Abstract
The 3-day lecture series will discuss fundamental principles and contemporary trends in statistical inference for finite populations. With some cases presented, participants will gain practical insights into real-world applications of sampling methodologies and their implications. With the discussion of current trends, this equips researchers with the knowledge and tools necessary to conduct inference in finite populations, ensuring the generation of valuable insights for informed decision-making in various disciplines.
Day 1: Jun 14 (2PM - 4PM), Basic Concepts in Survey Sampling. Sampling in finite populations is faced with challenges of balancing accuracy and efficiency in selecting representative samples. The pivotal role survey data plays in shaping official statistics, underscoring its influence on policy-making and decision processes is presented. This will also cover design-unbiased estimation principles, Horvitz-Thompson estimator, weight adjustment techniques, and variance estimation strategies for more reliable characterization of diverse population segments.
Day 2: Jun 15 (2PM - 4PM), Construction of Sampling Designs. Practical methodologies useful in the development of sampling designs will be presented. Monte Carlo Simulation is discussed to bridge the gap between the sampling frame and the actual target population in the assessment of efficiency of the proposed sampling design. The role of bootstrap methods in enhancing estimates (including estimation of their variance) will be discussed. Some cases will be to illustrate various methods are included as well.
Day 3: Jun 16 (2PM - 4PM), Analysis of Survey Data. Survey data are often integrated into data from diverse sources like big data and administrative reports. Methodologies for harmonizing disparate data streams to estimate some indicators for Sustainable Development Goals (SDGs), navigating statistical challenges and some solutions are discussed. Model-based estimation techniques, partial least squares-structural equation modeling (PLS-SEM), and other strategies in analysing survey data will be presented, integrating survey weights from data generated using complex survey designs. Through case studies, participants will learn practical implications of integrating multiple data sources, ensuring comprehensive insights into complex societal phenomena covered in the survey.
Brief Biography
Former professor at School of Statistics, University of the Philippines Diliman. Served as visiting professor at Karlstad University (Sweden) and as visiting researcher at Asian Development Bank Institute (Japan). His research interest focused on computational statistics, computational econometrics and nonparametric methods with applications in spatiotemporal modeling, time series models, big data analytics, financial markets, among others. Currently an associate editor of Communications in Statistical Applications and Methods (Korean Statistical Society and Korean International Statistical Society). An elected member of the International Statistical Insitute (2012), full member of Philippine American Academy of Science and Engineering. Also a member of: Econometric Society, East Asian Economic Association, and Philippine Statistical Association, Inc. Was elected into the board of directors of the International Association of Statistical Computing (2022-2025).