Methodology (stat.ME)

Abstract: We consider the problem of independence testing for two univariate random variables in a sequential setting. By leveraging recent developments on safe, anytime-valid inference, we propose a test with time-uniform type-I error control and derive explicit bounds on the finite sample performance of the test and the expected stopping time. We demonstrate the empirical performance of the procedure in comparison to existing sequential and non-sequential independence tests. Furthermore, since the proposed test is distribution free under the null hypothesis, we empirically simulate the gap due to Ville's inequality, the supermartingale analogue of Markov's inequality, that is commonly applied to control type I error in anytime-valid inference, and apply this to construct a truncated sequential test.

Abstract: Comparison and contrast are the basic means to unveil causation and learn which treatments work. To build good comparisons that isolate the average effect of treatment from confounding factors, randomization is key, yet often infeasible. In such non-experimental settings, we illustrate and discuss how well the common linear regression approach to causal inference approximates features of randomized experiments, such as covariate balance, study representativeness, sample-grounded estimation, and unweighted analyses. We also discuss alternative regression modeling, weighting, and matching approaches. We argue they should be given strong consideration in empirical work.

Abstract: We consider the problem of detecting causal relationships between discrete time series, in the presence of potential confounders. A hypothesis test is introduced for identifying the temporally causal influence of $(x_n)$ on $(y_n)$, causally conditioned on a possibly confounding third time series $(z_n)$. Under natural Markovian modeling assumptions, it is shown that the null hypothesis, corresponding to the absence of temporally causal influence, is equivalent to the underlying `causal conditional directed information rate' being equal to zero. The plug-in estimator for this functional is identified with the log-likelihood ratio test statistic for the desired test. This statistic is shown that is asymptotically normal under the alternative hypothesis and asymptotically $\chi^2$ distributed under the null, facilitating the computation of $p$-values when used on empirical data. The effectiveness of the resulting hypothesis test is illustrated on simulated data, validating the underlying theory. The test is also employed in the analysis of spike train data recorded from neurons in the V4 and FEF brain regions of behaving animals during a visual attention task. There, the test results are seen to identify interesting and biologically relevant information.

Abstract: We develop new methodology to improve our understanding of the causal effects of multivariate air pollution exposures on public health. Typically, exposure to air pollution for an individual is measured at their home geographic region, though people travel to different regions with potentially different levels of air pollution. To account for this, we incorporate estimates of the mobility of individuals from cell phone mobility data to get an improved estimate of their exposure to air pollution. We treat this as an interference problem, where individuals in one geographic region can be affected by exposures in other regions due to mobility into those areas. We propose policy-relevant estimands and derive expressions showing the extent of bias one would obtain by ignoring this mobility. We additionally highlight the benefits of the proposed interference framework relative to a measurement error framework for accounting for mobility. We develop novel estimation strategies to estimate causal effects that account for this spatial spillover utilizing flexible Bayesian methodology. Empirically we find that this leads to improved estimation of the causal effects of air pollution exposures over analyses that ignore spatial spillover caused by mobility.

Abstract: Propensity score matching (PSM) and augmented inverse propensity weighting (AIPW) are widely used in observational studies to estimate causal effects. The two approaches present complementary features. The AIPW estimator is doubly robust and locally efficient but can be unstable when the propensity scores are close to zero or one due to weighting by the inverse of the propensity score. On the other hand, PSM circumvents the instability of propensity score weighting but it hinges on the correctness of the propensity score model and cannot attain the semiparametric efficiency bound. Besides, the fixed number of matches, K, renders PSM nonsmooth and thus invalidates standard nonparametric bootstrap inference. This article presents novel augmented match weighted (AMW) estimators that combine the advantages of matching and weighting estimators. AMW adheres to the form of AIPW for its double robustness and local efficiency but it mitigates the instability due to weighting. We replace inverse propensity weights with matching weights resulting from PSM with unfixed K. Meanwhile, we propose a new cross-validation procedure to select K that minimizes the mean squared error anchored around an unbiased estimator of the causal estimand. Besides, we derive the limiting distribution for the AMW estimators showing that they enjoy the double robustness property and can achieve the semiparametric efficiency bound if both nuisance models are correct. As a byproduct of unfixed K which smooths the AMW estimators, nonparametric bootstrap can be adopted for variance estimation and inference. Furthermore, simulation studies and real data applications support that the AMW estimators are stable with extreme propensity scores and their variances can be obtained by naive bootstrap.

Abstract: Amortized variational inference produces a posterior approximator that can compute a posterior approximation given any new observation. Unfortunately, there are few guarantees about the quality of these approximate posteriors. We propose Conformalized Amortized Neural Variational Inference (CANVI), a procedure that is scalable, easily implemented, and provides guaranteed marginal coverage. Given a collection of candidate amortized posterior approximators, CANVI constructs conformalized predictors based on each candidate, compares the predictors using a metric known as predictive efficiency, and returns the most efficient predictor. CANVI ensures that the resulting predictor constructs regions that contain the truth with high probability (exactly how high is prespecified by the user). CANVI is agnostic to design decisions in formulating the candidate approximators and only requires access to samples from the forward model, permitting its use in likelihood-free settings. We prove lower bounds on the predictive efficiency of the regions produced by CANVI and explore how the quality of a posterior approximation relates to the predictive efficiency of prediction regions based on that approximation. Finally, we demonstrate the accurate calibration and high predictive efficiency of CANVI on a suite of simulation-based inference benchmark tasks and an important scientific task: analyzing galaxy emission spectra.