Justice HOBBS
delivered the Opinion of the Court.
In this opinion, we review an unpublished decision of the court of appeals.1 Petitioner Lucas Golob appeals his conviction for criminal mischief arising from the January 5, 2001 break-in at the Custer County Road and Bridge Shop in Wetmore, Colorado. Golob argues that the judgment of conviction should be reversed because the trial court erred by: (1) allowing one of the prosecution's expert witnesses to testify that his findings were independently verified by a non-testifying witness, and (2) improperly limiting the testimony of the defense's expert witness.
We hold that the trial court improperly limited the testimony of Golob's expert witness after the prosecution had opened the door to this testimony and this error was not harmless beyond a reasonable doubt. Thus, we reverse the judgment of the court of appeals and remand this case to the court of appeals with instructions to return it to the trial court for further proceedings consistent with this opinion.
I.
On January 5, 2001, an intruder broke into the Custer County Road and Bridge Shop in Wetmore, Colorado, and stole property valued at more than $1,000. The intruder also caused several thousands of dollars in property damage to the shop. During the burglary, the intruder discharged a fire extinguisher, leaving a powdery residue throughout the shop. The next morning, the shop owner discovered that several items were missing from the store, including a pair of binoculars, a transceiver, and a camera. In addition, the intruder took two orange mesh vests which had been in a toolbox contained in a county truck parked near the shop. The police found partial shoeprints, made in the fire extinguisher residue, on a few sheets of copy paper that had been scattered on the floor during the break-in.
On April 7, 2001, the county sheriff's office obtained and executed a search warrant on Golob's home. During the search, the sheriffs deputies discovered two orange mesh vests within a backpack in a closet. The backpack also contained a pair of camouflage pants which were coated with a powdery substance. The sheriff's deputies did not locate any of the items missing from the shop.
Ten days later, on April 17, 2001, a sheriff's deputy collected a sample of the fire extinguisher residue from the inside of a desk drawer in the shop. The sheriffs office submitted the sample to the Colorado Bureau of Investigation ("CBI") for comparison with the powdery substance found on Golob's camouflage pants.
Agent Crippen conducted the comparison for the CBI. At trial, the district court certified Crippen as an expert in trace analysis and general forensic analysis. He testified that both substances were composed of ammonium phosphate and that they had identical characteristics. He further testified that ammonium phosphate is a common ingredient in fire extinguisher residue, as well as in detergents, fertilizers, and fire retardants. [1009]*1009Crippen concluded that both samples were consistent with fire extinguisher powder and were chemically indistinguishable from each other. However, he could not determine whether the residue from the pants came from the discharged fire extinguisher.
With respect to the orange mesh vests, the shop supervisor testified that the recovered vests were similar to the vests missing from the truck, but could not positively identify them as the same vests.
The sheriffs office also submitted the shoeprints recovered from the crime seene to the CBI for comparison to the soles of a pair of boots owned by Golob. Agent Sollars conducted the comparison for the CBI. At trial, the district court certified Sollars as an expert in the examination and comparison of known footwear to footwear track impressions. Based on his comparison of Golob's boots to the partial shoeprints recovered from the crime scene, Sollars testified that Golob's left boot could not have made any of the shoeprints. However, he could not rule out the right boot as a possibility. As a result, he conducted further tests. He determined that Golob's right boot could have been the source of one of the recovered prints. He also determined that it was highly probable that the right boot made another of the recovered prints, although he could not make a definite match.
During his direct testimony, the prosecutor provided Sollars with an expert report prepared by the defendant's foot impression expert, Anthony Cox. The prosecutor then questioned Sollars about the report's findings:
Prosecution: The proverbial bottom line [in the Cox report] states what?
Sollars: [quoting the Cox report] "My conclusion is that the identifying marks on the soles from [Golobl's boots do not conclusively match the boot prints taken from the piece of paper."
Prosecution: In your opinion, does that conclusion contradict your analysis and your conclusion? How do you compare and contrast that?
Sollars: In some regards. I'm not sure if [Cox] looked at all the tracks because there were some that were eliminated. When it says do not conclusively match, . then to me that would fall into the probable range. Just another way of saying it.
Prosecution: You're not telling this jury that in either one [Sollars's or Cox's] of your analyses on the two [recovered shoeprints] that it conclusively was, in other words, the dead bang positive identification. Is that accurate?
Sollars: Yes, sir.
Sollars testified that it is CBI policy that any highly probable or positive track impression identification must "be independently verified [] by another examiner before any verbal or written release of the results." He also testified that another CBI examiner independently verified his conclusions that the right boot could have been the source of one of the recovered shoeprints and that it was highly probable that the right boot made another recovered shoeprint. Defense counsel did not object to this testimony.
During presentation of his case, Golob called Cox to testify about the recovered prints. Cox worked for several years as a cobbler at a boot repair shop and, at the time of trial, worked as a certified pedorthist.2 In both of those positions, he gained extensive knowledge about foot tread design, wear patterns, and analyzing foot impressions. However, Cox never received any formal forensic training in shoeprint comparison. Instead, his expertise was based on his years of practice constructing and repairing shoe soles, as well as his study of a reference book on the subject, Foot Impression Evidence by William Bodziak. The trial court certified Cox as an expert on sole impressions and the court permitted him to testify as to specific characteristics of Golob's boots. However, upon the prosecution's objection, the trial court excluded any testimony regarding his comparison of the prints recovered at the crime scene with the soles of Golob's boots.
The jury found Golob guilty of criminal mischief. He was acquitted of burglary and [1010]*1010theft charges. He was subsequently sentenced to four years probation. On appeal, the court of appeals affirmed the conviction.
IL.
Free access — add to your briefcase to read the full text and ask questions with AI
Justice HOBBS
delivered the Opinion of the Court.
In this opinion, we review an unpublished decision of the court of appeals.1 Petitioner Lucas Golob appeals his conviction for criminal mischief arising from the January 5, 2001 break-in at the Custer County Road and Bridge Shop in Wetmore, Colorado. Golob argues that the judgment of conviction should be reversed because the trial court erred by: (1) allowing one of the prosecution's expert witnesses to testify that his findings were independently verified by a non-testifying witness, and (2) improperly limiting the testimony of the defense's expert witness.
We hold that the trial court improperly limited the testimony of Golob's expert witness after the prosecution had opened the door to this testimony and this error was not harmless beyond a reasonable doubt. Thus, we reverse the judgment of the court of appeals and remand this case to the court of appeals with instructions to return it to the trial court for further proceedings consistent with this opinion.
I.
On January 5, 2001, an intruder broke into the Custer County Road and Bridge Shop in Wetmore, Colorado, and stole property valued at more than $1,000. The intruder also caused several thousands of dollars in property damage to the shop. During the burglary, the intruder discharged a fire extinguisher, leaving a powdery residue throughout the shop. The next morning, the shop owner discovered that several items were missing from the store, including a pair of binoculars, a transceiver, and a camera. In addition, the intruder took two orange mesh vests which had been in a toolbox contained in a county truck parked near the shop. The police found partial shoeprints, made in the fire extinguisher residue, on a few sheets of copy paper that had been scattered on the floor during the break-in.
On April 7, 2001, the county sheriff's office obtained and executed a search warrant on Golob's home. During the search, the sheriffs deputies discovered two orange mesh vests within a backpack in a closet. The backpack also contained a pair of camouflage pants which were coated with a powdery substance. The sheriff's deputies did not locate any of the items missing from the shop.
Ten days later, on April 17, 2001, a sheriff's deputy collected a sample of the fire extinguisher residue from the inside of a desk drawer in the shop. The sheriffs office submitted the sample to the Colorado Bureau of Investigation ("CBI") for comparison with the powdery substance found on Golob's camouflage pants.
Agent Crippen conducted the comparison for the CBI. At trial, the district court certified Crippen as an expert in trace analysis and general forensic analysis. He testified that both substances were composed of ammonium phosphate and that they had identical characteristics. He further testified that ammonium phosphate is a common ingredient in fire extinguisher residue, as well as in detergents, fertilizers, and fire retardants. [1009]*1009Crippen concluded that both samples were consistent with fire extinguisher powder and were chemically indistinguishable from each other. However, he could not determine whether the residue from the pants came from the discharged fire extinguisher.
With respect to the orange mesh vests, the shop supervisor testified that the recovered vests were similar to the vests missing from the truck, but could not positively identify them as the same vests.
The sheriffs office also submitted the shoeprints recovered from the crime seene to the CBI for comparison to the soles of a pair of boots owned by Golob. Agent Sollars conducted the comparison for the CBI. At trial, the district court certified Sollars as an expert in the examination and comparison of known footwear to footwear track impressions. Based on his comparison of Golob's boots to the partial shoeprints recovered from the crime scene, Sollars testified that Golob's left boot could not have made any of the shoeprints. However, he could not rule out the right boot as a possibility. As a result, he conducted further tests. He determined that Golob's right boot could have been the source of one of the recovered prints. He also determined that it was highly probable that the right boot made another of the recovered prints, although he could not make a definite match.
During his direct testimony, the prosecutor provided Sollars with an expert report prepared by the defendant's foot impression expert, Anthony Cox. The prosecutor then questioned Sollars about the report's findings:
Prosecution: The proverbial bottom line [in the Cox report] states what?
Sollars: [quoting the Cox report] "My conclusion is that the identifying marks on the soles from [Golobl's boots do not conclusively match the boot prints taken from the piece of paper."
Prosecution: In your opinion, does that conclusion contradict your analysis and your conclusion? How do you compare and contrast that?
Sollars: In some regards. I'm not sure if [Cox] looked at all the tracks because there were some that were eliminated. When it says do not conclusively match, . then to me that would fall into the probable range. Just another way of saying it.
Prosecution: You're not telling this jury that in either one [Sollars's or Cox's] of your analyses on the two [recovered shoeprints] that it conclusively was, in other words, the dead bang positive identification. Is that accurate?
Sollars: Yes, sir.
Sollars testified that it is CBI policy that any highly probable or positive track impression identification must "be independently verified [] by another examiner before any verbal or written release of the results." He also testified that another CBI examiner independently verified his conclusions that the right boot could have been the source of one of the recovered shoeprints and that it was highly probable that the right boot made another recovered shoeprint. Defense counsel did not object to this testimony.
During presentation of his case, Golob called Cox to testify about the recovered prints. Cox worked for several years as a cobbler at a boot repair shop and, at the time of trial, worked as a certified pedorthist.2 In both of those positions, he gained extensive knowledge about foot tread design, wear patterns, and analyzing foot impressions. However, Cox never received any formal forensic training in shoeprint comparison. Instead, his expertise was based on his years of practice constructing and repairing shoe soles, as well as his study of a reference book on the subject, Foot Impression Evidence by William Bodziak. The trial court certified Cox as an expert on sole impressions and the court permitted him to testify as to specific characteristics of Golob's boots. However, upon the prosecution's objection, the trial court excluded any testimony regarding his comparison of the prints recovered at the crime scene with the soles of Golob's boots.
The jury found Golob guilty of criminal mischief. He was acquitted of burglary and [1010]*1010theft charges. He was subsequently sentenced to four years probation. On appeal, the court of appeals affirmed the conviction.
IL.
We hold that the trial court improperly limited the testimony of Golob's expert witness after the prosecution had opened the door to this testimony and this error was not harmless beyond a reasonable doubt.
A.
Hearsay Verification of Prosecution's Expert Testimony
At trial, Agent Sollars testified about the shoeprint evidence for the prosecution. On direct examination, Sollars stated that he had received independent verification of his conclusions that Golob's right boot could have made one of the recovered prints and that it was highly probable that Golob's right boot was the source of another. He went on to explain that according to CBI lab protocol, "any results of highly probable or positive have to be independently verified [] by another examiner before [the results are released]." On redirect, the prosecutor again inquired about CBI standard operating procedures and whether Sollars had gotten a second opinion on his highly probable conclusion. At no time did Golob object to this testimony. Finally, during rebuttal closing argument, the prosecutor compared Sollars's testimony to Cox's testimony. He stated, "And ask yourself who had some independent verification. ... Who had the report verified? It was Agent Sollars."
The court of appeals held that Sollars's testimony regarding the independent verification of his results was inadmissible hearsay. However, the court of appeals also concluded that the error did not fundamentally undermine the fairness of the trial. The court reasoned that the guilty verdict did not depend on the independent verification, but on two related cireumstances: (1) the fact that the "bootprints recovered from the scene plainly matched prints made from Golob's boot," and (2) the items recovered from Golob's backpack-the orange mesh vests and the camouflage pants coated with the chemical residue-indicated that he had been at the crime scene.
Hearsay is "a statement other than one made by the declarant while testifying at the trial or hearing, offered in evidence to prove the truth of the matter asserted." CRE 801(c). Unless an exception applies, hearsay statements are generally inadmissible. CRE 802.
The prosecution contends that the references to independent verification of Sollars's conclusions were not hearsay because they were offered only to explain CBI's protocol for conducting shoeprint analysis. According to the prosecution, under CRE 703, this testimony is properly classified as non-hearsay because the independent verification was part of the information Sollars relied upon in forming his expert opinion and was not offered for the truth of the matter asserted. See CRE 708 (permitting an expert to testify to facts and data that are otherwise inadmissible in evidence so long as they formed the basis of the expert's opinion and are of the type reasonably relied upon by experts in the field); see also Houser v. Eckhardt, 168 Colo. 226, 233, 450 P.2d 664, 668 (1969) ("The rule prevails generally that expert witnesses may testify to the information upon which they have relied in reaching their conclusions.... When presented for this purpose, the statements are not evidence of the matters stated, and hence not hearsay, but are merely explanatory of the opinion, enabling the jury to weigh it in the light of its basis.").
We disagree. First, as the court of appeals explained in a well-reasoned argument addressing this precise issue: "The use of facts and data to which CRE 703 applies is distinct from the use here.... [The testifying expert did not use the peer's conclusions as a basis for her findings and opinions. The conclusions merely boistered her findings and opinions." People v. Griffin, 985 P.2d 15, 18 (Colo.App.1998) (emphasis in original). In his testimony, Sollars stated that he only sought independent verification after he already reached his own conclusions. Thus, the verification did not provide a basis for the expert opinion, it provided additional support. Accordingly, the prosecution cannot [1011]*1011rely on CRE 708 to justify the admission of this statement.
Moreover, during his rebuttal argument, the prosecutor pointed to the verification as evidence that Sollars's conclusions were correct and, thereby, sought to bolster Sollars's credibility. The prosecutor told the jury, "[Alsk yourself who had some independent verification . ... Who had the report verified? It was Agent Sollars." The prosecutor was not using the hearsay testimony merely to describe the CBI's internal procedures as providing a basis for Sollars's conclusions; he directly encouraged the jury to use this fact as a reason for relying on Sollars's conclusion and discounting the testimony of Golob's expert witness. Because the prosecutor used the non-testifying witness's independent verification to prove the truth of the matter asserted-that is, to argue in favor of the validity of Sollars's findings-this testimony was inadmissible hearsay. However, because we reverse the judgment of the court of appeals on the basis that the trial court improperly restricted the testimony of Go-lob's expert witness, we do not determine whether the admission of this hearsay testimony standing alone would require reversal.
B.
Defense's Expert Testimony
We now consider whether the court of appeals correctly affirmed the trial court's decision limiting the defense's expert testimony.
1.
Standard of Review
The trial court has broad discretion to determine the admissibility of expert testimony. People v. Ramirez, 155 P.3d 371, 380 (Colo.2007). We will not overturn its decision unless it is "manifestly erroneous." Id. "This deference reflects the superior opportunity of the trial judge to gauge the competence of the expert and the extent to which his opinion would be helpful to the jury." Id.
Although we accord the trial court broad deference to determine the admissibility of expert testimony, this deference is not unbounded. Two important considerations counsel against the exclusion of expert testimony relevant to the defense of a eriminal defendant. First, CRE 702, which governs the admission of expert testimony, requires a "broad" and "liberal" inquiry into the admissibility of expert testimony. People v. Shreck, 22 P.3d 68, Ti-78 (Colo.2001). Second, the "exelusion of relevant and competent evidence offered in defense of a criminal charge is a severe sanction, implicating as it does the defendant's right to present a defense and ultimately the right to a fair trial." People v. Hampton, 696 P.2d 765, 778 (Colo.1985) (citations omitted), appeal after remand, 758 P.2d 1344 (Colo.1988); see also People v. Williams, 790 P.2d 796, 798 (Colo.1990) (" 'A decision excluding expert testimony offered by a criminal defendant is perhaps somewhat more susceptible of reversal because of the courts' sensitivity to the defendant's need and lack of access to the personnel available to the state."" (quoting 3 Jack B. Weinstein & Margaret A. Berger, Wein-stein's Federal Evidence § 702.04 (Joseph M. McLaughlin, ed.1988))).
2.
Analysis
CRE 702 governs the admission of expert testimony. Shreck, 22 P.3d at 70. We have ruled that, in determining whether expert testimony is appropriately admitted, the trial court should consider "the reliability and relevance of the proffered evidence" and "determin[el ... (1) the reliability of the scientific principles, (2) the qualifications of the witness, and (8) the usefulness of the testimony to the jury." Id. Trial courts should conduct a broad inquiry that considers the totality of the cireumstances involved in each case. Id. Trial courts should also evaluate the testimony under CRE 408 to ensure that the probative value of the evidence is not "substantially outweighed by the danger of unfair prejudice," and the evidence is not confusing, misleading, or cumulative. CRE 403; Romirez, 155 P.8d at 879. Finally, trial courts should "issue specific findings as [they apply] the CRE 702 and 403 analy-ses." Shreck, 22 P.8d at 70.
[1012]*1012In determining whether evidence is reliable, a trial court should consider two factors: (1) whether the scientific principles as to which the witness is testifying are reasonably reliable, and (2) whether the witness is qualified to opine on such matters. Id. at 77. The second of these two reliability prongs is at issue in this case. CRE 702 controls the qualification of experts. The rule is broadly phrased and provides that an expert may be qualified by any one of the five factors specified in the rule: knowledge, skill, experience, training, or education. Huntoon v. TCI Cablevision of Colo., 969 P.2d 681, 690 (Colo.1998). We have reinforced this liberal qualification standard by holding that a qualified expert witness need not "hold a specific degree, training certificate, accreditation, or membership in a professional organization." Id. In making its determination, a trial court should assess the witness's qualifications in the context of the evidence that is presented to the jury. See Shreck, 22 P.3d at 77; Campbell v. People, 814 P.2d 1, 7-8 (Colo.1991) (holding that the trial court retains broad discretion to evaluate on a case-by-case basis whether the testimony in question would assist the trier of fact to understand the evidence or to determine a fact in issue).
Golob offered Anthony Cox as an expert in "sole impressions." The prosecution did not object and the trial court accepted Cox as an expert on that subject. During both direct and re-direct examination, defense counsel sought to elicit Cox's opinion regarding whether Golob's boots were the source of the recovered prints. On each occasion, the trial court sustained the prosecution's objection to such testimony. At the time, the trial court did not explain its reasoning; however, when later resolving a motion for a new trial, the trial court stated that Cox lacked the requisite expertise to compare the recovered prints to the sole of Golob's boot.
We acknowledge that the question of whether Cox was sufficiently qualified to compare the prints found at the seene to the soles of Golob's boots is a close one. Although Cox had extensive experience in shoe construction and repair, he lacked direct experience or formal training in the comparison of known footwear to recovered shoeprints. However, when the entirety of the expert testimony on this issue is considered, we determine that the trial court should have permitted Cox to offer his comparison testimony because the prosecution opened the door to allowing Cox to testify about his print comparisons.
During the prosecution's direct examination of its footwear expert, the prosecution asked Sollars about Cox's expert report. Specifically, the prosecution asked Sollars to compare the report's conclusion that the "identifying marks from [Golob]'s boots do not conclusively match the boot prints taken from the piece of paper," with his own conclusions on this issue. Sollars testified that he interpreted Cox's conclusion to be consistent with his finding that it was highly probable that the right boot made one of the prints.
The concept of "opening the door" represents an effort by courts to prevent one party in a eriminal trial from gaining and maintaining an unfair advantage by the selective presentation of facts that, without being elaborated or placed in context, create an incorrect or misleading impression. People v. Murphy, 919 P.2d 191, 195 (Colo.1996) (citation omitted). When a party opens the door to otherwise inadmissible evidence, his opponent may then inquire into the previously barred matter. Id.; see also 21 Charles A. Wright & Kenneth W. Graham, Jr., Federal Practice and Procedure § 5089.1 (2d ed.2005).
By questioning its own expert witness about Cox's conclusion regarding the prints, the prosecution sought to characterize Cox's findings as consistent with the findings of the prosecution's expert. However, Cox disagreed with this characterization.3 By limit[1013]*1013ing Cox's testimony, the trial court permitted the jury to hear only one side of this issue. Cox was unable to explain the means by which he reached his opinion or how it differed from Sollars's opinion. Once the trial court permitted Sollars to comment on Cox's expert opinion, it should have allowed Cox to testify on this subject to provide the full context of the print evidence to the jury. Accordingly, we conclude that the trial court erred in limiting the scope of Cox's testimony.
Because a criminal defendant has the right to call witnesses in his defense, abridgment of that right is subject to a constitutional harmless error analysis. See Hampton, 696 P.2d at 778 ("The exclusion of relevant and competent evidence offered in defense of a criminal charge is a severe sanction, implicating as it does the defendant's right to present a defense and ultimately the right to a fair trial."); see also People v. Pronovost, 773 P.2d 555, 559 (Colo.1989); Blecha v. People, 962 P.2d 931, 942 (Colo.1998).
Despite the prosecution's argument that Golob did not object to the exelusion of his defense expert's print comparison testimony and the plain error standard should apply, we read the record as including a sufficient objection. Defense counsel plainly attempted to have Cox's comparison testimony admitted to rebut the prosecution's characterization of that testimony; the prosecution objected and the trial court ruled the testimony inadmissible, disallowing defense counsel from eliciting the testimony. The applicable standard requires reversal unless a court is confident beyond a reasonable doubt that the error did not contribute to the guilty verdict. Bernal v. People, 44 P.3d 184, 200 n. 11 (Colo.2002); see also Bartley v. People, 817 P.2d 1029, 1084 (Colo.1991) ("A constitutional error is harmless when the evidence properly received against a defendant is so overwhelming that the constitutional violation was harmless beyond a reasonable doubt."). Under this standard, the prosecution bears the burden of demonstrating that the trial court's decision to limit the seope of Cox's testimony did not contribute to the guilty verdict in this case. Key v. People, 865 P.2d 822, 827 (Colo.1994). If we determine "there is a reasonable probability from a review of the entire record that the defendant could have been prejudiced, the error cannot be harmless." Id.
In this case, the print evidence linking Golob to the crime seene was central to the prosecution's case. There were no eyewitnesses to place Golob at the erime scene, nor did Golob admit to being there. Although there was other evidence linking Go-lob to the crime seene-primarily the powdery residue found on Golob's camouflage pants and the two orange vests discovered in Golob's backpack-the shoeprint evidence was the linchpin of the prosecution's case. In fact, the prosecutor stated as much during closing argument when he told the jury, "If those aren't the boots, then he is not guilty.... The key element ... is ... did we get the right guy."
If allowed to testify, Cox would have challenged Sollars's expert opinion regarding the recovered prints. Thus, Cox's testimony potentially undermined the most significant item of evidence against Golob. The trial court's ruling allowed Sollars's testimony on Cox's expert report to stand unchallenged and resulted in an unfair advantage to the prosecution through the selective presentation of expert opinion to the jury on this important evidence. The trial court denied Cox the opportunity to state or explain how his opinion and conclusions differed from Sol-lars's characterization of them. While we cannot speculate as to the ultimate effectiveness of this line of argument, we conclude that Golob should have been allowed to more fully develop Cox's testimony on this subject.
The trial court's admission of the independent verification hearsay compounded this error. Both errors affected the same fundamental issue: whether the recovered shoe-prints showed that Golob was at the crime scene. The independent verification hearsay bolstered Sollars's credibility, while the limitation on Cox's testimony eliminated the only means by which Golob could effectively chal[1014]*1014lenge Sollars's expert opinion. Viewed in isolation, the independent verification testimony error might not have been sufficient to warrant a new trial. However, that error enhanced the prejudicial impact of the trial court's decision to limit Cox's testimony. The effect of both errors was to compromise Golob's ability to mount an effective defense. Based on our review of the entire record, we find that there is a reasonable probability that the trial court's decision to limit Cox's testimony contributed to Golob's conviction, coming as it did on the heels of admitting inadmissible hearsay testimony of independent verification. Thus, we conclude that the error was not harmless beyond a reasonable doubt.
IIL.
Accordingly, we reverse the judgment of the court of appeals and remand this case to the court of appeals with instructions to return it to the trial court for further proceedings consistent with this opinion.
Justice COATS dissents, and Justice RICE joins in the dissent.
Justice EID does not participate.