STEPHEN H. ANDERSON, Circuit Judge.
Mel Lambert Velarde appeals his conviction by a jury on one count of aggravated sexual abuse of a child, in violation of 18 U.S.C. § 2241(c), an offense committed within Indian country under 18 U.S.C. § 1153. He was sentenced to 135 months imprisonment. Because we conclude that the district court erred in permitting certain expert testimony, and that error was not harmless, we reverse and remand for a new trial.
BACKGROUND
Mr. Velarde, a member of the Jicarilla Apache Indian tribe, ran his family’s large ranch near Dulce, New Mexico. In July 1997, he entered into a relationship with another Jicarilla tribe member, Angelita Veneno. Ms. Veneno lived with her three children, Jordan, Shane and L., in a trailer home in Dulce, which is within the Jicarilla Apache Indian Reservation.
On some weekends, Mr. Velarde stayed overnight at the trailer home. L., then 8 years old, was accustomed to sleeping with her mother in her mother’s bed; When Mr. Velarde spent the night, she had to share her brothers’ bedroom. There was testimony that she would get upset when Mr. Ve-larde visited, because she could not sleep with her mother. In fact, L. testified it made her “mad” when she could not sleep with her mother. R. Vol. Ill at 53.
Ms. Veneno testified that, until L. was five years old, L.’s father, Levi Vigil, who is also Shane’s and Jordan’s father, lived in the trailer home with them. He was physically and verbally abusive towards' Ms. Veneno, and L. witnessed at least one instance of physical abuse and saw bruises on her mother on a number of occasions. After he moved out of the trailer home, Mr. Vigil had little interaction with his children and apparently refused to acknowledge them. Ms. Veneno’s father also intermittently lived at the trailer home. There was testimony that he had a serious alcohol abuse problem.
Mr. Velarde stayed at Ms. Veneno’s trailer on the night of February 1, 1998. Mr. Velarde testified that, when he awoke in the middle of the night to use the bathroom, he noticed that the dome light was on in Ms. Veneno’s car parked outside. As he usually did when he stayed at Ms. Veneno’s trailer, he used Shane and Jordan’s bathroom down the hall. Ms. Vene-no testified that she was awake when Mr. Velarde left to use the bathroom and awoke when he returned.
Id.
at 80. Mr. Velarde testified that nothing unusual happened — he simply used the bathroom, noticed the car light on, and returned to bed. He testified he was gone from the bedroom “[c]lose to five minutes.” R. Vol. IV at 43.
Approximately a week later, L. told her mother a- very different account of the night of February 1. She told her mother that, on that night, she was sleeping in Jordan’s bed, on the top bunk of her brothers’ bunk bed, as Shane slept below. She testified that she woke up to discover Mr. Velarde pulling her off the bunk bed. She further testified that he took her out into the hall, pulled her pajama shorts down, placed his hand over her mouth, and “put his private part in [her] private part.” R. Vol. Ill at 40. She testified that it hurt a little. She also stated that she kicked at him. L. further stated that after Mr. Ve-larde got off her, he told her not to tell anyone. She testified that she saw the light on in her mother’s car parked outside the trailer home.
At Mr. Velarde’s trial, Ms. Veneno testified that L.’s behavior changed after February 1:
[F]or one week she just stayed in her room, she was just quiet, she didn’t come out until it was time to go to bed. And then after she told me, she started acting up, like always, crying. She was wetting the bed. She was getting nightmares. She was — she found a bottle that she started nursing again, and that’s about how — how she was.
Id.
at 81.
Upon hearing L.’s accusation of sexual abuse, Ms. Veneno called the police, who came to the trailer home and took L.’s statement. The next day, February 9, Ms. Veneno took L. to the San Juan General Hospital in Farmington, New Mexico. Naida Ayers, a nurse, testified that she examined L., took her medical history and asked L. what had happened. She testi
fied that L. told her that Mr. Velarde came into the bedroom, pulled her out into the hall, and “touched her private parts” or “put his private part into her.”
Id.
at 114. Ms. Ayers testified that her physical examination of L. revealed no bruising or tears or other evidence of physical injury. Ms. Ayers testified that L. was cooperative and not frightened during the examination. Dr. Allen Hurt similarly examined L., also found no physical evidence of any injury, and also testified that L. was cooperative and acted normal during the examination. Dr. Hurt testified that the examination neither proved nor disproved that sexual abuse had occurred.
On February 20, L. was examined by Dr. Renee Ornelas, a pediatrician at the University of New Mexico. Dr. Ornelas took a patient history, recorded any unusual symptoms, determined what further tests were needed, and conducted a physical examination. Dr. Ornelas testified she used a colposcope, which enabled her to magnify L.’s genital area. After examining L., Dr. Ornelas agreed with Dr. Hurt that the results of the physical examination were normal. She testified that L. had a “normal” hymen. R. Vol. V at 160.
Dr. Charlene McGiver, a clinical psychologist, interviewed L. and her mother, separately and together, on October 12 and November 2. Dr. McGiver testified that her interviews revealed that L. began having nightmares, experienced episodes of bed-wetting, feared men, and had angry outbursts after the alleged sexual abuse occurred.
A federal grand jury returned an indictment against Mr. Velarde. He pled not guilty. Prior to trial, Mr. Velarde, represented by an appointed public defender, filed a motion to exclude the proposed testimony of Dr. Ornelas that “a child’s statement as part of her medical history is consistent with child abuse,” and asserted that the government had violated Fed. R.Crim.P. 16 by failing to disclose the bases and reasons for her opinion. R. Vol. I, Tab 13 at 2. The court denied the motion.
Then, represented by new retained counsel, Mr. Velarde filed another motion in limine seeking to preclude or limit Dr. Ornelas’s testimony, challenging her proposed testimony that L.’s statements and behavior are consistent with her having been sexually abused. He argued that such testimony was not admissible expert testimony under
Daubert v. Merrell Dow Pharmaceuticals,
509 U.S. 579, 113 S.Ct. 2786, 125 L.Ed.2d 469 (1993),
and that such testimony would amount to impermissible vouching for the child’s veracity. He requested a hearing under
Daubert.
He also challenged Dr. Ornelas’s proffered testimony about the behavioral characteristics of child abuse victims as grounded in psychology and therefore outside of her area of expertise, and asserted that some of her proffered testimony was irrelevant. The district court denied this second motion.
Mr. Velarde then filed a motion to preclude Dr. McGiver’s testimony and sought a
Daubert
hearing to determine the reliability of the proposed testimony. The court denied the motion and determined that an evidentiary hearing was unnecessary.
Approximately a month before trial, the government gave notice that it intended to present testimony from Trudy Harrison, Mr. Velarde’s niece, who was prepared to testify that, twenty years before, when she was eight or nine years old, Mr. Velarde had sexually assaulted her. Mr. Velarde filed a motion in limine to preclude the presentation of this evidence, or to eontin- ' ue the case, asserting that the late disclosure of this witness prejudiced him. He also argued that the trial court must conduct a Fed.R.Evid: 403 balancing test to determine if the probative value of the evidence outweighs any prejudicial effect. Without explicitly conducting any Rule 403 balancing, the court held that the govern
ment had given adequate notice of its intent to call Ms. Harrison, and ruled that the government could call Ms. Harrison as part of its case-in-chief. Mr. Velarde made numerous additional pre-trial motions relating to the government’s proposed expert witnesses, all of which were denied. The court also denied his motions for an independent evaluation of L. and for disclosure of her psychological treatment records.
The case proceeded to trial. On the first day of trial, March 23, 1999, the Supreme Court issued its decision in
Kumho Tire Co. v. Carmichael,
526 U.S. 137, 119 S.Ct. 1167, 143 L.Ed.2d 238 (1999), holding that trial courts must conduct some kind of gatekeeping reliability determination under
Daubert
with respect to all expert testimony, not just to testimony based upon a particular scientific methodology. Mr. Velarde renewed his request for a
Daubert
hearing with respect to the proposed testimony of Dr. Ornelas and Dr. McGiver, and referred the court to
Kum-ho. See
R. Vol. V at 267. The court subsequently denied his renewed request, stating, “Well, I’m not going to hold a
Daubert
hearing. I’ve had this testimony before in trials, and it’s not new and novel.... ”
Id.
at 342. Accordingly, both Dr. Ornelas and Dr. McGiver testified as government experts. Ms. Harrison also testified that Mr. Velarde had sexually abused her twenty years before, when she was eight or nine years old. Mr. Velarde took the stand and unequivocally denied ever sexually abusing or touching either L. or Ms. Harrison. The jury found Mr. Ve-larde guilty.
He appeals his conviction, arguing: (1) the court erred in admitting the testimony of governmental experts Drs. Ornelas and McGiver and in failing to first determine the reliability and relevancy of that testimony; (2) the court erred in admitting Ms. Harrison’s testimony under Rule 414; (3) the admission of that Rule 414 evidence violated Mr. Velarde’s right to due process, equal protection and a fair trial; and (4) the court erred in failing to instruct the jury on the lesser included offense of abusive sexual contact. Because we find that the district court erred in failing to conduct, on the record, any kind of reliability determination with respect to the government’s expert witnesses, we reverse and remand.
DISCUSSION
I. Expert Testimony
The trial court’s admission of expert testimony, over a timely objection, is reviewed for an abuse of discretion.
Kumho Tire Co.,
526 U.S. at 152, 119 S.Ct. 1167;
United States v. Charley,
189 F.3d 1251, 1261 (10th Cir.1999),
cert, denied,
- U.S. -, 120 S.Ct. 842, 145 L.Ed.2d 707 (2000). Fed.R.Evid. 702, which governs the admission of expert testimony, provides as follows:
If scientific, technical, or other specialized knowledge will assist the trier of fact to understand the evidence or to determine a fact in issue, a witness qualified as an expert by knowledge, skill, experience, training, or education, may testify thereto in the form of an opinion or otherwise.
Rule 702 “imposes a special gatekeeping obligation on the trial judge to ensure that an opinion offered by an expert is reliable.”
Charley,
189 F.3d at 1266. As the Supreme Court made clear in
Kumho,
“where [expert] testimony’s factual basis, data, principles, methods, or their application are called sufficiently into question ... the trial judge must determine whether the testimony has ‘a reliable basis in the knowledge and experience of [the relevant] discipline.’ ”
Kumho,
526 U.S. at 149, 119 S.Ct. 1167 (quoting
DaubeH,
509 U.S. at 592, 113 S.Ct. 2786). The trial judge has broad discretion “to determine reliability in light of the particular facts and circumstances of the particular case.”
Id.
at 158, 113 S.Ct. 2786;
see also Charley,
189 F.3d at 1266. The trial judge enjoys equally broad discretion in both deciding how to assess an expert’s reliability, including what procedures to utilize in making that
assessment, as well as in making the ultimate determination of reliability.
Kumho,
526 U.S. at 152, 119 S.Ct. 1167;
see also Hynes v. Energy West, Inc.,
211 F.3d 1193, 1203-05 (10th Cir.2000). We review either exercise of discretion for abuse.
Mr. Velarde argues the trial court abandoned its gatekeeping function, as required by
Kumho
and
Daubert,
when it refused to inquire into the reliability of the proposed expert testimony of Dr. Ornelas and Dr. McGiver. He also argues that the testimony of both experts was, for a variety of reasons, unreliable.
A. Dr. Ornelas
Dr. Ornelas began her testimony by recounting her usual procedures in examining a child where sexual abuse has been alleged. When asked what information she typically needed to obtain about an alleged perpetrator of sexual abuse, she responded in part, “We have to make sure that the child is safe. So just who they are and is the child away from that person.” R. Vol. V at 147. Dr. Ornelas then testified that L. had a “normal” physical exam, which was, she opined, “consistent with what [L.] said happened.”
Id.
at 155. Dr. Ornelas testified that a normal exam is “the most common physical findings for a child who has been sexually abused.”
Id.
at 149. This is so, Dr. Ornelas opined, because “the type of contact that most commonly occurs between adults and children that’s sexual is oral kinds of contact, touching, and what’s called labial coitus.”
Id.
Dr. Ornelas also testified that “L.’s behavior of withdrawing from her family, staying inside of her room, not being communicative, not being her regular bubbly, running-around kind of self, and ... waking up at night and touching her mother to make sure that her mother was there in bed with her, ... having sleep disturbances and some behavioral changes,”
id.
at 156, was “consistent with child sexual abuse,”
id.
at 155. Finally, she testified as follows:
Q. Doctor, if you had those symptoms or some of those symptoms and a report by a child of sexual abuse, are you comfortable forming a diagnosis of child sexual abuse in such a case?
A. I would base that diagnosis on the child’s statements about what had happened to them.
Id.
at 187.
Mr. Velarde argues the trial court erred in making no reliability findings at all about Dr. Ornelas’s testimony, and he argues that, for numerous reasons, the testimony was in fact unreliable.
While we recognize that the trial court is accorded great latitude in determining how to make
Daubert
reliability findings before admitting expert testimony,
Kumho
and
Daubert
make it clear that the court must, on the record, make
some
kind of reliability determination. “[Tjrialcourt discretion in choosing the manner of testing expert reliability [] is not discretion to abandon the gatekeeping function.”
Kumho,
526 U.S. at 158-59, 119 S.Ct. 1167 (Scalia, J., concurring).
The record in this case reveals no such reliability determination. Even when
Kumho
was specifically called to the court’s attention, thereby removing any question that the expert testimony to be offered by Dr. Ornelas was subject to
Daubert /Kumho
reliability standards, the court made no reliability findings. Rather, the court seemed to assume that Dr. Ornelas’s proffered testimony fell within the category of testimony in “ordinary” cases where courts may “avoid unnecessary ‘reliability’ proceedings ... where the reliability of an expert’s methods is properly taken for granted.”
Kumho,
526 U.S. at 152,119 S.Ct. 1167. However, the court gave no indication why this case could be viewed' as such an “ordinary” case, or why Dr. Orne-
las’s methods could be “properly taken for granted,”
id.,
except for the court’s remark that “I’ve had this testimony before in trials, and it’s not new and novel,” R. Vol. IV at 75.
In several prior cases, we have reviewed testimony given by Dr. Ornelas.
See United States v. Komh,
No. 99-2138, 2000 WL 342252, at *2-3 (10th Cir., April 3, 2000);
Charley,
189 F.3d at 1265-68;
United States v. McHorse,
179 F.3d 889, 895 (10th Cir.),
cert. denied,
— U.S. -, 120 S.Ct. 358, 145 L.Ed.2d 280 (1999);
United States v. Pacheco,
154 F.3d 1236, 1238 (10th Cir.1998),
cert. denied,
525 U.S. 1112, 119 S.Ct. 886, 142 L.Ed.2d 785 (1999). In
Charley,
a case also involving allegations of sexual abuse of young girls, Dr. Ornelas gave testimony, similar to that given in this case, that the victims had normal physical examinations, and exhibited behaviors consistent with sexual abuse.
Moreover, as in this case, she expressed her belief that sexual abuse had in fact occurred because the victim alleged that such abuse had occurred. We stated as follows in
Charley:
Regardless of whether Dr. Ornelas based her conclusion on the girls’ allegation of sexual abuse, or on the girls’ history of atypical medical symptoms, her testimony is problematic. If the conclusion was based on the girls’ medical symptoms, questions surrounding the reliability of such a conclusion arise. On the other hand, if the conclusion was based on the girls’ allegations, Dr. Orne-las was merely vouching for the credibility of the child complainants.
Charley,
189 F.3d at 1266.
We held in
Charley
that the court erred in making no reliability determination “with respect to Dr. Ornelas’s unconditional opinion that [the alleged victims] were sexually abused.”
Id.
at 1266-67. We did not specify that any particular type of reliability determination must be made. However, some such reliability determination must be apparent from the record: the trial court’s “evidentiary decisions do not warrant reversal if it determined,
in some apparent manner,
that the expert testimony admitted was reliable.”
Id.
at 1261 n. 11 (emphasis added).
Here, even with
Kumho
squarely before it, the district court made no reliability determination with respect to Dr. Orne-las’s proposed testimony. We conclude that, having failed to do so, the court abused its discretion when it admitted that testimony. We need not reach the question of whether, assuming a proper reliability determination preceded Dr. Orne-las’s testimony, we would nonetheless find it error to admit particular portions of her testimony.
B. Dr. McGiver
Mr. Velarde challenges three parts of Dr. McGiver’s testimony as unreliable and inadmissible: (1) Dr. McGiver’s testimony that certain of L.’s behaviors (having nightmares, bed-wetting and angry outbursts) were “consistent with” having been sexually abused; (2) Dr. McGiver’s testimony that she recommended L. receive therapy; and (3) Dr. McGiver’s testimony that she found no evidence that L. “was subject to either lying or over exaggerated fantasizing in her life.” R. Vol. V at 202. Mr. Velarde sought and was denied a
Danbett
reliability hearing on Dr. McGiver’s proposed testimony. The record reveals no reliability determination with respect to that testimony. As with Dr. Ornelas, we conclude that the district court abused its discretion in admitting Dr. McGiver’s testimony after failing to make any kind of reliability determination with respect thereto. We need not address whether, had a proper reliability determination been made, any particular parts of Dr. McGiver’s testimony would have been impermissible for any reason. We note, however, that, as we stated in
Charley,
testimony which essentially simply vouches for the truthfulness of another witness is impermissible.
See Charley,
189 F.3d at 1267.
C. Harmless Error
Since we have concluded that the district court erred in permitting the testimony of Dr. Ornelas and Dr. McGiver without the requisite reliability determinations, we must decide whether that error was harmless. “A non-constitutional error, such as a decision whether to admit or exclude evidence, is considered harmless ‘unless a substantial right of [a] party is affected.’ ”
Charley,
189 F.3d at 1270 (quoting Fed.R.Evid. 103(a)). We have defined an error affecting a substantial right of a party as “an error which had a ‘ “substantial influence” on the outcome or [which] leaves one in “grave doubt” as to whether it had such an effect.’ ”
Id.
at 1270 (quoting
United States v. Rivera,
900 F.2d 1462, 1469 (10th Cir.1990) (en banc) (quoting
Kotteakos v. United States,
328 U.S. 750, 765, 66 S.Ct. 1239, 90 L.Ed. 1557 (1946))). We review the record as a whole. The government bears the burden of proving that an error is harmless.
Id.
After reviewing the entire record in this case, we conclude that the error in the admission of Dr. Ornelas’s and Dr. McGiver’s testimony was not harmless. In asserting harmlessness, the government relies upon
Charley,
where we found the error in the admission of certain expert testimony was harmless. This case is significantly different from
Charley,
both in the quality and quantity of the admissible evidence presented, and, more generally, in the entire presentation of the case.
In
Charley,
two girls, both older than L., testified as to numerous instances of sexual abuse over an extended period of time; detailed expert testimony from the girls’ treating physician was properly admitted, as was the lengthy medical history of one of the girls; at the time of the abuse, the defendant was on supervised release following a conviction for sexual abuse of a child; defendant’s testimony was “flatly
contradicted” in three respects by other witnesses. By contrast, in this case, a young girl testified to a single instance of alleged sexual abuse; there was relatively little other evidence, besides the testimony of Drs. Ornelas and McGiver, suggesting that the abuse had occurred; Mr. Ve-larde’s credibility was not obviously suspect; aside from Ms. Harrison’s accusation of abuse occurring twenty years earlier, Mr. Velarde has no prior history of sexual abuse. In sum, we conclude that the error in admitting the testimony of Dr. Ornelas and Dr. McGiver did indeed substantially affect the trial’s outcome. It was not harmless. We therefore reverse and remand for a new trial.
II. Rule 414 Evidence
As indicated, Mr. Velarde’s niece, Ms. Harrison, testified that Mr. Velarde had sexually abused her twenty years earlier, when she was eight or nine years old. The court permitted this testimony under Fed. R.Evid. 414, which provides in pertinent part:
(a) In a criminal case in which the defendant is accused of an offense of child molestation, evidence of the defendant’s commission of another offense or offenses of child molestation is admissible, and may be considered for its bearing on any matter to which it is relevant.
Fed.R.Evid. 414(a).
Mr. Velarde challenges the admission of Ms. Harrison’s testimony on two grounds: (1) the court failed to balance the relevance of that testimony against its potentially prejudicial effect, as required by Fed.R.Evid. 403; and (2) the admission of that testimony under Rule 414 violated his constitutional rights to due process, equal protection and a fair trial.
Mr. Velarde acknowledges that a panel of this court has recently upheld the constitutionality of Rule 414 and the admission of evidence of prior offenses of child molestation pursuant thereto.
See United States v. Castillo,
140 F.3d 874, 883-84 (10th Cir.1998);
see also Charley,
189 F.3d at 1259-60. He further acknowledges that one panfel of this court cannot overrule another.
See United States v. Foster,
104 F.3d 1228, 1229 (10th Cir.1997). This disposes of Mr. Velarde’s arguments about the constitutionality of Rule 414.
We observed in
Castillo
that Fed. R.Evid. 403 “applies to Rule 414 evidence.”
Castillo,
140 F.3d at 884;
see also United States v. Meacham,
115 F.3d 1488, 1492 (10th Cir.1997). Rule 403 permits the court to exclude evidence “if its probative value is substantially outweighed by the danger of unfair prejudice, confusion of the issues, or misleading the jury, or by considerations of undue delay, waste of time, or needless presentation of cumulative evidence.” Fed.R.Evid. 403. We have noted the importance of the Rule 403 balancing: “[B]ecause of the unique nature of character evidence, it is important that the trial court ‘make a reasoned, recorded’ statement of its 403 decision when it admits evidence under Rules 413-415.”
Castillo,
140 F.3d at 884 (quoting
United States v. Guardia,
135 F.3d 1326, 1332 (10th Cir. 1998));
see also United States v. Mann,
193 F.3d 1172, 1173 (10th Cir.1999),
cert. denied,
— U.S.-, 120 S.Ct. 1284, 146 L.Ed.2d 231 (2000) (“ ‘Courts are to “liberally” admit evidence of prior uncharged sex offenses,’ but cannot ignore the balancing requirement of Rule 403.”) (quoting
Meacham,
115 F.3d at 1492).
The government agrees with Mr. Velarde that the record reveals no Rule 403 balancing of the Rule 414 evidence admitted in this case, and concedes that a remand to the district court is necessary. We have already determined to reverse and remand this case for a new trial. On retrial, should the government seek again to present Ms. Harrison’s testimony, the court must, on the record, conduct the necessary Rule 403 balancing.
III. Lesser Included Offense Instruction
Finally, Mr. Velarde argues the district court erred in failing to instruct the jury on the lesser included offense of abusive
sexual contact. “Determining whether a defendant was entitled to a lesser included offense instruction is a question of law that we review de novo.”
Castillo,
140 F.3d at 886.
Mr. Velarde was charged with aggravated sexual abuse of a child under the age of twelve, in violation of 18 U.S.C. § 2241(c). He argues the evidence supports at most a charge of abusive sexual contact, in violation of 18 U.S.C. § 2244. Mr. Velarde concedes that we have recently held that “[b]ecause section 2244 contains a specific intent element that sections 2242 and 2243 do not have, the crime of abusive sexual contact is not a lesser included offense of the crime of sexual abuse.”
Castillo,
140 F.3d at 886. This reasoning equally applies to section 2241, which also lacks the specific intent element. Mr. Velarde acknowledges that we are bound by
Castillo. See Foster,
104 F.3d at 1229.
Mr. Velarde argues, however, that the better reasoned approach is that of the Eighth Circuit in
United States v. Demar-rias,
876 F.2d 674, 676 (8th Cir.1989), which concluded that one could be guilty of a “sexual act” under section 2243, or 2241 or 2242, without having the specific intent required under section 2244.
Castillo
specifically considered and rejected the Eighth Circuit’s reasoning in
Demarrias.
That rejection controls our resolution of this issue.
CONCLUSION
For the foregoing reasons, we REVERSE and REMAND this case for a new trial.