For best experience please turn on javascript and use a modern browser!

Memory detection techniques can be used to test whether someone has knowledge that only the perpetrator of a particular crime would know. Numerous experiments have corroborated the reliability of the results, but to what extent can the fruits of that research be applied in practice? PhD candidate Linda Geven studied the techniques in terms of their external validity. She looked at the impact of factors that differ in an artificial research situation as opposed to a realistic forensic context.

Memory detection could involve asking individuals to identify the murder weapon from a list of five, multiple-choice style, e.g. knife, firearm, hammer. Each answer is displayed in succession for ten to twenty seconds. It’s fairly easy to spot when the perpetrator recognises the actual murder weapon.

Linda Geven: “His or her hands start to sweat; the heartbeat slows down and the breathing rate starts to decrease. The latter is due to inhibition, i.e. to distract attention from the moment of recognition. If you aren’t flustered by an answer, none of that will happen.” The technique is good for the innocent; the chances that an individual will have an intense reaction to each answer (even though you don’t know what they actually do know) is statistically insignificant.

Comparison with real criminals

Until now, all experiments were held under laboratory conditions and with fictitious crimes. “People are cognizant of the fact that they need to remember every last detail; that makes it difficult to compare them with actual criminals who, after all, do things they shouldn’t do of their own accord.” As part of the PhD research project, participants were given a quiz with ten questions. They were told that they would get five euros if they answered all ten questions correctly. What the participants didn’t know, however, was that the last two questions could not be answered. Half of the participants were told that they could use Google; the other half weren’t given that instruction.

The last group is the true subject of the experiment. They can be compared to criminals. If they answered the last two questions correctly, that meant they had used Google to “cheat”. Geven’s research revealed that their reactions were identical to those of the participants in the first group. This means that there was little difference with the lab situation and that the test should also work in real life. “Although the conclusion is positive, it still has certain problems. It turns out that it is quite difficult to hold back details from the public that only the perpetrator would know. Media access allows the public to learn just about every detail of a crime; that can influence the results.

Cooperation with police and prosecutors

It is important to ask for detailed information. “An innocent person might know that an automobilewas stolen, but not which brand. The perpetrator, however, certainly has that bit of information. Further research will be needed to see whether this works in practice. Geven would like to work with police and public prosecutors. “The test would need a proper framework. There would need to be details known only to the perpetrator. And the test cannot be used for every type of crime. In a rape case, for instance, both the victim and the perpetrator can indicate that a sexual act took place. The question, however, is whether or not the act was voluntary. The test won’t work in that situation because there can be no recognition of details known only to the perpetrator, as opposed to a murder or a missing person.”

Memory detection measures recognition. The test does not determine innocence or guilt. “Although the test can serve as supporting evidence in court, it is particularly useful for investigations. The results allow you to dig deeper intoan investigation or an interrogation.” Geven was particularly interested in finding a valid method for separating the innocent from the guilty. “I want us to avoid making errors. What is the best way to truly separate the innocent from the guilty? The test can help in certain cases. Cooperation with the police and public prosecutorsis essential. I would really like to see this applied in practice, but only time will tell.”

Linda Geven will be awarded her PhD on 6 September 2019 at 14:00 in the Agnietenkapel.