[279]*279Judge GIERKE
delivered the opinion of the Court.
A military judge sitting as a general court-martial convicted appellant, contrary to his pleas, of two specifications of making a false official statement; taking indecent liberties; and communicating a threat, in violation of Articles 107 and 134, Uniform Code of Military Justice, 10 USC §§ 907 and 934, respectively. The adjudged and approved sentence provides for a bad-conduct discharge, confinement for 10 months, and reduction to the lowest enlisted grade. The convening authority waived the automatic forfeitures for 6 months from the date of his action. The Court of Criminal Appeals affirmed the findings and sentence in an unpublished opinion.1
This Court granted review of the following issues:
I
WHETHER THE MILITARY JUDGE ABUSED HIS DISCRETION BY DENYING THE ADMISSION OP THE DEFENSE EXPERT TESTIMONY, AND, IN DOING SO, PREVENTED APPELLANT PROM PUTTING ON A DEFENSE.
II
WHETHER THE APPLICATION OF ARTICLES 57(a) AND 58b, UCMJ, VIOLATES THE EX POST FACTO CLAUSE OP THE CONSTITUTION WITH RESPECT TO APPELLANT.
For the reasons set out below, we resolve Issue I by holding that the military judge did not abuse his discretion. We resolved Issue II in appellant’s favor in United States v. Gorski 47 MJ 370 (1997).
Factual Background
The court below summarized the events preceding appellant’s court-martial as follows:
In February of 1991, Wanda Griffin, wife of the appellant, heard their two-year-old daughter, F., and her husband laughing in the bathroom of their home. Wanda went in to investigate and discovered her husband and F in the bathtub together. Wanda saw F playing with appellant’s penis which was erect. Wanda grabbed F out of the bathtub, and Wanda and F then moved out of the home into a shelter. Appellant told his wife that if she took F he would kill her. The conduct in the bathtub ultimately led to an [Air Force Office of Special Investigations] investigation, in the course of which the accused made a statement denying that his daughter touched his genitalia. Subsequently, appellant’s commander advised him that the allegations had been unsubstantiated and the case was closed. Sometime later the wife and daughter moved back to reside with appellant.
Several years later, in 1994, appellant received permanent change of station orders from Whiteman Air Force Base to Minot Air Force Base. He was advised that upon changing stations security clearances are normally updated. In August of 1995, in the course of an investigation for his update, appellant was interviewed by Mr. Schmitt, an investigator for the Defense Investigative Service (DIS) .... After making a statement denying any misconduct took place in February 1991 when he was bathing his daughter, appellant agreed to take a polygraph.
At 0900 hours on 19 September 1995 appellant met with the polygrapher, Special Agent Snyder, also from DIS, who advised appellant of his rights as required by [Department of Defense] directive and appellant declined counsel. By 1045 hours of the same day appellant had signed a statement admitting that his previous statements were not completely correct and that his daughter had touched his erect penis in the bathroom on the occasion witnessed by his wife.
Unpub. op. at 2-3.
Appellant’s two denials of misconduct were the basis for two charges of making false [280]*280official statements (Charge I and its specifications). His conduct with his daughter in the bathtub in February of 1991 was the basis of two charges: committing indecent acts with his daughter (specification 1 of Charge II); and taking indecent liberties with his daughter (specification 2 of Charge II). His alleged threat to kill his wife if she took his daughter away was the basis of a charge of communicating a threat (specification 3 of Charge II).
At his court-martial, appellant made a timely motion to suppress his statement to Special Agent (SA) Snyder on the grounds that it was coerced and false. In response, the Government produced the testimony of SA Snyder. He testified that he interviewed appellant, using “a rapport based interview technique that calls for the assumption that the person tells the truth to the person that he likes or respects.” The entire interview, from initial rights’ advisement until execution of the signed written statement, took about an hour and 45 minutes. SA Snyder described the interview technique as follows:
When he first comes in — it begins when I greet them I identify myself, show them the badge and credentials, shake hands, and then start talking to them at that time about their background, where they’re from, what they like, what they don’t like, in an attempt to develop rapport from the very beginning.
SA Snyder testified that he did not remember if he told appellant anything about himself, although it was possible. Only he and appellant were in the interview room. SA Snyder testified that he was not sure if he had both of appellant’s previous statements. He thought that he only knew about the statement that appellant made to SA Schmitt denying the allegations.
SA Snyder testified that he looks for nonverbal cues during an interview, such as lack of eye contact. If he sees clusters of nonverbal cues that indicate that the subject is not being truthful, he tries “to get them to talk and open up more.”
SA Snyder testified that he tells the subjects that the adjudicator, and not he, makes the decision whether to grant a security clearance. He testified that he works for the Department of Defense, not the military services, and that the subject is advised in the Privacy Act statement that the results of the interview can be disclosed to the military department concerned.
SA Snyder testified that, at the end of his interview of appellant, he and appellant discussed the cryptic notes he had taken, and he asked appellant if he would provide a handwritten statement. SA Snyder reviewed his notes and told appellant what they reflected that he had said. After telling appellant “[t]he areas that we needed to cover,” he asked appellant “if there was anything else that he wanted to” include in the statement, and appellant added some words “on his own.”
When the interview ended, appellant did not “appear to be scared” of SA Snyder. SA Snyder did not remember appellant getting upset at any time during the interview. SA Snyder testified that he made no threats or promises during the interview.
Appellant described the interview differently. He testified that the interview began “very calm, very caring.” SA Snyder asked him if he had allowed his daughter to touch his genitals, and appellant calmly answered, “no.” SA Snyder asked another similar question and appellant again responded, “no.” Appellant testified that SA Snyder then said, “I’m getting the impression you’re not being completely truthful.” SA Snyder asked, “have you ever said with your daughter on your lap, she is making me horny?” Appellant testified that he was caught off guard and was disgusted, causing him to hesitate. SA Snyder shook his finger at appellant and said, ‘You hesitated; that means you’re not being honest.” Appellant testified that SA Snyder pointed to the polygraph machine and said, “I can’t in good conscience hook you up to that thing unless I know you’re being honest and I feel you can pass it.”
Appellant testified that when he “started getting flustered and anxious,” SA Snyder pointed to a chair in the corner and said, “See that chair over there. I’m not charging you with anything. If I was charging you [281]*281with anything I would have you sit in that chair and I would be reading you your rights.” Appellant also testified that SA Snyder told him that if he did not answer honestly, “the judiciaries” would take away his security clearance, and that appellant would be kicked out of the Air Force because he needed a security clearance to do his job.
Appellant testified that when he became so flustered that he started stuttering, SA Snyder “would back off and calm down.” SA Snyder began talking about appellant’s father, a retired Marine, rebuilding the rapport with appellant, calming him down, and getting him relaxed.
Appellant testified that SA Snyder told him that, if he was not honest, the Air Force would label him a “pedophore.” Appellant reacted with fear and anger, and again SA Snyder calmed him down. Finally, SA Snyder told appellant, “We don’t need to go on any further with this interview. There is no way you’re going to pass that test. I’m not even going to try to administer it to you.” SA Snyder also said, “Before you leave I need to get a statement from you.” Appellant responded that he wanted to speak to an attorney before making a written statement. According to appellant, SA Snyder responded that he “cannot and will not forbid [him] to see an attorney,” but that if he did not have a statement in an hour or two he would “make a statement for you.” Appellant then agreed to provide a written statement. Appellant testified that SA Snyder told him, “Now write down what I tell you to.” SA Snyder dictated the entire statement, one or two sentences at a time, and then told appellant to initial the beginning and the end and sign at the end of the statement.
In support of his contention that his confession was false and coerced, appellant proffered the testimony of Dr. Rex Frank, a psychologist. The Government filed a motion in limine, asking the military judge to exclude Dr. Frank’s testimony on the ground that he was not qualified to offer an opinion on the issue, that his testimony was not relevant, and that his conclusions were not reliable.
After a lengthy Article 39(a), UCMJ, 10 USC § 839(a), session, the military judge excluded Dr. Frank’s testimony. Thereafter, appellant requested trial by military judge alone. The military judge’s ruling is the basis of Issue I.
Dr. Frank is a licensed psychologist, with a doctorate in counseling psychology. He has prior military service as an Air Force psychologist, with experience in security and personal reliability (PRP) investigations.
In 1993 he began studying the areas of false confessions and coercion in interrogations. He examined about 20 years’ worth of research materials, including a study of 350 cases, conducted in 1987, where suspects had confessed but had later been determined to be innocent based on other evidence. Of those 350 cases, the study concluded that 49 involved coerced confessions. He testified that there was a problem with the study because “they did not differentiate between the issue of coercion and the issue of torture in the police interviews that resulted in a confession.” He also testified that research on false confessions is “relatively new,” dating back to the 1980s.
Dr. Frank testified that the thrust of the research into confessions looks at three issues: (1) the nature of the person providing a confession; (2) the nature of the allegations against the person; and (3) the nature of the interaction between the individual and his personality structure and the circumstances of the interview. The third issue is the primary focus of the research.
Dr. Frank testified that there are three components that affect vulnerability to making coerced confessions:
(1) intelligence, with persons of lower intelligence being more vulnerable;
(2) “a tendency towards compliance, or that is the tendency to give in to authorities’ demands”; and
(3) suggestibility, “the tendency for a person to change his answers” in response to “challenges by an interviewer.”
Dr. Frank administered a battery of psychological tests on appellant, reviewed his personnel records, interviewed him for a to[282]*282tal of about 6 hours, interviewed SA Snyder, interviewed another person who had been interrogated by SA Snyder, and listened to an interview of appellant’s spouse. Dr. Frank concluded that appellant is intelligent, not suggestible, but is a “highly compliant individual.” Dr. Frank also testified that military persons “are more compliant in general than people on the outside or they don’t stay here.”
Dr. Frank determined that appellant’s psychological test scores on the trait of compliance was “within a couple of tenths of a point [of] the average score” for “false confessors.” In Dr. Frank’s opinion, appellant’s confession is “consistent with a coerced compliant type of confession.” Dr. Frank defines a “coerced compliant confession” as “a form of false confession in which the individual comes to confess to things that he or she did not do and full well knows during the interview-interrogation-and-eonfession process that they are confessing to something that they did not do.”
Dr. Frank testified that psychologists have no scientific means of determining whether a confession is true or false. Thus, he testified that he could not “pass judgment on the truthfulness or lack of truthfulness of the words that [appellant] put on the paper.” He testified that psychologists can only “understand some of the characteristics that make a person vulnerable to provide information which is erroneous.” He explained that the issue of coercion is separate “from the veracity of the content of the confession.”
Although Dr. Frank based his opinion in part on his interviews of appellant regarding appellant’s perceptions of the interview by SA Snyder, he testified that he had no way of knowing the degree of contamination of his interview by prior interviews by others. He testified that he had several reservations about his conclusions, because the “normative standards” were based on British prisoners, not American military personnel. He admitted that it is extremely difficult to determine when an interview “becomes coercive.”
In response to questions from the military judge, Dr. Frank pointed to two examples of words in appellant’s statement that were inconsistent with appellant’s speaking style and vocabulary. The first was “pedaphore” instead of “pedophile.” The second was “judi-cators” instead of “adjudicators.”
In support of its motion in limine, the prosecution called Lieutenant Colonel (LtCol) Nancy Slicner, a licensed psychologist with a doctorate in psychology, serving as Chief, Behavioral Science Unit, Air Force Office of Special Investigations, Washington, D.C. LtCol Slicner testified that one of the problems with research in the area of false confessions is inability to control the variables: the interviewer, the interview environment, and the psychology and personality of the person being interviewed. Because of these variables, it is difficult to develop a research protocol to conduct a “controlled study.” She testified that there are no studies that show a “statistically significant correlation between ... a specific personality trait and the likelihood of somebody falsely confessing.” In response to a question from the military judge, LtCol Slicner testified that the subject area “has not reached scientific acceptability.”
The military judge granted the prosecution’s motion in limine to exclude the testimony of Dr. Frank. He had noted in an earlier ruling that “the accused’s version of events while in the presence of Special Agent Snyder is significantly different than that related by Special Agent Snyder.” The military judge then made the following findings of fact regarding the circumstances of the interview:
To the extent that the versions differ, I find Special Agent Snyder’s testimony to be far more credible and believable than the testimony of the accused. Specifically, I find that the evidence shows that the accused was properly advised of his Article 31 rights, that he waived his rights, and that he provided the information in Prosecution Exhibit 1 for Identification. I find the accused’s statements that he asked for an attorney and Special Agent Snyder proceeded with the interview anyway to be not credible and not believable.
[283]*283The military judge later explained his ruling as to Dr. Frank as follows:
After carefully considering all of the evidence, I conclude that Dr. Frank knows a lot about the subject, but that this is not a proper subject matter for expert testimony in that this information will be more confusing to the members than helpful to them. The evidence he has does not have the necessary reliability to be of help to the trier of fact. Finally, under [Mil.R.Evid.] 403, I conclude that any probative value of this evidence is substantially outweighed by the danger of confusion of the members and also by consideration of waste of time.
After the military judge so ruled, appellant withdrew his request for enlisted members and requested trial by military judge alone. The military judge asked appellant, “Does the fact that I have presided over this trial for two days and listened to all the evidence that came before me cause you any concern at this point about me sitting as the trier of fact in this case?” Appellant responded, “No, sir, it does not.” The military judge then granted the request for a bench trial.
Discussion
Appellant contends that the military judge abused his discretion by denying him the right to present expert scientific testimony. He argues that he has a constitutionally protected right to present legally and logically relevant evidence at trial, citing United States v. Woolheater, 40 MJ 170, 173 (CMA 1994). He argues that the expert testimony was a significant factor in the defense case because his only realistic hope of acquittal was to convince the military judge that his confession was either coerced or false.
The Government asserts that the military judge conducted a proper analysis under Daubert v. Meirell Dow Pharmaceuticals, Inc., 509 U.S. 579, 592-97, 113 S.Ct. 2786, 125 L.Ed.2d 469 (1993) and United States v. Houser, 36 MJ 392, 397-400 (CMA), cert. denied, 510 U.S. 864, 114 S.Ct. 182, 126 L.Ed.2d 141 (1993). Accordingly, the Government argues that the military judge did not abuse his discretion by excluding Dr. Frank’s testimony because it was unreliable, had minimal probative value, and would have been confusing and a waste of time.
Mil.R.Evid. 304(e), Manual for Courts-Martial, United States (1998 ed.)2 places the burden on the prosecution to establish admissibility of a confession, if the defense has made an appropriate motion or objection. Mil.R.Evid. 304(e)(1) requires a military judge to “find by a preponderance of the evidence that a” confession “was made voluntarily before it may be received into evidence.”
Dr. Frank’s testimony was offered as expert testimony under Mil.R.Evid. 702, which 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.
This rule is taken from Fed.R.Evid. 702 verbatim. See Drafters’ Analysis of Mil.R.Evid. 702, Manual, supra at A22-47.
In Houser, supra at 397, this Court set out 6 factors that must be established by the proponent of expert testimony:
(1) “the qualifications of the expert”;
(2) “the subject matter of the expert testimony”;
(3) “the basis for the expert testimony”;
(4) “the legal relevance of the evidence”;
(5) “the reliability of the evidence”; and
(6) probative value outweighing the other considerations outlined in Mil.R.Evid. 403.
Two months after Houser was decided, the Supreme Court decided Daubert, concerning “scientific” evidence offered under Fed.R.Evid. 702. The Supreme Court focused on the issues of reliability, 509 U.S. at 590, 113 S. Ct. 2786, and relevance, id. at 591, 113 S.Ct. 2786, holding that Fed.R.Evid. 702 assigns to the trial judge the duty to act as a gatekeeper, i.e., “the task of ensuring that an [284]*284expert’s testimony both rests on a reliable foundation and is relevant to the task at hand.” Id. at 597, 113 S.Ct. 2786. The Supreme Court, while disclaiming any attempt “to set out a definitive checklist or test,” listed the following 6 factors to be considered by the trial judge in determining whether scientific evidence meets the requirements for reliability and relevance:
(1) Whether the theory or technique “can be (and has been) tested”;
(2) Whether “the theory or technique has been subjected to peer review and publication”;
(3) The “known or potential” error rate;
(4) The “existence and maintenance of standards controlling the technique’s operation”;
(5) The degree of acceptance within the “relevant scientific community”; and
(6) Whether the “probative value” of the evidence “is substantially outweighed by the danger of unfair prejudice, confusion of the issues, or misleading the jury.” 509 U.S. at 593-95, 113 S.Ct. 2786.
Although Houser was decided before Daubert, the two decisions are consistent, with Daubert providing more detailed guidance on the fourth and fifth Houser prongs pertaining to relevance and reliability. Daubeti concerned “scientific” evidence, leaving open the question whether the same analysis applied to “technical or other specialized knowledge.” Nevertheless, this Court applies the Houser/Daubert analysis to “technical or other specialized knowledge.” See Houser, supra at 393 (evaluating testimony about rape trauma syndrome); United States v. St. Jean, 45 MJ 435, 444 (1996) (applying Houser to “psychological autopsy”); see also United States v. Hall, 93 F.3d 1337, 1342 (7th Cir.1996) (applying Daubert analysis to psychological evidence).
We review a military judge’s rulings under Mil.R.Evid. 702 for abuse of discretion. See United States v. Thomas, 49 MJ 200, 202 (1998). We review de novo the question whether the military judge properly followed the Daubert framework. See Hall, 93 F.3d at 1342. If the judge did so, we will not overturn the ruling unless it is “manifestly erroneous.” See Thomas, supra at 202, citing General Electric Co. v. Joiner, 522 U.S. 136, -, 118 S.Ct. 512, 517, 139 L.Ed.2d 508 (1997).
The defense offered the testimony of Dr. Frank to show that appellant’s confession was involuntary. Under Mil.R.Evid. 304(e), the burden was on the Government to show, by a preponderance of the evidence, that appellant’s confession was voluntary. See United States v. Cottrill, 45 MJ 485, 488 (1997), citing United States v. Martinez, 38 MJ 82, 87 (CMA 1993), cert. denied in Cottrill, 520 U.S. 1213, 117 S.Ct. 1698, 137 L.Ed.2d 824 (1997). The voluntariness of a confession is determined by examining “the totality of all the surrounding circumstances — both the characteristics of the accused and the details of the interrogation.” Id. Dr. Frank’s testimony was potentially relevant regarding “the characteristics of the accused,” i.e., that he had a compliant personality that made him susceptible to psychological coercion. The critical question was the reliability of the proffered testimony.
Regarding appellant’s claim that Dr. Frank’s testimony would have shown that his confession was false, we hold that the military judge did not abuse his discretion by excluding it. Dr. Frank testified that he could not opine on the question whether the confession was false. 50 MJ at 282. Even if he could, such testimony would run afoul of this Court’s decisions holding that an expert witness may not act as a “human lie detector.” United States v. Birdsall, 47 MJ 404, 410 (1998); United States v. Partyka, 30 MJ 242, 247 (CMA 1990); see United States v. Cacy, 43 MJ 214, 218 (1995); United States v. Harrison, 31 MJ 330, 332 (CMA 1990).
Regarding appellant’s claim that Dr. Frank’s testimony would have shown that his confession was coerced, we also hold that the military judge did not abuse his discretion by excluding it. Again, appellant overstates Dr. Frank’s testimony. Dr. Frank did not testify that appellant’s confession was coerced. At most, he could testify only that appellant’s confession was “consistent with a coerced [285]*285compliant type of confession.” 50 MJ at 282. Even this conclusion was based in part on appellant’s description of the interview, which the military judge found to be “not credible and not believable.” 50 MJ at 282. Dr. Frank did not, and perhaps could not, say whether it was also consistent with a voluntary confession.
Dr. Frank admitted that he had reservations about the normative-standards base on which he based his conclusions. 50 MJ at 282. There was no showing that the 1987 study of 49 coerced confessions among British prisoners could be reliably applied to American military personnel. Dr. Frank also admitted that the failure of the study to differentiate between torture and psychological coercion was a problem. 50 MJ at 281, 282. Thus, we conclude that the military judge did not abuse his discretion by ruling that the evidence was not sufficiently reliable.
Furthermore, the proffered evidence shed little light on the question whether appellant was coerced to confess. While Dr. Frank diagnosed appellant as having a compliant personality, he also admitted that most successful military persons are compliant. He also testified that appellant had two traits that made him less susceptible to coercion: high intelligence and low suggestibility. 50 MJ at 282. Thus, Dr. Frank’s testimony was of little value in determining whether appellant was coerced to confess.
Based on the foregoing, we conclude that the military judge properly exercised his “ga-tekeeping” function under Mil.R.Evid. 702 and Daubert, and that he did not abuse his discretion by excluding Dr. Frank’s testimony.
Decision
The decision of the United States Air Force Court of Criminal Appeals is reversed with respect to Issue II (ex post facto violation). In all other respects, it is affirmed. Collection of any forfeitures, and execution of the reduction in grade prior to the date of the convening authority’s action, are hereby declared to be without legal effect. Any forfeitures already collected from appellant, and any pay and allowances withheld because of the premature reduction in grade, will be restored. The record of trial is returned to the Judge Advocate General of the Air Force for appropriate action.