Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2024 Mar 5.
Published in final edited form as: Child Maltreat. 2022 Jan 13;28(1):55–65. doi: 10.1177/10775595211067208

Suppositional wh- questions about perceptions, conversations, and actions are more productive than paired yes-no questions when questioning maltreated children

Hayden M Henderson 1, Georgia M Lundon 1, Thomas D Lyon 1
PMCID: PMC10914390  NIHMSID: NIHMS1963523  PMID: 35025692

Abstract

Forensic interviewers are taught to pair yes-no questions with open-ended requests for recall in order to reduce the likelihood that they will be misled by false “yes” responses. However, yes-no questions may elicit false “no” responses. Questioning 112 6- to 11-year-old maltreated children about three innocuous events (outside activities, yesterday, last birthday), this study compared the productivity of paired yes-no questions about perceptions, conversations, and actions involving the hands and mouth (e.g., “Did you say anything?”) with wh- questions (e.g., “What did you say?”). The wh- questions presupposed that children had content to provide, but did not specify that content. Children were twice as likely to deny content and half as likely to provide novel information when interviewers asked them yes-no questions. Younger children were more inclined than older children to deny content and give unelaborated “yes” responses. The results support further research into the potential for suppositional wh- questions to increase child witnesses’ productivity.

Keywords: child maltreatment, child sexual abuse, forensic interviews, interviewing children, interview techniques

Although forensic interviewers strive to maximize their use of open-ended questions, guidelines typically acknowledge that interviewers may need to ask more direct questions about information that a child omits, either because the child fails to recall the information, fails to appreciate its importance, or is reluctant to disclose it (American Professional Society on the Abuse of Children [APSAC], 2012; Newlin et al., 2015). When interviewers feel compelled to ask yes-no questions probing for such content, they are advised to pair their yes-no questions with invitations requesting elaboration if the child answers affirmatively (e.g., “Did he say anything?”/“Yes”/ “Tell me more about that”; Lamb et al., 2018). Pairing yes-no questions with invitations is designed to reduce the dangers of false “yes” responses and to maintain the emphasis on eliciting as much information as possible using open-ended questions (Lamb et al., 2018).

An unexplored potential problem with pairing is that children may provide false “no” responses, leading interviewers to move on to other content. An alternative is to ask wh- questions that directly ask for content, but do not specify what that content might be (e.g., “What did he say?”). Although these questions are often called suppositional (i.e., “What did he say?” presupposes that he said something), their lack of specific content, the ease with which children can respond “nothing,” and their potential for eliciting innocuous content suggests that they should not be considered unduly suggestive. This study explored whether suppositional wh- questions about perceptions, conversations, and actions are more effective in eliciting novel details from children than paired yes-no questions. We compared the productivity of wh- questions and yes-no questions as part of narrative practice rapport building with 6- to 11-year-old maltreated children. In what follows, we will first describe the pros and cons of pairing, highlighting the dangers of false negative responses to yes-no questions, and then discuss how some yes-no questions might safely be replaced with suppositional wh- questions.

The Pros and Cons of Pairing Yes-No Questions with Invitations

Forensic interviewing protocols and guidelines universally emphasize that interviewers should maximize their use of open-ended questions, including invitations, which include questions about “what happened” and questions that ask children to “tell more” about mentioned details (APSAC 2012; Lamb et al., 2018; Lyon, 2014; Newlin et al., 2015). In turn, wh- questions (what, how, where, when, why), which probe for recall memory, are preferred to yes-no questions, which probe for recognition memory. Recognition questions increase the likelihood of false positives, and on a question-by-question basis elicit fewer details than recall questions (Lamb et al., 2018). On the other hand, recognition questions are often more sensitive to memory than recall questions because they serve as a cue to memory (Schneider, 2015).

In the Revised NICHD protocol, interviewers are advised to prioritize invitations over other question types and to move to yes-no questions “only if you have already tried other approaches,” including invitations and wh- questions (Lamb et al., 2018, p. 249). When interviewers ask yes-no questions, they are instructed to pair those questions with invitations “whenever possible” (Lamb et al., 2018, p. 249). The interviewer should remind the child of an “activity, object, feeling [or] thought” mentioned by the child and ask a yes-no question about a “detail for [the] child to confirm or deny” (Lamb et al., 2018, p. 250). If the child responds affirmatively, the interviewer follows up with an invitation (e.g., “Tell me more about that”). Advice that interviewers pair yes-no questions with invitations can also be found in other protocols and practice guides (APSAC, 2012; Home Office, 2011 [Achieving Best Evidence]). Reviewing seven studies examining adoption of the NICHD protocol, Lamb et al. (2018) found that 12 to 35% of the questions asked were option-posing, which includes yes-no questions.

Pairing May Reduce False Positives

Pairing is thought to ameliorate some of the limitations of recognition questions. Younger children’s “yes” responses to recognition questions tend to be unelaborated (Lyon et al., 2019), even when the questions implicitly request additional information, such as when interviewers ask “do you know wh-” questions (Evans, et al., 2014; 2017). Asking children to “tell me more” can assist them in elaborating. Furthermore, research suggests that young children may give false “yes” responses. Incomprehensible yes-no questions elicit “yes” responses among children under three-and-a-half-years of age (Fritzley & Lee, 2003; Fritzley et al., 2013), and young children are prone to give false “yes” responses to plausible content (Ornstein et al., 1992; Rocha et al., 2013). Pairing can reduce the likelihood that false “yes” responses will be misconstrued. Stolzenberg and colleagues (2017) examined 4- to 9-year-old’s responses to yes-no questions directly inquiring whether “something bad” had happened; children who truthfully responded “yes” were able to elaborate on their responses 85% of the time, compared to only 8% of the false “yes” responders.

Pairing May Increase False Negatives

However, a largely overlooked potential problem with pairing is that children may give false “no” responses to the initial yes-no question, undermining the interviewer’s opportunity to elicit additional information through follow-up invitations. Furthermore, pairing forces the child to utter an explicit denial regarding content that would otherwise remain unmentioned (Lyon & Henderson, 2021), which means that any subsequent disclosure of the content will contradict the child’s original statements.

Children are particularly likely to provide false “no” responses to questions about undesirable or incriminating conduct (Talwar & Crossman, 2012). In Stolzenberg et al. (2017), asking a series of six yes-no questions directly inquiring into wrongdoing elicited true disclosures from approximately half of children who had failed to disclose when asked recall questions, but elicited persistent false “no” responses from the other half. Indeed, a common finding in research examining children’s disclosures of transgressions is that recognition questions will elicit disclosures from some children who fail to disclose in response to recall questions, but at the same time elicit a high percentage of false negatives (Ahern et al., 2016 [51%]; Quas et al., 2018 [54%]; Stolzenberg et al., 2017 [53%]).

Furthermore, yes-no questions are often worded in ways that pull for “no” responses. For example, when interviewers ask children “do you know wh-” questions, children will frequently answer “no,” even though they would give accurate information in response to the wh- question (Evans et al., 2014; 2017; Lyon & Saywitz, 1999). In forensic interviews, “Do you know why you came to talk to me?” has been shown to be less effective in eliciting abuse disclosures than the wh- question, “Tell me why you came to talk to me” (Hughes-Scholes & Powell, 2013).

Questions With Polarity Items (some, any, ever)

Yes-no questions that contain “polarity items” may be particularly likely to pull for false “no” responses. Examining trial transcripts of 5- to 12-year-old children testifying in child sexual abuse cases, Sullivan and colleagues (in press) identified miscommunications between the attorney and child regarding the body mechanics of abuse. In one-third of the miscommunications, attorneys initially failed to elicit abuse information using vague terms such as “some,” “any,” and “ever” (e.g., “Did something happen to you in the bathroom?”), also known as “polarity items.”

Polarity items can be positive or negative. “Some” is a positive polarity item because it occurs most often in a positive context. For example, “I called somebody” is sensible (what linguists refer to as “felicitous”), but “I didn’t call somebody” is infelicitous. Negative polarity items (any, ever) are most often restricted to negative contexts. For example, “Nicholas didn’t say anything” is felicitous, but “Nicholas said anything” is not, and “Nicholas hasn’t ever spoken” is felicitous but “Nicholas has ever spoken” is not (Giannakidou, 2017).

Negative polarity items may lead to higher rates of false “no” responding than positive polarity items. Heritage and colleagues (2007) found that adult patients at a health clinic who previously identified more than one health concern provided affirmative responses 90% of the time when their doctor asked “Is there something else you want to address in this visit today?” but only 53% of the time when asked “Is there anything else you want to address in this visit today?” Others have anecdotally observed the unproductivity of “anything” questions when questioning children (Brubacher et al., 2019; Childs & Walsh, 2018; Walker, 2013).

Suppositional Wh- Questions

Because wh- questions tend to elicit more productive and accurate responses than yes-no questions, it would appear advantageous to substitute pairing yes-no questions with wh- questions. The problem is that unless a child has already mentioned a topic, wh- questions about that topic are often considered suggestive, and protocols and guidelines universally recommend that interviewers avoid suggestive questions (Newlin et al., 2015; Lamb et al., 2018). However, in practice, it is difficult to identify with precision what sort of questions should be called “suggestive” (Peterson et al., 1999).

In their definition of “suggestive,” Lamb and colleagues (2018) identified a “suppositional” subcategory, which they defined as “assuming or implying an undisclosed allegation-specific content” (p. 61; see also Stevens et al., 2021). The “suppositional” label may include a large number of questions that are technically suggestive, but do not increase error. In a lab study examining the accuracy and productivity of different question types, Brown and colleagues (2013) found that children answered suggestive questions as accurately as cued invitations, in which interviewers ask children to “tell more” about previously mentioned content. Lamb and colleagues (2018) speculated that this may have been because the questions were only mildly suggestive. Significantly, Brown and colleagues (2013) taught their interviewers to avoid option-posing questions and emphasized that their interviewers never used tag questions; therefore the “suggestive” questions would have been predominantly wh- questions. For example, Brown and colleagues (2013) coded the question “What other things did you wear?” as suggestive, because the question presupposed that the child wore something else.

Universally accepted invitations such as “What happened next?” and “Tell me more about [child-generated content]” presuppose that a child has more to offer but fail to suggest content and are readily rejected by children. Researchers have shown that children’s responses to wh- questions about perceptions, conversations, and actions are often as accurate as their responses to invitations (Brown & Pipe, 2003; Canning & Peterson, 2020; Kulkofsky, 2010; Stolzenberg et al., 2018). Therefore, questions that presuppose children have something to offer regarding perceptions, conversations, and actions may also be productive, and more to the point, more productive than paired yes-no questions.

The Current Study

This study examined the productivity of asking 6- to 11-year-old maltreated children suppositional wh- questions about perceptions, conversations, and actions with the hands and mouth, comparing them to yes-no questions paired with follow-up invitations. Children were first asked free recall questions about each of three topics often used in narrative practice (i.e., outside activities, yesterday, and last birthday), followed by paired yes-no questions or suppositional wh- questions. If children in the paired yes-no question condition (hereinafter the yes-no condition) gave an unelaborated “yes” response, they were asked to “tell more,” and if they answered “no,” the interviewer accepted the response and moved on. Children in the suppositional wh- questions condition (hereinafter the wh- condition) were asked the wh- questions without any follow up question. The questions asked about perceptions (e.g., “What did you see/hear?”), conversations (e.g., “What did you say?”), and actions (e.g., “What did you do with your hands/mouth?”). We hypothesized that compared to the yes-no questions, the wh- questions would 1) elicit fewer denials of content, 2) elicit fewer responses lacking substantive details (which included denials plus don’t know responses and other responses lacking content), and 3) elicit a higher percentage of answers with novel content. We also predicted age differences, whereby younger children would be more likely than older children to deny content and to provide unelaborated “yes” responses to the yes-no questions.

Method

Participants

The final sample consisted of 112 6- to 11-year-old (Mage = 8.4 years old, SD = 1.7, 47% males) victims of maltreatment. Children were evenly distributed across age groups: 20 6-year-olds; 20 7-year-olds; 19 8-year-olds; 18 9-year-olds; 18 10-year-olds; 17 11-year-olds. We excluded 29 children who originally assented (Mage = 7.8 years old, SD = 1.8, 63% males), either because of a script error (n = 4), the child withdrew assent (n = 7), or court business, such as an attorney interview, a family visit, or a court appearance (n = 18). Children were primarily from ethnic/racial minority backgrounds: 63% Latinx, 28% African American, 8% Caucasian, and 2% Asian. The children were recruited from the Los Angeles County Dependency Court and had been removed from the custody of their parents or guardians because of substantiated abuse or neglect. Because children had been removed from parental care, the Presiding Judge of Juvenile Court granted consent for all children to participate in the study. Children were ineligible if they were awaiting an adjudication or contested disposition hearing on the date of testing because they might be asked to testify in court.

Procedure

All sessions were videotaped and began with the child providing assent. Children were randomly assigned to either the yes-no or wh- condition, stratified by age, gender, and ethnicity. After assenting, the interviewer began a scripted narrative practice with the child in which three topics were discussed: things the child liked to do outside, what happened yesterday, and what happened on the child’s last birthday. The order in which children were asked about the three topics was counterbalanced. The interviewer began the first narrative by saying “So [child’s name], today I’m going to ask you about three different things. This is the very first time we have ever met, and I want to know more about you.” For the “outside” narrative, the interviewer first asked, “First, tell me about things you like to do outside.” Following the child’s response, the interviewer would select the activity most likely to involve others (e.g., soccer rather than drawing) and follow up with, “All right, [child’s name], you said you like to [X]. Tell me everything that happened the last time you [X].” For the “yesterday” and “birthday” narratives, the interviewer asked “Tell me about everything you did yesterday from the time you woke up” or “Tell me about your last birthday. Tell me everything that happened.” If the child initially failed to provide a sequence, the interviewer would follow up with “What is the first thing that happened?” The interviewer then continued to ask “what happened next” questions until the child stopped providing information, or until five minutes elapsed. If the child stopped providing information before five minutes had elapsed, interviewers spent the rest of the time following up on main events by asking “You said [X]. Tell me more about that.”

At the end of each narrative practice topic, the interviewer asked the scripted yes-no or wh- questions, which included content from the child’s narrative as cues. In the wh- condition, the interviewer asked the wh- questions without any requests for elaboration. In the yes-no condition, if the child answered with an unelaborated yes, then the interviewer followed up with “Tell me more about that.” (See Online Appendix 1 for a complete list of questions.) Unelaborated “yes” responses consisted solely of a “yes” response (including “mm-hmm” and head nods) with no additional information. If the child answered “no,” the interviewer moved to the next question. Because three of the scripted questions were dependent upon the presence of another individual, if the child had not mentioned another person when finishing narrative practice, the interviewer asked “Tell me who was there.” If the child mentioned a collective group of individuals (e.g., “my friends”), then the questions were adapted to refer to the group (e.g., “What did your friends do with their hands?”). If the child said that no one was there, then the three questions about another individual(s) were excluded (this occurred only six times across all children and topics).

Coding

Children’s responses included any verbal utterance or non-verbal communication (e.g., head nods, shrugs). We coded for denials, non-substantive responses, novel responses, and non-novel responses. Denials of content included unelaborated “no” responses in response to yes-no questions and “nothing” responses to wh- questions. Non-substantive responses included denials and additional types of responses that failed to provide substantive information, including non-responsive (i.e., silence, off-topic); “I don’t know/remember (IDK)” or uncertain (e.g., “I’m not sure”) responses without elaboration; clarification-seeking responses without elaboration; boundary markers without elaboration (which signal that the child had reported everything, e.g., “That’s all”); and underinformative responses without elaboration (in which the child provided a responsive but vague answer, such as “stuff”). A response was considered novel if it contained relevant information that was not mentioned during the free recall portion of narrative practice. Coders were instructed not to make any inferences, and thus coded liberally for novelty. For example, if the child said that she played with her brother and later said that she played football, but did not specify who she played with, the coder would not infer that the child played football with her brother (e.g., because there might have been multiple people there, the child could have played alone, etc.). Non-novel responses repeated detail(s) produced during free recall. (Response codes with definitions are in Online Appendix 2.)

If interviewers sought clarification or elicited additional information by going off-script (e.g., using echoes or facilitators during the scripted yes-no or wh- questions), children’s additional responses were excluded in order to maintain consistency in the sample. A small number of responses were inaudible, and these were also excluded. A reliability coder coded 20% of the sample and inter-rater agreement for all response types was high (i.e., for all variables, κ > 0.87 and percent agreement > 96%).

Analysis Plan

We calculated descriptives examining the prevalence of different responses in the yes-no and wh- conditions. We also calculated descriptives examining children’s responses to the specific prompt topics (see, hear, say, hands, mouth). Next, generalized linear mixed models (GLMMs) examined whether condition (yes-no, wh-), child’s age, and an interaction affected children’s responses (i.e., denials of content, non-substantive responses, and novel responses). Last, a GLMM examined whether age significantly affected children’s “yes” responses to yes-no questions (elaborated, unelaborated). Random effects for ‘child’ and ‘interviewer’ were included due to the repeated nature of questioning in order to account for any individual variation within the child or interviewer. Exploratory analyses examining narrative practice topic (outside, yesterday, birthday) revealed that narrative practice topic interacted with condition (yes-no vs. wh-) when examining denials and non-substantive responses (but not novel details); because the interactions never reversed the direction of the condition differences, narrative practice topic was not included in the models below (see Online Appendix 3).

Analyses were performed using the glmer function in the R package lme4 with the bobyqa optimizer (Bates et al., 2015). GLMMs combine the properties of linear mixed models (which incorporate random effects) and generalized linear models (which handle non-normal data) and are preferable to traditional analysis of variance (ANOVA) models because they have fewer assumptions, handle response variables from different distributions (e.g., binary), and maximize power while simultaneously estimating between-subject variance (Bates et al., 2015). Models were cross-validated in order to identify the best fit model, which was determined by the Akaike Information Criteria (AIC), an estimator of the relative quality of a model for a given set of data (Vrieze, 2012). The best fit models are reported below accompanied by the unstandardized fixed effect estimates (β), standard errors of the estimates (SE), and estimates of significance (Z andp values).

Results

The percentage of children’s responses that were denials of content, non-substantive, novel, or non-novel is in Table 1. Table 2 breaks these percentages down by the different question topics (see, hear, say, hands, mouth). The odds ratios reveal that wh- questions were consistently less likely than yes-no questions to elicit denials and non-substantive responses, and consistently more likely to elicit novel information.

Table 1.

Response Types by Condition

Paired Yes-No Suppositional Wh- OR 95% CI
N % N %
Denial 389 28 152 11 0.32 0.26, 0.39
Non-substantive 442 32 249 18 0.47 0.39, 0.56
 Non-responsive 7 1 4 <1 0.58 0.15, 1.98
 IDK/Uncertain 10 1 50 4 5.11 2.69, 10.78
 Clarification 31 2 28 2 0.91 0.54, 1.52
 Underinformative 2 <1 10 1 10.11 1.29, 79.09
 Boundary marker 3 <1 5 <1 1.64 0.39, 8.53
Novel 818 60 1007 73 1.9 1.62, 2.24
Non-novel 110 8 108 8 0.98 0.75, 1.3
Open in a new tab

Note. The reported numbers are unadjusted means; adjusted means are noted in the text. Non-substantive responses are the sum of denials and non-responsive, idk/uncertain, clarification, underinformative, and boundary marker responses.

Table 2.

Response Types by Prompt and Condition

Response Type Question Paired Yes-No Suppositional Wh- OR 95% CI
N % N %
Denial See 54 29 8 4 0.11 0.05, 0.23
Hear 39 23 13 8 0.28 0.14, 0.54
Say 97 28 45 13 0.39 0.26, 0.58
Hands 83 25 32 9 0.31 0.19, 0.48
Mouth 45 35 55 17 0.37 0.26, 0.54
Non-substantive See 72 40 34 19 0.36 0.22, 0.57
Hear 46 27 26 15 0.49 0.29, 0.84
Say 112 33 66 20 0.50 0.35, 0.71
Hands 89 27 51 15 0.49 0.33, 0.72
Mouth 123 37 72 22 0.47 0.34, 0.67
Novel See 86 47 126 69 2.4 1.63, 3.83
Hear 117 68 136 79 1.8 1.11, 2.95
Say 213 62 245 73 1.62 1.18, 2.50
Hands 217 65 262 78 1.87 1.33, 2.64
Mouth 185 55 238 71 2.01 1.46, 2.78
Non-novel See 27 15 24 13 0.88 0.48, 1.59
Hear 10 6 10 6 1.01 0.40, 2.54
Say 18 5 26 8 1.50 0.81, 2.85
Hands 29 9 25 7 0.63 0.37, 1.06
Mouth 26 8 23 7 0.88 0.49, 1.58
Open in a new tab

Did Wh- Questions Elicit Fewer Denials of Content Than Yes-No Questions?

Consistent with the first hypothesis, wh- questions (M = 8%, SE = 0.02) elicited fewer denials than yes-no questions (M = 24%, SE = 0.04, B = 1.20, S.E, 0.31, Z = 3.88, p < .001) regardless of children’s age. Also consistent with our first hypothesis regarding age effects, younger children were more likely to deny content than older children (B = −0.19, SE, 0.09, Z = −2.08, p = .04). A breakdown of the percentage of children’s denials at different ages in the two conditions is in Online Appendix 4.

Did Wh- Questions Elicit Fewer Non-substantive Responses Than Yes-No Questions?

Children might deny content less often in response to the wh- questions but still give a non-substantive response, rendering the wh- question ineffective. Children could, for example, answer “I don’t know,” since wh- questions are known to elicit higher rates of don’t know responding than yes-no questions among young children (Waterman et al., 2001). Indeed, whereas only 1% of children’s responses to yes-no questions were “I don’t know,” 4% of children’s responses to wh- questions were “I don’t know” (Table 1).

Therefore, we examined how often children in the two groups gave a non-substantive response, which included denials of content, don’t know responses, and some other rare response types. Consistent with our second hypothesis, wh- questions (M = 16%, SE = 0.03) elicited fewer non-substantive responses than yes-no questions (M = 29%, SE = 0.04, B = 0.77, S.E = 0.27, Z = 2.84, p = .005) regardless of child’s age. Children’s age was not significantly associated with whether children gave a non-substantive response (p = .12).

Did Wh- Questions Elicit More Novel Details Than Yes-No Questions?

Children might be more responsive to the wh- questions but simply repeat details from their free recall responses. Therefore, we examined how often children in the two groups gave responses that included novel details, which were details that had not been previously mentioned. Consistent with our third hypothesis, wh- questions were more likely to elicit novel details (M = 77%, SE = 0.03) than paired questions (M = 62%, SE = 0.04, B = −0.69, S.E = 0.24, Z = −2.83, p = .005) regardless of child’s age. Children’s age was not significantly associated with whether children provided novel details (p = .09).

Were Younger Children More Likely Than Older Children to Give Unelaborated “Yes” Responses to Yes-No Questions?

Consistent with our second age hypothesis, younger children were more likely to answer yes-no questions with unelaborated “yes” responses than older children (B = −0.45, SE, 0.19, Z = −2.39, p = .02). The mean age of unelaborated yes responses was 7.72 years old (SD = 1.62), while the mean age of elaborated yes responses was 8.64 years old (SD = 1.58). Whereas 38% of 6-year-olds’ yes responses were unelaborated, only 11% of the 11-year-olds’ were unelaborated.

Did Children Respond Differently to Questions about Actions with the Hands and Mouth?

Because wh- questions specifically asking about the hands and mouth have not been explored in prior research, we conducted exploratory analyses in which we tested whether the question topic (i.e., hands/mouth vs. see/hear/say) affected children’s rates of denials, non-substantive responses, and novel responses. We found that the condition effects (yes-no, wh-) were unchanged and that the question topic (hands/mouth vs. see/hear/say) did not significantly interact with question type (yes-no, wh-). However, we did find that hands/mouth questions elicited more denials (M = 17%, SE = 0.03) than see/hear/say questions (M = 13%, SE = 0.02) across condition (B = 0.33, SE = 0.11, Z = 2.87, p = .004). Hands/mouth questions did not significantly differ from see/hear/say questions in the rate of non-substantive or novel responses.

We developed a qualitative coding scheme to examine the type of content that children produced when asked the hands and mouth questions, including whether the questions elicited negatively valenced content. Two independent coders examined the responses to the hand/mouth questions (N = 980) and inter-rater reliability was high for topic of children’s responses to mouth questions (n = 463, k > .85, percent agreement > 96%), and the topic of children’s responses to hand questions (n = 517, k > .80, percent agreement > 97%), and negatively valenced responses (n = 12, Prevalence Adjusted Bias Adjusted Kappa > 0.98, percent agreement > 99%). All disagreements were discussed and resolved with 100% agreement.

Response topics were not mutually exclusive, and a child could describe multiple hand or mouth actions within a single response (e.g., playing a ball game and playing video games). Thirty-two percent of children’s responses to hand questions were coded as hands’ specific actions (n = 166; e.g., writing/drawing, holding/grabbing things, clapping/waving). Twenty-five percent of children’s hand responses included references to physical or outdoor activities (n = 131; e.g., playing ball games, climbing trees, riding a bike). Eighteen percent of children’s hand responses included references to pretend or fantasy play (n = 95, e.g., playing with dolls or playing video games). Fifteen percent of children’s hand responses included references to interpersonal touch (n = 75) and included actions such as hugging or playing tag. Five percent of children’s hand responses were about refraining from using the hands (n = 25; e.g., “I kept them by my side”), and three percent (n = 18) referred to completing chores. Finally, three percent (n = 18) of children’s hands’ responses were classified as ‘other’.

Seventy percent of children’s mouth responses included references to talking (n = 325), while 28% included references to eating (n = 129), and five percent were about refraining from using the mouth (n = 24; e.g., “I kept [my mouth] closed”). Seven percent of responses to the mouth question were coded as other (n = 34), and included topics such as brushing teeth, blowing out birthday candles, and using the mouth for laughing and smiling. Of all children’s responses to questions regarding the use of hands and mouth, less than 1% (n = 12) were negatively valenced.

Discussion

This study analyzed the productivity of questions about perceptions, conversations, and actions (with the hands and mouth) with 6- to 11-year-old maltreated children, comparing paired yes-no questions to suppositional wh- questions: wh- questions that presupposed that children had content to provide but did not specify what that content might be. We asked children to provide three practice narratives: their last birthday, what happened yesterday, and a recent activity they had nominated as something they liked to do outside. We then asked children questions about what they saw, what they heard, what was said, and what was done with hands and mouths. In the yes-no condition, we asked if the children had information to provide, using a negative polarity item (e.g., “Did you see anything?”), and if they gave an unelaborated “yes,” we asked for elaboration (“Tell me about that”). In the wh- condition, we presupposed that children had information to provide by asking a wh- question (e.g., “What did you see?”), without any request for elaboration.

We predicted that children’s responses to the yes-no questions would be less productive than their responses to the wh- questions. We also anticipated that younger children would be particularly inclined to deny content and to give unelaborated “yes” responses to the yes-no questions. Each hypothesis was supported. Children in the wh- condition 1) less often denied content, 2) less often gave nonsubstantive responses, and 3) more often provided novel information. Younger children were more likely than older children to give unelaborated responses to the yes-no questions; indeed, 84% of 6-year-olds’ responses to the yes-no questions were unelaborated. In what follows, we discuss the implications of the results for assessing the productivity of yes-no questions and suppositional wh- questions.

Yes-No Questions

The results highlight three problems with yes-no questions. First, children tend to give unelaborated responses, a tendency noted in observational work examining forensic interviews with children (Korkman et al., 2006). This largely explains why yes-no questions are found to be less productive than wh- questions and invitations (Lamb et al., 2018). Second, children’s unelaborated “yes” responses overlook implied requests for elaboration. The yes-no questions in this study imply the associated wh- question: for example, “Did you say anything?” implies “What did you say?” Nevertheless, children’s “yes” responses were often unelaborated, and this was particularly true among the 6-year-olds. A similar problem has been observed in experimental and observational work examining how children respond to “do you remember” and “do you know” questions that implicitly ask a wh- question (e.g., “Do you remember where it happened?”; Evans et al., 2014; 2017). Younger children are particularly likely to simply answer “yes,” whereas older children will answer the implied wh- question. Children’s tendency to give unelaborated responses to yes-no questions has been referred to as formal reticence, which refers to children’s tendency to provide the most easily retrievable and minimally sufficient responses to questions, based on their form (Lyon et al., 2019). Simply put, if a young child can simply answer a question “yes” or “no,” they will do so.

Formal reticence may also explain a third and often overlooked problem with yes-no questions observed here. Yes-no questions broadly asking about “any” perceptions, conversations, and actions are asked in order to uncover content children initially fail to produce in response to recall questions, and thus to reduce omissions. However, they often lead to false “no” responses, in which information is not merely omitted but overtly denied. 40% of children gave a non-substantive answer (usually a “no”) when asked “Did you see anything?” despite the fact that they surely saw something. Research with young children has shown that they answer yes-no questions more quickly than wh- questions (Williams et al., 2019). The ease with which children can answer yes-no questions may limit the time and effort they expend in memory search, and thus may predispose them to deny content. Conversely, the innocuous presupposition in suppositional questions that the child saw and heard things, that people spoke, and that people did things with their hands and mouth, may increase memory search and thus encourage more thoughtful and productive responses.

Generic Suppositional Wh- Questions

The results suggest a promising avenue for further research into suppositional wh- questions that may increase children’s productivity without increasing error, which could be called generic suppositional wh- questions. Questions could be classified as generic suppositional wh- questions, and distinguished from suggestive questions, if they fulfill several criteria. First, they must be easily answerable without acceptance of the presupposition. For example, children can easily answer “nothing” to “What did you do with your mouth?” just as they can answer “nothing” to “What happened next?” This helps to distinguish generic suppositional wh- questions from highly suggestive suppositional wh- questions (such as the classic “When did you stop beating your wife?” Williams, 1909).

Second, wh- questions should not be considered generic if a “yes” response to the corresponding yes-no question would constitute endorsement of a substantive detail. For example, unelaborated “yes” responses to sexual abuse questions such as “Did something come out of his penis?” or “Did he say what would happen if you told?” would constitute substantive details, such that their wh- counterparts (e.g., “What came out of his penis?” and “What did he say would happen if you told?”) are not generic. Conversely, because people routinely perceive innocuous details, say innocuous things, and use their hands and mouths in innocuous ways, questions about perceptions, conversations, and actions need not be considered inherently suggestive.

Third, in order for a type of wh- question to be a generic suppositional question, there should be research demonstrating that children’s responses are comparably accurate to their responses to other types of questions commonly recognized as non-suggestive, including invitations and cued invitations. With respect to perception questions, Poole and Lindsay (1995) found that “Tell me how everything looked” and “Tell me about all of the things that you heard” elicited large amounts of additional information from 3- to 6-year-olds who had witnessed a series of science demonstrations, and accuracy rates were as high as responses to free recall questions. Similarly, Elischberger & Roebers (2001) found that asking kindergartners and second graders to report everything they remembered seeing and everything they remembered hearing in a film increased productivity without compromising accuracy.

With respect to conversation questions and action questions, Saywitz and colleagues demonstrated the efficacy of the Narrative Elaboration procedure in a series of studies in which children were trained to provide additional information in recall when reminded by cards depicting different aspects of the events, including conversations (cards depicting people with talk bubbles) and specific actions (cards depicting a broken window or a person in motion; Saywitz & Camparo, 2014). Subsequent research showed equal increases in productivity with comparable accuracy when children were simply asked the corresponding wh- questions about what “people said” and “people did” (Brown & Pipe, 2003; Canning & Peterson, 2020; Kulkofsky, 2010), with the exception of one study examining preschool children with serious delays in verbal ability (Chae et al., 2014). Furthermore, with respect to conversation questions about specific people, Stolzenberg et al. (2018) found that 4- to 9-year-old’s responses to a series of questions about what the child, a confederate, and an interviewer had said were as accurate as free recall responses to questions about what had happened.

Limitations and Future Directions

An obvious limitation is that we were unable to assess accuracy. Although it seems clear that children who totally denied perceptions, conversations, and actions must have been falsely denying content, we cannot say that the content that children generated in response to the wh- questions was accurate. It is therefore possible that children asked the wh- questions provided less accurate reports than children asked the paired yes-no questions, and that the paired yes-no questions enabled children to screen out inaccurate information. Furthermore, although prior research has found that children’s responses to recall questions about actions in general are comparable in accuracy to their responses to invitations and cued invitations, the action questions asked in this study were specific to the hands and mouth, and this might raise concerns about suggestiveness. Our qualitative analyses of these questions showed that children had little difficulty in recalling innocuous content, recognizing that the hands and mouth are involved in a wide variety of activities. Nevertheless, before broader use of suppositional wh- questions is recommended to practitioners, experimental laboratory research is needed examining their potential to elicit errors, particularly the previously unexamined questions about actions with the hands and mouth.

Although laboratory research is needed, it is important to note that our naturalistic examination of maltreated children’s reports is nevertheless valuable. Inclusion of maltreated children is valuable for the obvious reason that they are the subject of investigations and thus are questioned about their experiences. Asking about naturalistic experiences rather than experiences in the lab enabled us to question children about events especially rich in detail and extended over time. These experiences are particularly valuable when assessing the productivity of children’s reports because of the potential for children to provide an almost unlimited number of details.

Future research should also further examine the factors that influence accuracy and productivity in response to paired yes-no questions. Several aspects of the questions are likely important. First, as noted in the introduction, yes-no questions often elicit high rates of false “no” responses when children find the content undesirable or incriminating (Talwar & Crossman, 2012). We deliberately questioned children about neutral or pleasant acts in order to specifically examine the effect of the form of questions on children’s productivity. This may underestimate the false “no” problem when children are questioned about maltreatment. Second, we combined the negative polarity item “any” with the vague term “thing” in the paired yes-no questions (e.g., “Did you see anything?”), and this likely increased children’s rejection rate (Heritage, 2007; Sullivan et al., in press). Future work can tease apart the various aspects of yes-no questions that pull for “no” responses. Will children reject other details if “any” is paired with more specific words, such as “anyone” or “anywhere”? Are children less likely to reject content if asked “Did you see something?” instead of “Did you see anything?”

Finally, future work should explore how well interviewers pair yes-no and direct questions with invitations in the field. It may be that we underestimated the productivity of paired yes-no questions because interviewers will ask follow-up invitations even when children elaborate their “yes” responses (we only paired the unelaborated “yes” responses). Conversely, we may have underestimated the benefits of wh- questions because of interviewers’ difficulty, documented in some observational research (Wolfman et al., 2016), in following up yes-no questions with anything other than additional yes-no questions.

Conclusion

This study has demonstrated the productivity of wh- questions that directly ask for content regarding perceptions, conversations, or actions, but do not specify what that content is. We questioned 6- to 11-year-old maltreated children about three narrative practice topics and, at the end of free recall, either asked paired yes-no questions (e.g., “Did you do anything with your hands?” / “Tell me about that”) or comparable wh- questions (“What did you do with your hands?”). Children were twice as likely to deny content when asked the paired yes-no questions, and twice as likely to provide novel information when asked the wh- questions. The results suggest that carefully crafted suppositional wh- questions using generic language may be preferable to yes-no questions when questioning children about previously unmentioned content.

Acknowledgments

The authors would like to thank Aga Nogalska and Owen Friend for their significant contributions. Additionally, the authors are grateful to Jina Hur, Hannah Fondacaro, Hannah Siepmann, Samantha Hardy, Rayna Enriquez, Hailey Konovalov, Tianyu Wang, and Scarlet Cho for their assistance.

This study was supported in part by the National Institute of Child Health and Development (Grant HD087685).

Online Appendices

Appendix 1. Paired Yes-No and Generic Suppositional Wh- Questions

Paired Yes-No Questions Generic Suppositional Wh- Questions
So you said [X]. Did you see anything? So you said [X]. What did you see?
For every prompt: If the child says no, move onto the next question.
If the child says yes: Tell me more about that.
Did you hear anything when you [X]? When you were [X], what did you hear?
Did you say anything when you [X]? When you were [X], what did you say?
When you [X], did you do anything with your hands? What did you do with your hands when you were [X]?
Did you do anything with your mouth when you [X]? When you were [X], what did you do with your mouth?
And you mentioned [Person 1] was there. Did [Person 1] say anything? And you mentioned [Person 1] was there. What did [Person 1] say?
Did [Person 1] do anything with his/her/THEIR hands? What did [Person 1] do with his/her/THEIR hands?
Did [Person 1] do anything with his/her/THEIR mouth? What did [Person 1] do with his/her/THEIR mouth?
Open in a new tab

Appendix 2. Children’s Response Types

Response Type Definition
Novel Response is on-topic and contains at least one detail that was not mentioned in the child’s previous responses
Non-novel Response is on-topic but not novel
Unproductive Rejection, Non-responsive, IDK/Uncertain, Clarification, Generic underinformative, Boundary marker
• Rejection An unelaborated “nothing” response to a wh- question or a “no” to a yes-no question
• Non-responsive Response is off-topic or child is silent
• IDK/Uncertain Child says “I don’t know”, “I’m not sure”, or “I forgot” with no additional elaboration
• Clarification Child asks a question that is intended to aid in understanding or answering the question (e.g., “You mean my dad?”)
• Underinformative Generic and vague response, such that the answer is responsive but does not add any information (e.g., “Stuff”).
• Boundary marker Child indicates that their report is complete (e.g. “That’s it”).
Open in a new tab

Appendix 3. Analyses Including Interaction between Condition and Narrative Practice Topic

GLMM Examining Whether Wh- Questions Elicited Fewer Denials of Content Than Yes-No Questions Including the Interaction Between Condition and Narrative Practice Topic

The best fit model included condition, age, narrative practice topic, and an interaction between condition and narrative practice topic. Consistent with the first hypothesis, wh- questions (M = 8%, SE = 0.02) elicited fewer denials than yes-no questions (M = 24%, SE = 0.04, B = 1.7420, SE, 0.351, Z = 4.933.88, p < .001) regardless of children’s age. Also consistent with our first hypothesis regarding age effects, younger children were more likely to reject content than older children (B = −0.19, SE, 0.09, Z = −2.08, p = .04). There was also a significant interaction between condition and narrative practice topic for both yesterday (B = −0.89, SE, 0.28, Z = −3.15, p = .002) and birthday (B = −0.74, SE, 0.29, Z = −2.56, p = .01) compared to outside. Post hoc analyses confirmed that for each narrative practice topic, wh- questions (Outside: 8%, SE = .02; Yesterday: 10%, SE = 0.02; Birthday: 7%, SE = .02) elicited fewer denials than yes-no questions (Outside: 33%, SE = .05; Yesterday: 20%, SE = 0.04; Birthday: 18%, SE = .04; see below for statistical output). Post hoc analyses also showed that there was no difference in rejection rates based on topic in the wh- condition; however, in the yes-no condition, questions about outside elicited significantly more denials than both yesterday (OR = 1.94, SE = 0.35, Z = 3.73, p < .001) and birthday questions (OR = 2.30, SE = 0.41, Z = 4.67, p < .001).

GLMM Examining Whether Wh- Questions Elicited Fewer Unproductive Responses Than Yes-No Questions Including the Interaction Between Condition and Narrative Practice Topic

The best fit model included condition, age, narrative practice topic, and an interaction between condition and narrative practice topic. Consistent with our second hypothesis, wh- questions (M = 16%, SE = 0.03) elicited fewer unproductive responses than yes-no questions (M = 29%, SE = 0.04, B = 1.150.77, SE = 0.3127, Z = 3.772.84, p <= .0015) regardless of child’s age. Children’s age was not significantly associated with whether children gave an unproductive response (p = .112). There was also a significant interaction between condition and narrative practice topic for both yesterday (B = −0.55, SE, 0.25, Z = −2.21, p = .03) and birthday (B = −0.64, SE, 0.25, Z = −2.58, p = .01) compared to outside. Descriptively, wh- questions (Outside: 16%, SE = .03; Yesterday: 17%, SE = 0.03; Birthday: 16%, SE = .03) elicited fewer unproductive responses than yes-no questions for each topic (Outside: 37%, SE = .05; Yesterday: 27%, SE = 0.04; Birthday: 24%, SE = .04). However, post hoc analyses indicated that the difference was only significant for outside (see below for statistical output). Post hoc analyses also showed that there was no difference in unproductive response rates based on topic in the wh- condition; however, in the yes-no condition, questions about outside elicited significantly more unproductive responses than both yesterday (OR = 1.57, SE = 0.26, Z = 2.72, p = .02) and birthday questions (OR = 1.87, SE = 0.31, Z = 3.75, p < .001).

Condition x Narrative Practice Topic Interaction Results

Response Type Narrative Practice Topic OR SE Z P
Denials Outside 0.18 0.06 −4.93 < .001
Yesterday 0.43 0.15 −2.38 0.02
Birthday 0.37 0.13 −2.78 0.005
Unproductive Outside 0.32 0.10 −3.77 < .001
Yesterday 0.55 0.17 −1.94 0.05
Birthday 0.60 0.18 −1.68 0.09
Open in a new tab

Appendix 4. Rates of Denial by Age and Condition

Age Group Paired Yes-No Suppositional Wh- OR 95% CI
N % N % N %
6-Year-Olds 109 40 32 15 0.26 0.16, 0.40
7-Year-Olds 75 34 27 11 0.23 0.14, 0.37
8-Year-Olds 81 31 29 15 0.38 0.23, 0.61
9-Year-Olds 40 20 26 11 0.46 0.27, 0.79
10-Year-Olds 47 19 13 7  0.3 0.15, 0.55
11-Year-Olds 38 22 25 10  0.4 0.23, 0.70
Open in a new tab

Note. Denial constituted an unelaborated “no” response in the paired yes-no condition, and a “nothing” response in the generic suppositional wh- condition (or an equivalent response, such as “I didn’t see anything”).

Footnotes

We have no known conflicts of interest to disclose.

References

  1. Ahern EC, Andrews SJ, Stolzenberg SN, & Lyon TD (2018). The productivity of wh- prompts in child forensic interviews. Journal of Interpersonal Violence, 33(13), 2007–2015. 10.1177/0886260515621084 [DOI] [PubMed] [Google Scholar]
  2. American Professional Society on the Abuse of Children. (2012). Practice guidelines: Forensic interviewing in cases of suspected child abuse. https://www.apsac.org/guidelines
  3. Bates D, Mächler M, Bolker B, & Walker S (2015). Fitting linear mixed-effects models using lme4. Journal of Statistical Software, 67(1), 1–48. 10.18637/jss.v067.i01 [DOI] [Google Scholar]
  4. Brown DA, Lamb ME, Lewis C, Pipe M, Orbach Y, & Wolfman M (2013). The NICHD investigative interview protocol: An analogue study. Journal of Experimental Psychology: Applied, 19(4), 367–382. 10.1037/a0035143 [DOI] [PubMed] [Google Scholar]
  5. Brown D, & Pipe ME (2003). Individual differences in Children’s event memory reports and the narrative elaboration technique. Journal of Applied Psychology, 88(2), 195–206. 10.1037/0021-9010.88.2.195 [DOI] [PubMed] [Google Scholar]
  6. Brubacher SP, Peterson C, La Rooy D, Dickinson JJ, & Poole DA (2019). How children talk about events: Implications for eliciting and analyzing eyewitness reports. Developmental Review, 51, 70–89. 10.1016/j.dr.2018.12.003 [DOI] [Google Scholar]
  7. Canning HS, & Peterson C (2020). Encouraging more open-ended recall in child interviews. Psychiatry, Psychology and Law, 27(1), 81–94. 10.1080/13218719.2019.1687045 [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Chae Y, Kulkofsky S, Debaran F, Wang Q, & Hart SL (2014). Low-SES Children’s eyewitness memory: The effects of verbal labels and vocabulary skills. Behavioral Sciences & the Law, 32(6), 732–745. https://doi-org.libproxy1.usc.edu/10.1002/bsl.2145 [DOI] [PubMed] [Google Scholar]
  9. Childs C, & Walsh D (2018). Paradoxical invitations: Challenges in soliciting more information from child witnesses. Research on Language and Social Interaction, 51(4), 363–378. 10.1080/08351813.2018.1524561 [DOI] [Google Scholar]
  10. Elischberger HB, & Roebers CM (2001). Improving young Children’s free narratives about an observed event: The effects of nonspecific verbal prompts. International Journal of Behavioral Development, 25(2), 160–166. 10.1080/01650250042000203 [DOI] [Google Scholar]
  11. Evans AD, Stolzenberg S, Lee K, & Lyon TD (2014). Young Children’s difficulty with indirect speech acts: Implications for questioning child witnesses. Behavioral Sciences & the Law, 32(6), 775–788. 10.1002/bsl.2142 [DOI] [PubMed] [Google Scholar]
  12. Evans AD, Stolzenberg SN, & Lyon TD (2017). Pragmatic failure and referential ambiguity when attorneys ask child witnesses “Do you know/remember” questions. Psychology, Public Policy, & Law, 23(2), 191–199. 10.1037/law0000116 [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Fritzley VH, & Lee K (2003). Do young children always say yes to yes-no questions? A metadevelopmental study of the affirmation bias. Child Development, 74(5), 1297–313. 10.1111/1467-8624.00608 [DOI] [PubMed] [Google Scholar]
  14. Fritzley VH, Lindsay RCL, & Lee K (2013). Young children’s response tendencies toward yes–no questions concerning actions. Child Development, 84(2), 711–725. https://dx.doi.org/10.1111%2Fcdev.12006 [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Giannakidou A (2017). Polarity in the semantics of natural language. In Oxford Research Encyclopedia of Linguistics. 10.1093/acrefore/9780199384655.013.7 [DOI] [Google Scholar]
  16. Heritage J, Robinson JD, Elliott MN, Beckett M, & Wilkes M (2007). Reducing patients’ unmet concerns in primary care: the difference one word can make. Journal of General Internal Medicine, 22(10), 1429–1433. 10.1007/s11606-007-0279-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Home Office. (2011). Achieving the best evidence in criminal proceedings: Guidance on interviewing victims and witnesses, and guidance on using special measures. http://www.cps.gov.uk/publications/docs/best_evidence_in_criminal_proceedings.pdf
  18. Hughes-Scholes CH, & Powell MB (2013). Techniques used by investigative interviewers to elicit disclosures of abuse from child witnesses: A critique. Police Practice and Research, 14(1), 45–52. 10.1080/15614263.2012.680716 [DOI] [Google Scholar]
  19. Korkman J, Santilla P, & Sandnabba NK (2006). Dynamics of verbal interaction between interviewer and child in interviews with alleged victims of child sexual abuse. Scandinavian Journal Of Psychology, 47(2), 109–119. 10.1111/j.1467-9450.2006.00498.x [DOI] [PubMed] [Google Scholar]
  20. Kulkofsky S (2010). The effects of verbal labels and vocabulary skill on memory and suggestibility. Journal of Applied Developmental Psychology, 31(6), 460–466. 10.1016/j.appdev.2010.09.002 [DOI] [Google Scholar]
  21. Lamb ME, Brown DA, Hershkowitz I, Orbach Y, & Esplin PW (2018). Tell me what happened (2nd ed.). [Google Scholar]
  22. Lyon TD (2014). Interviewing children. Annual Review of Law & Social Science, 10(1), 73–89. 10.1146/annurev-lawsocsci-110413-030913 [DOI] [Google Scholar]
  23. Lyon TD, & Henderson HM (2021). Increasing true reports without increasing false reports: Best practice interviewing methods and open-ended wh- questions. American Professional Society on the Abuse of Children Advisor, 33(1), 29–39. https://ssrn.com/abstract=3759908 [Google Scholar]
  24. Lyon TD, McWilliams K, & Williams S (2019). Child witnesses. In (Brewer N & Douglass AB, eds.) Psychological Science and the Law (pp. 157–181). [Google Scholar]
  25. Lyon TD, & Saywitz KJ (1999). Young maltreated children’s competence to take the oath. Applied Developmental Science, 3(1), 16–27. 10.1207/s1532480xads0301_3 [DOI] [Google Scholar]
  26. Newlin C, Steele LC, Chamberlin A, Anderson J, Kenniston J, Russell A, & Vaughan-Eden V (2015). Child forensic interviewing: Best practices. U.S. Department of Justice, Office of Justice Programs, Office of Juvenile Justice and Delinquency Prevention (OJJDP). https://ojjdp.ojp.gov/sites/g/files/xyckuh176/files/pubs/248749.pdf [Google Scholar]
  27. Ornstein PA, Gordon BN, & Larus DM (1992). Children’s memory for a personally experienced event: Implications for testimony. Applied Cognitive Psychology, 6(1), 49–60. 10.1002/acp.2350060103 [DOI] [Google Scholar]
  28. Peterson C, Dowden C, & Tobin J (1999). Interviewing preschoolers: Comparisons of yes/no and wh- questions. Law & Human Behavior, 23(5), 539–555. 10.1023/A:1022396112719 [DOI] [PubMed] [Google Scholar]
  29. Peterson CC, Peterson JL, & Seeto D (1983). Developmental changes in ideas about lying. Child Development, 54(6), 1529–1535. 10.2307/1129816 [DOI] [PubMed] [Google Scholar]
  30. Poole DA, & Lindsay DS (1995). Interviewing preschoolers: Effects of nonsuggestive techniques, parental coaching, and leading questions on reports of nonexperienced events. Journal of Experimental Child Psychology, (60(1), 129–154. 10.1006/jecp.1995.1035 [DOI] [Google Scholar]
  31. Quas J, Stolzenberg SN, & Lyon TD (2018). The effects of promising to tell the truth, the putative confession, and recall and recognition questions on maltreated and nonmaltreated children’s disclosure of a minor transgression. Journal of Experimental Child Psychology, 166(Pt 1), 266–279. 10.1016/j.jecp.2017.08.014 [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Saywitz KJ, & Camparo LB (2014). Evidence-based child forensic interviewing: The developmental narrative elaboration interview. [Google Scholar]
  33. Schneider W (2015). Memory development from early childhood through emerging adulthood. [Google Scholar]
  34. Stevens LM, Henderson HM, & Lamb ME (2021). Linguistically complex recognition prompts in pre-recorded cross-examinations. Behavioral Sciences & the Law, 39(3), 369–382. 10.1002/bsl.2504 [DOI] [PubMed] [Google Scholar]
  35. Stolzenberg SN, McWilliams K, & Lyon TD (2017). The effects of the hypothetical putative confession and negatively valenced yes/no questions on maltreated and nonmaltreated children’s disclosure of a minor transgression. Child Maltreatment, 22(2), 167–173. 10.1177/1077559516673734 [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Stolzenberg SN, McWilliams K, & Lyon TD (2018). Children’s conversational memory regarding a minor transgression and a subsequent interview. Psychology, Public Policy, & Law, 24(3), 379–392. 10.1037/law0000176 [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Sullivan C, St. George S, Stolzenberg SN, Williams S, & Lyon TD (in press). Imprecision about body mechanics when child witnesses are questioned about sexual abuse. Journal of Interpersonal Violence. 10.1177/0886260521997941 [DOI] [PubMed] [Google Scholar]
  38. Talwar V, & Crossman AM (2012). Children’s lies and their detection: Implications for child witness testimony. Developmental Review, 32(4), 337–359. 10.1016/j.dr.2012.06.004 [DOI] [Google Scholar]
  39. Vrieze SI (2012). Model selection and psychological theory: A discussion of the differences between the Akaike information criterion (AIC) and the Bayesian information criterion (BIC). Psychological Methods, 17(2), 228–243. 10.1037/a0027127 [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Walker AG (2013). Handbook on Questioning Children: A Linguistic Perspective (3d. Ed.). American Bar Association. [Google Scholar]
  41. Waterman AH, Blades M, & Spencer C (2001). Interviewing children and adults: the effect of question format on the tendency to speculate. Applied Cognitive Psychology, 15(5), 521–531. 10.1002/acp.741 [DOI] [Google Scholar]
  42. Williams I (1909). Hints to witnesses. Green Bag, 27(12), 620–624. [Google Scholar]
  43. Wolfman M, Brown D, & Jose P (2016). Talking past each other: Interviewer and child verbal exchanges in forensic interviews. Law and Human Behavior, 40(2), 107–117. 10.1037/lhb0000171 [DOI] [PubMed] [Google Scholar]

RESOURCES