Tid: 28 april 2021, kl. 13-14
Plats: This seminar is given online. E-mail Dan Hedlin if you want to attend.

Abstract

Given the growing popularity of nonprobability samples as a cost- and time-efficient alternative to probability sampling, a variety of adjustment approaches have been proposed to correct for self-selection bias in non-random samples. Popular methods such as inverse propensity-score weighting (IPSW) and propensity score adjustment by subclassification (PSAS) utilize a probability sample as a reference to estimate pseudo-weights for the nonprobability sample. A recent contribution, Kernel Weighting (KW), has been shown to be able to improve over IPSW and PSAS. However, the effectiveness of these methods for reducing bias critically depends on the ability of the underlying propensity model to reflect the true (self-)selection process, which is a challenging task with parametric regression. In this talk, we present a set of pseudo-weights construction methods, KW-ML, utilizing both machine learning (ML) methods (to estimate propensity scores) and kernel weighting (to construct pseudo-weights based on the ML-estimated propensity scores). Our results indicate that particularly boosting-based KW methods represent promising alternatives to logistic regression and result in estimates with lower bias in a variety of settings, without increasing variance. The remainder of the talk will showcase additional applications of machine learning methods in survey research.