The reconstructive nature of memory leads to the idea that memory is unreliable with concepts, such as the schema theory. Bartlett (1932) carried out a study to show the impact of prior knowledge (schemas) on memory. He made the suggestion that the process of remembering things is an active reconstruction of the bits that are stored. When you learn something, it is in fact only elements of the experience that are stored. Thus meaning that reconstructions are actually made by joining the real elements of a memory with your knowledge of the world (schemas). Schema theory suggests that prior expectations, such as prejudices and stereotypes will affect what we believe we have seen and how we recall the information as a result. This view has been supported by research by Bartlett and also Allport and Postman. In Bartlett's highly regarded study (in 1932 as mentioned before), he also showed how the cultural expectations (or stereotypes) we possess consequently lead to inevitable adjustments in memory, as participant's recall distorted the content and style of the original story which was chosen to cause conflict between its contents and the reader's own knowledge of the world. This shows that stereotypes (and prejudices) can influence memory, which is the main reason that it is believed to cause unreliability in eyewitness testimony. However, how do we know that this would not happen with an English story? Surely, some parts would become distorted also and so possibly their results cannot all be thought to be due to the different culture. Other factors should also be taken into account, such as was the sample chosen a biased group? For example, the group could have been young students who would most likely not have such an extensive vocabulary or memory storage as a group of people in their mid-twenties. If this were the case, then these results would possibly not be able to be taken as a generalisation of the population as a whole, but only a particular cross-section of the population instead.