Data collection

For those using a positivist framework to guide their qualitative study, data collected from a large sample through observation or through structured or semi-structured interviews are highly suitable (Su Reference Su, Cassell, Cunliffe and Grandy2018; Alharahsheh Reference Alharahsheh and Pius2020). Structured and semi-structured interviews fit well with the positivist approach of using systematic research techniques as, for structured interviews, researchers generate an interview schedule based on predetermined ideas and follow this schedule very closely to ensure as little variation as possible during interviews (Longhurst Reference Longhurst2003; Smith Reference Smith, Osborn and Smith2003). For semi-structured interviews, an interview schedule is still generated and followed but the researcher is led slightly more by the participant's response, for example they can delve deeper into what they perceive to be an interesting response (Longhurst Reference Longhurst2003; Smith Reference Smith, Osborn and Smith2003).

Data analysis

Analysis using a deductive approach, that is the use of a prespecified framework (based on previous literature or theory) to guide data analysis, is also highly suitable for those using a positivist framework (Alharahsheh Reference Alharahsheh and Pius2020). A content analysis is particularly suitable for this framework as the aim of this analytical approach is to uncover reality as it exists (Saraisky Reference Saraisky2016) and is based on the understanding that meaning can be coded and counted systematically (Saraisky Reference Saraisky2016). Content analysis is also often done using a codebook based on predetermined knowledge and theory which the researcher uses to map patterns and meaning (Saraisky Reference Saraisky2016). Owing to its theoretical flexibility, a thematic analysis is also suitable for this approach as it involves summarising data into overarching patterns or themes and can be carried out using a deductive approach (Braun Reference Braun and Clarke2006).

Use in practice

Most mental health service research adopts a positivist perspective but often researchers do not reference this approach, especially within qualitative studies (Kutney Reference Kutney2006). It is important that the framework used in a study is mentioned when reporting research to inform others of the approach guiding the methodology and to show other researchers how to effectively utilise such an approach.

To enhance understanding of how the positivist framework can be used to guide qualitative mental health research, we present an example question focused on mental healthcare and discuss how the framework can be used to answer this question.

Example: applying the positivist framework

How might caregiving burden be experienced by family caregivers of those using mental health services?

A researcher using a positivist framework would approach this question by exploring the connection between family caregivers’ experiences and caregiving burden as defined in the literature (a multidimensional response to stressors associated with the caregiving experience (Liu Reference Liu, Sun and Chen2022)). As positivism focuses on verifying a hypothesis, a positivist researcher could test a specific hypothesis using qualitative research. In this context for example, positivist researchers may test the hypothesis that caregiving burden comprises several domains. A large number of family caregivers could be recruited to include participants with different characteristics of interest. Researchers could carry out structured interviews with these caregivers, using questions focused on their experience of supporting a family member who uses mental health services and the stress they feel regarding this experience. The interview data could then be analysed in a deductive content analysis, using a predetermined framework based on a theory related to burden or stress, for example the theory of caregiver stress (Tsai Reference Tsai2003). For this, researchers could code the interview data based on the prespecified framework (highlighting parts of the text that align with specific concepts within the framework) and note how many times a related code appears. Overarching concepts (or patterns) could then be identified based on how many codes relate to specific domains. These overarching domains can be used as a formal framework.

The use of a positivist theoretical framework lends itself well to areas of mental health research where the credibility, transferability and development of tangible knowledge are important, for example research looking to develop or test theoretical models of mental illness, or when developing mental health interventions that can be applied generally and/or uniformly (Su Reference Su, Cassell, Cunliffe and Grandy2018; Park Reference Park, Konge and Artino2020).

However, it is important to recognise the potential shortcomings of this theoretical framework within qualitative research. One potential limitation is that, in attempting to simplify a concept into something that can be applied universally, this framework may overlook the complexity of the given concept, for example the experience of caregiving, which is inherently complex (Alharahsheh Reference Alharahsheh and Pius2020). The positivist framework would focus on generic concepts without attempting to interpret these as something that may have different meanings depending on their context. It also purposely ignores the potential influence of the individual (including the researcher) and their prior experience or knowledge when attempting to understand a specific concept (Alharahsheh Reference Alharahsheh and Pius2020). Therefore, it is not suited to studies that focus on an in-depth exploration of individuals’ experiences or perceptions of, for example, mental health conditions or mental health services.

Phenomenology

Phenomenology is a branch of the interpretivist framework that is commonly applied to qualitative research (Alase Reference Alase2017). It assumes that knowledge about a specific concept or event can be obtained through an individual's first-hand experience (Smith Reference Smith, Osborn and Smith2003). This can range from experiences of objects or events to how the individual experiences the self and those around them (Smith Reference Smith, Osborn and Smith2003). In essence, researchers employing a phenomenological approach attempt to study how an individual describes their experiences in order to gain deeper insight into how they understand these experiences.

Data collection

As the interpretivist framework places value on an individual's experiences and how they make sense of the world, interpretivist researchers aim to obtain a rich amount of data from a small group of individuals (Alharahsheh Reference Alharahsheh and Pius2020). These individuals generally share a particular background or experience (Alharahsheh Reference Alharahsheh and Pius2020). However, some researchers feel that assessing a small sample of individuals may not be enough to achieve data saturation – the point in data collection or analysis where little or no new data addressing the research question are found (Sebele-Mpofu Reference Sebele-Mpofu2020). Nevertheless, the way in which data saturation is defined and assessed varies considerably depending on the qualitative research design, with saturation being met through rich, in-depth data within the interpretivist framework (Sebele-Mpofu Reference Sebele-Mpofu2020).

Researchers using the interpretivist approach generally obtain data through in-depth semi-structured or unstructured interviews (interviews that do not have a predetermined set of questions and are instead based on participants’ responses), or diaries recording individuals’ feelings, perceptions or beliefs regarding their experience (Reeves Reference Reeves, Albert and Kuper2008; Alharahsheh Reference Alharahsheh and Pius2020).

Data analysis

Data obtained from studies using an interpretivist framework are generally analysed inductively, identifying meaning directly from the data without a prespecified framework (Reeves Reference Reeves, Albert and Kuper2008). Owing to its focus on obtaining knowledge through the way individuals make sense of their experience in a given setting, interpretive phenomenological analysis (IPA) is a particularly suitable methodology for research using an interpretivist framework (Smith Reference Smith, Osborn and Smith2003). Owing to its theoretical flexibility, thematic analysis can also be used in studies that employ an interpretivist framework by focusing the analysis on individuals’ meanings and experiences (Braun Reference Braun and Clarke2006). This type of analysis can emphasise the social and cultural contexts that influence individuals’ experiences (Braun Reference Braun and Clarke2006). As interpretivism recognises the influence that all individuals involved in the research (including researchers) can have on findings, a reflexive thematic analysis may be a particularly useful method for this approach (Braun Reference Braun and Clarke2019). This form of analysis recognises the influence that researchers can have on developing themes and allows for the researcher's own experience and understanding of a concept to inform the development of themes (Braun Reference Braun and Clarke2019). Within this analytical approach, researchers need to be reflective about how they may have shaped the analysis of the data and themes produced (Braun Reference Braun and Clarke2019). IPA also acknowledges the influence that researchers can have on data analysis using the concept of double hermeneutics: while the participant attempts to make sense of their personal and social world, the researcher is trying to make sense of the participant attempting to make sense of their personal and social world (Smith Reference Smith, Osborn and Smith2003). Essentially, the researcher is making sense of participants’ sense-making. IPA understands that such a process will be influenced by researchers’ preconceptions and therefore the objective of this approach is to obtain a description as close to the participant's view as possible (Smith Reference Smith, Osborn and Smith2003).

To gain further understanding of how the interpretivist framework can be used to guide qualitative mental health research, we again present our hypothetical question and explore how it can be answered using the interpretivist framework.

Example: applying the interpretivist framework

How might caregiving burden be experienced by family caregivers of those using mental health services?

An interpretivist researcher could approach this question by exploring how individual family caregivers make sense of their experience of supporting a family member who uses mental health services and any burden associated with this experience. By observing how they describe their experience, researchers could understand how individual family caregivers perceive that experience. To fully understand and appreciate an individual's experience, researchers using this approach could conduct in-depth interviews with a small sample of family caregivers. The questions provided in these interviews could capture caregivers’ individual narratives of their caregiving experience, including how they make sense of this experience and the meaning they ascribe to the concept of caregiving burden based on their experience. Researchers could analyse the data produced from these interviews inductively and could use IPA to obtain information about how family caregivers understand caregiving burden through their caregiving experience. For this, researchers could explore and identify emerging themes and the connections between these themes for each interview before then looking for patterns across all cases. This allows the researcher to gain in-depth insight into each individual as well as see how those in the same context (family caregivers currently supporting someone using a mental health service) make sense of the concept of caregiving burden. Researchers employing the interpretivist approach could also work to understand and reflect on their influence on the analysis from their previous experiences and context. It may be useful in this context to discuss the analysis of findings within a small research team whose members have different experiences. This can help to broaden understanding and raise awareness of potential biases and assumptions that could influence analysis. This reflection could be recorded and discussed as part of the study. The plausibility of study findings can also be checked by taking the findings back to participants or others with lived experience involved in the research process to assess. The findings from the study could provide detailed insight into the concept of caregiving burden as it is experienced by individuals currently supporting someone using a mental health service.

By obtaining a deeper understanding of an individual's experience, we can understand more about how specific social or cultural structures can influence individuals’ perspectives of their experience (Hiller Reference Hiller2016). This is particularly important when attempting to understand complex or emotional topics that are highly subjective and influenced by societal and personal contexts (Hiller Reference Hiller2016), including mental illness. Therefore, the interpretivist approach is particularly useful for attempting to understand these complex concepts on a deeper level and the potential influence that social and cultural structures may have on knowledge surrounding these topics.

However, a potential shortcoming of this approach is that it sees knowledge as purely subjective and dependent on individuals’ experiences alone (Nudzor Reference Nudzor2009). Therefore, it has been criticised for producing findings that lack reliability (Nudzor Reference Nudzor2009). By focusing entirely on individuals’ experiences, potentially inconsistent or contradictory explanations may be provided for certain concepts (Nudzor Reference Nudzor2009). Therefore, some researchers feel that an accurate and balanced representation of specific concepts cannot be obtained using an interpretivist framework (Nudzor Reference Nudzor2009). Of course, this is a critique from the positivist perspective, which assumes that researchers would strive for reliability in their findings. Interpretivists would feel that the findings produced are a plausible account of the experience of the individuals assessed, with the assumption that, if the sample is well situated, others could learn useful information that may be of relevance if they are in a similar situation or working with those in a similar situation (Nudzor Reference Nudzor2009; Hiller Reference Hiller2016).

The three domains of reality

Critical realism suggests that there are three domains of reality: the empirical, the actual and the real (Fletcher Reference Fletcher2017). The empirical domain describes what we perceive through our senses. The actual domain comprises the events that occur within reality that we may or may not be able to perceive through the senses. The real domain refers to the mechanisms or structures that create events within the actual domain. To obtain knowledge about these mechanisms or structures that cause an event, an understanding within the empirical domain is needed, which can be obtained through qualitative investigation. For example, a mental health condition can be observable at the empirical level through exploration of peoples’ experiences. The actual domain describes events that happen when peoples’ mental health conditions are either treated or neglected (some we may be able to perceive and some we may not). These events can only be explained with reference to the real domain where mechanisms/structures, for example gender or socioeconomic status, lead to an actual event that could then be assessed empirically. The goal of critical realism is to understand the causal mechanisms that lead to the occurrence of certain events (Fletcher Reference Fletcher2017).

Methodological approaches

As the critical realist framework lies between positivist and interpretivist approaches, it is more methodologically flexible (Lawani Reference Lawani2020). This means that a wider range of methodological approaches are available for researchers employing a critical realist framework, provided that the methodology has a clear focus on establishing the causality of events, such as understanding how and why certain events may occur (Lawani Reference Lawani2020). For example, some studies have used methodologies focused on explaining the events themselves, and others on explaining the structure and context within which an event occurs (Lawani Reference Lawani2020). Triangulation, defined as the use of several methods within a study (Noble Reference Noble and Heale2019), is commonly used by critical realist researchers to obtain a richer understanding of the causal mechanisms behind a particular event or concept (Fletcher Reference Fletcher2017; Lawani Reference Lawani2020).

A range of qualitative methods can be used by researchers employing a critical realist framework, including interviews (structured, semi-structured or unstructured), case studies, observations and written material such as diaries (Fletcher Reference Fletcher2017). A hybrid inductive–deductive analytical approach (analysis using a prespecified framework but with space to add to this framework any findings obtained directly from the data) fits well with the critical realist framework, although given its flexibility either approach (inductive or deductive) is suitable alone (Fletcher Reference Fletcher2017). A range of analytical techniques can also be used when applying this framework, although thematic analysis may be particularly beneficial for critical realist researchers because of its theoretical flexibility and the fact that it can provide a detailed, multidimensional account of data by examining the experience of individuals while also identifying general patterns within the data (Braun Reference Braun and Clarke2006; Lawani Reference Lawani2020).

To understand how the critical realist framework can be used to guide qualitative mental health research, we present our hypothetical question a final time and explore how it can be approached using this framework.

Example: applying the critical realist framework

How might caregiving burden be experienced by family caregivers of those using mental health services?

Using the critical realist framework, caregiving burden could be seen as the actual event that can be explored through family members’ caregiving experiences. Thus, a researcher adopting a critical realist approach will seek to explore the causal mechanisms behind caregiving burden, looking specifically into the structure and context in which family caregivers may feel this burden. Researchers can examine this using a combination of semi-structured interviews and diary entries to allow for triangulation of information. Interview questions could focus on how individuals’ caregiving experience may influence their feelings of caregiving burden or in what context individuals feel a sense of burden. For diary entries, caregivers could be asked to write about their daily experiences of supporting a family member and their feelings about providing that support to get an idea of what happens day to day and what feelings arise on certain days. A hybrid inductive–deductive thematic analysis could be used to analyse the data produced. Various theories could be used to partially inform this analytical process, including for example the theory of self-efficacy (Bandura Reference Bandura and Adams1977) to explore whether factors related to self-efficacy might influence family members’ feelings of caregiving burden. The data could also be explored in detail to determine whether any other potential factors influence caregiving burden, such as culture or family dynamics. Research using the critical realist framework could allow researchers to identify and understand the potential mechanisms that cause or influence caregiving burden. By identifying these potential influences, interventions targeting those mechanisms can be created and applied to reduce caregiving burden.

Research using a critical realist approach could also allow researchers to gain a deeper understanding of the influence of specific constructs, for example gender or socioeconomic status, on events. This could provide insight into how we could address issues related to these constructs that could, in turn, reduce any potential negative influence they may have.

The critical realist framework does, however, have a few shortcomings. The first is that critical realism often requires researchers to use their existing knowledge of possible social structures to explain causal mechanisms (Roberts Reference Roberts2014). This may limit the information that can be obtained about the causal mechanisms assessed and their potential influence on events (Roberts Reference Roberts2014). Additionally, the number of causal mechanisms that could influence certain events are highly complex (Roberts Reference Roberts2014). Because of this, the explanations generated by some critical realists may fail to capture the complex nature of peoples’ experiences of events (Roberts Reference Roberts2014).