StunningArt/ShutterstockPsychologists have intensively studied the factors that make both eyewitnesses and victims more or less susceptible to memory distortion. But to date there has been no experimental evidence comparing memory suggestibility between the complainants and accused in sexual assault cases.My recent study was the first to compare memory errors between complainants and the accused. The findings of this study, which examined the memories of people embroiled in a fictional sexual assault case, suggest that both parties are equally likely to misremember the details of what happened. Imagine that you have been called to serve on a jury, evaluating a case of alleged sexual assault. In this case, David and Rebecca are university classmates and went to a party one Saturday evening. Towards the end of the night, they went to an upstairs bedroom, where sexual activity took place. The next day, Rebecca went to police to lodge a complaint of sexual assault, stating that she did not consent to sex with David.In this trial, as is often the case in real court cases, the accused, David, contends that the sexual activity was entirely consensual. As a juror, this puts you in a difficult position. Both parties agree that sexual activity took place, so any physical evidence is probably going to be of limited use. To make your decision, you must rely on the testimony of David and Rebecca, based on their recollections of the evening. This means that you must evaluate the credibility of David and Rebecca’s memories.When a court case hinges on the memories of victims or eyewitnesses, expert witnesses are sometimes called to explain the science of memory to the jury. The issue is understudied, but evidence suggests that in sexual assault cases, these experts are almost always called by the defence rather than the prosecution. In the case I described above, this means that the expert testimony would be used to argue that Rebecca’s – but not David’s – memory might be distorted. Research into eyewitness memory literature is largely motivated by a desire to reduce eyewitness errors and avoid miscarriages of justice. As a direct result, the majority of evidence that the expert can rely upon in giving their testimony will have focused on memory distortion among witnesses and victims of crimes. This can give the impression that witnesses and complainants are particularly prone to memory errors, while the memory of the complainant is infallible.People who are accused of crimes are human too, and their memories are subject to the same reconstructive processes as anyone else. In response to this issue, my colleagues and I conducted a series of experiments – recently published in Scientific Reports – which showed that both parties in a “he said, she said” case are equally likely to suffer from memory distortion.In our study, participants were invited to imagine that they were going on a date with either a man or a woman. They then watched a video of scenes from the date, filmed from a first person perspective. After the video, participants were told that an accusation of sexual assault had been made, and were randomly assigned to the role of the complainant or the accused. Next, they were shown witness statements from a security guard, bartender and taxi driver that included some misleading descriptions of the date. For example, the statement suggested that the accused was plying the complainant with drinks, or that the complainant was sexually aggressive. Over three experiments, we found that the “accused” and the “complainant” were equally likely to incorporate these misleading details into their memory of the date.A lot of people tend to think of remembering as a simple act of accessing information, like pulling up a computer file. But research has shown we reconstruct each memory from the ground up every time we recall it, as though we are building a Lego tower out of individual bricks, rather than recalling the event as a whole. This reconstruction can be error prone, and we sometimes incorporate misinformation into our memories, like adding a brick to our tower where it shouldn’t be. The problem arises when we expect human beings to have machine-like recollection for the details of an event, and judge them harshly when they don’t. In court, the scrutiny is often focused on the alleged sexual assault victim. Gorodenkoff/Shutterstock These errors can have devastating consequences in judicial settings. The US-based independent non-profit the Innocence Project reported in 2014 that 72% of mistaken convictions that were later overturned when DNA evidence emerged had originally relied upon faulty eyewitness testimony. But psychologists have developed techniques that interrogators can use to obtain uncontaminated eyewitness testimony. For example, interviewers can be trained to extract eyewitness reports using techniques from the cognitive interview, a technique developed by psychologists to avoid introducing post-event misinformation and distorting witnesses’ memories. In this technique, the interviewer can help witnesses recall details by asking them to form an image of the original scene (such as the location of objects in a room), to comment on their emotional reactions at the time and to describe any sounds, smells and other physical conditions. When people ask why we didn’t evolve perfect memories, the answer is the same as when we ask why we didn’t evolve to be four metres tall or have hearts that beat 300 times per second: we didn’t need to. Evolutionary pressures pushed us to stand upright and reach a height that supported our ability to feed and defend ourselves, but once that need was met, natural selection no longer favoured ever-increasing height. In the same way, our memories evolved to support our daily lives – to help us make decisions and take action – not to be an infallible recording device.When it comes to eyewitness memory, we should treat it just like any other form of evidence, recognising its value but also understanding that it can be contaminated. In the case of sexual assault, it is important to understand that the factors that might undermine a victim’s account – including the passage of time since the event, alcohol consumption and exposure to post-event misinformation are just as likely to apply to the defendant too.Ciara Greene receives funding from Research Ireland, Science Foundation Ireland, the Health Research Board of Ireland, and AXA Insurance.