This code, written for my Master's thesis at the MCMP, puts eight agents to the task of determining the reliability of an advisor via expectation-based updating on multiple rounds of testimony. It is similar to the model used by Hahn et al. (2018) & Merdes et al. (2021) in that it places these focal agents in a situation without any direct access to actual outcomes. Inspired by Duijf (2021), my model disentangles reliability into comptency and alignment and evaluates agents in terms of their trustworthiness-accuracies in addition to their (change in) veritistic values (Goldman 1999). The Boolean BH-model was introduced Bovens & Hartmann (2004), the continuous OL-model by Olsson (2011) & Angere (2010), the Boolean AL-model by Assaad (2022) & Harris et al. (2016), and in creating the SD-model I was again inspired by Duijf (2021).
- Adam Harris, Ulrike Hahn, Jens Madsen & Anne Hsu (2016) “The appeal to expert opinion: Quantitative support for a Bayesian network approach.”
- Erik J. Olsson (2011) “A simulation approach to veritistic social epistemology.”
- Sofie Angere (2010) “Knowledge in a social network.”
- Leon Assaad (2022) “Trust in Information Sources: Reconstructing Worrisome Phenomena Using Bayesian Models of Source Reliability.”
- Luc Bovens & Stephan Hartmann (2004) “Bayesian epistemology.”
- Hein Duijf (2021) “Should one trust experts?”
- Ulrike Hahn, Christoph Merdes & Momme von Sydow (2018) “How good is your evidence and how would you know?”
- Christoph Merdes, Momme von Sydow & Ulrike Hahn (2021) “Formal models of source reliability.”
- Alvin Goldman (1999) “Knowledge in a social world.”