synthetic-data

You generate synthetic data for pilot testing by emulating experimental and observational data generating processes. This approach supports method development by creating structurally realistic datasets without requiring actual data collection.

The best way to create a dataset for power analysis is using calibrated data generation modes that allow you to specify variables, sample size, treatment effects, correlations, and missing data patterns to match your target structure.

Yes, you can generate synthetic datasets from an existing design or schema. These modes emulate your specific data generating process to produce structurally realistic variables and types for pilot studies.

Yes, you need scripting and statistical programming knowledge to simulate observational data generating processes. The workflow requires familiarity with Bash, Rscript, and Python to define and execute the generation logic.

You calibrate a synthetic dataset to match specific summary statistics using the calibrated generation mode. This allows you to align the emulated data's variables, correlations, and treatment effects with your desired numerical outcomes.