Partial source dependence and reliability revision: the impact of shared backgrounds
- Jens Madsen, School of Geography and the Environment, University of Oxford, Oxford, United Kingdom
- Ulrike Hahn, Centre for Cognition Computation and Modelling/Dept. off Psychological Sciences, Birkbeck, University of London, London, London, United Kingdom
- Toby Pilditch, Department of Experimental Psychology, University College London, London, United Kingdom
AbstractThe paper explores how people revise their belief in a hypothesis and the reliability of sources given independence of sources or partial dependence (e.g. the sources share a joint source). Specifically, we test a formal model of reliability revision. The study provides support for Bovens and Hartmann’s model of reliability revision. If a source provides a positive report for an unlikely hypothesis, participants initially revise the reliability of the source negatively. However, as additional positive reports emerge, they increase their estimate of the reliability of the source. Further, if it becomes known the sources are partially dependent (here, taught the same school of thought), the reliability of the source decreases again. Both of these findings are in line with the Bayesian approach to reliability revision.
Return to previous page