A central question in the field of cognitive psychology is the fact that the human mind constructs both false and true memories. Since the last century, researchers have found that memory can be surprisingly unreliable. Unfortunately, the research addressing false memories has been predominantly conducted on adults. In fact, researchers had essentially only began to study memory flaws in children over these past two decades. Indeed, most studies about a child’s production of spontaneous false memory still relies upon the Deese-Roediger-McDermott (DRM) paradigm, which was derived from semantic word associations made by adults. In summary, the DRM paradigm consists of individual lists of words which are semantically related to each other, and to an item, called a “critical lure” (CL) which is not present on the list. After being exposed to the word lists, the subject is requested to recall the presented words. This DRM paradigm, therefore, essentially seeks to the degree of CL during the recall phase. In other words, the recognition of CLs during the free recall or the recognition task constitutes a false memory.
In all the studies carried out on the DRM paradigm in children, there is near consensus that the young population is little affected because the semantic illusion of this paradigm increases with age. Only one study, Metzger & al. (2008), has focused on adapting DRM lists for children. In particular, the authors used a free association task to encourage children to produce associated words from lures. They determined children produced different word associations than adults. As for memorization, these researchers showed that correct responses increased in children that used the lists specifically adapted for them. Finally, as for the instances of false recall, researchers identified no differences compared to the adult lists. Despite these notable findings, one study does not warrant any conclusive conclusions about the functioning of a child’s semantic memory.
The objective of our study was to likewise adapt the DRM paradigm to children. Accordingly, we recruited children from 2 schools located near Nantes. The underlying research consisted of two phases. First, we offered a free association task to 119 children aged 8 to 11 years. We then requested the children utter the words that came to their minds from a presented item (lure). After the collection of the children’s words associations with lures, we created 15 lists, each of which included 9 words.
The second phase concerned the application of the word lists to 42 children (8-10 years). Our underlying goal was to assess the proportion of lures recalled for each word list to determine its sensitivity and thus definitively choose the 8 most appropriate DRM lists. The collected data was analyzed using descriptive statistics. Our results showed a significant variability in the apparition of LC between the lists. Of course, there are multiple variables that could influence the children’s memory, such as semantic and phonological characteristics, as well as the frequency and familiarity that children have with the words. Notably, this outcome demonstrated that the values of our most functional lists were close to those of adults. We could, therefore, make a hypothesis that children are sensitive to semantic distortions when lists are adapted to them.
Our study presents various limitations, including, most notably, its small sample size. In order to address this concern, perhaps a second study might be commissioned with the 8 selected DRM lists being presented to a larger number of children for testing.