Qualitative Research Part 2: Conducting qualitative research

This paper follows the paper exploring the place of qualitative research in mental health nursing practice. It moves from the methodologies to the method of conducting qualitative research. It will explore three key components of the research process: developing a research question, data collection, and data analysis. It also considers some of the limitations of qualitative research.

All qualitative research processes pivot on the research question. The question will determine both the methodology and the method of conducting the research. It encapsulates the object of the research inquiry and facilitates the research process. Ideally, its focus would be the patient's experience of some aspect of mental health nursing practice. A good place to start is to identify an area of mental health nursing interest or consider those taken-for-granted aspects of practice e.g., the patient's experience of admission, treatment, a specific intervention, the relationship with nurses, and the psychiatric process. The next step is to explore the literature on the topic and what studies have been undertaken.

The research needs to contribute to new knowledge, so it is important to identify gaps in the literature. For example, your interest may be the use of observations in an acute inpatient setting as a strategy to manage risk. You would need to explore the nursing literature and the broader literature on observation and risk management and the mental health nursing literature. Most research raises further questions, and it is these unanswered questions that require further investigation. There will be relevant studies that have focused on observation and risk management, but your goal is to identify those aspects of the findings that raise further questions. Studies may have examined patients' experiences of observations from different perspectives, and your question may develop from those findings e.g., one study found that patients' sleep was disturbed by nursing observations at night (Veale et al. 2020). You may choose to develop a subsequent question around what patients think may be more helpful or the patients' experience of sleep deprivation and its impact on recovery.

Another strategy for developing a research question is to focus your exploration of the literature on systematic and integrative reviews and qualitative meta-syntheses. Observations may be regarded as a form of restrictive practice so you would need to look at the literature reviews on this topic. For example, an integrative review of patient preferences related to restrictive practices (Lindekilde et al. 2024) reported that factors such as environment, communication and duration were found to influence patients' preferences. After exploring a range of literature, a research question could be developed inquiring about the types of communication that patients would prefer when they are being observed, or what factors in the environment could improve risk management strategies.

Once a research question has been developed the next step is to decide on a research strategy. A question needs to be developed and a methodology chosen that is congruent. For example, if it is the lived phenomenological experience the question might be what is the patient's lived experience of time when they are under constant observation. From a grounded theory perspective, the question might be what does it mean to be under constant observation? A discourse analytic approach would be focused on the discourses so might be what discourses patients draw on to describe their experiences of observation. For thematic analysis, the research question can be less directed but still needs to focus on the quality of the experience such as what is the patient's experience of observation.

Examples of research questions that drive the research process include: What are forensic psychiatric inpatients' experiences of being in a seclusion room? (Askew, Fisher, and Beazley 2020); what is the nature and meaning of engagement for practitioners and service users within assertive outreach services (Wright, Callaghan, and Bartlett 2011); what are the experiences of depressed adolescents who had recently engaged in a preferred intensity exercise intervention? (Carter et al. 2016).

The next steps involve deciding on eligibility criteria, a process for recruitment, ethical approval and data collection. All these steps are dependent on the research question and the decisions undertaken need to reflect the chosen qualitative methodology. The most common type of qualitative data collection is interviews either one-to-one or focus groups. Before collecting the data it is usual to develop semi-structured interview prompts. These are prompts to keep the focus on the research question. You need to consider the rationale for each prompt in terms of the overarching question. Interviews will need to be recorded and transcribed so that they become texts for analysis.

Data saturation is a term that often accompanies qualitative data collection i.e., data collection will continue until data saturation. This can be defined as information redundancy or the point at which no new themes or codes ‘emerge’ from data. However Braun and Clarke (2021) argue that this is not consistent with the values and assumptions of reflexive thematic analysis. They propose that meaning is generated through interpretation of, not excavated from, data, and therefore judgements about ‘how many’ data items, and when to stop data collection, are inescapably situated and subjective, and cannot be determined (wholly) in advance of analysis. Decisions about sample size are thus determined by the research question and situated within the context of the research process.

The analysis of data is determined by the research methodology; however, most qualitative data analysis is concerned with inductively identified themes. If a phenomenological approach is used the analysis may focus on data related to lived time, lived space, lived relations and lived body. In grounded theory, the analysis is iterative and uses a constant comparison technique to develop codes and relationships between codes. In discourse analysis, the analysis involves identifying the discourses being drawn upon in the transcript and an exploration of those discourses. In thematic analysis, a process of coding, grouping and interpreting themes is used from a specific theoretical position e.g., critical realism. In all qualitative analyses, the research question is central and not the person's responses to specific questions. Qualitative analysis aims to identify findings related to the research question, the interview prompts should be used as tools for collecting relevant data.

Some qualitative studies analyse the data according to the interview prompts but these should not be the focus. The interview prompts facilitate reflection and articulation of the research question but are not substitutes for the research question. Using interview questions as themes represents an underdeveloped analysis (Braun and Clarke 2006). Such analyses are better suited to content analysis. Most qualitative methods require thematic analysis (phenomenological, grounded theory, thematic analysis) that examines the relationships and meanings in the categories to identify themes.

Qualitative research is underpinned by a philosophical position that assumes that experiences can be accurately and fully articulated in language and that the person's perspective is a fixed representation of their subjective experience (Crowe 1998). It can be argued that these assumptions about the ability of language to accurately and fully reveal subjective experience are cultural rather than natural. Cultures determine what attributes of subjectivity and individual action are acceptable. Western culture has constructed ‘normal’ subjectivity as involving displays of unitariness, moderation, rationality and productivity (Crowe 2000). We are required to articulate our subjective experiences in ways that demonstrate an ability to distinguish an internal and external world while recognising the two as separate (unitariness). We are also required to provide some value to our society (productivity), regulate our impulses, speech and actions (moderation), and we demonstrate allegiance to social expectations of literalness, objectivity and fixity (rationality).

Qualitative research is a good fit with mental health nursing practice. It is also a good fit for understanding the people we care for. As this paper has emphasised the research question is central to all phases of the research process. It is important to spend time developing this question upon which the research will sit. The next paper in this series will describe the process of preparing the research for publication.

The authors declare no conflicts of interest.