(concurring, with whom Ireland, J., joins). I concur in the result reached by the court, but write separately to express my view on an issue that will arise on remand: the admissibility of expert testimony on the subject of cross-racial eyewitness identification.
In Commonwealth v. Walker, 421 Mass. 90 (1995), we held that a trial judge, in refusing to allow a defendant to present expert testimony concerning the unreliability of cross-racial identifications, “could reasonably have concluded in his discre*154tian that the subject of cross-racial identifications was not ‘one on which the opinion of an expert would have been of assistance to the jury.’ ” Id. at 96, quoting Commonwealth v. Francis, 390 Mass. 89, 98 (1983). Our conclusion rested on citations to two principal authorities: (1) an Appeals Court decision, Commonwealth v. Middleton, 6 Mass. App. Ct. 902 (1978), which reasoned that the rehabihty of cross-racial identifications was not “beyond the ordinary experience and knowledge of the average juror,” and (2) Massachusetts and Federal decisions stating that only “meager data” support the contention that cross-racial identifications. are unreliable.1 See Commonwealth v. Charles, 397 Mass. 1, 8 (1986); United States v. Telfaire, 469 F. 2d 552, 561 (D.C. Cir. 1972) (Leventhal, J., concurring); United States v. Brown, 461 F.2d 134, 145-146 n.l (D.C. Cir. 1971) (Bazelon, C.J., dissenting). In light of extensive scientific research in the field of behavioral psychology, both of these rationales are no longer persuasive as justifications for excluding expert testimony on cross-racial identification.
While the data supporting what social scientists call “own-race bias”2 may have been “meager” when the Telfaire court first weighed in on the theory in the early 1970’s, more than thirty years of research have provided an abundance of rehable data for judicial consideration. “[T]he evidence is now quite clear that people are better able to recognize faces of their own race or ethnic group than faces of another race or ethnic group.” G. L. Wells, Eyewitness Testimony, 54 Ann. Rev. Psychol. 277, 280-281 (2003). This phenomenon affects the vast majority of individuals in all racial groups. See MacLin, Race, Arousal, At-*155tension, Exposure, and Delay: An Examination of Factors Moderating Face Recognition, 7 Psychol., Pub. Pol’y & L. 134, 135 (2001); Meissner, Thirty Years of Investigating the Own-Race Bias in Memory for Faces: A Meta-Analytic Review, 7 Psychol., Pub. Pol’y & L. 3, 4 (2001) (Meissner) (summarizing developments in psychology of own-race bias from 1971 through 2001).
Own-race bias has what experts call a “mirror-effect pattern”: a witness is more likely to make an accurate identification of a person of her own race, but also more likely to make a misidentification of a person of another race. See Meissner, supra at 21. The magnitude of this misidentification is significant: a person is, according to one study, 2.2 times as likely to identify correctly a same-race suspect as a cross-racial suspect. Id. at 26. Moreover, the effect of own-race bias can be significant, particularly for defendant members of minority groups, because “the majority of errors for other-race faces are false alarms, that is, incorrectly identifying an other-race face,” id., while white witnesses are more likely to demonstrate own-race bias, particularly by offering false-alarm responses, id. at 21.3
Modem studies also persuasively undermine the contention in Commonwealth v. Middleton, supra, that the reliability of cross-racial identifications is not “beyond the ordinary experience and knowledge of the average juror.” Id. at 902. Rather, research has shown that “laypersons often predict higher identification accuracy rates than are generally found among participants of eyewitness research” — that is, the average person thinks that eyewitness testimony generally is more accurate than it actually is. Devenport, Eyewitness Identification Evidence, 3 Psychol. Pub. Pol’y & L. 338, 347 (1997). In the more specific context of cross-racial identifications, “laypersons . . . appeared to be insensitive to . . . the impact of cross-racial identifications on *156eyewitness accuracy.” Id. In sum, the unreliability of cross-racial identification is a subject “beyond the ordinary experience and knowledge of the average juror.”
In this case, one of the defendants, an African-American, sought funds to hire an expert to testify concerning the reliability of a cross-racial identification by a white eyewitness. The defendant and the eyewitness were not previously known to each other, and the identification occurred while the witness was riding through the defendant’s neighborhood in a police cruiser looking for the individuals who assaulted and robbed him two days earlier. The motion judge reasoned that the average person would not incur the expense of expert testimony, “given the questionable admissibility of this type of evidence.” It is my view that such evidence should be admissible, because it is relevant, rests on a reliable basis, is sufficiently tied to the facts of this case, and is a subject on which jurors need assistance. See Commonwealth v. Santoli, 424 Mass. 837, 844 (1997) (providing standards governing admissibility of expert testimony). More broadly, it is my view that expert testimony concerning cross-racial identification should generally be admissible to challenge the reliability of a.witness’s identification in circumstances where the identifying witness and the person identified are strangers to each other.4
It bears noting in passing that these two rationales — the “meager data” argument and the “average juror” theory — are in tension with one another. The “average juror” theory suggests that cross-racial identifications are so unreliable that their unreliability is common knowledge, while the “meager data” claim suggests that there is insufficient scientific evidence of unreliability to permit expert testimony.
As used by witness identification experts, the term “own-race bias” does not refer to racial prejudice. Rather, own-race bias describes “the frequently observed performance deficit of one ethnic group in recognizing faces of another ethnic group compared with faces of one’s own group.” Sporer, The Cross-Race Effect: Beyond Recognition of Faces in the Laboratory, 7 Psychol., Pub. Pol’y & L. 170, 170 (2001). Some experts also refer to this phenomenon as “cross-race recognition deficit” and “cross-race effect.” Id.
These “false-alarm” responses are of particular concern in criminal cases, because “the erroneous identification of an individual who is not the perpetrator” is a frequent cause of wrongful conviction. Meissner, Thirty Years of Investigating the Own-Race Bias in Memory for Faces: A Meta-Analytic Review, 7 Psychol., Pub. Pol’y & L. 3, 23 (2001). For example, in one study of convictions that were subsequently overturned when DNA evidence exonerated the defendants, false-alarm identifications were the “primary evidence” of guilt in eighty-five per cent of cases. Id. at 23-24.
Experts in the science of eyewitness identification also tend to support the use of expert testimony in trials involving cross-racial identifications. A recent survey found that ninety per cent of surveyed experts found that the phenomenon of own-race bias was “reliable enough for psychologists to present it in courtroom testimony,” and seventy-two per cent would themselves testify to the existence of own-race bias. Kassin, On the “General Acceptance” of Eyewitness Testimony Research: A New Survey of Experts, 56 Am. Psychologist 405, 412 (2001).