Misq methods

Research Design & Methods (misq-methods)

When to trigger

• You have a theory or design propositions but no defensible way to test/evaluate them
• The method may not match the tradition or the claim (e.g., a causal claim with a correlational design)
• A reviewer asks "how do you know the artifact works?" or "what identifies this effect?"
• You need to decide what fits inside the category page limit (which counts everything)

Match the design to the tradition and claim

MISQ spans four traditions, so there is no single mandated method. Pick the design that the claim requires.

Tradition	Typical designs	The design must establish
Behavioral	Lab/online experiment, field experiment, multi-wave survey, panel	Internal validity, construct validity, and (for surveys) procedural remedies for common-method bias
Design science	Build-and-evaluate of an IT artifact	That the artifact is novel and useful for a real problem — utility demonstrated, not asserted
Economics of IS	Natural experiment, DiD, IV, RD, structural model	A credible identification strategy for the causal/economic effect
Organizational	Case study, ethnography, mixed methods, longitudinal field	Trustworthiness, rich context, and a transparent path from data to constructs

Design science: plan the evaluation up front

A MISQ design-science paper lives or dies on evaluation. Following Hevner et al. (2004), decide before building how you will demonstrate utility: held-out benchmarks against credible baselines, a controlled experiment or A/B field deployment, simulation, or expert evaluation — and tie each back to the design propositions. State the problem's relevance, the artifact's novelty, and the evaluation criteria so reviewers can judge rigor and relevance. "We built it and it ran" is not an evaluation.

Behavioral and economics: design out the threats early

• Behavioral surveys: build in procedural separations against common-method bias — temporal/psychological/source separation, validated scales, attention/manipulation checks — because statistical fixes alone will not convince reviewers later.
• Economics of IS: anchor identification in a real source of exogenous variation (a platform/policy change, staggered rollout, a breach, a system go-live). Pre-commit the comparison and the assumptions you will defend.

Mind the page budget

Because supplementary materials are discouraged and the page limit counts text, tables, figures, references, and appendices, design a study whose evidence fits the chosen category. Scope the design to the page budget rather than planning to offload it to an online appendix.

Execution bridge (StatsPAI / Stata MCP)

For the empirical / causal lane, estimate and audit rather than only specify. Full map: execution-with-mcp. MISQ is empirical IS — surveys, econometric panels, experiments, and design science; the chain below serves the causal / econometric lane, while design-science artifacts use their own evaluation standards.

• detect_design → recommend → fit with as_handle=true → audit_result to enumerate the checks the design owes.
• Panel / staggered DiD: callaway_santanna / sun_abraham + bacon_decomposition
  • honest_did_from_result. IV: effective_f_test + anderson_rubin_ci. RDD: rdrobust + mccrary_test.
• Experiments: randomization-based inference and romano_wolf for the many-outcome family-wise correction reviewers expect.

Match the toolchain to the reviewer pool, and report the effect size the venue wants. A run end-to-end (synthetic data, real returns) is in the JF execution walkthrough.

Checklist

• Design matches the tradition and the strength of the claim
• Behavioral: validity threats and CMB designed out, not just measured
• Economics: a named source of exogenous variation and a defensible identification logic
• Design science: an evaluation plan tied to design propositions and credible baselines
• Qualitative: a transparent, traceable path from data to constructs
• The evidence fits the category page limit

Anti-patterns

• A causal IS claim resting on a cross-sectional correlation.
• A design-science artifact with no comparison and no real-problem evaluation.
• Single-source, single-wave self-report with no procedural CMB remedies.
• A design that only "fits" by exporting half the evidence to discouraged supplements.

Output format

【Tradition & design】experiment / survey / DiD-IV-RD / build-and-evaluate / qualitative
【Identification or evaluation】source of variation OR evaluation plan + baselines
【Validity threats handled】CMB / confounds / trustworthiness
【Fits page budget?】yes / trim
【Next step】misq-data-analysis