Applied Statistical Methods and Causal Analysis 应用统计方法与因果分析 (MAV553)

MAV553 Applied Statistical Methods and Causal Analysis introduces the concept of statistical inference, empirical methods and causal analyses that are hands-on and practitioner focused. The course will provide an understanding of the fundamental principles of various empirical methods for statistical inference and causal analysis. The course will begin by covering basic concepts in statistics and programming. It will focus on common confounding challenges of causal analysis, and how classical empirical approaches such as linear regression and panel data regression may address them. It will also cover modern approaches for causal analysis such as instrumental variable estimation and difference-in-differences estimation. The course will focus on equipping students with the sound intuition and practical research skills in conducting statistical and causal analysis to address relevant business problems. MAV553应用统计方法与因果分析介绍了统计推断、实证方法和因果分析的概念，注重实践操作，面向从业者。课程将帮助学生理解用于统计推断和因果分析的各种实证方法的基本原理。课程将从统计和编程的基本概念开始，重点关注因果分析中常见的混杂挑战，以及如何通过线性回归和面板数据回归等经典实证方法加以应对。课程还将涵盖现代因果分析方法，如工具变量估计和双重差分估计。该课程旨在培养学生在统计和因果分析方面的直观理解和实际研究技能，以解决相关的业务问题。

• Fundamental statistical concepts 基本统计概念
• Programming for statistics 统计编程
• Hypothesis testing and statistical inference 假设检验与统计推断
• Common issues in empirical design 实证设计中的常见问题
• Randomized controlled trials and quasi-natural experiments 随机对照试验与准自然实验
• Linear regression concepts 线性回归概念
• Linear regression model design 线性回归模型设计
• Panel data regression concepts 面板数据回归概念
• Panel data regression results and interpretation 面板数据回归结果与解释
• Regressions with dummy variables 虚拟变量回归
• Difference-in-differences estimation 双重差分估计
• Instrumental variable regression 工具变量回归

• Construct testable hypotheses from available data 从可用数据中构建可检验的假设
• Test various hypotheses with appropriate statistical tests 使用适当的统计测试检验各种假设
• Evaluate the suitability of various empirical approaches for different business problems 评估各种实证方法在不同业务问题中的适用性
• Assess the advantages and pitfalls of the various empirical approaches 评估各种实证方法的优势和局限性
• Design experiments, research to understand the relationship between variables of interest 设计实验和研究，以理解感兴趣变量之间的关系
• Construct a programming workflow to execute an empirical method 构建用于执行实证方法的编程工作流程
• Design an empirical method, interpret and deploy the results of the empirical analysis 设计实证方法，解释并应用实证分析的结果