Inclusions analysts actually ship
Tight feedback loops on interpretation, documented limitations beside estimates, and compact datasets you can rerun without hunting for secrets in footnotes.
- Portfolio artifacts with mentor rubrics, not anonymous quizzes.
- Regression appendices that admit clustering choices in plain language.
- Office hours biased toward storytelling, not software flexing.
mentor hours logged last term — evidence that interpretation clinics stay staffed, not decorative.
Thailand Metric Echelon trains graduate students, research assistants, and desk analysts to connect models to datasets they can defend in roomfuls of sceptical readers.
The flow — from question to accountable appendix
Translate the decision into estimands peers can recognize.
Merge checks, attrition notes, and documented transforms.
Diagnostics first, stars second, limitations in the same view.
Narrative + appendix that survives another analyst opening the folder.
Featured courses
Build defensible linear models, interpret coefficients in context, and document diagnostics for stakeholder briefings.
Model seasonality, compare smoothing approaches, and communicate forecast intervals for operations planning.
Estimate fixed and random effects, test assumptions, and interpret dynamic panels with care.
Map causal questions to estimators, design falsification tests, and report results without overreach.
From our cohorts — mixed formats on purpose
Instrumental Variables Masterclass forced my committee to ask better questions because the first-stage table finally sat where readers expect it.
Financial Econometrics for Analysts — the event-study checklist stopped us from cherry-picking windows on two FX briefings.
Client in logistics analytics: forecasting cohort’s revision log template is now how we explain bands to executives.
Panel Data Methods for Thai Industries — magnitude paragraph saved my thesis defense when reviewers pushed on economic relevance.
Signals from the blog
- Reading Standard Errors Alongside Coefficients Why Thailand Metric Echelon cohorts treat uncertainty as a first-class citizen in policy briefings.
- Forecast Revisions as a Feature, Not an Embarrassment How we document revision trails when new Thai releases land and why transparency beats false precision.
- Panel Data Stories Without Heroic Claims Translating fixed effects results for industry audiences without pretending identification solved everything.
Quick questions
Do you run admissions like a mass-market MOOC?
Cohorts stay small enough for written feedback on drafts. We disclose when a module is compressed and what to read next.
Are certificates automatic?
Completion requires submitted artifacts and mentor sign-off. We state what is not included (software licenses, proprietary data procurement).
Can employers sponsor teams?
Yes — we issue cohort invoices and can align milestones to internal review cycles when arranged in advance.
Syllabus drops · no spam folder theatrics
Monthly notes on cohort openings, method readings, and dataset hygiene — unsubscribe anytime, inbox stays light.