Baddeley (1966a)
AIM: To examine the coding of the STM store.
Â
METHOD: Baddeley conducted a series of three experiments. As all three of his experiments focused on the coding of the STM store and showed similar results, this summary will focus on experiment one. Baddeley used an independent groups design, where both conditions attempted to recall 24 sequences of five words. A total of 41 participants took part (20 in Condition A and 21 in Condition B). In Condition A, the 12 sequences were acoustically similar (e.g. mad, man, mat, cap, can, etc.) and 12 sequences were dissimilar (e.g. cow, day, bar, few, pen, etc.). In Condition B, 12 sequences were semantically similar (e.g. big, long, broad, great, large, etc.) and 12 were semantically dissimilar (e.g. old, deep, late, safe, hot, etc.).
Â
RESULTS: Table 2: The mean percentage of sequences correctly reproduced for acoustically similar/dissimilar and semantically similar/dissimilar.

CONCLUSION: The results clearly suggest that the STM store codes acoustically. Words which are acoustically similar are poorly recalled in comparison to words which are acoustically dissimilar or semantic in nature. When participants rehearse information in their STM store, they do so by acoustically repeating the word sequences in their head; the similarity makes this more difficult. The same effect is not seen in the other conditions, as the STM store is not confused by the acoustically dissimilar or semantically similar/different words in the same way.