Configure, run, and interpret
Each lab exposes its own configuration (variable mapping, method picker, optional Low-Code controls). Native Run executes the in-app engine. Pro-Code Python/R views are templates: Run code unavailable — edited code is not executed.
- Association is not causation unless a causal design and assumptions are explicit.
- Simulation and DSGE outputs are model-implied under stated parameters or calibrations.
- Some methods withhold confidence intervals or p-values (for example cluster CI below 30 entities).
- Experimental methods may return coefficients without full inferential certification.