Of the first 40 cases overturned by new DNA evidence, 36 of the people were defendants convicted on eyewitness testimony. Many errors are made when a witness of a crime or event attempts to recall the experience. In a police interview an eyewitness may relay false information in an attempt to be as helpful as possible. A court jury will often declare a person guilty based solely on an eyewitness testimony, although many mistakes are often made through memory. Elizabeth Loftus claims .
" a confident witness, particularly one who's got details and expresses that detail in a confident manner, can be very persuasive," although the details may not be entirely correct the witness may have persuaded themselves to believe it, which is precisely why police interviews need to be structured correctly so as to gather accurate information. A witness who honestly believes he or she has identified the right person or seen a weapon, for example, can lead to a conviction because juries tend to believe that confidence equals accuracy.
After observing a crime, a person will later reconstruct it to be able to talk about it. This is one reason why eyewitness testimony is unreliable. It is not that people are knowingly stating false information but rather they have convinced themselves that this false information is true. During reconstruction, events that are witnessed are fit into the existing knowledge or schemas of a person. Bartlett's Schema Theory (1932) states there are clusters of knowledge, or schemas, which are used to fit any new information into. In other words, instead of remembering an event exactly how it was seen, it will be intermixed with existing knowledge or template and past experiences to form a reconstructed memory. Bartlett conducted many experiments investigating how people tend to remember and recall information. One investigation consisted of Bartlett reading a Native-American story to a group of English participants.