Giving time a chance in the midsession reversal task

Area of Science:

• Behavioral neuroscience
• Animal cognition
• Learning and memory

Background:

• The midsession reversal task assesses how animals learn and adapt to changing reinforcement contingencies.
• Pigeons often appear to time the reversal point in this task, leading to errors around the expected changeover trial.

Purpose of the Study:

• To investigate the role of timing versus other strategies (counting, contingency learning) in pigeons' performance on the midsession reversal task.
• To examine the effect of manipulating the intertrial interval (ITI) on the accuracy and timing of the reversal response.

Main Methods:

• Pigeons were trained on a midsession reversal task with a constant intertrial interval (ITI).
• Test sessions involved manipulating the ITI (doubling or halving) to observe effects on reversal trial performance.
• Additional tests included conditions with constant reinforcement or delayed reinforcement of one stimulus to probe strategy use.

Main Results:

• Most pigeons' performance remained consistent with timing strategies even when the ITI was altered.
• Halving the ITI did not consistently double the reversal trial, suggesting factors beyond simple timing may be involved.
• Some individual pigeons demonstrated strategies based on counting or reinforcement contingencies rather than pure timing.

Conclusions:

• While timing is a prevalent strategy in the midsession reversal task, its dominance can be influenced by factors like ITI manipulation and individual differences.
• The observed asymmetry in ITI effects suggests that novel cues or learning strategies can interact with or override simple temporal prediction.