Department of Statistics, Computer Science, Applications “Giuseppe Parenti”, Università degli Studi di Firenze, Viale Morgagni 59, 50134 Firenze, Italy
Prof. Dr. Fabrizia Mealli
Department of Statistics, Computer Science, Applications “Giuseppe Parenti”, Università degli Studi di Firenze, Viale Morgagni 59, 50134 Firenze, Italy
Summary: The Rubin Causal Model (RCM) is a framework that allows to define the causal effect of an intervention as a contrast of potential outcomes. In recent years, several methods have been developed under the RCM to estimate causal effects in time series settings. None of these makes use of autoregressive integrated moving average (ARIMA) models, which are instead very common in the econometrics literature. In this paper, we propose a novel approach, named Causal-ARIMA (C-ARIMA), to define and estimate the causal effect of an intervention in observational time series settings under the RCM. We first formalise the assumptions enabling the definition, the estimation and the attribution of the effect to the intervention. We then check the validity of the proposed method with a simulation study. In the empirical application, we use C-ARIMA to assess the causal effect of a permanent price reduction on supermarket sales. The CausalArima R package provides an implementation of the proposed approach.
26 August 2019 in Journal of Business & Economic Statistics
This paper revisits results from a field experiment, conducted in Florence, Italy, to study the effects of incentives provided to high school teens to motivate them to visit art museums. In the experiment, different classes of students were randomized to three types of encouragement and were offered a free visit to a main museum in the city. Using the principal stratification framework, the paper explores causal pathways that may lead students to increase future visits, as induced by the encouragement received, or by the individual experience of the proposed free museum visit, or by the spillover of classmates’ experience. We do so by estimating and interpreting the causal effects of the three forms of encouragement within the principal strata defined by compliance behaviors. Bayesian inferential methods are used to derive the posterior distributions of weakly identified causal parameters.
Facing the SARS-CoV-2 epidemic requires intensive testing on the population to early identify and isolate infected subjects. During the first emergency phase of the epidemic, RT-qPCR on nasopharyngeal (NP) swabs, which is the most reliable technique to detect ongoing infections, exhibited limitations due to availability of reagents and budget constraints. This stressed the need to develop screening procedures that require fewer resources and are suitable to be extended to larger portions of the population. RT-qPCR on pooled samples from individual NP swabs seems to be a promising technique to improve surveillance. We performed preliminary experimental analyses aimed to investigate the performance of pool testing on samples with low viral load and we evaluated through Monte Carlo (MC) simulations alternative screening protocols based on sample pooling, tailored to contexts characterized by different infection prevalence. We focused on the role of pool size and the opportunity to develop strategies that take advantage of natural clustering structures in the population, e.g. families, school classes, hospital rooms. Despite the use of a limited number of specimens, our results suggest that, while high viral load samples seem to be detectable even in a pool with 29 negative samples, positive specimens with low viral load may be masked by the negative samples, unless smaller pools are used. The results of MC simulations confirm that pool testing is useful in contexts where the infection prevalence is low. The gain of pool testing in saving resources can be very high, and can be optimized by selecting appropriate group sizes. Exploiting natural groups makes the definition of larger pools convenient and potentially overcomes the issue of low viral load samples by increasing the probability of identifying more than one positive in the same pool.
5 January 2023 in Health Services and Outcomes Research Methodology
Clinical trials often allow patients in the control arm to switch to the treatment arm if their physical conditions are worse than certain tolerance levels. For instance, treatment switching arises in the Concorde clinical trial, which aims to assess causal effects on the time-to-disease progression or death of immediate versus deferred treatment with zidovudine among patients with asymptomatic HIV infection. The Intention-To-Treat analysis does not measure the effect of the actual receipt of the treatment and ignores the information on treatment switching. Other existing methods reconstruct the outcome a patient would have had if they had not switched under strong assumptions. Departing from the literature, we re-define the problem of treatment switching using principal stratification and focus on causal effects for patients belonging to subpopulations defined by the switching behavior under control. We use a Bayesian approach to inference, taking into account that (i) switching happens in continuous time; (ii) switching time is not defined for patients who never switch in a particular experiment; and (iii) survival time and switching time are subject to censoring. We apply this framework to analyze the synthetic data based on the Concorde study.