Evaluation-Oriented Mixed Methods Meta-Inference
Also known as: evaluation MMR meta-inference, evaluation-focused meta-inference, mixed methods evaluation inference, meta-inference in evaluation research
Evaluation-oriented mixed methods meta-inference is a rigorous concluding process in program evaluation research in which the researcher integrates inferences drawn from both quantitative and qualitative strands of a mixed methods study into a single, coherent, higher-order conclusion. This meta-inference is explicitly anchored to evaluation questions — such as program worth, merit, or impact — and is judged by dual quality criteria: inferential consistency and interpretive consistency across strands.
Read the full method
Sign in with a free account to read this section.
Method map
The neighbourhood of related methods — select a node to explore.
When to use it
Use evaluation-oriented mixed methods meta-inference when a program evaluation study combines quantitative outcome data with qualitative process or context data and requires a unified evaluative conclusion — for instance, when funders or policymakers need a single evidence-based judgment about program merit or worth. It is especially appropriate when quantitative and qualitative strands address different but complementary evaluation questions that must be reconciled. Do NOT use this approach when only one data type is collected, when the evaluation timeline prevents rigorous analysis of both strands before reporting, or when stakeholders require only descriptive reporting without integrative conclusions.
Strengths & limitations
- Produces a single, defensible evaluative conclusion grounded in both quantitative and qualitative evidence.
- Explicitly surfaces tensions or contradictions between strands, improving the honesty and usefulness of evaluation findings.
- Quality criteria for meta-inference (inferential and interpretive consistency) provide an auditable standard for evaluators and reviewers.
- Well-aligned with evaluation frameworks that require evidence of both outcomes and implementation processes.
- Increases stakeholder confidence by demonstrating that conclusions account for multiple types of evidence.
- Requires high competence in both quantitative and qualitative methods; few individual evaluators hold equal mastery of both.
- The integration step is time-consuming and cannot be rushed without undermining quality; adds cost and timeline to evaluations.
- When strands yield contradictory findings, reaching a coherent meta-inference requires explicit interpretive judgment that may be contested by stakeholders.
- No universally agreed protocol for producing meta-inference exists; evaluators must exercise substantial methodological discretion.
- Reporting meta-inferences accessibly to non-technical audiences without oversimplifying the evidence base is challenging.
Frequently asked
How is meta-inference different from simply summarizing both strands?
Summarizing strands produces two parallel conclusions side by side. Meta-inference is a higher-order, integrative step that produces a single conclusion by actively reconciling, comparing, and synthesizing what each strand found. It requires the evaluator to assess whether the strands are consistent, and if not, to explain why and what that means for the overall evaluative judgment.
What are inferential consistency and interpretive consistency?
Inferential consistency asks whether the conclusions from the quantitative and qualitative strands are logically compatible — they do not have to be identical, but they should not fundamentally contradict each other without explanation. Interpretive consistency asks whether the meta-inference makes sense as a coherent whole given the full body of evidence, considering context, stakeholder input, and the evaluation purpose.
What should I do when the strands contradict each other?
Contradictions between strands are scientifically valuable, not a problem to hide. The evaluator should report the divergence explicitly, investigate possible explanations (different populations sampled, different time points, different constructs measured), and craft a meta-inference that honestly characterizes what is known, what is uncertain, and what additional evidence would resolve the tension.
Does this design require a specific sequence — quantitative first, then qualitative, or vice versa?
No. Evaluation-oriented meta-inference is compatible with concurrent designs (both strands collected simultaneously), sequential designs (one strand informs the other), and multiphase designs. The timing is determined by the evaluation questions and context; the meta-inference step occurs after both strands are fully analyzed regardless of collection sequence.
Is a team required, or can a solo evaluator conduct this design?
A solo evaluator can in principle conduct both strands and produce the meta-inference, but mixed methods evaluation studies are commonly conducted by interdisciplinary teams because the dual competence requirement is demanding. When using a team, it is important to plan for joint integration sessions where quantitative and qualitative analysts together build the meta-inference rather than simply exchanging reports.
Sources
- Tashakkori, A., & Teddlie, C. (Eds.). (2010). SAGE Handbook of Mixed Methods in Social and Behavioral Research (2nd ed.). SAGE Publications. ISBN: 978-1412972666
- Mertens, D. M. (2009). Transformative Research and Evaluation. Guilford Press. ISBN: 978-1606230541
How to cite this page
ScholarGate. (2026, June 3). Evaluation-Oriented Mixed Methods Meta-Inference. ScholarGate. https://scholargate.app/en/research-design/evaluation-oriented-mixed-methods-meta-inference
Which method?
Set this method beside its closest kin and read them side by side — the library lays the books on the table; the choice is yours.
- Concurrent Triangulation Mixed Methods DesignResearch Design↔ compare
- Evaluation-focused exploratory sequential mixed methodsResearch Design↔ compare
- Evaluation-Focused Multiphase Mixed MethodsResearch Design↔ compare
- Mixed Methods Meta-InferenceResearch Design↔ compare
- Transformative Mixed Methods DesignResearch Design↔ compare