explanation/evidence for stores being separate in terms of encoding, capacity and duration
- found after learning 4 sets of words the accoustically sim set had worse recall if recall was immediate but sem sim set had worse recall if there was a delay before recall
immediate recall would be testing words still held in STM that hadn't decayed, acoustically sim words were confused = coding in STM is acoustic
whereas after a delay LTM is being tested
sem sim words were confused = LTM encoded semantically
two stores encode differently = supports MSM concept of diff encoding in diff stores