Title: Bayesian factor analysis for estimating a non-randomised intervention via multi-outcomes observational panel data
Authors: Silvia Montagna - University of Turin (Italy) [presenting]
Pantelis Samartsidis - University of Cambridge (United Kingdom)
Daniela de Angelis - University of Cambridge (United Kingdom)
Shaun Seaman - University of Cambridge (United Kingdom)
Andre Charlett - Public Health England (United Kingdom)
Matthew Hickman - University of Bristol (United Kingdom)
Abstract: The problem of estimating the effect of a non-randomized binary intervention on multiple outcomes of interest is addressed by using time series data on units that received the intervention (treated) and units that did not (controls). One popular estimation method in this setting is based on the factor analysis (FA) model within the Rubin causal inference framework. We propose a model that extends the FA model for estimating intervention effects by jointly modelling the multiple outcomes to exploit shared variability, and assuming an auto-regressive structure on factors to account for temporal correlations. Our simulation studies show that the proposed method can improve the precision of the intervention effect estimates and achieve better control of the type I error rate (compared with the FA model), especially when either the number of pre-intervention measurements or the number of control units is small. We apply our method to estimate the effect of stricter alcohol licensing policies on alcohol-related harms.