Developmental Challenges Underlying the Impeachment of Child Witnesses With Prior Inconsistent Statements

Abstract

One way attorneys challenge witnesses’ credibility is through impeachment by prior inconsistent statements. We examined the testimony of 6- to 12-year-old child witnesses in 120 sexual abuse prosecutions and identified 96 cases in which attorneys challenged children with prior testimony that was allegedly inconsistent with the child’s current statements. In 11% of the cases, the statements were not in fact inconsistent. In 94% of the remaining cases, we identified one or more of 12 developmental difficulties with the question-types or topics potentially leading to the inconsistencies. This included suggestive tag questions in 45% of the cases, which always co-occurred with other question-type or topic difficulties. The most common problems (appearing in more than 20% of cases) included questions about time and number, questions requiring the child to distinguish among individual incidents, sexual terminology, and questions using negative polarity items such as “any” and “ever.”

Keywords

child abuse child maltreatment court testimony interviewing children childhood sexual abuse

Challenges to the credibility of a witness in the law are known as impeachment. A standard form of impeachment is confronting the witness with prior inconsistent statements (Mueller et al., 2018). Criminal defendants have a constitutional right to cross-examine witnesses (Davis v. Alaska, 1974), and this includes impeaching witnesses with prior inconsistent statements (Chambers v. Mississippi, 1973; Federal Rules of Evidence 613). Surveys with jurors have found that inconsistencies in children’s reports between their pre-trial statements and their testimony during trial are a major factor in assessing their credibility (Myers et al., 1999).

Parties may confront their own witnesses with prior inconsistent statements (Federal Rules of Evidence 607), often in the hope that the witness will modify or correct their testimony to conform to their prior statements, or because the inconsistencies are themselves admissible as evidence (e.g., Federal Rules of Evidence 801(d)(1)(A)). In these cases, the party offering the prior inconsistent statement argues that the prior statement is more likely to be true than trial testimony.

In child sexual abuse prosecutions, and in child witness cases more generally, there are many opportunities for inconsistencies. Child victims are likely to be questioned multiple times, including by social workers, the police, medical personnel, and of course attorneys (Malloy et al., 2007). Children’s prior disclosures can corroborate their report, but can also reveal evidence of influence and suggestiveness.

Prior inconsistencies can occur at several points. They can occur prior to trial, in both formal and informal contexts. They can also occur during a preliminary hearing, at which children are often asked to testify (Cross & Whitcomb, 2017). Preliminary hearings are a common means by which prosecutors justify holding the defendant to answer (the other option is to seek an indictment by a grand jury; Cal. Const. art. I, Section 14; Cal. Penal Code Sections 737–738). Prior inconsistencies can also occur during trial, such as when a child’s testimony in direct examination is inconsistent with their testimony during cross-examination.

Questions about potentially inconsistent prior statements are common. Examining 72 6- to 16-year-olds testifying about sexual abuse in criminal trials in Los Angeles County, Stolzenberg and Lyon, 2014 found that 100% of children were asked about conversations with disclosure recipients, with the average child asked about five different people with whom they had spoken. Examining transcripts of 56 5- to 17-year-old children in Scotland, Andrews and Lamb (2019) found that 40% of defense attorney’s questions were about the child’s prior statements, including prior testimony (p. 1017; tbl 2).

Children may be inconsistent because they are dishonest or mistaken, which supports doubts about their credibility. However, inconsistencies may also be due to question-types or topics that are difficult for children because of their developmental limitations. In this study, we examined cases in which child witnesses were specifically asked about potentially inconsistent prior testimony, either at a preliminary hearing or at trial, which enabled us to examine the precise language in both attorneys’ questions and children’s answers and identify potential developmental challenges. (Prior inconsistencies can be introduced through witnesses who spoke to the child outside of the courtroom, but those witnesses may incorrectly recall children’s statements). In what follows we first discuss prior research, seeking to identify factors that lead to inconsistencies in children’s testimony. Taking a novel approach, we lay out a categorization scheme that identifies question-types and topics that research has found are particularly challenging for younger children and then examine whether these were present in children’s inconsistencies.

Research Identifying Inconsistencies in Children’s Testimony

Beginning with Brennan and Brennan (1988), a number of studies have examined attorney questioning of child witnesses. The research has uniformly criticized the questions as age-inappropriate, predominantly due to their complexity and the difficulty of the topics to which they refer (e.g., Powell et al., 2016; Zajac et al., 2003). A subset of the work on attorney questioning has also looked at children’s responses to attorneys’ questions. Although the work has consistently found that defense attorneys are more likely to elicit inconsistencies than prosecutors, it has failed to find age differences. Zajac et al. (2003) examined 21 transcripts of 5- to 13-year-olds testifying in New Zealand and found 76% of children changed their story during cross-examination. All inconsistencies were elicited by defense attorneys, and there were no age differences. Zajac & Cannan (2009) compared 15 transcripts of 5-to-12-year olds’ testimony to the testimony of 15 adults and found 93% of children changed their story at least once. However, so did 100% of the adult witnesses. Defense attorneys elicited more inconsistencies, but again there were no age differences. Andrews et al. (2015) found almost all children contradicted themselves at least once in both a sample of 120 6- to 12-year-old children testifying in Los Angeles County (95%), with attorney differences but no age differences, and in the aforementioned Scottish sample (100%, Andrews & Lamb, 2016), again with attorney differences but no evidence that younger witnesses were more inconsistent. Andrews et al. (2015) found that 91% of children contradicted themselves at least once in response to defense attorneys, but so did 86% when questioned by the prosecution.

It is unclear from the research whether the inconsistencies are undermining children’s credibility, given the lack of age differences, and the fact that inconsistencies are often elicited by prosecutors. As noted above, prosecutors may wish to elicit inconsistencies to set the record straight, that is, to elicit testimony from the child that is consistent with the initial charges. Moreover, reporting whether witnesses uttered any inconsistencies may exaggerate the problem; the percentage of responses that are inconsistent is necessarily much lower. In Andrews et al. (2015), for example, 6% of children’s responses to defense questions were inconsistent, 2% of their responses to prosecutors (the corresponding percentages in Zajac and Cannan (2009) were 4% and 2%).

Research Examining Correlates of Inconsistency

Another approach to understanding the source and significance of inconsistencies is to identify question-types and topics that are correlated with inconsistencies. Although the results are mixed, there is some evidence that age-inappropriate questions and suggestive questions are more likely to lead to inconsistency. Davies and Noon (1991) asked court observers to evaluate the age appropriateness of questions and correlated their judgments with inconsistencies. They found that when prosecutors asked age-inappropriate questions, inconsistencies were more likely, but they found no effect for defense attorneys.

In the New Zealand samples, Zajac et al. (2003) found 95% of inconsistencies were elicited by credibility challenges, suggestive questions, or both, but did not separately analyze the two types of questions. The authors did not define credibility challenging questions but gave an example: “I don’t think that’s what really happened.” It is unlikely that this question is common. It is objectionable in the United States as an impermissible expression of an opinion by the attorney (United States v. Young, 1985). A similar rule against attorneys expressing their opinions was enacted in New Zealand in 2008 (Mahoney, 2015). It is therefore unclear what sort of currently allowable questions led to inconsistencies.

Zajac and Cannan (2009) adopted a broader definition of credibility challenging questions, which included questions suggesting uncertainty, deception, or lacking memory, but also included references to delayed disclosure and a positive relationship with the accused, a quite different approach that relies on juror misconceptions about sexual victimization (Stolzenberg & Lyon, 2014). The authors found that both children and adults were influenced to change their story by credibility challenging questions. Surprisingly, only adults changed their story in response to suggestive questions.

In contrast to Zajac and colleagues’ (2003, 2009) inconsistent findings with respect to the effect of suggestive questions on children’s consistency, Andrews and colleagues’ (2015) and Andrews and Lamb’s (2016, 2021) larger samples, which examined both Scotland and Los Angeles, both found that suggestive questions increased the likelihood of self-contradictions, particularly when they were repeated. It thus seems fair to conclude that suggestive questions lead to inconsistencies.

Beyond suggestive questions, however, research has failed to find other predictors. Andrews and Lamb (2017) found no relation between question complexity and inconsistency. Andrews and Lamb (2019) failed to find that questions about peripheral details increased the likelihood of inconsistency. Rather, they were surprised to find that questions about central details did so. With respect to other possible correlates, the authors warned that “[t]he current study did not have enough statistical power to investigate differences in the rates of self-contradiction and uncertainty when responding to questions varying in topic” (p. 1023).

The Current Study

A novel approach is to start by identifying question-types and topics known to present difficulties for younger children based on experimental and observational work in developmental and applied psychology, and then examine how often inconsistencies include those difficulties. Table 1 identifies 12 question-types and topics we identified for this study. For each category, the table cites research and research reviews that have explicated the difficulties they present for children.

Table 1.

Categories and Definitions of Difficulties Potentially Leading to Inconsistency

Difficulty	Definition	Relevant developmental and applied literature
1. Tag questions	Leading questions in which a declarative statement is followed by a yes/no tag (e.g., “You were scared, weren’t you?).	Suggestiveness of tag questions (Behzadnia & Mehrani, 2018; Cassel et al., 1996; Greenstock & Pipe, 1996; Krackow & Lynn, 2003).
2. Time and number	Temporal location (when events occurred using conventional temporal information, such as days, months, years, or age); temporal distance (time between events or since present), duration, frequency or numerosity.	Acquisition of conventional temporal terms late in grade school (Friedman, 1992); difficulty in estimating temporal location and distance (Friedman & Lyon, 2005), duration (Block et al., 1999), and numerosity (Wandrey et al., 2012).
3. Distinguishing individual incidents	Recalling details from individual episodes of repeated events.	Difficulty in individuating repeated events (Woiwod et al., 2019).
4. Sexual terminology	Unclear sexual terminology, including possible differences in adult/child definitions of “touch.”	Difficulty in naming sexual parts (Szojka et al., 2023b); underextended understanding of “touch” (Sullivan et al., 2022a).
5. Negative polarity items	Questions about abuse using the terms any or ever.	“No” bias in responding to negative polarity items (Henderson et al., 2023; Szojka et al., 2023a).
6. Forcing a guess with option-posing questions	Asking an option-posing question after the child has given a don’t know response.	Tendency to guess in response to option-posing and closed-ended questions (McWilliams et al., 2021); failure to respond “neither” or “both” to forced-choice questions (Rocha et al., 2013).
7. Conceptual confusion	Difficulty in testifying about concepts and the distinctions among them.	Difficulty in understanding intending an action vs. intending the outcome (Rakoczy et al., 2015), assisting recall vs. coaching (Wylie et al., 2022) knowing vs. inferring (Pillow et al., 2000).
8. Anaphora and ellipsis	Questions containing anaphora or ellipsis. In anaphora, one uses a word or phrase as a substitute for a previously mentioned entity. In ellipsis, one elides a word or words.	Difficulty in interpreting and disambiguating anaphora (Ruiz-Earle et al., 2024; Sullivan & Stolzenberg, in press) and ellipsis (Sullivan et al., 2022b).
9. Language confusion	Confusing one word or phrase with another.	Difficulty in using spatial prepositions (Stolzenberg et al., 2017); before/after (Szojka et al., 2026); and right/left (Dellatolas et al., 1998; Rigal, 1996).
10. Referential ambiguity	When words or phrases in the attorney’s questions or child’s responses could refer to multiple properties or things.	Difficulty in responding to referentially ambiguous questions (Evans et al., 2017).
11. Grammatical confusion	Complex questions, including compound questions, questions with multiple clauses, and negative questions.	Difficulty in answering complex questions (Perry et al., 1995); ambiguity in negative questions (Wylie et al., 2026).
12. Reluctance	Child explains that they did not want to disclose the information.	Reluctance in disclosing sexual abuse (Lyon et al., 2026).

We examined trial transcripts of 6- to 12-year-olds testifying in child sexual abuse criminal trials and identified instances in which attorneys impeached child witnesses by confronting them with claimed inconsistencies. Reviewing the trial transcript and/or the preliminary hearing transcript, we determined whether the statements were indeed inconsistent, and if so, coded for developmentally challenging question-types and topics. We calculated descriptive statistics and qualitatively assessed individual cases. We present case studies in the results section that illustrate the difficulties, with a focus on two acquittals, enabling the reader to understand the cumulative pressures that the questioning placed on the children and the seriousness of the inconsistencies.

Method

Sample

We examined 120 criminal court trial testimonies of 6- to 12-year-old children (M = 9.58 years old, SD = 1.81 years) in 96 cases with felony sexual abuse charges filed in Los Angeles County from 1997 to 2001. Eighty-two percent of the victims were female. In 28% of the cases, the defendant was a father or father-figure (e.g. stepfather, foster father). In 66% of the trials, multiple acts of abuse were alleged and 44% of the trials alleged penetration. Overall, 74% of the trials resulted in a conviction.

Coding and Qualitative Content Analysis

First, we machine coded for question-answer pairs containing prior conversation words (e.g., said, testified, told) in trial testimony. Two research assistants reviewed those pairs for instances in which an attorney questioned the child about a prior inconsistency. A random 20% of the cases were selected and independently coded to assess inter-rater reliability. We used Cohen’s Kappa to measure agreement between coders; a Kappa of .92 was observed. The research team then resolved all discrepancies to ensure complete agreement.

Second, we reviewed the preliminary hearing testimony and trial testimony to identify relevant testimony related to the queried inconsistency, creating a comprehensive dataset of each inconsistency. We qualitatively examined the testimony and constructed a coding guide based on question-types and topics identified as presenting developmental difficulties in prior research (Table 1). Two research assistants then independently assessed the relevant testimony to categorize all potential reasons for the inconsistency. The categories were not mutually exclusive; rather, a case of impeachment could fall into more than one category. Because of their infrequency, some categories had high prevalence indices, making Cohen’s Kappa difficult to interpret. Therefore, we measured agreement between coders through Prevalence Adjusted, Bias Adjusted Kappa (PABAK; Brennan & Silman, 1992; Byrt et al., 1993). For all variables, a PABAK of .83 or higher was observed. The research assistants resolved all discrepancies after reaching reliability on all variables.

For discussion of individual cases, quotations from children were edited for clarity, and pseudonyms were created that matched the child’s gender and probable ethnicity (Namsor, n.d). In addition to providing the age of the child, the child’s subject number was noted, corresponding to our archive of court transcripts.

Results

Across the 120 testimonies examined, we identified a total of 96 instances in which attorneys confronted children about inconsistent statements. In 29% of the instances (n = 28) the inconsistency involved preliminary hearing testimony, and the remainder involved inconsistencies within trial testimony. In the cases where the inconsistency occurred between the preliminary hearing and trial, the average number of days between the preliminary hearing and the child’s testimony was 174 days (SD = 104, range = 55–545). This was comparable to the delays among all cases with preliminary hearings (M = 174 days, SD = 93.6, range = 14–545). The 96 instances of impeachment occurred across 43 cases (with 45 child witness testimonies). Of these 43 cases, 19% (n = 8) resulted in an acquittal, while the remaining 81% (n = 35) resulted in a conviction.

In 11% of the impeachment instances (n = 11), children were not in fact inconsistent but attorneys misstated or misrepresented their testimony. In two of the 11 instances, the prosecutor objected, and both objections were successful – in one instance, the judge sustained the objection, and in the other instance, the defense withdrew their question. In the remaining nine instances, no objections were raised. In three of these, the child stated they did not recall making the allegedly inconsistent statement, but no additional questioning, objections, or clarification followed.

In the remaining 89% of instances (n = 85), children’s statements were in fact inconsistent. Of those 85 instances, we identified one or more possible explanations for the inconsistency in 94% of instances (n = 80), based on question-type or topic (Table 2). All inconsistencies coded as including tag questions also fell into another category (or categories). For the non-tag categories, we calculated the percentage of instances in which tag questions were also used (Table 2).

Table 2.

Frequency of Potential Reasons for Inconsistency and Overlap With Tag Questions

Potential reason for inconsistency	% impeachment instances (N = 80)	% also containing Tag questions
1. Tag questions	45%
2. Time and number	30%	33%
3. Distinguishing individual incidents	24%	58%
4. Sexual terminology	23%	44%
5. Negative polarity items	21%	75%
6. Forcing a guess with option-posing questions	18%	36%
7. Conceptual confusion	16%	50%
8. Anaphora and ellipsis	11%	44%
9. Language confusion	10%	38%
10. Referential ambiguity	9%	14%
11. Grammatical confusion	8%	43%
12. Reluctance	5%	23%

In what follows, we provide examples of the different categories in individual cases. We first discuss two acquittals in which a large number of the potential causes of inconsistency were identified, and then provide individual examples from additional categories.

California v. Curtis H. Shae (subject no. 368) was a 10-year-old girl who testified to her stepfather, Curtis H., sexually abusing her on multiple occasions, usually while her mother was at work. Shae testified that one night Curtis H. came home drunk and made her sleep in bed with him; Shae’s mother walked in on him naked in bed with Shae. Shae’s mother testified to this incident, which prompted Shae to disclose previous acts of abuse. Shae alleged Curtis H. asked her if she wanted to see a magic trick and proceeded to pull down his pants and expose himself to her and masturbate in front of her. After the first incident, he warned her not to tell her mother or he would get in trouble and go to jail. Shae also described multiple incidents where Curtis H. pulled down her pants and rubbed her vagina with his hand and penis.

The jury acquitted. Several of the inconsistencies involved time and number. Shae was asked to recall what days of the week the abuse occurred, distinguishing between what are traditionally school and workdays (Monday to Friday) and the weekend (Saturday and Sunday). During the preliminary hearing, Shae reported that the abuse usually occurred “[a]fter school and when I have no school and on Saturday or Sunday when my mom isn’t there.” At trial, the defense asked, “did it happen, those one or two times a week, was it always like on a Monday, Tuesday, Wednesday, Thursday, or Friday?” and Shea responded, “it was on one of those days.” The defense attorney then asked a series of tag questions: “Q. One of those days. Monday through Friday, right? A. Yeah. Q. And it didn't ever happen, then, on a Saturday or Sunday? A. Yeah -- no. Q It did not; is that right? A. Yeah. Q Okay. So never on Saturday or Sunday, right? A. Yeah.” In addition to asking tag questions, the defense attorney used anaphora and ellipsis, either referring to the touching as “it” or omitting reference to the touching altogether (i.e., “one of those days”). This may have made it difficult for Shea to track what the defense attorney was referring to (Ruiz-Earle et al., 2024; Sullivan et al., 2022; Sullivan & Stolzenberg, in press).

Furthermore, the key question was grammatically confusing, because it was a negative declarative: “it didn’t ever happen…on a Saturday or Sunday?” This may have led to Shea’s uncertainty about whether to answer yes or no (“Yeah--no”). That is, if she wished to respond that it had occurred on a weekend, it was unclear whether she should respond “no, it did happen on a weekend” or “yes, it did happen on a weekend” (Wylie et al., 2026).

The defense attorney then read the inconsistent preliminary hearing testimony into the record and confronted Shea: “So what you just told me a couple minutes ago was wrong, that when you said it never happened on Saturday or Sunday?” and Shae responded, “usually, my Mom is, my Mom isn't home.” The defense attorney highlighted the inconsistency in his final argument: “So which choice of this do you believe? What days of the week did it happen on? Monday through Friday when her Mom wasn't home, that's what she testified to in court. At the preliminary hearing, she testified that it happened on Saturdays and Sundays as well. So – was it Monday through Friday, or weekdays. We've got two stories for everything, just depending upon when she's asked.”

Another difficulty with time and number occurred when Shae was asked to approximate what time her mom came home the first time that the abuse occurred, which also required distinguishing individual incidents. At the preliminary hearing, the defense asked Shae a series of questions about the first incident, including what time her mother came home, and Shae estimated “about one o'clock in her break.” At trial the prosecutor asked Shae “around what time” her mother came home and Shae estimated “3 p.m. or 4.” In cross-examination, the defense attorney confronted Shae with the inconsistency, and Shae offered an estimate about when her mother sometimes came home (Q: Do you remember testifying that other time and you said your Mom came home about 1 that day? A: Yeah. Q: So what time was it? Was it 1, 3? A: What was it, that question? Q: On the very first day when this happened, did your Mom come home at 1 or 3? A: It was, sometimes it was 2 or 3). The defense attorney admonished her that “I don’t want to know about sometimes. I want to know about the day this happened,” and then asked again, “what time did your Mom come home on the first day this happened?” and Shae answered “2.” The defense attorney emphasized the inconsistency (Q: Okay. So 2, not 3; right? A: Yeah. Q: And not 1? A: No). In his closing argument, the defense attorney cast the inconsistency, coupled with other testimony about uncertainty whether the mother was home when the abuse occurred: “So in the space of two minutes, she says her Mom was there when they got home, her Mom wasn't there when they got home. And just made up an answer. About when it happened the first and second time.”

Several times, Shae attempted to testify about what usually or sometimes happened, but both attorneys insisted on specifics about individual events. In their final argument, the prosecutor noted that if you did something every day, you might not recall exactly how it happened on a particular day. However, the prosecutor had elicited the timing inconsistency at trial by asking Shae when her mother came home the first time. The case illustrated how the defense attorney elicited inconsistencies through a combination of tag questions and temporal questions, requiring Shae to recall specific events.

California v. Larry R. Larry R. was a family friend of 12-year-old Bella (subject no. 369) and her 6-year-old brother Terrence (subject no. 370). Larry R. had worked with the children’s father and was a godfather to one of their younger children, who they named after him. Larry R. had stepped in to help with the children when the mother became sick and had lent the family money. Larry R. would often drive Bella and several of her brothers to lunch, to shop, and to his apartment to swim. The alleged abuse was discovered when Bella told her brothers, who then told their mother (her stepmother). Bella testified to five charged acts; two occasions when Larry R. allegedly put her hand on his penis (in the pool and in the car), two touches in the kitchen (her breasts and her vagina), and one occasion of touching her leg in the car. Terrence testified to two charged acts. Terrence stated that in Larry R.’s bedroom, Larry had laid Terrence across his knees, pulled down Terrence’s pants, licked his finger, and inserted it between Terrence’s buttocks. Terrence recalled that it hurt because of Larry R.’s fingernail. Terrence also testified to an occasion by the pool in which Larry R. had Terrence remove his swim trunks and sit naked on Larry R.’s lap.

Bella also testified to a number of uncharged acts, which were offered to prove sexual intent (St. George et al., 2024), including other touches in the car, a dream Larry R. told her about in which she was breastfeeding a baby he sired, an incident in which he had told her to go upstairs and take off her bathing suit, and an overnight stay in which he had asked her to sleep with him. In the preliminary hearing, she also testified that he had touched the top of her buttocks while she sat on his lap at the computer, and attempted to tongue kiss her on another occasion, but at trial she denied these touches, describing them as innocuous. At both the preliminary hearing and trial, the prosecutor transitioned to new events without asking about specific places or acts, but instead asked questions such as: “Can you think of another time and tell us about that?” Probably for that reason, Bella mentioned some incidents for the first time at trial, including touching of her leg in Larry R.’s living room and at Arby’s, and attempting to enter a dressing room at Macy’s while she was changing and a bathroom while she was showering.

The jury acquitted. In their final argument, the defense attorney cited the jury instruction that in assessing the believability of witnesses, jurors should consider a “statement previously made by a witness that is either consistent with or inconsistent with the testimony here at trial” (CALJIC 2.20, 2005). The defense then focused on several inconsistencies and suggested they were attributable to talking to the prosecutor between cross-examination and redirect.

The first inconsistency concerned Bella’s difficulty in distinguishing individual events, specifically whether one of the touching incidents occurred on the way to Arby’s or in Arby’s. At the preliminary hearing, Bella testified that “[w]henever I was in the car with my brothers and they weren’t paying attention he used to rub my leg.” She specifically mentioned Larry R. touching her leg on the way to swim, to Macy’s, and to Arby’s. At trial she referred to touching in the car on the way to lunch, but when the prosecutor asked her where, she responded “we went to--I don’t remember, we went many places to lunch.”

At trial, Bella testified to touching in Arby’s, describing how she sat on one side of the booth with Larry R. while three of her brothers sat on the other side, and how Larry drank soda with one hand while the other hand was under the table rubbing her leg until she “scooted” over to avoid him. In cross-examination, the defense attorney challenged her by asking her, “that’s not the way you described it at Alhambra [the site of the preliminary hearing], is that right?” and read the portion of her preliminary testimony when she responded, “Arby’s” when the prosecutor asked, “where were you going?” Pressed by the defense’s question, “what you told us at Alhambra wasn’t correct?” Bella responded, “I don’t remember what I told you.” The defense attorney then asked a forced choice question: “Did it happen on the car on the way to Arby’s, or did it happen in Arby’s” and she responded, “in Arby’s.” The forced-choice question may have suppressed a “both” response (Rocha et al., 2013). The attorney followed up with an elliptical negative question: “Not on the way?” and she responded, “yeah.”

In redirect, the prosecutor referred to her preliminary hearing testimony and Bella reaffirmed that he had also touched her on the way to Arby’s. In final argument, the defense attorney argued that, “she told us that nothing happened on the way” to Arby’s but, “after a discussion with the prosecutor, she testified, ‘oh yeah, that did happen on the way there.’”

Another apparent inconsistency concerned the contents of Larry R.’s dream. In the preliminary hearing, Bella initially described the part that, “he had a baby by me.” The prosecutor asked, “did he tell you more about the dream?” She added: “Yes. He told me that I was breastfeeding the baby, and I told him to leave me alone. Then I went back to play with my brothers.” At trial, Bella testified: “At the park I was playing soccer, soccer with my brothers, and he told me to come to him. And I walked over to him, and he told me that he dreamed that he was gonna have a baby with me. And I told him to leave me alone.” Because Bella had stopped narrating the dream at this point, she understandably answered “no”to the prosecutor’s negative polarity item question: “Did he say anything else?” A negative polarity item suggests a “no” response (Henderson et al., 2023; Szojka et al., 2023a). In redirect the next day, the prosecutor asked more pointedly, “do you remember saying [at preliminary hearing] what Larry said he saw you doing in the dream” and Bella recalled the breastfeeding. In their argument, the defense attorney recalled a more leading exchange: “On Friday, however, the prosecutor reminded her, ‘was there anything about breastfeeding?’ And on Friday, suddenly she responded, ‘Oh, yes. And he told me that he saw me breast feeding our baby.’”

At trial, Bella spontaneously mentioned the pool incident, initially stating that Larry R. “tried to make me touch his penis” and then explaining that “[h]e just took [my hand] very gently and put it on.” But with respect to touching in the car, after answering “no” to the negative polarity item questions “did he do anything else to you inside the car?” and “did you touch Larry anywhere on his body?”, Bella ultimately acknowledged a time when she “touched his penis,” describing how “he took my hand, and he put my hand on his penis, and he squeezed my hand again. And I took my hand off his penis really fast.” In the defense attorney’s words, Bella “had to be assisted,” it took “some prodding,” but “suddenly she remembered” what had occurred.

In what the defense attorney in argument referred to as a “dramatic change in testimony,” another apparent inconsistency concerned language confusion over the invasiveness of the genital touch in the kitchen. For legal purposes, genital touching over the clothes with sexual intent is still a felony (Cal. Penal Code Section 288), so any inconsistency was an attack on Bella’s credibility rather than evidence against illegality. In her direct testimony, Bella testified that the touching occurred when she was wearing a shirt, pants, and swimming suit bottoms, and when the prosecutor asked, “were his fingers actually touching the pants, the bathing suit, or something else,” Bella responded, “my bathing suit.” The touching thus may have occurred under the pants but over the bathing suit. In cross-examination, Bella agreed with the defense attorney’s yes-no question that the touching occurred “on the outside of your clothes.” In redirect, she agreed with the prosecutor’s yes-no questions that she remembered testifying at preliminary hearing “that Larry touched you under your clothing” and that he had in fact done so. When touching occurs under one piece of clothing but over another, it is both over and under the clothes, but the yes-no questions did not facilitate this response. In his final argument, the defense attorney mischaracterized Bella’s testimony: “You may remember that the prosecutor asked her, ‘was that over or under the clothes?’ And she said, ‘that was over my pants and over the swimming suit bottom that I had on.’”

Another possible reason for the apparent inconsistency regarding the invasiveness of touching in the kitchen was the fact that Bella appeared to recall two different occasions when the defendant touched her genitalia, and exhibited difficulty in distinguishing individual incidents. In the preliminary hearing, she testified that Larry was in the living room with her brothers and then came into the kitchen and touched her “under my underwear.” When asked what Larry R. had said to her in the kitchen, she testified that he told her to “stop twitching,” but then she corrected herself: “Actually this is not the time when he--my brothers weren’t in there. They went swimming.” The prosecutor failed to clarify whether she was now talking about a different episode, and whether the invasiveness of the touching was also different. At trial, she testified that Larry came into the kitchen and touched her on top of her bathing suit while the brothers were in the pool, and again recalled him telling her not to move. Whereas Bella only described the bathing suit episode at trial, she may have been describing two episodes in the preliminary hearing, skin to skin touching under her underwear when her brothers were in the living room and touching over her bathing suit when her brothers were in the pool.

With respect to Terrence, the defense attorney focused on an apparent inconsistency regarding sexual terminology, specifically, whether there was anal penetration. For legal purposes, anal penetration is not necessary for felony sexual abuse (Cal. Penal Code Section 288), so as with his sister’s testimony about clothing, any inconsistency in Terrence’s testimony solely affected his credibility. In direct examination, Terrence testified that when he was in Larry’s bedroom, Larry told Terrence to “lay down on his lap,” whereupon Larry “grabbed my pants, then he put them down.” He then testified, “then he put his finger in my…” but he was initially unable to continue. The prosecutor asked him about unrelated subjects (largely about testimonial competency), and then asked, “what did Papo do with his finger?” Terrence responded, “he went up my butt,” adding that Larry had licked his finger, and that Terrence felt a “hurt spot” because “he has a sharp nail;” “It was hurting me bad.”

Difficulties arose in distinguishing among Terrence’s buttocks, anus, and the toilet. In the preliminary hearing, the prosecutor asked Terrence, “do you know the hole where you go No. 2,” and he responded, “yeah.” Then the prosecutor asked, “did he put it in that hole,” and Terrence responded “no.” Terrence may not have been thinking of his anus. Indeed, when the prosecutor first asked him at trial, “where does No. 2 come from,” he responded, “the bathroom,” suggesting another hole where one goes No. 2.

Continuing at trial, when the prosecutor attempted to clarify by asking, “do you know the place in your body where No. 2 comes from,” Terrence responded, “behind,” and then affirmed that this was where “Papo put his finger.” Here, whether he distinguished between his buttocks and his anus was unclear. In cross-examination, the defense attorney paraphrased the prosecutor as simply asking, “whether he put his finger where No. 2 comes from” and then asked Terrence a series of elliptical and anaphoric questions: “Do you remember that?”/“Yes”; “Did he or didn’t he?”/“Huh?”; “Did he do that or not?”/“Yeah”; “Yes, he did?”/“Yes”; “Did he put it inside that place?”/“Yes.” The defense attorney then confronted Terrence with his preliminary hearing testimony, asking, “were you telling the truth then,” which Terrence affirmed. “So now you think maybe he put it in that hole, is that it?”/“Yes”…“And now you think that he put it in that hole after all?”/“Yes.” In their argument, the defense referred to the statements as “totally inconsistent:” and argued that Terrence “clearly said ‘no’ to that question” at the preliminary hearing, “whereas here he said ‘yes’ to that question. And when I asked him about that, he said it was true then.”

Additional examples: Sexual terminology. The difficulty in determining whether Terrence experienced anal penetration is just one example of difficulties in sexual terminology. Younger children are likely to deny that touching has occurred when the suspect touches the child with something other than their hand (Sullivan et al., 2022a). Kayla (subject no. 59) was 7 years old when she testified to sexual abuse by her uncle (Jacobo, conviction). She testified to several episodes in which she described how he “rubbed her up and down,” rubbing his stomach against her vaginal area. However, she repeatedly denied that he had “touched” her, except when describing one episode where he had also touched her with his finger.

Forcing a Guess. Eleven-year-old Reagan (subject no. 259) testified to repeated abuse by a family friend (Durado, conviction). In the preliminary hearing, she initially refused to respond to a question about number: “how many times do you think you saw” the defendant before abuse occurred. Even when the defense attorney persisted and asked her if she saw the defendant “over many different times,” she asked, “how much is many?” She finally relented when the defense attorney said, “You tell me how many times you think you saw him,” answering “probably ten…probably more than ten.” At trial, she testified that it was “probably more than five,” whereupon the defense attorney impeached her with her preliminary hearing testimony.

Conceptual Confusion. Children sometimes denied thinking or doing things they found undesirable, but which were side effects of their actions or a necessary means to their goals. In California v. David D. (acquittal), four children testified that they had been touched by a schoolteacher: 10-year-old Chloe (subject no. 325), 10-year-old Kali (subject no. 328), 9-year-old Sadie (subject no. 327), and 12-year-old Destiny (subject no. 326). In the preliminary hearing, 9-year-old Sadie initially acknowledged that when she went to the principal’s office to report the teacher’s behavior, she thought she “was getting Mr. D. in trouble,” though she did not want to. Later, when the defense attorney asked, “when you went to Mr. M. with Kali you thought you were getting him -- Mr. D. in trouble, right?” She initially answered, “no,” and then conceded. Sadie thus had difficulty acknowledging thoughts about consequences of her report that she did not intend.

Similarly, at trial 10-year-old Chloe acknowledged on direct examination that she hugged Mr. D. “when [she] would go in to borrow the ropes,” because Mr. D. oversaw the classroom where the jump ropes were stored. However, when the defense attorney asked her in the preliminary hearing whether she had “ever run up to Mr. D. and [given] him a hug,” she answered, “no.” When the defense attorney pressed and asked again, “did you ever run into his classroom and just run up to him and give him a hug to say hello” she answered, “I just went into his classroom to get ropes.” At trial, the defense attorney impeached her with her preliminary hearing testimony, and she could not explain the inconsistency. In final argument, the defense attorney argued her behavior was implausible (noting that she had prior experience with abuse): “Every day, 3 and 4 times a day…There was Chloe coming in, ‘Hi, Mr. D.’ Give him a hug.” Chloe found it difficult to acknowledge an action that she performed only reluctantly to achieve another goal: hugging Mr. D. because he controlled the jump ropes. In both testimonies, Sadie and Chloe at times denied thoughts or actions that were not their ultimate purpose.

Reluctance. Reluctance was rarely coded (Table 2, 5%) as a potential reason for inconsistency. Many of Bella and Terrence’s difficulties may have been due in part to their reluctance to describe sexual acts, but they failed to explicitly express reluctance. An exception is Rivera (conviction), which concerned 12-year-old Camila’s allegations of abuse. Camila partially recanted her allegations in direct examination. Recantations are most common in cases where the child is close to the alleged perpetrator and the mother is unsupportive of the allegations (Malloy et al., 2007). Consistent with this pattern, Camila accused her stepfather of abuse, and her mother was unsupportive. Her case illustrates the many inconsistencies caused by children’s reluctance to disclose. The abuse allegations first came to authorities’ attention when a friend reported to police that Camila had disclosed the abuse to her. When the patrol officer came to the house, Camila was “hysterically in tears,” telling them that her mother had asked her, “did you tell her something” when they first arrived, and that “I don’t want to be taken away from my family. Nothing happened, he didn’t touch me” (non-referenced quotations are from confidential reports in the dependency file, which we were able to obtain in addition to the appellate court opinion).

Alone with the officer, Camila disclosed fondling of her breasts, vaginal area, and buttocks, intercourse and sodomy, oral copulation by the defendant, oral copulation by the child, and fondling of the defendant’s penis. Camila’s report included the allegation that the defendant would force her to swallow his semen. Regarding her disclosure, the defendant had told her, “you better keep your mouth shut and don’t tell your mother” and had given her money to keep silent. At one point during the police interview, Camila “went to [her mother], buried her face in her lap, and asked for forgiveness. Mrs. Rivera did not answer her or look at her” (California v. Riviera, 2001, p. 3). The mother acknowledged to the police that Camila had disclosed abuse to her when Camila was approximately eight years old, but stated the defendant had denied the abuse.

Camila re-disclosed to an emergency response social worker the same day, who noted, “the mother is very dependent on the stepfather to provide for the family.” A day later, Camila told her uncle that she wished to withdraw the allegations. A few days after that, she redisclosed to a nurse practitioner and a police investigator. The defendant was arrested and jailed. A month later, Camila re-disclosed to a dependency investigator; she now stated that the defendant had only put his penis between her legs, though she realleged that he put “ugly stuff on me and my mouth, it made me sick, I would spit it out.” She told the investigator that she just wanted to forget about the abuse. The mother now no longer acknowledged that Camila had ever disclosed, stated that she doubted the allegations, and claimed she had never left Camila alone with the defendant.

Four months later, the criminal case came to trial. Waiting to testify, Camila sat alone in the courtroom hallway, her mother sitting apart with the other two children. The mother was the first to testify, and she denied that Camila had ever told her about the abuse. She also denied telling the police or the district attorney’s office that Camila had disclosed. She even denied the police had told her about the abuse allegations. Camila testified next, and she initially faltered when describing how her stepfather had touched her, testifying, “he was touching my chest and--can I have, please water?” She minimized or denied various sexual acts; for example, she testified the touching had occurred over the clothes. She denied making inconsistent statements to the police officer and nurse practitioner.

The defense attorney objected repeatedly to the prosecutor’s questions. During cross-examination, he reminded Camila that she had recanted to him in the kitchen of her apartment and in his office. He also pointed out inconsistencies between her testimony and prior reports. However, his aggressive approach appeared to backfire, because Camila began to disclose acts she had denied on direct examination, such as whether the defendant had forced her to touch his penis. She now clearly alleged that, “something came out of his penis” when he put it in her mouth. She told the defense attorney, “you don’t have to tell me if I am lying or not because you didn’t suffer anything” (the judge struck part of the response as non-responsive).

Toward the end of cross-examination, Camila said: “O.K. That’s it. I am tired of this. You want the truth?” In re-direct, the prosecutor offered her the opportunity to expand, and among other acts she described how “some white thing came out of his penis, and that went to my mouth. I tried to swallow, but I couldn’t. I throw it up. I was in the, I was throwing up in the restroom.” She explained that she had lied earlier in her testimony “because I wanted all of this [to] be over. Because I wanted to die. And I have to go on in my life.” As their next witnesses, the prosecution called the friend to whom Camila first disclosed, the patrol officer, the nurse practitioner, and the dependency investigator. They were allowed to repeat Camila’s disclosures under various hearsay exceptions.

In their final argument, the defense argued that Camila should not be believed: “A witness who contradicts themselves on the witness stand point after point after point cannot be trusted.” In contrast, the prosecutor emphasized the inconsistencies between the mother’s testimony and her prior statements, and focused on her unsupportiveness as the reason for Camila’s reluctance. The prosecutor argued that the jury’s verdict was the substitute for the mother’s unprotectiveness: “In the jury room and you’re talking to each other, you remember when Camila was telling the police officer what happened when she hears her mother crying out in the hallway. She goes to her mother. Her mother won’t look at her. Her mother won’t speak to her. Her mother will not comfort her. Her mother will not reassure her. You are what matters.” The case illustrates the many difficulties prosecutors face in prosecuting cases where the child’s reluctance leads to inconsistencies in their report.

Discussion

This study examined 96 instances in which attorneys impeached child witnesses with their inconsistent statements in child sexual abuse trials. Because the allegedly inconsistent statements had occurred at preliminary hearings or trial, we were able to examine the context in which the inconsistencies were elicited. In 11 of the cases, we found attorneys misrepresented children’s previous testimony, and opposing attorneys objected in only two of those cases. In 80 of the 85 remaining instances (94%), we identified one or more possible explanations for the inconsistency, based on prior research identifying question-types and topics that present difficulties for children. The most common possible cause was tag questions (45%), and in all of those instances, another possible cause was also identified, the most common questions about time and number (30%), questions requiring children to distinguish among individual incidents (24%), the use of sexual terminology (23%), and questions about abuse using negative polarity items, such as “any” and “ever” (21%). Seven additional potential causes were identified with less frequency. Overt references to reluctance as a potential cause of inconsistent testimony were quite rare (5%). These additional potential causes were often associated with inconsistencies even without the use of tag questions. Case studies including two acquittals highlighted the way in which the different potential causes often interacted, such as difficult question-types being used to ask about time and number or requiring children to distinguish among repeated events. The results suggest that although children’s inconsistencies may be due to untruthful testimony, they may also be attributable to the ways in which children are questioned about abuse, which overlook developmental limitations that lead to errors and miscommunication.

Limitations and Future Directions

As with other studies examining testimony, it is impossible to know with certainty whether the sexual abuse actually occurred. Similarly, whether cases resulted in acquittals or convictions cannot be attributed to the inconsistencies given the myriad case, attorney, jury and judge characteristics that influence verdicts. The cases were drawn from one jurisdiction (Los Angeles County, California) over twenty-five years ago, and of course questioning may be different in other jurisdictions and over time. However, research examining more recent court samples in other jurisdictions has continued to criticize the developmental inappropriateness of attorney questioning (e.g., Powell et al., 2016 [New South Wales, Australia, 2011-2015])

The results highlight the need to jointly consider question-type and topic in understanding the difficulties that children encounter when testifying. For example, experimental work could explore how difficult question-types may be particularly likely to undermine accuracy when children are asked about concepts such as time and number. Moreover, the fact that inconsistencies are not inevitably tied to tag questions highlights the need for research examining inconsistencies in forensic interviews, in which interviewers are taught to avoid tag and other suggestive questions. Ultimately, we may distinguish between inconsistencies that reflect on children’s credibility and those that reflect on adults’ inadequacies in questioning children.

Footnotes

Acknowledgements

Special thanks to Stephanie Morse, Elliana Lederman, Kayleigh Brehen, and Kaylan Gehr for coding support; Samantha Andrews and Zsofi Szojka for early intellectual contributions; and Sebastienne Leo for initial project assistance.

ORCID iDs

Colleen E. Sullivan

Stacia N. Stolzenberg

Shanna Dewit-Williams

Thomas D. Lyon

Ethical Considerations

The use of archived interviews for research purposes was approved by the University of Southern California Institutional Review Board as exempt (45 CFR Section 46.014(d)(4)(ii)).

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This project was supported by Eunice Kennedy Shriver National Institute of Child Health and Human Development Grant HD101617 and the National Science Foundation Postdoctoral Fellowship SPRF 2494704.

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Data Availability Statement

De-identified data are available to researchers upon request.*

References

Andrews

S. J.

Lamb

M. E.

(2016). How do lawyers examine and cross-examine children in Scotland? Applied Cognitive Psychology, 30(6), 953–971. https://doi.org/10.1002/acp.3286

Andrews

S. J.

Lamb

M. E.

(2017). The structural linguistic complexity of lawyers’ questions and children’s responses in Scottish criminal courts. Child Abuse & Neglect, 65, 182–193. https://doi.org/10.1016/j.chiabu.2017.01.022

Andrews

S. J.

Lamb

M. E.

(2019). Lawyers’ question content and children’s responses in Scottish criminal courts. Psychology, Crime and Law, 25(10), 1008–1027. https://doi.org/10.1080/1068316X.2019.1611829

Andrews

S. J.

Lamb

M. E.

(2021). Lawyers’ question repetition and children’s responses in Scottish criminal courts. Journal of Interpersonal Violence, 36(1-2), 276–296. https://doi.org/10.1177/0886260517725739

Andrews

S. J.

Lamb

M. E.

Lyon

T. D.

(2015). Question types, responsiveness and self-contradictions when prosecutors and defense attorneys question alleged victims of child sexual abuse. Applied Cognitive Psychology, 29(2), 253–261. https://doi.org/10.1002/acp.3103

Behzadnia

Mehrani

M. B.

(2018b). Young children’s yes bias in response to tag questions. Early Child Development and Care, 188(12), 1665–1674. https://doi.org/10.1080/03004430.2016.1278216

Block

R. A.

Zakay

Hancock

P. A.

(1999). Developmental changes in human duration judgments: A meta-analytic review. Developmental Review, 19(1), 183–211. https://doi.org/10.1006/drev.1998.0475

Brennan

(1988). Strange language: Child victims under cross-examination. Charles Sturt University-Riverina.

Brennan

Silman

(1992). Statistical methods for assessing observer variability in clinical measures. BMJ British Medical Journal, 304(6840), 1491–1494. https://doi.org/10.1136/bmj.304.6840.1491

10.

Byrt

Bishop

Carlin

J. B.

(1993). Bias, prevalence and kappa. Journal of Clinical Epidemiology, 46(5), 423–429. https://doi.org/10.1016/0895-4356(93)90018-V

11.

California v. Riviera , 2001 WL 1190640 (Cal. Ct. App., 2001).

12.

CALJIC . (2005). California jury instructions criminal. Superior court.

13.

Cassel

W. S.

Roebers

C. E. M.

Bjorklund

D. F.

(1996). Developmental patterns of eyewitness responses to repeated and increasingly suggestive questions. Journal of Experimental Child Psychology, 61(2), 116–133. https://doi.org/10.1006/jecp.1996.0008

14.

Chambers v. Mississippi , 410 U.S. 284 (1973).

15.

Cross

T. P.

Whitcomb

(2017). The practice of prosecuting child maltreatment: Results of an online survey of prosecutors. Child Abuse & Neglect, 69(7), 20–28. https://doi.org/10.1016/j.chiabu.2017.04.007

16.

Davies

G. M.

Noon

(1991). An evaluation of the live link for child witnesses: Home Office.

17.

Davis v. Alaska , 415 U.S. 308 (1974).

18.

Dellatolas

Viguier

Deloche

De Agostini

(1998). Right-left orientation and significance of systematic reversal in children. Cortex, 34(5), 659–676. https://doi.org/10.1016/S0010-9452(08)70771-8

19.

Evans

A. D.

Stolzenberg

S. N.

Lyon

T. D.

(2017). Pragmatic failure and referential ambiguity when attorneys ask child witnesses “Do you know/remember” questions. Psychology, Public Policy, and Law, 23(2), 191–199. https://doi.org/10.1037/law0000116

20.

Friedman

W. J.

(1992). Children’s time memory: The development of a differentiated past. Cognitive Development, 7(2), 171–187. https://doi.org/10.1111/j.1467-8624.1991.tb01520.x

21.

Friedman

W. J.

Lyon

T. D.

(2005). Development of temporal‐reconstructive abilities. Child Development, 76(6), 1202–1216. https://doi.org/10.1111/j.1467-8624.2005.00845.x

22.

Greenstock

Pipe

M. E.

(1996). Interviewing children about past events: The influence of peer support and misleading questions. Child Abuse & Neglect, 20(1), 69–80. https://doi.org/10.1016/0145-2134(95)00117-4

23.

Henderson

H. M.

Lundon

G. M.

Lyon

T. D.

(2023). Suppositional wh-questions about perceptions, conversations, and actions are more productive than paired yes-no questions when questioning maltreated children. Child Maltreatment, 28(1), 55–65. https://doi.org/10.1177/10775595211067208

24.

Krackow

Lynn

S. J.

(2003). Is there touch in the game of twister? The effects of innocuous touch and suggestive questions on children’s eyewitness memory. Law and Human Behavior, 27(6), 589–660. https://doi.org/10.1023/B:LAHU.0000004888.52210.35

25.

Lyon

T. D.

Wylie

B. E.

Szojka

Z. A.

(2026). Initial denial of childsexual abuse: Reluctance or suggestibility. In Goodman

(G. S.

Lyon

T. D.

Hobbs

S. D.

Enriquez

(Eds.), Handbookof child witnesses: Improving psychological science and legalpractice (pp. 97–118). American Psychological Association.

26.

Mahoney

(2015). Lawyers’ statements of personal opinion to the court. Otago Law Review, 14(1), 147–170. https://www.austlii.edu.au/nz/journals/OtaLawRw/2015/10.pdf

27.

Malloy

L. C.

Lyon

T. D.

Quas

J. A.

(2007). Filial dependency and recantation of child sexual abuse allegations. Journal of the American Academy of Child & Adolescent Psychiatry, 46(2), 162–170. https://doi.org/10.1097/01.chi.0000246067.77953.f7

28.

McWilliams

Williams

Stolzenberg

S. N.

Evans

A. D.

Lyon

T. D.

(2021). Don’t know responding in young maltreated children: The effects of wh-questions type and enhanced interview instructions. Law and Human Behavior, 45(2), 124–137. https://doi.org/10.1097/01.chi.0000246067.77953.f7

29.

Mueller

C. B.

Kirkpatrick

L. C.

Richter

L. L.

(2018). Evidence (6th ed.). Aspen.

30.

Myers

J. E.

Redlich

A. D.

Goodman

G. S.

Prizmich

L. P.

Imwinkelried

(1999). Jurors’ perceptions of hearsay in child sexual abuse cases. Psychology, Public Policy, and Law, 5(2), 388–419. https://doi.org/10.1037/1076-8971.5.2.388

31.

Namsor. (n.d.). Namsor

, for name origin, ethnicity and gender detection. Namsor Applied Onomastics. https://namsor.app/

32.

Perry

N. W.

McAuliff

B. D.

Tam

Claycomb

Dostal

Flanagan

(1995). When lawyers question children: Is justice served? Law and Human Behavior, 19(6), 609–629. https://doi.org/10.1007/BF01499377

33.

Pillow

B. H.

Hill

Boyce

Stein

(2000). Understanding inference as a source of knowledge: Children’s ability to evaluate the certainty of deduction, perception, and guessing. Developmental Psychology, 36(2), 169–179. https://doi.org/10.1037//0012-1649.36.2.169

34.

Powell

Westera

Goodman-Delahunty

Pichler

A. S.

(2016). An evaluation of how evidence is elicited from complainants of child sexual abuse: Royal Commission into Institutional Responses to Child Sexual Abuse.

35.

Rakoczy

Behne

Clüver

Dallmann

Weidner

Waldmann

M. R.

(2015). The side-effect effect in children is robust and not specific to the moral status of action effects. PLoS One, 10(7), e0132933. https://doi.org/10.1371/journal.pone.0132933

36.

Rigal

(1996). Right-left orientation, mental rotation, and perspective-taking: When can children imagine what people see from their own viewpoint? Perceptual and Motor Skills, 83(3), 831–842. https://doi.org/10.2466/pms.1996.83.3.831

37.

Rocha

E. M.

Marche

T. A.

Briere

J. L.

(2013). The effect of forced-choice questions on children’s suggestibility: A comparison of multiple-choice and yes/no questions. Canadian Journal of Behavioural Science/Revue Canadienne des Sciences du Comportement, 45(1), 1–11. https://doi.org/10.1037/a0028507

38.

Ruiz-Earle

Sullivan

C. E.

Stolzenberg

S. N.

(2024). Pronoun anaphora and children's developing abilities to backward reference in criminal cases of alleged child sexual abuse. Applied Cognitive Psychology, 38(5), e4257. https://doi.org/10.1002/acp.4257

39.

St George

Henderson

Flippin

Stolzenberg

S. N.

Lyon

T. D.

(2024). Appellate review of child sexual abuse convictions: The importance of character evidence. Child Maltreatment, 29(4), 661–671. https://doi.org/10.1177/10775595231196096

40.

Stolzenberg

S. N.

Lyon

T. D.

(2014). Evidence summarized in attorneys’ closing arguments predicts acquittals in criminal trials of child sexual abuse. Child Maltreatment, 19(2), 119–129. https://doi.org/10.1177/1077559514539388

41.

Stolzenberg

S. N.

McWilliams

Lyon

T. D.

(2017). Spatial language, question type, and young children’s ability to describe clothing: Legal and developmental implications. Law and Human Behavior, 41(4), 398–409. https://doi.org/10.1177/1077559514539388

42.

Sullivan

C. E.

St George

Stolzenberg

S. N.

Williams

Lyon

T. D.

(2022b). Imprecision about body mechanics when child witnesses are questioned about sexual abuse. Journal of Interpersonal Violence, 37(13–14), NP12375–NP12397. https://doi.org/10.1177/0886260521997941

43.

Sullivan

E. Stolzenberg

N. (in press) (Submitted for publication). Anaphora in attorney questions and children’s developing abilities to backward reference in criminal cases of alleged child sexual abuse. Annals of the New York Academy of Sciences. https://doi.org/10.1111/nyas.70332

44.

Sullivan

C. E.

Stolzenberg

S. N.

Williams

Lyon

T. D.

(2022a). Children’s underextended understanding of touch. Psychology, Public Policy, and Law, 28(4), 505–514. https://doi.org/10.1037/law0000374

45.

Szojka

Z. A.

Fondacaro

H. N.

Yalcinkava

S. R.

Lyon

T. D.

(2026). Difficulties in attorneys’ before/after questions in child sexual abuse trials. Psychology, Public Policy, and Law, 32(2), 218–230. https://doi.org/10.1037/law0000479

46.

Szojka

Z. A.

Henderson

H. M.

Hur

Siepmann

Lyon

T. D.

(2023a). Elaborations and denials in children’s responses to yes–no any/some questions in forensic interviews. Child Maltreatment, 28(3), 407–416. https://doi.org/10.1177/10775595231154552

47.

Szojka

Z. A.

Moussavi

Burditt

Lyon

T. D.

(2023b). Attorneys’ questions and children’s responses referring to the nature of sexual touch in child sexual abuse trials. Child Maltreatment, 28(3), 438–449. https://doi.org/10.1177/10775595231161033

48.

United States v. Young , 470 U.S. 1 (1985).

49.

Wandrey

Lyon

T. D.

Quas

J. A.

Friedman

W. J.

(2012). Maltreated children's ability to estimate temporal location and numerosity of placement changes and court visits. Psychology, Public Policy, and Law, 18(1), 79–104. https://doi.org/10.1037/a0024812

50.

Woiwod

D. M.

Fitzgerald

R. J.

Sheahan

C. L.

Price

H. L.

Connolly

D. A.

(2019). A meta-analysis of differences in children’s reports of single and repeated events. Law and Human Behavior, 43(1), 99–116. https://doi.org/10.1037/lhb0000312

51.

Wylie

Stolzenberg

Lyon

T. D.

(2026). The potential ambiguity of negative questions in children’s testimony. Applied Cognitive Psychology, 40(2), Article e70182. https://doi.org/10.1002/acp.70182

52.

Wylie

B. E.

George

S. S.

McWilliams

Evans

A. D.

Stolzenberg

S. N.

(2022). Children’s acquiescence to polysemous implicature questions about coaching: The role of parental support. Journal of Applied Developmental Psychology, 78, Article 101370. https://doi.org/10.1016/j.appdev.2021.101370

53.

Zajac

Canaan

(2009). Cross-examination of sexual assault complainants: A developmental comparison. Psychiatry, Psychology and Law, 16(sup1), S36–S54.

54.

Zajac

Gross

Hayne

(2003). Asked and answered: Questioning children in the courtroom. Psychiatry, Psychology and Law, 10(1), 199–209. https://doi.org/10.1375/pplt.2003.10.1.199