Abstracts 2005 - The Psychonomic Society

Friday Evening Posters 3105–3111
(3105)
Emotion, Experiential Congruity, and Content Type in Conditional
Reasoning. SHARON LEE ARMSTRONG, La Salle University—
Neutral content has been shown to produce superior validity judgments
in conditional arguments over content that is incongruent with
real-world experience. This study additionally examined the role of
the emotional content of conclusions. Forms of arguments and types
of congruity relationships were varied (e.g., psychological laws, social
rules, moral principles). Arguments with conditional premises
that were consistent with life experiences were easier to evaluate than
arguments posed in neutral terms, and difficulty increased when the
content was incongruent with experience. The nature of the incongruity
mattered; subjects were especially helped or hurt (relative to
congruity) by the moral rules. Although conclusion affect had no effect,
this does not mean that there were no emotional responses that
mediated performance. States of affairs that violate certain expectations
may evoke an emotional response that is more intense than the
response to states of affairs judged pleasant or unpleasant out of context.
(3106)
Conditional Reasoning, Conditional Probability, and the “Fast”
System. JOHN BEST, Eastern Illinois University—Dual-process theories
of reasoning suggest that conditional reasoning may be accomplished
by two systems: (1) a fast, associative system that operates
probabilistically and (2) a slower system that may be cognizant of
principles of logical necessity. When applied to conditional reasoning
(If P then Q), dual-process theorists have suggested that the fast system
may perform a conditional probability computation, converting
the statement above to a computation of the Prob (Q/P). However,
when people were invited to use conditional reasoning to draw deductions
from an array of information in which the conditional probability
of certain conclusions had been varied, the findings showed
that the variability with which people drew particular inferences was
not accounted for by variations in the actual conditional probability
of those inferences. Rather, people’s deductions were based on much
cruder measures of frequency and covariation, and such deductions
were as likely to be inaccurate as they were to be correct.
(3107)
Availability in Inductive Reasoning. PATRICK SHAFTO, Massachusetts
Institute of Technology, & JOHN D. COLEY & DAVID BALDWIN,
Northeastern University (sponsored by John D. Coley)—Previous research
presents conflicting evidence about novices’ use of context to selectively
guide induction. We propose that patterns of inference result
from the relative availability of different kinds of knowledge. We present
three experiments examining the use of taxonomic and ecological
knowledge in folk biological induction. In Experiment 1, familiarity with
taxonomic and ecological relations was not sufficient to induce selective
inferences about novel properties, diseases, and toxins shared by taxonomically
and ecologically related species; novice participants relied
disproportionately on taxonomic knowledge. In Experiment 2, novices’
judgments of similarity for ecologically, but not taxonomically, related
species were sensitive to priming. In Experiment 3, time pressure disproportionately
reduced novices’ ecological, but not taxonomic, inferences.
Together, these studies show that mere possession of relevant
knowledge is not sufficient for fostering context-sensitive inferences and
that differences in the availability of taxonomic and ecological knowledge
account for novices’ patterns of inference.
• JUDGMENT AND DECISION MAKING •
(3108)
Intuitive Judgments of Semantic Coherence Under Different Response
Deadlines. ANNETTE BOLTE, Technische Universität Braunschweig,
& THOMAS GOSCHKE, Technische Universität Dresden—
Intuition is the ability to judge stimulus properties on the basis of
information that is activated in memory but not consciously retrieved.
We investigated one central feature of intuitive judgments—namely,
103
their speed. Participants judged whether or not three clue words were
coherent, in the sense that they were weakly associated with a common
fourth concept. To restrict the time available for conscious deliberation
of possible solution words, participants had to synchronize
their judgments with a response signal appearing at different lags after
the clue words. In two experiments, participants discriminated coherent
and incoherent triads reliably better than chance even if they did
not consciously retrieve the solution word and even when the lag between
clue words and response signal was as short as 1.5 sec. Results
indicate that intuitive judgments can indeed be made very fast and
without extended conscious deliberation. Possible mechanisms underlying
intuitive judgments are discussed.
(3109)
Cheater Recognition and the Avoidance of Exploitation in Social
Exchange. DAN L. BOYLL, SONYA M. SHEFFERT, STEVEN M.
COLARELLI, & BRYAN GIBSON, Central Michigan University—
An understanding of natural selection as a gene-level phenomenon requires
that we understand group-directed behavior, such as cooperation,
as having evolved because it is individually advantageous. This
appears paradoxical in light of the possibility for ostensibly cooperative
individuals to “cheat” by exploiting others’ cooperative actions.
However, game theoretic models predict that cooperation can evolve
if cooperators selectively avoid attempting to cooperate with known
“cheaters.” Unfortunately, observed recognition biases favoring
cheaters do not speak to a human capacity for cheater avoidance.
However, these biases could nevertheless prove to have adaptive value
if they somehow serve to facilitate cheater avoidance. The present
study therefore attempted to replicate previous findings regarding two
types of cheater recognition biases, which differ importantly in their
operational definitions of “cheaters” and “noncheaters.” In addition,
the present study directly measured cheater avoidance, in order to assess
the possibility that such recognition biases could have adaptive
value.
(3110)
Age Differences in the Influence of Belief and Evidence on Contingency
Discrimination. SHARON A. MUTTER, LAURA M. STRAIN,
& LESLIE F. PLUMLEE, Western Kentucky University—This study
examined the impact of belief confirming and disconfirming evidence
on young and older adults’ discrimination of contingent and noncontingent
relationships. Participants completed contingency problems in
which prior beliefs about the relationship between two events (i.e.,
positive, negative, unrelated, none) were combined orthogonally with
new contingency evidence (i.e., positive, negative, zero). Baseline
contingency estimates were influenced only by prior belief. After receiving
new contingency evidence, older adults showed less accurate
contingency estimation, especially for negative contingencies. Younger
adults’ estimates improved when the evidence confirmed either a positive
or a negative belief, whereas older adults’ estimates improved
only when the evidence confirmed a positive belief. There were no age
differences when evidence disconfirmed a negative belief, but older
adults’ estimates were more biased when evidence disconfirmed a
positive belief. These findings suggest that older adults are both less
sensitive to negative contingencies and more influenced by prior positive
beliefs.
(3111)
Resonance-Based Retrieval and the Hick–Hyman Law. RANDALL
K. JAMIESON, McMaster University, & DOUGLAS J. K. MEWHORT,
Queen’s University—Response latency is a linear function of stimulus
uncertainty, a result known as the Hick–Hyman law. Although the
law is highly reliable, the mechanism underlying it is unknown. By
simulation, we show that a theory of retrieval based on a resonance
metaphor captures Hyman’s (1953) classic demonstration of the law.
The simulation demonstrates that complex rulelike behavior can be reproduced
by applying a resonance-based retrieval mechanism to an
instance-based record of experience.