Representational Constraints on Children's Suggestibility

Abstract

In a multistage experiment, twelve 4- and 9-year-old children participated in a triad rating task. Their ratings were mapped with multidimensional scaling, from which euclidean distances were computed to operationalize semantic distance between items in target pairs. These children and age-mates then participated in an experiment that employed these target pairs in a story, which was followed by a misinformation manipulation. Analyses linked individual and developmental differences in suggestibility to children's representations of the target items. Semantic proximity was a strong predictor of differences in suggestibility: The closer a suggested distractor was to the original item's representation, the greater was the distractor's suggestive influence. The triad participants' semantic proximity subsequently served as the basis for correctly predicting memory performance in the larger group. Semantic proximity enabled a priori counterintuitive predictions of reverse age-related trends to be confirmed whenever the distance between representations of items in a target pair was greater for younger than for older children.

Memory development: implications for adults recalling childhood experiences in the courtroom

“…Using multidimensional scaling analyses to calculate relational (or associative, that is, semantic) distances between these concepts, one can map how these distances change across age. When 4-year-olds and 9-year-olds rated the same set of concepts for their similarity, different scaling solutions were obtained ( Figure 2) 29 , indicating that older children's knowledge base is organized differently (i.e., concepts are interrelated in a qualitatively different manner) than younger children's in ways that can have direct behavioural consequences. For example, semantic distances between concepts for younger children predict confusions among those concepts for children of that age but not for older children.…”

“…For example, semantic distances between concepts for younger children predict confusions among those concepts for children of that age but not for older children. Indeed, these distances have been used to predict susceptibility to memory illusions (e.g., spontaneous false memories) [30][31][32] and to anticipate age-based changes in children's vulnerability to suggestion 29 . Interestingly, younger children (4-to 7-year-olds) generate fewer spontaneous false memories than older children (9-to 12-year-olds) and adults 28,[30][31][32] .…”

“…It is perhaps prudent, then, that legal proceeding where memories of such events constitute the main or only evidence heed the advice in a report by the British Psychological Society 25 that states, first, "In general the accuracy of memories dating below the age of 7 years cannot be established in the absence of independent corroborating evidence"; and second, "These findings lead to the conclusion that by approximately 9 to 10 years of age children have autobiographical memories that are adult in nature". Figure 2: Differences in (a) younger and (b) older children's knowledge base using nonmetric multidimensional scaling solutions derived from INDSCAL models of children's similarity ratings (adapted from Ceci et al 29 ). These solutions show Euclidean distances between concepts for 4-year-olds and 9-year-olds and illustrate changes with age in how concepts are interrelated in semantic memory (knowledge base).…”

“…Parallel developments also occur in the semantic components of children's autobiographical 1,[27][28][29] . One way to study these developmental changes in the organization of semantic knowledge is by asking children to make similarity judgments among concepts or objects.…”

The development of automatic associative processes and children’s false memories

“…For example, whenever the paradigm relies on internally generated semantic processes that occur outside of conscious awareness, false memory rates tend to increase with age (e.g., in the DRM task; also see Ceci et al, 2007). Alternatively, whenever the paradigm relies on semantic processes whose generated contents are subject to conscious awareness, false memory rates tend to decrease with age as consciously controlled strategies (e.g., source-monitoring, memory editing strategies) can be used to edit memory outputs.…”