The Promise & Limitations of Synthetic Data for Quantitative Research
Synthetic data, generated using AI/machine learning (ML) models, is intended to mimic the statistical structure of real-world data. The term “synthetic data” encompasses a wide variety of emerging techniques used in both qualitative and quantitative methodologies. These range from segment personas, digital twins, and sample augmentation to applications that support exploratory analysis, stress-testing, or sensitivity analysis, and simulation of concept or message performance under different assumptions.