| Summary: | This paper provides a novel interpretation of focal point
responses (0, 50, 100 percent) in terms of ambiguous beliefs dynamics that
arise in new developments of decision theory such as Choquet expected utility
theory. In particular, focal point responses that have been updated from
nonfocal responses can be interpreted as non-additive beliefs that account for
psychological bias. A focal point response of 100 that has been updated from a
nonfocal response can be represented by a non-additive belief that has been
updated according to the Overestimating Update Rule. A focal point response of
zero that has been updated from a nonfocal response can be represented by a
non-additive belief that has been updated according to the Underestimating
Update Rule. Focal point responses given consistently over time are not subject
to psychological bias, and can be represented by additive probability
distributions. Estimation results show such a model to be a very good fit to
the data.
|
|---|