PhD candidate: Laura Kolbe, MSc
Structural equation modelling (SEM) is becoming one of the central and arguably most popular statistical techniques in the social sciences. Many classical and modern techniques, such as regression analysis, analysis of variance, factor analysis, and item response theory can be formulated as structural equation models. Using fit-indices, researchers can evaluate whether the specified model is a good representation of the data. Obviously, researchers do not want to reject a well-fitting model or accept an ill-fitting model. Therefore, it is important to know the statistical behavior of existing fit-indices across different conditions, and to develop new fit-indices in cases where the existing fit-indices do not behave well. SEM is constantly evolving and extended to be used in non-standard situations, such as multigroup data, categorical data, and meta-analytic data. This PhD project concerns the evaluation of model fit in such non-standard situations.
This project runs from 2017 – 2021.