A table to show the results of all the groups, including variations of our method. Discussion The descriptive statistics would suggest that the self generation increased recall ability. The experimental hypothesis was accepted. The results were very significant when tested which would indicate that the self generation of words, or perhaps any kind of increased concentration on the material will improve memory for it. This is similar to the concept of mnemonics where mental imagery is used to increase learning. This can include organising the information and benefiting from retrieval cues which also increase learning and prevent cue-dependent forgetting.
Meaning is imposed on the material just like it has been in this investigation. However mnemonics do not really contribute to improving complex learning. Similarly, the results of this study have little meaning for important and complex memory tasks. The task presented to the participants was fairly simple and so was not very ecologically valid. There is little chance that someone would have to carry out that task in real life, but our results have been valuable in showing that, for example, a shopping list of around 11 or 12 items could be remembered using a rhyme or rhyming words that match even if interference occurs between the time of learning and the shop.
The study was also carried out in an artificial setting in quiet surrounding which, while reducing the confounding variables to the experiment, has also reduced the ecological validity. In further experiments it might be a good idea to try it in a naturalistic setting to see if memory is better or poorer. However, naturalistic settings have their limitations too
Another problem is that we included no measure of how long after the task the material was remembered. It is possible that the participants could have remembered the words for a long time after e.g. days or months, which would have been very6 interesting and had more implications for our study. This could be useful in a further replication. Another strength of this investigation was that counterbalancing was used to reduce order effects. Subjects were given the generated list and the read list in alternate orders. It should be brought to our attention that other confounding variables may have affected the accuracy of the results, such as subject variables and demand characteristics.
This type of memory experiment is fairly common in psychology and participants may have deliberately changed their memory in accordance with what results were expected. The study was also not ecologically representative. Opportunity samples are generally more representative than other sampling methods, but mostly students were used. Students are not representative of the population and there is evidence that they are not good participants. It is likely that the primary and recency effects have occurred in this study and also in Slamecka and Graf’s.
The study could become more reliable and be improved by using more people (and a more representative sample) and perhaps by using a larger variation of words (different lists) and then calculating the mean. It is possible that the words we used we easier to remember than others. It could also be made more real to life. The study shows that memory does depend on processes at the time of learning and that if we are able to generate the material we need to remember more then it is more likely to be stored for longer or transferred to the long term memory.