Baddeley (1966) and Reconstructive Memory

• Classic Study: Baddeley (1966) explored cognitive psychology.

Reconstructive Memory Theory

• Memories are reconstructed each time they are recalled, not perfectly encoded, stored, or retrieved.
• Memory reconstruction utilizes schemas to fill gaps.
• Schemas alter memory content, making it consistent with beliefs and experiences.

Baddeley (1966): Aim

• The study aimed to determine whether long-term memory (LTM) encodes information acoustically or semantically.
• Previous research indicated short-term memory (STM) encodes it acoustically, showing poorer recall for acoustically similar words.

Baddeley (1966): Procedure

• 72 volunteer participants were recruited from the Applied Psychology Research Unit subject panel at Cambridge University
• The experiment was conducted in a lab and tested how acoustically and semantically similar word lists affect sequential recall, not just word recall.
• List A contained 10 acoustically similar words (e.g., man, can, cat, map).
• List B contained 10 acoustically dissimilar words (e.g., pit, few, cow, mat).
• List C contained 10 semantically similar words (e.g., great, large, broad, big).
• List D contained 10 semantically dissimilar words (e.g., good, huge, deep, late).
• Each list was presented via projector at one word every three seconds.
• Participants completed six tasks involving memory for digits to prevent rehearsal; they wrote down eight random numbers three times.
• After each interference task, participants recalled the words in the correct order. There were a total of 4 learning trials to measure STM
• Participants were then given an interference task for 15 minutes and then had a surprise 5th LTM test, where they recalled words in sequence.

Baddeley (1966): Findings

• Acoustically similar words had a negative effect on STM, leading to poorer recall than dissimilar words in the initial learning phase across 4 trials.
• Acoustic similarity had no effect on LTM, with no significant difference in recall during a surprise retest.
• Semantic similarity had no effect on STM during the initial learning phase of 4 trials; semantically similar word recall was comparable with dissimilar word recall
• Semantic similarity had a negative effect on LTM, with a significant difference in recall between semantically similar and dissimilar words in the surprise retest.

Baddeley (1966): Conclusion

• STM encoding is largely acoustic, as acoustic similarity confused the participants during the initial learning phase.
• LTM encoding is semantic, as semantic similarity confused the participants on the retest, impairing recall.

Implications

• Social impact theory and social agency theory may be confused because semantic similarity in LTM makes them very similar.
• Well-known subjects become semantically dissimilar, reducing the likelihood of mixing them up.
• Milgram’s and Sherif’s experiments are easily recalled with less confusion because they are semantically very different in LTM.

Understanding the Findings

• Information sorted by sound gets confused when similar-sounding, complicating correct-order recall; hence, the poor recall means STM encoding is acoustic
• Information sorted by meaning gets confused when meanings are similar, complicating correct-order recall; hence, the poor recall means LTM encoding is semantic

Evaluations of Baddeley (1966)

• The study lacks mundane realism because real-life LTM does not only consist of remembering word lists in order
• Tasks such as revising, recalling conversations, or telephone numbers are more realistic.
• Memory researchers defend the study, noting that removing context simplifies memory to isolate key aspects.
• The study was scientific due to its controlled lab environment and standardized procedure, increasing reliability.
• It included word lists containing the same number of words (10) shown for 3 seconds via projector screen.
• Cause-and-effect was established between word list similarity (semantic/acoustic) and long-term memory.
• The study can be regarded as replicable as a result of the controlled elements.
• Baddeley's results can be applied to improve long-term recall in education
• Students benefit from reorganizing material and relating it to existing knowledge rather than rote rehearsal, which matches semantic encoding in LTM.
• Teachers can emphasize the key differences between concepts/theories to help students remember new content.