As we pointed out in an earlier review (Brainerd & Reyna, 1998), the study of children's false memories can be segmented into research on two classes of phenomena: spontaneous false memories and implanted false memories. The former arise from normal, uncontrolled distortion processes that all memory is heir to, whereas the latter are due to misinformation ("Remember when you drank that Pepsi at the game last week?"), whether accidental or deliberate, that is imposed between the time that an event occurred (e.g., drinking a Coke at the game) and the time that memory for the event is tested ("Did you have anything to drink at the game last week?"). This segmentation is somewhat arbitrary because, as Reyna and Lloyd (1997) showed, there is a gradation of misinformation in studies of children's false memories, and even an old-new recognition probe ("Did you drink a Pepsi at the game last week?") is somewhat suggestive (Lamb & Fauchier, 2001), more so when it is repeated (Ceci & Bruck, 1995). Nevertheless, the segmentation is highly serviceable because there are important differences between the methodologies and findings of studies that focus on spontaneous versus implanted false memories.
Our own research program has centered on spontaneous rather than implanted false memories, for both theoretical and forensic reasons. Theoretically, the paradigms that are used to study spontaneous false memories are more attractive because they are simpler than those that are used to study implanted false memories. Greater simplicity means that it is easier to test theoretical hypotheses because there is only one source of memory falsification to deal with—spontaneous distortion processes— and one need not control for possible interactions between spontaneous distortion processes and other distortion processes that are pursuant to misinformation. Most importantly, one need not worry about controlling for developmental interactions between spontaneous distortion processes and susceptibility to misinformation, interactions that are both ubiquitous and powerful (Brainerd & Reyna, 2005), because misinformation is not present. Forensically, the question of children's implanted false memories is already settled law. The law has already recognized, in a number of legal opinions, that children are at heightened risk of making false memory reports when they receive misinformation (e.g., in the form of suggestive questioning about their experiences; Ceci & Friedman, 2000). Therefore, the law concedes that it is particularly important not to subject child witnesses and victims to misinformation as part of investigative procedures. This does not mean that this prescription is followed in most cases in which children are involved, and indeed, the data indicate that it is not followed (e.g., Warren & McGough, 1996; Warren & Woodall, 1999). Because the prescription is settled law, however, when it is not followed, children's memory reports can be legally challenged on scientific grounds as being unreliable, both at trial and in pretrial procedures (see Rosenthal, 2002). Unlike implanted false memories, the question of spontaneous false memories is not settled law. For these reasons, the review that follows focuses on spontaneous rather than implanted false memories in children. Insofar as children's implanted false memories are concerned, multiple reviews of the relevant literature are already extant (e.g., Bruck & Ceci, 1997, 1999; Ceci & Bruck, 1993, 1995), and interested readers are directed to those documents.
The modern scientific literature on spontaneous false memories may properly be said to have begun with two articles, one by Underwood (1965) that focused on false memory for words and one by Bransford and Franks (1971) that focused on false memory for sentences from narratives. Bransford and Franks' article caused an immediate sensation, stimulated an extensive literature on false memory for narratives, and soon became a staple of undergraduate psychology textbooks. Underwood's article, on the other hand, was less well known initially, but it has become prominent in recent years as research on false memory for word lists has exploded, with the advent of the Deese/Roediger-McDermott (DRM) paradigm (Deese, 1959; Roediger & McDermott, 1995). In the review that follows, we focus first on developmental studies of false memories for information from narratives and then take up developmental studies of false memory for words. Our review concentrates on research that is conducted under controlled laboratory conditions, rather than on naturalistic studies, for the same reasons that motivate concentrating on spontaneous false memory: theoretical probity and forensic application. In the former connection, tests of theoretical hypotheses demand research designs in which causal factors are carefully separated, rather than confounded. As is well known, controlled laboratory research is unparalleled in its ability to achieve such separation and thereby to provide clear answers to alternative theoretical hypotheses. It is less well understood that forensic application demands comparable clarity and rigor. Naturalistic studies of false memory that confound multiple factors are of limited use in the courtroom (see Reyna, Mills, Estrada, & Brainerd, 2007). The law recognizes this fact and has enshrined it in the standards that courts have adopted for the introduction of scientific evidence. The most notable of these is the Daubert standard, which governs scientific evidence in all federal courts and many state courts. It is designed to ensure that such evidence approximates the ideal of controlled laboratory experimentation to the greatest extent possible.
Owing to this emphasis, readers of this volume may be disposed to wonder about the generality of the findings on children's false memory to real-life situations. Although we will return to this topic at the close, we offer two preliminary observations. First, developmental scientists are inclined to treat the notion of naturalism and ecological validity in research as though it were a self-evident truth, not because there is vast empirical support for it but because it resonates so well with some deeply Rous-seauian presuppositions (White, 1970). In fact, however, the empirical evidence, which ought to be the ultimate arbiter of our beliefs, is almost entirely on the other side of the issue. As Banaji and Crowder (1989), Ceci and Bronfenbrenner (1991), and other respected authorities have commented, there appears to be no extant example of a major laboratory principle of memory that does not hold or that holds in the opposite way in everyday remembering. To take just few examples, laws such as serial position, encoding variability, the forgetting curve, and massed practice operate in memory for birthday parties and classroom lectures as well as in memory for controlled laboratory tasks. Although quantitative results for these laws may be somewhat different for some everyday experiences versus some laboratory tasks, there is also quantitative variability among laboratory tasks. The evidence that laboratory memory laws operate in the same general ways in everyday remembering is so extensive that it is clear what our baseline assumption should be. That assumption should not be that well-replicated laboratory patterns are somehow suspicious because they are not based on real-life situations but, rather, that they apply to such situations unless there are positive data to the contrary.
The other observation, as Reyna, Holliday, and Marche (2002) have discussed, is that, ironically, the findings that are generated in laboratory experimentation are usually more applicable to legal cases than naturalistic memory research. The reason is simple: Carefully controlled laboratory work has as its hallmark the ability to identify causes of children's false memories and to rule out other possible explanations by disentangling various factors. Those same factors are mixed and confounded in everyday life, so that no clear conclusions about what causes what are possible. Moreover, it should be obvious that no single naturalistic study or collection of studies can hope to match the specific details of legal cases. What the law requires of science, as we have said, are well-supported general principles that cut across different remembering situations. That is the stock in trade of laboratory experimentation.
Was this article helpful?
Parenting is a challenging task. As a single parent, how can you juggle work, parenting, and possibly college studies single handedly and still manage to be an ideal parent for your child? Read the 65-page eBook Single Parenting Becoming The Best Parent For Your Child to find out how. Loaded with tips, it can inspire, empower, and instruct you to successfully face the challenges of parenthood.