filed a dissenting opinion.
The majority states that Applicant’s allegation regarding the unreliability of Dr. Coons’s expert testimony does not raise a cognizable claim on habeas corpus. Normally, the majority would be correct, but in this ease, I disagree. If the majority would have taken the time to research the record, it would have discovered that Applicant’s claim was raised but not addressed by this Court on direct appeal. Therefore, the majority’s entire analysis is based on a false premise. When addressing Dr. Coons’s expert testimony in the direct appeal opinion, this Court stated that Applicant was arguing that Dr. Coons *399was not qualified to testify as an expert witness on future dangerousness. The Court determined that Dr. Coons was qualified as an expert and explained:
The evidence was that Dr. Coons held both a law degree and a medical degree, served in the United States Army Medical Corps, and was a consultant for the Brook Army Medical Center. He was certified by the Board of Psychiatry and Neurology, trained in neurology and psychiatry, and had been in private practice since 1975. Dr. Coons had evaluated approximately 8,000 people for competency to stand trial, and had consulted on 150 capital cases for either the prosecution or the defense.
In evaluating the appellant for the special issue on future dangerousness, Dr. Coons examined “twenty pounds of printed material and quite a number of CDs regarding statements” as well as offense reports, pictures, and educational records. While he did not personally interview the appellant, the Rules do not require an expert to complete interviews in order to make such determinations.1
While this information is true and Dr. Coons is qualified as an expert, this was not Applicant’s argument at the Dau-bert/Kelly hearing or before this Court on direct appeal. Applicant’s argument in his appellate brief (although sometimes inart-fully worded) focused on the methodology of Dr. Coons’s testimony, not his qualifications:
... during the Daubert hearing, the State failed to show that Dr. Coons was qualified to give an opinion as to the future dangerousness of the Defendant. A witness is not qualified to give an opinion as to future dangerousness under Tex.R. Evid. 702 just because he has a degree and is licensed to practice medicine and/or psychiatry. See Holloway v. State, 613 S.W.2d 497, 500 (Tex.Crim.App.1981). An expert’s prediction of future dangerousness is inadmissible if “there is no evidence of technical or scientific support for it.” Id., at 501.
For Dr. Coons’ testimony, the following requirements must be met: (1) the expert’s testimony must be based on sufficient facts or data, (2) the expert’s testimony must be the product of reliable principles and methods, and (3) the expert must apply the principles and methods reliably to the facts of the case. See [E.I. du Pont de Nemours & Co. v.]Robinson, 923 S.W.2d [549] at 556-557 [ (Tex.1995) ]; Kelly, 824 S.W.2d at 571-572; Tex.R. Evid. 702. None of Dr. Coons’ testimony, during the Daubert hearing and trial on the merits, came close to meeting the above referenced standards. (38 C.R.R. 26-36, 65-84.) Dr. Coons presented no empirical data and/or scientific principles upon which to support his opinions and conclusions. In short, Dr. Coons’ testimony was based solely on supposition and a series of hypotheticals artfully supplied by the State.2
Applicant contended that the trial court erred in admitting the testimony of Dr. Coons regarding Applicant’s potential for future dangerousness and preserved this error at the trial level for review on direct appeal. However, although this argument is substantially similar to the appellant’s argument in Coble, this Court did not address it on direct appeal. The majority alleges that the issue of reliábility was not *400preserved at trial. This is incorrect. Most of the questions during the Dau-bert/Kelly hearing focused on Dr. Coons’s methodology. While the judge’s pronouncement at the conclusion of the hearing only addressed Dr. Coons’s qualifications, it is clear that Applicant argued methodology beginning with his motion and throughout the hearing. Applicant is not required to "correct the judge’s deficiency in ruling on qualifications only. Nevertheless, even if it had not been preserved at trial, the fact that it wasn’t preserved should have been addressed in' the direct appeal opinion. Instead this Court completely avoided the issue. In effect, this claim has not been raised and rejected on direct appeal and the failure of this Court to properly address the argument on direct appeal violated Applicant’s due process rights. We made a mistake, and now we have the opportunity and obligation to correct it.3 This is how I would have analyzed Applicant’s cognizable claim in light of Coble.
Arguments of the Parties
Applicant contends that the methodology used by Dr. Coons to assess future dangerousness in this case is indistinguishable from the methodology that was found to be unreliable in Coble. Applicant argues that when considered post-Coble, it was error to admit Dr. Coons’s testimony. Applicant further contends that this error denied his right to a fair sentencing trial and affected his substantial rights under Texas Rule of Appellate Procedure 44.2(b). Applicant asserts that the constitutional claims involved in this case distinguish it from Coble because they affect the reliability of the jury’s factual determination of the future-dangerousness special issue. More specifically, Applicant argues that the heightened reliability standard for death-penalty cases created by the Eighth Amendment applies to expert testimony and the admission of Dr. Coons’s testimony violated his due process rights. Applicant requests a new sentencing hearing.
The State responds that there was no violation of Applicant’s due process or Eighth-Amendment rights because there is no constitutional bar' to a psychiatrist’s prediction of future dangerousness. The State further argues that there is no heightened reliability standard for evidence in death-penalty cases. The State refers to our decision in Nenno v. State4 and posits that the reliability of Dr. Coons’s testimony should be assessed with less rigor than that required for hard-science methods of validation. The State also asserts that the admission of Dr. Coons’s testimony was harmless and made no material impact on the jury’s determination of the future dangerousness of Applicant.
Legal Principles
The legal framework surrounding expert psychiatric or psychological testimony with regard to the future-dangerousness special issue was established by the United States Supreme Court in Daubert v. Merrell Dow *401Pharmaceuticals.5 The admissibility of expert testimony lies within the discretion of the trial court and is subject to an abuse of discretion standard of review.6 Such testimony is governed by Texas Rule of Evidence 702, which places limits on the admissibility of scientific evidence to ensure that it is both relevant and reliable.7
In Daubert, the Supreme Court held that, in order for expert testimony to be reliable, the expert must have derived his or her conclusions from the scientific method.8 The Court provided illustrative examples of factors to be considered when determining acceptable scientific methods: (1) whether the theory or technique is refutable, testable, and falsifiable; (2) whether the findings have been subjected to peer review or published; (3) the error rate, including methodological standards; and (4) whether the theory or technique has been accepted by a relevant scientific field.9 These are flexible guidelines that have since been expanded to include technical and specialized-knowledge testimony.
In Kelly v. State,10 we determined that to be considered reliable, scientific-theory evidence must also satisfy three specific criteria relevant to its validity and application: “(a) the underlying theory must be valid; (b) the technique applying the theory must be valid; and (c) the technique must have been properly applied on the occasion in question.” 11 The function of the reliability inquiry is to assist trial courts in weeding out novel, unproven scientific theories so that only evidence based in sound scientific methodology is admitted.12 As we explained in Nenno v. State,13forensic psychiatry, although considered a “soft science,” is clearly a science and the general principles announced in Kelly (and Daubert.) apply.14 However, the use of “soft sciences” also requires the trial court to determine “(1) whether the field of expertise is a legitimate one, (2) whether the subject matter of the expert’s testimony is within the scope of that field, and (3) whether the expert’s testimony properly relies upon and/or utilizes the principles involved in the field.”15
I agree with the Supreme Court that predictions of future dangerousness are necessary and beneficial to the jury in capital cases; however, along with the Supreme Court and the American Psychological Association,16 this Court has recognized that psychiatric assessments of future dangerousness can be inaccurate.17 *402Because of this, the Daubert/Kelly/Nenno guidelines and the trial court’s role as gatekeeper are crucial when measuring the reliability of future-dangerousness testimony. The proponent of such testimony must establish its admissibility in each case.18 As we explained in Coble, the validity of the expert’s conclusions depends upon the soundness of the methodology.19
Application of the Daubert/KellylNenno factors to Coble and in this Case
In Coble, Dr. Coons testified that he developed his own methodology to come to a conclusion on future dangerousness.20 He explained that he considers factors such as the person’s criminal history, conscience, and attitude toward violence.21 After applying the Nenno factors to Dr. Coons’s testimony, we determined that the trial judge had abused his discretion by admitting the testimony.22 The issue under Rule 702 is whether his future-dangerousness testimony is based upon the scientific principles of forensic psychiatry. In Coble we said:'
From this record, we cannot tell what principles of forensic psychiatry Dr. Coons might have relied upon because he cited no books, articles, journals, or even other forensic psychiatrists who practice in this area. There is no objec-five source material in this record to substantiate Dr. Coons’s methodology as one appropriate in the practice of forensic psychiatry. He asserted that his testimony properly relied upon and utilized the principles involved in the field of psychiatry, but this is simply the ipse dixit of the witness. Dr. Coons agreed that his methodology is idiosyncratic and one that he has developed and used on his own for the past twenty to thirty years. Although there is a significant body of literature concerning the .empirical accuracy of clinical predictions versus actuarial and risk assessment predictions, Dr. Coons did not cite or rely upon any of these studies and was unfamiliar with the journal articles given to him by the prosecution.23
The record, including the Daubert/Kelly24 hearing conducted by the trial court in this case, indicates that Dr. ■ Coons used the same methodology for assessing future dangerousness as he did in Coble. When asked if his technique was objective or subjective during the Daubert/Kelly hearing Dr. Coons responded:
Well, I mean, I don’t know. I can’t sit here and quote an absolute piece of literature that says that, but I know that it is certainly well accepted in our field that people who are violent tend to — are *403more likely to commit violent acts than they are to be, to not in the future.
Dr. Coons was unfamiliar with the journal articles mentioned by the defense attorney and stated that it is impossible to have an established rate of error for his predictions.25 In regard to the underlying scientific or medical support for his methodology he admitted that “in terms of exactly where I have gotten my most recent material, I can’t tell you exactly.”
In Coble, as well as in this case, Dr. Coons’s testimony failed to meet the third Nenno standard, which requires “soft science” testimony to rely upon and use scientific principles relevant to the appropriate field of psychiatry. Based on these specific problems and omissions, I conclude that the prosecution did not satisfy its burden of showing the scientific reliability of Dr. Coons’s methodology for predicting future dangerousness by clear and convincing evidence during the Daubert/Kelly gatekeeping hearing in this case. I would hold that the trial judge abused his discretion in admitting Dr. Coons’s testimony before the jury.
Harm. Analysis
Having established error, it is necessary to determine if Dr. Coons’s inadmissible expert testimony affected Applicant’s substantial right to a fair sentencing trial. Relief should not be granted unless Applicant demonstrates that the error was harmful. If the error had a substantial and injurious effect or influence on the jury’s verdict, a substantial right has been affected and Applicant is entitled to relief.26 However, if the. inadmissible expert testimony had only a slight effect upon the jury’s deliberations, such non-constitutional error is harmless.27 My analysis of the harm associated with the erroneous admission of Dr. Coons’s testimony considers everything in the record, as each case must be examined on its own facts.28 In Jurek v. State,29 we determined that there are several factors that are important for a jury to consider when assessing the future dangerousness of the defendant.30 These factors include the defendant’s criminal record, age, the severity of any prior criminal conduct, and the presence of duress or other extreme form of emotional pressure at the time of the crime.31
*404In Coble, we determined that the admission of Dr. Coons’s testimony was harmless error and explained that:
(1) there was ample other evidence supporting a finding that there was a probability that appellant would commit future acts of violence; (2) the same basic psychiatric evidence of appellant’s character for violence was admissible and admitted, without objection, through other entirely objective medical sources; (3) Dr. Coons’s opinion was not particularly powerful, certain, or strong; (4) Dr. Coons’s testimony was effectively rebutted and refuted by Dr. Cunningham; (5) the State barely mentioned Dr. Coons during closing argument and did not emphasize him or his opinions.32
After reviewing the record, it is clear that the facts of Applicant’s case differ from Coble in each of these areas.
In this case, Dr. Coons was the only source of psychiatric evidence presented by the State regarding Applicant’s future dangerousness. The defense introduced two psychological reports made when Applicant was eight and eleven years old. These reports diagnosed him with ADHD and adjustment disorders, but unlike Co-ble, there was no evidence of “lifelong maladjustment” or any psychological or medical assessments that supported an opinion that Applicant has a character for violence.33 In Coble, we determined that Dr. Coons’s testimony was not powerful or strong because he simply stated — based on the State’s hypothetical — that there was a probability that Appellant would be a continuing threat to society.34 In this case, Dr. Coons agreed with the State that based on reasonable medical certainty Applicant would be a continuing threat to society. This indicates that Dr. Coons’s testimony in this case was likely to have a more powerful effect on the jury’s assessment of future dangerousness.
We also determined in Coble that the State did not emphasize Dr. Coons or his opinions during closing argument.35 This is not true in this case. In its closing argument, the State specifically stated, “Dr. Coons said this guy is a killing machine and he is going to kill again and he is going to be dangerous.” This clearly emphasized Dr. Coons’s testimony and again was likely to have a more powerful effect on the jury’s assessment of future dangerousness.36 Another difference between Coble and this case is that in Coble, Dr. Coons’s testimony was effectively rebutted and refuted by another psychiatric expert.37 The only other psychiatric expert to testify in this case, Dr. Kunick, did not offer any insight regarding the unscientific nature of Dr. Coons’s methodology. Instead, Dr. Kunick’s own methodology and lack of experience regarding predictions of future dangerousness were used by the State to impeach his testimony. The jury did not hear any evidence concerning the unscientific nature of Dr. Coons’s testimony from a qualified expert. The jury also heard evidence of Applicant’s criminal his*405tory; however, when considering his age, the severity of his prior criminal conduct, and his attitude about the offense as reflected in the statements he made to others, the totality of the evidence is not compelling enough to support a finding of future dangerousness absent the reliable testimony of a psychiatric expert.38 In Coble, the appellant had a long history of violence, this is simply not the case here.
Conclusion
It is clear that the testimony given by Dr. Coons concerning Applicant’s future dangerousness was improperly admitted at the punishment stage of Applicant’s trial. Because this error affected Applicant’s substantial right to a fair sentencing hearing,39 this Court should have remanded the case on direct appeal to the trial court for new punishment proceedings. Because the Court failed to address the correct issue on direct appeal, this Court violated Applicant’s right to due process. Unfortunately, the majority refuses to consider Applicant’s properly raised issue which makes it impossible for him to get relief.
With these comments, I respectfully dissent.
. Ramey v. State, No. AP-75,678, 2009 WL 335276, at *14, 2009 Tex.Crim.App. Unpub. LEXIS 124, at *44-5 (Tex.Crim.App. Feb. 11, 2009) (not designated for publication).
. Appellant’s Brief at 19-21. Ramey was "Appellant” on direct appeal and is now "Applicant” so these titles, in this case, may be used interchangeably.
. It seems apparent that the majority is going to great lengths to prevent Applicant from bringing forth his claim by unsubstantiated claims of non-preservation and cognizability. I have seen previous 11.071 writs with only a fraction of the amount of preservation evidence that this writ contains where we have not rejected the claim based on preservation or cognizability. This also seems particularly peculiar in light of this Court's previous scrutiny of Dr. Coons's methodology in Coble. Also, as our dissenting opinion further shows, the harm created by his testimony was clearly established in this case using the same analysis this Court used in Coble.
. 970 S.W.2d 549 (Tex.Crim.App.1998).
. 509 U.S. 579, 113 S.Ct. 2786, 125 L.Ed.2d 469 (1993).
. Joiner v. State, 825 S.W.2d 701, 708 (Tex.Crim.App.1992).
. Tex.R. Evid. 702.
. Daubert, 509 U.S. at 590, 113 S.Ct. 2786.
. Id. at 593-94, 113 S.Ct. 2786.
. 824 S.W.2d 568 (Tex.Crim.App.1992).
. Id. at 573.
. Jordan v. State, 928 S.W.2d 550, 555 (Tex.Crim.App.1996).
. 970 S.W.2d 549, 559-62 (Tex.Crim.App.1998).
. Id. at 560.
. Id. at 561.
. The American Psychological Association is a voluntary, non-profit scientific and professional organization with approximately 155,-000 members and affiliates.
. Barefoot v. Estelle, 463 U.S. 880, 897-99, 103 S.Ct. 3383, 77 L.Ed.2d 1090 (1983); Coble v. State, 330 S.W.3d 253, 274-75 (Tex. Crim.App.2010). The American Psychiatric Association (APA) filed an amicus curiae brief in the United States Supreme Court explaining that unstructured clinical assessments that are not grounded in science are less reliable than structured risk-assessment approaches. The APA described the unstructured clinical approach as a freeform ap*402proach to risk, management that is nothing more than a subjective assessment unaided by additional materials.
.See Druery v. State, 225 S.W.3d 491, 506-07 (Tex.Crim.App.2007) (explaining that the State has the burden of proving that there is a probability that the defendant would commit criminal acts of violence in the future, so as to constitute a continuing threat); Muniz v. State, 851 S.W.2d 238, 250 (Tex.Crim.App.1993) (emphasizing that the State is required to prove that the defendant "would, more likely than not, commit violent criminal acts in the future”).
. Coble, 330 S.W.3d at 276-77.
. Coble, 330 S.W.3d at 271.
. Id.
. Id. at 279-80.
. Id. at 277-78.
. See Daubert, 509 U.S. 579, 113 S.Ct. 2786; Kelly, 824 S.W.2d 568 (requiring a trial judge to conduct a "gatekeeping” hearing outside the presence of the jury to assess the reliability of scientific evidence).
. Determining an established rate of error is not impossible. As we explained in Coble, Dr. Coons cannot say what his accuracy rate is because he has never attempted to obtain records to check the accuracy of future-dangerousness assessments he has made in the past. Coble, 330 S.W.3d at 272.
. Tex.R.App. Proc. 44.2(b); Reese v. State, 33 S.W.3d 238, 243 (Tex.Crim.App.2000); Johnson v. State, 967 S.W.2d 410, 417 (Tex.Crim.App.1998).
. Id. See also Williams v. State, 273 S.W.3d 200, 219 (Tex.Crim.App.2008) (reiterating that if the State’s evidence on future dangerousness is overwhelming, error is more likely to be harmless). The future-dangerousness special issue depends on the jury's assessment of the defendant’s character. Id. at 219-20.
. Motilla v. State, 78 S.W.3d 352, 356 (Tex.Crim.App.2002).
. 522 S.W.2d 934 (Tex.Crim.App. 1975).
. Jurek v. State, 522 S.W.2d 934, 939 (Tex.Crim.App.1975).
. Id. at 939-40; See also Robinson v. State, 548 S.W.2d 63, 64 (Tex.Crim.App.1977). See also Roney v. State, 632 S.W.2d 598, 603 (Tex.Crim.App.1982) (finding that Appellant's young age, lack of a prior record or arrests, the absence of psychiatric evidence, and his surrender to the police did not support a verdict of future dangerousness); Wallace v. State, 618 S.W.2d 67, 69 (Tex.Crim.App.1981) (explaining that the death penalty must be set aside where "there was no evidence of prior convictions, no prior acts of violence, no character evidence, and no psychiatric evidence”).
. Coble, 330 S.W.3d at 286-87.
. See Id. at 281-82 (explaining that psychological evaluations developed for litigation are less reliable than those conducted independent of litigation by professionals acting in their normal capacity).
. Id. at 286 (emphasis added).
. Id. at 283-84.
. Psychiatric predictions of future violence often have an undue effect on the ultimate verdict. Barefoot, 463 U.S. at 929, 103 S.Ct. 3383. The Court further explained that even judges will accept psychiatric assessments with little regard for other testimony or cross-examination. Id.
. Coble, 330 S.W.3d at 286-87.
. In 2001, when he was 16 years old, Applicant received juvenile probation for sexual assault. He was charged with criminal mischief in 2003 when he was 18 years old and again received probation. After violating his probation, he was sentenced to 180 days in the Jackson County Jail. In 2004, when he was still 18 years old, he was charged with burglary of a building and served 180 days in the Jackson County Jail.
. Tex.R.App. P. 44.2(b).