GoITSystems by Dr. M. Gottlieb

Mixed-Method

Overview and Definition

The mixed methods research approach refers to a research strategy that systematically combines, analyzes, and integrates qualitative and quantitative methods within a single research endeavor, with the goal of achieving a more comprehensive understanding of a phenomenon than would be attainable through a monomethod approach alone (; ). Crucially, qualitative and quantitative data are not merely placed side by side; rather, they are actively linked at the level of research design, data collection, or interpretation of findings ().

Mixed methods is increasingly recognized in Anglo-American scholarly discourse as a third methodological paradigm that productively moves beyond the traditional divide between qualitative and quantitative research (; ). While quantitative approaches draw epistemologically on Critical Rationalism, and qualitative approaches align with Symbolic Interactionism or constructivist positions, mixed methods research is philosophically grounded primarily in American pragmatism — as developed by Charles S. Peirce, William James, and John Dewey (; ). Pragmatism does not treat scientific knowledge as an absolute representation of reality, but rather as a capacity for problem-solving; research questions — not epistemological presuppositions — guide the choice of methods.

The attribution of independent paradigm status to mixed methods is not, however, without controversy. Critical voices caution that pragmatically grounded mixed methods research may unreflectively adopt the positivist features of quantitative instruments — such as standardized questionnaires with closed-ended response formats — and thereby implicitly reproduce positivist positions despite claiming to represent a third way (). Against this backdrop, careful methodological reflection on one's own epistemological assumptions remains advisable even within a mixed methods framework.

Qualitative and Quantitative Methods Compared

A systematic distinction between the two constituent methodological strands is useful for understanding the logic of the mixed methods approach:

Qualitative methods — including semi-structured interviews, observation, focus groups, grounded theory, and ethnographic studies — aim at understanding subjective experiences, meaning-making processes, and social contexts. They facilitate the exploration of complex phenomena, the generation of new hypotheses, and the development of theory grounded in empirical material. Their strengths lie in conceptual depth and contextual sensitivity; their limitations in restricted statistical generalizability ().

Quantitative methods — such as standardized surveys, laboratory experiments, and cross-sectional or longitudinal statistical studies — enable the structured collection of numerical data from large samples. They are well suited to hypothesis testing, identifying statistically significant associations and causal relationships, and producing representative, generalizable findings. Their strengths lie in precision, replicability, and generalizability; their limitations in the frequent absence of contextual depth and the subjective perspective of participants ().

The mixed methods approach integrates both perspectives in order to combine their respective strengths and compensate for — rather than eliminate — their respective weaknesses. What is decisive is that integration goes beyond mere juxtaposition and gives rise to substantive conclusions that neither strand could have produced independently ().

Goals and Research Interests

The decision to adopt a mixed methods approach should always be guided by a clearly articulated research interest. Kuckartz (2014) recommends posing the following guiding questions in advance: Which aspects of the phenomenon would remain inaccessible if only one methodological strand were employed? And what concrete added value does the integration of both strands offer over a monomethod design?

The literature identifies five central research interests that may motivate mixed methods designs (; ):

Triangulation: Validation and mutual cross-checking of findings from both strands — do the qualitative and quantitative results converge, complement each other, or contradict one another?

Complementarity: Elaboration and illustration of findings from one strand through the other, in order to produce a richer overall picture.

Development: The findings of one strand are used to develop the other — for example, qualitative findings informing the construction of quantitative data collection instruments.

Initiation: The deliberate search for contradictions between strands in order to open up new research perspectives.

Expansion: Broadening the thematic scope of the study by deploying methodologically distinct strands for different sub-questions.

Design Types According to Creswell and Plano Clark (2011)

Creswell and Plano Clark (2011) distinguish six core mixed methods designs. The choice of design depends directly on the research question, the research interest, and the available resources ():

Convergent Parallel Design: Qualitative and quantitative data are collected simultaneously and independently, then merged to enable comparative interpretation of findings. This design is particularly well suited to triangulation and validation purposes.

Explanatory Sequential Design: A quantitative first phase is followed by a qualitative phase that deepens and explains surprising or unclear quantitative findings. The qualitative strand helps illuminate the "why" behind statistical patterns.

Exploratory Sequential Design: The qualitative phase comes first and serves to explore a little-researched phenomenon. Its findings guide the subsequent quantitative data collection — typically in the development and validation of new measurement instruments or scales.

Embedded Design: A secondary methodological strand is nested within a dominant, overarching design (e.g., qualitative interviews embedded within an experiment), without the two strands being treated as equally weighted.

Transformative Design: One of the four core designs is framed by an overarching theoretical or ideological perspective — such as Critical Theory, feminist theory, or participatory research — that substantially shapes the research goals and questions.

Multiphase Design: Multiple sequential or concurrent phases are combined over an extended period of time, most commonly within research programs comprising several interrelated studies.

Data Analysis and Integration: Mixing and Joint Displays

A central — and in practice frequently underestimated — step in mixed methods studies is integrative data analysis. Beyond the separate analysis of each strand, the process of "mixing" — that is, the active linking of both data strands — is of particular importance (). Three fundamental transformation strategies are distinguished:

Quantitizing: Qualitative data — such as coding frequencies derived from interview analyses — are transformed into quantitative variables and subjected to statistical processing.

Qualitizing: Quantitative data — such as scale scores or cluster assignments — are interpreted qualitatively and enriched with detailed case descriptions.

Joint Displays: Qualitative and quantitative findings are visualized and analyzed in parallel through integrative representations — such as side-by-side tables or typology matrices — in order to make convergences, complementarities, or contradictions visible (). Software such as MAXQDA has long offered dedicated functionality specifically for these integrative analytical steps.

Meta-Inferences as a Quality Criterion

The overarching quality criterion of a mixed methods study is the generation of meta-inferences: substantive conclusions that emerge from the integration of both methodological strands and that go beyond the individual findings of either strand in isolation (; ). Venkatesh et al. (2013) developed specific guidelines for IS researchers that address three core questions: (1) Is the mixed methods approach appropriate for the research question at hand? (2) How are valid meta-inferences derived from integrated findings? (3) How is the quality of these meta-inferences assessed and communicated?

To evaluate quality criteria in mixed methods studies, Tashakkori and Teddlie (2009) propose the concept of inferential quality, which encompasses dimensions such as design quality (the fit between research question and design), interpretive rigor (the consistency of conclusions with the data), and integrative efficacy (the degree to which both strands have been genuinely integrated).

Distinction from Triangulation

Mixed methods is frequently equated with triangulation in the literature — a conflation that is, however, imprecise. Triangulation in Denzin's sense refers primarily to a validation strategy in which the same phenomenon is examined from multiple perspectives in order to assess the reliability of findings. Mixed methods designs may incorporate triangulation, but they extend considerably beyond it in conceptual scope: they pursue not only validation, but also complementarity, development, initiation, and expansion; they require the integration of qualitative and quantitative data as a constitutive feature; and — unlike triangulation — they are discussed as an independent methodological approach in their own right (). Methodological triangulation is therefore best understood as one possible — but by no means necessary — element of mixed methods studies.

Relevance for Information Systems Research

In German-speaking Information Systems research (Wirtschaftsinformatik, WI), mixed methods designs are comparatively well established. An analysis of the WI methodological profile between 2007 and 2012 found that 31% of the multi-method contributions examined combined qualitative and quantitative methods — a markedly higher proportion than in Anglo-American Information Systems Research (ISR), where, according to Venkatesh et al. (2013), fewer than 5% of empirical publications employed a mixed methods design ().

In their landmark MIS Quarterly publication, Venkatesh et al. (2013) derived explicit guidelines for IS researchers aimed at enabling a methodologically rigorous application of mixed methods. These guidelines are particularly relevant for sociotechnical research endeavors — including studies on technology acceptance (e.g., TAM-based research), digital transformation, algorithmic decision support, or AI governance — since such topics require both statistical pattern identification and subjective user perspectives to arrive at a complete picture of the phenomenon under investigation ().

Critical Assessment and Limitations

Despite its advantages, the mixed methods approach entails substantial challenges that should be accounted for during research planning:

Methodological competence: Mixed methods studies require solid expertise in both methodological traditions. Researchers whose primary background lies in only one strand risk treating the other in a superficial or analytically inadequate manner.

Time and resource demands: Data collection, analysis, and integration across two strands are considerably more demanding than monomethod studies; this should be factored into project planning realistically.

The integration problem: Merely conducting both sub-studies does not in itself guarantee genuine integration. If qualitative and quantitative findings are ultimately reported in isolation, the core added value of the approach is never realized ().

Paradigmatic incoherence: Combining methods from ontologically incompatible paradigms may give rise to epistemological contradictions if the philosophical foundations of the respective methods are not sufficiently reflected upon ().

References

  • Venkatesh, V., Brown, S. A., & Bala, H. (2013). Bridging the qualitative–quantitative divide: Guidelines for conducting mixed methods research in information systems. MIS Quarterly, 37(1), 21–54. https://doi.org/10.25300/MISQ/2013/37.1.02
  • Creswell, J. W., & Plano Clark, V. L. (2011). Designing and conducting mixed methods research (2nd ed.). SAGE.
  • Creswell, J. W. (2003). Research design: Qualitative, quantitative, and mixed methods approaches (2nd ed.). SAGE.
  • Tashakkori, A., & Teddlie, C. (2009). Foundations of mixed methods research: Integrating quantitative and qualitative approaches in the social and behavioral sciences. SAGE.
  • Kuckartz, U. (2014). Mixed Methods: Methodologie, Forschungsdesigns und Analyseverfahren. Springer VS. https://doi.org/10.1007/978-3-531-93267-5
  • Johnson, R. B., & Onwuegbuzie, A. J. (2004). Mixed methods research: A research paradigm whose time has come. Educational Researcher, 33(7), 14–26. https://doi.org/10.3102/0013189X033007014
  • Hampson, T., & McKinley, J. (2023). Problems posing as solutions: Criticising pragmatism as a paradigm for mixed research. Studies in Second Language Acquisition. https://doi.org/10.1017/S0272263123000160
  • Guetterman, T. C., Creswell, J. W., & Kuckartz, U. (2015). Using joint displays and MAXQDA software to represent the results of mixed methods research. In M. T. McCrudden, G. J. Schraw, & C. W. Buckendahl (Eds.), Use of visual displays in research and testing (pp. 145–176). Information Age Publishing.
  • Schreiner, M., Hess, T., & Benlian, A. (2015). Gestaltungsorientierter Kern oder Tendenz zur Empirie? Zur neueren methodischen Entwicklung der Wirtschaftsinformatik (WIM Working Paper 1/2015). Ludwig Maximilian University of Munich. https://epub.ub.uni-muenchen.de/96484/

Publications

Social Networks