ÐÂÏã¸ÛÁùºÏ²Ê¿ª½±½á¹û

Regression analysis, one of the most important quantitative methods, has interrogated countless empirical and interdisciplinary questions using small- and large-scale datasets. This module will explore the development of regression models and contextualise the methodological development in their social and cultural context, using applications across a wide range of disciplines such as astronomy, biology, geography, social science, epidemiology, environmental science, artificial intelligence, and so on.

This module will start with the Ordinary Least Squares (OLS) regression, then Generalised Linear Models (GLM) and then explore contemporary models in machine learning. A series of real-world applications discussed in this module will enable students to understand the fundamentals of regression models; gain knowledge on selecting appropriate models and interpreting regression outputs. This module will discuss some key concepts and debates about causality. Examples of applications from various disciplines will be used to illustrate two different purposes of regression analysis: predictive and causal inference. This module combines lectures, seminars, group work and hands-on practicals using R. Students will gain skills to implement various regression models and draw conclusions from the results.

Teaching Delivery

This moudle is taught in 10 weekly lectures and 10 weekly coding practicals.

Indicative Topics *Based on module content in 2023/24, subject to possible changes

• Ordinary Least Squares (OLS) regression
• Logistic regression
• Moderation analysis
• Principal Component Analysis (PCA)
• Multilevel model
• Causation
• Directed Acyclic Graph (DAG)
• Mediation analysis

Module aims and objectives

The purpose of this module is to meet the current and future needs for applying quantitative methods to understand real-world phenomena by introducing a broad set of advanced algebraic, statistical models, statistical inference, predictive inference and causal thinking. This module provides students with the technical skills and critical thinking required for making connections across different subjects, identifying and answering interdisciplinary challenges and problems. By the end of this module, students will:Ìý

1. Understand the basic principles of several different regression models Ìý
2. Appreciate the context in which different regression models are best applied.Ìý
3. Gain a basic understanding of predictive and causal inference. Ìý
4. Gain practical experiences of applying the methods studied, using R programming for statistical computing and data visualisation.

Recommended Reading

• Kabacoff, Robert I. R in Action, Third Edition: Data Analysis and Graphics with R and Tidyverse. New York: Manning Publications Co. LLC, 2022. Print. (ÐÂÏã¸ÛÁùºÏ²Ê¿ª½±½á¹ûlibrary online)
• R for Data Science Ìý