Välkomna till ett seminarium med Mattias Forsgren, Uppsala universitet.
Titel
Naturally formed preferences are transitive
Abstract
The APA Dictionary of Psychology states that intransitive relationships 2appear to be illogical and inconsistent but are often found in matters of personal preference2. This received view seems to be built on studies demonstrating intransitive preferences for so-called monetary gambles. Recent work has criticised the statistical methods in those articles but persisted in employing monetary gambles. Such novel, artificial stimuli scarcely exist outside of the laboratory and are not representative of the objects people encounter in their everyday lives. To make general, externally valid statements, we should instead sample representative real-world options (``representative design''). Across two large-sample experiments, we find that while a small number of participants show intransitive preferences for monetary gambles, we never find evidence of intransitivity for any participant across ten categories of everyday objects. We suggest a theory where preferences for such objects are formed naturally through experience, as opposed to constructed through ad hoc weighting of attributes, and corrupted by Fechnerian noise. Present results are suggestive of such a theory. While not denying that people can be placed in situations where they have to construct preferences, which occasionally will violate transitivity, we argue that in many – perhaps most – situations people have naturally formed preferences that are well-informed and transitive. The conclusion that preferences are “often” intransitive appears to be driven by an exaggerated focus on a particular kind of artificial stimulus for which people lack such well-informed preferences.
För de som var på SweCog: på seminariet kommer Mattias visa upp nya analyser och fynd utöver de som presenterades då.