Appellant, Robert A. Russell, appeals from his convictions of armed carjacking, armed robbery, possession of a firearm during commission of a crime of violence (PFCV), unauthorized use of a motor vehicle (UUV), and theft.1 The sole issue raised by his appeal is whether the trial court abused its discretion in precluding appellant from presenting expert testimony concerning the reliability and accuracy of eyewitness testimony. Concluding that more recent developments in the scientific community and in our case law require more in-depth consideration of the factors affecting admissibility than was afforded here, we reverse and remand for further proceedings consistent with this opinion.
I.
A. Factual Background
The charges arose out of an armed carjacking and robbery of Damon Warren that occurred on December 23, 2006 at about 3:50 a.m. Having consumed some beer and concerned about a pending driving under the influence charge, Warren allowed his friend, Terence Childs to drive his Mercury Grand Marquis automobile to Childs’ home in the 1900 block of Jackson Street, N.E. Warren remained in the car while Childs went inside to get some of his belongings. Warren testified that he noticed two young black men wearing hoods and North Face jackets conversing and walking down Jackson Street. Warren testified that he did not feel threatened, and he got out of the car to urinate with his back to the two men. Warren then heard someone curse and demand that he “give it up.” Warren turned and saw the same two men, one of whom was masked, pointing guns in his face; they demanded his keys and identification. Warren threw his driver’s license, cell phone, and about $400 in cash to the ground and told them *584more than once that he did not have the keys. One of the robbers exclaimed that the keys were in the ignition, jumped in the vehicle, and drove off toward 18th Street. The masked man ran to a cream, white or light colored car that was parked up the street and drove away in front of Warren’s Mercury. Warren ran to Childs’ house and informed Childs, who called the police. Warren described the man who drove away in his car as weighing between 145 and 160 pounds, wearing dark jeans and a black North Face jacket with the hood on and no mask. Warren also said that the man was an inch or two taller than his own height (about 5' 10" to 5' 11"),2 “y0Uthful,” with braided hair hanging “[cjlose to the chin,” and “smooth” “cinnamon-brown skin.”
Metropolitan Police Department Officers Eric Hairston and Carlos Heraud, responding to a radio assignment for the car jacking, met Warren at the Jackson Street location. Appellant described his assailants to the police as two black males, both wearing black North Face jackets with hoods, blue or dark-colored jeans, and black boots. Warren told the police that the man who drove his car had dreadlocks, and the second man had hair twists. Officer Heraud broadcast a lookout for the suspects and the vehicles. A few minutes later, Officer Evely spotted the Mercury turning left onto Rhode Island Avenue, N.E. from 18th Street. Officer Evely followed the car, which made a right turn on 20th Street, and a left on Hamlin Street before both cars speeded away. The driver of the Mercury lost control of the car and crashed it into a tree. When Officer Evely was about 15 feet from the vehicle, a man, whom the officer identified in court as appellant, got out of the car, looked in his direction and ran down Hamlin Street. The officer chased appellant for three to four minutes before appellant ran into a wooded area near Franklin Street, N.E. Officer Terrence Liddell, who responded to Officer Evely’s request for K-9 assistance, found Warren’s driver’s license about halfway into the woods; however, he did not find the suspect.
About 5:10 a.m., Officers Ludovick Rock and Greg Pamer, who had participated in the search for the carjacking suspect earlier, spotted appellant, who matched the description, walking in the 2200 block of Douglas Street, N.E. Officer Rock noticed that appellant’s blue jeans were wet and muddy. At the officer’s request, appellant removed his hood, revealing that he had shoulder-length dreadlocks. Officer Rock reported that he had a suspect fitting the broadcast description, and Officer Evely went to the area and identified appellant as the person he had chased from the wrecked Mercury. The police first showed Warren a heavy-set, light skinned, unconscious man with braids or dreadlocks, wearing a black North Face jacket on Irving Street, N.E., but Warren said that he was not one of the carjaekers. Warren said that the police told him later that they had another person for him to look at. At that show-up, Warren remained in the police cruiser about 20 feet from where appellant was standing with Officers Rock and Palmer. A police officer told Warren that “we have an individual we want you to ID,” and that he should “look at him.” At first, Warren told the officer that he was not sure. Warren then asked to have appellant stand up with his hood up over his head. Once appellant pulled the hood up, Warren identified him as the unmasked carjacker.
*585B. Proffered Expert Evidence
At trial, defense counsel sought to call as an expert Dr. Lori Van Wallendael, a university professor of cognitive psychology with specialized experience in eyewitness perception and memory.3 He asserted that Dr. Wallendael would “testify about the factors present in this case that are known to affect adversely eyewitness perception, memory, and identification.” He also stated that the professor’s area of expertise was a well-established field, that the subject areas “are supported by extensive research and have broad acceptance in the scientific community,” and that she had been qualified to testify in these areas in the Superior Court of the District of Columbia and in other jurisdictions. The defense proffered that the expert would testify that: (1) there is little correlation between a witness’ level of confidence and the accuracy of the identification; (2) police officers perform no better than lay witnesses in witness identifications; (3) the presence of a weapon adversely affects the witness’ ability to focus and remember the face of the perpetrator; (4) presenting a suspect in clothing similar to that of the perpetrator undermines the reliability of the identification, an effect known as “clothing bias”; and (5) post-event information (contamination) leads to distortions and memory errors.
In the trial court, the government challenged the admissibility of appellant’s expert’s testimony principally on the ground that it did not meet the first Dyas factor. Specifically, the government argued that the subject matter of the proffered testimony was within the common experience of every juror and that defense counsel could probe on cross-examination and present argument on the matters he sought to bring out through the expert. The trial court concluded that none of the subject areas were beyond the ken of lay jurors to understand, evaluate, and to deal with effectively through cross-examination, and therefore, it excluded the evidence.
II.
Analysis
Appellant argues that reversal is required because the trial court erred in precluding him from presenting expert testimony on the psychological factors affecting the accuracy of eyewitness identifications. He contends that the court’s erroneous ruling deprived him of the opportunity to present his defense of misidentification where the government’s case rested entirely on the eyewitness identifications of two strangers. The government argues in response that appellant failed to lay the requisite foundation for admission of the proffered expert testimony.
The trial court has broad discretion to admit or exclude expert testimony. Oliver v. United States, 711 A.2d 70, 73 (D.C.1998) (citation omitted). However, “that discretion ‘must be weighed against the constitutional rights of a defendant to present a defense.’ ” Howard v. United States, 656 A.2d 1106, 1117 (D.C.1995) (quoting Clark v. United States, 639 A.2d 76, 81 (D.C.1993)) (citing Bassil v. United States, 517 A.2d 714, 716 (D.C.1986)). Generally, expert testimony should be admitted if it will aid the trier of fact in understanding the facts in issue. Coates v. *586 United States, 558 A.2d 1148, 1152 (D.C.1989) (citations omitted). We review such decisions for an abuse of discretion, and we will not disturb the trial court’s ruling unless manifestly erroneous. Id. (citing Salem v. United States Lines Co., 370 U.S. 31, 35, 82 S.Ct. 1119, 8 L.Ed.2d 313 (1962)) (other citation omitted).
It is now well-established that the admissibility or exclusion of expert eyewitness identification testimony is governed by a three-pronged test set forth in Dyas v. United States, 376 A.2d 827 (D.C.1977). That test requires the trial court to consider the following criteria:
(1) the subject matter “must be so distinctively related to some science, profession, business or occupation as to be beyond the ken of the average layman; ” (2) “the witness must have sufficient skill, knowledge, or experience in that field or calling as to make it appear that his opinion or inference will probably aid the trier in his search for truth ” and (3) expert testimony is inadmissible if “the state of the pertinent art or scientific knowledge does not permit a reasonable opinion to be asserted even by an expert.”
Id. at 832 (alteration in original) (quoting MoCormick on EvidenCE § 13 at 29-31 (E. CleaRY, 2d ed. 1972)).
In Dyas, this court held that the trial court did not abuse its discretion in excluding proffered expert eyewitness identification testimony concerning complications in the identification process misunderstood by the lay public that were somewhat similar to those proffered in this case. See Dyas, supra, 376 A.2d at 831-32. The opinions sought to be admitted in Dyas included that: (1) scientific literature supports that stress has an adverse impact on the accuracy of a witness’ observations and subsequent identifications; (2) suggestions from one in authority have a considerable impact on the identification process; and (3)a person’s public announcement of an opinion will motivate him or her to maintain that opinion in spite of evidence to the contrary. Id. at 831. This court held that the subject matter of the testimony was not beyond the ken of the average layperson and that the jury could evaluate the witness’ credibility based on information disclosed during cross-examination. Id. at 832.
More recently, however, this court reaffirmed that “ ‘Dyas and its progeny do not articulate a per se requirement that all expert testimony about the reliability of eyewitness identification must be excluded.’ ” Benn v. United States, 978 A.2d 1257, 1277 (D.C.2009) (Benn II) (quoting Green v. United States, 718 A.2d 1042,1051 (D.C.1998)). In Benn II, we observed that new developments in relevant scientific studies and case law concerning eyewitness identification may require more than cursory scrutiny of the first Dyas factor. Id. (citing Burgess v. United States, 953 A.2d 1055, 1063 n. 12 (D.C.2008)) (other citation omitted). This court remanded the case to the trial court for proper consideration of the proffered expert testimony where the court came close to employing a per se rule excluding the evidence and had overlooked more recent studies to the effect that scientific research had identified some factors not apparent to lay jurors that might cause them to question an eyewitness’ identification.4 Id. at 1261, *5871277, 1280. We observed that careful consideration of the proffer and a voir dire of the defense expert as requested would have provided evidence concerning whether the research underlying the proffered expert’s findings is generally accepted in the scientific community and whether the subject is beyond the ken of average lay persons. Id. at 1277. We admonished that “[wjithout such an inquiry and factual determination, the trial court is not free to disregard this court’s expressed views on the subject.” Id. at 1277-78 (footnote omitted).
The government argues that appellant’s meager proffer gave the trial court no basis to find that the proposed expert testimony would be beyond the ken of the average lay person and left the trial court with no alternative but to exclude the evidence. While appellant’s proffer was not as extensive as the one made in Benn II, it was sufficient to require further consideration before excluding it on an almost per se basis as within the ken of average lay persons.5 In making his oral presentation, appellant relied, in part, upon his Rule 16 letter. In that letter, appellant stated specifically the opinions that his expert would express which would be based upon “her own independent research and studies, and her knowledge of and familiarity with research and studies conducted by other experts in her field.... ” He represented that all of the areas that she would cover “are supported by extensive research and have broad acceptance in the scientific community.” There is attached to the Rule 16 letter an eight-page curriculum vitae outlining Dr. Van Wallendael’s education, training, and experience in her field. Defense counsel proffered that the factors affecting identification about which the expert would testify were not obvious, but beyond the ken of the average lay person, citing by way of example supporting cases for two proffered subject areas.6
In response to the court’s question, defense counsel represented that there are scientific studies supporting the opinions that he sought to present. He represented that the Public Defender Service (PDS) had an outside organization conduct a survey that showed that jurors do not understand the issues he sought to present through an expert and that the results of the study were consistent with findings of similar studies in other jurisdictions, and he offered to provide the study for the court. As in Benn II, there was no voir dire of the proffered expert to gain further information on the science and research underlying the proffered opinions and their usefulness to the jury in understanding the effect of particular phenomena on eyewitness identification. Rather, without the benefit of the detailed analysis on the subject provided in Benn II, and on an *588essentially per se basis, the trial court concluded that none of the subject areas “are beyond the [ken] of jurors to understand or evaluate and for [defense counsel] to effectively deal with in terms of cross-examination.”7 In light of more recent developments in the scientific community and more recent case law on the subject, more in-depth consideration of such proffers will be necessary before rejecting this type of expert evidence based upon previously accepted common sense notions of what matters are within the ken of the average lay person. Here, as in Benn II, a remand is required to allow the trial court to give individualized consideration to the defense’s proffer in the context of the facts in this case, applying the guidance provided in Benn II. See Benn II, supra, 978 A.2d at 1278-80 (remanding the case to the trial court for proper consideration of the defense’s proffer of expert identification evidence consistent with the opinion).8
The government argues that the trial court’s ruling can be affirmed on the alternative ground that appellant failed to show that the expert had sufficient skill, knowledge, or experience in the field such that it would probably aid the jury in its search for the truth or that the scientific basis for her opinions were not shown adequately. Appellant contends that the government failed to raise these challenges in the trial court, and therefore, they are waived. We are not persuaded on this record that the government either conceded or waived the question of whether appellant’s proffer was sufficient to meet the second and third Dyas factors. It appears that in ruling the proffered evidence inadmissible, the trial court focused only on the first Dyas factor (i.e., whether the subject is so distinctly related to some science as to be beyond the ken of the average layman). See Dyas, supra, 376 A.2d at 832. Having concluded that appellant could not meet the first prong of the Dyas analysis, the trial court apparently found it unnecessary to consider, and did not consider, the second two factors. Appellant will still be required to meet all three of the Dyas criteria before his expert’s evidence can be admitted. Therefore, upon remand, it will be up to the trial court to determine, after appropriate inquiry, whether appellant’s proffered evidence can meet each of the separate criteria identified in Dyas.
Finally, the government argues that even assuming that the trial court abused its discretion by excluding the proffered expert testimony, the error is harmless. Appellant argues that the error is of constitutional magnitude, thereby requiring the government to demonstrate beyond a reasonable doubt that the error was harmless, see Chapman v. California, 386 U.S. 18, 24, 87 S.Ct. 824, 17 L.Ed.2d 705 (1967), but that the government cannot show harmless error even under the less *589stringent standard set forth in Kotteakos v. United States, 328 U.S. 750, 765, 66 S.Ct. 1239, 90 L.Ed. 1557 (1946) (reversal appropriate where it can be concluded “with fair assurance ... that the judgment was not substantially swayed by the error”). There is substantial support for appellant’s argument that an erroneous trial court ruling that deprives a defendant of the opportunity to present non-eumula-tive evidence on an issue central to the defense is of constitutional magnitude and subject to harmless error analysis under Chapman. See, e.g., Harris v. United States, 834 A.2d 106, 127 (D.C.2003) (applying the Chapman standard where the trial court refused to enforce a subpoena for a witness who would have corroborated a part of a defendant’s self-defense theory); Benn v. United States, 801 A.2d 132, 145 (D.C.2002) (Benn I) (applying the Chapman standard where the trial court erroneously denied the defense’s request to recall an alibi witness). The government argues, however, that appellant was not deprived of a meaningful opportunity to present a defense because he fully and forcefully presented his misidentification defense through vigorous cross-examination, and therefore, the Kotteakos standard should be applied.
As appellant points out, the government’s case rested upon the eyewitness identifications of two strangers. There was no scientific evidence linking him to the crime. There was no evidence that appellant was apprehended with the weapon or proceeds of the crime, and he made no statements implicating himself. The proffered expert testimony, if admitted, would have provided a basis upon which the accuracy of the eyewitness’ identifications could have been challenged and made subject to doubt. Without the expert’s testimony, appellant had no factual underpinnings for the scientific theories he sought to present that might cast doubt on the eyewitness’ testimony. Appellant’s proffered scientific theories could not be developed on the cross examination of lay witnesses. See Benn II, supra, 978 A.2d at 1279 (noting the difference between the information that an expert can provide about research studies and that elicited from lay witnesses). In Benn II, we noted that while there may be certain eases where cross-examination might be sufficient to test the reliability of an eyewitness’ identification, it must be borne in mind that “the information that an expert can provide about research studies is different in nature and cannot be elicited from a lay witness during cross examination.” Id. “[Rjeliance on cross-examination cannot be automatic or reflexively adopted in lieu of proffered expert testimony.” Id. In this case, the excluded expert testimony was central to appellant’s mis-identification defense. We are not persuaded that the fact that appellant cross-examined vigorously the eyewitnesses and argued the shortcomings of the identifications to the jury to the extent that he could means that he had a meaningful opportunity to present his misidentification defense. See Scott v. United States, 975 A.2d 831, 839 (D.C.2009) (concluding that the exclusion of telephone records providing independent verification of an otherwise unsubstantiated account of defendant’s friend on a central issue deprived him of a meaningful opportunity to present a defense and not harmless under Chapman or Kotteakos). Different, relevant, and material evidence to the defense in this case was excluded. Assuming error, as the government requests, we cannot conclude that it was harmless under either the Chapman or the Kotteakos standard.
For the foregoing reasons, the judgment is vacated and the case is remanded for further proceedings consistent with this opinion. If the trial court determines on *590reconsideration that the expert testimony should have been permitted, it shall order a new trial. If the trial court adheres to its ruling that the testimony is inadmissible, it shall re-enter the judgment subject to appellant’s right to challenge the ruling in a renewed appeal.
So ordered.