Don't miss our trial consultant responses at the end of this article: Katherine James and Robert Galatzer-Levy.

Thousands of children testify each year in the United States (Ceci & de Bruyn, 1999). Children testify both as witnesses and victims in a variety of legal settings, including family court, dependency court, civil matters, and, most conspicuously, in criminal cases in which sexual abuse is alleged (Quas & Sumaroka, 2012). In many of these cases, much turns on the testimony of the child and whether jurors perceive it to be credible. Jurors often use heuristics or cues to evaluate credibility, such as facial expressions, eye contact, and the general demeanor of the child (Regan & Baker, 1998). Indeed, the United States Supreme Court held that children must testify in front of the jury, rather than behind screens or through the use of out-of-court statements, precisely because jurors need to view these cues in order to evaluate credibility (see, for e.g., Coy vs. Iowa, 1988).

Unfortunately, these expectations are not reflective of the actual way in which children testify.  For instance, studies indicate that jurors expect sexually abused children to cry and exhibit negative emotion when testifying about alleged abuse, and jurors tend to disbelieve child witnesses who do not emote in this way (Myers et al., 1999). But research indicates that children commonly do not cry or express negative emotions when describing sexual abuse (Sayfan et al., 2008), and there are a number logical of reasons for their unanimated testimony in general. For instance, children are often interviewed multiple times regarding the incident before testifying in court, or they may simply not have perceived the event as negative. What’s more, the emotion expressed by testifying children could be an artifact of the courtroom experience—i.e., being questioned by unfamiliar and potentially hostile attorneys—and have little to do with the alleged incident itself (Hill & Hill, 1987).

Improving the Quality of Child Witness Testimony

The discordance between what jurors expect and how children do testify could lead to the testimony being unfairly dismissed. As mentioned, the outcome of the case can largely turn on the credibility of the child’s testimony. There are (at least) two theoretical ways to augment the perceived credibility of child witness testimony. First, one could call an expert in developmental psychology to disabuse juror expectations and explain the usual range of emotion expressed by children.  Research on this prospect is not encouraging, as jurors tend to heavily discount this type of expert testimony and revert back to their preconceived expectations (see Kovera et al., 1997). The second prospect is by improving the substance and quality of the testimony itself.

An important aspect of credibility is the extent to which the witness describes his or her reactions to the event in question. According to the Story Model of Juror Decision Making, jurors are more likely to be persuaded by a coherent narrative, which consists of logically and sequentially connected events and the internal responses of the narrator (Pennington & Hastie, 1992).  Internal responses include a description of subjective feelings about the event; thus, it follows that describing subjective feelings about the event could augment credibility.

A small body of literature has examined how children respond to different types of questions. In general, open-ended questions tend to elicit longer and richer responses than close-ended questions, though close-ended questions are sometimes necessary when children are reticent (Lamb et al., 2008). A potential problem of close-ended questions is that they increase the likelihood of children acquiescing to (rather than producing) inaccurate information.  One study found that a particular type of open-ended question, namely “Wh-“ questions such as “what happened?” or “why did you feel that way?”, is likely to elicit more accurate information as well as greater details about the event in question compared to closed-ended questions (Lamb et al., 2008). This is exactly the type of information that is germane to a coherent narrative.

Nearly all of this research has examined forensic interviews of children who are suspected of being sexually abused. It is not clear whether the general finding—that different types of questions affect the rate at which children produce details about the event—would generalize to a trial context. There are major differences between forensic interviews, which tend to occur in private between a single interviewer and child after establishing rapport, and examining child witnesses in court, where numerous adults are congregated and ask questions. The present study examined whether different types of questions increased the production of details by children who testified in actual legal proceedings.

The Study

From January 1997 until November 2001 there were 3,622 felony sexual abuse charges filed in Los Angeles County. 309 of these cases went to trial, of which 82% resulted in a conviction and 17% in an acquittal (the others were ultimately plea-bargained). 218 of these cases had at least one witness under the age of 18 who testified as the victim. From this latter set, 80 cases were randomly selected, yielding a sample of child witnesses who ranged in age from 5–18 with an average age of 12. All of the questions asked of and answers provided by the witnesses were coded. There were 16,495 question/answer turns.

The questions were classified into one of three types: “option-posing” which are questions that can be answered ‘yes’ or ‘no’ (e.g., “Did you feel good?”); “Wh-“ which are questions that contain the stem ‘Wh-‘ (i.e., Who, What, Where, When, Why); and “How” which are questions prefaced with ‘How’ (e.g., “How did you feel?”). Responses were classified according to whether they contained an evaluative response (yes/no), which is defined as any emotional (e.g., “I hated him.”), cognitive (e.g., “I was confused.”), or physical (e.g., “It hurt”) response to the event in question.

Figure 1 depicts the percentage of the various types of questions that were posed to the witnesses. By and large, option-posing questions were the most common, while the least common (asked only 6% of the time) were How questions.


The prosecution asked a majority of the overall questions (62%), and was slightly more likely to ask an option-posing question (56% of all option-posing questions were from the prosecutor). 23% of all the Wh- questions and 34% of all How questions were asked by the defense. Overall, only 3.5% of the answers contained an evaluative response. However, this low percentage depended on the type of question asked, as depicted in Figure 2.[1]


Figure 2. displays the efficiency of the various types of questions in producing evaluative content. Only 1% of option-posing question yielded an evaluative response, compared to 7% for the Wh- questions and 11% for How questions. Thus, the low overall rate of evaluative responses (i.e., 3.5%) can be partially explained by the fact that option-posing questions predominate and option-posing questions are the least productive in eliciting evaluative content. Indeed, How questions were approximately 10 times more likely to elicit evaluative content than option-posing questions.

Bear in mind that this finding exists independent of the age of the witness, which was built into the statistical model.[2] In other words, it is not simply the case that older witnesses were asked more How questions since older witnesses are naturally more articulate. It is also noteworthy that this finding was replicated on a sample forensic interviews in which children were systematically asked the various types of questions, thus limiting the alternative explanation that articulate children were disproportionally asked How questions.

Implications for Practice

As a general matter, children provided few evaluative details while testifying in court. However, when asked an open-ended question, especially a How question, children were considerably more likely to provide evaluative content than when asked an option-posing question. The implication is obvious: attorneys ought to ask more open-ended questions of child witnesses. For the most part, this prescription can be easily implemented and involves a simple reframing of the question. For example, rather than asking, “Do you feel scared when he yells?” attorneys might ask, “How do you feel when he yells?” Consider the following dialog, which was elicited from a 10-year-old child using open-ended questions:

Q: How did you feel when he touched you?

A: Kind of angry at him cause he shouldn’t be doing that and sometimes I thought that he was doing that ’cause I wasn’t his daughter (oh, o.k.) I felt kind of mad, disappointed. ’Cause in front of my mom he always say that he love me really. And on my mind I say that if he loves me why was he doing that to me.

Q: Okay. How did you feel after he touched you?

A: I felt like nasty. Like dirty.

Q: Really. Tell me about that, dirty and nasty.

A: ’Cause he touch, if he touches me, he touch me, right. Then he just leaves and like if like if I didn’t work anymore just leave me like that (uh-huh). And I felt like mad and at the same time felt kind of dirty because he shouldn’t be doing that because I’m just a little girl.


There are several limitations of the reported study. First, the data are from sample of sexual abuse cases in Los Angeles, CA. Generalizing beyond this context (i.e., children testifying as the victim of a sexual offense) requires further study, though the replication with the forensic interview sample is highly encouraging in this regard. Second, it is assumed that providing evaluative content enhances the credibility of the child’s testimony. Although this is ultimately an empirical question that requires further study, there is no reason to believe that furnishing evaluative content would attenuate the credibility of children’s testimony. Finally, one might question the factual accuracy of the evaluative content. Ground truth is typically unknown and perhaps unknowable in many ecologically valid settings, as it was with this sample. But it is worth noting that the same pattern of findings emerged when the sample was restricted to cases that resulted in a conviction.

Final Remarks

Although it would be unethical to cajole child witnesses into emoting on the stand in order to satisfy jurors’ expectations, there is nothing improper about phrasing questions in such a way that is likely to yield valuable and persuasive testimony. The findings clearly indicate that How questions are relatively more productive of evaluative content.

Nicholas Scurich is Assistant Professor, Departments of Psychology & Social Behavior and Criminology, Law & Society at the University of California—Irvine. Professor Scurich earned his Ph.D. in Psychology from the University of Southern California in 2012, as well as a Ph.D. achievement award – a university-wide award for the most exceptional doctoral candidate. From 2010–11 he was a Fellow of the Saks Institute at the USC Gould School of Law. He is an expert on judgment and decision-making, especially within legal settings. He teaches undergraduate and graduate courses on Psychology & Law.


[1]A nested logistic regression indicated that the effect for the type of question is significant χ2 (83, N = 16,495) = 1371.36, p < .001. A complete explanation of the statistical model utilized for this analysis is available from the author upon request.

[2]In addition to the age of the witness, a dummy code for each participant was entered into the model in order to control for the possibility the certain children were highly articulate and thus more likely to have been asked open-ended questions as opposed to close-ended questions. The statistical approach is fully described in Lyon et al. (2012) at p. 450.


Ceci, S.J. & de Bruyn, E. (1993) Child witnesses in court: A growing Dilemma. Children Today, 22(1).

Coy vs. Iowa. (1988). 487 U.S. 1012.

Hill, P. E., & Hill, S. M. (1987). Videotaping children’s testimony: An empirical view. Michigan Law Review, 85,809–833.

Kovera, M., Gresham, A., Borgida, E., Gray, E., & Regan, P. (1997). Does expert psychological testimony inform or influence juror decision mak- ing? A social cognitive analysis. Journal of Applied Psychology, 82,178–191. doi:10.1037/0021–9010.82

Lamb, M. E., Hershkowitz, I. Orbach, Y., & Esplin, P. W. (2008). Tell me what happened: Structured investigative interviews of child victims and witnesses.West Sussex, UK: Wiley.

Lyon, T. D., et al. (2012). “How did you feel?”: Increasing child sexual abuse witnesses’ production of evaluative information. Law and Human Behavior, 36(5), 448–457. doi: 10.1037/h0093986

Myers, J. E., Redlich, A., Goodman, G., Prizmich, L., & Imwinkelreid, E. (1999). Juror’s perceptions of hearsay in child sexual abuse cases. Psychology, Public Policy, & Law, 5,388–419. doi:10.1037/1076- 8971.5.2.388

Pennington, N., & Hastie, R. (1992). Explaining the evidence: Tests of the story model for juror decision making. Journal of Personality and Social Psychology, 62,189–206. doi:10.1037/0022–3514.62.2.18

Quas, J. A., & Sumaroka, M. (2012). Consequences of legal involvement on  child victims of  maltreatment. To appear in M. Lamb, D. LaRooy, C. Katz, & L. Malloy  (Eds.), Children’s Testimony. Cambridge, England: University of Cambridge Press.

Regan, P. C., & Baker, S. J. (1998). The impact of child witness demeanor on perceived credibility and trial outcome in sexual abuse cases. Journal of Family Violence, 13,187–195. doi:10.1023/A:1022845724226

Sayfran, L., Mitchell, E. B., Goodman, G. S., Eisen, M. L., & Qin, J. (2008). Children’s expressed emotions when disclosing maltreatment. Child Abuse & Neglect, 32,1026–1036. doi:10.1016/j.chiabu .2008.03.004

We asked two trial consultants to respond to Dr. Scurich’s article on Child Witnesses. On the following pages, Katherine James and Robert Galatzer-Levy offer their comments.

Katherine James responds:

Katherine James, MFA is a trial consultant based in Culver City, CA. Her specialization is live communication skills. She specializes in making witnesses “not do that anymore and do this instead” in cases and attorneys to be the best they can be in live and virtual workshops. Read more about her company ACT of Communication at the website.

Thank you, Nicholas Scurich, for reminding us that children, nay all people, know that open-ended questions give the best information.

Of course jurors trust the answers to open-ended questions more than to closed ended questions. This is true of adults who are testifying – why should it be different for children who are testifying? Who doesn’t remember being a child and being forced to answer a “yes” or “no” question about something vital and feeling the need to please the adult in question rather than tell the truth.

In my experience, however, many attorneys find themselves feeling out of control with child witnesses. This tends to make attorneys want to ask kids closed ended questions that can only be answered “yes” or “no”.  Just like they do with adult witnesses. This leads to the attorney knowing exactly what he or she wants as testimony from the child witness and the child not only being clueless but “out of control”.

By the way, I find this true of all generations of attorneys with whom I practice – not just the ones who were raised in the “children should be seen and not heard” era. Like I was. I shall never forget the awkwardness of the agreement I made with my mother once when she said, “Now – you aren’t going to notice the giant tumor on Uncle Yalmer’s forehead, are you, Kathy?” I knew I was supposed to say, “No” and of course I did. But damn, once I got to his house and got a look at that giant messy thing on his face that was so hard to do. A jury such as the one that Nicholas Scurich was talking about would have read my body language as that of one coerced young lady, I can tell you that.

Take a young child I worked with – we’ll call her “Sally.” Her mother had been killed by a train. While role playing the direct examination, her attorney started out with the following question, “Sally – do you know who I am?” Sally stared at him quizzically. “See, this is why I hate putting kids on the stand,” he hissed at me over her head. “Sally is wondering why if you are so smart you don’t know who you are,” I hissed back. Solution? Sally talked about a picture she had drawn entitled “My Mommy”. The attorney asked questions filled with “Why?” and “How come?” and “How did that work?” and “How so?” Sally was not only able to talk a blue streak about the picture and everything in it, but the open-ended questions led to lots more information about Sally and her mother. The questions allowed her to laugh with memories, sigh with sadness, and finally to look at the jurors and say, “Gosh, I wish you had known her” with no prompting at all.

Daring to prompt them and then get out of the way of child witnesses and allow them to tell their stories takes effort. It is a million times easier to think of a closed-ended question. It feels really “in control”, especially with children, to ask that “yes or no” question. However, jurors and children alike know that the payoff of the truth as spoken by any witness is much more valuable to everyone.

Robert M. Galatzer-Levy responds:

Robert M. Galatzer-Levy, M.D. is a Clinical Professor of Psychiatry and Behavioral Neuroscience at the University of Chicago, member of the Faculty of the Chicago Psychoanalytic Institute and the author or editor of 5 books and more than 120 research publications. His practice includes child, adolescent and adult psychiatry, providing expert testimony in criminal and civil matters, and litigation consulting.

During the past half-century attitudes toward child witnesses, especially in sexual abuse cases, have swung wildly between extremes – from always believing the child to a belief children cannot differentiate fantasy and reality. Polemical professional studies supporting these positions have given way to efforts to develop systematic methods that differentiate credible and non-credible child testimony (Kuehnle, 2009; Lamb, 2008).  But the bottom line remains. It is often devilishly hard to assess where truth lies and triers of fact have an even harder time doing so than experts. As Scurich observes, juries are likely to focus on the narrative credibility of the child’s testimony, including the consonance between the child’s emotions and the content of what is reported. He shows that “how questions” are more likely to elicit richer and hence more emotionally believable responses than narrower questions.  He cautiously recommends such questions for this purpose.

I question the step from his empirical finding to his recommendation. Attorneys often prefer narrower questions. Such questions are less likely to lead the witness to impeachable elaborations. Notice that in Scurich’s example the witness volunteers potentially impeachable statements. She reports what the defendant said in front of the child’s mother. She states that the defendant leaves after the sexual act. The child says, “He shouldn’t be doing that because I’m just a little girl” which could easily be followed up on cross with questions suggesting that the child had been indoctrinated. People’s memories for facts are poor. Their memories for emotions, much less the causes of those emotions, are worse. While the child has perhaps provided more emotionally compelling testimony, she has also opened up several areas for forceful cross examination and impeachment.

As Scurich suggests the common finding that children’s narratives lack expectable emotional force derives from several sources including testifying in court, rehearsal effects, the child not regarding the behavior in a conventional light, and the child’s wish to please, or to tell the “truth,” i.e., what she believes the adults think is true. If the event was traumatic, the child may exhibit the emotional flattening common in recounting such events.  (See Gabourey Sidibe ’s brilliant performance in Precious.)

Some of these difficulties can be addressed if the jury sees videotapes of the initial interview with the child so jurors can assess how the child’s statements came to light and the child’s testimony is less influenced by the very strange situation of being in a courtroom. Well conducted interviews carefully and neutrally assess issues like the source of the child’s memories. They thus address juror’s well-founded concern that the child has been indoctrinated. The dissonance between the child’s emotion and jurors’ expectation is often addressed. (“How do you feel when you talk about these things?” “I just kind of feel dead inside — like there is nothing in me, like I’m empty.”)

The problem of rehearsal is enormous in children’s testimony. As Loftus (1997) famously demonstrated even mild questioning of a child can induce clear and vivid memories of events that never occurred.  Indeed, it seems very likely that the stiffness of some children’s testimony occurs because although the child recounts actual events, the source for memory is not the event itself, but the child’s previous telling of the story. (In psychotherapy it is not rare for patients to realize that a memory, whether or not it is true, derives not from the event but a previous narration of it.) A video recording of the child’s initial report can be very helpful in addressing these problems. However, it must be kept in mind that the strength of the video is also its weakness since it is likely to demonstrate any problems associated with the interview.

In jurisdictions where videos cannot be admitted directly into evidence, they may often be admissible as part of the basis of an expert witness opinion. This brings us to the question of experts, who are, as Scurich indicates, often ineffective.  In addition to problems common to all experts, like talking above the jury’s heads and using jargon rather than vivid specific language, experts on child abuse are all too often advocates for one of the extreme positions (children always tell the truth; children are never reliable) mentioned at the beginning of this discussion. Such experts are sometimes effective because of their passionate advocacy but their advocacy also makes them less credible. Worse, well prepared cross examination can often demonstrate that the testimony does not live up to any reasonable standard.


Kuehnle, K & Connel, M. (eds) 2009 The Evaluation of Child Sexual Abuse Allegations. New York: Wiley.

Lamb, M. E., Hershkowitz, I. Orbach, Y., & Esplin, P. W. (2008). Tell me what happened: Structured investigative interviews of child victims and witnesses.West Sussex, UK: Wiley.

Loftus, E. (1997) Creating False memories. Scientific American  277: 70–75.

Photos by lobo235 on Flickr. Graphic design by Brian Patterson of Barnes & Roberts.