About Thailand Metric Echelon
We formed around a stubborn observation: talented analysts rarely fail on algebra alone — they lose rooms when estimates cannot be narrated with limitations attached. The academy keeps interpretation clinics staffed, rubrics explicit, and datasets compact enough to audit without a subpoena.
Principles
- Limitations belong beside estimates, not in closing slides.
- Reproducibility beats performer charisma in live sessions.
- Certificates reward artifacts mentors can inspect, not click counts.
- Thai policy and finance contexts stay in examples, not decoration.
Timeline
Team
Sets cohort standards, mentors capstones, and keeps evaluation rubrics aligned with employer briefings.
Regression interpretation for policy audiences; publishes on labour transitions.
Forecasting and nowcasting bridges for Thai release calendars.
Reproducible pipelines, README discipline, and merge paranoia as a virtue.
Python adoption for teams migrating from point-and-click workflows.
Maps candidate goals to cohort fit without overselling intensity.
Coordinates employer letters and prerequisite checks for blended tracks.
Milestone nudges, accessibility accommodations, and peer-review logistics.
Nonlinear models and marginal effects for health and labour examples.
Panel data and firm dynamics with honest attrition appendices.
Causal designs for field teams; emphasizes falsification tests.
Cross-over labs between prediction metrics and policy questions.
Financial econometrics vocabulary that risk teams actually use.
Voices from cohorts
Short verdict: Bayesian Econometrics Intensive priors section reads like ethics, not decoration.
Discrete Choice Modeling nested logit lab clarified substitution language our old slides hand-waved through — still dense on notation.
Machine Learning for Econometrics leakage checklist caught a temporal split bug our data science pod missed; awkward Monday, useful Tuesday.
Data Wrangling for Economists (Python) — paranoid join worksheet is pinned above my desk; README exercise was humbling in a productive way.
Capstone milestone rubric stopped me from polishing slides before fixing the data audit; slower weeks, cleaner deliverable.
Causal Inference Workshop falsification appendix now travels with every internal memo — DAG session was sharper than expected.
Microeconometrics Lab nonlinear section: managers noticed marginal-effect charts before they noticed p-values.
Forecasting Macroeconomic Indicators fan-chart workshop changed how we argue about uncertainty — not about being louder, about being explicit.