Theoretical frameworks used to inform qualitative mental health research: a focus on positivism, interpretivism and critical realism

Summary

Qualitative research methods, defined as the collection and analysis of non-numerical data to understand concepts and experiences, are often used to inform mental health practice and policies. When utilising qualitative research methods, it is important that the researcher uses an explicit theoretical framework to guide the study. A theoretical framework informs how researchers engage with a topic or problem, report on their work with participants, describe key concepts and address assumptions within the research questions and procedures. In this article, we describe the basic concepts underpinning three of the most commonly used frameworks in mental health research: positivism, interpretivism and critical realism. We also describe how these theoretical frameworks may guide the qualitative process, including the theoretical and methodological approaches chosen and the ways in which these theoretical frameworks can be applied in practice. To enhance understanding of these frameworks, we include examples of how such frameworks can be used in qualitative mental health research.

LEARNING OBJECTIVES

After reading this article, you will be able to:

• • understand three popular theoretical frameworks that can be used in qualitative research

• • consider the relevance of each framework for your own qualitative mental health research

• • understand how these frameworks could be applied to your own qualitative mental health research.

Qualitative research involves the collection and analysis of non-numerical data and is commonly used to understand people's beliefs, experiences, attitudes, behaviours and interactions. This design gives voice to participants and enhances involvement of everyone related to the study (Pathak 2013). Although qualitative methods have a history of being criticised for their predominantly subjective nature and ‘lack of scientific accuracy’ (Johansson 2003), these methods can provide a deep understanding of a phenomenon (Pathak 2013). This has been increasingly recognised in various disciplines, including mental health research (Peters 2010). Qualitative methods are particularly relevant when attempting to understand sensitive issues such as mental illness and mental health service use (Badu 2019). The depth of understanding that qualitative research can provide on these issues can inform the development of clinical mental health policies and plans (Badu 2019). Qualitative findings have been used to inform the development of a new long-term plan for the UK's National Health Service (NHS) (Department of Health & Social Care 2023). By exploring individuals’ subjective experiences, we can understand how clinical mental health policies are enacted in practice and identify and address any potential issues (Brown 2001). Alternatively, exploring the experiences of those using mental health services can inform the design of clinical policies to address needs or issues that are being missed within mental health services (Brown 2001). As a result, qualitative research methods have been widely used in mental health service research (Palinkas 2014).

Terminology in qualitative research

To increase the quality and acceptability of qualitative research in mental health, a theoretical framework should guide the research. The use of terminology in qualitative research, specifically in relation to theoretical frameworks, research paradigms, theories, methodologies and methods (the two terms are often used interchangeably although they do have different meanings), can be difficult to navigate so we briefly define each of these concepts here to enhance understanding and clarity. Box 1 summarises these definitions, with examples that are mentioned later in the article.

BOX 1 Terminology in qualitative research

Theoretical frameworks (research paradigms)

The general beliefs and principles that shape how a researcher sees the world and informs their research investigation. It influences what is to be studied, how it can be studied and how results are to be interpreted. Examples include:

• • positivism

• • interpretivism

• • critical realism

• • phenomenology

• • narrative theory

• • grounded theory

• • ethnography.

Theories

Concepts and propositions that explain, understand or predict certain events or phenomena. Examples include:

• • attachment theory

• • social identity theory

• • attribution theory.

Methodologies

The overarching strategy that outlines how research is conducted to address study objectives and justifies the methods used. Examples include:

• • interpretive phenomenological analysis (IPA)

• • grounded theory

• • discourse analysis.

Methods

Specific tools, techniques and procedures used to answer research questions. Examples include:

• • in data collection – interviews (structured, semi-structured, unstructured), case studies, observation, diaries

• • in data analysis – content analysis, thematic analysis, discourse analysis.

Some examples can be utilised in various forms, e.g. grounded theory as both a theoretical framework and methodology, and discourse analysis as both a methodology and method. This can be due to the concepts having overarching principles that can guide researchers' investigation of phenomena as well as a specific set of guidelines for carrying out such an investigation.

A theoretical framework (also known as a research paradigm – the terms are used interchangeably across research (Rehman 2016; Kivunja 2018)) comprises the general beliefs and principles that shape how a researcher sees the world (Kivunja 2017). It offers a conceptual lens that guides how the researcher investigates specific phenomena (Kivunja 2017; Collins 2018). Clear understanding of theoretical frameworks offers key insights into conducting qualitative research. An explicit framework provides focus for organising a study and clarifies the nature of knowledge obtained (Collins 2018). A theoretical framework also increases the rigour and credibility of research findings, as well as facilitating the development of new concepts and the transferability of these concepts (Anfara 2014).

The use of theoretical frameworks influences what is to be studied, how it can be studied and how results are to be interpreted (Kivunja 2017). This includes how a researcher may use theory (or theories) to guide their research, including the methodology and methods used (Kivunja 2018). Theory can be defined as interrelated concepts and propositions that explain, understand or predict certain events or phenomena (Kivunja 2018). Researchers may use one or more theories within their framework to help make sense of the phenomena being studied (Collins 2018). The definitions for methodologies and methods vary, but a generally well accepted definition is that methodologies refer to the overarching research strategy that outlines how research is conducted to address study objectives and provides justification for the use of certain methods or a range of methods (Alharahsheh 2020). Methods refer to specific tools and procedures used to answer research question(s) (Alharahsheh 2020).

Scope of this article

In this article, we focus on three theoretical frameworks that are commonly used in mental health research (Denzin 2017; Alharahsheh 2020) – positivism (Park 2020), interpretivism (Smith 2003) and critical realism (Archer 2016) – and ways in which they may guide the qualitative research process. These three frameworks were chosen because first, they are well-known and widely used in qualitative mental health research and thus are likely to be of value to those reading or conducting research in this field (Hiller 2016; Fletcher 2017; Park 2020). Second, we as the authors have experience and expertise in using these frameworks and so can provide an informative, comprehensive overview of each, including how they could be used in qualitative mental health research to enhance the reader's understanding.

It is important to note that there are a number of other well-established theoretical frameworks in qualitative research (including phenomenology (Alase 2017), narrative theory (Creswell 2016), grounded theory (Chun Tie 2019) and ethnography (Savage 2000)) that the reader can explore, if interested. There are also many ways in which theoretical frameworks can be used in qualitative research, and we do not provide a full description or prescriptive recommendations on how they must be used as these are beyond the scope of the article. Rather, the aim of this article is to provide some insight into the use of theoretical frameworks in qualitative mental health research, based on our own experience and expertise, that clinicians and researchers can consider alongside other frameworks when choosing the most suitable strategies for their own qualitative service or research evaluations.

Additionally, as the primary focus of this article is to promote understanding of positivism, interpretivism and critical realism, we have not included the extensive number of theories that can also be used to guide qualitative mental health research and develop a deeper understanding of the phenomena being studied, as discussion of these would be outside this article's scope. Collins & Stockton (2018) provide in-depth insight into the use of theories in qualitative research.

Finally, although we have made reference to specific methodologies and methods that would be well-suited to the theoretical frameworks discussed in this article, we have not included all possible methodologies/methods that could be used. Key ones include grounded theory (Chun Tie 2019) or discourse analysis (Hodges 2008).

Table 1 provides a summary of the methodologies, methods and limitations of the three theoretical frameworks that we have chosen, and in the following sections we explore them in greater detail.

TABLE 1 Description of three qualitative theoretical frameworks, together with suitable methodologies and methods

The positivist framework

Positivism is based on the assumption that there is a single, tangible reality that exists and our knowledge of reality is based only on what we can objectively observe, measure and record (Park 2020). This framework attempts to understand society in a similar way to empirical science, with its goal to generate universal laws that can support or explain certain behaviours (Alharahsheh 2020). Such laws are to be tested systematically against the data collected (Park 2020). The knowledge and laws that we obtain about society using this framework are thought to be accurate and consistent with reality (Park 2020). As a result of these theoretical assumptions, positivism has commonly been used to guide quantitative studies. However, it can also be used to guide qualitative research (Su 2018).

Positivism relies on the hypothetico-deductive method to verify a hypothesis, meaning that its focus is on building a testable hypothesis and developing an empirical study to either confirm or reject that hypothesis (Park 2020). Qualitative research focusing on a positivist framework aims to explore relationships between different concepts and attempts to summarise patterns identified from data collected, intending to obtain generalisable findings through, for example, formal concepts or frameworks (Su 2018). Positivist researchers often use a highly structured methodology based on a defined set of principles, using systematic research techniques to generate knowledge regarding a specific concept (Su 2018).

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 2018; Alharahsheh 2020). 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 2003; Smith 2003). 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 2003; Smith 2003).

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 2020). A content analysis is particularly suitable for this framework as the aim of this analytical approach is to uncover reality as it exists (Saraisky 2016) and is based on the understanding that meaning can be coded and counted systematically (Saraisky 2016). 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 2016). 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 2006).

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 2006). 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 2022)). 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 2003). 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 2018; Park 2020).

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 2020). 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 2020). 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.

The interpretivist framework

Interpretivism was developed as a response to the criticisms surrounding positivism (Alharahsheh 2020). Unlike positivism, which suggests that knowledge is universal and can be objectively identified, interpretivism assumes that all knowledge is grounded in human experience (Alharahsheh 2020). Individuals are thought to construct their knowledge of the world through their experiences (Hiller 2016). For someone using an interpretivist framework, the aim is to understand a particular concept through the meaning that individuals ascribe to that concept based on their personal experiences in a specific setting (Hiller 2016).

Phenomenology

Phenomenology is a branch of the interpretivist framework that is commonly applied to qualitative research (Alase 2017). It assumes that knowledge about a specific concept or event can be obtained through an individual's first-hand experience (Smith 2003). This can range from experiences of objects or events to how the individual experiences the self and those around them (Smith 2003). 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 2020). These individuals generally share a particular background or experience (Alharahsheh 2020). 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 2020). 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 2020).

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 2008; Alharahsheh 2020).

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 2008). 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 2003). 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 2006). This type of analysis can emphasise the social and cultural contexts that influence individuals’ experiences (Braun 2006). 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 2019). 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 2019). Within this analytical approach, researchers need to be reflective about how they may have shaped the analysis of the data and themes produced (Braun 2019). 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 2003). 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 2003).

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 2016). This is particularly important when attempting to understand complex or emotional topics that are highly subjective and influenced by societal and personal contexts (Hiller 2016), 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 2009). Therefore, it has been criticised for producing findings that lack reliability (Nudzor 2009). By focusing entirely on individuals’ experiences, potentially inconsistent or contradictory explanations may be provided for certain concepts (Nudzor 2009). Therefore, some researchers feel that an accurate and balanced representation of specific concepts cannot be obtained using an interpretivist framework (Nudzor 2009). 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 2009; Hiller 2016).

The critical realist framework

Critical realism is an alternative framework that sits between positivist and interpretivist approaches (Archer 2016). Critical realism agrees with the positivist notion that an objective reality exists but assumes that our knowledge of reality is situated within social and cultural structures (Archer 2016). This means that, similar to the interpretivist framework, knowledge is thought to be based on how individuals view the world through their experience. However, critical realism goes beyond the interpretivist framework to suggest that the knowledge that can be obtained captures only a small part of a deeper and larger reality that exists beyond our comprehension (Fletcher 2017). Critical realists believe that some knowledge (developed through research) is closer to an objective reality and therefore assumed to be potentially more plausible than other knowledge (Archer 2016; Fletcher 2017).

The three domains of reality

Critical realism suggests that there are three domains of reality: the empirical, the actual and the real (Fletcher 2017). 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 2017).

Methodological approaches

As the critical realist framework lies between positivist and interpretivist approaches, it is more methodologically flexible (Lawani 2020). 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 2020). 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 2020). Triangulation, defined as the use of several methods within a study (Noble 2019), is commonly used by critical realist researchers to obtain a richer understanding of the causal mechanisms behind a particular event or concept (Fletcher 2017; Lawani 2020).

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 2017). 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 2017). 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 2006; Lawani 2020).

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 1977) 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 2014). This may limit the information that can be obtained about the causal mechanisms assessed and their potential influence on events (Roberts 2014). Additionally, the number of causal mechanisms that could influence certain events are highly complex (Roberts 2014). Because of this, the explanations generated by some critical realists may fail to capture the complex nature of peoples’ experiences of events (Roberts 2014).

Conclusions

The use of a theoretical framework to guide qualitative mental health research is important to ensuring good-quality studies with enhanced credibility. We hope that this article has given researchers further understanding of how three well-known frameworks – positivism, interpretivism and critical realism – and related methodological approaches can be applied to their own qualitative research, which could inform decisions regarding their use of a framework for their own investigations. Clinicians reading this article can gain an understanding of how findings from a qualitative mental health study were generated and increase their awareness of the value and limitations of a study.

It is important to note that there are a number of other well-established theoretical frameworks in qualitative research, including phenomenology, grounded theory and ethnography, that the reader can explore alongside this article to broaden their knowledge and inform their decisions in using a framework. Those who are intending to use qualitative methods for research will also have to grapple with questions concerning how to develop patient and public involvement within their research, what expertise should be available in the research team and what the intended audience is for the research. All these considerations are outside the scope of this article but may be influenced by the theoretical framework used.