Advertisement

Children’s Eyewitness Lineup Accuracy One Year Later: The Role of Social Support and Working Memory Capacity

  • Liana C. Peter-HageneEmail author
  • Kelly C. Burke
  • Bette L. Bottoms
  • Kari Nysse Carris
  • Andrew R. A. Conway
Original Paper
  • 56 Downloads

Abstract

Children’s ability to identify perpetrators who are strangers is crucial. Both failures to identify actual perpetrators and false accusations have serious costs for children’s personal security and wellbeing or for innocent suspects. In the current study, we investigated the accuracy of children’s person identification after a very long delay (one year) and, for the first time, as a function of interviewer-provided socio-emotional support, children’s pre-existing social support reserves, state anxiety, and individual differences in children’s working memory capacity. One year after a play session, 7- to 8-year-old children were asked to identify a stranger from a target-present lineup that included four pictures, a “She’s not here” card, and an “I don’t know” option in the form of a question mark card. Overall, children made few correct identifications, but also few false identifications, preferring to give “I do not know” responses or incorrectly stating the person was not in the lineup. Contrary to predictions, there were no main effects of interviewer-provided social support nor of social support reserves, but as predicted, children with higher working memory capacity performed significantly more accurately than did children with lower working memory capacity, regardless of interview condition. Children’s state anxiety and their confidence in the identification were unrelated to their performance, although interviewer support did decrease state anxiety. Implications of these findings for children’s personal safety are discussed.

Keywords

Children Eyewitness Lineup identification Social support Working memory Delay 

Children are often involved in the legal process as victims or bystander witnesses. Their testimony is central when their memory of the events is the only evidence (Goodman 2006; Memon et al. 2011; 2017; Myers et al. 1999). Although the vast majority of child sexual abuse cases involves perpetrators known to the child (e.g., Ullman 2007), the proportions are far from negligible. For example, stranger assaults made up 25% of child sexual assaults in a national sample of children (Finkelhor et al. 2008), and 35% of total victimization incidents in another sample (Gallagher et al. 2008). When a stranger sexually assaults a child, the child’s ability to accurately identify the stranger is an important concern. The accuracy of that identification has serious implications for justice. An accurate identification can assist law enforcement in stopping a dangerous criminal, but an inaccurate identification can lead to the prosecution of an innocent. Both failures to identify actual perpetrators and false accusations have serious personal and societal costs (Innocence Project 2019; Pipe et al. 2007; Wells et al. in preparation)—especially when pursuing the wrong suspect places other children at risk.

Actual cases illustrate both outcomes. For example, in August 1983, 3-year-old Lori Poland was kidnapped from her front yard, sexually assaulted, and left for dead at the bottom of a 10-foot deep outhouse in the Colorado mountains. A series of unlikely events led to her rescue and her attacker’s capture: Hikers heard Lori’s faint cries from the outhouse pit, leading to her rescue. Witnesses recalled part of the abductor’s license plate, leading police to a suspect. From her hospital bed, where she lay traumatized and suffering from exposure, the 3-year-old recalled details of her ordeal and correctly identified the suspect, Robert Paul Thiret, from a photo lineup. In addition, the child’s detailed account of the assault matched perpetrator’s, illustrating that even the memory of very young children can be accurate and helpful in criminal cases (Jones and Krugman 1986). Thieret served 6 years in prison. But the outcome of another child-witness identification was different. In 1988, an 11-year-old boy witnessed the murder of 16-year-old Felix Valentin in Chicago. The child said he had a “clear view” of the shooter. He even specified that he had seen the shooter before at a local park. He described the shooter in detail, including the colors of his clothing and distinctive dyed ponytail—colors that led police to suspect the Latin Kings gang. When police showed the child witness photographs of suspected gang members, he identified 23-year-old Jacques Rivera. Later, he picked Rivera out of a live lineup. Based on this identification, Rivera was convicted and served 13 years before the now-adult witness admitted that, days after the photo lineup, he had seen the real shooter and realized that he had identified the wrong man. (In fact, the boy had made both identifications from the photographs during the police interview.) Rivera received over $17 M in damages after the jury learned that a police detective had concealed the recantation of the identification and had pressured the child to identify Rivera (Center on Wrongful Convictions and Possley 2019).

Lineups (in person or photo arrays) should help children be more accurate compared to verbal descriptions, given that children often have difficulty providing sufficient details in response to open-ended recall questions such as “What did the person look like?” (e.g., Goodman et al. 1990; Karageorge and Zajac 2011; Pozzulo 2014; Schwartz-Kenney et al. 1996). In fact, children do provide more correct responses on photo identification tasks compared to recall tasks, although this improvement can also bring about more incorrect responses (e.g., false alarms or incorrect identifications; Rudy and Goodman 1991; Schwartz-Kenney et al. 1996). Indeed, Zajac and Karageorge (2009) found that “several children provided wildly inaccurate descriptions of the target, yet went on to correctly identify him from the lineup” (p. 365).

Research has addressed several factors affecting children’s identification accuracy, including child age, lineup type (target present vs. target absent), etc. (see Fitzgerald and Price 2015, for review). In the current study, we investigated the accuracy of children’s person identification after a very long delay (one year) and as a function of socio-emotional support (provided during a forensic interview and as a pre-existing feature of children’s social lives—social support reserves) and individual differences in children’s working memory capacity. To our knowledge, only two studies have tested the effects of social support on lineup identification (Goodman et al. 1991a; Rush et al. 2014), despite the robust evidence for its important role for children’s event recall accuracy (see Bottoms et al. 2007; Saywitz et al. 2019). We also investigated whether children’s state anxiety and self-reported confidence are related to their accuracy after long time delays. Such delayed identifications are understudied but important, because children often have to identify a suspect or remember details of an event long after the event has taken place, when their accuracy might draw particular scrutiny.

Factors Affecting Children’s Identification Accuracy

The literature on children’s identification accuracy is growing and mixed. Early work highlights that under some circumstances, children can perform as well as adults (who themselves are not always accurate) on target-present photo identification tasks after age 5 years (e.g., Goodman and Reed 1986; Pozzulo and Lindsay 1998), arguably because face perception processes are quite mature by the age of 5 (Crookes and McKone 2009) and because recognition memory develops earlier and is often better than recall memory (Kail 1990). (For the same reasons, photo lineups should be easier than open-ended descriptions for children.) Yet a more recent meta-analysis concluded that children are generally less accurate than young adults on target-present lineups: less likely to identify the target correctly, more likely to identify a filler (foil) incorrectly, and more likely to incorrectly reject all of the lineup members (Fitzgerald and Price 2015). Accuracy in response to a lineup task, for adults or children, involves not only cognitive abilities (memory), but also socioemotional factors, especially suggestibility. Pozzulo et al. (2012) demonstrated that even when children’s memory for a lineup target is accurate, they are still more likely to falsely identify a target than are adults, perhaps because they are affected more by the social context, which suggests to them that they are supposed to choose someone from the lineup, as opposed to saying “don’t know,” which researchers have tried to address (e.g., Schwartz-Kenney et al. 1996).

We were particularly interested in children’s person memory after a one-year delay. Over long delays, both adult and child memory accuracy tends to deteriorate (Flin et al. 1992; Wixted and Wells 2017), and neither is very accurate in identifying strangers from photo lineups. (An exception can be memory for personally relevant events and for stressful events, which is less likely to deteriorate over time and is resistant to suggestion—e.g., Goodman et al. 1990; Scrivner and Safer 1988; for review, see Goodman, Ogle, McWilliams, Narr and Paz-Alonso 2014). The research on children’s person identification reveals no effects of short delays (one to two weeks) on photo lineup accuracy among 5–11 year-olds (Karageorge and Zajac 2011) or 5–6-year-olds (Goodman et al. 1991b, Study 2), and fewer correct identifications (but not more false identifications) among 3–4 year-olds (Goodman et al. 1991b, Study 2). Delays of three to six weeks affected 4- to 12-year-olds’ photo identification accuracy compared to accuracy at 4–10 days (Oats and Shrimpton 1991), but did not affect 3- to 7-year-olds’ photo lineup accuracy compared to accuracy at two weeks (Goodman et al. 1991a). Delays did not decrease 5-year-olds’ photo lineup accuracy (Lindberg et al. 2001), but did hurt the accuracy of children in the latter study who were merely bystander witnesses. In the only study with a one-year delay, Goodman et al. (1991b, Study 4) found a decrease in 4–7-year-olds’ correct photo identification rates over time, but no significant change in false identifications.

In addition to testing accuracy after a long delay, we sought to test the effects of interviewer-provided social support on children’s accuracy—i.e., their ability to correctly identify the suspect from a target-present lineup, especially compared to making false identifications (i.e., identifying one of the fillers in the lineup). In addition to its direct effects on accuracy, we were also interested in whether interviewer support moderated the effects of individual differences in social support reserves and working memory capacity. Finally, we explored the relations between state anxiety, confidence in the photo identification, and children’s accuracy. We review literature relevant to these variables and their relations next.

Interviewer-Provided Social Support

Social support is a form of interaction that fosters feelings of well-being in the receiver (Burleson et al. 1994; Rosenthal and Jacobson 1968; Sarason et al. 1990; Wolchik et al. 1990). Warm, socially supportive interviewing is a central component of forensic interviewing techniques (e.g., the NICHD protocol, Lamb et al. 2008; Narrative Elaboration Interview, Saywitz and Camparo 2014; for review, see Sayitz et al. 2018). In general, when interviewers provide social support (e.g., exhibit warmth, friendliness, and open body posture; make eye contact; smile; establish rapport) versus not (e.g., exhibit formal, stern behavior; closed body posture; make minimal eye contact; fail to smile or establish rapport), children are more accurate (but see Imhoff and Baker-Ward 1999; for reviews, see Bottoms et al. 2007; Saywitz et al. 2019). Specifically, children who are interviewed by a socially supportive interviewer are less likely to make errors in response to specific nonsuggestive questions and more resistant to suggestive questions. Interviewer support appears to have somewhat less influence on children’s free recall in laboratory studies (e.g., Davis and Bottoms 2002; Eisen et al. 2019; Quas, Wallin, Papini, Lench and Scullin 2005; but see Carter et al. 1996; Goodman et al. 1991a). Yet in field studies involving children who are interviewed about actual alleged abuse, those who were interviewed in a more (vs. less) supportive manner were less reluctant in their responses, gave more complete and detailed reports, and were more likely to disclose abuse (Hershkowitz 2011; Hershkowitz et al. 2007, 2015).

Therefore, we reasoned that social support is also likely to increase children’s person identification accuracy. There have been only two studies we know of directly testing this. In the first, interviewer support had no significant effect on children’s identification accuracy after 2 or 4 weeks (Goodman et al. 1991a). This was the first empirical study of the impact of interviewer-provided social support. The operationalization of support was strengthened in later studies that relied upon clinical literature to bolster the technique (Carter et al. 1996). In fact, the effects in that study were largely positive, but as described by the authors, “sporadic” across other measures of memory and suggestibility. In the second, interviewer support had no significant effect on children’s identification accuracy after a 2-week delay; however, social support did help children who had been stressed during the encoding event perform as well as or better than children who suffered no stress (Rush et al. 2014). Thus, although it had no main effect, interviewer support had beneficial moderating effects on the relationship between stress and identification accuracy.

Three other laboratory studies with less direct tests also support the hypothesis that positive effects of interviewer-provided social support will generalize from forensic interview accuracy in general to accuracy on person identification tasks specifically. First, in a study conducted by Wilcock et al. (2018), children who were interviewed by a registered intermediary (“an impartial, trained professional who facilitates understanding and communication between vulnerable witnesses and members of the justice system,” p. 367) were more accurate in identifying a suspect than were children interviewed with existing procedures. Second, children interviewed by someone wearing casual attire (vs. a uniform) experienced less anxiety and were (marginally) more accurate in their identifications (Lowenstein et al. 2010). Finally, playing a game with an interviewer before a photo lineup also improved boys’ (but not girls’) performance (Rassin et al. 2006). Although these researchers did not conceptualize their manipulations specifically as interviewer-provided social support during a forensic interview, it is likely that the presence of a registered and trained intermediary, interviewers’ casual attire, and rapport-building gameplay improved children’s lineup accuracy because these circumstances created a warmer, more supportive interview. In contrast, Ricci, Beal, and Dekle et al. (1996) found that children were more relaxed, yet marginally less accurate when a parent administered a lineup than when a stranger did so. It is difficult to draw definitive conclusions from this, however, because interviewer identity was confounded with social support, and in some cases, parents did not appear to be supportive.

Davis and Bottoms (2002) argue that interviewer-provided social support might be particularly beneficial after long delays, because children are likely to have less accurate memories and be more suggestible. In support, Goodman et al. (1991a) found that after a 4- (vs. 2-) week delay, age differences in children’s responses to specific questions (i.e., older children performed better than younger children) disappeared, although younger (vs. older) children did make more omission errors on misleading questions. Even more benefits might result after a significant delay such as 1- year (as in our study), which would place additional pressure on children’s cognitive capabilities.

Why is interviewer support helpful for children’s eyewitness accuracy? In general, social support is associated with positive outcomes that, theoretically, should promote eyewitness accuracy, such as higher perceived ability to cope with stress, lower physiological response to stress, more accurate short-term recall, and lower levels of trauma-related mental health issues (e.g., Cohen and Wills 1985; Kawachi and Berkman 2001; Kelley and Gorham 1988; Ozbay et al. 2007; Sarason et al. 1990; Wolchik et al. 1990). Social support helps children perform at their highest cognitive levels (Fischer 1980; Fischer et al. 1993; Vygotsky 1934/1978). Children might also perceive a supportive interviewer as less intimidating, and less likely to respond negatively if the child provides information that contradicts the interviewer’s expectations (Goodman et al. 1991a). This encourages children to be less compliant, which could explain why children are less likely to acquiesce to suggestive questioning when interviewed in a supportive manner. When identifying suspects in a photo lineup, children might feel less pressured to make an identification regardless of whether they recognize the suspect—just to please the interviewer. Thus, social support would increase children’s cognitive ability for accurate recall, reflected in a higher correct-identification rate (compared to, for example, misses), but also children’s ability to resist real or perceived pressure to identify someone incorrectly—reflected in higher correct-identification versus false-identification rates.

Interviewer-provided social support might also be particularly helpful to children whose pre-existing cognitive and/or emotional characteristics place them at some type of relative disadvantage. For example, it might boost the performance of children with increased autonomic reactivity (i.e., greater sensitivity to stress in the environment, Quas et al. 2004) or insecure attachment style (Alexander, Quas and Goodman 2002; Davis and Bottoms 2002; Goodman, Quas, Batterman-Faunce, Riddlesberger and Kuhn 1997; see Bottoms et al. 2007 for a review). It is especially helpful to younger children (Quas and Lench 2007), who are less successful at resisting suggestive questioning compared to older children (e.g., Goodman et al. 1991b). Interviewer support can also reduce the detrimental effects of stress and anxiety on children’s performance (Rush et al. 2014; Saywitz et al. 2019). Further, non-supportive interviewing might decrease performance among children who are high in trait suggestibility (Quas et al. 2005) or physiological arousal (Quas and Lench 2007). In our study, we focused on whether interviewer support moderated the effects of two such pre-existing characteristics: social support reserves and working memory capacity.

Social Support Reserves

Social support reserves refers to the overall amount of perceived social support children receive in their daily lives from parents, family members, peers, teachers, etc. (Davis and Bottoms 2002; Zelkowitz 1989). Carter et al. (1996) first proposed that interviewer-provided social support would moderate the effect of social support reserves on children’s eyewitness suggestibility. Specifically, they suggested that children who have low social support reserves would benefit most from being interviewed in a supportive (vs. non-supportive) manner because a supportive interview style would provide the empowerment lacking in children with a weaker support network, helping them to resist misleading questions. Therefore, differences in children’s accuracy due to their social support reserves before the interview should be reduced when all children are empowered to perform well by the interviewer-provided support.

Davis and Bottoms (2002) were the first to test this hypothesis directly, yet found no moderation effect. The authors attributed the null findings to the social support measure they used. Specifically, parents reported their children’s social support reserves, which might not have accurately assessed children’s actual reserves due to the parents’ social desirability concerns when completing the measure, their lack of knowledge about peer support their children were actually receiving outside of their presence, etc. Even so, Davis and Bottoms found that interviewer support helps children with lower social support reserves in particular, as theorized: Children who were low in social support reserves made fewer commission errors on specific questions when interviewed in a supportive versus non-supportive manner. In contrast, children who were high in social support reserves responded similarly regardless of interviewer support. Again, there was no test of how this variable affected person identification accuracy.

In the current study, children had to identify a woman who had played the role of a sitter during a play session one year before the interview. We relied on children’s own assessment of their social support received from parents and peers, arguably a more accurate measure of children’s perceived social support reserves. We hypothesized, as did Carter et al. (1996) and Davis and Bottoms (2002), that when children are not interviewed in a supportive manner, pre-existing social support reserves would influence their accuracy: those with lower reserves would perform worse than those with higher reserves. But when children are interviewed in a supportive manner, the effect of social support reserves would be reduced—children low in support reserves would improve to perform as well as those high in reserves. Further, we expected this effect to be most marked when comparing children’s correct identification rates to false-identification rates, where children did not recognize the sitter but felt perhaps pressured or thought the adult wanted them to make an identification.

Working Memory Capacity

Another potential individual characteristic with implications for identification accuracy is working memory capacity (Bottoms et al. 2007)—and we predicted that interviewer support would moderate its role as well. Working memory is a dynamic system involving both storage and control processes designed to keep information in mind while performing complex cognitive tasks such as reasoning, comprehension, or learning (Baddeley 2010; Conway and Engle 1996). Specifically, working memory is the cognitive ability to maintain and to manipulate task-relevant information (Owens et al. 2014). Working memory capacity involves efficient attention control processes (Unsworth et al. 2014). Thus, people who have high (vs. low) working memory capacity are better at controlling and focusing on information that is relevant to the task at hand, despite the presence of distractions. In contrast, because people with low working memory have less efficient control processes, they are less effective at controlling and focusing on relevant information, and are more likely to be distracted by irrelevant information (Conway and Engle 1996). For example, they are more likely to experience the classic cocktail party effect than those with high working memory (Conway et al. 2001).

Working memory capacity develops through childhood (e.g., Nevo and Breznitz 2011), and it is likely to be instrumental to cognitive activity such as remembering or recognizing information during forensic interviews. During a forensic interview, high working memory capacity should enable children to focus more effectively on the content of questions and be more efficient at ignoring misleading information. Low working memory capacity might be related to children’s lower ability to concentrate on the central content of questions, and therefore to increased distractibility and susceptibility to misleading information (Bottoms et al. 2007; but see Roebers and Schneider 2005). Support for these predictions comes from literature investigating other related measures of cognitive functioning, including short-term memory, a separate but related construct that reflects capacity for short-term storage of activated information, but without the attention/central-executive component (Baddeley and Hitch 1974; Engle et al. 1999; Engel de Abreu et al. 2010). Specifically, among children who were being assessed for child abuse, greater cognitive functioning (consisting of a composite measure of short-term memory, intelligence, and language comprehension) predicted better memory performance on free recall, open-ended, specific, and misleading questions (Eisen, Goodman, Qin, Davis and Crayton 2007), and higher (vs. lower) short-term memory was associated with fewer total memory errors (Eisen, Qin, Goodman and Davis 2002). Nonetheless, neither the composite measure of cognitive functioning nor short-term memory predicted better photo lineup accuracy, perhaps because the measures tapped into verbal abilities, which were less relevant to the identification task (vs. other tasks).

Therefore, we predicted that children with high (vs. low) working memory capacity would be less suggestible overall, but that this effect would be moderated by interviewer-provided support. Children with low working memory capacity would be less accurate than children with high working memory when interviewed in a non-supportive manner. But when interviewed by a supportive interviewer, children with low working memory would be able to perform more like those with high working memory capacity. Thus, we expected supportive interviewing to give children with low working memory the boost they need to resist misleading information, and perform similarly to children with high working memory—a pattern similar to the one for social support reserves.

State Anxiety

Research on the relationship between anxiety and memory is abundant, but it presents contradictory findings and warrants further research. Children’s state anxiety during retrieval of information might be particularly important to study in the context of interviewer support. Interviewer-provided social support might enhance children’s accuracy because it reduces children’s state anxiety (Davis and Bottoms 2002). Saywitz et al. (2019) also suggested this mechanism, drawing upon attentional control and processing efficiency theories (Eysenck and Calvo 1992; Eysenck et al. 2007) and theories of stress and coping (Cohen 2004). Alleviating the negative cognitive effects of anxiety (i.e., cognitive load and cognitive interference) would allow children to allocate more cognitive resources to attention, processing, and retrieval of information (Derakshan and Eysenck 2009; Eysenck and Calvo 1992; Eysenck et al. 2007) such as a suspect’s image from memory, and to the task of comparing that memory to photographs in a lineup. Indeed, children who are interviewed in a supportive manner exhibit lower levels of anxiety (Davis and Bottoms 2002; Rush et al. 2014), and also show improved memory performance (Almerigogna et al. 2007; Quas et al. 2014) and deeper processing of an event that was stressful at the time of encoding—which could, arguably, induce some level of anxiety when children are asked to remember that event (Klemfuss, Milojevich, Yim, Rush and Quas 2013). Yet, in the first study to test the relation between interviewer-provided social support and state anxiety at retrieval, the mediation process was not supported: Anxiety was not a significant mediator of the effect of social support on suggestibility (Davis and Bottoms 2002).

Evidence that anxiety is relevant to children’s memory processes comes from another set of studies focused on children’s anxiety during encoding—that is, anxiety during the event children have to remember. Although anxiety might have different effects during encoding (vs. retrieval), this research is relevant to our question about state anxiety at retrieval because children might experience heightened state anxiety when they have to remember details of an event that elicited anxiety during encoding. In forensic contexts, which often involve highly stressful events children have to remember, this is especially plausible. When information is personally distressing and meaningful, higher state anxiety during the event can result in better memory accuracy and less suggestibility (Goodman et al. 1991b; Quas et al. 2014; for review, see Goodman et al. 2014). Other times, however, anxiety during the event hurts children’s memory and increases suggestibility (Almerigogna et al. 2007; Eisen et al. 2007). This conflicting pattern is reflected in lineup identification research. Some found no relation between anxiety and children’s eyewitness identification accuracy (Eisen et al. 2002, 2007; Oats and Shrimpton 1991), others found that anxiety hurts accuracy (Deffenbacher et al. 2004), and others that it improves it (Goodman et al. 1991b, Study 2).

In the present study, we measured state anxiety during the interview and predicted that it would (a) decrease children’s accuracy, such that increased anxiety would reduce correct response rates; (b) be increased by non-supportive versus supportive interviewing; and (c) act as a mediator between interviewer support and lineup identification accuracy.

Children’s Confidence

Although researchers have studied the confidence-accuracy relation extensively in adults’ eyewitness identification, less attention has focused on this relation among children (Fitzgerald and Price 2015). In adults, research initially suggested that confidence and accuracy were unrelated, bringing into question the Supreme Court explicit recommendation for jurors to consider eyewitness certainty as a cue to accuracy (Neil v. Biggers1972). If confidence was an unreliable marker of witness’s identification accuracy and if, as revealed by other research, jurors are most likely to believe confident witnesses, miscarriages of justice surely abound. As a result, several states have since ruled that jurors should be explicitly warned against relying on witness confidence (Georgia in Brodes v. State2005, Kansas in State v. Mitchell2012, New Jersey in State v. Henderson2011). Yet more recent work demonstrates that confidence and accuracy are not as weakly related as previously thought—in fact, the lack of relationship could have been an artifact of the improper lineup procedures that were also being studied (Wixted and Wells 2017). In this recent body of work, confidence is positively related to accuracy when optimal identification procedures are used and when confidence ratings are made immediately after the identification (Brewer and Wells 2006). Are the same patterns true for children, especially in light of their limitations relative to adults in terms of metacognitive monitoring and awareness (Keast et al. 2007)?

The few studies of the relation between children’s confidence and accuracy on photo lineups reveal that, overall, young children’s confidence ratings are not predictive of their lineup accuracy (e.g., Brackmann et al. 2019; Humphries and Flowe 2015), and children tend to be overconfident in their abilities (Brewer and Day 2005; Keast et al. 2007;), just as the classic adult findings. Little is known about moderators of this relation, but Bruer and Price (2017) found that, across different lineup procedures, greater confidence predicted better accuracy among older (vs. younger) children for choosing a suspect from a lineup (but not for rejecting a lineup). We therefore examined the relation between children’s confidence and their accuracy in response to lineups as a secondary question in our study.

Overview

In summary, children’s person identification accuracy after a long delay would be predicted by interviewer-provided social support (versus lack thereof), social support reserves, and individual differences in working memory capacity. We tested children’s photo lineup accuracy one year after exposure to a stranger female in a laboratory. During the initial event, children engaged with a stranger (i.e., the female babysitter) during a play session and were interviewed about the event but did not perform a lineup identification of the babysitter at that time. We hypothesized that interviewer-provided social support and higher levels of working memory would be associated with increased accuracy and decreased false identifications. We used a target-present lineup where children could identify the babysitter (correct response), identify a filler (false identification), reject the lineup (miss), or indicate they did not know. We expected interviewer support to increase the correct identification rates compared to all other possible responses (i.e., improve memory), but to increase correct identifications especially compared to false identifications, because false identifications might be reflections of children trying to please the interviewer by choosing a picture even though they did not recognize the babysitter. Therefore, we expected social support to both improve memory and reduce suggestibility.

We also hypothesized that interviewer support would moderate the effects of working memory capacity and of pre-existing social support reserves on accuracy, such that children with a pre-existing disadvantage (low social support reserves or low working memory ability) would be most helped by supportive interviewing. Separately, we also tested the effect of state anxiety on children’s accuracy—given the contradictory findings about this variable. We hypothesized that anxiety would predict lower correct response rates compared to all other types of responses. Finally, we explored the relationship between confidence and accuracy in children’s identification.

Method

Participants

Participants were 72 7- to 8-year-old (age range = 85–106 months, M = 94.97 months, Mdn = 94 months; 36 girls, 36 boys) Chicago-area children who had participated one year earlier in a child witness laboratory study (Davis and Bottoms 2002). Nine children from the first study did not participate: one could not be located, two had moved out of state, and six declined participation due to scheduling difficulties or child’s refusal to participate. One participant was African American, 65 were Caucasian, 3 were Latino, and 3 were of other ethnicities. Children were of upper-middle socioeconomic status (Hollingshead 1975). Children received $25, a toy, and a certificate of achievement for their participation.

Materials

A-State Scale of the State-Trait Anxiety Inventory for Children

We used an 11-item subscale of the A-State scale of the State-Trait Anxiety Inventory for Children (STAIC) (Spielberger 1979), a measure appropriate for children 5 years old and older when administered orally (Papay and Spielberger 1986) that measures transitory anxiety by asking children to indicate how they feel at a particular moment. The A-State scale is reliable (α = .79 in this sample), internally consistent, and correlates significantly with other indices of anxiety such as the Children’s Manifest Anxiety Scale for Children (Castaneda et al. 1956; Spielberger 1979). For each item (i.e., calm, upset, nervous, scared, relaxed, worried, frightened, happy, good, bothered, nice), children pointed to one of three statements that best represented how they felt during the forensic interview; for example, “I felt 1 (not scared) 2 (scared) or 3 (very scared)” and “I felt 1 (very happy) 2 (happy) or 3 (not happy).” Some items were recoded so that higher numbers always indicated elevated anxiety, M = 1.39, SD = 0.24.

Social Support Scale for Children

The Social Support Scale for Children (Harter 1985) includes four 6-item subscales measuring the perceived level of social support (i.e., positive regard) children receive from parents, teachers, peers, and classmates. Each item consists of two statements that describe two types of children: children who receive support from others and children who do not receive support. For example, “Some kids have parents who don’t really understand them, but some kids have parents who really do understand them. What kind of kid are you?” Children choose one of the two statements, then indicate by pointing whether that was “really true for me” or “sort of true for me.” Each item is scored on a scale from 1 to 4, where 1 represents the lowest level of support (i.e., choosing the first child in this example plus “really true for me”) and 4 represents the highest level of support.

Because the Social Support Scale for Children was designed for children above 7 years old, an experimenter read all of the items to the children. Harter (1985) reported that 8-year-old children did not differentiate between peers and classmates, thus children completed only parent and peer subscales. Responses for the remaining 12 items were averaged across the subscales for each child, yielding an overall social support score. This scale (collapsed across the two subscales) was reliable and internally consistent, Cronbach’s alpha = .79, M = 3.56, SD = 0.45.

Counting Span Task

Working memory capacity was measured with the counting span task, which is a standard measure used in both adults and children (Case et al. 1982; Engel de Abreu et al. 2010; Towse and Hitch 1995). It is a computer-based task that requires the participant to count and remember the number of targets (here, blue circles) embedded in a display of targets and non-targets (here, red circles). In this task, each display contained three to seven targets presented in a trial block of two, three, or four displays. Children completed three practice trial blocks to familiarize themselves with the task and then completed nine trial blocks, 3 of each display length, for a maximum possible score of 27. Children were told to count aloud and remember the number of blue circles they counted in each display. After counting the last display in a trial block, children were instructed to recall the number of blue circles in each display in correct serial order. Children’s responses were scored according to a partial credit scoring key by awarding points equal to the number of correctly recalled displays in each trial block—one point for each display. For example, in a trial block with two displays of 3 and 5 blue circles, children received two points if they correctly recalled “3” and “5,” respectively. To receive any points, children had to recall the number of blue circles in the order presented. Scores ranged from 0 to 24, M = 10.40, SD = 5.02; we included all children in the analyses related to this variable, but all results were the same when the two children who scored 0 were excluded.

Target-Present Photo Lineup

Children saw a 7-card target-present person identification lineup that contained five Polaroid photos, one card labeled “She’s not here,” and one card labeled “I don’t know” with a big question mark. The latter two cards were included because of Karen Saywitz’s illustrations of the importance of non-leading external memory aids in her Narrative Elaboration interviewing protocol research (e.g., Saywitz and Camparo 2014; Saywitz and Snyder 1996;), and because of the use of a question mark card by Schwartz-Kenney et al. (1996, Study 2). The target was a woman who was introduced to children as “the babysitter” who stayed with the children in “the dinosaur room” during the staged play session of the first study one year earlier. All 4 foils wore the same long-sleeved shirt worn by the target during the play session. Foils matched the target on ethnicity, hair length, and hair color. To insure that the lineups were unbiased, during pilot testing, the photographs were shown to 10 other 7- to 8-year-old children, who were asked, “Which one of these looks most like a babysitter?” Separate one-sample t-tests comparing the proportion of children who selected each picture to the proportion expected by chance (20%) revealed that children selected each woman’s picture at chance levels, all ts(9) < 1.22, ps > .25.

Confidence-in-Photo-ID Response Measure

Confidence was assessed on a 4-point scale from 1 (just guessing) to 4 (really, really sure), M = 2.20, SD = 1.00. Children who picked “I don’t know” did not answer the confidence question.

Procedure

One year after children engaged in the laboratory play session with the stranger/babysitter, they returned for the current study, where they were asked various questions about the original play session and asked to identify the babysitter from a target-present lineup. Children returned on average 13.13 months after their initial play session (SD = 0.70).

Study 1: Engaging Interactions with the Target Babysitter

Interactions with the babysitter during the first study are described in detail by Davis and Bottoms (2002). In brief, children were escorted alone to the “dinosaur room” in the laboratory, where the child played with a woman research assistant introduced as the “babysitter.” (Three women alternated playing this role.) The babysitter engaged the child in various activities with pictures, stories, balloons, bubbles, a Barbie doll (which “accidentally” broke during the play session), a Polaroid camera, a staged puppet show (where a ladybug puppet bit another puppet on the behind), and activities that involved innocuous touching (the babysitter traced the child’s body on a large piece of paper and touched the child’s biceps and shoulder). The child’s parent(s) watched the interaction surreptitiously from another room. This part of the session lasted approximately 20 minutes.

Then the babysitter took the child to a new room for a mock forensic interview by a man interviewer. He explained that he was the babysitter’s boss and wanted to make sure the babysitter did a good job. He administered the interview under one of two support conditions modeled on the techniques used by Carter et al. (1996): (a) socially supportive: he introduced himself, spent several minutes building rapport, maintained supportive eye contact, smiled often, used a warm voice, and sat close to the child with a relaxed body posture; or (b) non-supportive: he began the interview without introducing himself nor building rapport, then sat for a few moments in an intimidating silence while flipping through his interview pages then, throughout the interview, avoided eye contact, did not smile, spoke monotonously, and maintained formal body posture. After the interview and a set of measures with another experimenter, children and parents were debriefed, children were given two toys and a certificate of achievement, and parents were compensated. The entire session lasted approximately 40 minutes.

Current Study: The Photo-ID Interview, One Year Later

Separately, parent/child pairs from the first study came back to the same university laboratory one year later, where they were greeted by an experimenter who obtained parental informed consent while children played with a research assistant. Parents were shown forensic interview questions (part of a larger study not reported herein) and the Social Support Scale for Children items (Harter 1985). Children learned that they would talk to some adults and signed Child Assent Forms, which were read aloud. Then the children were escorted to an interview room and introduced to a woman research assistant. The assistant reminded each child that he or she had visited the “dinosaur room” a year ago, and she asked the child to recall one thing that had occurred in the dinosaur room. Then children were re-introduced to the same man who had interviewed them one year prior. He explained that he had lost everything that the child had told him last year, and then asked the child the same questions from the first interview session. Then, the research assistant came back and the interviewer left the room. The research assistant administered the STAIC then left when the interviewer came back. At this point, the male interviewer asked the child to identify the babysitter from the original play session using the target-present photo lineup. The interview was conducted under one of the two support conditions used during Study 1. Half of the children received the same support manipulation that they had gotten during Study 1; half received the other support manipulation. Thus, 18 children experienced only supportive interviews during both Study 1 and the current follow-up study, 19 children experienced only non-supportive interviews at both, 17 children experienced a supportive Study 1 interview and a non-supportive current-study interview, and 18 children experienced a non-supportive Study 1 interview and a supportive current-study interview. Child gender and age were approximately balanced across conditions.

Specifically, the interviewer told the child that he would show the child five pictures of women, one of whom might be the babysitter from the dinosaur room. The child was instructed, “Take your time and look closely at each one of the pictures. Be careful, because all of them look a lot alike. The babysitter’s picture might be here, or it might not be here at all. If you think her picture is here, I want you to pick it out.” The interviewer also showed the child the two labeled cards and said, “Along with the pictures, there are two cards. One of them says ‘She’s not here.’ If you think her picture is not here, I want you to pick the card that says ‘She’s not here.’ Now, there’s one more card. It says “I don’t know” and it has a question mark on it. If you just don’t know whether her picture is here or whether it’s not here, then you pick the card that says, ‘I don’t know.’ The interviewer asked the child to say whether he or she understood the task just before or as he was opening the folder and showing the child the photos. A folder was used that allowed the interviewer to present the photo array “blindly,” without seeing the pictures himself. Due to experimenter error, the “I don’t know card” was not included in the lineups presented to 7 children. Analyses conducted with and without these children produced similar results, so we report the results of analyses for all children.

Children’s responses to the Photo ID task were coded into four categories: correct responses (if the child identified the babysitter correctly), misses (if the child reported the babysitter was not in the lineup by picking the “She’s not here” card), false identifications (if the child identified one of the women who was not the babysitter), and do not know (if the child picked the “I don’t know” card). Children who chose a photo were then asked to indicate how sure they were that the woman they picked was the babysitter. Children who selected the “She’s not here” card were asked how sure they were that the babysitter’s picture was not there. After the photo ID task, the research assistant returned and administered the C-SPAN Counting Task, then gave the child a short cookie and juice break, then administered the Social Support for Children Scale.

The interview was videotaped and parents watched via closed-circuit television. Parents could see, but not hear, their child’s responses to the Social Support Scale for Children; to protect the children’s confidentiality and encourage their frankness—children were told that the microphone to the videocamera was turned off during this portion of the session.

Finally, children and parents -were debriefed, and children picked out a toy and received $25 and a certificate of achievement. Children interviewed in the non-supportive condition were reintroduced to the interviewer so that they would know that he was a “nice guy.” The entire session lasted approximately 1 hour and 15 minutes. All study procedures and materials from both Study 1 and the current study were reviewed and approved by the university Institutional Review Board.

Results

Overall, children were not good at identifying the babysitter from the previous year’s play session, which was not surprising considering the long period between the event and the interview. Specifically, 10 children (14%) correctly identified the babysitter, 23 children (32%) failed to identify her even though she was in the lineup, 11 children (15%) identified another woman as the babysitter (false identification), and 28 children (39%) said they did not know. Children rendered correct and false identification responses at rates significantly lower than chance, and missed the babysitter or said they did not know at rates significantly higher than chance, χ2 (3, N = 72) = 13.22, p = .004. See Table 1 for response frequencies per experimental condition.
Table 1

Frequencies of each type of lineup identification response for supportive and non-supportive interview conditions

 

Non-supportive

Supportive

n

%

n

%

Correct identification

5

13.9

5

13.9

Miss

13

36.1

10

27.8

False identification

4

11.1

7

19.4

Do not know

14

38.9

14

38.9

Total

36

100.0

36

100.0

To test the hypotheses that (a) interviewer-provided social support during the current study would improve children’s performance by increasing the likelihood of a correct response (especially at the expense of false-identification responses that might be due to children’s suggestibility and desire to please the adult), and (b) the beneficial effect of interviewer support would be stronger for children low (vs. high) in working memory capacity and in social support reserves, we used multinomial logistic regression analyses. Multinomial logistic regression is a generalization of logistic regression for binary dependent variables to nominal variables with more than two categories. This approach is akin to estimating logistic regression equations for each possible pair of categories simultaneously—resulting in k -1 comparisons (where k is the number of categories in the dependent variable)—using one category as the reference group (e.g., Brunner et al. 2007). In this study, we chose correct responses as the reference group. First we tested the main effects of interviewer-provided social support, individual differences in working memory capacity, and individual differences in social support reserves—that is, we tested whether these variables significantly influenced the likelihood of children giving a correct response versus all other types of responses.

Second, we tested the hypothesized interactions of interviewer-provided social support with social support reserves and working memory capacity. Specifically, we tested the prediction that interviewer-provided support would help most those children who had lower levels of social support reserves and lower working memory capacity, whereas children who were already high in these individual difference variables would already have higher correct response rates regardless of interviewer support. In all analyses, we controlled for interview type (supportive or non-supportive) given to children during the first study, which had no significant relation to the current study’s photo-identification responses, all Bs < − 0.92, all ps > .27, all OR 95% CIs include 1 (Table 2). Therefore, in all results below, “interviewer support” refers to the support provided during the current study (not the original study).
Table 2

Multinomial logistic regression coefficients for interviewer support on correct versus other response rates

 

B

Wald

p

OR

OR 95% CI

LL

UL

Interviewer support only

  Miss

Intercept

0.53

0.71

.40

   

Interviewer support study 1

0.86

1.22

.27

2.35

0.51

10.75

Interviewer support current study

− 0.29

0.14

.71

0.75

0.17

3.38

  False ID

Intercept

− 0.70

0.73

.39

   

Interviewer support study 1

0.95

1.11

.29

2.59

0.44

15.16

Interviewer support current study

0.53

0.35

.55

1.71

0.29

9.91

  Do not know

Intercept

1.10

3.41

.06

   

Interviewer support study 1

− 0.18

0.06

.81

0.83

0.19

3.67

Interviewer support current study

0.01

0.00

.99

1.00

0.24

4.26

Interviewer-Provided Social Support

The multinomial logistic regression model testing the main effect of interviewer support on children’s responses was not significant; that is, the model including interviewer support was not significantly different than the intercept-only model, χ2 (6, N = 72) = 5.81, p = .45, meaning that support did not explain the correct response rates beyond chance. Specifically, interviewer support did not increase children’s likelihood of giving correct responses versus misses, correct responses versus false identifications, or correct responses versus don’t know answers, all ps > .71.

Working Memory and Interaction

Having established that there was no main effect of social support on children’s accuracy, we next tested the influence of the other variables, alone and in interaction with social support. We added working memory capacity as a continuous predictor (centered) into our regression model. The model including working memory capacity was significantly stronger in predicting responses compared to a reduced model (i.e., a model without working memory), χ2 (3, N = 72) = 8.85, p = .031, indicating that working memory capacity had a significant main effect on children’s responses. Specifically, the higher children’s working memory capacity, less likely they were to give a false-identification response (the worst type of error, in legal terms) versus a correct response. In other words, as working memory capacity increased, the odds of giving a false versus correct response decreased, B = − 0.28, Wald = 6.34, p = .012, OR = 0.76, 95% CI [0.61, 0.94] (Table 3, Main Effects Model). The other comparisons—correct responses versus misses and don’t know responses—were not significant.
Table 3

Multinomial logistic regression coefficients for interviewer support and working memory capacity effects on correct versus other response rates

 

B

Wald

p

OR

OR 95% CI

LL

UL

Main effects model–working memory capacity

  Miss

Intercept

0.63

0.93

.33

   

Interviewer support study 1

0.86

1.21

.27

2.36

0.51

10.86

Interviewer support current study

− 0.33

0.18

.67

0.72

0.16

3.28

Working memory capacity

− 0.05

0.47

.49

0.95

0.81

1.10

  False ID

Intercept

− 0.87

0.94

.33

   

Interviewer support study 1

0.84

0.78

.38

2.31

0.36

14.89

Interviewer support current study

0.49

0.27

.60

1.64

0.25

10.51

Working memory capacity

− 0.28

6.34

.01

0.76

0.61

0.94

  Do not know

Intercept

1.22

3.89

.05

   

Interviewer support study 1

− 0.18

0.05

.82

0.84

0.19

3.75

Interviewer support current study

− 0.06

0.01

.94

0.94

0.22

4.10

Working memory capacity

− 0.09

1.34

.25

0.92

0.79

1.06

Interaction model interviewer support × working memory capacity

  Miss

Intercept

0.75

1.13

.29

   

Interviewer support study 1

0.76

0.91

.34

2.13

0.45

10.15

Interviewer support current study

− 0.47

0.33

.57

0.62

0.12

3.12

Working memory capacity

− 0.09

0.80

.37

0.91

0.75

1.11

Support × working memory

0.10

0.36

.55

1.11

0.80

1.53

  False ID

Intercept

− 0.49

0.28

.59

   

Interviewer support study 1

0.68

0.50

.48

1.98

0.30

13.11

Interviewer support current study

− 0.06

0.00

.95

0.94

0.11

7.68

Working memory capacity

− 0.21

2.03

.15

0.81

0.60

1.08

Support × working memory

− 0.08

0.14

.71

0.92

0.59

1.43

  Do not know

Intercept

1.32

3.88

.05

   

Interviewer support study 1

− 0.26

0.11

.73

0.77

0.17

3.54

Interviewer support current study

− 0.17

0.05

.83

0.84

0.18

3.96

Working memory capacity

− 0.12

1.53

.22

0.88

0.73

1.07

Support × working memory

0.09

0.32

.57

1.10

0.80

1.51

Adding the interaction term between the centered working memory measure and interviewer support did not improve fit, with the model not significantly different from the intercept-only model, χ2 (12, N = 72) = 15.92, p = .195; the interaction terms did not significantly predict the likelihood of correct responses versus each of the other types of responses, all ps > .547 (Table 3, Interaction Model for specific coefficients for each comparison). In other words, children with higher working memory capacity were more accurate—that is, more likely to correctly identify the sitter instead of one of the fillers—than children with lower working memory capacity, regardless of whether the interviewer behaved in a supportive manner or not.

Social Support Reserves and Interaction

Next, in separate analyses, we tested the main and moderating effects of social support reserves. We added the centered social support reserves variable to the multinomial regression that included the experimental manipulations of interviewer-provided support. The main effects model with social support reserves was not significantly different from the reduced model, χ2 (3, N = 72) = 0.78, p = .85: Individual differences in children’s level of social support reserves had no significant main effect on children’s responses, all ps > .40 (Table 4, Main Effects Model). Adding the interaction between social support reserves and interviewer support, however, significantly improved model fit compared to a model without the interaction term, χ2 (3, N = 72) = 8.65, p = .03. As Table 4 illustrates, the improved fit was due primarily to a significant social support reserves by interviewer support interaction on the likelihood that children would give a correct response versus a false identification, B = 8.46, Wald = 4.76, p = .029, OR = 470.45, 95% CI [2.39, 938.17]. None of the other comparison interactions were significant.
Table 4

Multinomial logistic regression coefficients for interviewer support and social support reserves effects on correct versus other response rates

 

B

Wald

p

OR

OR 95% CI

LL

UL

Main effects model–social support reserves

  Miss

Intercept

0.59

0.84

.36

   

Interviewer Support Study 1

0.78

0.99

.32

2.19

0.47

10.19

Interviewer Support Current Study

− 0.27

0.13

.72

0.76

0.17

3.44

Support Reserves

− 0.58

0.36

.55

0.56

0.09

3.71

  False ID

Intercept

− 0.65

0.61

.55

   

Interviewer Support Study 1

0.89

0.95

.43

2.44

0.41

14.57

Interviewer Support Current Study

0.54

0.36

.33

1.72

0.29

10.01

Support Reserves

− 0.46

0.17

.55

0.63

0.07

5.51

  Do not know

Intercept

1.16

3.66

.06

   

Interviewer Support Study 1

− 0.29

0.14

.71

0.75

0.17

3.40

Interviewer Support Current Study

0.03

0.00

.97

1.03

0.24

4.39

Support Reserves

− 0.78

0.69

.41

0.46

0.07

2.89

Interaction model interviewer support × social support reserves

  Miss

Intercept

0.63

0.91

.34

   

Interviewer support study 1

0.67

0.68

.41

1.96

0.39

9.75

Interviewer support current study

0.72

0.33

.57

2.06

0.17

24.28

Support reserves

0.25

0.06

.81

1.28

0.17

9.64

Support × support reserves

− 4.40

1.81

.18

0.01

0.00

7.54

  False ID

Intercept

− 1.00

0.98

.32

   

Interviewer support study 1

0.85

0.77

.38

2.34

0.35

15.60

Interviewer support current study

1.85

1.61

.20

6.33

0.37

109.18

Support reserves

2.90

1.67

.20

18.13

0.22

1472.50

Support × support reserves

− 8.46

4.76

.03

0.00

0.00

0.42

  Do not know

Intercept

1.20

3.70

.05

   

Interviewer support study 1

− 0.37

0.20

.65

0.69

0.14

3.40

Interviewer support current study

1.03

0.68

.41

2.81

0.24

32.35

Support reserves

0.43

0.17

.68

1.53

0.20

11.91

Support × support reserves

− 5.51

2.89

.09

0.00

0.00

2.32

To test the hypothesis that interviewer-provided social support helped those children who needed it most (i.e., children lower in pre-existing social support reserves), we tested the simple slope of interviewer support at high (1 SD above the mean) and low (1 SD below the mean) levels of support reserves, by computing new support reserves variables, centered at −/+ 1 SD above the mean, and new interaction terms between these and interviewer support. We found that, for children high in social support reserves, the effect of interviewer-provided support on likelihood of giving a correct response versus false identification was not significant, B = 1.99, Wald = 2.34, p = .126, OR = 7.32, 95% CI [0.57, 93.54]. For children low in social support reserves, the effect of interviewer support was borderline significant, B = − 5.68, Wald = 3.71, p = .05, OR = 0.003, 95% CI [0.000, 1.11] (Fig. 1). The negative sign suggests that low-reserves children were less likely to give a false-identification (vs. correct) response when interviewed in a supportive (vs. non-supportive) manner—in other words, for children low in support reserves, interviewer support decreased the odds of false identifications versus correct responses, as predicted. Even so, children who should have performed the best (high in support reserves, received supportive interview), did not—looking much like those who were low in reserves and received a non-supportive interview.
Fig. 1

Effect of interviewer-provided social support and pre-existing support reserves on children’s odds ratio of providing a correct response versus false identification. The graph represents the simple slopes of support reserves when children were interviewed in a supportive versus non-supportive manner

State Anxiety After the Interview

Because prior literature is inconclusive on whether children’s anxiety during forensic interviews is related to the accuracy of their responses (e.g., Almerigogna et al. 2007; Davis and Bottoms 2002), we tested the main effect of state anxiety on children’s photo ID responses in the present study. Contrary to predictions, children's state anxiety levels were not related to their likelihood of selecting correct versus other types of responses, all Bs < − 0.30, all ps > .84, all OR 95% CIs include 1. Together with the lack of interviewer support main effect on accuracy, these findings precluded us from specifically testing for mediation. We did find, however, that children were significantly less anxious when interviewed in a supportive (M = 1.33, SD = 0.22) versus non-supportive (M = 1.45, SD = 0.26) manner, F(1, 71) = 4.63, p = .035, η2p = .06, which is in line with prior findings illustrating that interviewer support can either reduce anxiety or reduce the negative effects of anxiety on children’s performance. This has potential implications for other areas where anxiety hurts children’s accuracy in forensic contexts.

Confidence in Lineup Identification

Children were not particularly confident in their identification, with 34% indicating they were “just guessing,” 18% that they were “not so sure,” 41% that they were “kind of sure,” and 7% that they were “really, really sure.” Identification accuracy and confidence were not significantly related, χ2 (6, N = 44) = 7.64, p = .27, nor was children’s confidence related to interviewer support, F(1, 43) = 0.55, p = .462.

Discussion

We investigated the accuracy of 7- to 8-year-old children’s person identification after a very long delay (one year) and, for the first time, as a function of interviewer-provided socio-emotional support, children’s pre-existing social support reserves, state anxiety, and individual differences in children’s working memory capacity. With one exception (i.e., Goodman et al. 1991b, Study 4), ours is the only study we know of investigating children’s photo identification accuracy after such a long delay.

Overall, children made few correct identifications, but also few false identifications—which is both good and bad news for the legal system. That is, their accuracy over this long period of time (when no initial identification was made immediately after the event) was not good—like adults’ facial recognition performance, which is also drastically impaired over long periods (Deffenbacher et al. 2008). Their self-rated confidence was also not a good indicator of their accuracy. Even so, the children were cautious to not mis-identify the wrong person, preferring to give “do not know” responses.

Our results highlight the importance of having an explicit “I don’t know” response option, particularly when memory becomes unreliable due to time delays. Providing children with a question-mark card that allowed them to engage in the same behavioral response as they would for identifying a suspect—pointing—facilitated children’s endorsement of this response. Our approach was inspired by Saywitz’s pioneering work illustrating the importance of non-leading visual cues when it comes to interviewing young children. As part of her Narrative Elaboration interviewing protocol, external memory aids were demonstrated to improve free and guided recall without generating unintended errors (e.g., Saywitz and Camparo 2014; Saywitz and Snyder 1996).

Providing such intuitive, visual cues is important because children might not perceive “I don’t know” to be an option unless told so explicitly, which is why they tend to not give this response without a specific prompt (Schwartz-Kenney et al. 1996). In support, Dekle et al. (1996) found that children were less likely than adults to use an “I don’t know” option (see also Roebers and Schneider 2001), and Schwartz-Kenney et al. found that children were more likely to respond with “I don’t know” when it was presented as a physical option (vs. not). Dekle et al. and others theorize that children might provide incorrect (i.e., false identification) responses because they do not recognize their own uncertainty as well as adults do (e.g., Fitzgerald and Price 2015). Tobey and Goodman et al. (1992) found that children asked about a person’s age during a verbal recall task were less accurate than children who were given an age photo lineup task, but the latter made more commission errors. Children who truly do not know/remember are susceptible to giving false identification responses, presumably because a line up is inherently suggestive of the need to pick something. That is, with a typical photo lineup, children might assume that it is better to guess than give no response or say “I don’t know” if they are uncertain of the correct response (Schwartz-Kenney et al; King and Yuille 1987). In addition, developmental differences in moral decision-making and perspective taking can help explain younger children’s (vs. older children’s and adults’) tendency to make false identifications, as young children might fail to recognize the moral consequences of falsely identifying someone or might perceive it as worse not to identify anyone at all (Spring et al. 2013). In our study—where children had the question-mark response option—children had high rates of “I don’t know” and low false identification rates.

The Role of Social Support

We investigated the role of interviewer-provided social support in increasing accuracy, being especially interested in whether it could increase correct identification rates compared to false identifications. Prior forensic interview analogue studies have shown that interviewer-provided social support improves accuracy in two ways (for review, see Bottoms et al. 2007; Saywitz et al. 2019): (a) by improving children’s cognitive/memory performance, reflected for example in reduced number of inaccuracies during free recall after a delay (e.g., Goodman et al. 1991a), and (b) by reducing children’s suggestibility, reflected for example in decreased commission-error rates in response to misleading questions after no delay (Carter et al. 1996; Davis and Bottoms 2002).

Although the published literature on interviewer social support and lineup identifications is surprisingly sparse (e.g., Goodman et al. 1991a; Rush et al. 2014), we reasoned that the identification task is suggestive in and of itself due to children’s assumption that the interviewer is showing them a lineup that includes the target person, and so their job is to identify that target. This assumption might increase children’s perceived social pressure to choose someone, which children are less able to resist compared to adults (Fitzgerald and Price 2015). We expected interviewer support to increase the correct identification rates compared to all other possible responses (i.e., improve memory), but especially compared to false identifications, because false identifications might be reflections of children trying to please the interviewer by choosing a picture even though they did not recognize the babysitter. Therefore, we expected social support to both improve memory and reduce suggestibility.

Our results did not support these hypotheses—interviewer support had no main effect on correct identification rates after a one-year delay. In fact, Goodman et al. (1991a) and Rush et al. (2014), the only other direct investigations of interviewer support effects on identification accuracy, also failed to find a main effect of support. Instead, in their study, interviewer support only helped children who had been in a high-stress condition at encoding—suggesting that interviewer support matters the most when children’s accuracy is threatened by some pre-existing disadvantage, a theory discussed by Bottoms et al. (2007). In fact, in keeping with that theory, the Rush et al. results, and our hypotheses, we did find a significant interaction between social support reserves and interviewer support: Interviewer support helped children who were at a disadvantage—those who had lower social support reserves—by greatly increasing their correct response versus false identification rates. But interviewer support had no significant effect on children with higher social support reserves—and indeed, the relationship trended in the opposite direction, with interviewer-provided support slightly decreasing their accuracy. We will not speculate much about this trend, given that it was not significant. We believe it indicates that more studies are needed to confirm that the helpful effects of social support on children’s memory for events extends to their lineup identification accuracy.

Working Memory Capacity

As hypothesized, working memory capacity matters in children’s photo lineup identifications: Children with higher working memory capacity made more correct versus false identifications, an effect that could be attributed to decreased suggestibility. It was surprising that higher working memory was not related to increased correct responses versus misses, an effect that could be attributed more to cognitive mechanisms (memory). The latter could be due to the high miss versus correct response overall, given the long time delay. Although the relationship in this study, as in others, is correlational, children with higher working memory capacity might be better not only at selecting, storing, and retrieving information appropriately, but also better able to rely on meta-cognitive processes that would inform them that they are unsure and would prevent false identifications. In support, working memory capacity in children is related to metacognitive abilities (Alloway et al. 2009; Freeman et al. 2017), and working memory affects performance when metacognitive capabilities are also high (Whitebread 1999). Thus, working memory capacity significantly predicted children’s correct (versus false alarm) responses either through direct, causal mechanisms or because it is part of a cluster of cognitive and cognitive characteristics related to lineup identification performance.

State Anxiety

Contrary to our predictions, state anxiety during retrieval was unrelated to children’s accuracy. This lack of a correlation could reflect the low anxiety level overall be, regardless of experimental conditions. Children were not extremely stressed by the play session a year before, nor were they stressed during the interview—likely not enough for anxiety-caused cognitive load and cognitive interference to distract from attention and retrieval processes. Yet it is important to continue including anxiety in studies involving children and social support in particular, because, as we found in the current study, social support does decrease anxiety and this might be crucial (a) in actual forensic contexts when children might be more anxious/stressed, (b) for children who become anxious easily, and (c) for children testifying about a stressful experience.

In addition, anxiety has been proposed not only as a mediator (i.e., state anxiety during the interview), but also as a moderator (i.e., trait or independently manipulated anxiety) of support effects on children’s accuracy. Social support fosters effective appraisal and coping strategies, which can improve children’s emotional response to stressful situations and can minimize the disruptive influence of threat-related thoughts that might otherwise impede performance (Cohen 2004; Derakshan and Eysenck 2009; Saywitz et al. 2019). Thus, interviewer support might be beneficial to children who experienced anxiety during encoding (Rush et al. 2014) because it reduces state anxiety experienced when children are asked to recall a stressful event. In our study, we relied on measured state anxiety conceptualized as a causal result of social support and predictor of accuracy—but its role is potentially more complex and task-dependent. Future research on delayed interviewing could combine manipulations of anxiety during encoding with measures of state anxiety during retrieval to test whether interviewer support ameliorates negative effects of encoding-anxiety by helping reduce anxiety during retrieval.

Confidence and Accuracy

Some research suggests that children have significant metacognitive limitations compared to adults, and we believe this is reflected by the lack of relationship between confidence and accuracy of photo lineup identifications (see also Keast et al. 2007). That is, a strong relationship between confidence in one’s response and the accuracy of that response indicates adequate monitoring of cognitive processes and increased awareness of one’s cognitive limitations or strengths—in short, strong metacognitive abilities. In contrast, a weak relationship between confidence and accuracy denotes poor metacognitive abilities. As have others before (e.g., Brackmann et al. 2019; Humphries and Flowe 2015), we found that children’s perceived confidence is not indicative of their actual accuracy—while it is also unrelated to the false identification rates. Unlike others, however, we cannot attribute the confidence-accuracy discrepancy to children’s overconfidence (Brewer and Day 2005; Keast et al. 2007): only 7% of children in our study indicated they were “really, really sure.” Therefore, we conclude that children were not overly confident in their ability to identify the babysitter after prolonged delay. Nevertheless, our findings illustrate that children’s confidence—such as it is—should not be interpreted by investigators and interviewers as a heuristic cue for accuracy.

Implications

Our study extends extant literature on children’s accuracy in forensic contexts in two ways. First, it is one of the few studies to investigate children’s accuracy after a prolonged delay—one year after the event. Often, child abuse goes undiscovered for long periods because children tend to be secretive about the abuse for various reasons (Bottoms et al. 2016; London et al. 2005). There is, in fact, evidence that among child abuse cases referred for prosecution, many children disclosed 6 months or one year after the abuse took place (Elliott and Briere 1994; Goodman et al. 1992). Thus, it is plausible that many child victims end up performing suspect identifications after extremely long delays—yet their performance in such circumstances is understudied. From our research (and that of Goodman et al. 1991b), we conclude that children’s memory is poor after such long delays—children most often fail to identify anyone or say they do not know. Of course, our generalization to actual cases is limited by the fact that our original event involving the babysitter differs from an abusive incident in important ways: Children were not distressed or anxious, they were introduced to a nice young woman in a toy room, and knew their parent was right next door throughout the interaction.

Second, ours is only the third study (to our knowledge) to investigate the effects of interviewer-provided social support on children’s lineup identifications. Although we now have a robust body of evidence for the importance of social support for children’s eyewitness accuracy in general (Bottoms et al. 2007; Saywitz et al. 2019), future studies should focus more specifically on its role in lineup identifications. As much as recalling accurate details about specific events, children’s ability to correctly identify a suspect—and to correctly recognize when they cannot do so—has tremendous implications for the administration of justice. Often, those who abuse children do so in secret, taking advantage of the children’s naiveté, desire to please adults, and/or reduced physical strength. When the abusers are strangers, children’s ability to identify them can be the only conclusive evidence for ensuring indictment and/or conviction in a court of law.

Although it can be argued that most abuse is perpetrated by people known to the child (e.g., Ullman 2007), incidents of attempted and completed stranger sexual assault are unfortunately all too common. For example, stranger assaults made up 35% of total victimization incidents in one sample of children (Gallagher et al. 2008). In these cases, investigators must rely on the child’s recall of what happened, but also on their ability to identify a stranger suspect. Thus, it is important to test whether commonly endorsed interviewing techniques—such as visual aids and social support—are likely to increase correct identification rates. We found tentative support for both these techniques. Specifically, children in our study were clearly comfortable endorsing an “I don’t know” response, despite prior research illustrating that children are generally reluctant to volunteer this response verbally. This was likely due to our inclusion, among lineup pictures, of the question-mark card that signaled to children “I don’t know” is a normative, acceptable response (Schwartz-Kenney et al. 1996). Further, we found that interviewer support might help children who have low social support reserves. This is important, because the literature on childhood victimization illustrates just how vulnerable to abusers children become when they lack adequate social support from parents and peers. Social isolation is a strong predictor of child sexual abuse—for example, not having somebody in whom to confide (Fleming et al. 1997), and growing up in unhappy families (Finkelhor et al. 1990). Thus, if children who lack social support reserves are at higher risk of victimization, it is these very children who would benefit the most when socially supportive forensic interviewing helps them correctly identify those harmed them.

Conclusion

Cognitive (working memory capacity) and socio-emotional (social support) factors contribute to children’s ability to identify “suspects” from target-present lineups after a long, ecologically valid and realistic delay of one year. Research on best practices for forensic interviews in general and lineup identifications in particular can inform thoughtful employment of resources to further children’s physical and psychological security (Melton 2018). Most importantly, when forensic interviews help children identify perpetrators of abuse, those abusers are apprehended and the threat they pose to children’s safety is reduced. In addition, when children identify the wrong suspect, the real perpetrator remains at large and continues to pose a threat to children. Also of importance, actors in the legal system must be mindful of the psychological effects of forensic interviews on children’s psychological wellbeing. We have found that supportive interviews reduce children’s anxiety even during a relatively benign laboratory study where children had to remember a pleasant play session with a babysitter. Such supportive methods would prove even more beneficial when children who have suffered traumatic events must re-experience them in order to help law enforcement. Thus, a secondary benefit of supportive interviewing to children’s psychological safety is attempting to reduce children’s negative experiences when they come in contact with the legal system. Finally, our research indicates that interviewer support is particularly important for children who lack social support reserves—perhaps improving children’s ability to identify a perpetrator for those who need it most.

Notes

Acknowledgments

We thank Tamara Haegerich and Suzanne Davis for early leadership on this project. We thank Jason Rohacs, Alon Stein, Rich Reyes, Lisa Tockman, Kelly Ricketts, Danielle Brandstetter, Sonja Veille, Kara Doering, Matthew Badanek, Nora O’Malley, Jessica Dilley, Michelle Prestige, Michelle Gap, Erika Chen, Laurie Guy, and Elaine Shreder for valuable research assistance. We also extend thanks to the families who participated in the study. Some toys and prizes were provided by Chicago International Children’s Film Festival, Cineplex Odeon, Little Ceasars, and Taco Bell.

Funding Information

Funding for this research was provided by a BSTART grant to Bette L. Bottoms from the National Institute of Mental Health.

Compliance with Ethical Standards

Conflict of Interest

The authors declare that they have no conflict of interest.

References

  1. Alexander, K. W., Quas, J. A., & Goodman, G. S. (2002). Theoretical advances in understanding children’s memory for distressing events: The role of attachment. Developmental Review, 22, 490–519.Google Scholar
  2. Alloway, T. P., Gathercole, S. E., Kirkwood, H., & Elliott, J. (2009). The cognitive and behavioral characteristics of children with low working memory. Child Development, 80, 606–621.  https://doi.org/10.1111/j.1467-8624.2009.01282.x.CrossRefGoogle Scholar
  3. Almerigogna, J., Ost, J., Bull, R., & Akehurst, L. (2007). State of high anxiety: How non-supportive interviewers can increase the suggestibility of child witnesses. Applied Cognitive Psychology, 21, 963–974.  https://doi.org/10.1002/acp.1311.CrossRefGoogle Scholar
  4. Baddeley, A. (2010). Working memory. Current Biology, 20, R136–R140.  https://doi.org/10.1016/j.cub.2009.12.014.CrossRefGoogle Scholar
  5. Baddeley, A. D., & Hitch, G. (1974). Working memory. In G. A. Bower (Ed.), The psychology of learning and motivation (Vol. 8, pp. 47–89). New York: Academic Press.Google Scholar
  6. Bahrick, H. P., Bahrick, P. O., & Wittlinger, R. P. (1975). Fifty years of memory for names and faces: A cross-sectional approach. Journal of experimental psychology: General, 104, 54–75.  https://doi.org/10.1037/0096-3445.104.1.54.CrossRefGoogle Scholar
  7. Bandura, A. (1982). Self-efficacy mechanism in human agency. American Psychologist, 37, 122–147.  https://doi.org/10.1037/0003-066X.37.2.122.CrossRefGoogle Scholar
  8. Bandura, A. (1986). Social foundations of thought and action: A social cognitive theory. Englewood Cliffs: Prentice-Hall.Google Scholar
  9. Bandura, A. (1997). Self-efficacy: The exercise of control. New York: Freeman.Google Scholar
  10. Bohannon, J. N. (1988). Flashbulb memories for the space shuttle disaster: a tale of two theories. Cognition, 29, 179–196.  https://doi.org/10.1016/0010-0277(88)90036-4.CrossRefGoogle Scholar
  11. Bottoms, B. L., Quas, J. A., & Davis, S. L. (2007). The Influence of interviewer-provided social support on children’s suggestibility, memory, and disclosures. In M. E. Pipe, M. E. Lamb, Y. Orbach, & A. C. Cederborg (Eds.), Child sexual abuse: Disclosure, delay, and denial (pp. 135–158). Mahwah: Erlbaum.Google Scholar
  12. Bottoms, B. L., Peter-Hagene, L. C., Epstein, M. A., Wiley, T. R. A., Reynolds, C. E., & Rudnicki, A. G. (2016). Abuse characteristics and individual differences related to disclosing childhood sexual, physical, and emotional abuse and witnessed domestic violence. Journal of Interpersonal Violence, 31, 1308–1339.  https://doi.org/10.1177/0886260514564155.CrossRefGoogle Scholar
  13. Brackmann, N., Sauerland, M., & Otgaar, H. (2019). Developmental trends in lineup performance: adolescents are more prone to innocent bystander misidentifications than children and adults. Memory & Cognition, 47, 428–440.  https://doi.org/10.3758/s13421-018-0877-6.CrossRefGoogle Scholar
  14. Brewer, N., & Day, K. (2005). The confidence-accuracy and decision latency-accuracy relationships in children's eyewitness identification. Psychiatry, Psychology and Law, 12, 119–128.  https://doi.org/10.1375/pplt.2005.12.1.119.CrossRefGoogle Scholar
  15. Brewer, N., & Wells, G. L. (2006). The confidence-accuracy relationship in eyewitness identification: effects of lineup instructions, foil similarity, and target-absent base rates. Journal of Experimental Psychology: Applied, 12, 11–30.  https://doi.org/10.1037/1076-898X.12.1.11.CrossRefGoogle Scholar
  16. Brodes v. State, 279 Ga. 435, 614 S.E.2d 766 (2005).Google Scholar
  17. Bruer, K. C., & Price, H. L. (2017). A repeated forced-choice line-up procedure provides suspect bias information with no cost to accuracy for older children and adults (Confidence-accuracy relationship) [Supplemental material]. Applied Cognitive Psychology, 31, 448–466.  https://doi.org/10.1002/acp.3342.CrossRefGoogle Scholar
  18. Brunner, R., Parzer, P., Haffner, J., Steen, R., Roos, J., Klett, M., & Resch, F. (2007). Prevalence and psychological correlates of occasional and repetitive deliberate self-harm in adolescents. Archives of Pediatrics & Adolescent Medicine, 161, 641–649.  https://doi.org/10.1001/archpedi.161.7.641.CrossRefGoogle Scholar
  19. Burleson, B. R., Albrecht, T. L., Goldsmith, D. J., & Sarason, I. G. (1994). Introduction: The communication of social support. In B. R. Burleson, T. L. Albrecht, & I. G. Sarason (Eds.), Communication of social support: Messages, interactions, relationships, and community (pp. xi–xxx). Thousand Oaks: Sage.Google Scholar
  20. Carter, C. A., Bottoms, B. L., & Levine, M. (1996). Linguistic and socioemotional influences on the accuracy of children’s reports. Law and Human Behavior, 20, 335–358.  https://doi.org/10.1007/BF01499027.CrossRefGoogle Scholar
  21. Case, R., Kurland, D. M., & Goldberg, J. (1982). Operational efficiency and the growth of short-term memory span. Journal of Experimental Child Psychology, 33, 386–404.  https://doi.org/10.1016/0022-0965(82)90054-6.CrossRefGoogle Scholar
  22. Castaneda, A., McCandless, B. R., & Palermo, D. S. (1956). The children’s form of the Manifest Anxiety Scale. Child Development, 27, 317–326.  https://doi.org/10.2307/1126201.CrossRefGoogle Scholar
  23. Center on Wrongful Convictions, & Possley, M. (2019, January 29). Jacques Rivera. The National Registry of Exonerations. Retrieved from http://www.law.umich.edu/special/exoneration/Pages/casedetail.aspx?caseid=3815
  24. Cohen, S. (2004). Social relationships and health. American Psychologist, 59, 676–684.  https://doi.org/10.1037/0003-066X.59.8.676.CrossRefGoogle Scholar
  25. Cohen, S., & Wills, T. A. (1985). Stress, social support, and the buffering hypothesis. Psychological Bulletin, 98, 310–357.  https://doi.org/10.1037/0033-2909.98.2.310.CrossRefGoogle Scholar
  26. Conway, A. R. A., & Engle, R. W. (1996). Individual differences in working memory capacity: more evidence for a general capacity theory. Memory, 4, 577–590.  https://doi.org/10.1080/741940997.CrossRefGoogle Scholar
  27. Conway, A. R. A., Cowan, N., & Bunting, M. F. (2001). The cocktail party phenomenon revisited: the importance of working memory capacity. Psychonomic Bulletin & Review, 8, 331–335.  https://doi.org/10.3758/BF03196169.CrossRefGoogle Scholar
  28. Crookes, K., & McKone, E. (2009). Early maturity of face recognition: no childhood development of holistic processing, novel face encoding, or face-space. Cognition, 111, 219–247.  https://doi.org/10.1016/j.cognition.2009.02.004.CrossRefGoogle Scholar
  29. Davis, S. L., & Bottoms, B. L. (2002). Effects of social support on children’s eyewitness reports: a test of the underlying mechanism. Law and Human Behavior, 26, 185–215.  https://doi.org/10.1023/A:1014692009941.CrossRefGoogle Scholar
  30. Deffenbacher, K. A., Bornstein, B. H., Penrod, S. D., & McGorty, E. K. (2004). A meta-analytic review of the effects of high stress on eyewitness memory. Law and Human Behavior, 28, 687–706.  https://doi.org/10.1007/s10979-004-0565-x.CrossRefGoogle Scholar
  31. Deffenbacher, K. A., Bornstein, B. H., McGorty, E. K., & Penrod, S. D. (2008). Forgetting the once-seen face: estimating the strength of an eyewitness's memory representation. Journal of Experimental Psychology: Applied, 14, 139–150.  https://doi.org/10.1037/1076-898X.14.2.139.CrossRefGoogle Scholar
  32. Dekle, D. J., Beal, C. R., Elliott, R., & Huneycutt, D. (1996). Children as witnesses: a comparison of lineup versus showup identification methods. Applied Cognitive Psychology, 10, 1-12. doi: https://doi.org/10.1002/(SICI)1099-0720(199602)10:13.0.CO;2-Y
  33. Derakshan, N., & Eysenck, M. W. (2009). Anxiety, processing efficiency, and cognitive performance: new developments from attentional control theory. European Psychologist, 14, 168–176.  https://doi.org/10.1027/1016-9040.14.2.168.CrossRefGoogle Scholar
  34. Eisen, M. L., Qin, J., Goodman, G. S., & Davis, S. L. (2002). Memory and suggestibility in maltreated children: Age, stress arousal, dissociation, and psychopathology. Journal of Experimental Child Psychology, 83, 167–212.Google Scholar
  35. Eisen, M. L., Goodman, G. S., Qin, J., Davis, S., & Crayton, J. (2007). Maltreated children's memory: Accuracy, suggestibility, and psychopathology. Developmental Psychology, 43, 1275.Google Scholar
  36. Eisen, M. L., Goodman, G. S., Diep, J., Lacsamana, M. T., Olomi, J., Goldfarb, D., & Quas, J. A. (2019). Effects of interviewer support on maltreated and at-risk children’s memory and suggestibility. International Journal on Child Maltreatment: Research, Policy and Practice, 1–24.  https://doi.org/10.1007/s42448-019-00016-7.CrossRefGoogle Scholar
  37. Elliott, D. M., & Briere, J. (1994). Forensic sexual abuse evaluations of older children: disclosures and symptomatology. Behavioral Sciences & the Law, 12, 261–277.  https://doi.org/10.1002/bsl.2370120306.CrossRefGoogle Scholar
  38. Engel de Abreu, P. M., Conway, A. R., & Gathercole, S. E. (2010). Working memory and fluid intelligence in young children. Intelligence, 38, 552–561.  https://doi.org/10.1016/j.intell.2010.07.003.CrossRefGoogle Scholar
  39. Engle, R. W., Tuholski, S. W., Laughlin, J. E., & Conway, A. R. (1999). Working memory, short-term memory, and general fluid intelligence: a latent-variable approach. Journal of Experimental Psychology: General, 128, 309–331.  https://doi.org/10.1037/0096-3445.128.3.309.CrossRefGoogle Scholar
  40. Eysenck, M. W., & Calvo, M. G. (1992). Anxiety and performance: the processing efficiency theory. Cognition and Emotion, 6, 409–434.  https://doi.org/10.1080/02699939208409696.CrossRefGoogle Scholar
  41. Eysenck, M. W., Derakshan, N., Santos, R., & Calvo, M. G. (2007). Anxiety and cognitive performance: attentional control theory. Emotion, 7, 336–353.  https://doi.org/10.1037/1528-3542.7.2.336.CrossRefGoogle Scholar
  42. Finkelhor, D., Hotaling, G., Lewis, I. A., & Smith, C. (1990). Sexual abuse in a national survey of adult men and women: prevalence, characteristics, and risk factors. Child Abuse & Neglect, 14, 19–28.  https://doi.org/10.1016/0145-2134(90)90077-7.CrossRefGoogle Scholar
  43. Finkelhor, D., Hammer, H., & Sedlack, A. J. (2008). Sexually assaulted children: National estimates and characteristics. Juvenile Justice Bulletin – NCJ214383 (pp. 1-12). Washington, DC: US Government Printing Office.Google Scholar
  44. Fischer, K. W. (1980). A theory of cognitive development: the control and construction of hierarchies of skills. Psychological review, 87, 477–531.  https://doi.org/10.1037/0033-295X.87.6.477.CrossRefGoogle Scholar
  45. Fischer, K. W., Bullock, D. H., Rotenberg, E. J., & Raya, P. (1993). The dynamics of competence: How context contributes directly to skill. In R. H. Wozniak & K. W. Fischer (Eds.), Development in context: Acting and thinking in specific environments (pp. 93–117). Hillsdale: Lawrence Erlbaum.Google Scholar
  46. Fitzgerald, R. J., & Price, H. L. (2015). Eyewitness identification across the life span: a meta-analysis of age differences. Psychological Bulletin, 141, 1228–1265.  https://doi.org/10.1037/bul0000013.CrossRefGoogle Scholar
  47. Fleming, J., Mullen, P., & Bammer, G. (1997). A study of potential risk factors for sexual abuse in childhood. Child Abuse & Neglect, 21, 49–58.  https://doi.org/10.1016/S0145-2134(96)00126-3.CrossRefGoogle Scholar
  48. Flin, R., Boon, J., Knox, A., & Bull, R. (1992). The effect of a five-month delay on children's and adults' eyewitness memory. British Journal of Psychology, 83, 323–336.  https://doi.org/10.1111/j.2044-8295.1992.tb02444.x.CrossRefGoogle Scholar
  49. Freeman, E. E., Karayanidis, F., & Chalmers, K. A. (2017). Metacognitive monitoring of working memory performance and its relationship to academic achievement in Grade 4 children. Learning and Individual Differences, 57, 58–64.  https://doi.org/10.1016/j.lindif.2017.06.003.CrossRefGoogle Scholar
  50. Gallagher, B., Bradford, M., & Pease, K. (2008). Attempted and completed incidents of stranger-perpetrated child sexual abuse and abduction. Child Abuse and Neglect, 32, 517–528.  https://doi.org/10.1016/j.chiabu.2008.02.002.CrossRefGoogle Scholar
  51. Goodman, G. S. (2006). Children's eyewitness memory: a modern history and contemporary commentary. Journal of Social Issues, 62, 811–832.  https://doi.org/10.1111/j.1540-4560.2006.00488.x.CrossRefGoogle Scholar
  52. Goodman, G. S., Rudy, L., Bottoms, B. L., & Aman, C. (1990). Children’s concerns and memory: Issues of ecological validity in the study of children’s eyewitness testimony. In R. Fivush & J. A. Hudson (Eds.), Knowing and remembering in young children (pp. 249–284). Cambridge: Cambridge University Press.Google Scholar
  53. Goodman, G. S., Bottoms, B. L., Schwartz-Kenney, B. M., & Rudy, L. (1991a). Children’s testimony about a stressful event: improving children’s reports. Journal of Narrative and Life History, 1, 69–99.  https://doi.org/10.1075/jnlh.1.1.05chi.CrossRefGoogle Scholar
  54. Goodman, G. S., Hirschman, J. E., Hepps, D., & Rudy, L. (1991b). Children’s memory for stressful events. Merrill-Palmer Quarterly, 37, 109–157.Google Scholar
  55. Goodman, G. S., Ogle, C. M., McWilliams, K., Narr, R. K., & Paz-Alonso, P. M. (2014). Forensic Context.Google Scholar
  56. Goodman, G. S., Taub, E. P., Jones, D. P., England, P., et al. (1992). Testifying in criminal court: Emotional effects on child sexual assault victims. Monographs of the Society for Research in Child Development, 57(5, Serial No. 299). doi: https://doi.org/10.2307/1166127 CrossRefGoogle Scholar
  57. Goodman, G. S., & Reed, R. S. (1986). Age differences in eyewitness testimony. Law and Human Behavior, 10, 317–332.Google Scholar
  58. Goodman, G. S., Quas, J. A., Batterman-Faunce, J. M., Riddlesberger, M. M., & Kuhn, J. (1997). Children's reactions to and memory for a stressful event: Influences of age, anatomical dolls, knowledge, and parental attachment. Applied Developmental Science, 1, 54–75.Google Scholar
  59. Harter. S. (1985). The social support scale for children. Unpublished manuscript, University of Denver.Google Scholar
  60. Hershkowitz, I. (2011). Rapport building in investigative interviews of children. In M. E. Lamb, D. J. La Rooy, L. C. Malloy, & C. Katz (Eds.), Children’s testimony: A handbook of psychological research and forensic practice (pp. 109–128). Chichester: Wiley.  https://doi.org/10.1002/9781119998495.ch6.CrossRefGoogle Scholar
  61. Hershkowitz, I., Orbach, Y., Lamb, M. E., Sternberg, K. J., Pipe, M. E., & Horowitz, D. (2007). Suspected victims of abuse who do not make allegations: An analysis of their interactions with forensic interviewers. In M. E. Pipe, M. E. Lamb, Y. Orbach, & A. C. Cederborg (Eds.), Child sexual abuse: Disclosure, delay, and denial (pp. 97–114). Mahwah, NJ: Erlbaum.Google Scholar
  62. Hershkowitz, I., Lamb, M. E., Katz, C., & Malloy, L. C. (2015). Does enhanced rapport-building alter the dynamics of investigative interviews with suspected victims of intra-familial abuse? Journal of Police and Criminal Psychology, 30, 6–14.  https://doi.org/10.1007/s11896-013-9136-8.CrossRefGoogle Scholar
  63. Hollingshead, A. B. (1975). Four factor index of social status. Unpublished manuscript.Google Scholar
  64. Humphries, J. E., & Flowe, H. D. (2015). Receiver operating characteristic analysis of age-related changes in lineup performance. Journal of Experimental Child Psychology, 132, 189–204.  https://doi.org/10.1016/j.jecp.2014.12.009.CrossRefGoogle Scholar
  65. Imhoff, M. C., & Baker-Ward, L. (1999). Preschoolers’ suggestibility: effects of developmentally appropriate language and interviewer supportiveness. Journal of Applied Developmental Psychology, 20, 407–429.  https://doi.org/10.1016/S0193-3973(99)00022-.CrossRefGoogle Scholar
  66. Innocence Project. (2019). Eyewitness identification reform. Retrieved from https://www.innocenceproject.org/eyewitness-identification-reform/
  67. Jones, D. P., & Krugman, R. D. (1986). Can a three-year-old child bear witness to her sexual assault and attempted murder? Child Abuse & Neglect, 10, 253–258.  https://doi.org/10.1016/0145-2134(86)90086-4.CrossRefGoogle Scholar
  68. Kail, R. (1990). The development of memory in children. New York: Freeman.Google Scholar
  69. Karageorge, A., & Zajac, R. (2011). Exploring the effects of age and delay on children’s person identifications: verbal descriptions, lineup performance, and the influence of wildcards. British Journal of Psychology, 102, 161–183.  https://doi.org/10.1348/000712610X507902.CrossRefGoogle Scholar
  70. Kawachi, I., & Berkman, L. F. (2001). Social ties and mental health. Journal of Urban Health, 78, 458–467.  https://doi.org/10.1093/jurban/78.3.458.CrossRefGoogle Scholar
  71. Keast, A., Brewer, N., & Wells, G. L. (2007). Children’s metacognitive judgments in an eyewitness identification task. Journal of Experimental Child Psychology, 97, 286–314.  https://doi.org/10.1016/j.jecp.2007.01.007.CrossRefGoogle Scholar
  72. Kelley, D. H., & Gorham, J. (1988). Effects of immediacy on recall of information. Communication Education, 37, 198–207.  https://doi.org/10.1080/03634528809378719.CrossRefGoogle Scholar
  73. King, M. A., & Yuille, J. C. (1987). Suggestibility and the child witness. In S. J. Ceci, M. P. Toglia, & D. F. Ross (Eds.), Children’s eyewitness memory (pp. 24–35). New York: Springer-Verlag.CrossRefGoogle Scholar
  74. Klemfuss, J. Z., Milojevich, H. M., Yim, I. S., Rush, E. B., & Quas, J. A. (2013). Stress at encoding, context at retrieval, and children’s narrative content. Journal of Experimental Child Psychology, 116, 693–706.Google Scholar
  75. Lamb, M. E., Hershkowitz, I., Orbach, Y., & Esplin, P. W. (2008). Tell me what happened: Structured investigative interviews of child victims and witnesses. West Sussex: Wiley.  https://doi.org/10.1002/9780470773291.CrossRefGoogle Scholar
  76. Lindberg, M. A., Jones, S., McComas Collard, L., & Thomas, S. W. (2001). Similarities and differences in eyewitness testimonies of children who directly versus vicariously experience stress. The Journal of Genetic Psychology, 162, 314–333.  https://doi.org/10.1080/00221320109597486.CrossRefGoogle Scholar
  77. London, K., Bruck, M., Ceci, S. J., & Shuman, D. W. (2005). Disclosure of child sexual abuse: What does the research tell us about the ways that children tell? Psychology, Public Policy, and Law, 11, 194–226.  https://doi.org/10.1037/1076-8971.11.1.194.CrossRefGoogle Scholar
  78. Lowenstein, J. A., Blank, H., & Sauer, J. D. (2010). Uniforms affect the accuracy of children’s eyewitness identification decisions. Journal of Investigative Psychology and Offender Profiling, 7, 59–73.  https://doi.org/10.1002/jip.104.CrossRefGoogle Scholar
  79. Melton, G. B. (2018). Children’s right to personal security: the foundation for a new journal. International Journal on Child Maltreatment: Research, Policy and Practice, 1, 1–10.  https://doi.org/10.1007/s42448-018-0007-x.CrossRefGoogle Scholar
  80. Memon, A., Havard, C., Clifford, B., Gabbert, F., & Watt, M. (2011). A field evaluation of the VIPER system: a new technique for eliciting eyewitness identification evidence. Psychology, Crime & Law, 17, 711–729.  https://doi.org/10.1080/10683160903524333.CrossRefGoogle Scholar
  81. Myers, J. E. B., Redlich, A. D., Goodman, G. S., Prizmich, L. P., & Imwinkelried, E. (1999). Jurors' perceptions of hearsay in child sexual abuse cases. Psychology, Public Policy, and Law, 5, 388–419.  https://doi.org/10.1037/1076-8971.5.2.388.CrossRefGoogle Scholar
  82. Neil v. Biggers, 409 U.S. 188 (1972).Google Scholar
  83. Nevo, E., & Breznitz, Z. (2011). Assessment of working memory components at 6 years of age as predictors of reading achievements a year later. Journal of Experimental Child Psychology, 109, 73–90.  https://doi.org/10.1016/j.jecp.2010.09.010.CrossRefGoogle Scholar
  84. Oats, K., & Shrimpton, S. (1991). Children’s memories for stressful and non-stressful events. Medical Science and Law, 31, 4–10.Google Scholar
  85. Owens, M., Stevenson, J., Hadwin, J. A., & Norgate, R. (2014). When does anxiety help or hinder cognitive test performance? The role of working memory capacity. British Journal of Psychology, 105, 92–101.  https://doi.org/10.1111/bjop.12009.CrossRefGoogle Scholar
  86. Ozbay, F., Johnson, D. C., Dimoulas, E., Morgan III, C. A., Charney, D., & Southwick, S. (2007). Social support and resilience to stress: from neurobiology to clinical practice. Psychiatry, 4, 35–40.Google Scholar
  87. Papay, J. P., & Spielberger, C. D. (1986). Assessment of anxiety and achievement in kindergarten and first- and second-grade children. Journal of Abnormal Child Psychology, 14, 279–286.  https://doi.org/10.1007/BF00915446.CrossRefGoogle Scholar
  88. Pipe, M. E., Lamb, M. E., Orbach, Y., & Cederborg, A. C. (Eds.). (2007). Child sexual abuse: Disclosure, delay and denial. Mahwah: Erlbaum.Google Scholar
  89. Pozzulo, J. (2014). Person description and identification by child witnesses. In R. C. L. Lindsay, D. F. Ross, J. D. Read, & M. P. Toglia (Eds.), The handbook of eyewitness psychology Vol. II: Memory for people (pp. 283–308). New York: Psychology Press.Google Scholar
  90. Pozzulo, J. D., & Lindsay, R. C. L. (1998). Identification accuracy of children versus adults: a meta-analysis. Law and Human Behavior, 22, 549–570.  https://doi.org/10.1023/A:1025739514042.CrossRefGoogle Scholar
  91. Pozzulo, J. D., Dempsey, J., Bruer, K., & Sheahan, C. (2012). The culprit in target-absent lineups: understanding young children’s false positive responding. Journal of Police and Criminal Psychology, 27, 55–62.  https://doi.org/10.1007/s11896-011-9089-8.CrossRefGoogle Scholar
  92. Quas, J. A., & Lench, H. C. (2007). Arousal at encoding, arousal at retrieval, interviewer support, and children's memory for a mild stressor. Applied Cognitive Psychology, 21, 289–305.  https://doi.org/10.1002/acp.1279.CrossRefGoogle Scholar
  93. Quas, J. A., Bauer, A., & Boyce, W. T. (2004). Physiological reactivity, social support, and memory in early childhood. Child Development, 75, 797–814.  https://doi.org/10.1111/j.1467-8624.2004.00707.x.CrossRefGoogle Scholar
  94. Quas, J. A., Rush, E. B., Yim, I. S., & Nikolayev, M. (2014). Effects of stress on memory in children and adolescents: testing causal connections. Memory, 22, 616–632.  https://doi.org/10.1080/09658211.2013.809766.CrossRefGoogle Scholar
  95. Quas, J. A., Wallin, A. R., Papini, S., Lench, H., & Scullin, M. H. (2005). Suggestibility, social support, and memory for a novel experience in young children. Journal of Experimental Child Psychology, 91, 315-341.Google Scholar
  96. Rassin, E., Van Der Sleen, J., Van Amelsvoort, A. G., & Buttinger, E. (2006). Boys will be boys: a pre-line-up gameplay has beneficial effects on boys' but not girls’ identification performance. Journal of Investigative Psychology and Offender Profiling, 3, 157–163.  https://doi.org/10.1002/jip.50.CrossRefGoogle Scholar
  97. Roebers, C. M., & Schneider, W. (2001). Memory for an observed event in the presence of prior misinformation: developmental patterns of free recall and identification accuracy. British Journal of Developmental Psychology, 19, 507–524.  https://doi.org/10.1348/026151001166227.CrossRefGoogle Scholar
  98. Roebers, C. M., & Schneider, W. (2005). Individual differences in young children's suggestibility: relations to event memory, language abilities, working memory, and executive functioning. Cognitive Development, 20, 427–447.  https://doi.org/10.1016/j.cogdev.2005.05.006.CrossRefGoogle Scholar
  99. Rosenthal, R., & Jacobson, L. (1968). Pygmalion in the classroom: Teacher expectation and student intellectual development. New York: Holt, Rinehart, & Winston.CrossRefGoogle Scholar
  100. Rudy, L., & Goodman, G. S. (1991). Effects of participation on children's reports: implications for children's testimony. Developmental Psychology, 27, 527–538.  https://doi.org/10.1037/0012-1649.27.4.527.CrossRefGoogle Scholar
  101. Rush, E. B., Quas, J. A., Yim, I. S., Nikolayev, M., Clark, S. E., & Larson, R. P. (2014). Stress, interviewer support, and children’s eyewitness identification accuracy. Child Development, 85, 1292–1305.  https://doi.org/10.1111/cdev.12177.CrossRefGoogle Scholar
  102. Sarason, B. R., Sarason, I. G., & Pierce, G. R. (Eds.). (1990). Social support: An interactional view. New York: Wiley.Google Scholar
  103. Sayitz, K. J., Lyon, T. D., & Goodman, G. S. (2018). When interviewing children: A review and update. In J. B. Klika & J. B. Conte (Eds.), The APSAC handbook on child maltreatment (pp. 310–329). Thousand Oaks: Sage.Google Scholar
  104. Saywitz, K. J., & Camparo, L. B. (2014). Evidence-based child forensic interviewing: The developmental narrative elaboration interview. New York: Oxford University Press.Google Scholar
  105. Saywitz, K. J., & Snyder, L. (1996). Narrative elaboration: Test of a new procedure for interviewing children. Journal of Consulting and Clinical Psychology, 64, 1347–1357.  https://doi.org/10.1037/0022-006X.64.6.1347.CrossRefGoogle Scholar
  106. Saywitz, K. J., Wells, C. R., Larson, R. P., & Hobbs, S. D. (2019). Effects of interviewer support on children’s memory and suggestibility: systematic review and meta-analyses of experimental research. Trauma, Violence, & Abuse, 20, 22–39.  https://doi.org/10.1177/1524838016683457.CrossRefGoogle Scholar
  107. Schwartz-Kenney, B. M., Bottoms, B. L., & Goodman, G. S. (1996). Improving children’s person identification. Child Maltreatment, 1, 121–133.  https://doi.org/10.1177/1077559596001002004.CrossRefGoogle Scholar
  108. Scrivner, E., & Safer, M. A. (1988). Eyewitnesses show hypermnesia for details about a violent event. Journal of Applied Psychology, 73, 371–377.  https://doi.org/10.1037/0021-9010.73.3.371.CrossRefGoogle Scholar
  109. Spielberger, C. D. (1979). State-trait anxiety inventory for children. Palo Alto: Consulting Psychologists Press.Google Scholar
  110. Spring, T., Saltzstein, H. D., & Peach, R. (2013). Children’s eyewitness identification as implicit moral decision-making. Applied Cognitive Psychology, 27, 139–149.  https://doi.org/10.1002/acp.2871.CrossRefGoogle Scholar
  111. State v. Henderson, 208 N.J. 208 (2011).Google Scholar
  112. State v. Mitchell, 294 Kan. 469, 275 P.3d 905 (2012).Google Scholar
  113. Towse, J., & Hitch, G. J. (1995). Is there a relationship between task demand and storage space in tests of working memory capacity? The Quarterly Journal of Experimental Psychology A: Human Experimental Psychology, 48a, 108–124.  https://doi.org/10.1080/14640749508401379.CrossRefGoogle Scholar
  114. Ullman, S. E. (2007). Relationship to perpetrator, disclosure, social reactions, and PTSD symptoms in child sexual abuse survivors. Journal of child sexual abuse, 16, 19–36.CrossRefGoogle Scholar
  115. Unsworth, N., Fukuda, K., Awh, E., & Vogel, E. K. (2014). Working memory and fluid intelligence: Capacity, attention control, and secondary memory retrieval. Cognitive Psychology, 71, 1–26.  https://doi.org/10.1016/j.cogpsych.2014.01.003.CrossRefGoogle Scholar
  116. Vygotsky, L. (1934/1978). Mind in society: The development of higher psychological processes. Cambridge: Harvard University Press.Google Scholar
  117. Wells, G. L., Kovera, M. B., Douglass, A. B., Brewer, N., Meissner, C., & Wixted, J. (in preparation). Policy and procedure recommendations for the collection and preservation of eyewitness identification evidence. American Psychology-Law Society (Draft White Paper). Retrieved from http://apls.wildapricot.org/resources/Documents/Feb42019EWwhitepaper.pdf
  118. Whitebread, D. (1999). Interactions between children’s metacognitive abilities, working memory capacity, strategies and performance during problem-solving. European Journal of Psychology of Education, 14, 489–507.  https://doi.org/10.1007/BF03172975.CrossRefGoogle Scholar
  119. Wilcock, R., Crane, L., Hobson, Z., Nash, G., Kirke-Smith, M., & Henry, L. A. (2018). Supporting child witnesses during identification lineups: Exploring the effectiveness of registered intermediaries. Applied Cognitive Psychology, 32, 367–375.  https://doi.org/10.1002/acp.3412.CrossRefGoogle Scholar
  120. Wixted, J. T., & Wells, G. L. (2017). The relationship between eyewitness confidence and identification accuracy: A new synthesis. Psychological Science in the Public Interest, 18, 10–65.  https://doi.org/10.1177/1529100616686966.CrossRefGoogle Scholar
  121. Wolchik, S. A., Sandler, I. N., & Braver, S. L. (1990). Social support: Its assessment and relation to children's adjustment. In N. Eisenberg (Ed.), Contemporary topics in developmental psychology (pp. 319–349). New York: Wiley.Google Scholar
  122. Zajac, R., & Karageorge, A. (2009). The wildcard: A simple technique for improving children’s target-absent lineup performance. Applied Cognitive Psychology, 23, 358–368.  https://doi.org/10.1002/acp.1511.CrossRefGoogle Scholar
  123. Zelkowitz, P. (1989). Parents and children as informants concerning children’s social networks. In D. Belle (Ed.), Children’s social networks and social supports (pp. 221–237). New York: Wiley.Google Scholar

Copyright information

© Springer Nature Switzerland AG 2019

Authors and Affiliations

  1. 1.Department of Psychology (MC 6502)Southern Illinois UniversityCarbondaleUSA
  2. 2.University of Illinois at ChicagoChicagoUSA
  3. 3.NORC at the University of ChicagoChicagoUSA
  4. 4.Claremont Graduate UniversityClaremontUSA

Personalised recommendations