Beyond Pragmatism: Epistemological Plurality and Ontological Unity in Mixed Methods Research

Martin Robson, PhD

Introduction

Mixed methods research has become established across the social sciences, education, health and organisational studies (Creswell & Plano Clark, 2018; Tashakkori & Teddlie, 2003). Its practical appeal is clear. Organisational turnover, educational achievement, patient experience or leadership, for example, may involve measurable patterns as well as meanings, interpretations and lived experiences. A single form of inquiry may therefore illuminate only part of the phenomenon.

However, mixed methods research presents a philosophical problem. Quantitative and qualitative traditions have historically been associated with different assumptions about reality, knowledge, the relationship between researcher and researched, and the criteria by which claims are warranted (Lincoln & Guba, 1985; Crotty, 1998). The so-called paradigm wars made these differences explicit and produced the incompatibility thesis: if the underlying paradigms make contradictory ontological and epistemological claims, combining their methods within a single study risks philosophical incoherence.

Pragmatism became an influential solution because it shifts attention from philosophical allegiance to the research problem and its consequences. This has been highly productive for mixed methods research. Yet the practical permission to mix methods does not by itself answer a prior question: if epistemological assumptions contribute to the rigour of single-method inquiry, why should they become dispensable merely because methods are combined? A mixed methods design should not become a paradigmatic exemption zone.

This paper advances two related arguments. First, methodological pluralism need not require epistemological compromise. The strongest requirement for epistemological coherence occurs in the generation of evidence: the relationship among researcher, theory, participant and data should remain consistent with the epistemological orientation under which that evidence is produced. Integration may occur iteratively during data collection and analysis, provided that it does not convert an interpretivist, participant-led process into a researcher-imposed test of prior constructs, or otherwise displace the epistemological logic of either component. Second, if different epistemologies are retained within one study, their coexistence and integration require philosophical justification. The paper proposes that Monistic Idealism can provide one coherent overarching ontology within which apparently opposed positions can be understood as legitimate engagements with different aspects or levels of a unified reality. The central proposition is therefore epistemological plurality within ontological unity. The argument is offered not as the only possible philosophical foundation, but as a contribution that sits productively alongside dialectical pluralism, transformative approaches and critical realism (see, for example, Greene, 2007; Johnson, 2017; Shan, 2024).

Paradigms, Epistemology and the Mixed Methods Problem

Ontology concerns what exists or the nature of reality; epistemology concerns what can be known and the relationship between the knower and what is known. According to Crotty (1998), methodology concerns the logic and strategy through which knowledge is generated, while methods are the specific techniques used to collect and analyse data. These distinctions matter because a method is not, by itself, a paradigm. An interview is not intrinsically interpretivist, just as a questionnaire or statistical procedure is not intrinsically positivist. The philosophical character of a component depends on the purpose for which the method is used, the claims being made and the assumptions governing the inquiry.

Nevertheless, paradigmatic assumptions have consequences, especially for how evidence is generated. A quantitative study designed to estimate relationships among variables, test hypotheses or generalise from a sample is characteristically researcher-led and deductive: constructs are specified, operationalised and measured under assumptions concerning regularity, inference, researcher independence and warranted claims beyond the immediate observations. An interpretivist study is characteristically participant- or meaning-led and inductive: it seeks to understand meanings as situated, recognises the researcher's involvement in knowledge production, emphasises context and reflexivity, and remains open to concepts emerging from participants' accounts. These are characteristic rather than exclusive logics—both traditions may employ induction, deduction and interpretation—but they establish importantly different relationships among prior theory, researcher and evidence.

The philosophical difficulty in mixed methods research is therefore not simply that different techniques are placed beside one another. It is that a study may seek to produce different kinds of knowledge under different assumptions. The important question is whether these differences must be eliminated before integration can occur. Liu (2022) has argued that paradigmatic compatibility remains a neglected criterion of research quality, and that ignoring ontology and epistemology in favour of a purely methodological stance risks undermining the defensibility of mixed designs. The present paper takes that concern seriously.

Pragmatism: A Practical Solution with a Remaining Philosophical Question

Pragmatism has been widely adopted as a philosophical home for mixed methods research. Its major contribution is to make the research problem primary. Rather than allowing prior allegiance to a method or paradigm to dictate design, the researcher asks what combination of approaches will best address the problem. This has helped move mixed methods research beyond sterile methodological divisions (Tashakkori & Teddlie, 2003; Creswell & Plano Clark, 2018).

The criticism advanced here is not that pragmatism makes mixed methods invalid. On the contrary, its practical flexibility is one reason mixed methods research has flourished. The concern is that pragmatism can be used as a philosophical shortcut. Hampson and McKinley (2023), for example, question its use as a convenient solution to deeper paradigmatic problems, identifying features that render it unfit as a robust paradigm for mixed research: its character as a “paradigm of convenience,” its consequentialist view of truth and good research, and its prioritisation of the research question over ontology or epistemology. If pragmatism is invoked simply to say that researchers may use whatever works, the question of how different forms of knowledge are philosophically related can be deferred rather than resolved.

Proposition One: Preserve Epistemological Integrity in the Generation of Evidence

The first proposition is that mixed methods research is strengthened when the epistemological integrity of each form of evidence is preserved, particularly at the point at which data are generated. Where a quantitative component is framed in realist or post-positivist terms, its operationalisation, sampling, measurement and claims should remain coherent with those assumptions. Where a qualitative component is framed interpretively, participant engagement, questioning, reflexivity and the treatment of context and meaning should remain coherent with that stance. The crucial issue is not whether the eventual data are numerical or textual, but whether the process that produced them is consistent with the epistemological relationship between knower and known that the study claims to adopt.

The above approach avoids an unnecessary compromise. A researcher undertaking interpretivist inquiry cannot simply import a positivist logic of hypothesis testing into participant interviews without changing the epistemological character of the evidence being generated. If predetermined theoretical categories structure the encounter so that participants are primarily asked to confirm or disconfirm the researcher's theory, participant meanings risk being subordinated to researcher-imposed constructs. Conversely, a quantitative component designed to test theoretically specified relationships should not abandon the standards of operationalisation, measurement and inference that make its claims warranted in its own terms. Epistemological plurality therefore does not mean that anything can be combined in any manner; it requires discipline in how each form of evidence is produced.

Epistemological differentiation is not the same as temporal separation. Evidence generated under different epistemological assumptions may interact throughout a study, including during data collection, provided that the interaction does not displace the epistemological logic governing the generation of either form of evidence. The challenge is therefore not to identify a single point at which previously isolated components become integrated, but to explain the philosophical basis on which differently generated forms of knowledge can enter into an iterative relationship.

Proposition Two: Ontological Unity as the Basis for Epistemological Plurality

Philosophical coherence in mixed methods research cannot be achieved by imposing a single epistemology across all components. It can instead be achieved at the ontological level. Different epistemologies may be legitimate if they engage with different aspects, strata or levels of a reality that is ultimately unified.

The above argument draws on an earlier paper in which Monistic Idealism was used to reconcile apparently incompatible realist and idealist positions (Robson, 2009). That paper argued that many apparent contradictions arise because phenomena can be described simultaneously at different levels. Scientific realism may provide a useful account of the observable macro-world (classical physics) even where a deeper account of reality (quantum physics) challenges some of its metaphysical assumptions. Similarly, individual consciousness can be described as distinct at one level while participating in a more fundamental unity of consciousness at another, explained further below. The present paper extends that reasoning to mixed methods research: positions that appear contradictory at one level may be complementary when located within a wider ontology that permits multiple levels of description.

Monistic Idealism as an Overarching Ontology

Monistic Idealism, as developed in Robson (2009) drawing particularly on Goswami (1995), proposes consciousness as the ground of existence from which matter and individual consciousness emerge. It is monistic because the apparent division between mind and matter is ultimately located within a single ground, while remaining differentiated at other levels of description. Robson (2009) combined this idea with a stratified or layered account of reality influenced by analytical dualism and systems thinking.

The argument is grounded in a contrast between ontological assumptions associated with classical and quantum accounts of physical reality. Drawing on Capra’s (1996) account of the shift from the classical worldview to the conceptual implications of modern physics, the important point is that the two frameworks rest on markedly different, and in important respects opposing, assumptions about the character of reality. Yet the success of classical description at one level is not erased by a deeper account that challenges its ultimate assumptions. Monistic Idealism provides a way of understanding this relationship through layering: apparently opposed descriptions can retain their distinctiveness while belonging to a more encompassing unity. The opposites need not be collapsed in order to constitute the whole.  

The parallel for mixed methods research is that apparently opposed positivist and interpretivist ontologies need not be weakened or collapsed in order to coexist within a more encompassing account of reality. Their ontological differentiation instead provides the basis for corresponding epistemological differentiation. In brief, the position holds that: (1) there is a fundamental unity of reality grounded in consciousness; (2) this unity differentiates into multiple levels or strata of manifestation (physical, biological, psychological, social); and (3) each level can be investigated with methods and assumptions appropriate to its properties, without requiring that those methods or assumptions be identical across levels. A measurable organisational structure and an employee’s lived interpretation of that structure need not compete for the status of the one real account; they can be understood as knowledge generated at different levels or from different relations between knower and known.

For mixed methods research, the crucial implication is not that every researcher must adopt the same metaphysical commitments. It is that Monistic Idealism offers one philosophical model under which ontological unity does not require epistemological uniformity. Under this framework, post-positivism and interpretivism need not be collapsed into one another. A realist or post-positivist component can make appropriately qualified claims about regularities, relationships or structures. An interpretivist component can make claims about meaning, experience and situated understanding. Their findings can be different without being mutually exclusive because the ontology allows reality to be differentiated while remaining part of a larger whole. The layered character of this account has affinity with the stratified ontology of critical realism (see Bhaskar, 1975, 1993). The present argument differs in locating those differentiated levels within a Monistic Idealist ontology in which consciousness is ultimately primary.

Integration without Epistemological Homogenisation

Integration need not mean forcing the outputs of different components into a common epistemological language, nor need it be confined to a single point of interface. It means bringing differently generated forms of knowledge into relation while preserving awareness of the epistemological conditions under which each was produced. Bazeley (2024) has argued that integration is constituted by interdependence that is connected, transactional, transformative, coherent and dynamic. The present account can accommodate such iterative integration, provided that interdependence does not erase the epistemological logic governing data generation.

The researcher has a central synthesising role, but synthesis may begin before all data collection is complete. Grounded theory illustrates the point. Analysis of early participant data can legitimately inform subsequent interviews and theoretical sampling because the developing concepts arise from participants’ accounts rather than being imposed a priori as hypotheses to be tested. By the same logic, quantitative evidence may enter the developing inquiry at such a point, provided that it has itself been generated in a manner coherent with the positivist or post-positivist assumptions under which its claims are warranted. Insights from one component may therefore shape subsequent inquiry in another without requiring epistemological homogenisation.

In practice, integration can proceed iteratively through deliberate moves: (1) making explicit the epistemic status of each emerging finding and the assumptions under which its evidence was generated; (2) allowing findings from one component to inform questions, sampling or analytic attention in another where appropriate; (3) ensuring that such interaction does not replace participant-led meaning generation with researcher-imposed testing, or vice versa; (4) interpreting convergences, divergences and complementarities in light of the layered ontology; and (5) constructing an integrated account that retains the distinctiveness of each contribution while showing how they jointly illuminate the phenomenon. Integration is therefore constrained less by timing than by epistemological integrity.

Implications for Mixed Methods Design

Consider an organisational study of employee turnover. A quantitative component might ask which variables predict intention to leave and use a large sample, validated measures and regression analysis. If the purpose is explanatory or predictive in a post-positivist sense, that component should be designed and evaluated accordingly—with attention to measurement validity, sampling adequacy, statistical assumptions and the limits of inference. Data collection remains largely researcher-led. A qualitative component might ask how employees experience organisational culture and how they interpret the events leading to a decision to leave. If that component is interpretivist, interviews, reflexivity and contextual interpretation should be conducted consistently with that stance: data collection is participant-led, the researcher avoids imposing theoretical categories or leading questions, and remains open to the meanings participants themselves construct. Monistic Idealism provides an ontological warrant for combining these differently led forms of inquiry within a single study, because researcher-led and participant-led approaches are understood as engaging different levels or aspects of a unified reality.

Integration need not wait for a final synthesis stage. Early qualitative interviews might, for example, reveal an unanticipated participant-defined meaning of “supportive supervision.” That insight could shape subsequent interpretivist interviews while remaining participant-led, much as iterative analysis operates in grounded theory. It might also prompt the researcher to examine an existing quantitative measure or design a subsequent quantitative component. Conversely, a quantitative pattern might identify a subgroup or relationship warranting deeper interpretive exploration. What matters is that the quantitative evidence is generated and interpreted consistently with its post-positivist warrant, while the qualitative inquiry remains open to participant meanings rather than becoming a disguised test of the quantitative model.

A Practical Framework for Researchers and Students

A mixed methods researcher working from this position can make the philosophical architecture of the study explicit by asking the following questions:

• What is the overall research problem, and why does it require more than one form of evidence?
• What specific question is each component intended to answer?
• What kind of knowledge is each component intended to produce?
• Which epistemological and paradigmatic assumptions are appropriate to those knowledge claims?
• Are the design, data collection, analysis and quality criteria of each component internally coherent with those assumptions?
• How can the components interact iteratively during data gathering and analysis without compromising the epistemological integrity of either form of evidence?
• What ontological position justifies treating differently generated findings as contributions to the same overall inquiry?

The above sequence preserves the practical strength of mixed methods while refusing to treat philosophy as optional. It also avoids requiring students to choose one paradigm for an entire project before they have considered the different questions being asked. The overall study can be ontologically coherent while its components remain epistemologically distinct.

Conclusion

If the assumptions that underpin quantitative and qualitative inquiry matter when each is conducted on its own, they should not simply disappear when the two are combined. Mixed methods research should preserve epistemological integrity in the generation of evidence: positivist or post-positivist inquiry remains characteristically researcher-led and deductive, while interpretivist inquiry remains characteristically participant- or meaning-led and inductive. These differences need not prevent iterative integration, provided the distinct logics of evidence generation are respected.

Epistemological plurality is coherent when located within an ontology that allows apparently opposed descriptions of reality to operate at different levels of a unified whole. Monistic Idealism supplies one such ontology. Difference is retained within unity; the opposites constitute the whole. Pragmatism solved an important practical problem for mixed methods research. The argument offered here addresses the philosophical problem that remains.

References

Bazeley, P. (2024). Conceptualizing integration in mixed methods research. Journal of Mixed Methods Research, 18(3), 225–234. https://doi.org/10.1177/15586898241253636

Bhaskar, R. (1975). A realist theory of science. Harvester Press.

Bhaskar, R. (1993). Dialectic: The pulse of freedom. Verso.

Capra, F. (1996). The web of life: A new synthesis of mind and matter. HarperCollins.

Creswell, J. W., & Plano Clark, V. L. (2018). Designing and conducting mixed methods research (3rd ed.). Sage.

Crotty, M. (1998). The foundations of social research: Meaning and perspective in the research process. Allen & Unwin.

Goswami, A. (1995). The self-aware universe: How consciousness creates the material world. Tarcher/Putnam.

Greene, J. C. (2007). Mixed methods in social inquiry. Jossey-Bass.

Hampson, T., & McKinley, J. (2023). Problems posing as solutions: Criticising pragmatism as a paradigm for mixed research. Research in Education, 116(1), 124–138. https://doi.org/10.1177/00345237231160085

Johnson, R. B. (2017). Dialectical pluralism: A metaparadigm whose time has come. Journal of Mixed Methods Research, 11(2), 156–173. https://doi.org/10.1177/1558689815607692

Lincoln, Y. S., & Guba, E. G. (1985). Naturalistic inquiry. Sage.

Liu, Y. (2022). Paradigmatic compatibility matters: A critical review of qualitative-quantitative debate in mixed methods research. SAGE Open, 12(1). https://doi.org/10.1177/21582440221079922

Robson, M. J. (2009). Towards the reconciliation of idealism and realism through monistic idealism: Bridging the gap. Proceedings of the 4th Annual Postgraduate Research Conference, University of New England.

Shan, Y. (Ed.). (2024). Philosophical foundations of mixed methods research: Dialogues between researchers and philosophers. Routledge.

Tashakkori, A., & Teddlie, C. (Eds.). (2003). Handbook of mixed methods in social and behavioral research. Sage.