What were the results of Baddeley's coding experiment?
When recalling after short periods of time, participants tended to make acoustic errors (e.g. 'fountain' instead of 'mountain'). This suggests that we code the words based on sound for STM.
When recalling after long periods of time, participants tended to make semantic errors (e.g. 'hill' instead of 'mountain'). This suggests that we code the words based on their meaning for LTM.