GANTS, C.J.  In the early morning of September 10, 2011, the defendant slashed the face of the victim, Zachary Sevigny, with a box cutter while the victim was sitting in the driver’s seat of his vehicle.  A Superior Court jury found the defendant guilty of mayhem, in violation of G. L. c. 265, § 14; assault and battery by means of a dangerous weapon, in violation of G. L. c. 265, § 15A (b); and breaking and entering a vehicle in the nighttime with the intent to commit a felony, in violation of G. L. c. 266, § 16.[1]  On appeal, the defendant claims that the judge erred by giving the model jury instruction regarding eyewitness identification that we adopted in Commonwealth v. Rodriguez, 378 Mass. 296, 310-311 (Appendix) (1979), rather than the instruction he requested, which would have informed the jury about various scientific principles regarding eyewitness identification.  We conclude that the judge did not err by declining to instruct the jury about these principles where the defendant offered no expert testimony, scholarly articles, or treatises that established that these principles were “so generally accepted that . . . a standard jury instruction stating [those principles] would be appropriate.”  Commonwealth v. Santoli, 424 Mass. 837, 845 (1997), citing Commonwealth v. Hyatt, 419 Mass. 815, 818-819 (1995).  Therefore, we affirm the convictions of mayhem and of breaking and entering.[2]

However, now that we have the benefit of the Report and Recommendations of the Supreme Judicial Court Study Group on Eyewitness Evidence (Study Group Report),[3] and the comments in response to it,[4] we conclude that there are scientific principles regarding eyewitness identification that are “so generally accepted” that it is appropriate in the future to instruct juries regarding these principles so that they may apply the principles in their evaluation of eyewitness identification evidence.  We include as an Appendix to this opinion a provisional jury instruction regarding eyewitness identification evidence, and we invite comments regarding its content and clarity before we declare it a model instruction.[5]  This provisional instruction should be given, where appropriate, in trials that commence after issuance of this opinion until a model instruction is issued.

Background.  At approximately 1:30 A.M. on September 10, 2011, the defendant, who appeared intoxicated, walked into a gasoline station convenience store in Pittsfield, bumped into a customer, Lindsay Holtzman, and asked the employee who was working the cash register, Jordan Wilson, for a box of matches.  Wilson asked the defendant to apologize to Holtzman.  In response, the defendant cursed and stared at Wilson, and challenged him to a fight.  Wilson laughed and gave the defendant a box of matches.  The defendant left the store, but continued to yell at Wilson to meet him outside to fight.

When the defendant left the store, the victim and his friend, Gerald Mortensen, were sitting in the victim’s automobile, parked in a well-lit portion of the store’s parking lot, approximately ten to fifteen feet from the front door of the store.  The victim was in the driver’s seat with the window down.  After the victim and the defendant made eye contact, the defendant approached the vehicle and said to the victim, “What the fuck are you looking at, tough guy?”  The victim responded, “I’m not looking at anything.”  The defendant then pulled a box cutter from his back pocket, reached inside the vehicle with his left arm, and slashed the victim with the blade behind the victim’s ear and down his face.

Mortensen, who was sitting on the passenger’s side, ran inside the store, said that his friend had been cut, and told a store clerk to call for help.  Mortensen and Holtzman then left the store and watched as the defendant walked backwards toward a corner of the parking lot, still staring at the victim.  The victim was taken to a hospital where he received approximately thirty stitches.

On September 15, Wilson went to the Pittsfield police station to meet with Detective Timothy Koenig.  Wilson said that he had seen the person who injured the victim before and could identify him.  Detective Koenig created a pool of 975 archived photographs that fit Wilson’s description of the person.  Wilson used a computer, which displayed twelve photographs per page, to look through the pool.  He eventually selected the defendant’s photograph.  When he made the identification, he reported that he was “110 per cent positive.”

Detective Koenig then created a simultaneous array containing eight photographs, one of which depicted the defendant, and presented the array that same day to Mortensen and the victim separately.[6]  Mortensen stated that none of the photographs showed the assailant.  The victim said that he did not think the assailant “[wa]s anyone in these photos,” but added that if he had to choose somebody, it would be the man with a chin similar to that of the assailant; that man was the defendant.  Holtzman did not view a photographic array, although Detective Koenig attempted to reach her by telephone more than once to do so.

On September 18, Holtzman, Mortensen, and the victim were driving together, and stopped for gasoline at a different service station in Pittsfield.[7]  Holtzman and the victim entered the convenience store while Mortensen stayed inside the victim’s vehicle.  The victim testified that he briefly left the store to retrieve exact change from his automobile to purchase drinks and cigarettes.  When he reentered the store, he immediately saw the defendant and recognized him as the assailant.  After he put down his change, the victim and Holtzman left the store together, and confirmed with each other that the man in the store was the assailant.  The victim then told Mortensen that the defendant was inside the store.  When the defendant left the store, Mortensen agreed that the defendant was the assailant and the victim called the police on his cellular telephone.[8]

As the victim spoke with the police, the defendant left the gasoline station in someone’s automobile.  The victim and Mortensen followed the defendant to an apartment complex in Pittsfield, with the victim communicating the defendant’s location to the police as he was driving.  Shortly thereafter, the police arrived at the apartment complex and asked the victim and Mortensen to perform a showup identification; they identified the defendant as the assailant.  After the defendant’s arrest, Detective Koenig interviewed Holtzman, who confirmed that the person in the convenience store on September 18 had been the same person she saw at the other convenience store on September 10.

Before trial, the defendant filed motions to suppress Holtzman’s and the victim’s pretrial identification of the defendant.[9]  The trial judge denied the motions.  At trial, the defendant argued that he had been mistakenly identified as the assailant, and offered the testimony of his father, Earl Kirchner, who said that he lived with the defendant and that the defendant did not leave his apartment on the evening of the attack.

Discussion.  1.  The defendant’s requested eyewitness identification instruction.  The defendant requested that the judge provide a jury instruction regarding eyewitness identification that essentially mirrored a model instruction that had become effective in New Jersey approximately one week before the defendant’s trial commenced.[10]  The proffered jury instruction was considerably longer and more detailed than the Rodriguez instruction.  It would have instructed the jury on various principles regarding eyewitness identification and human memory, most importantly that (1) human memory does not operate like a video recording that a person can replay to recall what happened;[11] (2) a witness’s level of confidence in an identification may not indicate its accuracy;[12] (3) high levels of stress can reduce the likelihood of making an accurate identification;[13] (4) information from other witnesses or outside sources can affect the reliability of an identification and inflate an eyewitness’s confidence in the identification;[14] and (5) viewing the same person in multiple identification procedures may increase the risk of misidentification.[15]

The judge denied the request and gave an identification instruction consistent with the Rodriguez instruction.  The judge reasoned that the principles included in the defendant’s request were more appropriate for expert testimony or for closing argument.[16]  Furthermore, the judge explained:

“[T]his [proposed instruction] adds facts in.  The process of remembering consists of three stages.  That may be true.  That may not be true.  I have no idea myself but there is no information given to the jury that that is in fact accurate.  So I cannot instruct them as a matter of law that that’s what the law is.”

The defendant objected to the omission of that part of his requested instruction, which recited these five scientific principles, so we review for prejudicial error.  See Commonwealth v. Cruz, 445 Mass. 589, 591 (2005).

The issue before us is not whether the judge had the discretion to give the proffered instruction, but whether he abused his discretion by refusing to do so.  See Hyatt, 419 Mass. at 818-819 (no error in declining to instruct on cross-racial identification, but giving proposed instruction “may be appropriate in the judge’s discretion”).  We conclude that, given the record before him, the judge did not abuse his discretion in denying the defendant’s proposed jury instruction.

We have long recognized that “a principle concerning eyewitness identifications may become so generally accepted that, rather than have expert testimony on the point, a standard jury instruction stating that principle would be appropriate.”  Santoli, 424 Mass. at 845.  See Hyatt, supra (“We recognize that, based on a trial record or on the published results of studies, or both, some new principle concerning the process of eyewitness identification may become sufficiently reliable so as to justify formulating a jury instruction that should be given in particular circumstances on request, in addition to those instructions that we identified in [Rodriguez, 378 Mass. at 310-311,] and Commonwealth v. Pressley, 390 Mass. 617, 619-620 [1983]“).  The defendant here did not provide the judge with any expert testimony, scholarly articles, or treatises that would reasonably have enabled the judge to determine whether the principles in the defendant’s proposed instruction were “so generally accepted” that it would be appropriate to instruct the jury regarding them.[17]  Where the defendant failed to furnish such information, and where there was an instruction approved by this court that was not erroneous but, at worst, inadequate and incomplete, the judge did not abuse his discretion in denying the proffered instruction and charging the jury in accordance with the Rodriguez instruction.  See Cruz, 445 Mass. at 595 n.4, 598, 600 (no error in judge’s refusal to give jury instruction that “there is no proven relationship between a witness’[s] confidence in his identification and the accuracy of the witness’[s] identification” where defendant did not call expert witness and “there was no hearing or testimony regarding the reliability of these scientific studies or their general acceptance in scientific community”); Hyatt, 419 Mass. at 818 (“The defendant points to no relevant empirical study that assessed the relative reliability of cross-racial and non-cross-racial identifications in confrontations of the sort involved here”).

Although we conclude that the judge in this case did not abuse his discretion, and therefore affirm the defendant’s convictions of mayhem and of breaking and entering a vehicle in the nighttime with intent to commit a felony, we take this opportunity to revisit our jurisprudence regarding eyewitness identification jury instructions in general and the Rodriguez instruction in particular.  In Commonwealth v. Walker, 460 Mass. 590, 604 n.16 (2011), we recognized that “eyewitness identification is the greatest source of wrongful convictions but also an invaluable law enforcement tool in obtaining accurate convictions,” and declared our intention to convene the Study Group to consider, among other matters, “whether existing model jury instructions provide adequate guidance to juries in evaluating eyewitness testimony.”  We noted that our creation of the Study Group reflected “our willingness to revisit our jurisprudence” regarding eyewitness identification evidence.  Id. at 606.  With the Study Group Report completed and the comments to that report received, it is now time to do what we declared we were willing to do with respect to eyewitness identification jury instructions.[18]

2.  Model jury instruction.  The Rodriguez instruction derives from the model set forth in United States v. Telfaire, 469 F.2d 552, 555 (D.C. Cir. 1972), which recognized the “special problems” with the reliability of eyewitness identifications and the need for an identification instruction that “emphasizes to the jury the need for finding that the circumstances of the identification are convincing beyond a reasonable doubt.”  See Rodriguez, 378 Mass. at 302.  We adopted the Telfaire model “to assist a jury in evaluating the reliability of a positive identification of the defendant as the perpetrator of the crime by a witness.”  Commonwealth v. Franklin, 465 Mass. 895, 910 (2013).  Over time, we have modified and supplemented it.  See Commonwealth v. Cuffie, 414 Mass. 632, 640 (1993) (removing language that risked suggesting that witness’s first sighting of offender was always accurate); Santoli, 424 Mass. at 845 (omitting language emphasizing “strength of the identification”).  See also Pressley, 390 Mass. at 620 (establishing supplemental instruction on “possibility of an honest but mistaken identification”); Franklin, 465 Mass. at 912 (judge should provide, on request, identification instruction where eyewitness gave partial identification).  At its core, though, the Rodriguez instruction delineates factors for the jury to consider when evaluating an eyewitness identification, such as (1) the opportunity the witness had to observe the offender; (2) the length of time between the crime and the identification; (3) the witness’s prior familiarity with the offender; (4) the circumstances surrounding any identification procedure; (5) whether the identification procedure was a lineup or photographic array rather than a single-person showup; (7) whether the witness failed to make an identification or made an inconsistent identification before identifying the defendant; and (8) the credibility of the witness.[19]  It focuses the jury on factors they “should consider” that may affect the accuracy of an eyewitness’s positive identification of the defendant, and poses questions the jury should ask themselves.  It generally does not instruct the jury as to how those factors may affect the accuracy of the identification.

The New Jersey model instruction, as earlier noted, goes well beyond the Rodriguez instruction by telling the jury what principles have emerged from the research regarding eyewitness identification.  We now consider, first, what it means for a principle of eyewitness identification to be “so generally accepted” that it is appropriate to include in a model instruction, and, second, whether the five principles at issue in this case are “so generally accepted” that it is appropriate that they now be included in a revised model jury instruction.

a.  “So generally accepted.”  The phrase “so generally accepted” sounds like the test in Frye v. United States, 293 F. 1013, 1014 (D.C. Cir. 1923), for the admissibility of expert testimony based on scientific knowledge, which asks “whether the community of scientists involved generally accepts the theory or process,” Commonwealth v. Lanigan, 419 Mass. 15, 24 (1994), quoting Commonwealth v. Curnin, 409 Mass. 218, 222 (1991), and which was once the exclusive test governing the admissibility of expert testimony.  See Lanigan, supra at 25-26 (adopting standard in Daubert v. Merrell Dow Pharms., Inc., 509 U.S. 579 [1993], while maintaining Frye standard as alternative means to establish reliability of expert testimony).  But satisfaction of the Frye test meant only that expert testimony would be admissible in evidence.  It did not mean that the jury were required to accept the scientific principles that had gained general acceptance in the relevant scientific community.  See Commonwealth v. Hinds, 450 Mass. 1, 12 n.7 (2007) (model instruction on expert testimony, stating, “it is completely up to [the jury] to decide whether [they] accept the testimony of an expert witness, including the opinions that the witness gave”).  In contrast, where a principle is included in a jury instruction, it becomes part of a judge’s instructions of law, which the jury generally must accept.  See Commonwealth v. Johnson, 441 Mass. 1, 7 (2004); Commonwealth v. Watkins, 425 Mass. 830, 840 (1997) (“We presume that a jury follow all instructions given to [them] . . .”).  Therefore, the Frye test cannot define “so generally accepted” in this context; the standard for including a principle of eyewitness identification in a model jury instruction must be higher than a standard that would simply permit a judge to admit expert testimony.[20]

To determine when a principle of eyewitness identification is “so generally accepted” that it is appropriate to incorporate into a model instruction, we focus on the instruction’s underlying purpose and the concerns it is intended to alleviate.  The accuracy of an eyewitness identification is often the critical issue in a criminal case, the difference between a conviction and an acquittal.  See State v. Cabagbag, 127 Haw. 302, 313 (2012) (“Without appropriate instructions from the court, the jury may be left without sufficient guidance on how to assess critical testimony, sometimes the only testimony, that ties a defendant to an offense”).  We have long recognized that the mistaken eyewitness identification of a defendant whom the witness had never seen before the crime “is the primary cause of erroneous convictions, outstripping all other causes combined.”  Commonwealth v. Martin, 447 Mass. 274, 293 (2006) (Cordy, J., dissenting).[21]  See Franklin, 465 Mass. at 909; Irwin v. Commonwealth, 465 Mass. 834, 848-849 (2013); Commonwealth v. Francis, 390 Mass. 89, 100 (1983).

Our jury instructions are intended to provide the jury with the guidance they need to capably evaluate the accuracy of an eyewitness identification.  See Francis, 390 Mass. at 101 (“We permit, indeed require, the judge to instruct the jury concerning factors that bear on the reliability of eyewitness identification”); Commonwealth v. Rodriguez, 6 Mass. App. Ct. 738, 742 (1978), S.C., 378 Mass. 296 (1979).  If we were to define “so generally accepted” so narrowly that none of the scientific principles regarding eyewitness identification could survive the test, we would continue to use the Rodriguez instruction, which generally identifies factors a jury may consider in applying their common sense, and would require the results of the relevant research to be communicated to the jury solely through expert testimony, where such testimony is offered.  The problem with this approach is that the research makes clear that common sense is not enough to accurately discern the reliable eyewitness identification from the unreliable, because many of the results of the research are not commonly known, and some are counterintuitive.  See State v. Guilbert, 306 Conn. 218, 234-235 (2012) (there is “near perfect scientific consensus” that “eyewitness identifications are potentially unreliable in a variety of ways unknown to the average juror”); Henderson, 208 N.J. at 274 (juror surveys and mock-jury studies “reveal generally that people do not intuitively understand all of the relevant scientific findings”).  See also Perry v. New Hampshire, 132 S. Ct. 716, 739 (2012) (Sotomayor, J., dissenting) (“Study after study demonstrates that eyewitness recollections are highly susceptible to distortion by postevent information or social cues; that jurors routinely overestimate the accuracy of eyewitness identifications; that jurors place the greatest weight on eyewitness confidence in assessing identifications even though confidence is a poor gauge of accuracy; and that suggestiveness can stem from sources beyond police-orchestrated procedures” [footnotes omitted]).[22]  If the research regarding eyewitness identification could be communicated to the jury only through expert testimony, very few juries would hear it, because expert testimony is not often proffered in cases where eyewitness identification is at issue, and because the admission of expert testimony is left to the sound discretion of the trial judge.  See Commonwealth v. Watson, 455 Mass. 246, 257 (2009) (“[E]xpert testimony concerning the reliability of eyewitness identification is not admissible as of right, but is left to the discretion of the trial judge”).

Having balanced the importance of instructing juries about the generally accepted principles that can inform their understanding of eyewitness identification with the risks of requiring them to accept principles that may still be suspect or in flux, we conclude that a principle is “so generally accepted” that it is appropriate to include in a model eyewitness identification instruction where there is a near consensus in the relevant scientific community adopting that principle.

After reviewing the scholarly research, analyses by other courts, amici submissions, and the Study Group Report and comments, we conclude that there are various principles regarding eyewitness identification for which there is a near consensus in the relevant scientific community and that it is appropriate to revise the Rodriguez instruction to include them.  See Study Group Report, supra at 17 (“The scientific studies have produced a consensus among experts about the . . . variables that have been shown to affect the reliability of eyewitness identification”).  See also Guilbert, 306 Conn. at 234-236; Cabagbag, 127 Haw. at 310-311; State v. Lawson, 352 Or. 724, 740 (2012); State v. Clopten, 223 P.3d 1103, 1108 (Utah 2009); Report of the Special Master, State vs. Henderson, N.J. Supreme Ct., No. A-8-08, at 14 (June 18, 2010), available at http://www.judiciary.state.nj.us/pressrel/henderson%20final%20brief%20.pdf%20%2800621142%29.pdf [http://perma.cc/EA3S-453F] (last visited Jan. 8, 2015) (Special Master’s Report).[23,24]

We are not alone in concluding that certain scientific principles should be incorporated into a model jury instruction on eyewitness identification.  New Jersey has done so most comprehensively, promulgating a ten-page model instruction after concluding that its previous model, which was similar to the Rodriguez instruction, see Henderson, 208 N.J. at 226-227, “overstate[d] the jury’s inherent ability to evaluate evidence offered by eyewitnesses who honestly believe their testimony is accurate.”  Id. at 218, 298-299.  See National Research Council of the National Academies, Identifying the Culprit:  Assessing Eyewitness Identification 28 (2014) (pending publication) (National Academies) (“The New Jersey instructions adopted, following the Henderson decision, are by far the most detailed set of jury instructions regarding eyewitness identification evidence”).  Other States have also incorporated scientific principles of eyewitness identification into model jury instructions.  See, e.g., Cabagbag, 127 Haw. at 314; Connecticut Criminal Jury Instruction 2.6-4 Identification of Defendant (2013), available at http://www.jud.ct.gov/ji/criminal/part2/2.6-4.htm [http://perma.cc/B9PS-DS8X] (last visited Jan. 8, 2015); 1-6 Maine Jury Instruction Manual § 6-22A (4th ed. 2012); Model Utah Jury Instructions, Second Edition, CR404 Eyewitness Identification (2014), available at http://www.utcourts.gov/resources/muji/index.asp?page=crim&view=all_crim [http://perma.cc/X9V3-2759] (last visited Jan. 8, 2015).

We recognize that even a principle for which there is near consensus is subject to revision based on further research findings, and that no principle of eyewitness identification should be treated as if set in stone.  Therefore, we acknowledge the possibility that, as the science evolves, we may need to revise our new model instruction’s description of a principle.[25]  We also recognize the possibility that a party may offer expert testimony at trial that properly may persuade a trial judge to depart from the model instruction.  See Lawson, 352 Or. at 741 (“[A]cknowledgment of the existence of th[is] research . . . is not intended to preclude any party in a specific case from validating scientific acceptance of further research or from challenging particular aspects of the research described in this opinion”).

b.  Five generally accepted principles regarding eyewitness identification.  We turn now to the five principles at issue in this case that we determine to have achieved a near consensus in the relevant scientific community and therefore are “so generally accepted” that it is appropriate that they now be included in a revised model jury instruction regarding eyewitness identification.  We also summarize the research that informed our conclusions as to each generally accepted principle.[26]

i.  Human memory does not function like a video recording but is a complex process that consists of three stages: acquisition, retention, and retrieval.  The central principle that has emerged from over 2,000 published studies over the past thirty years is that “memory does not function like a videotape, accurately and thoroughly capturing and reproducing a person, scene or event. . . .  Memory is, rather[,] a constructive, dynamic and selective process.”  Study Group Report, supra at 15, quoting Special Master’s Report, supra at 9.  See E.F. Loftus, J.M. Doyle, & J.E. Dysart, Eyewitness Testimony:  Civil and Criminal § 2-2, at 14 (5th ed. 2013); Brigham, Wasserman, & Meissner, Disputed Eyewitness Identification Evidence:  Important Legal and Scientific Issues, 36 Ct. Rev., no. 2, 1999, at 13.  Rather, memories are made through a three-stage process:  “acquisition — ‘the perception of the original event’; retention [or storage] — ‘the period of time that passes between the event and the eventual recollection of a particular piece of information’; and retrieval — the ‘stage during which a person recalls stored information.’”  Study Group Report, supra at 16, quoting Henderson, 208 N.J. at 245.

ii.  An eyewitness’s expressed certainty in an identification, standing alone, may not indicate the accuracy of the identification, especially where the witness did not describe that level of certainty when the witness first made the identification.  We have long questioned the reliability of a witness’s certainty as a reflection of accuracy.  See Commonwealth v. Jones, 423 Mass. 99, 110 n.9 (1996); Santoli, 424 Mass. at 846 (“[T]here is significant doubt about whether there is any correlation between a witness’s confidence in her identification and the accuracy of her recollection”); Cruz, 445 Mass. at 597-600 (court stated it was prepared to consider in future whether weak confidence-accuracy relationship warrants instruction).  Our doubts are now supported by the research.  “[S]tudies show that, under most circumstances, witness confidence or certainty is not a good indicator of identification accuracy.”  Lawson, 352 Or. at 777 (Appendix).  See Study Group Report, supra at 19.[27]

This does not mean that eyewitness certainty is never correlated with accuracy; it means simply that the existence and strength of the correlation depends on the circumstances.  After viewing the crime but before the identification procedure, an eyewitness’s expressed level of certainty does not correlate with accuracy.  See Study Group Report, supra; Henderson, 208 N.J. at 254 n.7.[28]  Where an eyewitness makes a positive identification and expresses a level of certainty immediately after the identification procedure, there is some correlation between certainty and accuracy, but there is not yet a near consensus regarding the strength of that correlation.[29]  There is, however, a near consensus in the research that, where an eyewitness during an identification procedure did not express certainty when first asked to make an identification, a subsequent claim of certainty by that witness deserves little weight in evaluating the accuracy of that identification.  See Henderson, 208 N.J. at 254 (“Confirmatory feedback can distort memory.  As a result, to the extent confidence may be relevant in certain circumstances, it must be recorded in the witness’[s] own words before any possible feedback”); Lawson, 352 Or. at 745 (“Retrospective self-reports of certainty are highly susceptible to suggestive procedures and confirming feedback, a factor that further limits the utility of the certainty variable”).[30,31]

Although the research regarding the correlation (or lack of correlation) between eyewitness certainty and accuracy is complex and still evolving, it is necessary to inform a jury about this tenuous relationship because there is a near consensus that jurors tend to give more weight to a witness’s certainty in evaluating the accuracy of an identification than is warranted by the research.  See Commonwealth v. Collins, ante 255, 264 n.14 (2014), quoting Study Group Report, supra at 20 (“Studies show that eyewitness confidence is the single most influential factor in juror determinations regarding the accuracy of an eyewitness identification”); Cabagbag, 127 Haw. at 311; Clopten, 223 P.3d at 1108 (“Indeed, juries seemed to be swayed the most by the confidence of an eyewitness, even though such confidence correlates only weakly with accuracy”).[32]  Therefore, it is necessary to inform the jury that an eyewitness’s expressed certainty in an identification, standing alone, may not indicate the accuracy of an identification, and that this is especially true where the witness did not describe that level of certainty when the witness first made an identification.

iii.  High levels of stress can reduce an eyewitness’s ability to make an accurate identification.  “[A]n eyewitness under high stress is less likely to make a reliable identification of the perpetrator.”  Special Master’s Report, supra at 43.  “[H]igh levels of stress significantly impair a witness’s ability to recognize faces and encode details into memory.”  Lawson, 352 Or. at 769 (Appendix).  There is “considerable support for the hypothesis that high levels of stress negatively impact both accuracy of eyewitness identification as well as accuracy of recall of crime-related details.”  Deffenbacher, Bornstein, Penrod, & McGorty, A Meta-Analytic Review of the Effects of High Stress on Eyewitness Memory, 28 Law & Hum. Behav. 687, 699 (2004) (Deffenbacher et al.).  See Study Group Report, supra at 29 n.27, citing Deffenbacher et al., supra at 695 (thirty-nine per cent of participants under high-stress conditions correctly identified suspect in target-present lineups compared to fifty-nine per cent of participants under low-stress conditions).[33]  This principle is counterintuitive to the “common misconception that faces seen in highly stressful situations can be ‘burned into’ a witness’s memory.”  Lawson, 352 Or. at 770 (Appendix).  See Morgan, Hazlett, Doran, Garrett, Hoyt, Thomas, Baranoski, & Southwick, Accuracy of Eyewitness Memory for Persons Encountered During Exposure to Highly Intense Stress, 27 Int’l J.L. & Psychiatry 265, 274 (2004) (rejecting “popular conception that most people would never forget the face of a clearly seen individual who had physically confronted them and threatened them”).  Therefore, it is important to inform the jury of this principle lest they evaluate an identification made under high stress based on the “common misconception.”

iv.  Information that is unrelated to the initial viewing of the event, which an eyewitness receives before or after making an identification, can influence the witness’s later recollection of the memory or of the identification.  “An extensive body of studies demonstrates that the memories of witnesses for events and faces, and witnesses’ confidence in their memories, are highly malleable and can readily be altered by information received by witnesses both before and after an identification procedure.”  Special Master’s Report, supra at 30-31.  See B.L. Garrett, Convicting the Innocent:  Where Criminal Prosecutions Go Wrong 48-49 (2011) (reviewing trial records for 161 wrongful convictions involving eyewitness misidentification and finding that seventy-eight per cent involved police contamination of identification).  This outside information, known as “feedback,” affects witnesses’ memory differently depending on whether the witness receives feedback before or after making an identification.  See Study Group Report, supra at 21-22; Henderson, 208 N.J. at 253.  “Jurors, however, tend to be unaware of . . . how susceptible witness certainty is to manipulation by suggestive procedures or confirming feedback.”  Lawson, 352 Or. at 778 (Appendix).

Preidentification feedback may contaminate the witness’s memory.  For instance, suggestive wording and leading questions prior to participating in an identification procedure can influence the process of forming a memory.  See Study Group Report, supra at 21; Lawson, 352 Or. at 786-788 (Appendix).[34]  Postidentification feedback is information unrelated to the witness’s actual memory that suggests to the witness that he or she correctly identified the suspect.  See Study Group Report, supra at 22; Henderson, 208 N.J. at 255; Lawson, 352 Or. at 744.  This confirmatory information may boost the witness’s level of certainty without increasing the likelihood of an accurate identification.  See Lawson, supra; Special Master’s Report, supra at 33 (“A number of studies have demonstrated that witnesses’ confidence in their identifications, and their memories of events and faces, are readily tainted by information that they receive after the identification procedure”).[35]

Although police officers are common potential sources of feedback, feedback from cowitnesses and other private actors can also influence a witness’s memory.  “When a witness is permitted to discuss the event with other witnesses or views another witness’s identification decision, the witness may alter his or her own memory or identification decision to conform to that of the cowitness.”  Lawson, 352 Or. at 788 (Appendix).  See Henderson, 208 N.J. at 268-271.[36]

v.  A prior viewing of a suspect at an identification procedure may reduce the reliability of a subsequent identification procedure in which the same suspect is shown.  A prior viewing of a suspect in an identification procedure raises doubts about the reliability of a subsequent identification procedure involving the same suspect.  See Study Group Report, supra at 25, citing Special Master’s Report, supra at 27-28.  “[S]uccessive views of the same person can make it difficult to know whether the later identification stems from a memory of the original event or a memory of the earlier identification procedure.”  Henderson, 208 N.J. at 255.  See Collins, supra at 262 n.9, citing Study Group Report, supra at 78-79 (“An eyewitness may recall the defendant’s face, but not recall that the source of the eyewitness’s memory was the defendant’s presence in a pretrial lineup or photographic array rather than the defendant’s presence at the scene of the crime”); Commonwealth v. Scott, 408 Mass. 811, 826 (1990) (“danger of misidentification is increased if the photograph of the same individual is included in different arrays”); Lawson 352 Or. at 784 (Appendix).

One form of this source memory problem is “mugshot exposure,” where a witness’s viewing of an innocent suspect’s mugshot can heighten the chances of a later misidentification.  See Study Group Report, supra at 25, citing Henderson, supra at 256.  A meta-analysis of eleven published articles showed that “prior mugshot exposure decreases accuracy at a subsequent lineup, both in terms of reductions in rates for hits and correct rejections as well as in terms of increases in the rate for false alarms.”  Deffenbacher, Bornstein, & Penrod, Mugshot Exposure Effects:  Retroactive Interference, Mugshot Commitment, Source Confusion, and Unconscious Transference, 30 Law & Hum. Behav. 287, 306 (2006).  See id. at 299 (fifteen per cent of subject witnesses misidentified innocent person in lineup when seeing person for first time, while thirty-seven per cent of witnesses with mugshot exposure misidentified innocent person).[37]

c.  Provisional model jury instruction.  After evaluating the scientific evidence and concluding that the aforementioned principles are so generally accepted that they may be stated in a model jury instruction, we propose in the Appendix to this opinion a new provisional jury instruction regarding eyewitness identification.  We have made the jury instruction provisional to allow for public comment and possible future revision before we declare it a model, but it should be given, where appropriate, in trials that commence after issuance of this opinion until a model instruction is issued.  We intend the new instruction to have no retroactive application.  See Santoli, 424 Mass. at 845 (declining retroactively to apply new rule to omit “strength of the identification” language).  See also Commonwealth v. Ashley, 427 Mass. 620, 628 (1998) (declining retroactively to apply Santoli); Commonwealth v. Payne, 426 Mass. 692, 698 (1998) (same).

Our provisional instruction updates the Rodriguez instruction with principles relevant to the evaluation of eyewitness testimony for which there is at least a near consensus in the relevant scientific community.  It will provide juries with more comprehensive guidance to evaluate and weigh eyewitness identifications, but we recognize that not every principle regarding eyewitness identification that has attained a near consensus in the relevant scientific community is included; nor are the included principles set forth in great detail.  We aspired in drafting the instruction for clarity, brevity, and balance, recognizing that an eyewitness identification instruction is only one of many instructions in a jury charge.  We also understand that the longer the jury instruction, the greater the risk that it will implicitly communicate the message that all eyewitness identifications should be viewed as unreliable rather than simply evaluated with caution and care, so we have balanced this risk with the need to educate jurors.[38]  See National Academies, supra at 29 (noting concern that “jury instructions cause jurors to become more suspicious of all eyewitness identification evidence”).  The provisional instruction is longer than the Rodriguez instruction, but it will be the rare case where the entirety of the instruction need be given, because a judge need only give the portions of the provisional instruction that are relevant to the eyewitness identification evidence involved in the case.

We expect the new model instruction will provide at least one source of reliable information in cases where expert testimony is not offered.  Jury instructions offer certain advantages over expert testimony:  “they are focused and concise, authoritative (in that juries hear them from the trial judge, not a witness called by one side), and cost-free; they avoid possible confusion to jurors created by dueling experts; and they eliminate the risk of an expert invading the jury’s role or opining on an eyewitness’[s] credibility.”  Henderson, 208 N.J. at 298.  See United States v. Jones, 689 F.3d 12, 19 (1st Cir. 2012).  But see Clopten, 223 P.3d at 1110 (research “has shown that a cautionary instruction does little to help a jury spot a mistaken identification”).

Nevertheless, our provisional instruction is not intended in any way to preclude expert testimony regarding eyewitness identification or to discourage judges from exercising their discretion to permit such expert testimony.  Cf. Clopten, supra at 1107 (“It was never the intent of this court to establish cautionary instructions as the sole means for educating juries about eyewitness fallibility”).  Expert testimony may be important to elaborate on the generally accepted principles in a model instruction and to explain how other variables relevant to the particular case can affect the accuracy of the identification.  A judge may also allow an expert to challenge the generally accepted principles we incorporated, and, where the judge finds the expert’s challenge to be persuasive, the judge may modify the model instruction accordingly.  See part 2.a, supra.

Conclusion.  In the circumstances of this case, based on the record before him, the judge did not abuse his discretion in declining to give the New Jersey model jury instruction regarding eyewitness identification and instead giving the Rodriguez instruction.  Therefore, we affirm the defendant’s judgments of conviction of mayhem and breaking and entering a motor vehicle in the nighttime with intent to commit a felony.  We remand the case to the Superior Court to vacate the defendant’s judgment of conviction and sentence for assault and battery by means of a dangerous weapon as duplicative of the mayhem conviction.  Because the sentence to be vacated was less than the sentence of mayhem, and was ordered to be served concurrent with that sentence, we do not order resentencing of the defendant.

So ordered.

Appendix.[1]

One of the most important issues in this case is whether the defendant is the person who committed [or participated in the commission of] the crime[s].  The Commonwealth has the burden of proving beyond a reasonable doubt that this defendant was in fact the perpetrator of the crime[s] alleged in the indictment[s].

The identification of the defendant as the person who committed [or participated in the commission of] the crime[s] may be proved by direct evidence or circumstantial evidence, or by some combination of direct and circumstantial evidence, but it must be proved beyond a reasonable doubt.  If you are not convinced beyond a reasonable doubt that the defendant is the person who committed [or participated in the commission of] the crime[s], you must find the defendant not guilty.

In evaluating eyewitness identification testimony, it is not essential that a witness be free from doubt as to the correctness of his or her identification of the defendant.  However, you, the jury, must be satisfied beyond a reasonable doubt, based on all of the credible evidence, that this defendant is the person who committed [or participated in the commission of] the crime[s] before you may convict him/her.

As with any witness, you must determine the credibility of a witness identifying the defendant as the offender.  If you conclude that the witness is not telling the truth regarding the person’s identification, you shall disregard that testimony.  If you conclude that the witness intended to tell the truth, you must also consider the possibility that the witness made a good faith error in identification.  That is, you should consider whether the witness could be honestly mistaken in his or her identification of the defendant.

Human beings have the ability to recognize other people from past experiences and to identify them at a later time, but research has shown that people sometimes make mistakes in identification.  That research has focused on the factors that may affect the accuracy of an identification, including the nature of human memory.

Research has shown that human memory is not like a video recording that a witness need only replay to remember what happened.[2]  Memory is far more complex.  The process of remembering consists of three stages:  first, a person sees or otherwise acquires information about the original event; second, the person stores in the brain the information about the event for a period of time until, third, the person attempts to recall that stored information.[3]  At each of these stages, memory can be affected by a variety of factors.[4]

Relying on some of the research that has been done in this area, I am going to list some specific factors you should consider in determining whether the identification testimony is accurate.  By instructing you on the factors to consider, I am not expressing any opinion about the accuracy of any specific memory of any particular witness.  You, the jury, must decide whether the witness’s identification is accurate.

(1)  The witness’s opportunity to view the event.  You should consider the opportunity the witness had to observe the offender at the time of the offense, how good a look the witness had of the offender, the degree of attention the witness was paying to the offender at that time, the distance between the witness and the offender, how good the lighting conditions were, and the length of time the witness had to observe the offender;

ADD ONLY IF RELEVANT TO THE EVIDENCE IN THE CASE:

[IF DISGUISE WAS INVOLVED OR FACE WAS OBSCURED] whether the offender was disguised or had his/her features obscured in some way;[5] [IF PERPETRATOR HAD DISTINCTIVE FACE OR FEATURE] whether the perpetrator had a distinctive face or feature;[6] [IF A WEAPON WAS INVOLVED] and whether the witness saw a weapon during the event — the visible presence of a weapon may reduce the reliability of an identification if the crime is of short duration, but the longer the event, the more time the witness has to adapt to the presence of the weapon.[7]

(2)  Characteristics of the witness.  You should also consider characteristics of the witness when the observation was made, such as the quality of the witness’s eyesight, whether the witness knew the offender, and, if so, how well,[8] and whether the witness was under a high degree of stress — high levels of stress, compared to low to medium levels, can reduce an eyewitness’s ability to accurately perceive an event;[9]

ADD ONLY IF RELEVANT TO THE EVIDENCE IN THE CASE:

[IF DRUGS OR ALCOHOL WERE INVOLVED] whether the witness at the time of the observation was under the influence of alcohol or drugs, and if so, to what degree;

[IF WITNESS AND OFFENDER ARE OF DIFFERENT RACES] and whether the witness and the offender are of different races — research has shown that people of all races may have greater difficulty in accurately identifying members of a different race than they do in identifying members of their own race.[10]

(3)  The time elapsed.  You should consider how much time elapsed between the event observed and the identification.  Generally, memory is most accurate right after the event and begins to fade thereafter.[11]

(4)  Witness’s expressed certainty.  Research shows that a witness’s expressed certainty in an identification, standing alone, may not be a reliable indicator of the accuracy of the identification,[12] especially where the witness did not describe that level of certainty when the witness first made the identification.[13]

(5)  Exposure to identification information from others.  A person’s memory may be affected by information the person received between the incident and the identification,[14] as well as after the identification,[15] and the person may not realize that his or her memory has been affected.[16]  You may consider whether the witness was exposed to identifications made by other witnesses, to opinions or descriptions given by others, including police officers, or to any other information or influence.[17]  Such exposure may affect the independence and reliability of a witness’s identification, and may inflate the witness’s confidence in the identification.[18]

An identification that is the product of some suggestive conduct by the police or others should be scrutinized with special caution and care.  The risk that suggestion will affect the identification is greater where the witness did not get so good a look at the offender, because a witness who got a good look is less likely to be influenced by suggestion.[19]

ADD ONLY IF RELEVANT TO THE EVIDENCE IN THE CASE:

[IF THERE WAS A PHOTOGRAPHIC ARRAY OR LINEUP] An identification may occur as part of the police investigation through the showing of an array of photographs or through a lineup of individuals.  You may take into account that any identification that was made by picking the defendant out of a group of similar individuals is generally more reliable than one which results from the presentation of the defendant alone to a witness.

You should consider whether the police in conducting the photographic array or lineup followed established or recommended procedures that are designed to diminish the risk of suggestiveness.[20]  If there was evidence that any of those procedures were not followed, you should evaluate the identification with particular care and consider whether the failure to follow the procedure affected the reliability of the identification.

Where a witness identified the defendant in a photographic array [or in a lineup], you should consider the number of photographs in the array [or individuals in the lineup],[21] whether there was anything about the defendant’s photograph [or the defendant’s appearance in the lineup] that made him/her stand out from the others,[22] whether the person administering the photographic array [or lineup] did not know who was the suspect and therefore could not influence the witness’s identification,[23] and whether anything was said to the witness that would suggest that the suspect was among the persons shown in the photographic array [or lineup], or that would suggest that the witness should identify the suspect.[24] [IF THERE WAS A SHOWUP]  An identification may occur as part of the police investigation through what is known as a showup, where a suspect is shown alone to a witness.  An identification procedure in which a witness selects a person from a group of similar individuals in a photographic array or a lineup is generally less suggestive than a showup, which is to some degree inherently suggestive.[25]  You should consider how long after the initial event the showup took place, as a fresh memory of an event that occurred only a few hours earlier may reduce the risks arising from the inherently suggestive nature of a showup.[26]

You should consider whether the police, in conducting the showup, followed established or recommended procedures that are designed to diminish the risk of suggestiveness.  If any of those procedures were not followed, you should evaluate the identification with particular care and consider whether the failure to follow the procedure affected the reliability of the identification.

ADD ONLY IF RELEVANT TO THE EVIDENCE IN THE CASE:

[IF THERE WERE MULTIPLE VIEWINGS BY THE SAME WITNESS]  You should consider whether the witness viewed the defendant in multiple identification procedures or events.  When a witness views the same person in more than one identification procedure or event, it may be difficult to know whether a later identification comes from the witness’s memory of the actual, original event, or from the witness’s observation of the person at an earlier identification procedure or event.[27]

(6)  Failure to identify or inconsistent identification.  You may take into account whether a witness ever tried and failed to make an identification of the defendant, or made an identification that was inconsistent with the identification that such witness made at trial.

(7)  Totality of the evidence.  You should consider all the relevant factors that I have discussed, viewed in the context of the totality of the evidence in this case, in evaluating the accuracy of a witness’s identification testimony.  Specifically, you should consider whether there was other evidence in the case, direct or circumstantial, that tends to support or not to support the accuracy of an identification.  If you are not convinced beyond a reasonable doubt that the defendant was the person who committed [or participated in the commission of] the crime[s], you must find the defendant not guilty.

footnotes

[1] The judge sentenced the defendant to concurrent State prison terms of from eight to twelve years on the mayhem conviction, from seven to ten years on the conviction of assault and battery by means of a dangerous weapon, and from three to five years for breaking and entering a vehicle in the nighttime with intent to commit a felony.

[2] We vacate the defendant’s conviction and sentence on the charge of assault and battery by means of a dangerous weapon.  The defendant contends on appeal that his convictions of mayhem and for assault and battery by means of a dangerous weapon were based on the same conduct, the defendant’s slashing of the victim’s face, and that the convictions are duplicative because assault and battery by means of a dangerous weapon is a lesser included offense of the theory of mayhem presented to the jury.  The Commonwealth agrees that the convictions are duplicative, and so do we.  “A crime is a lesser-included offense of another crime if each of its elements is also an element of the other crime.”  Commonwealth v. Martin, 425 Mass. 718, 722 (1997), quoting Commonwealth v. Perry, 391 Mass. 808, 813 (1984).  “Mayhem (second theory) is essentially an assault and battery by means of a dangerous weapon, with the additional aggravating factors of a specific intent to maim or disfigure, and certain forms of resultant physical injury.  Therefore, the latter is a lesser included offense of the former.”  Martin, supra.  See Commonwealth v. Ogden O., 448 Mass. 798, 808 (2007).  “The appropriate remedy for the imposition of duplicative convictions is to vacate both the conviction and sentence on the lesser included offense, and to affirm the conviction on the more serious offense.”  Commonwealth v. Mello, 420 Mass. 375, 398 (1995).

[3] See Supreme Judicial Court Study Group on Eyewitness Evidence:  Report and Recommendations to the Justices (July 25, 2013) (Study Group Report), available at http://www.mass.gov/courts/docs/sjc/docs/eyewitness-evidence-report-2013.pdf [http://perma.cc/WY4M-YNZN] (last visited Jan. 8, 2015).

[ 4] The comments in response to the Study Group Report can be found at http://www.mass.gov/courts/docs/sjc/docs/eyewitness-evidence-report-comments.pdf [http://perma.cc/UF62-STVZ] (last visited Jan. 8, 2015).

[5] We acknowledge the amicus briefs submitted by the Innocence Network; the American Psychological Association and the Center for Law, Brain & Behavior; the District Attorney for the Suffolk District; and the Massachusetts Association of Criminal Defense Lawyers.

[6] Before presenting the array to both Gerald Mortensen and the victim, Detective Timothy Koenig read nine advisements to the witnesses:  (1) “I am going to show you a group of photos that are in random order”; (2) “[t]he person who committed the crime may or may not be included, so you should not feel compelled to make an identification”; (3) “[i]t is just as important to clear innocent people as it is to identify possible perpetrators”; (4) “[w]hether or not you identify someone, the police will continue to investigate”; (5) “[a]fter you are done, I will not be able to provide you with any feedback or comments on the results of the process”; (6) “[p]lease do not discuss this identification procedure, or the results, with other witnesses in this case or with the media”; (7) “[p]eople may not appear exactly as they did at the time of the event, because features such as clothing or head/facial hair are subject to change”; (8) “[a]s you look at each photo, if you see someone that you recognize, please tell me how you know the person, and in your own words, how sure you are of the identification”; and (9) “[i]f you identify someone, I will ask you to place your initials and the date on a form I will give you, clearly marking your selection.”

[7] Lindsay Holtzman did not know the victim or Mortensen before the incident on September 10.

[8] Holtzman’s recollection of this event differed slightly from that of the victim.  She testified that she recognized the defendant standing three or four feet behind her in the store.  She said something to the victim and may have nudged his arm to alert him to the defendant’s presence.  The victim then left the store while she waited in line and finished making her purchases.  After leaving the store, she walked over to the vehicle as the victim was telephoning the police.

[9] No motion was filed with respect to Mortensen.

[10] In State v. Henderson, 208 N.J. 208, 219, 228-229 (2011), the New Jersey Supreme Court, having earlier remanded the case to a special master who considered more than 200 published scientific studies on human memory and eyewitness identification during a ten-day hearing, rendered a landmark decision regarding eyewitness identification where it concluded that “the court system should develop enhanced jury charges on eyewitness identification for trial judges to use.”  The court delegated to its criminal practice committee and committee on model criminal jury charges the task of drafting the revised model jury instructions.  Id. at 298-299.  On July 19, 2012, the court released the model instructions, which became effective on September 4, 2012.  See Press Release, Supreme Court Releases Eyewitness Identification Criteria for Criminal Cases (July 19, 2012), available at http://www.judiciary.state.nj.us/pressrel/2012/pr120719a.htm [http://perma.cc/VQF3-SXH4] (last visited Jan. 8, 2015).  The New Jersey model instructions can be found at http://www.judiciary.state.nj.us/criminal/charges/idinout.pdf [http://perma.cc/4BE2-F79V] (last visited Jan. 8, 2015).  Although the defendant cited the Henderson opinion in his request for a jury instruction, the defendant did not inform the judge that the instruction he proffered was a model jury instruction in New Jersey; his attorney merely told the judge that the proffered instruction “seem[ed] to be an appropriate instruction in New Jersey.”

[11] The defendant’s proffered jury instruction provided:

“Human memory is not foolproof.  Research has revealed that human memory is not like a video recording that a witness need only replay to remember what happened.  Memory is far more complex. . . .  The process of remembering consists of three stages:  (1) acquisition — the perception of the original event; (2) retention — the period of time that passes between the event and the eventual recollection of a piece of information; and (3) retrieval — the stage during which a person recalls stored information.  At each of these stages, memory can be affected by a variety of factors.”  (Citation omitted.)

[12] The proffered jury instruction provided:

“Although nothing may appear more convincing than a witness’s categorical identification of a perpetrator, you must critically analyze such testimony.  Such identifications, even if made in good faith, may be mistaken.  Therefore, when analyzing such testimony, be advised that a witness’s level of confidence, standing alone, may not be an indication of the reliability of the identification.”

[13] The proffered jury instruction provided:

“Even under the best viewing conditions, high levels of stress can reduce an eyewitness’s ability to recall and make an accurate identification.”

[14] The proffered jury instruction provided:

“You may consider whether the witness was exposed to opinions, descriptions, or identifications given by other witnesses, to photographs or newspaper accounts, or to any other information or influence, that may have affected the independence of his/her identification.  Such information can affect the independent nature and reliability of a witness’s identification and inflate the witness’s confidence in the identification.”

[15] The proffered jury instruction provided:

“When a witness views the same person in more than one identification procedure, it can be difficult to know whether a later identification comes from the witness’s memory of the actual, original event or of an earlier identification procedure.  As a result, if a witness views an innocent suspect in multiple identification procedures, the risk of mistaken identification is increased.  You may consider whether the witness viewed the suspect multiple times during the identification process and, if so, whether that affected the reliability of the identification.”

[16] Before trial, the judge allowed the defendant’s motion for funds to obtain an expert on the reliability of eyewitness identification evidence.  The judge denied the Commonwealth’s motion to exclude expert testimony, but the defendant never called an expert at trial.

[17] The only citations to scientific studies in the record are located in the disclosure of the defendant’s proffered expert witness on eyewitness identification, regarding the subject matter of his proposed testimony.  The defendant made no reference to this document in requesting his proposed jury instruction on identification testimony.

[18] We thank the Study Group for its thorough review of the research regarding eyewitness identification and its thoughtful recommendations.  We also thank those who submitted comments regarding the Study Group Report.

[19] The instruction, as set forth in Commonwealth v. Franklin, 465 Mass. 895, 910 n.24 (2013), states:

“One of the most important issues in this case is the identification of the defendant as the perpetrator of the crime. The Government has the burden of proving identity beyond a reasonable doubt.  It is not essential that the witness himself be free from doubt as to the correctness of his statement.  However, you, the jury, must be satisfied beyond a reasonable doubt of the accuracy of the identification of the defendant before you may convict him.  If you are not convinced beyond a reasonable doubt that the defendant was the person who committed the crime, you must find the defendant not guilty.

“Identification testimony is an expression of belief or impression by the witness.  Its value depends on the opportunity the witness had to observe the offender at the time of the offense and to make a reliable identification later.

“In appraising the identification testimony of a witness, you should consider the following:

“Are you convinced that the witness had the capacity and an adequate opportunity to observe the offender?

“Whether the witness had an adequate opportunity to observe the offender at the time of the offense will be affected by such matters as how long or short a time was available, how far or close the witness was, how good were lighting conditions, whether the witness had had occasion to see or know the person in the past.

“In general, a witness bases any identification he makes on his perception through the use of his senses. Usually the witness identifies an offender by the sense of sight — but this is not necessarily so, and he may use other senses.

“Are you satisfied that the identification made by the witness subsequent to the offense was the product of his own recollection?  You may take into account the circumstances under which the identification was made.

“If the identification by the witness may have been influenced by the circumstances under which the defendant was presented to him for identification, you should scrutinize the identification with great care.

“You may also consider the length of time that lapsed between the occurrence of the crime and the opportunity of the witness, some time after the occurrence of the crime, to see and identify the defendant as the offender, as a factor bearing on the reliability of the identification.

“You may also take into account that an identification made by picking the defendant out of a group of similar individuals is generally more reliable than one which results from the presentation of the defendant alone to the witness.

“You may take into account any occasions in which the witness failed to make an identification of [the] defendant, or made an identification that was inconsistent with his identification at trial.

“Finally, you must consider the credibility of each identification witness in the same way as any other witness, consider whether he is truthful, and consider whether he had the capacity and opportunity to make a reliable observation on the matter covered in his testimony.

“I again emphasize that the burden of proof on the prosecutor extends to every element of the crime charged, and this specifically includes the burden of proving beyond a reasonable doubt the identity of the defendant as the perpetrator of the crime with which he stands charged.  If after examining the testimony, you have a reasonable doubt as to the accuracy of the identification, you must find the defendant not guilty.”

In addition, “[f]airness to a defendant compels the trial judge to give an instruction on the possibility of an honest but mistaken identification when the facts permit it and when the defendant requests it.”  Id., quoting Commonwealth v. Pressley, 390 Mass. 617, 620 (1983).

[20] Nor can we look to the standard for judicial notice to define the meaning of “so generally accepted” in this context.  A court may take judicial notice of adjudicative facts that are “not subject to reasonable dispute in that it is either (1) generally known within the territorial jurisdiction of the trial court or (2) capable of accurate and ready determination by resort to resources whose accuracy cannot reasonably be questioned.”  Mass. G. Evid. § 201(b) (2014).  Matters of common knowledge may be judicially noticed, see Commonwealth v. Hartman 404 Mass. 306, 313 n.9 (1989), but “[f]acts which ordinarily are not known without the aid of expert testimony or other proof cannot be said to be matters of common knowledge.”  Id., quoting Mady v. Holy Trinity Roman Catholic Polish Church, 223 Mass. 23, 26 (1916).  The principles at issue in eyewitness identification are not matters of common knowledge.  Nor can these principles be readily looked up in an authoritative source; rather, they require review of the considerable scientific literature and published research studies regarding eyewitness identification.  Therefore, these principles, no matter how well accepted they may be in the relevant scientific community, are not the type of adjudicative facts of which a court generally may take judicial notice.  Moreover, “[i]n a criminal case, the court shall instruct the jury that they may, but are not required to, accept as conclusive any fact which the court has judicially noticed.”  Mass. G. Evid. § 201(e) (2014).  See Commonwealth v. Kingsbury, 378 Mass. 751, 755 (1979).

[21] According to the Innocence Project, “Eyewitness misidentification is the single greatest cause of wrongful convictions nationwide, playing a role in 72% of convictions overturned through [deoxyribonucleic acid] testing”).  Innocence Project, Eyewitness Misidentification, http://www.innocenceproject.org/understand/Eyewitness-Misidentification.php [http://perma.cc/XAQ2-4QJG] (last visited Jan. 8, 2015).  The National Registry of Exonerations has recorded 522 known exonerations of persons whose cases involved at least one witness who mistakenly identified the exoneree as the perpetrator of the crime.  See National Registry of Exonerations, Exoneration Detail List, http://www.law.umich.edu/special/exoneration/Pages/detaillist.aspx [http://perma.cc./DPD3-BJBB] (last visited Jan. 8, 2015).  See also Connors, Lundregan, Miller, & McEwen, U.S. Department of Justice, Convicted by Juries, Exonerated by Science:  Case Studies in the Use of DNA Evidence to Establish Innocence After Trial 15-17, 24 (1996), at https://www.ncjrs.gov/pdffiles/dnaevid.pdf [http://perma.cc/RUA3-8NKW] (last visited Jan. 8, 2015) (“[E]yewitness testimony was the most compelling evidence” in majority of twenty-four sexual assault cases reviewed where defendants were convicted and later exonerated); B.L. Garrett, Convicting the Innocent:  Where Criminal Prosecutions Go Wrong 48 (2011) (analyzing 250 wrongful convictions and finding 190 involved eyewitness misidentification).  There have been forty exonerations in Massachusetts since 1990, and twenty of those cases involved mistaken eyewitness identification.  See National Registry of Exonerations, supra.  See also Irwin v. Commonwealth, 465 Mass. 834, 849 n.25 (2013), citing Fisher, Convictions of Innocent Persons in Massachusetts:  An Overview, 12 B.U. Pub. Int. L.J. 1, 64 & n.278 (2002) (“A summary of several studies of erroneous convictions in Massachusetts concluded that, in over half of the cases where convicted defendants were later officially exonerated, the convictions involved mistaken identifications by eyewitnesses, including by multiple eyewitnesses who had had ample opportunity to observe the perpetrator”).

[22] See Benton, Ross, Bradshaw, Thomas, & Bradshaw, Eyewitness Memory Is Still Not Common Sense:  Comparing Jurors, Judges and Law Enforcement to Eyewitness Experts, 20 Applied Cognitive Psychol. 115, 119 (2006) (survey found that jurors and experts differed on eighty-seven per cent of survey’s statements about eyewitness identification); Schmechel, O’Toole, Easterly, & Loftus, Beyond the Ken?  Testing Jurors’ Understanding of Eyewitness Reliability Evidence, 46 Jurimetrics 177, 204 (2006) (“a substantial number of jurors come to each trial with basic misunderstandings about the way memory works in general and about specific factors that can affect the reliability of eyewitness identifications”).

[23] In a 2001 survey of experts in the field of psychology, researchers found that at least eighty-seven per cent of experts believed the following principles were reliable enough to be presented in court:  “[a]n eyewitness’s confidence can be influenced by factors that are unrelated to identification accuracy” (ninety-five per cent), “[e]xposure to mug shots of a suspect increases the likelihood that the witness will later choose that suspect in a lineup” (ninety-five per cent), “[e]yewitness testimony about an event often reflects not only what they actually saw but information they obtained later on” (ninety-four per cent), and “an eyewitness’s confidence is not a good predictor of his or her identification accuracy” (eighty-seven per cent).  Kassin, Tubb, Hosch, & Memon, On the “General Acceptance” of Eyewitness Testimony Research:  A New Survey of the Experts, 56 Am. Psychol. 405, 407-412 (2001).  See Malpass, Ross, Meissner, & Marcon, The Need for Expert Psychological Testimony on Eyewitness Identification, in Expert Testimony on the Psychology of Eyewitness Identification 15 (2009) (“[I]t would be very difficult to sustain the position that many of the findings in research on eyewitness memory lack general agreement within the scientific community”).

[24] We note that the instruction we adopted in Commonwealth v. Rodriguez, 378 Mass. 296, 310-311 (Appendix) (1979), already essentially includes two principles on which there is at least near consensus in the relevant scientific community, that is, “that an identification made by picking the defendant out of a group of similar individuals is generally more reliable than one which results from the presentation of the defendant alone to the witness,” and that where the “identification by the witness may have been influenced by the circumstances under which the defendant was presented to him for identification, [the jury] should scrutinize the identification with great care.”  Therefore, it is more accurate to say that we are adding scientific principles to our eyewitness identification instruction rather than incorporating such principles into our instruction for the first time.

[25] We will look to our newly reconstituted Supreme Judicial Court Committee on Eyewitness Identification to assist us in recognizing the need for such revision.

[ 26] This list of generally accepted principles is not intended to be exhaustive, as we only address the principles most relevant to the case before us.  Therefore, the exclusion of a principle should not be construed to suggest that it is not so generally accepted as to be worthy of inclusion in a model jury instruction on eyewitness identification.  In fact, the provisional jury instruction we include in the Appendix to this decision incorporates principles beyond the five addressed here.

[27] See Commonwealth v. Crayton, ante 228, 239 n.15 (2014), quoting Wells, Memon, & Penrod, Eyewitness Evidence:  Improving Its Probative Value, 7 Psychol. Sci. in the Pub. Interest 45, 66 (2006) (“Even among ‘highly confident witnesses, [studies] indicate that 20 to 30% could be in error’”); Crayton, supra, quoting Wells & Quinlivan, Suggestive Eyewitness Identification Procedures and the Supreme Court’s Reliability Test in Light of Eyewitness Science:  30 Years Later, 33 Law & Hum. Behav. 1, 11-12 (2009) (“the less-than-perfect correlation between height and gender in humans is ‘considerably greater’ than the correlation between certainty and accuracy in eyewitness identifications”).

[28]See Cutler & Penrod, Forensically Relevant Moderators of the Relation Between Eyewitness Identification Accuracy and Confidence, 74 J. Applied Psychol. 650, 652 (1989) (meta-analysis showing that pre-lineup confidence “certainly should not be used in the evaluation of eyewitness identification accuracy”).

[29] Compare Study Group Report, supra at 19, quoting Report of the Special Master, State vs. Henderson, N.J. Supreme Ct., No. A-8-08, at 34 (Special Master’s Report) (“confidence expressed immediately after making an identification has only a low correlation to the accuracy of the identification”), with Wells & Olson, Eyewitness Testimony, 54 Ann. Rev. Psychol. 277, 283 (2003) (more recent studies “indicate that the

certainty-accuracy relation is stronger” if analysis is restricted to witnesses who actually made identifications, thereby excluding witnesses who did not identify anyone).  See also Sporer, Read, Penrod, & Cutler, Choosing, Confidence, and Accuracy:  A Meta–Analysis of the Confidence–Accuracy Relation in Eyewitness Identification Studies, 118 Psychol. Bull. 315, 322 (1995).

[30] See Wells & Bradfield, Distortions in Eyewitnesses’ Recollections:  Can the Postidentification–Feedback Effect Be Moderated?, 10 Psychol. Sci. 138, 138 (1999) (“The idea that confirming feedback would lead to confidence inflation is not surprising.  What is surprising, however, is that confirming feedback that is given after the identification leads eyewitnesses to misremember how confident they were at the time of the identification”).  See also Commonwealth v. Collins, ante 255, 263 n.10 (2014), quoting National Research Council of the National Academies, Identifying the Culprit:  Assessing Eyewitness Identification 75 (2014) (pending publication) (“[I]n-court confidence statements may . . . be less reliable than confidence judgments made at the time of an initial out-of-court identification . . . . The confidence of an eyewitness may increase by the time of the trial as a result of learning more information about the case, participating in trial preparation, and experiencing the pressures of being placed on the stand”).

[31] “Because ‘a witness’s confidence in the accuracy of his identification grows once he learns that the police believe he made the correct identification,’ we have previously announced that we ‘expect’ police to use protocols for photographic arrays that include a ‘procedure requir[ing] the administrator to ask the witness to state, in his or her own words, how certain he or she is of any identification.’” Collins, supra at 263 n.11, quoting Commonwealth v. Silva-Santiago, 453 Mass. 782, 791, 798 (2009).

[32] See Cutler, Penrod, & Dexter, Juror Sensitivity to Eyewitness Identification Evidence, 14 Law & Hum. Behav. 185, 190 (1990) (mock-jury experiment showed jurors “gave disproportionate weight to the confidence of the witness”); Wells, Lindsay, & Ferguson, Accuracy, Confidence, and Juror Perceptions in Eyewitness Identification, 64 J. Applied Psychol. 440, 446 (1979) (“The data indicate that although jurors’ decisions to believe the witness are highly related to their ratings of the witnesses’ confidence, the confidence-accuracy relationship is very poor”).

[33] In another experiment, 509 active duty military personnel in military survival school training were subjected to high- or low-stress interrogations.  See Morgan, Hazlett, Doran, Garrett, Hoyt, Thomas, Baranoski, & Southwick, Accuracy of Eyewitness Memory for Persons Encountered During Exposure to Highly Intense Stress, 27 Int’l J.L. & Psychiatry 265, 267-268 (2004).  When subjects were asked to identify the interrogator in a lineup or photographic array, “the accuracy of eyewitness recognition . . . for the interrogator appeared to be greater for the low-, compared to the high-stress condition.”  Id. at 272.  “These data provide robust evidence that eyewitness memory for persons encountered during events that are personally relevant, highly stressful, and realistic in nature may be subject to substantial error.”  Id. at 274.  See Morgan, Southwick, Steffian, Hazlett, & Loftus, Misinformation Can Influence Memory for Recently Experienced, Highly Stressful Events, 36 Int’l J.L. & Psychiatry 11, 16 (2013) (similar study of military personnel at survival school found that “human memory for realistic, recently experienced stressful events is subject to substantial error.  In addition,  . . . memories for stressful events are also highly vulnerable to modification by exposure to misinformation”).

[34] See also Loftus & Zanni, Eyewitness Testimony:  The Influence of the Wording of a Question, 5 Bull. Psychonomic Soc’y 86, 88 (1975) (changing wording of question from “[d]id you see a broken headlight” to “[d]id you see the broken headlight” led to more false recognitions [emphasis added]); Loftus, Leading Questions and the Eyewitness Report, 7 Cognitive Psychol. 560, 566 (1975) (after watching videotape of vehicle driving on road where there was no barn, 17.3 per cent of participants who were asked to estimate vehicle’s speed “when it passed the barn” claimed to see barn, compared to 2.7 per cent of participants whose question did not mention barn).

[ 35] In one experiment, witnesses who made false identifications at a target-absent lineup were given either confirming feedback (“Good.  You identified the actual suspect”), disconfirming feedback (“Actually, the suspect was number _”), or no feedback.  Wells & Bradfield, “Good, You Identified the Suspect”:  Feedback to Eyewitnesses Distorts Their Reports of the Witnessing Experience, 83 J. Applied Psychol. 360, 363 (1998).  Not only did confirmatory feedback affect witness reports of how certain they were at the time of the identification, but it also distorted “their reports of the witnessing experience.”  Id. at 367.  Witnesses receiving confirming feedback reported “a better view of the culprit, a greater ability to make out details of the face, greater attention to the event, [and] a stronger basis for making an identification,” compared to witnesses receiving no feedback.  Id. at 366.  Additionally, a meta-analysis of ten published and four unpublished studies, totaling approximately 2,400 participants, showed that participants who received confirming feedback “expressed significantly more retrospective confidence in their decision compared with participants who received no feedback” and “significantly inflate[d] their reports to suggest better witnessing conditions at the time of the crime, stronger memory at the time of the lineup, and sharper memory abilities in general.”  Douglass & Steblay, Memory Distortion in Eyewitnesses:  A Meta–Analysis of the Post-identification Feedback Effect, 20 Applied Cognitive Psychol. 859, 863–865 (2006).  See Crayton, supra at 239 n.15, quoting Wells & Quinlivan, Suggestive Eyewitness Identification Procedures and the Supreme Court’s Reliability Test in Light of Eyewitness Science:  30 Years Later, 33 Law & Hum. Behav. 1, 12 (2009) (suggestive confirmatory effect “is stronger for mistaken eyewitnesses than it is for accurate eyewitnesses, thereby making inaccurate eyewitnesses look more like accurate eyewitnesses and undermining the certainty-accuracy relation”).

[36] When pairs of subjects viewed a crime and discussed who they believed was the culprit, researchers concluded that “post-identification feedback does not have to be presented by the experimenter or an authoritative figure (e.g. police officer) in order to affect a witness’[s] subsequent crime-related judgments.”  Skagerberg, Co–Witness Feedback in Line-Ups, 21 Applied Cognitive Psychol. 489, 494 (2007).  When the cowitnesses agreed with one another, they reported having better views of the culprit, higher certainty, and more willingness to testify compared to cowitnesses who disagreed on the culprit’s identity — even though none of the photographic arrays showed the actual suspect.  Id. at 493-495.

[37] “Unconscious transference” is a similar phenomenon that occurs “when a witness confuses a person seen at or near the crime scene with the actual perpetrator.”  Study Group Report, supra at 31, quoting Special Master’s Report, supra at 46.  In one experiment, witnesses were asked to identify the assailant from a target-absent lineup containing an innocent bystander they had seen previously near the crime scene; witnesses “were nearly three times more likely to misidentify the bystander than were control subjects.”  Ross, Ceci, Dunning, & Toglia, Unconscious Transference and Mistaken Identity:  When a Witness Misidentifies a Familiar but Innocent Person, 79 J. Applied Psychol. 918, 923 (1994). “Regardless of the content of the lineup (bystander present or assailant present), a majority of the transference subjects thought the assailant and the bystander were the same person who was seen in two different places.”  Id. at 924.  However, we recognize that there is less conclusive support for unconscious transference, and it is not clear still how or why it occurs.  Id. at 919, 929-930.

[38] A recent experimental study of the New Jersey model jury instructions revealed that they did not improve jurors’ ability to distinguish between “weak” and “strong” eyewitness testimony; rather, the enhanced instructions “caused jurors to indiscriminately discount testimony.”  Papailiou, Yokum, & Robertson, The Novel New Jersey Eyewitness Instruction Induces Skepticism But Not Sensitivity, Arizona Legal Studies Discussion Paper No. 14-17, at 22 (Aug. 2014).  “[U]se of the novel New Jersey instruction substantially reduced the likelihood that the defendant would be found guilty, but its reducing effect was the same regardless of whether the eyewitness identification testimony was weak or strong.” Id. at 12-13.  See also Vermont Model Criminal Jury Instructions, Reporter’s Note (Aug. 2012), available at http://vtjuryinstructions.org/?page_id=662 [http://perma.cc/8WFD-42AF] (last visited Jan. 8, 2015) (drafters of Vermont model instructions recognized that “the general approach to eyewitness identification may be evolving” but cautioned “against using a longer instruction on eyewitness identification”).  Our provisional jury instruction is approximately 1,000 words shorter than the comparable New Jersey model jury instruction.

footnotes for Appendix 

[1] The following jury instruction has not been adopted as an official model.  Rather, it is a provisional instruction that trial courts should use until we adopt a model instruction after soliciting comments from the public.

[2] See Supreme Judicial Court Study Group on Eyewitness Evidence:  Report and Recommendations to the Justices 15 (July 25, 2013), available at http://www.mass.gov/courts/docs/sjc/docs/eyewitness-evidence-report-2013.pdf [http://perma.cc/WY4M-YNZN] (last visited Jan. 8, 2015) (Study Group Report), quoting Report of the Special Master, State vs. Henderson, N.J. Supreme Ct., Docket No. A-8-08 (June 10, 2010), at 9 (Special Master’s Report) (“The central precept is that memory does not function like a videotape, accurately and thoroughly capturing and reproducing a person, scene or event. . . .  Memory is, rather[,] a constructive, dynamic and selective process”); State v. Henderson, 208 N.J. 208, 245 (2011); State v. Lawson, 352 Or. 724, 771 (2012) (Appendix).  See also E.F. Loftus, J.M. Doyle, & J.E. Dysart, Eyewitness Testimony:  Civil and Criminal § 2-2, at 14 (5th ed. 2013) (Loftus et al.); Brigham, Wasserman, & Meissner, Disputed Eyewitness Identification Evidence:  Important Legal and Scientific Issues, 36 Ct. Rev., no. 2, 1999, at 13.

[3] See Study Group Report, supra at 16, quoting Henderson, 208 N.J. at 245 (“Three stages are involved in forming a memory: stages:  ‘acquisition — “the perception of the original event”; retention — “the period of time that passes between the event and the eventual recollection of a particular piece of information”; and retrieval — the “stage during which a person recalls stored information”‘”).

[4] For a detailed discussion of the three stages of memory and how those stages may be affected, see Study Group Report, supra at 16; National Research Council of the National Academies, Identifying the Culprit:  Assessing Eyewitness Identification 40-46 (2014) (pending publication) (National Academies) (“Encoding, storage, and remembering are not passive, static processes that record, retain, and divulge their contents in an informational vacuum, unaffected by outside influences”).  See also State v. Guilbert, 306 Conn. 218, 235-236 (2012); Henderson, 208 N.J. at 247; Loftus et al., supra at § 2-2, at 15 (“Numerous factors at each stage affect the accuracy and completeness of an eyewitness account”).

[5] See Study Group Report, supra at 30, quoting Lawson, 352 Or. at 775 (Appendix) (“[S]tudies confirm that the use of a disguise negatively affects later identification accuracy.  In addition to accoutrements like masks and sunglasses, studies show that hats, hoods, and other items that conceal a perpetrator’s hair or hairline also impair a witness’s ability to make an accurate identification”); Henderson, 208 N.J. at 266 (“Disguises and changes in facial features can affect a witness’[s] ability to remember and identify a perpetrator”); State v. Clopten, 223 P.3d 1103, 1108 (Utah 2009) (“[A]ccuracy is significantly affected by factors such as the amount of time the culprit was in view, lighting conditions, use of a disguise, distinctiveness of the culprit’s appearance, and the presence of a weapon or other distractions”); Wells & Olson, Eyewitness Testimony, 54 Ann. Rev. Psychol. 277, 281 (2003) (Wells & Olson) (“Simple disguises, even those as minor as covering the hair, result in significant impairment of eyewitness identification”).  See also Cutler, A Sample of Witness, Crime, and Perpetrator Characteristics Affecting Eyewitness Identification Accuracy, 4 Cardozo Pub. L. Pol’y & Ethics J. 327, 332 (2006) (“In data from over 1300 eyewitnesses, the percentage of correct judgments on identification tests was lower among eyewitnesses who viewed perpetrators wearing hats [44%] than among eyewitnesses who viewed perpetrators whose hair and hairlines were visible [57%]“); Patterson & Baddeley, When Face Recognition Fails, 3 J. Experimental Psychol. 406, 410 (1977).

[6] See Study Group Report, supra at 30-31, quoting Lawson, 352 Or. at 774 (“Witnesses are better at remembering and identifying individuals with distinctive features than they are those possessing average features”); Clopten, 223 P.3d at 1108; Wells & Olson, supra at 281 (“Distinctive faces are much more likely to be accurately recognized than nondistinctive faces”); Shapiro & Penrod, Meta–Analysis of Facial Identification Studies, 100 Psychol. Bull. 139, 140, 145 (1986) (meta-analysis finding that distinctive targets were “easier to recognize than ordinary looking targets”).

[7] See Study Group Report, supra at 29, quoting Henderson, 208 N.J. at 262-263 (“‘Weapon focus’ can . . . impair a witness’s ability to make a reliable identification and describe what the culprit looks like if the crime is of short duration”); Guilbert, 306 Conn. at 253; Lawson, 352 Or. at 771-772 (Appendix).  See also Kassin, Hosch, & Memon, On the “General Acceptance” of Eyewitness Testimony Research:  A New Survey of the Experts, 56 Am. Psychol. 405, 407-412 (2001) (Kassin et al.) (in 2001 survey, eighty-seven per cent of experts agree that principle that “[t]he presence of a weapon impairs an eyewitness’s ability to accurately identify the perpetrator’s face” is reliable enough to be presented in court); Maass & Köhnken, Eyewitness Identification:  Simulating the “Weapon Effect,” 13 Law & Hum. Behav. 397, 405-406 (1989); Steblay, A Meta–Analytic Review of the Weapon Focus Effect, 16 Law & Hum. Behav. 413, 415–417 (1992) (meta-analysis finding “weapon-absent condition[s] generated significantly more accurate descriptions of the perpetrator than did the weapon-present condition”); id. at 421 (“To not consider a weapon’s effect on eyewitness performance is to ignore relevant information.  The weapon effect does reliably occur, particularly in crimes of short duration in which a threatening weapon is visible”); Wells & Quinlivan, Suggestive Eyewitness Identification Procedures and the Supreme Court’s Reliability Test in Light of Eyewitness Science:  30 Years Later, 33 Law & Hum. Behav. 1, 11 (2009).  But see National Academies, supra at 64 (recent meta-analysis “shows that the effect of a weapon on accuracy is slight in actual crimes, slightly larger in laboratory studies, and largest for simulations”).

[8] See Study Group Report, supra at 135 (recommending instruction stating, “If the witness had seen the defendant before the incident, you should consider how many times the witness had seen the defendant and under what circumstances”); Commonwealth v. Adams, 458 Mass. 766, 770-771 (2011) (“Traditional identification procedures such as photographic arrays, showups, and lineups were designed primarily for witnesses who had never before seen a particular individual, or who may have seen the individual previously but on a limited basis.  They are not normally used, and are not required, for witnesses who know an individual well”).  See also Commonwealth v. Pressley, 390 Mass. 617, 619 (1983) (“There may be cases in which the parties are so well known to each other or so closely related that under sufficient lighting and with appropriate physical proximity, the identification by the victim is either true or the victim is lying”); Commonwealth v. Stoddard, 38 Mass. App. Ct. 45, 48 (1995) (no error in omitting “honest but mistaken” language where “victim knew the defendant as a regular customer of the [gasoline] station and had encountered him numerous times over a year and one-half”).  But see Pezdek & Stolzenberg, Are Individuals’ Familiarity Judgments Diagnostic of Prior Contact?, 20 Psychol. Crime & L. 302, 306 (2014) (twenty-three per cent of study participants misidentified subjects with unfamiliar faces as familiar, and only forty-two per cent correctly identified familiar face as familiar).

[9] See Study Group Report, supra at 29, quoting Special Master’s Report, supra at 43 (“The scientific literature reports that, while moderate levels of stress improve cognitive processing and might improve accuracy . . . , an eyewitness under high stress is less likely to make a reliable identification of the perpetrator”); Lawson, 352 Or. at 769 (Appendix).  See also Deffenbacher, Bornstein, Penrod, & McGorty, A Meta-Analytic Review of the Effects of High Stress on Eyewitness Memory, 28 Law & Hum. Behav. 687, 699 (2004) (finding “considerable support for the hypothesis that high levels of stress negatively impact both accuracy of eyewitness identification as well as accuracy of recall of crime-related details”); Morgan, Hazlett, Doran, Garrett, Hoyt, Thomas, Baranoski, & Southwick, Accuracy of Eyewitness Memory for Persons Encountered During Exposure to Highly Intense Stress, 27 Int’l J. L. & Psychiatry 265, 272-274 (2004).  But see Study Group Report, supra, quoting Henderson, 208 N.J. at 262 (“There is no precise measure for what constitutes ‘high’ stress, which must be assessed based on the facts presented in individual cases”).

[10] See Study Group Report, supra at 31 (“A witness may have more difficulty identifying a person of a different race or ethnicity”); Kassin et al., supra at 407-412 (in 2001 survey, ninety per cent of experts agree that principle that “[e]yewitnesses are more accurate when identifying members of their own race than members of other races” is reliable enough to be presented in court); Meissner & Brigham, Thirty Years of Investigating the Own-Race Bias in Memory for Faces:  A Meta-Analytic Review, 7 Psychol., Pub. Pol’y, & L. 3, 15 (2001) (meta-analysis of thirty-nine research articles concluding that participants were “1.4 times more likely to correctly identify a previously viewed own-race face when compared with performance on other-race faces” and “1.56 times more likely to falsely identify a novel other-race face when compared with performance on own-race faces”); Wells & Olson, supra at 280-281.  See also Commonwealth v. Zimmerman, 441 Mass. 146, 154-155 (2004) (Cordy, J., concurring); State v. Cabagbag, 127 Haw. 302, 310-311 (2012); Lawson, 352 Or. at 775 (Appendix); National Academies, supra at 66, citing Grimsley, Innocence Project, What Wrongful Convictions Teach Us About Racial Inequality, Innocence Blog (Sept. 26, 2012, 2:30 P.M.), at http://www.innocenceproject.org/Content/What_Wrongful_Convictions_Teach_Us_About_Racial_Inequality.php [http://perma.cc/KX2J-XECN] (last visited Jan. 9, 2015) (“Recent analyses revealed that cross-racial [mis]identification was present in 42 percent of the cases in which an erroneous eyewitness identification was made”).

[11] See Study Group Report, supra at 31-32, quoting Lawson, 352 Or. at 778 (Appendix) (“The more time that elapses between an initial observation and a later identification procedure [a period referred to in eyewitness identification research as a ‘retention interval’] . . . the less reliable the later recollection will be. . . . [D]ecay rates are exponential rather than linear, with the greatest proportion of memory loss occurring shortly after an initial observation, then leveling off over time”); National Academies, supra at 11 (“For eyewitness identification to take place, perceived information must be encoded in memory, stored, and subsequently retrieved.  As time passes, memories become less stable”).

[12] See Study Group Report, supra at 19 (“Social science research demonstrates that little correlation exists between witness confidence and the accuracy of the identification”); Lawson, 352 Or. at 777 (Appendix) (“Despite widespread reliance by judges and juries on the certainty of an eyewitness’s identification, studies show that, under most circumstances, witness confidence or certainty is not a good indicator of identification accuracy”); Clopten, 223 P.3d at 1108.  See also Commonwealth v. Cruz, 445 Mass. 589, 597-600 (2005); Commonwealth v. Santoli, 424 Mass. 837, 845-846 (1997); Commonwealth v. Jones, 423 Mass. 99, 110 n.9 (1996).

[13] See Commonwealth v. Crayton, 470 Mass. 228, 239 (2014) (“Social science research has shown that a witness’s level of confidence in an identification is not a reliable predictor of the accuracy of the identification, especially where the level of confidence is inflated by its suggestiveness”); Henderson, 208 N.J. at 254 (“Confirmatory feedback can distort memory.  As a result, to the extent confidence may be relevant in certain circumstances, it must be recorded in the witness’[s] own words before any possible feedback”); Lawson, 352 Or. at 745 (“Retrospective self-reports of certainty are highly susceptible to suggestive procedures and confirming feedback, a factor that further limits the utility of the certainty variable”); Wells & Bradfield, Distortions in Eyewitnesses’ Recollections:  Can the Postidentification–Feedback Effect Be Moderated?, 10 Psychol. Sci. 138, 138 (1999) (“The idea that confirming feedback would lead to confidence inflation is not surprising.  What is surprising, however, is that confirming feedback that is given after the identification leads eyewitnesses to misremember how confident they were at the time of the identification”).

[14] See Study Group Report, supra at 21-22; Special Master’s Report, supra at 30-31 (“An extensive body of studies demonstrates that the memories of witnesses for events and faces, and witnesses’ confidence in their memories, are highly malleable and can readily be altered by information received by witnesses both before and after an identification procedure”); Lawson, 352 Or. at 786 (Appendix) (“The way in which eyewitnesses are questioned or converse about an event can alter their memory of the event”).

[15] See Study Group Report, supra at 22, quoting Henderson, 208 N.J. at 255 (postidentification feedback “affects the reliability of an identification in that it can distort memory, create a false sense of confidence, and alter a witness’[s] report of how he or she viewed an event”); Special Master’s report, supra at 33 (“A number of studies have demonstrated that witnesses’ confidence in their identifications, and their memories of events and faces, are readily tainted by information that they receive after the identification procedure”).  See also Commonwealth v. Collins, 470 Mass. 255, 263 (2014) (“Where confirmatory feedback artificially inflates an eyewitness’s level of confidence in his or her identification, there is also a substantial risk that the eyewitness’s memory of the crime at trial will ‘improve’”).

[16] See Study Group Report, supra at 117, 136 n.4, citing Principles of Neural Science, Box 62-1, at 1239 (Kandel, Schwartz, & Jessell eds., 2000); Clark, Marshall, & Rosenthal, Lineup Administrator Influences on Eyewitness Identification Decisions, 15 J. Experimental Psychol.:  Applied 63, 72 (2009) (“Most witnesses appeared to be unaware of the influence” of lineup administrator in staged experiment).

[17] See Henderson, 208 N.J. at 253 (“Confirmatory or post-identification feedback presents the same risks.  It occurs when police signal to eyewitnesses that they correctly identified the suspect”); Lawson, 352 Or. at 777-778 (Appendix); Hope, Ost, Gabbert, Healey, & Lenton, “With a Little Help from My Friends . . .”:  The Role of Co–Witness Relationship in Susceptibility to Misinformation, 127 Acta Psychologica 476, 481 (2008); Skagerberg, Co–Witness Feedback in Line-ups, 21 Applied Cognitive Psychol. 489, 494 (2007) (“post-identification feedback does not have to be presented by the experimenter or an authoritative figure [e.g. police officer] in order to affect a witness’ subsequent crime-related judgments”).

[18] See Study Group Report, supra at 21-22; Henderson, 208 N.J. at 255; Lawson, 352 Or. at 744.  See also Douglass & Steblay, Memory Distortion in Eyewitnesses:  A Meta–Analysis of the Post-Identification Feedback Effect, 20 Applied Cognitive Psychol. 859, 863–65 (2006) (participants who received confirming feedback “expressed significantly more retrospective confidence in their decision compared with participants who received no feedback”); Wells & Bradfield, “Good, You Identified the Suspect”:  Feedback to Eyewitnesses Distorts Their Reports of the Witnessing Experience, 83 J. Applied Psychol. 360, 366-367 (1998) (witnesses receiving confirming feedback reported “a better view of the culprit, a greater ability to make out details of the face, greater attention to the event, [and] a stronger basis for making an identification” compared to witnesses receiving no feedback); Wells & Bradfield, Distortions in Eyewitnesses’ Recollections:  Can the Postidentification–Feedback Effect Be Moderated?, 10 Psychol. Sci. 138, 138 (1999); National Academies, supra at 64 (“Research has . . . shown that . . . if an eyewitness hears information or misinformation from another person before law enforcement involvement, his or her recollection of the event and confidence in the identification can be altered . . .”).

[19] See Steblay, Wells, & Douglass, The Eyewitness Post Identification Feedback Effect 15 Years Later:  Theoretical and Policy Implications, 20 Psychol. Pub. Pol. & L. 1, 10 (2014) (significant but smaller postidentification feedback effect on accurate eyewitnesses compared to inaccurate eyewitnesses).  See also Allan, Midjord, Martin, & Gabbert, Memory Conformity and the Perceived Accuracy of Self Versus Other, 40 Memory & Cognition 280, 285 (2011) (study participants with least amount of time to view initial event, and who were told that their partner had twice as long to view same event, were most likely to conform their memory to their partner’s recollection); Deffenbacher, Bornstein, & Penrod, Mugshot Exposure Effects:  Retroactive Interference, Mugshot Commitment, Source Confusion, and Unconscious Transference, 30 Law & Hum. Behav. 287, 288 (2006) (bias from mugshot exposure “is all the more problematic when viewing of the perpetrator has occurred under less than optimal viewing conditions”).  Cf. Wells & Olson, supra at 283 (when accuracy is low due to poor witnessing conditions, certainty-accuracy relationship is less correlated).

[20] See Commonwealth v. Silva-Santiago, 453 Mass. 782, 797-798 (2009) (“What is practicable in nearly all circumstances is a protocol to be employed before a photographic array is provided to an eyewitness, making clear to the eyewitness, at a minimum that he will be asked to view a set of photographs; the alleged wrongdoer may or may not be in the photographs depicted in the array; it is just as important to clear a person from suspicion as to identify a person as the wrongdoer; individuals depicted in the photographs may not appear exactly as they did on the date of the incident because features such as weight and head and facial hair are subject to change; regardless of whether an identification is made, the investigation will continue; and the procedure requires the administrator to ask the witness to state, in his or her own words, how certain he or she is of any identification”).

[21] See Commonwealth v. Walker, 460 Mass. 590, 604 (2011) (“Unless there are exigent or extraordinary circumstances,” photographic array should not contain fewer than five fillers for every suspect).  See also Henderson, 208 N.J. at 251 (live lineups should also employ minimum of five fillers).

[22] See Wells & Olson, supra at 287 (“Ideally, lineup fillers would be chosen so that an innocent suspect is not mistakenly identified merely from ‘standing out,’ and so that a culprit does not escape identification merely from blending in”); Silva-Santiago, 453 Mass. at 795, quoting Commonwealth v. Melvin, 399 Mass. 201, 207 n.10 (1987) (“we ‘disapprove of an array of photographs which distinguishes one suspect from all the others on the basis of some physical characteristic’”).  See also Henderson, 208 N.J. at 251; Lawson, 352 Or. at 781; Malpass, Tredoux, & McQuiston-Surrett, Lineup Construction and Lineup Fairness, in 2 Handbook of Eyewitness Psychology 156 (2007) (“Decades of empirical research suggest that mistaken eyewitness identifications are more likely to occur when the suspect stands out in a lineup”).

[23] See Silva-Santiago, 453 Mass. at 797 (“we acknowledge that [a double-blind procedure] is the better practice because it eliminates the risk of conscious or unconscious suggestion”); Guilbert, 306 Conn. at 237-238 (courts across country accept that “identifications are likely to be less reliable in the absence of a double-blind, sequential identification procedure”); Henderson, 208 N.J. at 249 (“The consequences are clear:  a non-blind lineup procedure can affect the reliability of a lineup because even the best-intentioned, non-blind administrator can act in a way that inadvertently sways an eyewitness trying to identify a suspect”).  See also National Academies, supra at 18 (“As an alternative to a ‘double-blind’ array, some departments use ‘blinded’ procedures.  A blinded procedure prevents an officer from knowing when the witness is viewing a photo of the suspect, but can be conducted by the investigating officer”); id. at 73 (“The committee recommends blind [double-blind or blinded] administration of both photo arrays and live lineups and the adoption of clear, written policies and training on photo array and live lineup administration.  Police should use blind procedures to avoid the unintentional or intentional exchange of information that might bias an eyewitness”).

[24] See Clark, Marshall, & Rosenthal, Lineup Administrator Influences on Eyewitness Identification Decisions, 15 J. Experimental Psychol. Applied 63, 74 (2009) (subtle, nondirective statements by lineup administrator “can lead a witness to make an identification, particularly when the perpetrator was not present”); Malpass & Devine, Eyewitness Identification:  Lineup Instructions and the Absence of the Offender, 66 J. Applied Psychol. 482, 486-487 (1981) (where subject witnesses were asked to identify assailant in staged experiment, “[c]hanging the instruction from biased [suspect is present in lineup] to unbiased [suspect may or may not be present] resulted in fewer choices and fewer false identifications without a decrease in correct identifications”).

[25] See Study Group Report, supra at 26, citing Special Master’s Report, supra at 29 (showups carry their own risks of misidentification “due to the fact that only one person is presented to the witness”); Lawson, 352 Or. at 742-743 (“A ‘showup’ is a procedure in which police officers present an eyewitness with a single suspect for identification, often [but not necessarily] conducted in the field shortly after a crime has taken place.  Police showups are generally regarded as inherently suggestive — and therefore less reliable than properly administered lineup identifications — because the witness is always aware of whom police officers have targeted as a suspect”); Dysart & Lindsay, Show-up Identifications:  Suggestive Technique or Reliable Method?, in 2 Handbook of Eyewitness Psychology 141 (2007) (“Overall, show-ups [fare] poorly when compared with line-ups.  Correct identification rates are equal and false identification rates are about two to three times as high with show-ups compared with line-ups”).  See also Silva-Santiago, 453 Mass. at 797; Commonwealth v. Martin, 447 Mass. 274, 279 (2006) (“One-on-one identifications are generally disfavored because they are viewed as inherently suggestive”).

[26] See Crayton, 470 Mass. at 235-236 (“there is generally ‘good reason’ [to conduct showup] where the showup identification occurs within a few hours of the crime, because it is important to learn whether the police have captured the perpetrator or whether the perpetrator is still at large, and because a prompt identification is more likely to be accurate when the witness’s recollection of the event is still fresh”); Study Group Report, supra at 141 n.30, quoting Special Master’s Report, supra at 29 (“The research shows, in fact, that the risk of misidentification is not heightened if a showup is conducted immediately after the witnessed event, ideally within two hours:  the benefits of a fresh memory seem to balance the risks of undue suggestion”).  See also Dysart & Lindsay, The Effects of Delay on Eyewitness Identification Accuracy:  Should We Be Concerned?, in 2 Handbook of Eyewitness Psychology 370 (2007) (showups become particularly unreliable after twenty-four hours, rather than two hours).

[27] See Study Group Report, supra at 25, quoting Special Master’s Report, supra at 27-28 (“The problem is that successive views of the same person create uncertainty as to whether an ultimate identification is based on memory of the original observation or memory from an earlier identification procedure”); Henderson, 208 N.J. at 255; Deffenbacher, Bornstein, & Penrod, Mugshot Exposure Effects:  Retroactive Interference, Mugshot Commitment, Source Confusion, and Unconscious Transference, 30 Law & Hum. Behav. 287, 306 (2006) (“prior mugshot exposure decreases accuracy at a subsequent lineup, both in terms of reductions in rates for hits and correct rejections as well as in terms of increases in the rate for false alarms”).  See also Simmons v. United States, 390 U.S. 377, 383-384 (1968); Commonwealth v. Scott, 408 Mass. 811, 826 (1990).