From entrepreneurship-theory-and-practice-skills
Executes and defends ETP manuscript analysis: survival models, endogeneity/IV, SEM/CFA, and qualitative coding rigor for new-venture inference.
How this skill is triggered — by the user, by Claude, or both
Slash command
/entrepreneurship-theory-and-practice-skills:etp-data-analysisThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
- The estimator is chosen but endogeneity, selection, or survivorship is not yet addressed in the numbers
ETP wants analysis that the theory can stand on and that survives the new-venture inference traps. Because the journal is method-plural, "analysis" differs by branch — but every branch must (a) match the estimator to the outcome and the entrepreneurial data structure, (b) confront endogeneity/selection head-on, and (c) report uncertainty honestly. ETP house style follows APA: report effect sizes and confidence intervals, not a forest of significance asterisks standing in for substance.
ETP's dual mandate reaches the results: translate coefficients into the venture-relevant scale (a hazard ratio as "ventures with X fail 30% faster," a marginal effect as "one more co-founder shifts funding probability by Y points"). A practitioner implication needs a magnitude, not a p-value.
A team wants to test whether accelerator participation raises venture survival, using cohorts admitted across several years and a binary "survived to year 3" outcome. The first draft runs OLS on the 0/1 outcome with year and region controls. Three ETP-specific upgrades: (1) the outcome is fundamentally time-to-event — recast as a discrete-time hazard or Cox model with competing risks (acquired vs. shut down vs. still operating), recovering the timing and censoring OLS discards; (2) accelerators select promising ventures, so survival differences may be selection, not treatment — exploit a plausibly exogenous admission threshold (a scoring cutoff supports a regression-discontinuity or fuzzy-RD design) rather than controls alone; (3) because cohorts enter in staggered years and the program changed over time, a naive two-way fixed-effects "treatment" coefficient can be biased — use a modern staggered-DID estimator with a pre-trend check. Finally, report the hazard ratio with a CI and translate it: "admitted ventures fail roughly 25% slower over three years," a magnitude an accelerator director can act on.
【Journal】Entrepreneurship Theory and Practice
【Branch】quantitative / SEM / qualitative
【Outcome→estimator】outcome type + matched model
【Selection/survivorship】how addressed in the numbers
【Endogeneity】IV / control-function / lagged / dynamic panel + exclusion logic
【Inference】effect sizes + CIs (APA); CMB if self-report
【Magnitude for practice】coefficient translated to venture scale
【Next skill】etp-contribution-framing
npx claudepluginhub brycewang-stanford/awesome-journal-skills --plugin entrepreneurship-theory-and-practice-skillsRuns and reports statistical analysis for JBV entrepreneurship manuscripts, selecting estimators (survival, selection, panel, experiment) and addressing survivorship, attrition, and endogeneity.
Helps design and defend research methods for Entrepreneurship Theory and Practice (ETP) manuscripts, addressing selection bias, survivorship, and process observation issues in new venture studies.
Runs and defends quantitative (regression/SEM/robustness) or qualitative (coding/abduction/trustworthiness) analysis for JMS manuscripts, focusing on credibility of results.