Poster 33.pptx (1.141Mb)

Download Report

Transcript Poster 33.pptx (1.141Mb)

Clara Bergene, Olga F. Lazareva (Drake University), Regina Paxton Gazes (Zoo Atlanta), Robert Hampton (Emory University)
Background
o Transitive inference is the ability to deduce that if A > B,
and B > C, then A > C
Results: Sequential backward training
Results: List Linking
(averaged across A…G list and H…M list)
o Example of deductive reasoning
o In nonverbal tasks, can be explained by appealing
to reinforcement history
o Reinforcement-based models (Wynne, Siemann-Delius)
have been used to accurately predict pigeons’
performance on the tasks
o Less is known about the ability of reinforcement-based
model to predict primate data
Our goal
o Test predictive ability of Siemann-Delius and Wynne
models using primate data for seven-item series with
backward training and list linking training
Training details
o Data collected at
Hampton’s lab (Emory
University)
o Captures difference between the end-anchor and the inner pairs
o Captures the difference between end-anchor pairs and inner pairs
o Predicts opposite to symbolic distance effect
o Correctly predicts between-list pairs and within-list pairs for List 1 (A…G)
o Possibly due to the lack of correction trials or to the backward order of
presentation
Sequential backward training with mock correction trials
o Randomly added 1-3 correction trials after each incorrect choice and
repeated simulations
o Results similar to initial simulations
o Does not predict within-list pairs for List 2 (H…M)
o Possibly due to list linking procedure
List linking, mock List 2 training only
o Run simulations using List 2
training sequence without list
linking procedure, and only with
initial training
o 12 adult rhesus
macaques (Macaca
mulatta)
o Predicts better performance, but
still well below monkeys’
performance
o Seven-item sequential
backward training
o Low performance in List 2 cannot
be attributed to list linking
procedure alone
o F+ G-, E+ F-, D+
E-, C+ D-, B+ C-,
A+ Bo Tested using all
novel pairs
o List linking training
o F+ G-, E+ F-, D+ E-, C+ D-, B+ C-, A+ Bo M+ N-, L+ M-, K+ L-, J+ K-, I+ J-, H+ Io Linking pair: G+ Ho Tested using within-list pairs (e.g., BD or IM) and
between-list pairs (e.g., BI)
Conclusions
Sequential backward training in mock forward order
o Coded the data in reversed order (F+ G- as A+ B-, and so on) and repeated
simulations
o Does not predict any difference between end-anchor pairs and inner pairs
o Correctly predicts the direction of symbolic distance effect
o Backward order of presentation in primates appears to provide a good control
for reinforcement history
Simulation details
o Used both Wynne model and Siemann-Delius model (only
Siemann-Delius presented)
o Associative models can fit sequential backward training data obtained with
pigeons, but not with primates
o Possibly due to quicker learning in primates
o In list linking, the models provide a good fit for between-list pairs that were
assumed to be challenging
o They do not, however, predict performance in within-list pairs for List 2 (H…M)
o This failure cannot be attributed solely to list linking procedure
o Pattern of results suggests contribution from associative values augmented by
another process
o Selected a combination of free parameters that produced
best fit for training data
o Used these parameters and accrued associative values to
predict testing performance on subject-by-subject basis
Acknowledgments
This research was supported by the A&S Drake University Faculty Development Research grant to OFL