Parameter Recovery Checker

Parameter recovery is a simulation study that tests if a model's true parameters can be accurately retrieved from simulated data. It validates model identifiability and guards against misinterpreting fitted values.

To check parameter identifiability, define ground-truth parameters, simulate data, fit the same model using your exact pipeline, and compute recovery metrics like r, bias, and RMSE to evaluate recoverability.

Yes, you can run a parameter recovery study for computational cognitive models including DDM, RL, and ACT-R. The workflow evaluates parameter tradeoffs and model recoverability across these specific frameworks.

Parameter tradeoffs and bias occur when true parameters cannot be independently recovered from the data. A parameter recovery study diagnoses this by reporting r, bias, and RMSE between ground-truth and fitted values.

You should conduct a parameter recovery study before data collection to evaluate model recoverability and plan experiments. It validates novel models and audits analyses by ensuring fitted values accurately reflect ground truth.