--- name: etp-methods description: Use when designing or defending the research design for an Entrepreneurship Theory and Practice (ETP) manuscript — matching a method to an entrepreneurial question and handling the selection, survivorship, and process-observation problems unique to new ventures. Designs the study; it does not estimate it (etp-data-analysis) or frame the contribution (etp-contribution-framing). --- # Methods & Research Design (etp-methods) ## When to trigger - You are choosing a design and unsure it can observe the entrepreneurial process you theorize - A reviewer flags selection into entrepreneurship, survivorship bias, or an undocumented sampling frame - The design is cross-sectional but the theory is a venturing *process* - You are weighing inductive qualitative work, experiments, archival venture panels, or mixed methods - Construct validity for an entrepreneurial measure (alertness, effectuation, SEW, passion) is in doubt ## ETP is method-plural but theory-first ETP publishes strong **quantitative, qualitative, and mixed-methods** work; it has no single preferred method. The constant is that the design must let you **build or test entrepreneurship theory** and that it confronts the inference problems that are *endemic to studying new ventures*. Choose the design that can actually observe the mechanism, then design against the entrepreneurship-specific threats below — these are the things ETP referees reflexively probe. | Entrepreneurial question / claim | Fitting design | |----------------------------------|----------------| | How ventures/opportunities emerge; sensemaking; identity work | Inductive qualitative (Gioia, process), longitudinal cases | | Founder decision logic / cognition under uncertainty | Experiments, conjoint / policy-capturing, verbal protocols | | Antecedents/consequences across many ventures | Archival venture panels (Crunchbase, PitchBook, GEM, PSED/KFS, registries) | | Founding, exit, IPO, failure, time-to-event | Survival / event-history models | | Financing signals (VC, angels, crowdfunding) | Field/natural experiments; panels with funding events | | Family-firm behavior (SEW) | Survey + archival; sometimes dyadic/family data | | Theory-building plus generalization | Mixed methods (qual to build, quant to test) | ## The four threats ETP referees reflexively probe - **Selection into entrepreneurship.** Founders self-select; comparing them to non-founders, or treating "who became an entrepreneur" as random, invites a Heckman/Roy or design-based response. Ignoring selection is a near-automatic flag. - **Survivorship.** New-venture samples lose failed ventures fast; a sample of survivors silently conditions on success and biases any "what makes ventures succeed" claim. Push the frame upstream to nascent/registry panels (PSED-style) that capture pre-founding and failed ventures. - **Process observability and temporal precedence.** To claim antecedents or a process, the design must order constructs in time — longitudinal waves, pre/post a shock, staged measurement. A single cross-section cannot carry a venturing-process claim. - **Construct validity for entrepreneurial constructs.** Effectuation, alertness, passion, SEW, and entrepreneurial orientation are contested measures; cite the validated scale, report reliability, and show discriminant validity from the generic relabel (e.g., effectuation ≠ improvisation; SEW ≠ ownership concentration). ## Qualitative rigor at ETP For inductive work, ETP expects a visible **trustworthiness** apparatus: a clear data structure (1st-order codes → 2nd-order themes → aggregate dimensions), an audit trail, representative quotations tied to informants, and a process model as the output. Sampling logic (theoretical, not convenience) and saturation should be argued. A rich story without this scaffolding reads as journalism to a methods-attentive ETP reviewer. ## Execution bridge (StatsPAI / Stata MCP) For the **empirical / causal lane**, estimate and audit rather than only specify. Full map: [`execution-with-mcp`](../../../shared-resources/empirical-methods/execution-with-mcp.md). ETP is entrepreneurship, where selection and survival bias are pervasive — foreground identification and selection corrections. - `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](../../../Journal-of-Finance-Skills/resources/worked-examples/02-execution-walkthrough.md). ## Checklist - [ ] The design can actually observe/test the theorized entrepreneurial mechanism - [ ] Selection into founding is modeled or addressed by design - [ ] Survivorship/attrition handled in the sampling frame, not patched statistically later - [ ] Temporal precedence supports any antecedent/process claim - [ ] Entrepreneurial constructs use validated scales with reliability + discriminant validity - [ ] Qualitative: data structure, audit trail, theoretical sampling, quotations, process model - [ ] Sample frame, coverage, and construction documented (especially for novel datasets) ## Anti-patterns - **Survivor-only sample** treated as representative of venture creation - **Cross-sectional self-report** used to claim a dynamic entrepreneurial process - **Selection ignored** — founders compared to non-founders as if assignment were random - **Construct relabeling** — a generic measure renamed as an entrepreneurial one - **Convenience startup sample** with an undocumented frame - **Thin qual** — vivid quotes, no data structure or audit trail ## Worked vignette (illustrative) A study claims that founders' prior startup failure *causes* faster pivoting in their next venture, using a panel of funded startups. An ETP methods reviewer sees three traps at once. **Survivorship**: the panel only includes ventures that secured funding and survived long enough to be observed pivoting — failures that died before pivoting are invisible, so "failure speeds pivoting" may be an artifact of who survived. **Selection**: founders who start again after failure differ systematically from those who quit; the comparison is not random. **Temporal precedence**: pivoting must be measured *after* the prior-failure exposure, not concurrently. The design fix moves the frame upstream to a nascent/registry panel (PSED-style) that captures ventures from formation regardless of survival, models the founding/restart decision (Heckman or a design-based shock to who restarts), and times the pivot measurement strictly after the failure. Only then does the antecedent claim survive — and the analysis hands off to an event-history model in `etp-data-analysis`. ## Output format ```text 【Journal】Entrepreneurship Theory and Practice 【Question→design fit】design + why it observes the entrepreneurial mechanism 【Selection】founding-choice strategy (model / design-based / N/A + why) 【Survivorship/attrition】sampling-frame handling 【Temporal precedence】how constructs are ordered in time 【Construct validity】scale + reliability + discriminant evidence 【Qual rigor】data structure / audit trail / saturation (if applicable) 【Next skill】etp-data-analysis ```