Audio Communication Group

Factor analysis and structural equation modeling 2022

An introduction to latent variable analysis using Jamovi and R-lavaan.

Dr. Steffen Lepa

October 10-12, 2022, 10-12h and 13-16h respectively.
(in presence considering possible corona protection regulations of the TU Berlin valid at that time).

Room H 3001, main building of the TU Berlin, Straße des 17. Juni 135.

In social science research projects with music- or acoustics-related questions, people become their own observers in experiments or surveys, who then fill out questionnaires. In many cases, this results in measurements that are subject to random or systematic errors. This can quickly lead to problems, especially with small samples and complex theoretical models. Structural equation modeling offers an effective and in psychology and sociology meanwhile widely used simple statistical method to estimate arising measurement errors and to calculate them out of construct variables, to optimize survey instruments thereby and to test even very complex model concepts, which go far beyond the standard procedures of statistical software packages. The block seminar offers a simple introduction to the possibilities of structural equation modeling in Jamovi and R, this is done using numerous examples from music and acoustics research, topics covered are:

  • Path analysis
  • Mediation analysis
  • Exploratory factor analysis
  • Confirmatory factor analysis
  • Structural equation models (multiple regression / analysis of variance)
  • Invariance tests (e.g. for the evaluation of language versions of questionnaires)
  • Factor analysis and structural equation models for longitudinal data
  • Data-based development and optimization of measurement models / inventories

The course combines theoretical (morning) and practical parts (afternoon). Participation requires a basic knowledge of statistics (significance testing, regression analysis) and your own notebook computer (current Windows, MacOS or Linux). For the exercises, sample data sets will be provided, but you can also use your own data. External participants will receive a certificate of attendance and students of the TU Berlin can receive credit for the course in the elective section.

Due to limited room capacities, we ask for a binding registration by August 1, 2022 via email to steffen.lepa(at)tu-berlin.de.

You will then receive a final confirmation of participation at the beginning of September.

The seminar is free of charge for members (employees, students, PhD students) of the TU Berlin. For external participants we charge a fee of 600 EUR (reduced to 300 EUR for students and PhD students of other universities).