In industry, practitioners often stitch together incomplete datasets from disparate sources to get a more complete picture and collect additional data if their existing set is not sufficient for rigorous modeling. To facilitate analysis in situations where users must decide which data source to use and which to sample next, Gupta et al. propose Online Moment Selection. Online moment selection applies the generalized method of moments (GMM) to estimate a statistical parameter (identified by a set of moment conditions) and to decide which data sources to query. They also present two adaptive strategies – explore-then-commit (OMS-ETC) and explore-then-greedy (OMS-ETG) to select data sources based on the estimated asymptotic variance of the target parameter.