Baddeley (1966b)

Aim: To investigate how information is coded in long-term memory (LTM).

Method: Baddeley conducted three experiments investigating coding in LTM. This summary focuses on the third experiment, which used interference tasks to reduce the influence of STM. Baddeley used an independent groups design in which each participant completed one of four conditions:

• 15 participants learned a list of 10 acoustically similar words

• 20 participants learned a list of 10 acoustically dissimilar words

• 16 participants learned a list of 10 semantically similar words

• 21 participants learned a list of 10 semantically dissimilar words

The participants completed four learning trials. After each presentation of the word list, they performed a short-term memory task to prevent rehearsal. They were then given one minute to write the 10 words in the correct order.

After the four trials, participants completed a self-paced digit-copying task for 15 minutes. They were then asked to recall the word sequence again. This delayed retest examined the coding of information in LTM.

Results: Baddeley found that:

• there was no significant difference in delayed recall between acoustically similar and acoustically dissimilar words

• participants recalled significantly fewer semantically similar words than semantically dissimilar words

• semantic similarity therefore had a greater effect on LTM recall than acoustic similarity

Conclusion: Baddeley concluded that information in LTM is coded semantically. Semantically similar words were more difficult to recall because their similar meanings caused confusion in LTM. Acoustically similar words did not produce the same effect, suggesting that the sound of information is less important for long-term retention.