Design controversies and the generation effect: support for an item-order hypothesis

Area of Science:

• Cognitive Psychology
• Human Memory
• Learning and Cognition

Background:

• Previous research by Nairne, Riegler, and Serra (1991) suggested that item generation impairs long-term retention of serial order.
• The 'generation effect' typically shows better memory for self-generated items compared to read items, but its impact on serial order memory is debated.

Purpose of the Study:

• To investigate the disruptive effect of item generation on long-term serial order retention.
• To clarify the conflicting findings in the literature regarding the generation effect and its influence on different memory components (item vs. order).

Main Methods:

• Three experiments were conducted using reading and item generation tasks.
• List structures (pure vs. mixed) and learning procedures (incidental learning) were varied.
• Memory was assessed using reconstruction tests and free recall, with some conditions including immediate serial order reconstruction post-recall.

Main Results:

• Experiment 1: Reading showed an advantage over generation in pure lists for reconstruction, but not in mixed lists.
• Experiment 2: The standard generation advantage was observed in free recall for mixed lists, but not pure lists.
• Experiment 3: Replicated Experiment 1 findings using incidental learning, confirming the pattern observed with intentional learning.

Conclusions:

• The results support the hypothesis that generation has dissociative effects, impacting item memory and order memory differently.
• These dissociative effects provide an explanation for the inconsistencies and controversies observed in previous generation effect research.
• Generation specifically disrupts the memory for the sequence of items, while potentially preserving item memory itself.