The Finite Selection Model (FSM) was developed by Carl Morris in the 1970s for the design of the RAND Health Insurance Experiment (HIE), one of the largest and most comprehensive social science experiments conducted in the U.S. The idea behind the FSM is that each treatment group takes turns selecting units in a fair and random order to optimize a common assignment criterion. At each of its turns, a treatment group selects the available unit that maximally improves the combined quality of its resulting group of units in terms of the criterion. In the HIE and beyond, we revisit, formalize, and extend the FSM as a general tool for experimental design. Leveraging the idea of D-optimality, we propose and analyze a new selection criterion in the FSM. The FSM using the D-optimal selection function has no tuning parameters for covariate balance, is affine invariant, and when appropriate, retrieves several classical designs such as randomized block and matched-pair designs. For multi-arm experiments, we propose algorithms to generate a fair and random selection order of treatments. We demonstrate FSM's performance in a case study based on the HIE and in ten randomized studies from the health and social sciences. On average, the FSM achieves 68% better covariate balance than complete randomization and 56% better covariate balance than rerandomization in a typical study. We recommend the FSM be considered in experimental design for its conceptual simplicity, efficiency, and robustness.
Accurate estimates of migrant population stocks by age and sex are essential for demographic analysis and policy. This study develops a hierarchical Bayesian model that integrates survey data from the Labour Force Survey with digital trace data from the Facebook Advertising Platform to improve estimation of migrant age-sex distributions. The approach applies reduced Rogers-Castro age schedules as first-stage smoothing and harmonizes profiles within a multinomial-Dirichlet-Dirichlet framework. We illustrate the model using data on the 10 largest European migrant groups in the United Kingdom in 2018-2019. Results reveal 3 broad patterns: younger age structures among Western and Southern European migrants, slightly older working-age profiles for Central and Eastern Europeans, and a predominantly older Irish migrant population. External validation against Census 2021 country-of-birth data for England and Wales shows that the proposed model more closely matches benchmark age-sex distributions than a specification inspired by the area model of Wi & sacute;niowski et al. (2016. Integrated modelling of age and sex patterns of European migration. Journal of the Royal Statistical Society: Series A (Statistics in Society), 179(4), 1007-1024. https://doi.org/10.1111/rssa.12177). Facebook data improve representation of younger migrants, who are often underrepresented in surveys, while the Labour Force Survey provides broader demographic structure. Overall, the findings demonstrate the value of combining traditional and digital data to address gaps in migration statistics and produce timely, detailed population estimates.
This study examines the effects of gender-marked bilateral development aid on women’s legal rights and economic participation in aid-eligible countries from 1990 to 2021. Using a panel dataset that combines gender-marked aid flows with data from the World Bank’s Women, Business and the Law database, we employ two complementary empirical strategies—a panel fixed effects model with unit-specific trends and a shift-share instrumental variable approach. Our results indicate that gender-marked aid contributes to legal reforms that expand women’s economic rights. We also find evidence of a positive link to women’s employment in the industrial sector, while finding no comparable effects in the services or agricultural sectors. Overall, the findings suggest that targeted development aid can play a significant role in advancing gender-equal legal frameworks and promoting women’s economic participation.
It is common for tennis players to set up a winning shot via a calculated previous shot. To study this phenomenon we employ data from five Grand Slam matches that include Roger Federer, Rafael Nadal, Novak Djokovic, and Juan del Potro on the three major surfaces (hard court, clay, and grass) to infer shot location based on a tennis court that has been divided into six zones in a novel way. Exploratory analysis alone highlights player-specific shot location preferences based on receipt location before a given shot. However, we propose a Bayesian hierarchical model and testing scheme that allows for greater understanding of player shot location preference through posterior distributions on each shot location probability estimate. Specifically, we use a Multinomial-Dirichlet model to estimate expected shot locations along with corresponding uncertainty quantification. We also utilize simultaneous posterior probability estimation methods to perform sequential testing and identify differences in player behaviour over each surface, receipt location, and shot location.
This paper investigates the feasibility of using earlier provisional data to improve the now- and forecasting accuracy of final and official statistics. We propose the use of a multivariate structural time series model, which includes common trends and seasonal components to combine official statistics series with related auxiliary series. In this way, more precise and more timely nowcasts and forecasts of official statistics can be obtained by exploiting the higher frequency and/or the more timely availability of the auxiliary series. The proposed method can be applied to different data sources consisting of any number of missing observations both at the beginning and at the end of the series simultaneously. Two empirical applications are presented. The first one focuses on fatal traffic accidents and the second one on labour force participation at the municipal level. The results demonstrate the effectiveness of our proposed approach in improving forecasting performance for the target series and providing early warnings to policy-makers.
