Examining Heterogeneity in the Impacts of Socio-Emotional Curricula in Preschool: A Quantile Treatment Effect Approach

We examine treatment effect heterogeneity using data from the Head Start CARES study, in which a sample of preschool centers was randomly assigned to either one of three curricula interventions targeting socio-emotional (SE) skills (i.e., emotional knowledge, problem-solving skills, and executive functions) or to continue using their “business-as-usual” curriculum. Most existing research estimates only mean differences between treatment and control groups, and uses simple subgroup analyses to assess treatment heterogeneity, which may overlook important variation in treatment effects across the ex post outcome distribution. We use quantile treatment effects analyses to understand the impacts of these curricular interventions at various parts of the outcome distribution, from the 1st percentile to the 99th percentile, to understand who benefits most from SE curricula interventions. Results show positive impacts of the curricula interventions on emotional knowledge and problem-solving skills, but not equally across the full skill distribution. Children in the upper half of the emotional knowledge distribution and at the higher end of the problem-solving skills distribution gain more from the curricula. As in the study’s original mean-comparison analyses, we find no impacts on children’s executive function skills at any point in the skills distribution. Our findings add to the growing literature on the differential effects of curricula interventions for preschool programs operating at scale. Importantly, it provides the first evidence for the effects of SE curricula interventions on SE outcomes across children’s outcome skill levels. We discuss implications for early education programs for children with different school readiness skills.

Debiased/Double Machine Learning for Instrumental Variable Quantile Regressions

In this study, we investigate the estimation and inference on a low-dimensional causal parameter in the presence of high-dimensional controls in an instrumental variable quantile regression. Our proposed econometric procedure builds on the Neyman-type orthogonal moment conditions of a previous study (Chernozhukov et al. 2018) and is thus relatively insensitive to the estimation of the nuisance parameters. The Monte Carlo experiments show that the estimator copes well with high-dimensional controls. We also apply the procedure to empirically reinvestigate the quantile treatment effect of 401(k) participation on accumulated wealth.

Vitamin C may increase the recovery rate of outpatient cases of SARS-CoV-2 infection by 70%: reanalysis of the COVID A to Z Randomized Clinical Trial

The COVID A to Z trial (JAMA Netw Open. 2021;4:e210369) is important as it focused specifically on SARS-CoV-2 coronavirus patients and examined a high dose of vitamin C which was previously predicted to reduce the duration of respiratory virus infections by about 20%. Unfortunately, there are several limitations in the trial methods. The COVID A to Z trial was “stopped early for futility”. In the sample size calculation, the authors assumed a 1.0 day shorter symptoms by intervention. Duration of symptoms was reduced by 1.2 days in the vitamin C arm compared with the usual care arm. Given that the observed vitamin C effect was 20% greater than the expected effect (1.2 vs. 1.0), it is illogical to have stopped the trial early because of “futility”. In this reanalysis we calculated the rate ratio of recovery between the vitamin C and usual care arms and found that vitamin C increased the rate of recovery by 70% (95% CI 6.8% to 170%, P = 0.025). Furthermore, we calculated quantile treatment effect of vitamin C. At the 60th percentile level of symptom distribution, duration was 9 days in the usual care arm, and 6 days in the vitamin C arm, which corresponds to reduction in symptom duration by 3 days (95% CI 3 to 4.6 days; P < 0.001). The analysis of the quantile treatment effect indicates that there may be around 30% reduction in symptom duration in patients with the longest symptoms. Our reanalysis indicates a need for methodologically sound trials with larger numbers of patients to investigate the treatment effects of vitamin C against SARS-CoV-2.

Moving Beyond Linear Regression: Implementing and Interpreting Quantile Regression Models with Fixed Effects

Quantile regression (QR) provides an alternative to linear regression (LR) that allows for the estimation of relationships across the distribution of an outcome. However, as highlighted in recent research on the motherhood penalty across the wage distribution, different procedures for conditional and unconditional quantile regression (CQR, UQR) often result in divergent findings that are not always well understood. In light of such discrepancies, this paper reviews how to implement and interpret a range of LR, CQR, and UQR models with fixed effects. It also discusses the use of Quantile Treatment Effect (QTE) models as an alternative to overcome some of the limitations of CQR and UQR models. We then review how to interpret results in the presence of fixed effects based on a replication of Budig and Hodges's (2010) work on the motherhood penalty using NLSY79 data.

Assessing Impact of BRRI Released Modern Rice Varieties Adoption on Farmers’ Welfare in Bangladesh: Application of Panel Treatment Effect Model

This study assesses the impact of Bangladesh Rice Research Institute (BRRI) released modern wet (Aman) season rice variety adoption on farmers‟ well-being in Bangladesh. Bangladesh Integrated Household Survey (BIHS) data collected by IFPRI were used for this study. The study applied difference-in-difference treatment effect and difference-in-difference quantile treatment effect models using unbalanced panel data to achieve the set objectives. Analysis revealed that BRRI released wet (Aman) season rice technology has a robust and positive effect on small farmers‟ welfare in Bangladesh as indicated by the level of increases in per capita household real income, increases in real aman rice income, and also increases in yield and decreases both in poverty gap and squared poverty gap over time. The marginal and near landless farmers have not gained significantly through adopting BRRI released modern variety over non-adopters in terms of all the indicators except aman rice yield. However, only yield of BRRI released modern wet (Aman) season rice technology has positive and significant impact on the marginal and near landless farmers. As such, BRRI variety adoption seemed to be conducive in increasing the level of yield of marginal and near-landless farms but it hardly helps them to overcome the poverty level, unless other equity-enhancing policy measures are undertaken. Overall, there was large scope for enhancing adoption of BRRI released rice variety in order to reduce the poverty level in rural areas. The current rice policy (rice self-sufficiency) appears to be supportive to help Bangladesh rice sector for achieving food security in the country. Bangladesh Rice j. 2019, 23(1): 1-11

INSTRUMENTAL VARIABLE QUANTILE REGRESSION WITH MISCLASSIFICATION

This article investigates the instrumental variable quantile regression model (Chernozhukov and Hansen, 2005, Econometrica 73, 245–261; 2013, Annual Review of Economics, 5, 57–81) with a binary endogenous treatment. It offers two identification results when the treatment status is not directly observed. The first result is that, remarkably, the reduced-form quantile regression of the outcome variable on the instrumental variable provides a lower bound on the structural quantile treatment effect under the stochastic monotonicity condition. This result is relevant, not only when the treatment variable is subject to misclassification, but also when any measurement of the treatment variable is not available. The second result is for the structural quantile function when the treatment status is measured with error; the sharp identified set is characterized by a set of moment conditions under widely used assumptions on the measurement error. Furthermore, an inference method is provided in the presence of other covariates.

Quantile treatment effects and bootstrap inference under covariate‐adaptive randomization

In this paper, we study the estimation and inference of the quantile treatment effect under covariate‐adaptive randomization. We propose two estimation methods: (1) the simple quantile regression and (2) the inverse propensity score weighted quantile regression. For the two estimators, we derive their asymptotic distributions uniformly over a compact set of quantile indexes, and show that, when the treatment assignment rule does not achieve strong balance, the inverse propensity score weighted estimator has a smaller asymptotic variance than the simple quantile regression estimator. For the inference of method (1), we show that the Wald test using a weighted bootstrap standard error underrejects. But for method (2), its asymptotic size equals the nominal level. We also show that, for both methods, the asymptotic size of the Wald test using a covariate‐adaptive bootstrap standard error equals the nominal level. We illustrate the finite sample performance of the new estimation and inference methods using both simulated and real datasets.