Introduction to Structural Equation Modelling

Dr James Hall · University of Southampton

An introduction to Structural Equation Modelling for researchers with a background in regression or factor analysis, covering path analysis, latent variable estimation, measurement and structural models, model fit, and basic SEM implementation in R using the lavaan package and the HolzingerSwineford1939 dataset.

Introduction to Assessment Data: Classical Test Theory and Item Response Theory

Prof Christian Bokhove · University of Southampton

In this session we explore how we can analyse assessment data within the frameworks of Classical Test Theory and Item Response Theory. We start with simple data and models, and then slowly build our understanding towards more complex models. We do this within the statistical package R within Rstudio. It is useful to have this open source package available, as well as the script and sample datasets in the zip file on the right, so you can try out the analyses yourself as well.

Introduction to Social Network Analysis

Dr Natalia Lavrushkina · University of Bournemouth

An applied introduction to Social Network Analysis for researchers interested in studying relationships, collaboration, advice, communication and knowledge-sharing. The session covers core SNA concepts, network visualisation, basic metrics, ethical issues in network data collection, and includes a guided Polinode practical using an example email communication dataset.

Preparation for delegates: This session will include a guided practical using Polinode. No software installation is required, as Polinode is web-based. Delegates are asked to acquaint themselves with the example dataset and data description in advance: en.wikipedia.org/wiki/Enron_Corpus and, where possible, set up a free individual account with Polinode before the session: polinode.com. This will help to make the most of the limited workshop time.

Introduction to Multilevel Modelling

Dr Laone Maphane · NCRM / University of Southampton

An introduction to multilevel modelling for researchers working with nested or hierarchical data, such as students within schools. The session covers the rationale for multilevel approaches, intraclass correlation, random intercept models, pupil- and school-level predictors, within- and between-group decomposition, and model diagnostics, with a guided practical in R using the lme4 package and a real secondary school examination dataset.

Introduction to Machine Learning

Mr Brody Hannan · University of Southampton

A hands-on introduction to machine learning for researchers with a traditional statistics background, covering decision trees, random forests, gradient boosting, and clustering using real-world census data. Includes a lecture covering theory and key concepts, before a guided workshop using US Census data.

Introduction to Modelling Causal Inference

Dr Yin Wang · NCRM / University of Southampton

An introduction to causal inference for researchers working with observational data, with a focus on Synthetic Difference-in-Differences (SDID). The session covers the causal identification problem, the difference-in-differences framework, how SDID constructs a weighted synthetic counterfactual, uncertainty estimation via placebo reassignment, and result visualisation, with a guided practical in R using the synthdid package and the California Proposition 99 cigarette consumption dataset.