Children are poor witnesses. Or are they?

The vulnerability of children’s testimony

Early research has pointed out weaknesses in children’s memory and its implications for the legal field (e.g., Loftus, 2005). The underlying question was whether and under which circumstances children are capable of giving an accurate report of what they have witnessed. Among the first researchers who systematically and experimentally analysed children’s memory in a way that has implications for the legal context were Binet (1900) and Varendonck (1911) at the beginning of the twentieth century (as cited in Ceci & Bruck, 1993). They found that the type of questioning has an influence on children’s answers. Specifically, invitations to freely recall what had happened led to the most accurate responses, whereas (mis)leading questions merely elicited conformal answers (Binet, 1900). In Varendonck’s field study, 17 of 22 children even reported details about an unknown suggested person. Subsequent studies adapted the methods used by these pioneers. A review of studies published between 1979 and 1992 comparing the suggestibility of young children vs. older children and adults found a decrease in the formation of false memories with age in 83% of the studies (Ceci & Bruck, 1993). Even though individual differences exist, young children were generally found to be more vulnerable to suggestive interviews and erroneous information than older children and adults (Bruck & Ceci, 1999).

The research reported above was, among others, stimulated by legal cases in which the authenticity of children’s testimony was doubted (e.g., Ceci & Bruck, 1993). Examples are the American day care child sexual abuse cases of Kelly Michaels and the McMartin Preschool that occurred in the late 1980s. These two cases had in common that the police interviews with the children were conducted in a highly suggestive manner merely suited to confirm the interrogators’ prevailing hypothesis of the factual occurrence of abuses, but not to obtain truthful and accurate testimony (Schreiber et al., 2006). Interviewers’ techniques included (1) the introduction of new information that had not previously been mentioned by the child witness, (2) positive reinforcement of answers that were in accordance with police's prior expectations and beliefs of what had happened, (3) showing scepticism towards answers that were not in accordance with those prior expectations and beliefs, (4) applying conformity pressure by referring to statements made by other children, and (5) invitations to speculate about the alleged events (Schreiber et al., 2006).  

Importantly, such cases were not exclusively American. In Germany, the Wormser- and the Montessori-trials are notable examples of cases in which children’s statements were likely the result of suggestive interviewing techniques (Schade & Harschneck, 2000). Furthermore, in the Netherlands, the alleged ritual abuse of about 50 children between the ages of three to 11 by strangers in the city Oude Pekela was highly debated (Jonker & Jonker-Bakker, 1997). Notably, all these cases started with a vague initial suspicion that evolved into serious and precise allegations with life-changing consequences for the suspect.

The question can be raised whether children’s vulnerability as eyewitnesses is due to memory impairments or adults creating an atmosphere of speculation and rumour transmission. Important in this branch of research is Principe, in whose studies children watched a magic show where the rabbit-out-of-hat trick failed (Principe & Schindewolf, 2012). The children were then assigned to different groups among which one group overheard an adult conversation about why the trick failed and how the rabbit escaped. This group and another group of their classmates were the rumour groups. After a two-week-delay, their statements were compared to a group that actually witnessed a rabbit running around after the magic show. Intriguingly, no differences were present between these groups on reporting the loose rabbit (Principe, Kanaya, Ceci, & Singh, 2006). Thus, children can remember a witnessed event, but false memories might be due to their vulnerability to elaborate on rumours.  

The bottom-line message emerging from the legal cases and the related research was that false memories are more prevalent among children than adults, and hence children’s testimonial accuracy is inferior to that of adults. However, this view has recently been challenged by a series of studies showing age-related increases in false memories. For example, younger children were less likely to infer and elaborate on the loose rabbit rumour when only clues (i.e., nibbled carrot) suggested its escape than older children (Principe, Guiliano, & Root, 2008). These diametrical findings challenge the child’s supposed inability to make credible statements. It even suggests that not children but adults are more susceptible to certain false memories

article author(s)