Tuesday, 7 October 2014

Longitudinal statistics seminar


The link above leads to the poster for this seminar on Friday 24th October.

The very successful longitudinal data analysis course will run on Friday 24th October from 9.30-4.30pm led by CCACE statistician Dr Mike Allerhand and will take place in S1, Department of Psychology, 7 George Square, Edinburgh.

The course is £100 for external applicants. Payment is due by Friday 10th October.
Places are limited - please register by emailing Anna Sim on ccace@ed.ac.uk.
If you have any questions, please email caroline.brett@ed.ac.uk.

Longitudinal data enables researchers to understand how cognitive and physical function changes over time and disentangle individual and environmental influences on that change. However, analysis of longitudinal data can be difficult and daunting.
This 1-day course aims to explain the complexities of longitudinal data analysis in simple statistical terms. It will describe methods  and models for exploring and analysing longitudinal data and their assumptions, strengths, and weaknesses, equipping participants with sufficient knowledge  to make appropriate modelling choices and interpret results.

The course does not require skills with any particular program. The course is not about programming, and is not a demonstration of analyses that can be done using these programs. There should be something for users of most statistics programs.
Models described will include:

Fixed effects regression models
Multi-level (mixed effects) models
Latent growth curve models
Autoregressive and latent change score models
Cross-lagged joint models of bivariate longitudinal data The Methods described will include:
Time-varying and time-invariant covariates
Parametric and non-parametric curve fitting
Semi-parametric generalized additive models
Segmented models
FIML and REML parameter estimation"


If you attend, let me have your comments on what promises to be a very useful seminar. If you cannot attend, demand that that course be written up as a booklet, (with a copy sent to me).

1 comment: