Jennifer Lynn Hill

New York University | NYU · Department of Applied Statistics Social Science and Humanities

This analysis estimates the impact on basic reading assessments for Syrian refugee children of attending a non-formal remedial support program that was infused with SEL practices. Both the treatment effects of high versus low attendance and the average dosage-response functions for continuous attendance were estimated using BART for causal inferenc...

Visual diagnosis of radiographs, histology and electrocardiograms lends itself to deliberate practice, facilitated by large online banks of cases. Which cases to supply to which learners in which order is still to be worked out, with there being considerable potential for adapting the learning. Advances in statistical modeling, based on an accumula...

Bayesian additive regression trees (BART) provides a flexible approach to fitting a variety of regression models while avoiding strong parametric assumptions. The sum-of-trees model is embedded in a Bayesian inferential framework to support uncertainty quantification and provide a principled approach to regularization through prior specification. T...

Response to discussion of Dorie (2017), in which the authors of that piece express their gratitude to the discussants, rebut some specific criticisms, and argue that the limitations of the 2016 Atlantic Causal Inference Competition represent an exciting opportunity for future competitions in a similar mold.

Available here: http://www.tandfonline.com/eprint/ygfsdhqPzcSD7j9v9pxm/full Maternal education is one of the strongest predictors of children's academic outcomes. One possible explanation for this is that more highly educated mothers more frequently engage in parenting practices that may promote children's later cognitive development; however, mo...

This study evaluates the effect of attending a U.S. public middle or junior high school as compared with a K-8 school on eighth graders’ academic and psychosocial outcomes. In a national sample, we conducted propensity score weighted regression analysis. Initial findings indicated that for eighth-grade students, attending a middle or junior high sc...

Statisticians have made great strides towards assumption-free estimation of causal estimands in the past few decades. However this explosion in research has resulted in a breadth of inferential strategies that both create opportunities for more reliable inference as well as complicate the choices that an applied researcher has to make and defend. R...

Supporting info item

In the analysis of causal effects in non-experimental studies, conditioning on observable covariates is one way to try to reduce unobserved confounder bias. However, a developing literature has shown that conditioning on certain covariates may increase bias, and the mechanisms underlying this phenomenon have not been fully explored. We add to the l...

When estimating causal effects, unmeasured confounding and model misspecification are both potential sources of bias. We propose a method to simultaneously address both issues in the form of a semi-parametric sensitivity analysis. In particular, our approach incorporates Bayesian Additive Regression Trees into a two-parameter sensitivity analysis s...

A major obstacle to developing evidenced-based policy is the difficulty of implementing randomized experiments to answer all causal questions of interest. When using a nonexperimental study, it is critical to assess how much the results could be affected by unmeasured confounding. We present a set of graphical and numeric tools to explore the sensi...

Randomized experiments are considered the gold standard for causal inference because they can provide unbiased estimates of treatment effects for the experimental participants. However, researchers and policymakers are often interested in using a specific experiment to inform decisions about other target populations. In education research, increasi...

This article discusses causal inference in statistics. It describes the theoretical framework and notation needed to formally define causal effects and the assumptions required to identify them nonparametrically. This involves definition of potential outcomes that represent the potential value of the outcome across different treatment exposures. De...

Social-Emotional Learning (SEL) programs aim to improve students' social-emotional competencies in order to enhance their achievement. Although SEL programs typically implement classroom curricula, some programs also include a component for parents. Yet, little is known about the types of parents likely to participate in services, and whether paren...

R package for Data Analysis using multilevel/hierarchical model

We consider the relative performance of two common approaches to multiple imputation (MI): joint multivariate normal (MVN) MI, in which the data are modeled as a sample from a joint MVN distribution; and conditional MI, in which each variable is modeled conditionally on all the others. In order to use the multivariate normal distribution, implement...

Introduction: School transitions are critical periods for intervention to promote youth achievement and reduce maladjustment. Although school transitions in 6th or 7th grade occur at the same time as significant, often disruptive, developmental shifts, this period receives less attention in prevention and policy than earlier or later transitions. T...

Causal inference in observational studies typically requires making com-parisons between groups that are dissimilar. For instance, researchers inves-tigating the role of a prolonged duration of breastfeeding on child outcomes may be forced to make comparisons between women with substantially dif-ferent characteristics on average. In the extreme the...

Participant attrition may be a significant threat to the generalizability of the results of educational research studies if participants who do not persist in a study differ from those who do in ways that can affect the experimental outcomes. A multi-center trial of the efficacy of different computer-based instructional strategies gave us the oppor...

We compared differences in the hospital charges, length of hospital stay, and mortality between patients with healthcare- and community-associated bloodstream infections, urinary tract infections, and pneumonia due to antimicrobial-resistant versus -susceptible bacterial strains. A retrospective analysis of an electronic database compiled from labo...

Our mi package in R has several features that allow the user to get inside the imputation process and evaluate the reasonableness of the resulting models and imputations. These features include: choice of predictors, models, and transformations for chained imputation models; standard and binned residual plots for checking the fit of the conditional...

Dual threats of injection drug use and risky sexual practices continue to increase transmission of HIV and other sexually transmitted Infections (STIs) among drug-using couples in low-income communities in the United States. Two hypotheses were tested: (1) "intervention effect"-whether the HIV risk-reduction intervention provided to the couple or i...

This article explores some of the challenges that arise when trying to implement propensity score strategies to answer a causal question using data with a large number of covariates. We discuss choices in propensity score estimation strategies, matching and weighting implementation strategies, balance diagnostics, and final analysis models. We demo...

Researchers have long struggled to identify causal effects in nonexperimental settings. Many recently proposed strategies assume ignorability of the treatment assignment mechanism and require fitting two models—one for the assignment mechanism and one for the response surface. This article proposes a strategy that instead focuses on very flexibly m...

Iterative imputation, in which variables are imputed one at a time each given a model predicting from all the others, is a popular technique that can be convenient and flexible, as it replaces a potentially difficult multivariate modeling problem with relatively simple univariate regressions. In this paper, we begin to characterize the stationary d...

Comment on "The Essential Role of Pair Matching in Cluster-Randomized Experiments, with Application to the Mexican Universal Health Insurance Evaluation" [arXiv:0910.3752] Comment: Published in at http://dx.doi.org/10.1214/09-STS274A the Statistical Science (http://www.imstat.org/sts/) by the Institute of Mathematical Statistics (http://www.imstat....

Conditional cash transfer programmes have expanded in developing countries as a way to foster human capital accumulation. Despite evidence of these programmes' positive impact on school enrolment, little is known about their impact on school achievement. This study estimated the effect of Familias en Accion on school achievement. It found that the...

Applied researchers often find themselves making statistical inferences in settings that would seem to require multiple comparisons adjustments. We challenge the Type I error paradigm that underlies these corrections. Moreover we posit that the problem of multiple comparisons can disappear entirely when viewed from a hierarchical Bayesian perspecti...

The hypothesis that marriage increases men's earnings has contributed to legislative support for the Healthy Marriage Initiative (HMI). However, previous studies of this phenomenon have not controlled for many relevant characteristics that select men into marriage, nor have they focused on low-income, unmarried fathers-the population targeted by HM...

Data Analysis Using Regression and Multilevel/Hierarchical Models, first published in 2007, is a comprehensive manual for the applied researcher who wants to perform data analysis using linear and nonlinear regression and multilevel models. The book introduces a wide variety of models, whilst at the same time instructing the reader in how to fit th...

The female condom remains the only female-initiated method for preventing pregnancy and STDs, including HIV. Innovative methods for promoting its use, and for involving male partners in its use, are needed. A sample of 217 women and their main male sexual partners were randomly assigned to one of three study conditions: a six-session relationship-b...

In causal studies without random assignment of treatment, causal effects can be estimated using matched treated and control samples, where matches are obtained using estimated propensity scores. Propensity score matching can reduce bias in treatment effect estimators in cases where the matched samples have overlapping covariate distributions. Despi...

The employment rate for mothers with young children has increased dramatically over the past 25 years. Estimating the effects of maternal employment on children's development is challenged by selection bias and the missing data endemic to most policy research. To address these issues, this study uses propensity score matching and multiple imputatio...

Experimental sampleConstructed observational studyConcluding remarks

Intimate partner violence (IPV) has been recognized as a risk factor for HIV and sexually transmitted infections (STIs) among women, particularly among those who are drug involved. This study examines the temporal relationships between sexual and/or physical partner violence (IPV) and sexual risk of HIV/STI transmission in a longitudinal study with...

This randomized clinical trial examined the relative efficacy of a relationship-based HIV/STI prevention intervention for women and their regular male sexual partners at 12 months post-intervention. A total of 217 couples were randomized to (1) a six-session intervention provided to the woman and her sexual partner together (n=81); (2) the same int...

We examined whether frequent drug use increases the likelihood of subsequent sexual or physical intimate partner violence (IPV) and whether IPV increases the likelihood of subsequent frequent drug use. A random sample of 416 women on methadone was assessed at baseline (wave 1) and at 6 months (wave 2), and 12 months (wave 3) following the initial a...

This paper uses data from the National Longitudinal Survey of Youth to explore links between mothers' returns to work within 12 weeks of giving birth and health and developmental outcomes for their children. OLS models and propensity score matching methods are utilised to account for selection bias. Considerable associations between early returns t...

Although several studies have examined the relationship between intimate partner violence (IPV) and drug use among women in drug treatment programs, more information is needed to delineate differences, as a function of the specific drug used. Data from a random sample of 416 women attending methadone programs were analyzed to elucidate the differen...

Effects of high participation in the Infant Health and Development Program (IHDP), an 8-site randomized trial that targeted low-birth-weight (LBW) premature infants (N=1,082), were estimated. Children in the treatment group were offered high-quality center-based care in their 2nd and 3rd years of life (full-day care, 50 weeks per year). High-dosage...

This study examined the efficacy of a relationship-based HIV/sexually transmitted disease prevention program for heterosexual couples and whether it is more effective when delivered to the couple or to the woman alone. Couples (n = 217) were recruited and randomized to (1) 6 sessions provided to couples together (n = 81), (2) the same intervention...

The precarious state of the educational system in the inner cities of the United States, as well as its potential causes and solutions, have been popular topics of debate in recent years. Part of the difficulty in resolving this debate is the lack of solid empirical evidence regarding the true impact of educational initiatives. The efficacy of so-c...

This paper provides a survey on studies that analyze the macroeconomic effects of intellectual property rights (IPR). The first part of this paper introduces different patent policy instruments and reviews their effects on R&D and economic growth. This part also discusses the distortionary effects and distributional consequences of IPR protection a...

In policy research a frequent aim is to estimate treatment effects separately by subgroups. This endeavor becomes a methodological challenge when the subgroups are defined by post-treatment, rather than pre-treatment, variables because if analyses are performed in the same way as with pre-treatment variables, causal interpretations are no longer va...

In one of the most influential works in the public opinion literature, Philip Converse proposed a "black-and-white" model that divided respondents into two groups: opinion holders and unstable opinion changers. We extend the model by allowing for a group that makes rational opinion changes over time. This enables us to (I) explore e hypotheses abou...

Popular theories in political science regarding opinion-changing behavior postulate the existence of one or both of two broad categories of people: those with stable opinions over time; and those who appear to hold no solid opinion and, when asked to make a choice, do so seemingly at random. The model presented here explores evidence for a third ca...

Popular theories in political science regarding opinion-changing behavior postulate existence of one or both of two broad categories of people: those who hold their opinions over time; and those that hold no solid opinion and, when asked to make a choice, do so seemingly at random. This study explores evidence for a third category: durable changers...

The precarious state of the educational system existing in the inner-cities of the U.S., including its potential causes and solutions, has been a popular topic of debate in recent years. Part of the difficulty in resolving this debate is the lack of solid empirical evidence regarding the true impact of educational initiatives. For example, educatio...

We illustrate the use of a class of statistical models, finite mixture models, that can be used to allow for differences in model parameterizations across groups, even in the ab- sence of group labels. We also introduce a methodology for fitting these models, data augmentation. Neither finite mixture models nor data augmentation is routine in the w...

INTRODUCTION One of Don Campbell's many influential contributions was to the design of studies to estimate causal effects (e.g., Campbell & Stanley, 1966). He had particular interest in the trade-offs between matching and covariance adjustments (e.g., Campbell & Erlebacher, 1970; Cook & Campbell, 1979). One of the authors (Rubin), in fact, had his...

Randomized experiments suffering from missing data and noncompliance are 4 recurring problem for experimenters whose subjects are human. Until recently, analysts of such broken randomized experiments were largely forced to squeeze the data into the idealized template of the randomized experiment with neither noncompliance missing data Such practice...