Retention Interval Titration

An adaptive staircase links each trial’s next retention interval to the preceding response. Correct responding can support lengthening the delay, whereas an incorrect response can prompt shortening it, according to the experimenter’s predefined rule. Repeating these adjustments focuses testing around the individual’s target performance level and helps estimate memory persistence without relying on one fixed delay.

The performance criterion establishes the target level of accuracy that the participant must maintain while delays change. It gives the experimenter a consistent rule for deciding whether the retention interval should become longer or shorter. Because every condition is evaluated against the same target, the resulting interval estimate can support meaningful comparisons across learning conditions or experimental variables.

The longest interval maintained at the target accuracy provides an estimate of how persistently learned information supports later responding. A longer qualifying interval indicates that the individual can meet the performance criterion after more time has passed, whereas a shorter interval reflects a lower measured persistence under the tested conditions. The estimate therefore summarizes delayed performance rather than initial learning alone.

Reinforcement and training history are experimental variables that may influence how well responding is maintained across increasing delays. Retention Interval Titration makes those effects visible by comparing the intervals reached under different learning experiences or conditions. If the performance criterion is maintained for different lengths of time, the contrast can indicate that prior training or reinforcement affected memory persistence or delayed responding.

The sequence begins with an initial learning event, followed by a later test after a selected retention interval. The experimenter records whether the response is correct or incorrect, then applies the predefined adjustment rule to lengthen or shorten the next interval. Repeated trials continue until the procedure provides an estimate of the longest interval associated with the target accuracy.

This method is useful when researchers need to quantify how long information continues to support accurate behavior after learning. It can compare memory persistence across learning conditions and examine the effects of reinforcement, training history, or other experimental variables. Because the outcome is tied to delayed responding, the procedure connects learning experiences with measurable changes in behavior over time.