Joshua Tokle is a senior researcher and data scientist at AIR. He works to develop and implement statistical and machine learning methods with applications to science policy research, particularly in the areas of statistical matching and network analysis. He previously worked at the Center for Statistical Research and Methodology the U.S. Census Bureau, where his research areas included record linkage and nonparametric Bayesian imputation. His projects there included a record linkage error estimation study using Census Coverage Measurement data and a pilot study on adaptive design for survey nonresponse using American Communities Survey paradata. In the past Joshua worked as a consultant for FutureAdvisor where he developed algorithms for income projection and retirement portfolio diversification. Joshua received a PhD in mathematics from the University of Washington, where his research focus was stochastic analysis and probability theory.