Human-AI collaboration and decision fatigue
Requirements
To conduct this thesis you need experience in these areas:
- programming to develop a tool to test (prefered: streamlit with python)
- reading skills to dive into human-AI publications to understand your topic
Depending on the focus we choose:
- experience in creating AI models, so that you can test a "real" AI
- experience in conducting empirical research (study design, statistics etc.)
Contents
The following description serves the purpose to shortly give you an impression of what a topic regarding these key topics could be.
If you are interested in the research question or a similar one, please contact Ulrike Schäfer for more information.
Key Words:
- Human-AI collaboration
- Decision-making
- Decision fatigue
Research Question:
Does the length of interaction/amount of decision affect decision fatigue depending on having AI assistance or not?
Contribution:
- Literature research on the key words and specify research question
- Develop a scenario to test your question
- Create a test environment (application)
- Qualitative interview to understand, what specifically AI assisants changed. (You will collect more data than you have to assess in detail as this study is part of a bigger project on investigating human-AI interaction.)
Independant variable:
- Multiple groups: short and long duration and with and without AI
Dependant variable:
- Performance
- Mental Model: Interview questions to understand how the AI was perceived
Literature:
-
Decision fatigue: A conceptual analysis https://journals.sagepub.com/doi/pdf/10.1177/1359105318763510
-
Human Error, Decision Making and Fatigue https://www.dcs.gla.ac.uk/~johnson/teaching/safety/powerpoint/16_Human_error.pdf
-
Human Performance Consequences of Automated Decision Aids in States of Fatigue https://journals.sagepub.com/doi/pdf/10.1177/154193120905300435 (TU Berlin)
-
(decison making general) Gloria Phillips-Wren & Monica Adya (2020) Decision making under stress: the role of information overload, time pressure, complexity, and uncertainty, Journal of Decision Systems, 29:sup1, 213-225, DOI: 10.1080/12460125.2020.1768680