Baddeley 1966

Baddeley 1966

Aim

• 
• To investigate the influence of acoustic and semantic word similarity on learning and recall in short-term and long-term memory.

Procedure

• 72 Participants - men and women, recruited from the Applied Psychology Research Unit subject panel.
• Independent groups design, 4 conditions: Acoustically similar words, Acoustically dissimilar words, Semantically similar words and Semantically dissimilar words.
• The words were presented via projector at a rate of one word every 3 seconds in the correct order. After the presentation the participants completed interference tasks involving memory for digits, before having to recall the list of words in order. This was repeated over 4 trials.
• The word list, not in order, was made visible to the participants as the test was about order recall, not word recall.
• After the 4 trials, the participants completed an interference task for 15 minutes and then were given a surprise re-test of the word lists. This means the total number of trials was 4 + 1.
• The 4 initial trials showed short-term memory, whereas the re-test showed long-term memory.

Results and Conclusion

• Recall of the order or acoustically similar word lists was worse in the 4 initial trials than the dissimilar sounding words.
• Recall of the order of semantically similar word lists was worse in the re-test (after the 15 min interval task).
• Short-term memory is largely acoustic, so acoustically similar words are more difficult to encode in the short-term memory.
• Semantically similar words were more difficult to recall in the re-test, so the long-term memory is largely semantic.

Evaluation

• Strengths:
• Large sample - anomalies can be averaged out - generalisable.
• Standardised procedure - reliable.
• Baddeley improved reliability by projecting the words as hearing difficulties or speech would not have affected results. Also improved validity as it reduced the risk that some words would be harder/easier to recall because of unfamiliarity or associations.
• Supported by Working Memory Model
• The re-test simulated real life as you are not usually expecting to have to recall information - increases ecological validity.
• Weaknesses:
• After being split into 4 conditions group sizes were small (15-20) and so anomalies my have affected results.
• Ethnocentric
• Volunteer/Opportunity sample - certain type of person - university subject panel - young - intelligent.
• Low ecological validity - recalling word lists is unrealistic - artificial.
• They may have guessed the re-test because of the final interval task - reduces validity.