Baddeley

• Baddeley - AIM

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

  • 72 men and women recruited from Cambridge Applied Psychology research unit. P's divided into 4 groups​
  • Group A - 10 acoustically similar ​
  • Group B - 10 acoustically dissimilar ​
  • Group C - 10 semantically similar ​
  • Group D - 10 semantically dissimilar ​
  • Words presented on projector 1 word every 3s. After, P's did 6 distractor tasks (recalling digit sequences) then asked to recall by listing words in correct order​
  • Repeated 3 times, then given 15min interference task (copying sequences of digits)​
  • Surprise retest on recall of order​
• Baddeley – FINDINGS​

  • At first, harder to remember acoustically similar than dissimilar​
  • At end, easier to remember semantically similar than dissimilar
• Baddeley – CONCLUSION​

  • STM primarily encodes info acoustically​
  • LTM primarily encodes info semantically​
  • Results support MSM
• Baddeley :) Standardised Procedure


  highly controlled lab, ppants had same number of words, presented with word every 3s, all had distractor tasks, all repeated 3x. T/f reliable as can easily be replicated to find similar results about...
• Baddeley :) Practical applications


   found LTM encodes semantically, can be used in teaching to attach meanings to theories so students do better. T/f leads to better memory so do better in exams​
• Baddeley :( Demand Characteristics


  highly controlled lab, ppants couldve guessed aim was to...so may have changed their behaviour. T/f cannot establish cause and effect, lacks internal validity​
• Baddeley :( Population validity


  opportunity sample of 72 men and women from Cambridge Applied Psych Research Unit, so is ethnocentric as all ppants were from Cambridge. T/f sample is unrepresentitive and results on...cannot be generalised to a wider population
• Baddeley - Improvements

  • Make task more realistic, present info in lesson setting, after 1h give test then another 1wk later​
  • Increase pop validity, using sample of 300 from a range of backgrounds