364-1131-00L Methods in Management Research: Quantitative Research - Multilevel and Structural Equation Modelling
Semester | Spring Semester 2020 |
Lecturers | S. Raeder |
Periodicity | yearly recurring course |
Language of instruction | English |
Comment | If you have already successfully completed "364-1020-04L Methods in Management Research: Quantitative Research - Multilevel Analysis" and / or "364-1020-05L Methods in Management Research: Quantitative Research - Structural Equation Modelling", then you will not be permitted to attend this course. |
Abstract | Multilevel modelling and structural equation modelling are two regression-based methods of data analysis that are increasingly used in applied fields of Management and Organizational Behaviour. The course provides basic knowledge about both methods (e.g., design, analysis, reporting) and explains more advanced models (e.g., moderation, mediation, longitudinal). |
Learning objective | After this course, students will be able to: - design a multilevel model and a structural equations model, - calculate a multilevel model and a structural equations model, - interpret model results, - report model results, - assess models reported in existing research. |
Content | The course provides skills and knowledge for the design and analysis of multilevel models and structural equation models (SEM). Multilevel analysis is required for data collected in clustered samples for which sampling decisions were taken in several steps (e.g. first choosing firms, then employees in firms). Structural equation modeling (SEM) is a technique to build models and test causal relationships including latent variables, several outcome variables and intervening variables. The course teaches basic skills and advanced models for both methods. This allows students to compare both methods and their use and to choose the appropriate method for their own research. The basic knowledge in multilevel modelling covers: building the statistical multilevel model, calculating a multilevel model in SPSS, reporting of results and required sample size. The basic knowledge in SEM include: model identification and model fit, measurement model and structural model, calculating a SEM in Mplus and reporting of results. Advanced topics for both methods refer to moderation (i.e., interaction effects), mediation (i.e., intervening variables) and longitudinal analysis with three or more measurement waves. Comparing options provided by the different methods allows us to understand strengths and weaknesses of both methods in relation to research goals. Students work on six assignments during the course. In two assignments, students find sample papers from their field of research applying each of the methods. Two assignments consist of an analysis with each of the methods. One assignment refers to designing a multilevel model within a student’s own field of research. The final assignment requires students to report an analysis for presentation in a scientific paper. Students can use their own data for the assignments requiring data analysis. Basic knowledge in regression analysis is necessary for following the course. The course uses SPSS for multilevel modelling and Mplus for SEM. |