Two experiments were conducted to investigate the relationship between nodal distance, response accuracy, and response latency during testing for emergent relations. In both experiments, undergraduate subjects first learned A-B, B-C, C-D, and D-E constituent relations of six 5-member equivalence classes. In Experiment 1, only selected tests of trained and of 0-, 1-, and 2-node tests of emergent relations were carried out in order to avoid testing of 0- or 1-node relations that might form constituents of the 2-node relations which were tested. In Experiment 2, all possible trained and derived relations were tested in random order. Although considerable individual variability was observed in both response times and accuracy for the 14 subjects completing Experiment 1, latencies for correct responses generally increased and response accuracy decreased as a function of nodal distance. There was no nodal distance effect for latencies of incorrect responses. In Experiment 2, these relationships between response times, response accuracy, and nodal distance were observed for 3-node relations in 5 out of 6 subjects. Analysis of error response latencies for 2 subjects who made sufficient errors revealed a nodal distance effect for 1 subject but not for the other. In both experiments, response times decreased as testing progressed, but response accuracy increased during testing only in the second experiment.
Bentall, R. P.; Jones, R. M.; and Dickins, D. W.
"Errors and Response Latencies as a Function of Nodal Distance in 5-Member Equivalence Classes,"
The Psychological Record: Vol. 49
, Article 7.
Available at: http://opensiuc.lib.siu.edu/tpr/vol49/iss1/7