Quantitative Methods in Education Group
We bring together researchers to advance quantitative methods in education research, policymaking, and professional practice.
Masterclasses
Free online sessions on a single quantitative method, led by QMEG members and invited specialists. Most sessions introduce the method and then demonstrate it in software such as R, Stata or Python.
Upcoming
Sessions run on Microsoft Teams and last about two hours. Register on Eventbrite to receive the joining link.
An Introduction to MAIHDA for Studying Intersectional Inequalities
Prof George Leckie
University of Bristol
MAIHDA is a multilevel approach to studying inequalities across intersections of characteristics such as gender, race and socioeconomic background. The session applies it to PISA attainment data, with demonstrations in R and Stata.
Register on EventbriteIntroduction to Causal Inference with Machine Learning
Mr Brody Hannan
Southampton Education School
An introduction to Double Machine Learning, which uses flexible machine learning models to control for confounding while still giving valid effect estimates. Includes a worked example in Python. Suitable if you know basic regression; no causal inference experience needed.
Register on EventbritePast masterclasses
Recordings and slides from earlier sessions are free to view.
Longitudinal Growth Modelling with SEM
Dr James Hall
Southampton Education School
An introduction to longitudinal growth modelling with structural equation modelling, followed by a worked demonstration in R using lavaan.
View SlidesInternational Large-scale Assessment (ILSA) Data with R
Prof Christian Bokhove
Southampton Education School
Analysing data from international large-scale assessments such as PISA and TIMSS in R.
Watch RecordingMachine Learning and Traditional Statistical Models in Complex Survey Data
Dr Yin Wang
National Centre for Research Methods
How machine learning methods compare with traditional statistical models when analysing data from complex survey designs.
Watch RecordingDynamic Structural Equation Modelling
Prof Lars Malmberg
University of Oxford
Dynamic structural equation modelling (DSEM) combines multilevel, time-series and structural equation models for intensive longitudinal data, such as daily diaries or repeated measures within lessons.
Watch RecordingStay Up to Date
Join the QMEG mailing list to receive updates on upcoming events, new resources, and collaboration opportunities.
Or contact us directly: J.E.Hall@soton.ac.uk