Memory performance is enhanced when a certain task during encoding correlates with the type of task that occurs during retrieval. A transfer appropriate processing experiment varies the type of task used for encoding and retrieval.
Donald Morris conducted an experiment in 1977 which displayed this phenomenon. In this experiment, subjects were presented 32 sentences in the encoding part of the experiment. Two tasks were conducted during this encoding phase, the meaning task and the rhyming task. In the meaning task the participants heard fill in the blank sentences in which the word "blank" replaced the target word. After the sentence, subjects had to respond with "yes" if they thought it fit into the sentence and "no" if they thought it didn't. The rhyming task involved statements involving rhyming. In these sentences the "blank" replaced the target word. Following the sentence, subjects were asked to respond with "yes" or "no" if the statement was correct or not. The results of this experiment showed that subjects were able to better remember words when rhyming was used in the encoding stage. This is not a result of levels of processing theory, the key to better performance with rhyming is the transfer-appropriate processing where both encoding and retrieval are based on sound. Memory not just depends on levels of processing, but how well the conditions at encoding and retrieval match up.