MEAFA was so overwhelmed with the interest for the workshop on Multilevel/mixed Models Using Stata, 10-12 Feb 2010, that a second identical workshop was run back-to-back during 8-10 Feb 2010. In the end, all places were booked out within three days of the initial announcement (max 34 per day).
The attendance exceeded our expectations with participants arriving from different states of Australia and from New Zealand, from academia (PhD students and academics from a wide range of disciplines), governmental bodies and the private sector.
On the first day of the workshop Demetris provided a crash course on Stata 11, covering reproducible and tractable routines, ways to customise the software, basic data management techniques, basic statistical and estimation commands and how to produce tables and graphs. The targeted audience were those who were not familiar with Stata or wanted become more efficient in using Stata 11 for their work.
Bobby carried out the next two days of the workshop impeccably, engaging theory and application of multilevel/mixed models on real data. The Multilevel/mixed Models sessions started with the classic random-intercept linear model, where several approaches for fitting this type of model, along with the associated benefits and assumptions were discussed. The second session had the focus on random coefficients and the various covariance structures that can be imposed with multiple random-effects terms. On his second day, Bobby covered various methods for fitting more complex models, including crossed-effects models, growth curve models, and models with complex and grouped constraints on covariance structures. In the last session, the predictions, model diagnostics, and other post-estimation tasks were discussed along with other types of responses, in particular binary and count responses.
We thank all participants for their support and comments.
Please keep visiting MEAFA's website for future workshops.