Results of Experiment 3: Is consistency the key?

I know it’s been a while since my previous lab note, but it’s been a busy summer of conference travel, teaching, and research.

But the academic term is starting up again, and amidst preparing for teaching Comparative Psychobiology and starting up some new research projects, I managed to find some time to squeeze in another lab note to reveal more of our results.

The previous lab note discussed the effect of increasing attention to the stimulus display on object encoding. In a nutshell, attention to the object led the pigeons to encode the object. But, binding went away! That’s probably not surprising because the attention manipulation reduces the conditions that would naturally promote object-place binding.

The question we address now is: Is a consistent sequence order during training necessary for binding of object-location associations?

To answer this question, we trained 6 new birds on task. Three of the pigeons received the Both Consistent training procedure used in Experiment 1. The other three pigeons also received training on stimuli for which the same object always appeared at the same location, thus, we call this treatment group the Binding condition. The difference is that the individual items were not presented in a consistent order as shown in the figure below.

 

Will keeping the same object in the same location across all trials be sufficient for the formation of an object-location association which in turn promotes binding? Or is it necessary to also have the objects and locations appear in a consistent sequence order across all presentation sequences?

The results we found provide a clear answer.

 

For subjects that received the Both Consistent training, we replicated the object binding effect with the smallest RT cost on Both probe trials, for which object-location binding was preserved. The largest RT cost occurred on All probe trials, for which the object-location binding was broken. For subjects that received Binding training only, but without a consistent object-location sequence, no evidence for binding was found.

There you have it, folks! Having the objects and locations occur in a consistent sequence was necessary for the binding of object-location associations! This may provide a clue as to why prior studies of binding in pigeons failed to find any evidence. Those prior studies failed to present the stimuli in a consistent sequence order.

In the spirit of true comparative analysis, the final lab note which should go up soon will report on a replication of Experiment 1 in humans!

Aaron Blaisdell