Abstracts 2005 - The Psychonomic Society

Papers 126–131 Friday Afternoon
hand (placed in proximity to an unseen matching real hand), most observers
report that they feel tactile and/or thermal sensations from the
“touch” of the light on the rubber hand. The phenomenon occurs despite
conscious awareness that the hand is not one’s own and that
lights do not produce tactile sensations. Three experiments examine
the effects of the appearance and orientation of the false hand and the
method of occluding the real hand (a mirror is best) on cross-sensory
“incorporation.” The results emphasize the dissociation between conscious
cognitive understanding and the unconscious perceptual reality
of the body schema, for which intensity of light seems to be sufficient
to produce neural activity that is experienced as touch.
4:50–5:05 (126)
Viewpoint and Orientation Influence Picture Recognition in the
Blind. MORTON A. HELLER, Eastern Illinois University, JOHN M.
KENNEDY, University of Toronto, Scarborough, & MELISSA R. MC-
CARTHY, ASHLEY CLARK, NICOLE KAFFEL, & TARA RIDDLE,
Eastern Illinois University—Subjects felt solid geometrical forms and
matched raised-line pictures to the objects. Performance was best in
Experiment 1 for top views (83.3% correct), with shorter response latencies
than for side views, front views, or 3-D views with foreshortening.
In Experiment 2 with blind participants, matching accuracy was
not significantly affected by prior visual experience. Speed advantages
were found for top views, with 3-D views also yielding better
matching accuracy than did side views. There were no performance
advantages for pictures of objects with a constant cross-section in the
vertical axis. The objects were rotated to oblique orientations in Experiment
3. Early blind participants performed worse than the other
subjects given object rotation. Visual experience with pictures of objects
at many angles could facilitate identification at oblique orientations.
Alternative theoretical implications of the results are discussed.
5:10–5:25 (127)
Adaptation to Textured Surfaces: A Comparison of Direct and Indirect
Touch. MARK HOLLINS & FLORIAN M. LORENZ, University
of North Carolina, Chapel Hill—Participants examined and
gave magnitude estimates of the roughness of textured surfaces (arrays
of truncated pyramids, with spatial periods ranging from 124 to
1,416 microns). Test trials followed adaptation to (1) a surface with a
spatial period of 416 microns or (2) a smooth, featureless surface.
Some participants used direct touch to examine both the adapting and
the test surfaces, whereas others touched all surfaces indirectly, through
a stylus. With indirect touch, there was a significant interaction of test
surface and adapting condition: Texture adaptation lowered the roughness
of fine surfaces more than that of coarse ones. This result is consistent
with vibrotactile mediation of roughness perception during indirect
touch. Texture adaptation had no significant effect under conditions
of direct touch. We discuss the implications of these findings for theories
of roughness coding.
Information and Belief
Conference Rooms B&C, Friday Afternoon, 3:30–5:30
Chaired by David V. Budescu
University of Illinois, Urbana-Champaign
3:30–3:45 (128)
To Bayes or Not To Bayes? A Comparison of Models of Information
Aggregation. DAVID V. BUDESCU & HSIU-TING YU, University of
Illinois, Urbana-Champaign—We model the aggregation process used
by individual decision makers (DMs) who obtain probabilistic information
from multiple, possibly nonindependent sources. We distinguish
between three qualitatively different approaches: choosing a single
“best” advisor, “compromising” by averaging the advisors’ opinions,
and combining the forecasts according to Bayes’s rule. The DMs in
our studies had access to forecasts provided by J = 2 or 3 advisors who
judged multiple cases. Our data are unusually rich in many respects,
20
since the studies involve natural sampling of cues with various levels
of and types of dependence and various patterns of information overlap.
This provides an excellent opportunity to compare the quality of
these models. Overall, the DMs’ judgments are closest to the averaging
model, but clearly, they do not use one model exclusively. In particular,
their aggregates are more in line with Bayes when the advisors
are (1) highly consistent with each other and (2) provide extreme
forecasts.
3:50–4:05 (129)
The Perception of Correlation From Scatterplots. MICHAEL E.
DOHERTY, RICHARD B. ANDERSON, ANDREA M. ANGOTT, &
DALE S. KLOPFER, Bowling Green State University—Research on
the perception of correlation from scatterplots suggests that people’s
estimates of the Pearson r tend to be lower than the calculated value
and that the estimated r is a positively accelerated function of the calculated
value. This suggests that scatterplots representing high correlations
should be more discriminable than those representing low correlations,
given equal differences in calculated r. Experiment 1 used
the classical psychometric method of rank order across 21 values of
r. Subjective rank orders correlated with objective rank orders, but
confusability was greater for low than for high correlations. Experiment
2 used a yes/no task and showed that d′ was higher for pairs of
scatterplots representing high correlations, rather than for pairs representing
low correlations. Subjects in both experiments were also
asked to make an absolute judgment of correlation. For those judgments,
high correlations tended to be underestimated, but low correlations
tended to be overestimated.
4:10–4:25 (130)
Temporal and Probability Discounting of Money, Alcohol, and Food
Rewards. JOEL MYERSON, LEONARD GREEN, SARA J. ESTLE,
DANIEL D. HOLT, & TESSA MAZZOCCO, Washington University—
Using hypothetical delayed rewards, previous research has shown that
drug abusers discount their drug of abuse more steeply than money.
This could reflect specific characteristics of drugs or drug abusers or,
as Odum and Rainaud (2003) suggested, a general property of consumable
rewards. To test their suggestion, we compared temporal and
probability discounting of a nonconsumable reward (money) and three
consumable rewards (candy, soda, beer). With delayed outcomes,
money was discounted less steeply than the consumable rewards,
which were discounted at similar rates. With probabilistic outcomes,
there was no effect of reward type on discounting rate. Larger probabilistic
amounts were discounted more steeply than smaller amounts,
regardless of the type of reward. Larger delayed monetary amounts
were discounted less steeply than smaller monetary amounts, but no
such magnitude effect was observed with the consumable rewards, potentially
explaining the absence of magnitude effects in studies of
temporal discounting by nonhuman animals.
4:30–4:45 (131)
Further Tests of an Information Leakage Account of Attribute
Framing Effects. CRAIG R. MCKENZIE, University of California,
San Diego—When Levin and Gaeth (1988) described ground beef as
“75% lean,” rather than as “25% fat,” participants rated it as superior
on various dimensions. Evidence is presented that supports an “information
leakage” account of this well-known framing effect: Speakers
were more likely to describe ground beef as being “75% lean”
(rather than “25% fat”) when it was believed to be relatively lean than
when it was believed to be relative fatty. Furthermore, listeners presented
with the “75% lean” frame were more likely to infer that the
beef was relatively lean. A speaker’s choice of frame leaks relevant information
about whether the speaker believes that the beef is lean or
fatty, and listeners “absorb” this information. Also of interest is that
listeners were sensitive to the source of the frame: The absorption effect
almost disappeared when the frames were presented in the context
of an advertisement (rather than a conversation).