This thesis is composed of three experiments designed to verify predictions of the Discrimination-Analysis (DA) hypothesis about the effectiveness of the Many-to-One (MTO) and One-to-Many (OTM) training structures in producing equivalence classes.
In the MTO training structure, the baseline relations are interrelated by a common comparison stimulus and in the OTM, they are interrelated by a common sample. The “common” stimulus is called node.
Experiment 1 verified whether mixing baseline trials among emergent test trials would enhance the yields of emergence of three 5-member classes following the OTM training structure. Twenty adults were exposed to two conditions:
- Simultaneous Protocol with Baseline trials Intermixed (MixB)
- With Post-test of Baseline trials
Condition MixB had little (and negative) effect on the yields of immediate emergence (that is, on the number of participants who responded in accordance with equivalence).
Experiments 2 and 3 verified whether the MTO would be more effective than the OTM to produce the emergence of three 7- and 9-member classes, respectively.
Thirty-four adults participated in Experiment 2 and 42 participated in Experiment 3; half of them were exposed to the MTO and another half to the OTM. In Experiment 2, 88% of the participants passed the test for emergent relations in both groups. In Experiment 3, 86% and 48% passed the test in the MTO and OTM groups, respectively.
The second objective of Experiments 2 and 3 was to identify baseline performances predictive of success or failure in the test. In both experiments, the requirement of a great amount of trials to learn the MTO baseline was predictive of failure. In the OTM group in Experiment 3, the data showed that differences in the frequency of selection of the comparisons, that led to discrepancies in the overall number of reinforcements per node, was predictive of failure.
Keywords: training structures, Many-to-One, One-to-Many, equivalence classes
Vanessa Ayres Pereira disputerer for ph.d.-graden i atferdsvitenskap med avhandlingen "Emergence of Large Equivalence Classes as a Function of Different Training Structures".
Mandag 25. november kl 10.
Tittel: "The role of theory in behavior-analytic philosophy of science and experimental design."
Doktoranden vil offentlig forsvare avhandlingen 25. november kl 12.15.
Vi ber om at publikum finner sine plasser i god tid før disputasen begynner.
- Førsteopponent: Professor Carol Pilgrim, PhD, University of North Carolina Wilmington, Department of Psychology, USA
- Andreopponent: Associate Professor Russell W. Maguire, PhD, Department of Behavior Analysis, Simmons University, USA
- Leder av komiteen: Førsteamanuensis Børge Strømgren, Institutt for atferdsvitenskap, Fakultet for helsevitenskap, OsloMet
Leder for disputasen
Professor Ingunn Sandaker, OsloMet
- Hovedveileder: Professor Erik Arntzen, Institutt for atferdsvitenskap, Fakultet for helsevitenskap, OsloMet
- Medveileder: Førsteamanuensis Torunn Lian, Institutt for atferdsvitenskap, Fakultet for helsevitenskap, OsloMet