Causal inference with binary treatments from randomization versus binary treatments from categorization.

Advertising effectiveness measurement is a fundamental problem in online advertising. Various causal inference methods have been employed to measure the causal effects of ad treatments. However, existing methods mainly focus on linear logistic regression for univariate and binary treatments and are not well suited for complex ad treatments of multi-dimensions, where each dimension could be discrete or continuous. In this paper we propose a novel two-stage causal inference framework for assessing the impact of complex ad treatments. In the first stage, we estimate the propensity parameter via a sparse additive model; in the second stage, a propensity-adjusted regression model is applied for measuring the treatment effect. Our approach is shown to provide an unbiased estimation of the ad effectiveness under regularity conditions. To demonstrate the efficacy of our approach, we apply it to a real online advertising campaign to evaluate the impact of three ad treatments: ad frequency, ad channel, and ad size. We show that the ad frequency usually has a treatment effect cap when ads are showing on mobile device. In addition, the strategies for choosing best ad size are completely different for mobile ads and online ads.

Causal inference using observational data on multiple treatments is an important problem in a wide variety of fields. However, the existing literature tends to focus only on causal inference in case of binary or multinoulli treatments. These models are either incompatible with multiple treatments, or extending them to multiple treatments is computationally expensive. We use a previous formulation of causal inference using variational autoencoder (VAE) and propose a novel architecture to estimate the causal effect of any subset of the treatments. The higher order effects of multiple treatments are captured through a task embedding. The task embedding allows the model to scale to multiple treatments. The model is applied on real digital marketing dataset to evaluate the next best set of marketing actions. For evaluation, the model is compared against competitive baseline models on two semi-synthetic datasets created using the covariates from the real dataset. The performance is measured along four evaluation metrics considered in the causal inference literature and one proposed by us. The proposed evaluation metric measures the loss in the expected outcome when a particular model is used for decision making as compared to the ground truth. The proposed model outperforms the baselines along all five evaluation metrics. It outperforms the best baseline by over 30% along these evaluation metrics. The proposed approach is also shown to be robust when a subset of the confounders is not observed. The results on real data show the importance of the flexible modeling approach provided by the proposed model.

Machine learning for causal inference, particularly at the individual level, has attracted intense interest in many domains. Existing techniques focus on controlling differences in distribution between treatment groups in a data-driven manner, eliminating the effects of confounding factors. However, few of the current methods adequately discuss the difference in treatment group sizes. Two approaches, a direct and an indirect one, deal with potential missing data for estimating individual treatment with binary treatments and different treatment group sizes. We embed the two methods into certain frameworks based on the domain adaption and representation. We validate the performance of our method by two benchmarks in the causal inference community: simulated data and real-world data. Experiment results verify that our methods perform well.

Methodology for causal inference based on propensity scores has been developed and popularized in the last two decades. However, the majority of the methodology has concentrated on binary treatments. Only recently have these methods been extended to settings with multi-valued treatments. We propose a number of discrete choice models for estimating the propensity scores. The models differ in terms of flexibility with respect to potential correlation between treatments, and, in turn, the accuracy of the estimated propensity scores. We present the effects of discrete choice models used on performance of the causal estimators through a Monte Carlo study. We also illustrate the use of discrete choice models to estimate the effect of antipsychotic drug use on the risk of diabetes in a cohort of adults with schizophrenia.

There is growing interest in developing causal inference methods for multi-valued treatments with a focus on pairwise average treatment effects. Here we focus on a clinically important, yet less-studied estimand: causal drug-drug interactions (DDIs), which quantifies the degree to which the causal effect of drug A is altered by the presence versus the absence of drug B. Confounding adjustment when studying the effects of DDIs can be accomplished via inverse probability of treatment weighting (IPTW), a standard approach originally developed for binary treatments and later generalized to multi-valued treatments. However, this approach generally results in biased results when the propensity score model is misspecified. Motivated by the need for more robust techniques, we propose two empirical likelihood-based weighting approaches that allow for specifying a set of propensity score models, with the second method balancing user-specified covariates directly, by incorporating additional, nonparametric constraints. The resulting estimators from both methods are consistent when the postulated set of propensity score models contains a correct one; this property has been termed multiple robustness. In this paper, we derive two multiply-robust estimators of the causal DDI, and develop inference procedures. We then evaluate their finite sample performance through simulation. The results demonstrate that the proposed estimators outperform the standard IPTW method in terms of both robustness and efficiency. Finally, we apply the proposed methods to evaluate the impact of renin-angiotensin system inhibitors (RAS-I) on the comparative nephrotoxicity of nonsteroidal anti-inflammatory drugs (NSAID) and opioids, using data derived from electronic medical records from a large multi-hospital health system.

Propensity score weighting is an important tool for comparative effectiveness research. Besides the inverse probability of treatment weights (IPW), recent development has introduced a general class of balancing weights, corresponding to alternative target populations and estimands. In particular, the overlap weights (OW) lead to optimal covariate balance and estimation efficiency, and a target population of scientific and policy interest. We develop the R package [PSweight](https://CRAN.R-project.org/package=PSweight) to provide a comprehensive design and analysis platform for causal inference based on propensity score weighting. PSweight supports (i) a variety of balancing weights, (ii) binary and multiple treatments, (iii) simple and augmented weighting estimators, (iv) nuisance-adjusted sandwich variances, and (v) ratio estimands. PSweight also provides diagnostic tables and graphs for covariate balance assessment. We demonstrate the functionality of the package using a data example from the National Child Development Survey (NCDS), where we evaluate the causal effect of educational attainment on income.

Propensity score methods are a part of the standard toolkit for applied researchers who wish to ascertain causal effects from observational data. While they were originally developed for binary treatments, several researchers have proposed generalizations of the propensity score methodology for non-binary treatment regimes. Such extensions have widened the applicability of propensity score methods and are indeed becoming increasingly popular themselves. In this article, we closely examine two methods that generalize propensity scores in this direction, namely, the propensity function (PF), and the generalized propensity score (GPS), along with two extensions of the GPS that aim to improve its robustness. We compare the assumptions, theoretical properties, and empirical performance of these methods. On a theoretical level, the GPS and its extensions are advantageous in that they are designed to estimate the full dose response function rather than the average treatment effect that is estimated with the PF. We compare GPS with a new PF method, both of which estimate the dose response function. We illustrate our findings and proposals through simulation studies, including one based on an empirical study about the effect of smoking on healthcare costs. While our proposed PF-based estimator preforms well, we generally advise caution in that all available methods can be biased by model misspecification and extrapolation.