delivered the opinion of the court.
Specialist Arturo Cano entered pleas of not guilty to one specification of sodomy with a child under twelve, three specifications of indecent acts with a child under sixteen, and one specification of indecent liberties with a child under sixteen in violation of Articles 125 and 134 of the Uniform Code of Military Justice (UCMJ), 10 U.S.C. §§ 925, 934 (2000). He was tried and convicted by a military judge sitting as a general court-martial and sentenced to a dishonorable discharge, confinement for seventeen years, forfeiture of all pay and allowances, and a reduction in grade to E-l. The convening authority approved the sentence and the findings and sentence were subsequently affirmed by the Army Court of Criminal Ap*75peals. United States v. Cano, No. 20010086 (A.Ct.Crim.App. Feb. 4, 2004) (unpublished).
“Where an appellant demonstrates that the Government failed to disclose discoverable evidence in response to a specific request ... the appellant will be entitled to relief unless the Government can show that nondisclosure was harmless beyond a reasonable doubt.” United States v. Roberts, 59 M.J. 323, 327 (C.A.A.F.2004) (citing United States v. Hart, 29 M.J. 407, 410 (C.M.A.1990)). During discovery, Cano specifically requested that the Government produce the clinical psychologist’s medical records compiled during her therapy sessions with the victim. After reviewing the materials in camera, the military judge released a portion of the material and sealed the remaining documents. The Court of Criminal Appeals found that the military judge erred in withholding the materials but held that Cano was not prejudiced by the error. We granted review to determine whether the Court of Criminal Appeals erred in finding that Cano was not prejudiced by the withholding of these documents.1 We hold that the nondisclosure of these materials was harmless beyond a reasonable doubt and therefore affirm the decision of the Court of Criminal Appeals.
BACKGROUND
On August 9, 2000 Cano’s eleven year-old stepdaughter, DH, had a sleepover with two of her friends at the Cano home. The girls stayed up late and around midnight Cano came into DH’s room and asked the girls if they wanted to take some pictures with his computer. He took DH’s two friends, one at a time, into his computer room where he removed their clothes, touched each of them on their breasts and one of them on her vaginal area, made one of the girls touch his penis, and took pictures of each of them. After the girls went back to DH’s room they told DH what happened. DH tried to tell her mother but Cano prevented her from doing so.
One of the girls lived nearby, so the two girls left Cano’s house and went to the girl’s home where they told her mother what had happened. The incident was reported to Army Criminal Investigation Command (CID), which interviewed DH as part of its investigation. During that interview DH gave a sworn statement in which she alleged that Cano had sexually abused her for six years.
Some time after DH made this statement she met with Dr. DeeAnn Lau, a clinical psychologist from the on-post medical facility. DH was treated by Dr. Lau from September 2000 through December 2000. At some point between the August 9 incident and Cano’s trial, Cano’s wife (DH’s mother) told Dr. Lau that DH had recanted her story about having been abused by Cano. Mrs. Cano also made other statements regarding the possibility that DH was lying.
At trial DH testified regarding various incidents in which Cano had touched her, had forced her to rub his penis, had forced her to place his penis in her mouth, and had inserted his penis into her vagina. DH’s mother testified for the defense. She said that on the evening in question she had been in bed with Cano and had been awake until approximately 1 a.m. On cross-examination the Government pointed out that this was inconsistent with her statement on the night of the incident in which she said that she had fallen asleep at midnight. Mrs. Cano said that she had no explanation for the discrepancy and was also unaware that the pictures taken of the girls that had been found on Cano’s computer were time-stamped at approximately 12:30 a.m. She further testified that after she and DH watched a movie about a man who was falsely convicted for something he had not done, DH told her that “she was lying” because of the way Cano treated her and that she “said that stuff’ because Cano spanked her.
*76Prior to trial the defense made a specific discovery request for the collection of notes taken by Dr. Lau during her therapy sessions with DH between August and December 2000. The military judge rejected the Government’s contentions that the notes were privileged and reviewed the materials in camera. Following his review, the military judge released that portion of the materials he deemed relevant to the case and sealed the remainder. The Army Court of Criminal Appeals found that although the military judge erred in failing to release all of the materials, that error was not prejudicial to Cano.
In evaluating the impact of the requested evidence, the Army court tested for materiality on the basis of whether the evidence “might have affected the outcome of the trial.” The Army court found that the withheld evidence would have allowed for some minor impeachment of DH when she testified and “would have provided data relevant to the defense’s concern that DH’s testimony was of a memory created or implanted during the therapy process.” However, after evaluating DH’s testimony the Army court found that overall she was “direct,” “clear and forthcoming, largely consistent with her” earlier signed statement, and “credible.” It also noted that her testimony was “subject to reasonable scrutiny by the defense on cross-examination.” The court concluded that “the undisclosed evidence would not, in fact, have made any difference in the outcome of this case.”
Before this court Cano argues that the Army court erred in concluding that he was not prejudiced by the military judge’s erroneous decision to deny him the opportunity to review the withheld documents. He argues that the lower court applied the wrong standard for evaluating prejudice and also erred in failing to consider the impeachment value of the withheld evidence. Cano argues that the withheld notes demonstrate inconsistencies in DH’s account of what allegedly transpired between DH and Cano and support the possibility that Dr. Lau’s counseling techniques influenced DH’s account of events. Cano asserts that denying him access to these records undermined the defense’s ability to prepare for trial effectively, and more specifically, to prepare its cross-examination of DH.
The Government responds by arguing that the military judge’s error in failing to disclose this evidence was harmless beyond a reasonable doubt and therefore did not prejudice Cano. It contends that the evidence presented against Cano at trial was overwhelming, noting the strength of DH’s testimony and the defense’s failures in its attempts to impeach her credibility and her memory. Additionally, the Government argues that the evidence in question is not evidence of inconsistencies or coaching, but rather shows that DH was consistent in her statements regarding the abuse she suffered at the hands of her stepfather. It concludes that the undisclosed evidence in this case was of minimal evidentiary value and would not have affected the fact-finder’s decision-making process.
DISCUSSION
The Army Court’s Prejudice Standard
In Roberts this court clarified that “[wjhere an appellant demonstrates that the Government failed to disclose discoverable evidence in response to a specific request ... the appellant will be entitled to relief unless the Government can show that nondisclosure was harmless beyond a reasonable doubt.” 59 M.J. at 327 (citing United States v. Hart, 29 M.J. 407, 410 (C.M.A.1990)). The Army court’s opinion in this case, issued prior to this clarification, concluded that “the undisclosed evidence would not, in fact, have made any difference in the outcome of this case[,]” and therefore applied a lesser standard than required by Roberts. As we review issues of prejudice from erroneous evidentiary rulings de novo, this court can apply the correct “harmless beyond a reasonable doubt” standard in our review. See United States v. Diaz, 45 M.J. 494, 496 (C.A.A.F.1997).
Inconsistencies in DH’s Testimony
We first turn to Cano’s claim that Dr. Lau’s notes could have been used to show inconsistencies in DH’s testimony. The most *77obvious difference between DH’s earlier statement and her testimony at trial was her statement at trial that Cano had penetrated her, while her earlier statement suggested that he had only touched her with his penis. Dr. Lau’s notes do contain a statement noting that DH said Cano had penetrated her, and therefore DH’s assertions regarding penetration would have been revealed to the defense prior to trial had the materials been turned over.
Cano’s defense counsel became aware of this inconsistency when DH testified at trial and properly impeached her during cross-examination. By cross-examining on that point, the defense counsel demonstrated to the military judge that DH’s earlier statement was not as complete as her trial testimony. There was little more that he could have done even if he had received the notes at an earlier date.
Dr. Lau’s notes also indicated that during treatment, DH made inconsistent statements concerning the timing and location of incidents of abuse. For example, the notes stated that DH “denied having been abused since she was 10 y.o.” Cano points out that DH was, in fact, eleven years and five months old at the time she had earlier claimed she was last molested. Similarly, Cano argues that the evidence shows an inconsistency with regard to the location in which DH claimed incidents of abuse took place. There is a notation in the therapy notes that in talking to Dr. Lau, DH “recalled that the abuse occurred ‘downstairs in the computer room,’” but at trial DH testified that the abuse took place in the bathroom and in her parent’s bedroom.
We agree with the defense that these are inconsistencies that could have been used by the defense at trial. However, it was obvious throughout DH’s testimony, both on direct and cross-examination, that her sense of the timing and locations of various instances of abuse was not always entirely clear. Defense counsel himself showed at trial that DH’s understanding of time was not exact. He began his cross-examination by asking DH how many days were in a week, to which she replied “like 5 days,” and then asked how many hours in a day, to which she answered “I think 26.”
Inconsistencies such as these are not uncommon when child abuse victims testify:
[T]he evidence ... is underscored by the fact that the persuasive testimony is from a child, from whom gathering more exact details as to when the sexual conduct precisely began is an unreasonable expectation and a formidable hurdle. Any person who suffers from some type of traumatic experience, adult or child, may have difficulty relating that experience in a chronological, coherent and organized manner. See Kermit V. Lipez, The Child Witness in Sexual Abuse Cases in Maine: Presentation, Impeachment, and Controversy, 42 Me. L.Rev. 283, 345 (1990).
Paramore v. Filion, 293 F.Supp.2d 285, 292 (S.D.N.Y.2003).
While the additional information from Dr. Lau’s notes could have been used to demonstrate that DH did suffer from some confusion, that fact was already obvious from her testimony at trial. In United States v. Santos, 59 M.J. 317, 322 (C.A.A.F.2004), we found that withheld evidence might have been used to impeach an important Government witness, but concluded that the error in withholding it was harmless beyond a reasonable doubt because it was “largely cumulative of other information available to Appellant” at trial, was vague, and concerned a collateral issue. Conversely, in United States v. Jackson, 59 M.J. 330, 335-36 (C.A.A.F.2004), we found that erroneously withheld evidence was “critical on a pivotal issue on the ease” and that because it was so significant the error in withholding it was not harmless beyond a reasonable doubt.
As found by the Army court, DH’s “testimony was reasonably direct, with a minimum of leading. She was mainly clear and forthcoming, and largely consistent with her signed statement of 10 August 2000 to the CID, made on the morning she reported the abuse.” At trial the defense was able to explore the major inconsistency in DH’s testimony as to whether Cano had penetrated her with his penis or simply touched her. The defense was able to show that DH was *78confused concerning the timing and location of various encounters as well as raise the possibility that DH was lying through the testimony of her mother.
Unlike the evidence in Jackson, the notes in question here would not have been “critical on a pivotal issue in the case” because the defense already had the opportunity to attack DH’s credibility and the notes would not have provided any new ammunition with which to do so. 59 M.J. at 335-36. We find that any inconsistencies revealed in the withheld evidence in this case are harmless beyond a reasonable doubt because they are cumulative of other evidence available at trial, easily explained based on DH’s age and maturity, and are not significant in relation to DH’s overall testimony.
Suggestive Therapy
We next turn to Cano’s argument that the withheld evidence could have been used to demonstrate that DH’s testimony was the result of coaching by Dr. Lau in their counseling sessions. The record reveals that DH’s testimony concerning several of the more serious allegations of abuse at trial was consistent with the statement she gave to CID investigators immediately following the August 9 incident. That statement was made well before DH had met Dr. Lau or started her therapy. In the August statement she said that Cano had touched her “chest” and “private area” with his hand, and that he had her put her hands around “the part he goes to the bathroom with and ... shake it.” She also said he would “have me put it in my mouth sometimes.” At trial, she testified that Cano: touched her chest; put his finger in her vagina; put his “front bathroom part” in her mouth; and put his “front bathroom part” inside her; and that she had to rub Vaseline on his “front bathroom part” so “it can be wet and moisty.” While DH’s trial testimony was more detailed than the earlier statement, the main allegations in the earlier statement were largely consistent with her testimony at trial.
We disagree with Cano’s assertion that the withheld evidence showed Dr. Lau to be “more of a zealot who was prone to overreaching in interviewing an alleged child sex abuse victim than she was a therapist whose goal it was to simply treat a child sex abuse victim.” It has been noted in the context of rape counseling that the role of a therapist is:
not [to] probe inconsistencies in their clients’ description of the facts of the incident [or to] conduct independent investigations to determine whether other evidence corroborates or contradicts their clients’ renditions. Because their function is to help their clients deal with the trauma they are experiencing, the historical accuracy of the clients’ descriptions of the details of the traumatizing events is not vital in their task.
People v. Bledsoe, 36 Cal.3d 236, 203 Cal. Rptr. 450, 681 P.2d 291, 300 (1984).
There is a good deal of scholarly debate in the area of child suggestibility and its effect on the reliability of the testimony of a child victim.2 However, scholars agree that the danger of false testimony from a child is greater when the child is subjected to highly suggestive interviewing techniques such as “closed” (yes/no) questions and “multiple interviews with multiple interviewers.”3
Dr. Lau was the sole therapist who treated DH during this period and her notes make it clear that she was acting in accordance with her role as a counselor. She does indicate that DH required “prompting” in telling her story. However, the notes surrounding this statement reflect that the prompting came in the form of open-ended questions because the quoted responses by DH are in the form of factual descriptions of events rather than yes/no answers. Dr. Lau’s therapy questions are consistent with her role in counseling DH, a victim of trauma, and do not establish that Dr. Lau was a “zealot” or that she was *79coaching DH. Because we find that the notes do not provide evidence of suggestive questioning or coaching, we conclude that withholding the material was harmless beyond a reasonable doubt.
DECISION
The decision of the United States Army Court of Criminal Appeals is affirmed.
. We granted review of the following issue:
WHETHER THE ARMY COURT OF CRIMINAL APPEALS ERRED WHEN IT DETERMINED THAT THE MILITARY JUDGE'S ERROR IN NOT DISCLOSING MENTAL HEALTH RECORDS OF A VICTIM DID NOT MATERIALLY PREJUDICE APPELLANT. SEE UNITED STATES V. ROBERTS, 59 M.J. 323, 327 (C.A.A.F.2004).
. See Thomas D. Lyon, The New Wave of Child Suggestibility Research: A Critique, 84 Cornell L.Rev. 1004 (1999); Stephen J. Ceci and Richard D. Friedman, The Suggestibility of Children: Scientific Research and Legal Implications, 86 Cornell L.Rev. 33 (2000); David A. Martindale, On the Importance of Suggestibility Research in Assessing the Credibility of Children’s Testimony, 30 Ct. Rev. 8 (2001).
. Ceci, supra, at 86; Lyon, supra, at 1070-72.