Assessing Models of Subjective Probability Judgment.
Barch, Daniel.
2015
-
Abstract: Across domains, individuals tend to make judgments and choices about
probability that are apparently incongruent with the information they are given. This
dissertation examines both the precise nature of these departures from veridical
probability estimation and situations where individuals demonstrate accurate judgments.
Experiments 1 and 2 apply a method of model selection - assessing ... read morethe logarithmic
derivatives of competing models of risky choices - that is novel to the field of judgment
and decision-making and also introduce a new candidate model to the literature. Several
candidate models for probability judgment in risky choice are rejected, and two models are
shown to be superior. Experiments 3-6 assess memory for probability judgments in cases
where all of the information needed to make such judgments is presented at once and risk is
not an issue. This set of experiments examines memory for both simple probability judgments
(in which individuals are asked about one feature of a problem) and for conjunction
probability judgments (in which individuals are asked about multiple features of a
problem). Riskless probability judgments are remarkably accurate when mnemonic interference
is minimal. As interference increases, patterns of misestimation emerge that likely result
from a mixture of guesses and confident judgments, and this pattern is better fit with a
linear model than with a high-performing model of risky weighting.
Thesis (Ph.D.)--Tufts University, 2015.
Submitted to the Dept. of Psychology.
Advisor: Richard Chechile.
Committee: Ayanna Thomas, Raymond Nickerson, and Marc Howard.
Keywords: Psychology, and Cognitive psychology.read less - ID:
- rr172894m
- Component ID:
- tufts:21353
- To Cite:
- TARC Citation Guide EndNote
