Course content

The course introduces students to structural equation modeling (also referred to as latent variable modeling). Using the traditional regression analysis as a basis, the course goes through the most common analytical approaches used in the domain of structural equation modeling (SEM). The course first treats the subjects of specification and estimation of standard confirmatory factor models as well as full SEM models. Subsequently, the course goes on with more advanced SEM techniques such as multi-group, higher-order confirmatory factor, multilevel, and growth modeling. The main software to be used are STATA and Mplus. R (lavaan) codes will also be provided depending on need.

Learning outcome

Knowledge:
Students have theoretical knowledge about the rationale and statistical assumptions for estimating SEM models as well as interpretation and communication of SEM-results in a scientific format.
Skills:
Students can on their own manage the course software (STATA/Mplus) to employ the analytical approaches listed in the course content.
Competence:
Students have obtained theoretical/practical competence and accordingly can apply SEM to various research problems and associated research instruments.

Learning methods and activities

Lectures and lab exercises.The course consists of 30 hours and is arranged over a period of five days. The form of assessment is an individual paper of 4000-5000 words. The paper will be assessed by using the passed/not passed grading option. The paper should demonstrate that the candidate has been able to employ one or several of the techniques treated in the course. The candidate can use primary and/or secondary data. The paper cannot be identical to an article that is included in the doctoral thesis. The course requires a minimum of five registered PhD candidates to be held.

Compulsory assignments

• Deltakelse på forelesninger og lab øvelser

Recommended previous knowledge

Basic knowledge of multiple regression and/or factor analysis is recommended. Furthermore, experience with statistical software such as STATA, Mplus etc. may be useful.

Required previous knowledge

Completed Master's degree or similar. The course is limited to a maximum of 25 candidates. Candidates admitted to a PhD programme have priority.

Course materials

The recommended literature will be sent to the participants before the course starts.