Passage IV

Scientists investigated whether sleep improves procedural memory (memory of skilled movements).

Subjects trained to tap their fingers in 1 or 2 sequences: Sequence X only, or Sequence X followed by Sequence Y (see Figure 1). To train for a sequence, the subjects attempted to tap the sequence for 30 sec and then rested for 30 sec, 12 times in succession. During the last 3 of these 12 cycles, the subjects' accuracy was tested.

The subjects' accuracy was retested—either immediately after training or after a full night's sleep—as they attempted to tap the sequence during 3 cycles.

Six groups, each consisting of 15 people, were subjected to different protocols:

Group 1 trained for Seq X and was immediately retested.

Group 2 trained for Seq X, slept overnight, and was retested the next day.

Group 3 trained for Seq X, immediately trained for Seq Y and was immediately retested.

Group 4 trained for Seq X, immediately trained for Seg Y slept overnight, and was retested the next day.

Group 5 trained for Seq X, waited for 6 hr, trained for Seq Y, slept overnight, and was retested the next day.

Group 6 trained for Seq X, waited for 9 hr, trained for Seq Y, slept overnight, and was retested the next day.

The table shows, for each group, the percent change in the average accuracy of the subjects between their initial testing (during training) and their retesting.