===
one form
Paired associates learning is a memory task where individuals learn to link pairs of unrelated items, such as shapes or words, and recall one when shown the other. Probabilistic reversal learning tests cognitive flexibility by requiring adaptation to changing reward probabilities between stimuli after initial learning.
Participants view pairs of items, like a pattern in a box, then select the correct location or associate when prompted. Errors trigger reminders, with difficulty increasing by adding more pairs. This assesses visual episodic memory and new learning, sensitive to medial temporal lobe issues.
Subjects first discriminate stimuli with different reward probabilities, then adapt when contingencies reverse, such as the high-reward option becoming low-reward. It measures behavioral flexibility, perseveration, and integration of probabilistic feedback over trials. Reinforcement learning models often reveal higher learning rates and exploration during reversals.