Training courses

There is an ESF-network on Quantitative Methods in the Social Sciences in place. In general there are very interesting seminars and workshops going on, so please have a look:

a) Seminar 1: Multilevel analysis and multigroup comparison in cross-nation research

This seminar will deal with the actual discussion about the number of groups (countries) needed for valid, reliable and stable conclusion about the impact fixed and random effects in multi-level models. Several studies using simulations based on realistic data will be presented. Power issues are also subject of this seminar. What are the advantages of multi-level analysis over multi-group comparisons of latent means and structural relations in many groups (countries). What method should be used in order to obtain theoretical relevant conclusions depending on the kind of data. The comparative analysis dealt with in this seminar concerns also time series related issues. What are the conditions for using these kinds of analysis? Whenever possible, excellent examples of substantive research in will be used (in education, political science). The seminar will be held in Leuven, preferably in June 2010. First call is foreseen in January 2010.

b) Seminar 2: The effects of sample size and missing data in comparative analysis

This seminar will discuss and evaluate the attempts that are made to make samples comparable over countries (the concept of effective sample size), the adjustment for different sample designs (the computation and application of design weights), the implications of missing data (unit and item nonresponse), and the ways of dealing with these (imputation, weighting for nonresponse). Examples of ESS will be used and the implications of the solutions for statistical inference will be studied. When and how using what kind of weights? The seminar will be held in Vienna in 2011, preferably in September. First call for participants in January 2011.

c) Workshop 1. The quality of composite scores in cross-nation research

The following concepts are subject of this training course: measurement error in single items, invariance testing, weighs for composite scores, quality of composite scores, correction for measurement error in composite scores, comparison between latent means and component scores, comparing means and relations.

There are six sessions for teaching in the morning and five training sessions. Two experts are invited for providing evening lectures.

The workshop will be held at Vienna University or in Cologne in second half of August 2010 with call in January 2010, and is related to the 2nd seminar of 2009.

d) Workshop 2. Multilevel analysis and multiple group comparison in cross-nation research

The combination of these two approaches is very relevant for PhD work which is increasingly based on multiple group datasets. The problems detected in the 3th seminar are material for this workshop that will be prepared shortly after this seminar in 2010.
The seminar is planned for second half of August 2011 in Utrecht. First call for participation in January 2011.