10 Dual processes in the development of reasoning: the memory side of the theory – The Development of Thinking and Reasoning

10
Dual processes in the development of reasoning

The memory side of the theory

Carlos F. A. Gomes and Charles J. Brainerd

Theories of cognitive development that unify findings and concepts from different domains (e.g., reasoning and memory) are more parsimonious and fundamental than theories that are confined to a particular domain. Piaget’s constructivist view of cognitive development (Piaget, 1970), and his work on memory-reasoning relations in particular (Piaget & Inhelder, 1973), constitutes an early attempt to develop such a unified theory. Similarly, fuzzy-trace theory (Brainerd & Reyna, 1990; Reyna & Brainerd, 1995) began as an attempt to rethink earlier approaches to the development of memory and reasoning that could not predict or explain, for instance, statistical dissociations between the two (Brainerd & Reyna, 1992). In fuzzy-trace theory, this rethinking took the form of a dual-process theory that posited two basic mechanisms – verbatim and gist – as the principle underlying developmental trends in the memory and reasoning domains. The study of false memory (i.e., a memory about an event that has never happened) has been instrumental in revealing the developmental trajectories of verbatim and gist processes (Brainerd & Reyna, 2005). In particular, the counterintuitive finding that false memory can increase with age (Brainerd, Reyna, & Ceci, 2008; Brainerd, Reyna, & Zember, 2011) imposes major constraints on theoretical explanations of cognitive development and, therefore, will be the focus of this chapter.

In this chapter, we review the theoretical bases for predicting age increases in false memory, a developmental trend that is so counterintuitive it has been called a developmental reversal prediction (Brainerd, Yang, Reya, Howe, & Mills, 2008), and the data that support this prediction. The chapter is organized as follows. We start with an operational definition of false memory and present the standard experimental paradigm used to study false memories. We then present two theoretical bases for predicting developmental reversals in false memory. The first derives from Piaget’s view of the memory-reasoning relation, which we refer to as constructivism. The second derives from fuzzy-trace theory. In addition to predicting developmental reversals, fuzzy-trace theory specifies the conditions under which reversals occur. Next, we compare Piagetian constructivism with fuzzy-trace theory in the reasoning and memory domains. Although the two differ in important respects, growth of meaning making plays a central role in both theories and has direct implications for age increases in false memory. We then review data on developmental reversals and, finally, we present a two-pronged strategy for testing the ontogenesis of gist memory as the principle underlying developmental reversals.

False memory: definition and measurement

True and false memories are both episodic memory phenomena. True memory, on the one hand, refers to an episodic memory in which a person remembers a particular event as having happened that was in fact experienced. False memory, on the other hand, refers to an episodic memory in which a person remembers a particular event as having happened that was not in fact experienced. Although false memories are a form of memory distortion, they are not abnormal but, rather, a common feature of normal cognition (Grassi-Oliveira, Gomes, & Stein, 2011), and because they are fictional and are therefore “inferential,” it has traditionally been thought that false memories are closely connected to reasoning (Piaget, 1968; Reyna & Brainerd, 1995).

The experimental paradigms that are used to investigate false memories also allow the investigation of true memories, although the inverse is not necessarily true. In a standard false memory experiment, subjects are presented with a set of stimuli to study (e.g., words, sentences, objects, photographs), and later, they are asked either to recognize the studied stimuli among new ones or to recall them. Suppose, for instance, that subjects are presented with a list of words such as apple, banana, pear, lemon, chair, bed, desk, and sofa. After studying all these words, they perform a recognition test or a recall test. On a recognition test, there are three types of test probes: targets, related distractors, and unrelated distractors. Targets are studied items (e.g., apple, lemon, chair, sofa). Related distractors are items previously not presented but related to targets. In the working example, related distractors would be taxonomically related to targets (e.g., orange, table). Unrelated distractors are items previously not presented that do not preserve any salient features of targets (e.g., car, shovel) and, therefore, provide a baseline measure of response bias to which acceptance of targets and related distractors can be compared. Consequently, true memory is measured as the bias-corrected acceptance rate of targets, whereas false memory is measured as the bias-corrected acceptance rate of related distractors. Examples of bias-free measures of false memory include the signal detection d’ statistic for false items and its non-parametric analogue, A’ (Snodgrass & Corwin, 1988). In addition, subjects performing a recognition test are instructed to accept targets and reject both related and unrelated distractors. This procedure ensures that false memory is not an artifact of experimental design, such as accepting memory probes solely on the basis of meaning resemblance (Alba & Hasher, 1983).

If a free recall test is administered, subjects are simply asked to recall as many items from the study list as possible. At times, subjects will recall items that were not presented, which are called intrusions. Intrusions can be separated into two types: (a) those that preserve salient features of targets (e.g., semantic, phonological, or orthographical features) and (b) those that are unrelated to targets. Consequently, false memory is measured as the number of intrusions in a recall protocol that preserve a particular salient feature of targets (usually, meaning; but see Dewhurst & Robinson, 2004), whereas true memory is measured as the number of correctly recalled items.

Theoretical bases for predicting developmental reversals in false memory

Constructivism

The idea that knowledge results from a continuing interaction between intelligence and the environment is a core characteristic of Piaget’s view of cognitive development (Piaget, 1970), one that emphasizes the constructive nature of learning. In this approach, memory is described as an integral part of reasoning rather than being dissociated from it (Piaget & Inhelder, 1973). Therefore, one’s recollections about past experiences reflect what was understood about them rather than their literal form. Independent support for the notion that memory is constructive comes from Bransford and Franks’ (1971) classic experiment on memory for sentences. Bransford and Franks presented sentences such as “the ants ate the sweet jelly” and “the sweet jelly was on the table” to young adults. According to constructivism, adults’ memory for the sentences is based on the general understanding of the meaning of all sentences rather than the verbatim form of each sentence. Consequently, erroneous recognition of new sentences that integrate the meaning of two or more studied sentences should be similar to correct recognition of old sentences because general meaning content provides support for both old and new related sentences. Bransford and Franks’ experiment confirmed this prediction. Adults falsely recognized new sentences that preserved the meaning of studied ones (e.g., the ants ate the sweet jelly which was on the table) at nearly the same level as they correctly recognized old sentences, and their levels of confidence in the two types of responses were indistinguishable.

A key implication of Piaget’s perspective on the memory-reasoning relation is that memory depends on reasoning. More specifically, as one’s cognitive structures undergo qualitative changes from early childhood to young adulthood, there is an accompanying qualitative difference in the way in which children versus adults remember past experiences. In line with that, research on memory for seriated objects has shown that children who have developed the concept of seriation are better able to correctly recall or recognize an array of seriated sticks one week after its presentation than children who have not developed an understanding of seriation (Altemeyer, Fulton, & Berney, 1969; Piaget & Inhelder, 1973). Similar findings have been reported in which memory for objects was related to concepts other than seriation, such as horizontality and verticality (Furth, Ross, & Youniss, 1974; Liben, 1974; for a review, see Liben, 1977), suggesting that remembering is influenced by current levels of conceptual understanding.

A corollary of the constructivist view of the development of reasoning operations and its relation with memory is that developmental reversals in false memory will occur whenever constructive processing distorts memory for actual events in a direction that is consistent with subjects’ underlying conceptual understanding of the stimuli. For example, consider a list learning paradigm in which subjects are first presented with lists of taxonomically related words (apple, banana, pear, lemon, chair, bed, desk, sofa) and are asked to recall the studied words later on. Extraction of meaning relations from lists (e.g., that apple, banana, pear, and lemon are all types of fruit) requires concepts that detect global patterns, which are not developed until the formal operational stage in Piaget’s theory (Piaget, 1970). Therefore, if false memory is measured as the number of intrusions whose meaning overlaps with the studied words (e.g., recall of the words orange and table), then the constructivist framework would predict age increases in false memories.

Fuzzy-trace theory

In fuzzy-trace theory, the development of memory and reasoning are related by the ontogenesis of verbatim and gist processes (Reyna & Brainerd, 1995). Verbatim processes generate representations of an item’s surface features, such as its form, size, color, or quantity. Gist processes generate representations of an item’s semantic features, such as its bottom-line meaning. Because verbatim and gist processes are independent of each other, extraction of meaning content, for example, does not depend on storage of an item’s verbatim trace. Moreover, the principle of verbatim-gist independence explains why memory for an item’s surface information is often dissociated from memory for its meaning (Nelson, McGivney, Gee, & Janczura, 1998; Reyna & Lloyd, 1997; Shiffrin, 2003). Similarly, in fuzzy-trace theory, memory-reasoning dissociations (Brainerd & Kingma, 1984, 1985), which are a major challenge to early views of memory-reasoning relationship, can be explained in terms of preferred modes of processing when performing a memory task (e.g., verbatim) relative to performing a reasoning task (e.g., gist; for a review, see Brainerd & Reyna, 2002). More importantly, however, the principle of process independence allows for separate developmental trajectories in verbatim and gist memory.

Developmental variability in verbatim and gist memory

Multiple lines of evidence indicate that both verbatim and gist processes exhibit variability from early childhood to adulthood (Bouwmeester, Vermunt, & Sijtsma, 2007; Brainerd & Reyna, 2004; Reyna, Holliday, & Marche, 2002), with verbatim memory developing more rapidly than gist (Brainerd & Reyna, 2005). For instance, developmental improvements in verbatim memory have been demonstrated in source memory experiments (Drummey & Newcombe, 2002) and memory for nonsense words (Brainerd, Stein, & Reyna, 1998). Drummey and Newcombe (2002) tested source memory in groups of four-, six-, and eight-year-olds. Their subjects were presented with and questioned about ten facts (e.g., Q: “What animal cannot make any sounds?” A: “A giraffe cannot make any sound”) by either the experimenter or a puppet. One week later, memory for facts (e.g., a giraffe is an animal that cannot make any sound) and sources (e.g., this fact was presented by a puppet) were tested. Recall of both facts and sources improved with age. However, age improvements were different for facts and sources. Recall of facts improved steadily from four-year-olds (23%) to six-year-olds (32%) to eight-year-olds (51%), but source accuracy roughly doubled between four-year-olds (24%) and six-year-olds (47%) and did not improve thereafter. In the same vein, Brainerd et al. (1998) presented nonsense words (e.g., neppez, wux) to groups of seven- and ten-year-olds and asked them to recognize targets and new distractors immediately after study. According to fuzzy-trace theory, only verbatim traces of nonsense words are stored because nonsense words have no semantic content. Therefore, age increases in bias-corrected acceptance of nonsense target words ought to reflect age improvements in verbatim memory. Under standard recognition instructions to accept only targets and reject new distractors, Brainerd et al. found age increases in true memory for nonsense words between seven-year-olds and ten-year-olds.

Gist memory has been shown to develop independently of verbatim memory. One line of evidence comes from experiments that investigated semantic clustering in free recall (Bjorklund, 1987, 2004). Semantic clustering is a form of spontaneous organization in which items are output next to each other according to one or more types of semantic relations (e.g., taxonomy, antonymy, and synonymy). Development of gist memory is characterized by increasing ability to extract semantic relations among different items, which is a necessary condition for semantic clustering in free recall. Therefore, age improvements in semantic clustering provide evidence of gist memory development. In Bjorklund and Jacobs’ (1985) experiment, children of four age levels, 8-, 10-, 12-, and 14-year-olds, were presented with exemplars of four categories (animals, occupation, seasoning, and weapons) and then asked to recall as many exemplars as possible. The results revealed a steady increase in category clustering from 8-year-olds to 14-year-olds. Another line of evidence comes from experiments that investigated the development of meaning connection among words and sentences (Reyna & Kiernan, 1994, 1995). In such experiments, subjects study short sentences, such as “the cat is under the table,” and then perform a recognition test in which they are instructed to accept sentences whose meaning is consistent with studied sentences (e.g., the table is above the cat) and reject sentences whose meaning is inconsistent with studied sentences (e.g., the cat is on the table). Reyna and Kiernan (1994) measured meaning connection in groups of six- and nine-year-olds at two different levels. At the level of word to sentence connection, meaning connection was measured by presenting sentences during test that paraphrased studied sentences (the table is above the cat). At the level of phrase to story connection, meaning connection was measured by presenting sentences during test that combined information from two studied sentences. For instance, the sentences “the cage is on the table” and “the cat is under the table” were presented during study and the meaning-consistent test probe was “the cage is above the cat.” When the probability of both levels of meaning connection was computed for each age group, Reyna and Kiernan found age increases in meaning connection between six-year-olds (sentence =.26, story =.34) and nine-year-olds (sentence =.51, story =.60) at both the sentence and the story levels, supporting the hypothesis that gist memory improves with age.

Opponent processes in false memory

Fuzzy-trace theory posits that gist traces support both true and false memories, whereas verbatim traces support true and suppress false memories (through an operation known as recollection rejection; Brainerd, Reyna, Wright, & Mojardin, 2003). Suppose that subjects in a memory experiment are presented with the words apple, banana, pear, and lemon. If a forced-choice recognition test is administered after study, in which lemon (a target), orange (a related distractor), and shovel (an unrelated distractor) are presented as alternatives, the theory explains performance as follows. Acceptance of lemon is supported by retrieval of either a verbatim trace (lemon) or a gist trace (fruit). Acceptance of orange is also supported by retrieval of a gist trace, but orange can be rejected by retrieving an exclusionary verbatim trace (“orange was not presented because I only saw apple, banana, pear, and lemon”). When verbatim and gist retrieval both fail, either shovel or orange or lemon could be accepted on the basis of response bias. Consequently, conditions that foment retrieval of verbatim traces tend to increase true memory and reduce false memory, whereas conditions that foment retrieval of gist traces tend to increase both true and false memory.

An algorithm for producing developmental reversals

The principles of opponent processes in false memory and developmental variability in verbatim and gist memory have an important implication in predicting counterintuitive developmental reversals in false memory. Specifically, age increases in false memory are expected whenever retrieval of gist traces is maximized by the experimental design (because gist memory waxes with age and supports false memories) and retrieval of verbatim traces is minimized (because verbatim memory waxes with age as well and suppresses false memories). Therefore, two things must be accomplished in order to produce developmental reversals. First, in the gist sphere, false memories must be based on semantic abilities that are poorly developed in children but well developed in adults. Tasks in which the creation of false memories require formation of semantic connections among events, called connected-meaning tasks, are prime candidates for producing developmental reversals because the capacity to connect the meaning across events develops slowly (Brainerd et al., 2008). Second, in the verbatim sphere, recollection rejection of false but gist-consistent items should be hard to perform, thereby neutralizing the influence of age increases in verbatim memory. Candidate tasks that meet this criterion are ones in which false items preserve the meaning of several targets because retrieval of a few targets’ verbatim traces will not be diagnostic of whether a certain gist-consistent item is false. For example, suppose you are presented with a story about someone who takes a wine tour in France, sampling Bordeaux, Cabernet Sauvignon, Gamay, Grenache, Malbec, Mourvèdre, Pinot Noir, Rhône, and Shiraz. On a later memory test, retrieval of a few wines’ verbatim traces (“the person drank Cabernet Sauvignon, Malbec, and Shiraz”) is not sufficient to suppress false memory for Merlot (“but many other wines were drunk as well”).

Piagetian constructivism and fuzzy-trace theory

Reasoning

In the Piagetian approach to the development of reasoning, biases, and fallacies in decision making ought to decrease with age as young children’s thinking operations become more logical and less intuitive. This is the opposite of fuzzy-trace theory’s prediction (e.g., Reyna & Brainerd, 2011). In fuzzy trace-theory, both intuitive and analytical reasoning develop, as the forms of representation that underlie each (gist and verbatim memory) have been shown to improve between early childhood and adulthood. Contrary to Piagetian constructivism, fuzzy-trace theory holds that most adults prefer to reason with gist on decision problems because although gist is more susceptible to biases and fallacies, it enjoys great advantages in trace availability and malleability and leads to more flexible and less effortful reasoning (intuitive) than verbatim memory (analytical; Brainerd & Reyna, 1990).

When it comes to empirical findings, the rational adult described by Piagetian constructivism seems not to resemble the adult described by heuristics and biases research (Reyna & Brainerd, 1994, 1995). In fact, biases and fallacies in decision making have been shown to increase with age rather than decrease (Reyna & Brainerd, 2011). For instance, Reyna and Ellis (1994) investigated framing biases (i.e., illogical changes in risk preference depending on whether a problem is described as yielding gains or losses) in children ranging from 4- to 11-years-old. Subjects were presented with such problems and had to choose between an uncertain (probabilistic) and a certain outcome. For example, in gain frame problems, children had to decide between winning one prize with probability one (the certain outcome) or taking a chance of winning three prizes with probability one-third (the uncertain outcome, as there is a probability equal to two-thirds of winning nothing). In loss frame problems, children received an endowment of three prizes and had to decide between losing two of them with probability one or taking a chance of losing all three prizes with probability two-thirds (consequently, a one-third probability of losing nothing). From a statistical standpoint, there should not be a preference for choosing a probabilistic outcome when problems are framed as losses, or preference for choosing a certain outcome when problems are framed as gains (i.e., framing effects) because the expected values for the certain and probabilistic outcomes are always equal, regardless of whether problems are framed as gains or losses. According to the Piagetian theory, such framing effects should decrease with age, as reasoning operations conforms more closely to the rules of logic. Contrary to that prediction, Reyna and Ellis found that framing effects increased with age. That result has been replicated by several investigators (Reyna & Brainerd, 2011).

Memory

As a one-process theory, constructivism posits that true and false memories are based on a single semantic code that stores the meaning of past experiences rather than their verbatim form. Therefore, any factor that affects true memory ought to affect false memory in the same direction. Interestingly, age increases in false memory can also be predicted on the ground that this semantic code improves with age (Brainerd, Reyna, & Howe, 2009). However, the first prediction, that true and false memory must be positively associated owing to their shared semantic code, has been challenged by findings from a multitude of experimental studies showing dissociations between the two (McDermott & Watson, 2001; Reyna & Kiernan, 1994, 1995; Roediger, Watson, McDermott, & Gallo, 2001; for a review, see Brainerd & Reyna, 2005). In fuzzy-trace theory, these dissociations are actually predicted by the principle of opponent processes. More specifically, conditions that improve retrieval of verbatim traces should produce dissociations between true and false memory because verbatim traces affect them in opposite ways.

A common ground

Developmental variability in cognitive structure (Piagetian constructivism) or in gist memory (fuzzy-trace theory) affects the way one comprehends the world and, consequently, developmental changes in memory and decision making. In Piagetian constructivism, the development of logically organized structures authorize inferences about past experiences that produce false memories, such as when one falsely recognizes or recalls the word orange following a list that connected the words apple, banana, and pear (i.e., the past occurrence of orange is inferred from knowledge of having studied several fruit names). Similarly, according to the fuzzy-trace theory, the development of gist memory provides a basis for false recognition or false recall. Therefore, in both approaches to cognitive development, age increases in false memory are rooted in the growth of understanding conceptual relations between superficially distinct events.

Developmental reversals in false memory

We have seen that predictions of age increases in false memory follow naturally from principles in both Piagetian constructivism and fuzzy-trace theory. However, compelling support for this prediction did not emerge until 2002 (Brainerd, Reyna, & Forrest, 2002). Before then, belief in monolithic developmental declines in false memory had been a dominant hypothesis for decades (e.g., Bruck & Ceci, 1999; Ceci, Ross, & Toglia, 1987). This belief was challenged when developmental studies using an experimental paradigm called the Deese–Roediger–McDermott (DRM; Deese, 1959; Roediger & McDermott, 1995) illusion started to show the opposite pattern (Brainerd et al., 2008, 2011; Brainerd, Holliday, Reyna, Yang, & Toglia, 2010). We describe this paradigm next.

DRM: A technique for producing developmental reversals

In the DRM illusion, subjects study lists of 12 to 15 words (e.g., note, sound, piano, sing, radio, band, melody, horn, concert, instrument, symphony, jazz) that are semantically associated to unpresented critical words (music). After study, subjects perform an episodic memory test in which false memory is measured as either the bias-corrected rate of false recognition of unpresented critical words on recognition tests, or as the intrusion rate of unpresented critical words on free recall tests. Independent of the type of memory test, false memory levels are often very high and similar to true memory levels (Roediger et al., 2001).

The DRM illusion is a type of connected-meaning task that satisfies fuzzy-trace theory’s algorithm for producing developmental reversals. First, targets in each word list are dense in semantic relations between each other and with unpresented critical words (Brainerd et al., 2008), particularly taxonomic relations (words share a common taxonomic category, such as a type of fruit or bird), situational relations (words share a common situation, such as an action, function, or location), and entity relations (words share a common entity in which one refers to the entity itself [e.g., chair] and the other refers to a property of the entity [e.g., wood or metal]). This allows subjects in a DRM experiment to form strong gist memories for lists that can be retrieved during test to support false memories. Second, each DRM list is composed of 12 to 15 words, which makes retrieval of a few targets’ verbatim traces (note, sound, band, jazz) insufficient to demonstrate that an unpresented but gist-consistent item (music) is false. In other words, recollection rejection should become more difficult as list length increases, and DRM lists are long.

Review of data

The search for developmental reversals in false memory using the DRM paradigm started with a confusing picture. In 2002, two studies conducted by independent groups of researchers (Brainerd et al., 2002; Ghetti, Qin, & Goodman, 2002) reported opposite developmental trends in false memory. On the one hand, Brainerd et al. (2002) reported age increases in false memory. On the other hand, Ghetti et al. (2002) reported age declines in false memory.

Ghetti et al. (2002) presented ten DRM lists composed of seven words apiece to groups of 5-year-olds, 7-year-olds, and young adults. After each list, subjects performed a free recall test, and after the presentation of all lists, subjects performed a recognition test. The results revealed that false memory was age invariant on the recognition test, but decreased with age on the recall test. However, as Reyna, Mills, Estrada, and Brainerd (2007) pointed out, the length of the DRM lists used by Ghetti et al. was much shorter than usual. For this reason, Ghetti et al.’s experiment does not meet the second criterion of the algorithm for producing developmental reversals in false memory, namely that recollection rejection should be hard to perform. Moreover, Ghetti et al.’s false recall measure was the ratio between the number of unpresented critical words recalled and the number of targets plus unpresented critical words recalled, rather than the standard measure of the ratio of unpresented critical words recalled to the total number of DRM lists presented. As a result, Ghetti et al.’s false recall measure will increase as true recall decreases, which is problematic when investigating developmental trends because young children’s true memory is usually lower than older children’s and young adults’ (Brainerd et al., 2009). A reanalysis of Ghetti et al.’s experiment using the standard measure revealed that false memory was actually age invariant on both free recall and recognition tests (Reyna et al., 2007), thus rendering their results inconclusive concerning developmental reversals in false memory.

Brainerd et al. (2002) presented ten DRM lists composed of twelve words each to groups of 5-year-olds, 11-year-olds, and young adults. Similarly to Ghetti et al.’s (2002) experiment, all age groups performed both a free recall and a recognition test. Contrary to Ghetti et al.’s experiment, false memory on both the recall and the recognition tests increased with age. On the free recall test, the intrusion of unpresented critical words increased from 5-year-olds (11%) to 11-year-olds (23%) to young adults (37%). On the recognition test, a bias-free statistic of false memory (A’) also increased from 5-year-olds (.76) to 11-year-olds (.84) to young adults (.86). Therefore, the first experiment that satisfied the algorithm for producing developmental reversals actually produced such a trend.

In the years following 2002, more than 50 experiments have been published in which the DRM illusion was used to investigate developmental variability in false memory and to evaluate several questions of theoretical importance. For instance, Sugrue and Hayne (2006) presented eight DRM lists, half composed of seven words (short lists) and half composed of fourteen words (long lists), to groups of children (9- and 10-year-olds) and young adults. As before, both recall and recognition tests were administered. The results in the short and long list conditions replicated the findings reported by Ghetti et al. (2002) and Brainerd et al. (2002), respectively. For short lists, the results showed age invariance in false memory; for long lists, the results showed age increases in false memory. Therefore, developmental increases in the DRM illusion can be masked by conditions in which operations that suppress false memories, such as recollection rejection, are facilitated. In addition to list length, another variable that should affect recollection rejection is the presentation rate of DRM lists, with slower rates favoring recollection rejection (Brainerd et al., 2003). Carneiro and Fernandez (2010) presented DRM lists to groups of young (4- and 5-year-olds) and older children (11- and 12-year-olds) at three different presentation rates, 2 sec., 4 sec., and 8 sec., and asked them to free recall the studied words. Under standard recall instructions, Carneiro and Fernandez found that developmental reversals in false memory decreased as the presentation rate increased. Specifically, the difference in false recall between older and young children decreased from 18% in the 2 sec. condition to 7% in the 8 sec. condition.

Another question that emerged in studies of developmental reversals was whether age increases in the DRM illusion were artifacts of how DRM lists were generated in the first place (e.g., Carneiro, Albuquerque, Fernandez, & Esteves, 2007). DRM lists are generated using a free association procedure in which a large number of subjects are instructed to say out loud, or write down, the first word that comes to mind when presented with a cue word. For instance, when young adults are presented with the cue music, the first word that most often comes to mind is note. Therefore, note is a forward associate of music. In the case of DRM lists, each list is generated by selecting the first 15 forward associates to the unpresented critical word. Because most developmental DRM experiments used lists constructed from adult association norms (Roediger et al., 2001; Stadler, Roediger, & McDermott, 1999), a natural question is whether developmental reversals are due to differences in the type of associations generated by adults versus children rather than to age variability in meaning connection among words. To investigate this question, Metzger et al. (2008), Anastasi and Rhodes (2008), and Carneiro et al. (2007) created child-normed DRM lists and presented them to different age levels. However, despite using lists normed for children, all experiments showed developmental reversals in false memory, thus ruling out the hypothesis that developmental reversals are word knowledge artifacts.

Recently, Brainerd et al. (2011) reviewed 55 developmental DRM experiments published in English language journals and found that 96% showed developmental increases in false memory from early childhood to adulthood. In addition, age increases in false memory have been reported using other connected-meaning tasks, such as lists of categorized words (Brainerd, Holliday, & Reyna, 2004) or pictures (Sloutsky & Fisher, 2004), and misinformation designs (Ceci, Papierno, & Kulkofksy, 2007; Fazio & Marsh, 2008). Therefore, a large amount of evidence indicates that developmental increases in false memory can occur when memory tasks maximize reliance on gist processing and minimize reliance on verbatim processing. Nonetheless, fundamental questions about the causes of developmental reversals remain, and in the next section we explore a two-pronged strategy for testing the hypothesis that the development of gist memory is crucial to developmental reversals.

Does meaning connection cause developmental reversals?

The key goal of experimental work in cognitive development is to establish cause-effect relations. In the case of developmental reversals (a behavioral pattern), our goal is to establish whether meaning connection (a theoretical process) causes them or not. According to fuzzy-trace theory, developmental reversals in false memory are caused by children’s limitations in forming meaning connections among distinct events. If meaning connection indeed underlies developmental reversals, then age increases in false memory can be eliminated (a) by interfering with older children’s and adults’ superior ability to connect meaning across events or (b) by enhancing younger children’s impoverished ability to connect meaning across events. In the first case, the idea is to reduce older children’s and adults’ ability to make meaning connections and, consequently, shrink developmental reversals by reducing false memory at older age levels. Manipulations that ought to do this are called necessity manipulations because they answer the question of whether meaning connection is necessary to produce age increases in false memory. In the second case, the idea is to supply younger children with the missing ability to make meaning connections and, consequently, shrink developmental reversals by increasing false memory at younger age levels. Manipulations that ought to do this are called sufficiency manipulations because they answer the question of whether meaning connection is sufficient to produce age increases in false memory. Evidence that meaning connection is both necessary and sufficient to produce age increases in false memory establishes a direct causal relation between the two.

Necessity manipulations

In order to provide an answer to the necessity question, the experimental design must include a condition in which subjects’ ability to connect the meaning across events (e.g., that note, sing, symphony, and jazz are all related to music) is interfered with. If meaning connection is necessary to produce age increases in false memory, then such manipulations will suppress false memory more in older than in younger subjects, thereby shrinking developmental reversals. However, if age increases in false memory do not contract under conditions that ought to impair meaning connection, then meaning connection is not necessary to produce such increases.

One way of interfering with subjects’ ability to make meaning connections in DRM lists is to direct their attention away from the semantic content of the words toward their surface features. In line with that, Holliday, Brainerd, and Reyna (2011) compared 7- and 11-year-olds’ false memory rates under two different conditions. In one, subjects were presented with word fragments (e.g., not_, sin_, symphon_) and instructed to fill in the missing letter of each word. In a second condition, subjects were presented with whole words (e.g., note, sing, symphony). After the presentation of all lists, subjects performed a recognition test. The results are shown in Panel A of Figure 10.1. As expected, the bias-corrected acceptance rate of critical words increased between 7-year-olds and 11-year-olds in the standard condition. However, the bias-corrected acceptance rate of critical words decreased between 7-year-olds and 11-year-olds when DRM words were presented as fragments. In the same vein, Howe (2008) presented DRM lists either as words, line drawings, or photographs to groups of 5-, 7-, and 11-year-old children, and asked them to perform a free recall test after the presentation of each list. One hypothesis is that age increases in false memory will be reduced or reversed when DRM lists are presented as line-drawings or photographs because pictures focus processing on attention-capturing visual details (Schacter, Israel, & Racine, 1999). This hypothesis was supported in Howe’s experiment. The results revealed age increases in false recall for DRM lists presented as words, but not for lists presented either as line-drawings or photographs with heterogeneous backgrounds.

A second way of interfering with subjects’ ability to make meaning connections in DRM lists is to reduce the salience of the semantic relations among list words. In the standard DRM paradigm, the presentation of targets during study is blocked by list (e.g., note, sound, …, jazz; web, insect, …, animal) rather than randomly distributed (e.g., sound, web, animal, jazz, insect, note, …). The blocked design makes the semantic relations among targets more salient and, therefore, random presentation of targets should interfere with meaning connection. This manipulation has been used in experiments with adults (McDermott, 1996; Toglia, Neuschatz, & Goodwin, 1999), the main finding being that random presentation decreases false memory relative to blocked. In developmental research, blocked versus random

Figure 10.1 The effects of necessity manipulations on false memory as a function of chronological age. Panel A shows bias-corrected false recognition rates for the whole and fragmented word conditions in an experiment reported by Holliday, Brainerd, and Reyna (2011). Panel B shows levels of false memory for the blocked and random presentation conditions in an experiment reported by Lampinen, Leding, Reed, and Odegard (2006).

presentation was used by Lampinen, Leding, Reed, and Odegard (2006). In their experiment, 6- and 8-year-old children and young adults (aged 20 years on average) studied DRM words either blocked by list or randomly distributed, and then performed a recognition test. The main finding was that false memory increased with age when words were blocked by list but not when they were randomly distributed. As can be seen in Panel B of Figure 10.1, a bias-free measure of false recognition (d’) in the blocked condition increased with age, but it decreased with age in the random condition.

Sufficiency manipulations

In the previous section, meaning connection was shown to be necessary to produce age increases in false memory, but necessity does not imply sufficiency. In other words, meaning connection can be necessary but not actually sufficient to produce developmental reversals. If that is the case, additional processes are required to explain age increases in false memory. The sufficiency question can be answered by studying conditions in which subjects are supplied with prosthetics that help them to connect the meaning across events. If meaning connection is sufficient to produce age increases in false memory, then this should cause such increases to shrink.

One method of improving subjects’ ability to form meaning connections is by cuing them about the relations among events (Bjorklund, 1987). Odegard, Holliday, Brainerd, and Reyna (2008) created a manipulation of this sort for the DRM illusion, in which 11-year-olds and young adults (aged 24 years on average) were cued either toward a meaning consistent with an unpresented critical word or away from it. The manipulation consisted of modifying each DRM list so that each target was paired with a novel word that was not a forward associate of the list’s critical word. In the context-toward condition, the word paired with each target had a meaning congruent with the critical word (e.g., note – rap, sing – drums, …, jazz – dance). In the context-away condition, the word paired with each target had a meaning incongruent with the critical word (e.g., note – apple, sing – juice, …, jazz – orange). After the presentation of all context-toward and context-away DRM lists, subjects performed an associative recognition test in which targets (note – rap), critical items (jazz – music), and unrelated items (car – shovel) of both conditions were presented. The results are presented in Panel A of Figure 10.2. Under standard recognition instructions to accept only old items and reject new ones, Odegard et al. found age invariance in bias-corrected false recognition rates (A’) between 11-year-olds and young adults in the context-toward condition. In the context-away condition, however, they found age increases in bias-corrected false recognition rates between 11-year-olds and young adults.

A second method of improving subjects’ ability to form meaning connections in the DRM illusion consists of presenting targets embedded in a context that reinforces the gist of a particular DRM list. Dewhurst, Pursglove, and Lewis (2007) proposed an ingenious method of presenting each DRM list in a way that makes its theme salient even to young children. The method consists of presenting lists in narrative contexts, such as in the following example taken from Dewhurst et al. in which the underlined words are all forward associates of the critical word doctor:

The nurse had written a prescription for Sally because she was sick. Her mum, who was a lawyer, told Sally she had to take the medicine because it would improve her health. She said if Sally did not take it she would have to go to the hospital. Sally hated them more than the dentist. Sally saw a physician the

Figure 10.2 The effects of sufficiency manipulations on false memory as a function of chronological age. Panel A shows levels of false memory for the context-away and context-toward conditions in an experiment reported by Odegard, Holliday, Brainerd, and Reyna (2008). Panel B shows bias-corrected false recognition rates for the standard and narrative conditions in an experiment reported by Dewhurst, Pursglove, and Lewis (2007).

last time she was ill. She went into his office and he listened to her heart with a stethoscope. She then went to a different clinic where she saw a surgeon who gave her the treatment she needed to cure her. (p. 378)

Dewhurst et al. presented eight DRM lists either embedded in such narratives or following the standard presentation format to groups of 5-, 8-, and 11-year-olds. After the presentation of each list, subjects performed a recognition test for both old and new words. The main finding, which is shown in Panel B of Figure 10.2, was that false memory increased from 5-year-olds to 11-year-olds in the standard condition but did not increase in the narrative format. Narrative presentation completely compensated for developmental differences in meaning connection ability.

Summary of the necessity and sufficiency manipulations

In order to establish a causal relation between meaning connection and developmental reversals in false memory, we considered a two-pronged strategy in which meaning connection was tested for both necessity and sufficiency. In the necessity prong, we discussed results for two different manipulations. The first aimed at directing subjects’ attention away from the semantic features of words toward their surface features by presenting DRM lists as word fragments. The second aimed at reducing the salience of semantic relations among targets by presenting DRM words in a random order rather than blocked by list. The results of both manipulations converged on the conclusion that interfering with subjects’ ability to connect the meaning across events shrinks developmental reversals. In the sufficiency prong, we discussed results for two other manipulations. The first aimed at cuing subjects about the relations among targets by pairing targets with additional words that biased them toward the theme of DRM lists. The second embedded targets in meaning-salient contexts by presenting DRM lists as short stories. The results of both manipulations showed that enhancing subjects’ ability to form meaning connections across events shrinks developmental reversals. Therefore, necessity and sufficiency tests supported the original theoretical bases for predicting developmental reversals in false memory, namely meaning connection causes this phenomenon.

Concluding comments

The possible developmental trajectories of false memory supply an arena for testing hypotheses about the ontogenesis of processes underlying memory and reasoning. In this chapter, we reviewed data on age increases in false memory and presented two theoretical bases for predicting such increases, Piagetian constructivism and fuzzy-trace theory. Although constructivism and fuzzy-trace theory differ in some fundamental respects (e.g., the first is a one-process theory and the second is a dual-process theory), the two share the common principle that false memories are rooted in the developmentally variable process of meaning making (understanding). According to constructivism, age increases in false memory emerge when constructive processes distort memory for actual events. According to fuzzy-trace theory, age increases in false memory emerge when two conditions are satisfied, both of which are common in everyday life. First, false memories are created as by-products of the formation of meaning connections among events. Second, it is difficult to suppress false memories by retrieving verbatim traces of actual experience.

The research reviewed in this chapter provides compelling evidence that false memories can increase between early childhood and young adulthood, thus challenging the long-held belief in monolithic developmental declines in false memory. In particular, we showed that age increases in false memory occur when the experimental task meets the two conditions specified by fuzzy-trace theory’s developmental reversal algorithm (i.e., maximize gist processing and minimize verbatim processing). The DRM paradigm satisfies both conditions and, for this reason, was the main focus of this chapter. However, developmental reversals in false memory have also been reported using connected-meaning tasks other than the DRM illusion, such as lists of categorized words (Brainerd et al., 2004) and videos of crimes (Ross et al., 2006). Ross et al. (2006) presented a 3 min. videotape to children aged between 5 and 12 years in order to investigate their susceptibility to misidentifying a familiar bystander as the culprit of a crime. The videotape started by showing a male preschool teacher (the innocent bystander) reading a story to a group of children. Later, a female teacher (the victim) was shown entering a cafeteria and sitting next to another man (the culprit) who matched the male preschool teacher’s gender, age, build, and ethnicity; that is, a male who was gist-consistent with the innocent bystander. She then put her wallet over the table, turned her back to the man sitting next to her, and walked to a vending machine. At that moment, the man sitting at the table examined her wallet, removed money from it, and left the cafeteria. After watching the videotape, subjects were presented with a photo-lineup composed of the bystander and four other foils and were told that the culprit might or might not be present in the lineup. The subjects’ task was to say whether the culprit was present in the lineup by selecting either one suspect or none of them. The results revealed age increases in false identification of the bystander as the culprit. Specifically, the percentage of children who misidentified the bystander increased from 18% to 64% between 5-year-olds and 12-year-olds.

The DRM paradigm has been instrumental in validating theoretical explanations of developmental reversals. We sketched some findings produced by a two-pronged strategy for testing the necessity and sufficiency of meaning connection in producing age increases in false memory. In the necessity prong, we reviewed studies that addressed this question by impairing subjects’ ability to form meaning connections across events, which should affect older subjects more than younger ones. In the sufficiency prong, we reviewed studies that addressed this question by helping subjects to form meaning connections, which should affect younger subjects more than older ones. The data reviewed indicated that meaning connection is both necessary and sufficient to produce developmental reversals in false memory.

This conclusion, however, should not be overgeneralized. Although there is now a good empirical case that meaning connection plays a major role in producing age increases in false memory, it is very likely that other processes also influence age variability in false memory. Variability in item-level and strategic suppression of false information are obvious candidates, but research on the developmental trajectories of both forms of suppression has been scarce (Bouwmeester & Verkoeijen, 2010; Brainerd et al., 2004; Ghetti et al., 2002; Odegard et al., 2008). Similarly, how the development of working memory and word association (Howe, Wimmer, Gagnon, & Plumpton, 2009; Metzger et al., 2008) affect age variability in false memory is still a matter of debate. Therefore, a relevant goal for future research is to address how processes other than meaning connection influence the development of false memory.

References

Alba, J. W., & Hasher, L. (1983). Is memory schematic? Psychological Bulletin, 93, 203–231. doi:10.1037/0033–2909.93.2.203

Altemeyer, R., Fulton, D., & Berney, K. (1969). Long-term memory improvement: Confirmation of a finding by Piaget. Child Development, 40, 845–857. doi:10.2307/1127193

Anastasi, J. S., & Rhodes, M. G. (2008). Examining differences in the levels of false memories in children and adults using child-normed lists. Developmental Psychology, 44, 889–894. doi:10.1037/0012–1649.44.3.889

Bjorklund, D. F. (1987). How changes in knowledge base contribute to the development of children’s memory. Developmental Review, 7, 93–130. doi:10.1016/0273–2297(87)90007–4

Bjorklund, D. F. (2004). Children’s thinking: Cognitive development and individual differences. Belmont, CA: Wadsworth.

Bjorklund, D. F., & Jacobs, J. W. (1985). Associative and categorical processes in children’s memory: The role of automaticity in the development of organization in free recall. Journal of Experimental Child Psychology, 39, 599–617. doi:10.1016/0022–0965 (85)90059–1

Bouwmeester, S., & Verkoeijen, P. P. J. L. (2010). Latent variable modeling of cognitive processes in true and false recognition of words: A developmental perspective. Journal of Experimental Psychology: General, 139, 365–381. doi:10.1037/a0019301

Bouwmeester, S., Vermunt, J. K., & Sijtsma, K. (2007). Development and individual differences in transitive reasoning: A fuzzy-trace theory approach. Developmental Review, 27, 41–74. doi:10.1016/j.dr.2006.08.001

Brainerd, C. J., Holliday, R. E., & Reyna, V. F. (2004). Behavioral measurement of remembering phenomenologies: So simple a child can do it. Child Development, 75, 505–522. doi:10.1111/j.1467–8624.2004.00689.x

Brainerd, C. J., Holliday, R. E., Reyna, V. F., Yang, Y., & Toglia, M. P. (2010). Developmental reversals in false memory: Effects of emotional valence and arousal. Journal of Experimental Child Psychology, 107, 137–154. doi:10.1016/j.jecp.2010. 04.013

Brainerd, C. J., & Kingma, J. (1984). Do children have to remember to reason? A fuzzy-trace theory of transitivity development. Developmental Review, 4, 311–377. doi:10.1016/0273–2297(84)90021–2

Brainerd, C. J., & Kingma, J. (1985). On the independence of short-term memory and working memory in cognitive development. Cognitive Psychology, 17, 210–247. doi: 10.1016/0010–0285(85)90008–8

Brainerd, C. J., & Reyna, V. F. (1990). Gist is the grist: Fuzzy-trace theory and the new intuitionism. Developmental Review, 10, 3–47. doi:10.1016/0273–2297(90)90003-M

Brainerd, C. J., & Reyna, V. F. (1992). Explaining “memory free” reasoning. Psychological Science, 3, 332–339. doi:10.1111/j.1467–9280.1992.tb00042.x

Brainerd, C. J., & Reyna, V. F. (2002). Fuzzy-trace theory: Dual processes in memory, reasoning, and cognitive neuroscience. Advances in Child Development and Behavior, 28, 41–100. doi:10.1016/S0065–2407(02)80062–3

Brainerd, C. J., & Reyna, V. F. (2004). Fuzzy-trace theory and memory development. Developmental Review, 24, 396–439. doi:10.1016/j.dr.2004.08.005

Brainerd, C. J., & Reyna, V. F. (2005). The science of false memory. New York: Oxford University Press.

Brainerd, C. J., Reyna, V. F., & Ceci, S. J. (2008). Developmental reversals in false memory: A review of data and theory. Psychological Bulletin, 134, 343–382. doi:10.1037/0033–2909.134.3.343

Brainerd, C. J., Reyna, V. F., & Forrest, T. J. (2002). Are young children susceptible to the false memory illusion? Child Development, 73, 1363–1377. doi:10.1111/1467– 8624.00477

Brainerd, C. J., Reyna, V. F., & Howe, M. L. (2009). Trichotomous processes in early memory development, aging, and neurocognitive impairment: A unified theory. Psychological Review, 116, 783–832. doi:10.1037/a0016963

Brainerd, C. J., Reyna, V. F., Wright, R., & Mojardin, A. H. (2003). Recollection rejection: False-memory editing in children and adults. Psychological Review, 110, 762–784. doi:10.1037/0033–295X.110.4.762

Brainerd, C. J., Reyna, V. F., & Zember, E. (2011). Theoretical and forensic implications of developmental studies of the DRM illusion. Memory and Cognition, 39, 365–380. doi:10.3758/s13421–010–0043–2

Brainerd, C. J., Stein, L., & Reyna, V. F. (1998). On the development of conscious and unconscious memory. Developmental Psychology, 34, 342–357. doi:10.1037/0012– 1649.34.2.342

Brainerd, C. J., Yang, Y., Reyna, V. F., Howe, M. L., & Mills, B. A. (2008). Semantic processing in “associative” false memory. Psychonomic Bulletin & Review, 15, 1035–1053. doi:10.3758/PBR.15.6.1035

Bransford, J. D., & Franks, J. J. (1971). The abstraction of linguistic ideas. Cognitive Psychology, 2, 331–380. doi:10.1016/0010–0285(71)90019–3

Bruck, M., & Ceci, S. J. (1999). The suggestibility of children’s memory. Annual Review of Psychology, 50, 419–439. doi:10.1146/annurev.psych.50.1.419

Carneiro, P., Albuquerque, P., Fernandez, A., & Esteves, F. (2007). Analyzing false memories in children with associative lists specific for their age. Child Development, 78, 1171–1185. doi:10.1111/j.1467–8624.2007.01059.x

Carneiro, P., & Fernandez, A. (2010). Age differences in the rejection of false memories: The effects of giving warning instructions and slowing the presentation rate. Journal of Experimental Child Psychology, 105, 81–97. doi:10.1016/j.jecp.2009.09.004

Ceci, S. J., Papierno, P. D., & Kulkofksy, S. (2007). Representational constraints on children’s suggestibility. Psychological Science, 18, 503–509. doi:10.1111/j.1467– 9280.2007.01930.x

Ceci, S. J., Ross, D. F., & Toglia, M. P. (1987). Suggestibility in children’s memory: Psycholegal implications. Journal of Experimental Psychology: General, 116, 38–49. doi:10.1037/0096–3445.116.1.38

Deese, J. (1959). On the prediction of occurrence of certain verbal intrusions in free recall. Journal of Experimental Psychology, 58, 17–22. doi:10.1037/h0046671

Dewhurst, S. A., Pursglove, R. C., & Lewis, C. (2007). Story contexts increase susceptibility to the DRM illusion in 5-year-olds. Developmental Science, 10, 782–786. doi:10. 1111/j.1467–7687.2007.00592.x

Dewhurst, S. A., & Robinson, C. A. (2004). False memories in children: Evidence for a shift from phonological to semantic associations. Psychological Science, 15, 782–786. doi:10.1111/j.0956–7976.2004.00756.x

Drummey, A. B., & Newcombe, N. S. (2002). Developmental changes in source memory. Developmental Science, 5, 502–513. doi:10.1111/1467–7687.00243

Fazio, L. K., & Marsh, E. J. (2008). Older, not younger, children learn more false facts from stories. Cognition, 106, 1081–1089. doi:10.1016/j.cognition.2007.04.012

Furth, H., Ross, B., & Youniss. (1974). Operative understanding in children’s immediate and long-term reproductions of drawings. Child Development, 45, 63–70. doi:10.2307/1127750

Ghetti, S., Qin, J., & Goodman, F. S. (2002). False memories in children and adults: Age, distinctiveness, and subjective experience. Developmental Psychology, 38, 705–718. doi:10.1037/0012–1649.38.5.705

Grassi-Oliveira, R., Gomes, C. F. A., & Stein, L. M. (2011). False recognition in women with a history of childhood emotional neglect and diagnose of recurrent major depression. Consciousness and Cognition, 20, 1127–1134. doi:10.1016/j.concog. 2011.03.005

Holliday, R. E., Brainerd, C. J., & Reyna, V. F. (2011). Developmental reversals in false memory: Now you see them, now you don’t! Developmental Psychology, 47, 442–449. doi:10.1037/a0021058

Howe, M. L. (2008). Visual distinctiveness and the development of children’s false memories. Child Development, 79, 65–79. doi:10.1111/j.1467–8624.2007.01111.x

Howe, M. L., Wimmer, M. C., Gagnon, N., & Plumpton, S. (2009). An associative-activation theory of children’s and adults’ memory illusions. Journal of Memory and Language, 60, 229–251. doi:10.1016/j.jml.2008.10.002

Lampinen, J. M., Leding, J. K., Reed, K. B., & Odegard, T. N. (2006). Global gist extraction in children and adults. Memory, 14, 952–964. doi:10.1080/09658210601008957

Liben, L. (1974). Operative understanding of horizontality and its relation to long-term memory. Child Development, 45, 416–424.

Liben, L. (1977). Memory in the context of cognitive development: The Piagetian approach. In R. V. Kail & J. W. Hagen (Eds.), Perspective on the development of memory and cognition (pp. 297–331). Hillsdale, NJ: Lawrence Erlbaum Associates.

McDermott, K. B. (1996). The persistence of false memories in list recall. Journal of Memory and Language, 35, 212–230. doi:10.1006/jmla.1996.0012

McDermott, K. B., & Watson, J. M. (2001). The rise and fall of false recall: The impact of presentation duration. Journal of Memory and Language, 45, 160–176. doi:10.1006/jmla.2000.2771

Metzger, R. L., Warren, A. R., Price, J. D., Reed, A. W., Shelton, J., & Williams, D. (2008). Do children “DRM” like adults? False memory production in children. Developmental Psychology, 44, 169–181. doi:10.1037/0012–1649.44.1.169

Nelson, D. L., McGivney, V. M., Gene, N. R., & Janczura, G. A. (1998). Interpreting the influence of implicitly activated memories on recall and recognition. Psychological Review, 105, 299–324. doi:10.1037/0033–295X.105.2.299

Odegard, T. N., Holliday, R. E., Brainerd, C. J., & Reyna, V. F. (2008). Attention to global-gist processing eliminates age effects in false memories. Journal of Experimental Child Psychology, 99, 96–113. doi:10.1016/j.jecp.2007.08.007

Piaget, J. (1968). On the development of memory. Worcester, MA: Clark University Press. Piaget, J. (1970). Piaget’s theory. In P. Mussen (Ed.), Carmichael’s manual of child psychology (pp. 703–772). New York: John Wiley & Sons.

Piaget, J., & Inhelder, B. (1973). Memory and intelligence. New York: Basic Books.

Reyna, V. F., & Brainerd, C. J. (1994). The origins of probability judgment: A review of data and theories. In G. Wright & P. Ayton (Eds.), Subjective probability (pp. 239–272). New York, NY: Wiley.

Reyna, V. F., & Brainerd, C. J. (1995). Fuzzy-trace theory: An interim synthesis. Learning and Individual Differences, 7, 1–75. doi:10.1016/1041–6080(95)90031–4

Reyna, V. F., & Brainerd, C. J. (2011). Dual processes in decision making and developmental neuroscience: A fuzzy-trace model. Developmental Review, 31, 180–206. doi:10.1016/j.dr.2011.07.004

Reyna, V. F., & Ellis, S. C. (1994). Fuzzy-trace theory and framing effects in children’s risky decision making. Psychological Science, 5, 275–279. doi: 10.1111/j.1467– 9280.1994.tb00625.x

Reyna, V. F., Holliday, R., & Marche, T. (2002). Explaining the development of false memories. Developmental Review, 22, 436–489. doi:10.1016/S0273–2297(02)00003–5

Reyna, V. F., & Kiernan, B. (1994). Development of gist versus verbatim memory in sentence recognition: Effects of lexical familiarity, semantic content, encoding instructions, and retention interval. Developmental Psychology, 30, 178–191. doi:10.1037/0012– 1649.30.2.178

Reyna, V. F., & Kiernan, B. (1995). Children’s memory and metaphorical interpretation. Metaphor and Symbolic Activity, 10, 309–331. doi:10.1207/s15327868ms1004_5

Reyna, V. F., & Lloyd, F. (1997). Theories of false memory in children and adults. Learning and Individual Differences, 9, 95–123. doi:10.1016/S1041–6080(97)90002–9

Reyna, V. F., Mills, B., Estrada, S., & Brainerd, C. J. (2007). False memory in children: Data, theory, and legal implications. In M. P. Toglia, J. D. Read, D. F. Ross, & R. C. L. Lindsay (Eds.), Handbook of eyewitness psychology (pp. 479–507). Mahwah, NJ: Erlbaum.

Roediger, H. L., III & McDermott, K. B. (1995). Creating false memories: Remembering words not presented in lists. Journal of Experimental Psychology: Learning, Memory, and Cognition, 21, 803–814. doi:10.1037/0278–7393.21.4.803

Roediger, H. L., III, Watson, J. M., McDermott, K. B., & Gallo, D. A. (2001). Factors that determine false recall: A multiple regression analysis. Psychonomic Bulletin & Review, 8, 385–407. doi:10.3758/BF03196177

Ross, D. F., Marsil, D. F., Benton, T. R., Hoffman, R., Warren, A. R., Lindsay, R. C. L., & Metzger, R. (2006). Children’s susceptibility to misidentifying a familiar bystander from a lineup: When younger is better. Law and Human Behavior, 30, 249–257. doi:10.1007/s10979–006–9034-z

Schacter, D. L., Israel, L., & Racine, C. (1999). Suppressing false recognition in younger and older adults: The distinctiveness heuristic. Journal of Memory & Language, 40, 1–24. doi:10.1006/jmla.1998.2611

Shiffrin, R. M. (2003). Modeling memory and perception. Cognitive Science, 27, 341–378. doi:10.1016/S0364–0213(03)00027–2

Sloutsky, V. M., & Fisher, A. V. (2004). Induction and categorization in young children: A similarity-based model. Journal of Experimental Psychology: General, 133, 166–188. doi:10.1037/0096–3445.133.2.166

Snodgrass, J. G., & Corwin, J. (1988). Pragmatics of measuring recognition memory: Applications to dementia and amnesia. Journal of Experimental Psychology: General, 117, 34–50. doi:10.1037/0096–3445.117.1.34

Stadler, M. A., Roediger, H. L., & McDermott, K. B. (1999). Norms for words that create false memories. Memory & Cognition, 27, 494–500. doi:10.3758/BF03211543

Sugrue, K., & Hayne, H. (2006). False memories produced by children and adults in the DRM paradigm. Applied Cognitive Psychology, 20, 625–631. doi:10.1002/acp.1214

Toglia, M. P., Neuschatz, J. S., & Goodwin, K. A. (1999). Recall accuracy and illusory memories: When more is less. Memory, 7, 233–256. doi:10.1080/741944069