HSSH organizes 3 workshops on multilevel modelling

HSSH is organizing a series of three online workshops on multilevel modelling on November 2nd, 16th, and 30th at 11–12:30. The teacher is professor Lars-Erik Malmberg from the University of Oxford. The language of the workshops is English. The course exercises are carried out in MPlus and R. The course is free for the Centre Campus SSH-researchers.

The topics are:

2.11.: Multilevel Models and Multilevel Structural Equation Models

16.11.: Specification of Measurement Models for Single and Multilevel SEM

30.11.: Time-series modelling using the Bayesian estimator in Dynamic Structural
Equation Model (DSEM).

Each workshop consists of a pre-recorded lecture followed by a live online discussion and Q&A with professor Malmberg. The course also includes a package of written materials and exercises on the topics.

If you wish to participate in the workshop series, please sign up by e-mailing sointu.leikas@helsinki.fi to receive the links to the online lectures and materials.

**********************

Workshop Details:

Part I (2.11.): From the multilevel model (MLM) to MSEM: Hierarchically nested data. Fixed effects and random effects. Within-level and between-level predictors and outcomes. Similarities and differences between MLM and MSEM using 
manifest indicators.

Worked examples in Mplus and R lmer

Online workshop 1.5h, questions and answers, elaborations, discussion, and possible applications to participants’ own work. This first workshop focuses on participants’ questions regarding their own datasets and statistical problems.  You can submit your questions before the workshop.

Part 2 (16.11.): pre-recorded lecture and worked examples: Specification of measurement models for single and multilevel SEM. Goodness-of-fit indices in single-level and multilevel CFA. Matrix notation and estimation. Examples using time-points in students (Intraindividual research on students’ learning experiences), ISEM in Mplus.

Worked examples in Mplus and descriptives in R

Online workshop 1.5h, questions and answers, elaborations, discussion, and possible applications to participants’ own work

Part 3 (30.11.): pre-recorded lecture and worked examples: Time-series modelling using the Bayesian estimator has been incorporated into the Dynamic Structural Equation Model (DSEM). DSEM allows researchers to investigate processes over time under the assumptions of stationarity (no mean-trend) and equidistant lags between time-points. Individual differences in means, (in)stability over time, and innovations (unexplained variance) are possible to investigate. 

Worked examples in Mplus and descriptives in R

Online workshop 1.5h, questions and answers, elaborations, discussion, and possible applications to participants’ own work

Practical issues:

  • The analysis exercises are carried out in R (and RStudio) and MPlus, so you need to have access to these programs in order to get full benefits from the workshops. You can download R and RStudio from the UH Software Center App.
  • Unfortunately MPlus is not among the University of Helsinki default software, but some faculties carry MPlus licenses. If you don’t currently have MPlus, please ask your faculty’s service coordinator or your supervisor about a license. 
  • Alternatively, if you have a personal computer that is not under UH central maintenance, you can download a demo version of MPlus for free at https://www.statmodel.com/demo.shtml
  • The demo version of MPlus can be used for free for one month, so for these workshops you should download it on October 31st the earliest
  • You can also attend without participating in the practical exercises 
  • Please note that the course materials are only for your personal use and their distribution is not allowed.