POSTPONED Trajectory clustering analysis of longitudinal/ repeated measurements

Please note: This training event has been postponed until further notice.

This Advanced Quantitative Methods Training event has been jointly organised between the WRDTP and the Scottish DTP, and is open to all ESRC funded and non-ESRC funded students across the partner universities. This event forms part of a 3 day series of events – students can attend events individually or attend all 3.

This course is split into two sessions:

Session 1: Introduction to longitudinal clustering analysis

The participants will be introduced to longitudinal data structures, and how to manipulate and visualise them in the R environment. This will be followed by an introduction to two widely used clustering methods: group-based trajectory modelling and non-parametric k-means for identifying distinct trajectory groupings in a longitudinal dataset. Participants will then carry out a class exercise focusing on longitudinal crime data from Greater Manchester.

Session 2: The non-parametric ‘akmedoids’ approach

Here, we introduce a new R-package called ‘akmeans’ specifically designed for identifying long-term trends in longitudinal datasets based on pre-determined slopes. A comparative analysis of the akmeans and the existing methods (in session 1) will be carried out. This will include an overview of the pros and cons of each method as well as their potential implications for a wide range of domains.

By the end of the course, participants should be able to:

Gain an understanding of longitudinal data structures and learn how to manipulate and visualise them in the R environment
Carry out trajectory clustering analysis of longitudinal data in order to identify distinct trajectory groupings
Customise the non-parametric k-means method in order to fit a specific research objective
Visualise clustering outcomes and plot their spatial manifestations
Structure clustering results for causal analysis

Workshop organiser/ leader

Additional speaker

Dr Monsuru Adepeju

Senior Research Associate, Manchester Metropolitan University

Dr Adepeju is a Research Associate at the Manchester Metropolitan University Crime and Well-Being Big Data Centre. He obtained his PhD degree in GIS and Crime Science fro m University College London where he was able to develop a number of crime predictive methods, some of which are now being applied in a real policing environment.

His research interest covers a wide range of spatial and quantitative techniques for crime and policing data analytics, census/neighbourhood analysis, transport and urban dynamics. Some of his recent research has focused on investigating the relationships between socioeconomic inequalities and crime risk distribution across the UK.

Sam Langton

Research Associate, Manchester Metropolitan University

Sam Langton is a PhD candidate at Manchester Metropolitan Crime and Well-Being Big Data Centre. His research focuses on the geographic distribution of offender residences using longitudinal data from a large metropolitan area in England. Sam is supervised by Jon Bannister (Sociology), Gary Pollock (Sociology) and Liangxiu Han (Computing). He obtained a BSc in Social Policy and Government from the London School of Economics in 2012. After working in accounting for two years, he moved to the Netherlands to complete a two-year MSc in Sociology and Social Research at Utrecht University. Whilst there, he worked as an intern at the Netherlands Institute for the Study of Crime and Law Enforcement (NSCR) in Amsterdam, where he completed his master’s dissertation with Utrecht University on residential burglary target selection. Sam is particularly interested in the impact of spatial scale, longitudinal methods (growth trajectories, clustering) and data visualisation using open software (R and QGIS).

This 3 day training event is linked to the Understanding Inequalities Project, a three-year project funded by the ESRC. Further information can be found by clicking on the link on the right.

All attendees should have a basic familiarity with R Studio.

There are 30 places available at this AQM workshop

PLEASE NOTE: Students are responsible for arranging travel to and from this Advance Quantiative Methods Training session. The WRDTP cannot reimburse travel costs to this session.