I didn't say I was an expert. I said I did a dissertation on memory (academic). I am quite aware that eye witness accounts can be faulty. These types of experiments tend to revolve around how many people spotted the gorilla whilst they counted a basketball being bounced around by a team, or similar party tricks. Also key are events that happen towards the beginning or a sequence or the end. HOWEVER, when it comes to traumatic incidents, actually your memory is likely to be OVER-vivid coming to hit you as 'flashbacks'. This is seen in post-traumatic stress disorder. For example, soldiers from war zones, whenever they hear a loud noise may trigger unpleasant memories of gunshot in the war zone. If an event has an emotional impact on a survivor (as a rapidly sinking ship will and seeing your fellow passengers vanish before your eyes, presumed drowned, then you are likely to remember events clearly, as though in slow motion. It is nonsense to claim the passenger survivors' memories might be poor when the reverse is probably true.