The role of traditional health practitioners in Rural KwaZulu-Natal, South Africa: generic or mode specific?. 2016.
• ARRIVE Guidelines 2.0 for Reporting Animal Research. PLOS Biology, 2020.
• EQUATOR Network Reporting Guidelines. EQUATOR Network.
• MSD Veterinary Manual, Professional Edition. MSD Veterinary Manual.

Related Articles

• Reporting Qualitative Research in Veterinary Science
• Using Mixed Methods in Veterinary Research
• Conducting Pharmacovigilance Studies in Veterinary Medicine
• How to Write a Research Protocol for Veterinary Studies
• Applying Competing Risks Analysis in Veterinary Research

This article is educational professional reference material for veterinary audiences. It is not a substitute for veterinary diagnosis, individual clinical judgment, current product labeling, or applicable regulatory requirements.