We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

econ.GN

Change to browse by:

References & Citations

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Economics > General Economics

Title: Overinference from Weak Signals, Underinference from Strong Signals, and the Psychophysics of Interpreting Information

Abstract: Numerous experiments have found that when people receive signals that would lead a Bayesian to substantially revise beliefs, they underinfer. This paper experimentally considers inference from a wider range of signal strengths and finds that subjects overinfer when signals are sufficiently weak. A model of cognitive imprecision adapted to study misperceptions about signal strength explains the data well. As the theory predicts, subjects with more experience, greater cognitive sophistication, and lower variance in answers, infer less from weak signals and infer more from strong signals. The results also relate misperceptions to demand for information and inference from multiple signals.
Subjects: General Economics (econ.GN)
Cite as: arXiv:2109.09871 [econ.GN]
  (or arXiv:2109.09871v1 [econ.GN] for this version)

Submission history

From: Michael Thaler [view email]
[v1] Mon, 20 Sep 2021 22:33:20 GMT (6847kb,D)

Link back to: arXiv, form interface, contact.