Bachelorarbeitsthemen

Beschreibung: Regularization methods have gained prominence in psychological research as tools to address challenges such as high-dimensional data, multicollinearity, and overfitting. These techniques, which originated in fields like machine learning and biostatistics, are now being adapted to improve predictive accuracy and enhance model interpretability in psychological studies. Applications range from feature selection in large-scale datasets to stabilizing estimates in complex models.

This bachelor thesis will focus on the use of regularization methods in psychology, offering a literature review of current approaches and their potential to address unique challenges in the field. The review may include applications in subfields like decision science, clinical psychology, or personality assessment, as well as discussions on general trends and usage patterns. The aim is to critically review how regularization is used and can enhance or hinder psychological research methodologies, potentially considering insights from related disciplines such as biostatistics. The thesis can be written in English or German.

Literatur:

Friedrich, S., Groll, A., Ickstadt, K., et al. (2023). Regularization approaches in clinical biostatistics: A review of methods and their applications. Statistical Methods in Medical Research, 32_(2), 425?440. https://doi.org/10.1177/09622802221133557

Hoffmann, J. (2024). Decisions as ill-posed problems: A scoping review of regularization methods in decision science. Decision. Advance online publication. https://doi.org/10.1037/dec0000248
Kontakt: Charles Driver, E-Mail

Beschreibung: Deep learning has emerged as a powerful tool for predictive modeling in fields such as computer vision and natural language processing. While its applications in psychology remain limited, its potential for advancing psychological research is considerable. Deep learning techniques, including feedforward neural networks (FNNs), recurrent neural networks (RNNs), and convolutional neural networks (CNNs), offer innovative approaches to analyzing complex psychological data. These methods allow for capturing patterns and relationships that traditional statistical models may overlook.

This bachelor thesis will explore the application of deep learning in psychological research. The focus may be on specific application areas, such as personality psychology, cognitive modeling, or mental health diagnostics, or on general trends and usage patterns of deep learning across the discipline. The aim is to provide a comprehensive literature review, considering the benefits, limitations, and challenges of integrating deep learning techniques into psychological research. Practical examples and discussions from existing literature will serve as a foundation for understanding how these methods can address prediction-focused research questions in psychology. The thesis can be written in English or German.

Literatur:

Urban, C. J., & Gates, K. M. (2021). Deep learning: A primer for psychologists. _Psychological Methods, 26_(6), 743?773. https://doi.org/10.1037/met0000374

Kontakt: Charles Driver, E-Mail

Beschreibung: Dynamic systems theories (DST) offer a powerful framework for understanding psychological processes, emphasizing the complexity and adaptability of human behavior. These theories are particularly relevant for exploring developmental processes, social interactions, and personality dynamics, providing insights into how patterns emerge and evolve over time.

This bachelor thesis will focus on dynamic systems theories in psychology, allowing students to explore a specific area of interest. Potential focus areas include developmental psychology, social psychology, or personality research, among others. The task involves a thorough literature review, with an emphasis on theoretical frameworks, methodological innovations, and practical applications of DST. Students are encouraged to critically evaluate how DST contributes to understanding dynamic processes in their chosen domain and discuss its implications for future research.

The thesis can be written in English or German.

Literatur:

1. van Geert, P. (2011). The contribution of complex dynamic systems to development. _Child Development Perspectives._ [https://doi.org/10.1111/j.1750-8606.2011.00197.x](https://doi.org/10.1111/j.1750-8606.2011.00197.x)

2. Handbook of Research Methods in Social and Personality Psychology: Complex Dynamical Systems in Social and Personality Psychology. Cambridge University Press. https://www.cambridge.org/core/books/handbook-of-research-methods-in-social-and-personality-psychology/complex-dynamical-systems-in-social-and-personality-psychology/FCC5DED25E00FD30832969E11EE27684
Kontakt: Charles Driver, E-Mail

Beschreibung: Cross-lagged panel modeling (CLPM) is a statistical technique widely used in psychology to investigate reciprocal causal relationships in longitudinal data. Historically favored for its simplicity, the CLPM has faced criticism for its inability to distinguish within-person dynamics from stable between-person differences. This has led to the development of alternative approaches, such as the random-intercept CLPM, which aim to address these limitations by incorporating stable-trait components. The controversy surrounding CLPM centers on its relevance and appropriateness for answering various psychological research questions. Advocates of the CLPM argue it is well-suited for studying stable, individual-level processes, while critics highlight its susceptibility to biased or spurious causal effect estimates.

This bachelor thesis will explore the ongoing debate about the CLPM and its alternatives, critically examining the theoretical and methodological implications of these approaches. The focus will include understanding the conditions under which different models perform best, the impact of research design choices, and the role of timescale in studying psychological processes. The thesis aims to provide a comprehensive literature review on the topic, offering insights into how these methodological controversies influence the interpretation of longitudinal psychological data. The thesis can be written in English or German.

Literatur:

Hamaker, E. L. (2023). The within-between dispute in cross-lagged panel research and how to move forward. Psychological Methods. Advance online publication. https://doi.org/10.1037/met0000600

Lucas RE. Why the Cross-Lagged Panel Model Is Almost Never the Right Choice. Advances in Methods and Practices in Psychological Science. 2023;6(1). https://doi.org/10.1177/25152459231158378
Kontakt: Charles Driver, E-Mail