Abschlussarbeiten

(with Rasmus Bruckner)

Adaptive behavior requires dynamic learning processes. New information that can be relevant to our choices is often uncertain and ambiguous. For example, stock market data vary to some degree from day to day, and investment decisions should be made based on the average underlying price (e.g., averaged across a certain period). In other scenarios, the environment can change more fundamentally, such as after an economic crisis, which requires stronger behavioral adjustments (e.g., selling stocks to avoid losses).

In our project, we are interested in how humans adjust their learning and decision-making behavior to such uncertain and changing environments. We are currently developing a task that allows us to measure how humans dynamically regulate their learning behavior. This project is very well suited for a Bachelor's thesis, where students would work on a behavioral experiment with different versions of our task. Students would be involved in the data collection and learn behavioral analyses, also providing some insights into state-of-the-art computational modeling.

Suggested literature:

• Bruckner, R. & Nassar, M. R. (2024). Decision-making under uncertainty. Accepted for publication in Encyclopedia of the Human Brain, 2nd edition (Academic Press).
  https://osf.io/preprints/psyarxiv/ce8jf
• Nassar, M. R., Bruckner, R., & Frank, M., J. (2019). Statistical context dictates the relationship between feedback-related EEG signals and learning. eLife, 8:e46975.
  https://elifesciences.org/articles/4697

(with Jordan Deakin)

Shopping can be enjoyable, but when faced with countless options, how do we efficiently decide on what to buy? Every product has multiple attributes—such as screen size, storage, and price when buying a phone—that must be evaluated and integrated to make the best choice. Recently, we proposed a Bayesian computational model that predicts how humans efficiently allocate their attention in order to make these choices. The model suggests that individuals sample information about each attribute over time, using it to continuously update their overall belief in an option’s value. A choice is made once a person is sufficiently confident that one option is more valuable than the others.

This project aims to further investigate the cognitive mechanisms involved in this decision-making process. To validate the model's assumptions, we will employ a combination of eye-tracking and EEG to capture decision-making in real time. Participants will complete a task in which they choose between different options, while eye movement and EEG data are recorded. Open to Bachelor’s and Master’s students, this project provides valuable hands-on experience with the collection and analysis of EEG and eye-tracking data, as well as the opportunity to apply computational modelling to interpret the findings.

Suggested Reading:

• Busemeyer, J.R., Gluth, S., Rieskamp, J., and Turner, B.M. (2019). Cognitive and neural bases of multi-attribute, multi-alternative, value-based decisions. Trends in Cognitive Sciences, 23, 251–263
• Gluth, S., Rieskamp, J., and Büchel, C. (2013). Classic EEG motor potentials track the emergence of value-based decisions. NeuroImage, 79, 394–403.
• Gluth, S., Rieskamp, J., and Büchel, C. (2012). Deciding when to decide: time-variant sequential sampling models explain the emergence of value-based decisions in the human brain. Journal of Neuroscience, 32, 10686–10698.
• Ting, C. C., & Gluth, S. (2024). Unraveling information processes of decision-making with eye-tracking data. Frontiers in Behavioral Economics, 3, 1384713.

(with Sebastian Gluth)

Assume you are asked to choose your preferred fruit out of two options, an apple or an orange. Now assume you are asked to select from the same two options the one you like less. In theory, the second decision should simply be a mirror of the first one (picking the apple is equivalent to rejecting the orange). However, previous research has indicated that choosing and rejecting have their own specific dynamics, and that people sometimes even choose and reject the same option (Shafir, 1993). In this eye-tracking project, we are studying search and decision dynamics when choosing and rejecting options with positive and negative attributes. Participants will be asked to choose or reject options of monetary rewards (positive attribute) that are obtained via exerting cognitive effort (negative attribute), while their eye movements are being recorded to study how they distribute attention among these attributes, depending on whether their goal is to select or de-select an option.

Bachelor and Master students can take part in this project for their theses, with the only difference being the amount of study support and the complexity expected of the student’s output. Thesis can emerge from analyzing behavioral data, eye-tracking data, and – if interested – applying cognitive modeling to these data.

References

• Sepulveda, P., Usher, M., Davies, N., Benson, A. A., Ortoleva, P., & De Martino, B. (2020). Visual attention modulates the integration of goal-relevant evidence and not value. eLife, 9, e60705.
  https://elifesciences.org/articles/60705
• Shafir, E. (1993). Choosing vs. rejecting: While some options are both better and worse than others. Memory & Cognition, 21(4), 546–556.
  https://link.springer.com/article/10.3758/BF03197186
•  Westbrook, A., Lamicchane, B., & Braver, T. (2019). The subjective value of cognitive effort is encoded by a domain-general valuation network. Journal of Neuroscience, 39(20), 3934–3947.
  https://www.jneurosci.org/content/39/20/393

(With Mrugsen Gopnarayan; mrugsen.gopnarayan@uni-hamburg.com)

Human decision-making in social settings is not only a matter of isolated cognition, but also of dynamically coupled processes between interacting individuals. Hyperscanning EEG simultaneously recording neural activity from two participants, provides a way to study this coupling directly. In our current study, we extend the cooperative buyer–seller bargaining task to a hyperscanning setting. A buyer evaluates multi-attribute products, while a seller infers the buyer’s preferences from choices, response times, and eye movements. By recording both brains simultaneously, we can investigate whether inter-brain synchrony supports successful preference inference and leads to better bargaining outcomes.

Practical Application: If successful, this study could provide insights into how shared neural dynamics facilitate cooperation in real-world negotiations, training, or adaptive human–AI systems.

For Master’s Thesis: Conducting cooperative bargaining studies with hyperscanning EEG. Behavioral and neural data analysis (linking decision-making models with inter-brain connectivity). Example hypothesis: Do sellers who show higher neural synchrony with buyers achieve more efficient agreements?

Suggested Literature:

• Babiloni, F., & Astolfi, L. (2014). Social neuroscience and hyperscanning techniques: Past, present and future. Neuroscience & Biobehavioral Reviews, 44, 76–93.
• Pan, Y., Novembre, G., Song, B., Li, X., & Hu, Y. (2018). Interpersonal synchronization of inferior frontal cortices tracks social interactive learning of a song. NeuroImage, 208, 116432.

(with Barbara Oberbauer)

A majority of decision-making research has been concerned with studying how individuals make choices, such as selecting their preferred item from several options. However, in everyday life, we frequently postpone decisions, a behavior that has received comparatively little attention in the literature and is challenging to study in laboratory settings, where choices are often structured as forced decisions. The aim of this study is to address this gap and investigate the cognitive mechanisms of choice deferral.

More specifically, we aim to provide evidence that helps reconciling choice deferral with one of two concurrent accounts of the underlying cognitive processes that have been proposed in the literature. One line of reasoning proposes that when participants are asked to choose between making or deferring a decision, they pre-play the decision-making process and, in case of deferral, terminate it based on a certain criterion (e.g., a time criterion) (Bhatia & Mullett, 2016; Busemeyer & Rapoport, 1988; Gluth et al., 2013; Jessup et al., 2009). A concurrent line of reasoning suggests that participants might be exploring the absolute desirability of the choice options first when considering to defer, as proposed by two-stage models, such as White and colleagues’ (2015) model, rather than pre-playing the decision process.

To understand which of these accounts best conceptualizes choice deferral, we will collect choice, rating, response time, and eye-tracking data from participants completing a choice task, a choice deferral task, as well as an appraisal task.

This project is equally suited for Bachelor and Master theses, where students may work with behavioral and/or eye-tracking data and with computational models when interested. Students are expected to assist with data collection.

Suggested literature:

• Bhatia, S., & Mullett, T. L. (2016). The dynamics of deferred decision. Cognitive psychology, 86, 112-151.
• White, C. M., Hoffrage, U., & Reisen, N. (2015). Choice deferral can arise from absolute evaluations or relative comparisons. Journal of Experimental Psychology: Applied, 21(2), 140.

(with Chih-Chung Ting)

In addition to task difficulty (e.g., the difference in preference between options), recent studies have shown that people tend to make faster decisions when choosing between two options that are both highly attractive, compared to when both options are less attractive (Ting and Gluth, 2025; Shevlin et al., 2022). Interestingly, this negative relationship between overall value (OV: the sum preference of all available options) and response times (RTs) can be reversed when the decision goal shifts from choosing the most attractive option to choosing the least attractive one (Sepulvedat et al., 2020; Romy et al., 2019). This goal-dependent OV effect on RTs suggests that people are flexible in their decision-making and can adapt their choices to different goals. However, in daily life, our decisions often involve other types of goals. For example, sometimes we need to choose an option that is merely acceptable (neither the best nor the worst), or we may need to evaluate several options as a set rather than selecting a single one. 1. How do people search for information when they are asked to choose a single option compared to when they are asked to appraise a set of options? 2. How does overall value affect choice, response time, and eye movements when the decision goal is changed?

This project aims to address these questions by systematically manipulating decision goals and overall value (OV) in both perceptual tasks (e.g., brightness discrimination) and preferential tasks (e.g., snack choices). We have conducted online study and are preparing a laboratory experiment combining behavioral and eye-tracking data. Bachelor and master students are welcome to join this project by analyzing existing eye-movement datasets or/and collecting new data to test research questions formulated in their theses. Master students, in particular, are encouraged to use computational models to develop hypotheses and analyze the data.

References:

• Ting, C. C., & Gluth, S. (2025). High overall values mitigate gaze-related effects in perceptual and preferential choices. Journal of Experimental Psychology: General.
• Shevlin, B. R., Smith, S. M., Hausfeld, J., & Krajbich, I. (2022). High-value decisions are fast and accurate, inconsistent with diminishing value sensitivity. Proceedings of the National Academy of Sciences, 119(6), e2101508119.
• Sepulveda, P., Usher, M., Davies, N., Benson, A. A., Ortoleva, P., & De Martino, B. (2020). Visual attention modulates the integration of goal-relevant evidence and not value. elife, 9, e60705.
• Frömer, R., Dean Wolf, C. K., & Shenhav, A. (2019). Goal congruency dominates reward value in accounting for behavioral and neural correlates of value-based decision-making. Nature communications, 10(1), 4926.

(with Chih-Chung Ting)

Confidence plays a critical role in shaping our strategies for processing information and belief updating. For instance, individuals tend to search for less information when they feel more confident about making a correct decision (Desender et al., 2019). However, confidence can be biased by the anticipated outcome (e.g., rewards vs. losses). A growing body of research has shown that confidence is generally higher when learning to gain rewards compared to avoiding losses, despite similar learning performance in both contexts (Lebreton et al., 2019; Ting et al., 2023). This raises two questions: (1) Whether fixation pattern is also biased by the valence of outcome? (2) To what extent does valence-induced confidence bias can explain the variance of fixation patterns during outcome evaluation?

This project aims to investigate the relationship between fixation patterns, task performance, and confidence across different learning contexts. To address these questions, we will use a learning task in which potential outcomes are manipulated as rewards or losses, paired with eye-movement measurement (or manipulation). Both bachelor and master studentsstudents are welcome to join this project, and will be expected to assist with setting up the experiment, collecting data, and analyzing the results.

References:

• Desender, K., Murphy, P. R., Boldt, A., Verguts, T., & Yeung, N. (2019). A post-decisional neural marker of confidence predicts information-seeking in decision-making. Journal of Neuroscience, 17(2620), 3309–3319.
  https://doi.org/10.1101/433276
• Lebreton, M., Bacily, K., Palminteri, S., & Engelmann, J. B. (2019). Contextual influence on confidence judgments in human reinforcement learning. PLoS computational biology, 15(4), e1006973.
• Ting, C. C., Salem-Garcia, N., Palminteri, S., Engelmann, J. B., & Lebreton, M. (2023). Neural and computational underpinnings of biased confidence in human reinforcement learning. Nature Communications, 14(1), 6896.

(with Maryam Tohidimoghaddam)

Everyday decisions can be grouped into two types: within-domain choices, where options share common characteristics (e.g., choosing between two vacation destinations), and across-domain choices, where options are fundamentally different and lack a basis for direct comparison (e.g., celebrating your birthday with friends vs. going on vacation).

Common decision theories posit that decisions are made by first assigning subjective values to different options and then selecting the option with the highest subjective value (Levy & Glimcher, 2012; Vlaev et al., 2011). Because options vary in attribute and domain (within vs. across), the brain must represent the value of different types of options on a common scale for comparison and choice (Perkins et al., 2024). However, opposing theories suggest that the representation of subjective value in the brain may be unnecessary (Hayden & Niv, 2021; Walasek & Brown, 2023). Previous research has proposed that subjective value is represented in the ventromedial prefrontal cortex (vmPFC) (Bartra et al., 2013; Padoa-Schioppa, 2011) and the orbitofrontal cortex (OFC). In this study, we first aim to investigate the value representation in within-domain and across-domain choices using the fMRI technique. We will simultaneously record gaze positions and behavioral signatures to further investigate the information search strategy in within-domain vs. across-domain choices.

Master and also Bachelor students can be involved in this project for their theses, the only difference being the expected complexity of the output. For your thesis, you will be expected to formulate hypotheses about the behavior of the participants and analyze behavioral and eye-tracking data that has already been collected. For a master's thesis, you can also analyze fMRI data that has already been collected as well. Optionally, if you are interested, you may also apply existing computational models to the data.

Suggested reading:

• Levy, D. J., & Glimcher, P. W. (2012). The root of all value: A neural common currency for choice. Current Opinion in Neurobiology, 22(6), 1027–1038. 
  https://doi.org/10.1016/j.conb.2012.06.001
• Hayden, B. Y., & Niv, Y. (2021). The case against economic values in the orbitofrontal cortex (or anywhere else in the brain). Behavioral Neuroscience, 135(2), 192–201. 
  https://doi.org/10.1037/bne0000448