Sessions and Keynotes
In eight sessions and two symposia, more than 180 participants were able to discuss this year's main topic of the NEPS conference. In total, there were 32 contributions on various aspects of the methodological practice of data collection and evaluation in large-scale assessments in educational science.
The opening talk of the conference was given by Dr. Suzanne Yak from the University of Amsterdam (NL) on how to model survey data when respondents are grouped within larger clusters (e.g., schools). In her talk, she presented interpretive difficulties associated with existing two-stage models for cluster-level constructs and discussed possible alternative options for modeling.
The keynote address was given by Prof. Matthias von Davier from Boston College (USA). He presented various models that can be used to compare and contrast latent variables for longitudinal approaches. He presented examples and application areas and addressed the tension between model fitting and predictions.
This year, the NEPS Publication Award, which comes with a prize money of 1,000 EUR, was awarded to Sönke Matthewes. In his publication "Better Together? Heterogeneous Effects of Tracking on Student Achievement," published in The Economic Journal, the researcher at the Social Science Research Center Berlin (WZB) addresses the question of how higher- and lower-achieving students can benefit from joint instruction in a comprehensive school. Using NEPS data, he shows that joint schooling in early secondary education can reduce inequality. Early and rigid forms of achievement segregation, on the other hand, can reduce the efficiency of school systems.
Special Issue "Large-scale Assessments in Education"
This year's NEPS conference will be followed by a special issue of the journal "Large-scale Assessments in Education." For this issue, contributions are sought from a variety of disciplines that address methods of data collection and analysis in educational studies.
The special issue, edited by a team from LIfBi and the University of Bamberg, focuses on problem statements and innovative solutions for the analysis of psychological constructs in educational longitudinal studies such as NEPS. This includes in particular empirical applications of new methodological approaches, statistical simulations, or software reviews of methods in large-scale assessments of educational research. The contributions sought do not have to deal specifically with NEPS, but should be applicable to longitudinal studies in educational research and must be submitted by November 30, 2021. The special issue will also include selected NEPS conference papers.