Improving out-of-population prediction: The complementary effects of model assistance and judgmental bootstrapping
International Journal of Forecasting
Given an old domain Dold and a new domain Dnew, model-assisted judgmental bootstrapping is a five-step process. (a) Based on the target case in the new domain, an expert identifies a set of predictor values to input into the old-domain model in order to obtain a prediction that the expert can consult later. (b) The old-domain model creates a prediction for the input designated. (c) The expert reviews the output in light of the inputs and records their best estimate for the target case in the new domain. (d) When a sufficient number of pairs of expert estimates and new cases are in hand, they are used to train a judgmental bootstrapping model to predict the expert's estimates. (e) The bootstrapping model is ready to forecast new cases in the new domain.
Abstract
We propose and test a method for out-of-population prediction termed model-assisted judgmental bootstrapping, which leverages a predictive model from one domain combined with expert judgment to generate training data and subsequently a predictive model for a new domain. In a preregistered experiment (N=1440), we assessed the predictive accuracy of this method in increasingly challenging environments. We also analyzed the individual contributions of two techniques that underlie the method: model-assisted estimation and judgmental bootstrapping. Our findings revealed that both techniques significantly improved predictive accuracy. Furthermore, their impacts were complementary: model-assisted estimation provided the largest accuracy gains in the least demanding environment, while judgmental bootstrapping did so in the most challenging environment. Our results suggest that model-assisted judgmental bootstrapping is a promising technique for creating predictive models in domains in which outcome data are not available.
Citation
BibTeX
@article{hardy2025improving,
title={Improving out-of-population prediction: The complementary effects of model assistance and judgmental bootstrapping},
author={Hardy, Mathew D and Zhang, Sam and Hullman, Jessica and Hofman, Jake M and Goldstein, Daniel G},
journal={International Journal of Forecasting},
volume={41},
number={2},
pages={689--701},
year={2025},
publisher={Elsevier}
}