Misleading eyewitness information can be affected by proactive interference (knowledge and expectations from before the event interfere with memory retrieval), or retroactive interference (post-event information acquired after the event can affect memory retrieval). The most prominent research conducted by Loftus & Palmer (1974), showed how leading questions can affect memory retrieval of events. Participants who were shown videos of a road traffic incident were asked how fast the car was travelling when in hit the other vehicle. However, the verb used within the leading question was changed for each condition. Participants within the 'smashed' condition reported the car to be travelling at 40.8 mph versus 34 mph with the 'hit' verb. This research highlights how leading questions can distort eyewitness testimonies by cues provided in the questions. Despite this finding, it can be argued that the research lacked ecological validity because the car accident was shown on a video clip rather than occurring in real life. This means that the emotional trauma that comes with witnessing car accidents would be absent and therefore mean participants responses were likely to be more accurate. The effect that real life events have on memory retrieval was unsupported by Yuille & Catshall (1986), who found that misleading information did not alter the memory of people who had experienced an armed robbery. Therefore, more research needs to be conducted into the effect that misleading questions can have on eyewitness testimonies. Memory retrieval is also affected by post-event information on the accuracy of eyewitness testimony as shown by Gabbert et al (2003). Participants witnessed a girl stealing money in a co-witness group or individually. The co-witness group although told they had watched the same video, had actually viewed it from different perspectives. They found that 71% of the co-witness group recalled information they had not actually seen with 60% agreeing that the girl was guilty, despite not actually seeing her commit the crime accused of. This research highlights that post-event information through discussion can reduce the accuracy of eyewitness testimonies.