Symbolbild: Das Wort "Forschung", zusammengesetzt aus Buchstaben-Stempeln

Research projects

Symbolbild: Das Wort "Forschung", zusammengesetzt aus Buchstaben-Stempeln
Image: Jan-Peter Kasper (University of Jena)
This page has been machine translated.

Ongoing third-party funded projects

  • COSDIMH - Combining Surveys and Digital Tracking Data for Mental Health Research from a Computational Social Science Perspective

    Third-party funding provider: German Research Foundation (DFG)

    Koch, Tobias, Friedrich Schiller University Jena
    Domahidi, Emese, Ilmenau University of Technology

    Despite increasing academic attention to the relationship between mental health (MH) and ubiquitous digital media use (DMU), there are notable methodological challenges in this subject area|field. These include a lack of systematic comparisons of surveys using traditional questionnaires and objective digital data sources, an insufficient understanding of the temporal dynamics between DMU and MH, and a lack of experimental findings to clarify the causal relationship between the two variables.

    In the proposed interdisciplinary project COSDIMH, we will focus on combining different DMU and MH measures in survey data and digital tracking data. We will compare measures within and between survey and digital tracking data to analyse their validity and reliability. We investigate the short-term dynamic relationships between DMU and MH measures by combining different measures from multiple surveys and digital tracking data. Finally, we analyse the causality of the relationship between DMU and MH in an experimental setting and implement personalised interventions. We identify innovative procedures, methods and potential pitfalls for such combinatorial research approaches.

  • Enhanced Assessment of Social and Health-Related Processes in Panel Studies through Event-Contingent Multimethod Experience Sampling Designs (SHERPA)

    Third-party funding provider: German Research Foundation (DFG)

    Prof. Dr Tobias Koch, Prof. Dr Michaela Riediger

    Social interactions have profound effects on human well-being, with stable attachments promoting mental and physical health and social isolation leading to negative outcomes. Our project focuses on the crucial developmental phase of young adulthood, which is characterised by navigating new social dynamics and building important social networks. Through the use of innovative technologies, such as experience sampling methods and Bluetooth-enabled devices, we aim to capture real-time social interactions in various real-world contexts. Through integration with the German National Educational Panel Study (NEPS), we also aim to investigate the complex relationship between social interactions, health outcomes and developmental predictors. This project contributes to the Infrastructure Priority Programme "New Data Spaces for the Social Sciences" (SPP 2431) and advances Research Area 4 by providing new insights through multi-method ambulatory assessment studies. By developing innovative technologies and new statistical methods, we enable researchers to explore with high precision and validity how social interactions influence health and development.

    Further partner institutions|cooperation partners: Prof. Dr Franz Neyer (FSU Jena), Prof. Dr Jana Holtmann (Leipzig University), Prof. Dr David Martínez Iñigo (Rey Juan Carlos University), Prof. Dr Francisco Serradilla (Universidad Politécnica de Madrid)

  • Definition and estimation of causal effects in latent state trait models (CaST)

    Third-party funding provider: German Research Foundation (DFG)

    Prof Dr Tobias Koch, Prof Dr Manuel Völkle (HU Berlin)

    LST theory is one of the most influential measurement theory approaches in Psychology. It opens up the possibility of investigating processes of change and variability in psychological characteristics using suitable statistical models. Unanswered questions relate to the conditions under which LST models can be causally interpreted in quasi-experimental designs. Our project addresses this issue from different angles: We compare statistical with causal models, discrete-time with continuous-time models, and stochastic with graph-based causal theories. As part of the project, we are investigating how causal effects can be precisely defined and estimated in LST models with autoregressive effects.

    Further partner institutions|cooperation partners: Dr Christian Gische, Minne Hagel, Moritz Ketzer, Fabian Münch

  • Collaborative project: "Thuringian University Initiative for AI in Higher Education - THInKI" (Prof. Koch), Federal Ministry of Education and Research

    Third-party funding provider: Federal Ministry of Education and Research (BMBF)

    As part of the project, interdisciplinary self-study courses on "computational communication science and social sciences" (computational methods in Communication Science and Social Sciences) are being developed in cooperation with the Ilmenau University of Technology under the direction of Prof Dr Domahidi. The project focuses on the development of established methods of machine learning and automatic text analysis.
    You can find more information here: https: //tzlr.de/projects/thinki/External link

    Further partner institutions|cooperation partners: Christian Bloszies

Completed third-party funded projects

  • Analysing assessor (method) effects in empirical educational research (ABeBi)
  • Development of mixed distribution item response models for the analysis of cross-classified multirater data and their application in teaching evaluation research (Prof. Koch), German Research Foundation

    Third-party funding provider: German Research Foundation (DFG)

    The project aims to extend item response theory (IRT) models to analyse complex cross-classified multirater (CCM) data using mixed distribution approaches to identify unknown groups. It builds on current research findings in the field of multi-method research and extends existing models to CCM data that occur in various areas of Psychology and empirical educational research. A concrete example is teaching evaluations at institutions of higher education, where the same students evaluate several teachers, resulting in a CCM data structure. The newly developed models will be applied to teaching evaluation data in order to analyse the convergent and discriminant validity of teaching evaluations in detail. The statistical performance of the models will also be evaluated in Monte Carlo simulation studies. This approach contributes to making the review of teaching quality at institutions of higher education more precise and providing scientifically sound findings.

    Further partner institutions|cooperation partners: R. Maximilian Bee

Permanent projects

  • Central Office for Teaching Evaluation (ULe)

    The Central Office for Teaching Evaluation (ULe) supports the quality development process at the University of Jena on two levels through systematic and scientifically sound teaching evaluation procedures: (a) at the level of the individual course|class and (b) at the level of an entire degree programme/subject.
    Feedback instruments have been developed for both levels to support and promote dialogue between students and teaching staff.

    Learn more de
  • kompetenztest.de

Research areas

  • Change measurement (esp. latent state-trait models)
  • Multimethod research (esp. multitrait-multimethod analyses)
  • Multilevel analyses in Psychology
  • Applications of G-factor models
  • Bayesian estimation techniques
  • Psychometrics (classical and probabilistic test theory)