On the other hand, the study fell short of demonstrating a group effect on memory - both groups showed significant improvements both in accuracy and speed on a standard test. The apparent lack of effect, Kramer told MedPage Today, is probably a statistical artifact that results from large individual differences within the groups.