Research Interests
Primary: Economics of Artificial Intelligence and Robotics, Gender and Gendered Language
Secondary: Experimental Methodology, Networks, Contest
Publications
How Strength Asymmetries Shape Multi-Sided Conflicts, joint with Sebastiàn Cortes-Corrales, Economic Theory, 2024
Our older working paper version, including a proof showing that there exist no closed-form solutions to the model for non-trivial graphs, can be found here: Generalising Conflict Networks
Reproducibility in Management Science, by Fišar, M., Greiner, B., Huber, C., Katok, E., Ozkes, A., and the Management Science Reproducibility Collaboration, Management Science, 70(3), pp. 1343-2022, 2024, Note: Contributed as a Member of the Management Science Reproducibility Collaboration.
Prosocial behavior among human workers in robot-augmented production teams—A field-in-the-lab experiment, joint with Benedikt Renner and Louis Schäfer, Frontiers in Behavioral Economics (Section Culture and Ethics), 2, 1220563, 2023, preregistration
Competition and moral behavior: A meta-analysis of forty-five crowd-sourced experimental designs, joint with Christoph Huber, Anna Dreber, Jürgen Huber, Magnus Johannesson, Michael Kirchler, Utz Weitzel, Felix Holzmeister and others, Proceedings of the National Academy of Science (PNAS), 120(23), e2215572120, 2023
Working Papers
The Gender of Opportunity: How Gendered Job Titles Affect Job Seeker Attraction joint with Petra Nieken and Martin Trenkle (Job Market Paper, soon to be submitted)
Labor shortages have intensified employers’ efforts to attract more applicants. We therefore study how the use of gender-fair titles in job advertisements affects application behavior. Our analysis combines evidence from three complementary studies. The core of our project is a large-scale randomized controlled trial conducted in cooperation with a major online job platform in Germany. On this platform, we experimentally varied whether job titles used the generic masculine or a gender-fair formulation across the three job categories IT & Development (ID), Business & Management (BM), and Marketing & Sales (MS). We find significant positive effects on the number of applications and views of job ads in BM and for ID positions advertised by non-IT firms. In contrast, we find no consistent effect in MS. When pooling across all job categories, gender-fair titles do not significantly affect the total number of applications or the number of female applicants. To understand the underlying mechanisms, we complement the field experiment with an online study of hiring experts and a laboratory experiment with job seekers. The hiring expert study shows that the choice of gender-fair titles is systematically associated with support for gender diversity, suggesting an actual signaling value of gender-fair titles for job seekers. The lab study, which incorporates eye-tracking, allows us to investigate how job seekers process and respond to gender-fair titles. Our studies shed light on how context moderates gender segregation in the labor market and how a virtually costless intervention–changing three characters in the title–can affect job seeker attraction.
Adaptivity and Revealed Robot Aversion in Human-Robot Collaboration: A Field-in-the-Lab Experiment joint with Louis Schäfer (submitted)
We study human-robot collaboration in a controlled experiment run in a realistic production environment. Participants completed a sequential task in pairs, where one worker (Worker 1 ) decided whether to pass intermediate components to a coworker or not. Depending on the treatment, the coworker was either another human participant or a physical industrial robot. The coworker-setup was either static or adaptive, with adaptive coworkers’ productivity being influenced by Worker 1’s performance in the task. We find strong evidence of robot aversion: workers were significantly less likely to pass intermediate products to their coworkers in the robotic as compared to the human treatments. This was despite overall productivity was identical across treatments. In a subsequent responsibility attribution task, participants also attributed greater responsibility to the robots, indicating a systematic bias in social evaluation of machine coworkers. Adaptivity only marginally affected these outcomes. Our results demonstrate that cooperation and responsibility attribution in hybrid teams depend not only on performance but also on social perceptions of artificial agents, highlighting behavioural frictions that may constrain the effective integration of robots into human work environments.
Predicting Generosity: The Impact of Personalized AI and Human Oversight on Altruistic Behavior., joint with Eva Groos and Christina Strobel (under review)
Regulators emphasize the importance of human autonomy and oversight in AI-assisted decision-making. Since predictions are the foundation of AI tools, a key question is how personalized predictions affect individual choices. We study this in the context of altruistic decisions, relevant to donations and charitable giving, where AI increasingly suggests charities and amounts. Using data from dictator games, we trained an AI tool to generate personalized predictions of dictators’ choices. In one treatment, dictators received predictions before deciding; in another, a “human overseer” decided whether to share the prediction; the baseline received none. We find that participants sent less (≈ 4pp) when they received predictions, with the strongest reduction when the overseer withheld the prediction. These findings highlight the need to consider human reactions to AI predictions when assessing both the accuracy and impact of such tools, as well as the unintended consequences of human oversight.
Gendered Language, Economic Behavior, and Norm Compliance joint with Petra Nieken and Karoline Ströhlein (under review, draft available upon request)
We conducted a controlled experiment to examine how the gender frame of instructions—male, female, or gender-inclusive—along with the presence or absence of a prescriptive norm (norm salience) influences norm compliance. Participants played three standard two-player economic games focusing on sharing, cooperation, and honesty. Whereas we find no clear ordering in norm compliance across our male, gender-inclusive, and female gender frames, we do find a statistically significant negative effect of the gender-inclusive frame on norm compliance for almost all participants and a statistically significant negative effect of the male gender frame on men's norm compliance. Additionally, norm salience does not appear to increase norm compliance. These findings contribute to the debate on gender-inclusive language by highlighting its unintended behavioral effects.
Previously, this project circulated as two separate papers, The Effects of Gendered Language on Norm Compliance and He, She, They? The Impact of Gendered Language on Economic Behavior.
Feedback in the Factory—A Novel Field in-the-Lab Experiment joint with Petra Nieken and Karoline Ströhlein (soon to be submitted, draft available upon request)
We study how the mode of information provision affects productivity in a realistic production environment. In a field-in-the-lab experiment conducted in a learning factory, participants perform a multitasking real-effort task that requires assembling two product variants and repeatedly deciding when to retool a machine. We vary whether operational information is unavailable (Baseline), pushed automatically at fixed intervals (Push), or pulled on demand (Pull). Information affects output through its impact on the timing of variant switches: while treatments do not directly change retooling frequency, they alter how workers align retooling with input availability. Pull information increases output for participants who retool infrequently, whereas the effect reverses for frequent retoolers. Overall, Pull outperforms Baseline, while Push shows no systematic advantage. The results highlight that autonomy in information timing can increase output when workers use information strategically, but may decrease output when task- switching is frequent.
We analyse the correlation between job satisfaction and automatability - the degree to which an occupation can be or is at risk of being replaced by computerised equipment. Using multiple survey datasets matched with various measures of automatability from the literature, we find a negative and statistically significant correlation that is robust to controlling for worker and job characteristics. Depending on the dataset, a one standard deviation increase in automatability leads to a drop in job satisfaction of about 0.64% to 2.61% for the average worker. Unlike other studies, we provide evidence that it is not the fear of losing the job that mainly drives this result, but the fact that monotonicity and low perceived meaning of the job drive both automatability and low job satisfaction.
Conference Proceedings (peer-reviewed)
Seeing Is Feeling: Emotional Cues in Others’ Heart Rate Visualizations joint with Anke Greif-Winzrieth, Verena Dorner, Fabian Wuest, and Christof Weinhardt
Learning Factory Labs as Field-in-the-Lab Environment – An Experimental Concept for Human-Centred Production Research joint with Magnus Kandler, Louis Schäfer, Gisela Lanza, Petra Nieken, and Karoline Ströhlein
Decision Experiments in the Learning Factory: A Proof of Concept joint with Karoline Ströhlein, Magnus Kandler, Petra Nieken, Louis Schäfer, and Gisela Lanza
Human-Oriented Design of Andon-Boards 4.0 – Promoting Decentralized Decisions on the Shopfloor and Acceptance by Employees joint with Magnus Kandler, Karoline Ströhlein, Sebastian Riedinger, Petra Nieken, and Gisela Lanza
Work in Progress
Assessing the Human Premium: Task Allocation Preferences in a Hybrid Workforce, joint with Aleksandr Alekseev and Mikhail Anufriev (draft in preparation)
Predict, Advise, or Perform?—The Role of Automated Systems in Human Interaction, joint with Emike Nasamu and Mengjie Wang (draft in preparation)
FrISB-BE — A Framework for Integrating Sensordata and Biosignals in Behavioral Experiments, joint with Fabian Wüst, Anke Greif-Winzrieth, Petra Nieken, and Niklas Busse (Beta-Version and first conceptual paper soon to be released)
Do Personalized AI Predictions Change Subsequent Decision-Outcomes? Artificial versus Swarm Intelligence, joint with Christina Strobel (data gathering)
Audience Effects in the Overconfidence Gender Gap, joint with Kevin Grubiak
Exploring Distributional Biases in Responses of Generative AI, joint with Petra Nieken, Abdolkarim Sadrieh, and Frederic Sadrieh
Ongoing contributions to crowd-sourced projects