Multivariate Statistical Analysis 1: Dimension Reduction

In multivariate statistics we simultaneously model and estimate variability in more than one variable often in order to examine the relationship between variables. This workshop examines the key aspects of moving from univariate to multivariate analysis, and the situations and scenarios where multivariate analysis is typically applied. The focus will be on practical application of concepts through examples.

Topics covered will include:

• Motivations for undertaking multivariate analysis
• Statistical principles for multivariate analysis
• Dimension reduction techniques including principal components analysis (PCA), factor analysis (FA), correspondence analysis (CA) and non-metric multidimensional scaling (nMDS)

This is just one workshop from a modular training programme made up of 1.5 hour workshops, each focusing on a single statistical method offered by Statistical Consulting within the Sydney Informatics Hub. Statistical Workflows giving practical step-by-step instructions applicable in any software are used and include experimental design, exploratory analysis, modelling, assumption testing, model interpretation and presentation of results. Researchers are encouraged to design a custom programme tailored to their research needs.

Pre-requisites

Prior experience with statistical modelling is assumed, as the basics of regression modelling will not be covered. Please consider attending Linear Models 1 and/or Linear Models 2 workshops to come up to speed beforehand. Note that this workshop does not require knowledge of or use of specific statistics software. The analysis methods may be performed using a wide range of commonly available software.

Attendance options

You will get the most out of this session by attending in person. This way allows you to interact with the presenters, ask questions directly, and catch up after the session. Please only choose online if circumstances prevent you from attending campus.

In person: Choose the In person ticket option. University rules governing COVID related requirements will apply in the venue. In the event that in person delivery is not possible, you will be provided with the Zoom link for remote attendance.

Online: Choose the Online Zoom ticket option. It is recommended to have a dual monitor setup to view and participate remotely in the session. Registered attendees will receive a link to the Zoom event closer to the date.

Sydney Informatics Hub

• Open to: Academic and Professional Staff, research students, and affiliates of The University of Sydney (with a valid UniKey)
• Please use your University of Sydney email address to register i.e. @sydney.edu.au, @uni.sydney.edu.au, etc
• If you do not have a UniKey, please contact us to confirm your position after registration.

Registrations from open web email clients (@gmail.com, @hotmail.com etc) will AUTOMATICALLY be cancelled, without us contacting you!!!