carmichael (1932)

reconstructive memory - info is sorted and when it is remembered it is rebuilt so can be affected by extra info and by ideas (like schemas) we might already have

aim: wanted to find out wheteher words shown with pictures would affect the way the pictures were remembered. 

• lab experiment
• independent groups design
• 95 participants shown 12 pictures (the stimulus material)
• independant variable: what word they heard
• between each picture experiementer said 'the next picture resembles.." followed by a picture.
• participants then asked to draw pictures they had seen
• drawings compared to original (dependant variable)

two lists

• the drawing produced by people who heard words from list one very different to list two
• drawigns look like words they heard

concluded memory for pictures is reconstructed and the verbal cotext affects recall because memory of the word alters the way the picture is represented

strengths:

• used control group so could be sure that not all drawings were distorted in the same way
• 2 different lists showed verbal labels affected peoples drawings
• 12 pictures and many participants gave lots of evidence to make sure findings were reliable

weaknesses:

• in real life things are not generally ambiguous as the stimulus figures shown
• prentice (1954) tested affect of verbal labels  on recognition rather than recall and found that verbal lablels didnt effect recognition, this means carmichaels findings did not apply widly