Experimental and observational studies often lack validity due to untestable assumptions. We propose a double machine learning approach to combine experimental and observational studies, allowing practitioners to test for assumption violations and estimate treatment effects consistently. Our framework proposes a falsification test for external validity and ignorability under milder assumptions. We provide consistent treatment effect estimators even when one of the assumptions is violated. However, our no-free-lunch theorem highlights the necessity of accurately identifying the violated assumption for consistent treatment effect estimation. Through comparative analyses, we show our framework's superiority over existing data fusion methods. The practical utility of our approach is further exemplified by three real-world case studies, underscoring its potential for widespread application in empirical research.

dame-flame is a Python package for performing matching for observational causal inference on datasets containing discrete covariates. This package implements the dynamic almost matching exactly (DAME) and fast, large-scale almost matching exactly (FLAME) algorithms, which match treatment and control units on subsets of the covariates. The resulting matched groups are interpretable, because the matches are made directly on covariates, and high-quality, because machine learning is used to determine which covariates are important to match on instead of human inputs. The package provides several adjustable parameters to adapt the algorithms to specific applications, and can calculate treatment effects after matching. The most recent source code of the implementation is available at https://github.com/almost-matching-exactly/DAME-FLAME-Python-Package.

Even if a model is not globally sparse, it is possible for decisions made from that model to be accurately and faithfully described by a small number of features. For instance, an application for a large loan might be denied to someone because they have no credit history, which overwhelms any evidence towards their creditworthiness. In this work, we introduce the Sparse Explanation Value (SEV), a new way of measuring sparsity in machine learning models. In the loan denial example above, the SEV is 1 because only one factor is needed to explain why the loan was denied. SEV is a measure of decision sparsity rather than overall model sparsity, and we are able to show that many machine learning models - even if they are not sparse - actually have low decision sparsity, as measured by SEV. SEV is defined using movements over a hypercube, allowing SEV to be defined consistently over various model classes, with movement restrictions reflecting real-world constraints. We propose algorithms that reduce SEV without sacrificing accuracy, providing sparse and completely faithful explanations, even without globally sparse models.

Flexibly modeling how an entire density changes with covariates is an important but challenging generalization of mean and quantile regression. While existing methods for density regression primarily consist of covariate-dependent discrete mixture models, we consider a continuous latent variable model in general covariate spaces, which we call DR-BART. The prior mapping the latent variable to the observed data is constructed via a novel application of Bayesian Additive Regression Trees (BART). We prove that the posterior induced by our model concentrates quickly around true generative functions that are sufficiently smooth. We also analyze the performance of DR-BART on a set of challenging simulated examples, where it outperforms various other methods for Bayesian density regression. Lastly, we apply DR-BART to two real datasets from educational testing and economics, to study student growth and predict returns to education. Our proposed sampler is efficient and allows one to take advantage of BART's flexibility in many applied settings where the entire distribution of the response is of primary interest. Furthermore, our scheme for splitting on latent variables within BART facilitates its future application to other classes of models that can be described via latent variables, such as those involving hierarchical or time series data.

dame-flame is a Python package for performing matching for observational causal inference on datasets containing discrete covariates. This package implements the Dynamic Almost Matching Exactly (DAME) and Fast Large-Scale Almost Matching Exactly (FLAME) algorithms, which match treatment and control units on subsets of the covariates. The resulting matched groups are interpretable, because the matches are made on covariates, and high-quality, because machine learning is used to determine which covariates are important to match on. DAME solves an optimization problem that matches units on as many covariates as possible, prioritizing matches on important covariates. FLAME approximates the solution found by DAME via a much faster backward feature selection procedure. The package provides several adjustable parameters to adapt the algorithms to specific applications, and can calculate treatment effect estimates after matching. Descriptions of these parameters, details on estimating treatment effects, and further examples, can be found in the documentation at https://almost-matching-exactly.github.io/DAME-FLAME-Python-Package/

BACKGROUND:ABO-incompatible heart transplant is a method to increase the infant donor pool. However data on long-term survival and rejection after ABO-incompatible heart transplant in recent era are limited. METHODS:The United Network for Organ Sharing database was queried for infant heart transplants performed from January 2008 to March 2020. Patient demographics and known risk factors for posttransplant mortality were collected. Statistical analysis using Bayesian additive regression trees was performed to evaluate the association of ABO incompatibility and overall survival, graft survival, acute rejection episodes, and length of stay. RESULTS:Of 1368 included infants (age < 1 year), 280 (20.47%) were ABO incompatible. ABO incompatibility was not associated with increased all-cause mortality, acute rejection episodes, or length of stay, whereas extracorporeal membrane oxygenation and intubation status of the recipient at the time of transplantation were associated with increased all-cause mortality and graft failure. Idiopathic cardiomyopathy was associated with a decreased likelihood of posttransplant all-cause mortality. One-, 5-, and 10-year survival rates among compatible vs incompatible transplants were estimated to be 90% vs 88%, 82% vs 79%, and 77% vs 73%, respectively. CONCLUSIONS:ABO-incompatible infant heart transplant does not affect posttransplant survival, incidence of rejection, or postoperative length of stay. Therefore it remains a viable and important strategy to increase the infant donor pool.

AME-GUI (Almost Matching Exactly Graphical User Interface) is an interactive web-based application that allows users to perform matching for causal inference on large, complex datasets with categorical covariates. The application is powered by the Fast Large-Scale Almost Matching Exactly (FLAME) algorithm (Wang et al., 2021), which matches treatment and control units in a way that is (i) interpretable, because the matches are made directly on covariates, (ii) high-quality, because machine learning is used to determine which covariates are important to match on, and (iii) scalable, using techniques from data management. The graphical user interface highlights the utility of this algorithm and uses a suite of visualization tools to facilitate easy and interactive exploration of treatment effect estimates, as well as of the created matched groups that they depend on. The application gives a quick and simple overview of the open-source Python and R packages dame-flame and FLAME, and the range of functionality they provide for interpretable and efficient causal inference.

We propose a matching method that recovers direct treatment effects from randomized experiments where units are connected in an observed network, and units that share edges can potentially influence each others' outcomes. Traditional treatment effect estimators for randomized experiments are biased and error prone in this setting. Our method matches units almost exactly on counts of unique subgraphs within their neighborhood graphs. The matches that we construct are interpretable and high-quality. Our method can be extended easily to accommodate additional unit-level covariate information. We show empirically that our method performs better than other existing methodologies for this problem, while producing meaningful, interpretable results.

We propose a matching method for observational data that matches units with others in unit-specific, hyper-box-shaped regions of the covariate space. These regions are large enough that many matches are created for each unit and small enough that the treatment effect is roughly constant throughout. The regions are found as either the solution to a mixed integer program, or using a (fast) approximation algorithm. The result is an interpretable and tailored estimate of a causal effect for each unit.