IV Memory in children

• Previous Page
• Table of Contents
• Next Page

4.1 General memory processes: “Children cannot report what they cannot remember”

A child’s ability to describe events that they have experienced or witnessed will be a function of developmental, social, and cognitive processes. Here, we focus on a key cognitive consideration, memory, as we highlight some important foundations of what can be expected in recall.

Memory is a subjective experience. Most memory researchers agree that the purpose of the memory system is to understand and interpret the world around us and to help us make predictions about the future (e.g., Howe, 2011). Thus, the primary purpose of memory is not to retain precise and detailed information about prior experiences, but instead to remember them more generally. Naturally, this tendency to extract meaning creates a tension with the legal system wherein precision is often required of witnesses. For each personal experience, a person brings their prior experiences, thoughts, emotions, attention, cognitions, and motivation. As a result, the same event attended by many people will be experienced differently for each of them. Consider a traditional wedding ceremony. A fashion designer may focus on what the attendees are wearing and will later be able to provide rich descriptions of materials and designs. Conversely, a woodworker may notice the building structure and chairs but have little recollection of any of the clothing worn by participants. A child attending a wedding for the first time may be most intrigued by the unusual behaviours of his family members and be concerned about why some are crying. Because of the individual variability in experiences of the same event, each attendee who attended the same event will be able to provide information, but the nature and detail of that information will be very different. As a result of the varying experiences people bring to an event, memory is often described as a subjective experience, not a direct record of an experience. A key to understanding what children will be able to remember, and how forensic interviewers should account for children’s memory capabilities during interviews, is to recall the fundamental principle that “children cannot report what they cannot remember” (Ornstein, Larus, & Clubb, 1991).

Memory is reconstructive, fallible, and malleable. When a person tries to recall a memory, they will typically retrieve some of the stored pieces of a memory and this will be combined with general knowledge about similar events as well as current thoughts, emotions, among other factors (Howe, Knott, & Conway, 2018). As a result of this process of filling in the blanks of the original memory (“reconstruction”), as well as normal processes of forgetting, memory is fallible and prone to error (see Howe et al., 2018). Further, memory can change over time. Each time we reactivate a memory by thinking about it or discussing it with others, it is subjected to potential sources of contamination and may be forever changed as a result. Thus, memory is also malleable.

At its most basic description, memory has three phases: encoding, storage, and retrieval. Encoding is the taking in, or experiencing, of the event. Storage is the period of time after encoding until an attempt to recall the memory (also sometimes called the delay period). Retrieval is the attempt to recall the memory. We refer to these processes as we describe the understanding of children’s ability to report on their prior experiences.

Some of the original memory theorists to engage in research on children’s testimony articulated several relevant themes of memory that can be used to understand children’s memories of their experiences (see Baker-Ward & Thomas, 2021; Ornstein, Baker-Ward, Gordon, & Merritt, 1997). These themes provide a useful structure for interpreting children’s evidence:

1. Not everything gets into memory. For a detail to be encoded in memory, a person must attend to that detail. If a person does not encode a detail, it is not possible to retrieve it. Children may attend to different details than adults do and there are many reasons why a detail may not be encoded (e.g., prior knowledge of an event, stress, attention). Just because something was experienced does not mean that it was encoded. Thus, specific questions might target recall of information that was never encoded in the first place (one of the reasons that open-ended questions are so strongly advised, see Sections III and VI).
2. What gets into memory may vary in strength. Many of the same factors that influence whether or not details enter memory will also influence the strength of the memory representation. Factors like age (see Section 4.2) and experience with an event (see discussion of fixed versus variable details in Section 4.7) will influence the strength of a memory representation.
3. The status of information in memory changes over time. There are many events that can influence the status of memory. These events can be internal (e.g., thoughts, emotions, fantasies) or external (e.g., encountering new information) to the rememberer. Imagine a negative interaction with a neighbour that made you grumpy and led to a perception that the neighbour was unkind. When you later learn that the neighbour has just lost his wife, you might reconsider your original recollection of the experience in light of this new information. In re-thinking the event, you may recall signs that are consistent with your new knowledge, like that your neighbour looked tired or had bloodshot eyes. Of course, the accuracy of new information can vary considerably, but the impact on memory may be the same. Further, memory can be strengthened through rehearsal or practice remembering an event (e.g., talking with others about the experience, looking at pictures; Davis & Loftus, 2007; Howe et al., 2018), but such conversations may also serve as potential memory interference if new or conflicting information is encountered (see Section 4.3). Conversely, memory is generally weakened by time, with delays between the experience and recall typically resulting in a loss of information (see Section 4.5). Because many forensic interviews will take place long after the experience that brings a child to the justice system, some memories will fade and normal memory processes allow some information to become inaccessible. We cannot remember every detail that has been encoded throughout our lives and it is normal for many experiences to fade in memory. Other memory details may not have been experienced as particularly salient or notable (e.g., Harvey et al., 2025). If a person does not take particular note of an event or detail, the memory for it is likely to decay faster.
4. Retrieval is not perfect. This theme has been the subject of perhaps the largest body of research (see Section VI on investigative interviewing) and has clearly demonstrated that memory retrieval conditions have considerable impact on what is reported. In addition to questioning techniques, there is also the question of which memories are retrieved, given the array of available memories that could have been retrieved. This is a complex issue, and we are not in a position to judge whether or not there are more memories “in there” for retrieval. However, knowing that people do not remember everything in a single retrieval attempt–no matter how thorough the attempt–means that new information (i.e., reminiscence; La Rooy, Pipe, & Murray, 2005) should not be discounted simply because it was not reported initially.
5. Not everything that can be retrieved is reported. Similar to the prior theme, even if a memory is retrieved and a child is actively thinking about a particular detail, that does not mean that the information will inevitably be reported to a questioner. There are many circumstances under which people will retrieve information in memory and decide not to report it to a conversational partner. Details may be deemed unimportant, not well understood, or perhaps there are reasons that make a child reluctant to report information. Attempts to overcome such vetted reporting of information have been incorporated into forensic interviewing guidelines (e.g., an instruction that the interviewer is “ignorant” or does not know the circumstances of the event and thus all information is relevant and important to share).

Finally, a sixth theme was added (see Crossman et al., 2002) that focuses on reports of details that were never experienced:

6. Not everything that is reported was experienced. This theme will be expanded upon further in Section 4.8 (Children’s Suggestibility) and is a crucial consideration when evaluating children’s evidence. Importantly, it is not only suggestive questioning that can lead children to report false accounts: even everyday conversations that involve hearing another person’s version of an event can contribute to children (and adults!) reporting things that never happened (e.g., Principe, 2021; Principe & London, 2022; Principe, Ornstein, Baker-Ward, & Gordon, 2000). Such conversations may be particularly problematic when they take place with a parent (Lawson, Rodriguez Steen, London, & Colemen, 2024; Principe et al., 2013, 2017) and there is evidence that parents who question their children about abuse do so in a way that may inadvertently facilitate some false reporting (e.g., Korkman, Juusola, & Santilla, 2014).

4.2 Age as a factor in memory

A general conclusion about age differences in memory reports is that children of all ages can be highly accurate when recounting past events, but younger children typically provide less information than older children (e.g., Jack, Leov, & Zajac, 2014). The reduced volume of information is likely attributable to the relatively weaker memory traces in younger children’s less developed memory systems (Baker-Ward, Gordon, Ornstein, Larus, & Clubb, 1993; Ornstein, 1995) but may also be a result of less developed verbal skills (Jack et al., 2014) and the resulting capacity to describe their experiences. Caution is warranted in questioning young children about things they may not remember (see Sections 4.8 and VI). Nonetheless, an important take home message about age differences in memory reports is that even children as young as 3 years of age are, under the right circumstances, able to provide coherent narratives about their prior experiences (e.g., Jones & Krugman, 1986).

4.3 Prior knowledge and memory

Because the events that bring children to court are often unusual for children to experience, they may not fully understand the nature of the event. For instance, most parents will not teach their children about sexual activity in explicit detail when they are very young. If the event is unusual and the child does not have a knowledge structure with which to interpret it, children may not understand events they experience. When a child does not understand the event, it is more difficult to encode an experience because they do not have an existing knowledge structure that helps them to identify familiar concepts. Thus, when a child has prior knowledge, it typically facilitates remembering (e.g., Clubb et al., 1993; Ornstein et al., 2006).

Importantly, the influence of knowledge on recall is not restricted to only that which is acquired prior to an experience. There is also evidence that knowledge changes can lead to a “reworking” of memories to facilitate or increase compatibility with current knowledge (Baker-Ward, Ornstein, & Starnes, 2009; Baker-Ward & Thomas, 2021; Greenhoot, 2000).

Unfortunately, although prior knowledge has a clear benefit to remembering an event, it is also likely to contribute to errors in recall after memory fades. In such cases, children’s memory for what they experienced may conflict with their knowledge and expectations about similar experiences, and memory constructive processes lead to inferences about what likely happened (Ornstein, Ceci, & Loftus, 1998). Experiencing a similar event repeatedly can create prior knowledge and cause children to report details that are consistent with the event in general but may be inaccurate when describing a specific occurrence (see Section 4.7). Such instances of memory contamination or interference can be indistinguishable from actual memory.

4.4 Impact of anxiety/stress on memory

When considering the influence of stress on memory in a forensic context, it is first important to establish that children who are involved in justice processes will have had varied reactions to the events that bring them in contact with investigators. Some children will be traumatized by what they have experienced, others will be upset, and still others may not have experienced stress or do not understand why they are expected to have experienced stress. Here, we discuss the influence of anxiety on memory, but we do not presume that all children will have experienced anxiety or stress prior to giving evidence.

Research on children’s memory for stressful events is difficult to conduct for ethical reasons. However, creative researchers have found ways to study children’s memory for shootings (e.g., Pynoos & Nader, 1989), natural disasters (e.g., Bahrick, Parker, Fivush, & Levitt, 1998; Fivush, Sales, Goldberg, Bahrick, & Parker, 2004), medical procedures (Ornstein, Gordon, & Larus, 1992), emergency room visits (see Peterson, 2012), and abuse (Alexander et al., 2005), among other stressful experiences. The natural variability in the nature of the experiences and how children responded emotionally to the experience makes drawing firm conclusions challenging. For instance, even if we could determine precisely how stressful an experience was for an individual child (which we cannot), we would have to then relate that level of stress to the very particular details of the experience we sought to evaluate. This degree of control requires laboratory experiments that would clearly violate ethical standards. Despite these difficulties, there are still some valuable highlights that can be drawn from this body of work.

Stressful events may be more salient than non-stressful events, which may make the former more likely to be remembered (e.g., Fivush, 2002; Ornstein, 1995). However, stressful events are still subject to the same interference and contamination and basic memory processes as all other memories. Thus, while they may be remembered better due to their salience, they are not immune to the same memory problems as non-stressful events. Further, very high levels of stress might result in memory impairment (e.g., Bahrick et al., 1998; Fivush et al., 2004; Goodman & Quas, 1997; Morgan & Southwick, 2014; Pynoos & Nader, 1989), possibly from the narrowing of attention that high levels of stress may bring.

In a series of studies examining children’s memory for emergency room visits up to five years after the emergency, Peterson concluded that children typically remember the core features of a stressful event well, and for a long period of time, but that although children were quite accurate, the amount of information deteriorated with time (see Peterson, 2012 for a review; Goodman, Quas, Goldfarb, Gonzalves, & Gonzalez, 2019). This finding is generally in keeping with the broader literature on children’s memory for stressful events: it is possible and perhaps likely that children will remember the central features of a stressful event, but just like with everyday memories, the memories are subjected to the same processes of forgetting and interference.

4.5 Long-term memory

Because delayed reporting of some crimes (e.g., child sexual abuse) is normative (see Section V), and because justice will often move slowly, many cases that bring children to the justice system will involve a lengthy delay between the experience and the recounting of that experience. Thus, it is important to understand the influence of a long delay on recall. Several processes take place in the delay period, the most prominent of which are forgetting and interference. It is normal and expected that a person will forget things over time. It is also normal that with a long delay period, there will be increased opportunities to encounter information that interferes with the original memory for the event (e.g., conversations, media, thoughts, fantasies).

What exactly a person will remember depends on many factors, including several that have been discussed earlier (e.g., age at encoding, event salience, understanding of the event, rehearsal of the event in memory). A general conclusion that can be made is similar to the conclusion that was made about memory for stressful events: Children are capable of providing coherent accounts of prior experiences after a long delay, but generally the volume of details fades with time and delays provide greater opportunities for memory contamination (Peterson, 2002).

Adults recalling events from when they were children (Historic Child Sexual Abuse: HCSA). It is worth discussing a particular circumstance related to long-term memory that has seen an increase in the justice system over the past several years: historic child sexual abuse cases (Woiwod & Connolly, 2017). Sometimes a result of legal changes, other times a result of evolving societal views, there has been an increase in the number of people who, as adults, report sexual abuse they experienced as a child. Such cases raise interesting memory issues for consideration. While memory for sexual abuse should not differ qualitatively from memory for other salient and stressful events, the circumstances surrounding sexual abuse may introduce unique elements. Researchers have cogently argued that we should be highly skeptical of the content of memories that are reported to have been forgotten for many years, only to be retrieved with “memory work” or the help of a motivated therapist (e.g., Howe, 2012; Howe et al., 2018), and of reports of memories that were reported to be “repressed” and unavailable to access for many years (see McNally & Geraerts, 2009). These reports typically include circumstances that raise considerable questions about the reliability of the memories. However, there is also evidence that some people can remember salient events that occurred a long time ago with accuracy (e.g., Goldfarb, Goodman, Larson, Eisen, & Qin, 2019). The issue is complex and beyond the scope of the current report, but interested readers are referred to several important discussion papers (Dodier, Otgaar, & Manguilli, 2024; Goodman et al., 2003; Ornstein et al., 1998; Otgaar, Howe, & Patihis, 2022; Patithus, Ho, & Loftus, 2021).

4.6 Source monitoring

Source monitoring is a term used to describe the process by which we decide where our knowledge and memories came from (Johnson et al., 1993). This process can help us differentiate memories of actual experiences from things we have read in books, seen on television, heard others say, or imagined. Being able to identify the source of information can help us make judgments about its reliability. Source monitoring decisions are made based on memory characteristics such as sensory details like sights and sounds, spatial and temporal information (where, when), and emotions (Johnson et al., 1993). Memories with plentiful sensory characteristics are more likely to be experienced than imagined (Foley & Johnson, 1985). Perceptual characteristics can be used to judge which of two speakers told a joke, or what news outlet delivered the information. It can also be used to decide whether a memory detail actually happened or was suggested in an interviewer’s question. Making source monitoring decisions is often not easy, and because these skills take time to develop, it is even more difficult for children (Roberts, 2002; Roberts & Powell, 2001). The source monitoring process is also not always purposeful. In fact, many source attributions are made without awareness of decision making (Lindsay, 2002).

The greatest developments in children’s abilities to make decisions about the sources of their knowledge and memories happen between 3 to 8 years of age (Gopnik & Graf, 1998; Lindsay, 2002; Poole & Lindsay, 2001; Roberts, 2002). However, research has identified further refinements in source monitoring ability into adolescence and early adulthood (Sprondel, Kipp, & Mecklinger, 2011; Raj & Bell, 2010). Like many other developments, the relationship between age and source monitoring ability is not straightforward because memory for source is not a single skill (Lindsay, 2002). The ability to make source decisions relies on other cognitive skills that change and develop alongside age, such as the ability to integrate new information with existing knowledge, and the awareness of the representational nature of the mind; that is, that other people may have different beliefs, intentions, and access to knowledge (theory of mind, Section 2.1; Flavell, 1999; Gopnik & Graf, 1988; Welch-Ross, 1995). For example, compare a 4- and 7-year-old child trying to decide how they already knew what was in the wrapped-up birthday present for Mom. The younger child has an underdeveloped sense of how knowledge is acquired, and he exclaims, “I just know it!” The older child is aware that she has not always known the answer and has not seen the gift. She accurately reasons that her dad told her what he got. Source monitoring ability will differ from child to child, and from situation to situation (Lindsay, 2002; Roberts, 2002).

Although there is no need to correctly judge the source of many everyday memories, in the legal system it is often critical. Children have to separate what they have experienced themselves from other sources of memory contamination, including what others have said. They also may have to discriminate between multiple instances of similar events, discussed next.

4.7 Memory for repeated events

Similar experiences that occur routinely (or more than once) are often referred to in the research literature as repeated events. In children’s lives, these experiences might include going to school, partaking in hobbies, attending birthday parties and–for some children–abuse, maltreatment, and other victimization. In these cases, legal professionals often need to elicit information from children about individual occurrences to aid investigation and lay appropriate charges.

Repeated events are particularly prone to source monitoring confusions because the occurrences have similar perceptual features (Lindsay, 2002; Lindsay, Johnson, & Kwon, 1991). For example, the events may have the same pattern, occur in the same place, or on the same day of the week. Of course, occurrences of repeated events vary in their degree of similarity to one another, as in the following two cases: In Case A, a young child’s stepfather touches her inappropriately during a bathtime game every Friday night when her mother works an evening shift. In Case B, a boy’s uncle engages in a variety of opportunistic offences including exposure, production of child sexual exploitation material, and fondling when he can get the boy alone, and this has happened in a variety of locations. In Case A, the characteristics are very similar and there are few cues to differentiate the occurrences. In Case B, the nature of the abuse differs, as well as the location and temporal details. One laboratory study demonstrated that 5- to 9-year-old children were more accurate in their descriptions of an individual episode of a repeated event when the episodes contained more, rather than fewer, differences (Danby, Sharman, Brubacher, & Powell, 2019). This finding parallels early source-monitoring research showing that 4- and 6-year-old children (and in some cases, adults too) had greater difficulty discriminating between highly similar versus more distinct sources (Lindsay et al., 1991, Experiments 1-2). As sources increase in distinctiveness, there are more cues to help separate them in memory.

How repeated experiences are organized in memory. Memory for repeated experiences is complex and often represented by a script, a temporally organized framework that includes the typical components of an event (Schank & Abelson, 1977). For example, your script for going to a restaurant might include choosing food and drink, consuming it, and paying for it. Our ability to rely on memory scripts develops when we are young. Even 2.5-year-olds show awareness of similar event structures presented repeatedly (Bauer & Fivush, 1992). Scripts help us make meaning of our experiences and make future encounters more predictable. For this reason, a rudimentary script can develop after just a single experience. For example, after the first day of kindergarten, children were already able to report “what happens” (Fivush, 1984). With repeated experience, children extract a more complete script–a general event representation–which becomes more detailed and sophisticated over time (Farrar & Boyer-Pennington, 1999; Farrar & Goodman, 1990, 1992; Hudson et al., 1992). You can tell when a child is reporting a script because they use timeless present-tense language and impersonal “you” pronouns (e.g., in a birthday party script, “You eat cake and open presents.”), as well as optional and conditional statements (e.g., “You can sing Happy birthday if you want to. You might get a treat bag.”).

Some event details occur in a similar way each time and are represented at the script level (often referred to in the literature as fixed details). Memory for fixed details across repeated events (e.g., the perpetrator’s identity) is very strong, and children and adults are rarely confused about these details. They are resistant to misleading suggestions about them (Connolly & Lindsay, 2001). However, fixed details are associated with a unique type of error when they are not present in an event, such as if the detail occurred every time except for one incidence. In one formative study, 5- to 6- and 9- to 10-year-old children watched slide sequences of a visit to McDonald’s with details missing (Erskine, Markham, & Howie, 2001). (It was previously confirmed that children in this age range would have strong scripts for going to McDonald’s). For some children, the slide sequences omitted three central activities (arriving by car, waiting in line to order, and paying for the food). A week later, children’s memories were tested. Very few (about 3%) erroneously reported any missing details in response to an open-ended invitation to report everything they remembered about what happened in the slides (this is a feature of desirable open-ended questions–see Section 6.1). In response to specific yes/no questions, however, children incorrectly agreed, with confidence, that they had seen the central details, and this was more common among the younger children (older children agreed to about 1.5 of the 3 central details, on average, while younger children agreed to more than 2.5 of the 3). In summary, children’s memories for the typical (fixed) details are very strong, although they may err in claiming that a routinely-experienced true detail happened during a time when it was absent. Further, these results are also a good reminder that children make far more errors in response to closed than open questions (Brubacher, Peterson, La Rooy, Dickinson, & Poole, 2019; Section 6.1).

In contrast to fixed details, scripts also contain representations of details that vary predictably (e.g., clothing worn, time of day, type of food chosen). These variable details are highly prone to source monitoring errors (Powell et al. 1999; See section 4.6 for more details) and suggestion (Connolly & Lindsay, 2001) because the details could plausibly have occurred any time (e.g., the red shirt could have been worn the day the hamburger was chosen or the day the hotdog was chosen). This makes children with repeated experience generally more suggestible than children with a single experience (Connolly & Lindsay, 2001; Connolly & Price, 2006) and the former may also be perceived as less credible (Connolly, Price, Lavoie, & Gordon, 2008). The most common mistake that children make when recounting repeated experiences is to confuse the occurrence in which true things actually happened (Powell, Roberts, Ceci, & Hembrooke, 1999). That is, a child will remember what happened, but not when it happened. If children’s accuracy about what happened during a repeated event is judged based on their ability to attribute details to the correct occurrence in which they were present, children with repeated experience will be less accurate than children with a single experience. However, if accuracy is judged by examining experienced details (irrespective of when they occurred), children with repeated and single experience will be equally accurate (Woiwod et al., 2019).

Script for a visit to a McDonald’s restaurant.

Our understanding of children’s memories for repeated events is largely driven by laboratory studies (Connolly, Gordon, Woiwod, & Price, 2016; Connolly & Gordon, 2014; Connolly & Price, 2006; Powell et al., 1999; Powell, Thomson, & Ceci, 2003; Powell & Thomson, 1997, 2003; Price, Connolly, & Gordon, 2006, 2016; Price & Connolly, 2013; Price & Goodman, 1990). This is because staging events allows researchers to have baseline truth and thus make conclusions about children’s accuracy. Information about what actually happened is rarely available in real life cases of repeated abuse. In one field study, however, there was corroborating evidence in the form of audiotapes and photographs of repeated sexual abuse (Bidrose & Goodman, 2000). Four girls, aged between 8 to 15 years at the time they reported their abuse, worked as cleaners for a man who engaged them in sexual activities with himself and other men. Together, the girls made 246 allegations, all of which were either supported by the evidence (79%) or could not be confirmed (21%). None of the allegations were disconfirmed by the evidence. Research of this nature provides further support for the notion that–while confusion across occurrences of repeated events is common and normal–memory for the central elements of what happened is quite accurate.

Memory for details that deviate from the event script. Sometimes details occur during a repeated event that are not part of the typical event script. These could be atypical but not implausible (e.g., someone playing live music at a restaurant) or they could be quite unusual (e.g., someone wearing a gorilla costume at a restaurant). The more implausible, the better the details will be recalled (Davidson, 2006; Davidson & Hoe, 1993). However, research with 4- to 8-year-old children who participated in four similar activity sessions found that even mundane deviations from the event script were reported in response to open-ended questions, but only by the older children (Brubacher, Glisic, Roberts, & Powell, 2011). Half of the 7- to 8-year-old children reported these deviations, while almost none of the younger children did. The older children who reported the atypical details were also very accurate in attributing them to the correct occurrence.

The effect that a deviation has on memory for the instance that contains it may depend on a variety of factors. In some research, a deviation improved recall memory of 8-year-olds for all details contained in the deviation instance (Connolly et al., 2016, Experiments 1 and 2). The deviation was a (staged) argument between a magician and another researcher and occurred towards the end of the last of four event sessions. Children were found to recall all event details of that fourth event session containing the argument at higher rates than the prior event sessions. In another experiment (Connolly et al., 2016, Experiment 3), 6- to 11-year-old children watched the same magic show, but for some children, a deviation occurred at the beginning of the last session–another magician requesting help. After help was given, the first magician went on with the show. For some children, everything proceeded as normal. For others, the interruption caused the magician to fumble through the show, making mistakes. The first type of deviation is called discrete: it does not affect other parts of the event. The second deviation is called continuous: it affects how the rest of the event unfolds. In this experiment, the continuous deviation led to better memory for the whole instance containing the deviation, and the other instances, compared to the discrete deviation and no deviation. This was especially true for the 6- to 8-year-old children.

Recommendations for questioning children about repeated experiences. The research literature has generated some recommendations for talking to children and adults about repeated events. These recommendations include allowing children to provide their scripts for “what usually happens” if they are inclined to do so, adopting their names (labels) for individual occurrences, and ensuring that the language used in the question is appropriately matched to the type of information the interviewer seeks (Brubacher, Powell, & Roberts, 2014; Brubacher & Earhart, 2019).

Three laboratory studies have demonstrated that children may provide more information without a corresponding decrease in accuracy if they are questioned about their event scripts (i.e., what usually happens) early in an interview (Brubacher, Roberts, & Powell, 2012; Brubacher et al., 2018; Connolly & Gordon, 2014). Children with multiple abusive experiences are more likely to naturally provide information about “what usually happens” than describe specific occurrences (Brubacher et al., 2013). Interviewers can attempt to direct children’s attention to individual episodes by adopting their labels (Brubacher et al., 2012; 2013; Brubacher & La Rooy, 2014). Children’s labels for episodes (e.g., “The time in the tent”, “the day I was home sick from school”) often arise spontaneously in their responses to open-ended questions and when they provide their event scripts (Brubacher et al., 2013, 2018). When no labels arise, interviewers are advised to ask children about the last time and first time because these are generally best recalled in laboratory settings (Connolly et al., 2016; Powell et al., 2003; Roberts et al., 2015). Yet, interviewers are cautioned that the first time may be very long ago for some children, or difficult to identify in cases of grooming.

Interviewers can query about children’s scripts or episodes by altering the way in which they phrase their questions; when interviewers use generic language (“What happens after school?”) children typically provide script information (i.e., what usually happens). In contrast, episodic prompts (e.g., “What happened after school yesterday”) are more likely to obtain details about a specific episode (Brubacher, Glisic, Roberts, & Powell, 2011; Brubacher et al., 2012; Schneider, Price, Roberts, & Hedrick, 2011).

When repeated experiences evolve over time: Grooming. The term grooming is often used with child sexual abuse but can refer to any situation in which someone builds a relationship with another person over time, in order to manipulate, exploit, and/or abuse them (Berens, Bruer, Schick, Evans, & Price, 2023; Craven et al., 2006; Williamson, 2022). Typically, perpetrators seek victims who are vulnerable due to a lack of support networks, mental health challenges, or other reasons. Their strategies include befriending the child, creating opportunities and excuses for the inappropriate behaviours, and ensuring the child’s continued cooperation through threats, secrecy pacts, bribery, or emotional manipulation (Berliner & Conte, 1990; De Santisteban et al., 2018; McAlinden, 2012). They also make efforts to distance or isolate the victim from any existing sources of support (Ringenberg et al., 2022). In sexual abuse, there are a number of stages in the grooming process, which may or may not be employed in each case. For example, in intrafamilial grooming, there is no need to befriend the child. Across the literature, different research groups have characterized the stages in different ways (see Bennett & O’Donahue, 2014, for review). Here we provide an amalgamated overview of grooming stages:

Many grooming behaviours, particularly in early phases, are legal and may appear innocuous but are aimed to prepare the child for abuse and frequently increase in severity over time (Bennett & O’Donahue, 2014; Winters & Jeglic, 2017). They might include playing a ‘game’ that initially does not include sexual contact, which evolves into sexualized horseplay. If the victim expresses discomfort, the grooming process may slow down, or the nature of coercion may change. As such, the grooming relationship will include positive and negative experiences and emotional manipulation (Ringenberg et al., 2022).

The emergence of the internet and ubiquitous nature of personal mobile devices resulted in increased ease of access for offenders to vulnerable and isolated victims. Although the grooming process online is similar in many ways to the process in real life, researchers have identified several differences (Berens et al., 2023; Ringenberg et al., 2022). For example, during the friendship forming stage online, offenders need to determine whether the victim is a real child or an undercover law enforcement officer. During this risk assessment period (O’Connell, 2003), which often occurs very early in the conversation (Black et al., 2015), groomers ask questions about the child’s location and environment (Williams et al., 2013). The sexualization of the relationship naturally takes a different form online compared to real life. While some offenders are very direct in introducing sexual language very early on (Kloess et al., 2017; Winters et al., 2017), others indirectly introduce such topics by initially giving compliments that are increasingly sexual in nature. It is also critical to recognize that grooming may take place both online and in-person. Berens et al. (2023) found that many grooming cases involving both online and in-person grooming involved an authority figure in a community organization. They argued that education and policies that prevent inappropriate online contact with children may help to prevent such behaviours from progressing. It is clear that further education about the grooming process is required for all adults who have contact with children (see Berens et al., 2023).

Because grooming inherently involves contact over time, it can be considered a repeated event. It also often occurs within the context of another repeated event (e.g., over the course of gymnastics classes, the coach befriends and grooms the child). This makes the boundaries of the event difficult to recognize (e.g., what defined “the first time?”), and may affect how the abuse script develops and is represented in memory. Thus, grooming is very different in nature from the structure of repeated events that have been the focus of most research. As a result, we have a limited understanding of how the grooming process affects memory for the series of events (Deck et al., 2025).

4.8 Suggestibility of children’s memory

Concerns about the suggestibility of children’s memory arose from some high-profile cases in the 1980s and 1990s, where interviewers obtained stories of extreme and implausible satanic and ritualistic mass child abuse (Bruck & Ceci, 1999; Ceci, Ross, & Toglia, 1987). Over time, it became clear that the interviewing methods may have driven children to report things that had not happened (Bruck & Ceci, 1999; Lamb, Orbach, Hershkowitz, Esplin, & Horowitz, 2007). These interviews included asking children to speculate what might have happened, selectively reinforcing desirable responses, bribery, peer pressure, and criticizing or disagreeing with the child (Garven et al., 1998, 2000).

To contaminate children’s reports, suggestions need not be coercive or explicit (Poole, Dickinson, & Brubacher, 2014; Price & Ornstein, 2022) like the well-known leading tag question (e.g., You know what that is, don’t you?). In fact, any question can contain suggestive or leading information if it includes details not previously raised by the child. For example, imagine that an interviewer asked a child, “Tell me what happened when Leo hurt you,” but the child previously only said he got hurt playing with Leo. What was an unintentional injury when Leo’s knee connected with the interviewed child’s forehead has now turned into malicious battering in the interviewer’s report. A child’s initial response to a leading or misleading question may be one of acquiescence (the child does not want to contradict the interviewer), ignorance (the child may not even attend to the incorrect information), or contradiction (the child corrects the interviewer). If the child does not correct the interviewer, the information persists in the record. Whether or not it changes the child’s actual memory for the event may depend on a variety of factors.

Most experts agree that central and significant details are resistant to suggestion as are details that occur the same way in repeated events. Details that are peripheral, or those that vary, can become confused and prone to error. For example, in one study, 6- to 7-year-olds with repeated experience were more suggestible than those with a single experience, especially when the varying alternatives were highly associated (e.g., the sticker children received at every session was always a mode of transportation). This was not true for the 4- and 5-year-olds (Connolly & Price, 2006).

In one of the most well-known studies on children’s suggestibility, 3- to 8-year-olds participated in science demonstrations and their parents read them stories about those demonstrations three months later, three times on consecutive days (Poole & Lindsay, 2001). Some of the details in the stories had happened in the science demonstrations, and others had not. The children were interviewed about the science demonstrations 1) immediately after the demonstrations, 2) soon after their parents read the stories, and 3) again a month later. In the immediate interview, open-ended (“Tell me everything that happened when you were playing with Mr. Science.”) and sensory focus (e.g., “Tell me how everything looked.”) prompts were asked. Children had not been exposed to misinformation and their responses were highly accurate. In the second and third interviews, the open-ended and sensory focus prompts were asked along with ten direct question pairs that asked about the experienced and non-experienced (suggested) details. In the second interview, after exposure to misinformation from their parents, 21% of what children reported to the open-ended and sensory focus prompts was the non-experienced suggested details. This percentage dropped to 10% in the third interview, as might be expected with overall recall decrease over time. There were no age differences in the children’s susceptibility to including suggested details in their recall reports to open-ended and sensory focused prompts. Age differences did emerge when considering the direct questions. Here, younger children were more prone to agree that non-experienced details happened, thus falsely reporting suggested details. Young children are generally more prone than older children and adults to accepting information suggested to them by others (Ceci et al., 1987), but there are situations in which there are no differences or even developmental reversals (e.g., Brainerd, Reyna, & Ceci, 2008).

Finally, the children were asked source monitoring questions (Section 4.6) about all of the items they agreed had happened, to give them the opportunity to reject any items. They were told that some of the things they heard in the stories really happened when they visited Mr. Science, but other things were only in the story. Children were told to say “no” if they didn’t remember something or it was only in the story, (e.g., “Did Mr. Science really show you how to make an eyedropper go up and down in a bottle?” p. 31). These questions did not benefit the 3- and 4-year-old children greatly, but they did decrease the false reports of many of the older children. Nevertheless, there remained even some older children who still claimed they really experienced the suggested events even after interviewers warned them that some details only came from the story, gave them permission to say “no”, and gave them the opportunity to reject their earlier answer.

• Previous Page
• Table of Contents
• Next Page