BATTAGLIA, J.
This case involves an attempted robbery in which a cross-racial identification 1 was made by the victim of two defen[470]*470dants, Jason Mack and James Smith, as the perpetrators. The trial court denied defense counsel’s request to argue the difficulties of cross-racial identification in closing argument and to have the jury instructed concerning the cross-racial nature of the identifications. Based upon the circumstances of this case, we hold that the trial court erred in prohibiting defense counsel from commenting on cross-racial identification in their closing arguments and reverse the defendants’ convictions. We do not reach the jury instruction question.
Facts
On May 8, 2002, at approximately 10:30 p.m., Christine Crandall, a white female, parked her car in front of her residence in the Fells Point neighborhood of Baltimore City, when she noticed two black males walking toward her on the other side of the street. Ms. Crandall said “hey guys” to the two men, after which one of the men pointed a gun at her and stated “give me your keys, bitch. I’ll shoot you.” Shortly thereafter, the second man attempted to grab the keys from Ms. Crandall’s hand, but she maintained a tight fist around them, repeatedly stating “you don’t want to do this,” to the men.
During the altercation, one of Ms. Crandall’s neighbors, Mary Jo Slowey, looked out a nearby second-floor window and asked Ms. Crandall if she was okay. In response, Ms. Crandall yelled, “call 911, they have a gun.” At that point, the two men started to walk away, but the gunman allegedly turned and again pointed his weapon at Ms. Crandall, before leaving the scene.
Police officers responded to Ms. Slowey’s call but were unable to locate the two men. Ms. Crandall provided a general description of them and stated that the man with the gun had “dreds.” Two days after the incident, Detective Randolph Wynn, an officer with the Baltimore City Police Department assigned to the case, met with Ms. Crandall at [471]*471the police station and showed her an array of photographs based upon her description of the perpetrators. After viewing the photo arrays, Ms. Crandall wrote on the back of the last photo, “Out of these 6 photos, I do not recognize the ones who attempted to car jack me.”
Thereafter, Detective Wynn continued to investigate and prepared two additional photo arrays, which included pictures of the petitioners, Mr. Smith and Mr. Mack. On May 23, 2002, about two weeks after the incident, Ms. Crandall was shown the photos, and she identified Mr. Mack as the man who held the gun, although she noted that his hair was different because the man in the picture appeared to have his hair in “cornrows” rather than “dreds.” Ms. Crandall then wrote on the back of Mr. Mack’s photo, “He looks very much like the man who had the gun and attempted to rob me. The hair is changed but still looks like the man.” Ms. Crandall also chose Mr. Smith’s photo and wrote on the back, “This looks like the man wearing the hat that attempted to rob me. He tried to take the keys from my hand while the other man held the gun to me.” Based upon Ms. Crandall’s identification of Mr. Smith and Mr. Mack, both men were arrested on June 5 and 6, 2002, respectively.
Procedural History
Messrs. Mack and Smith were charged with Attempted Armed Robbery,2 First and Second Degree Assault,3 Carrying [472]*472a Handgun and Use of a Handgun in Commission of a Crime,4 [473]*473and Attempted Theft.5 Prior to the jury trial that began on February 24, 2003, in the Circuit Court for Baltimore City, Mr. Smith’s counsel submitted a motion in limine, in which Mr. Mack’s counsel orally joined, requesting that the jury be instructed on cross-racial eyewitness identification to enable the parties to raise the issue in opening statements. After hearing arguments by the parties, the judge denied the motion and stated that the issue of cross-racial identification could not be argued during opening statements, but offered that defense counsel could state that the defendants were black and that the case was based upon a single eyewitness identification, even though defense counsel did not make such a request.
After the trial court denied the requested jury instruction and prohibited the mention of cross-racial identification in opening statements, the State produced three witnesses to testify in support of its case-in-chief. Detective Wynn testified that he had been assigned to investigate an attempted [474]*474robbery of Ms. Crandall and that she had provided a general description of the suspects. He explained that he had prepared several photo arrays for review by Ms. Crandall, who upon seeing the photos, had stated with certainty that Mr. Smith and Mr. Mack were the two people who had pointed the gun at her and tried to take her keys, respectively.
Ms. Slowey also testified and stated that while she was watching television, she heard a noise in the street, opened the window, and saw Ms. Crandall standing on the sidewalk with two people who she could not describe by race or gender. Ms. Slowey declared that by the time she reached the sidewalk, the individuals were at the end of the block with their backs turned.
Finally, Ms. Crandall testified to the facts recounted here and, in addition, stated that during the incident she was able to observe both men. She described both of them as appearing “normal” and that they “looked great. They looked fine.” Ms. Crandall testified that there were street lights enabling her to see Mr. Smith’s face, which she described as “pretty distinctive looking.” Ms. Crandall further stated that Mr. Smith had been wearing a gray baseball hat, a “grayish long baggie sweat shirt and long over-sized pants” on the night in question. When asked if she could describe Mr. Smith’s facial features, Ms. Crandall declared that he had an “oval face” and “darker skin.” She also testified that Mr. Smith’s behavior was not as aggressive as Mr. Mack and that there was “a quieter look to him.”
Ms. Crandall also provided a description of Mr. Mack and testified that he was wearing a “dark bluish” sweatshirt, jeans and no hat and that she also was able to observe Mr. Mack’s face while he pointed the gun at her. In describing Mr. Mack’s facial features, Ms. Crandall stated that she “thought he was really handsome, and he had dreds. They were down, slightly trim and shorter, and full face and nice teeth.” Ms. Crandall further testified that Mr. Mack had a “strong posture.”
[475]*475Finally, counsel for the State asked Ms. Crandall why she was so sure that the defendants were the two individuals who had attempted to rob her, to which she replied that she was “extremely good with faces.” She further opined:
Well, I am a teacher and I watch lay mannerisms. I’ve been studying art and painting people since I was a little girl. I’m obsessed with people’s postures and the way you’[re] looking [at] them and seeing what’s going on.
And so, 1 think [of] myself [as] very, very good with people. [I] study faces and I have, I look for features on people that make them more distinct, Adams apple—
On cross-examination, defense counsel questioned Ms. Crandall about the sequence of events and her ability to identify the defendants, but did not ask specific questions relating to cross-racial identification.
During a recess of the proceedings, defense counsel requested permission to comment on the difficulty of the cross-racial identifications during closing arguments, and the following dialogue ensued:
[MR. MACK’S COUNSEL]: Your Honor, I understand you will not be reading an instruction to the jury on cross-racial identification. Nonetheless, will the defense counsel be permitted to argue to the jury on cross-racial identification.
[THE COURT]: The defense counsel will not be able to argue on cross-racial identification, ... there is not evidence in this case to that effect. But defense certainly could say my client is black, victim is white.
[MR. MACK’S COUNSEL]: But are we free to argue one reasonable inference, identification would not be as strong as if the complainant and defendant were the same race?
[THE COURT]: You can argue the facts that are in evidence. That is not a fact that is in evidence.
[MR. MACK’S COUNSEL]: But it’s an inference that can be drawn.
[476]*476[THE COURT]: You can argue he is black, and the victim is white. Anything else?
[STATE’S COUNSEL]: No, Your Honor.
[MR. SMITH’S COUNSEL]: No, Your Honor.
[MR. MACK’S COUNSEL]: No, Your Honor.
Thereafter, the defense called Officer Kevin Evans of the Baltimore City Police Department to testify as their only witness. Officer Kevin Evans briefly testified that when he responded to the call and arrived on the scene, Ms. Crandall had stated that she could draw a sketch of the man who had pointed the gun but could not sketch the unarmed man who also was present.
Prior to closing arguments, counsel for Mr. Mack renewed his request to argue cross-racial identification in his closing. The trial judge again denied the request:
[THE COURT]: Okay. So it is perfectly clear, I’m denying your request, but I would permit you to say that your client is black, victim is white, but I will not let you refer to cross-racial identification.
[MR. MACK’S COUNSEL]: And I take exception.
[THE COURT]: Certainly.
During closing arguments neither defense counsel raised the cross-racial identification issue, nor did either counsel mention the races of the defendants or the victim.
Subsequently, the jury found Mr. Smith guilty of attempted robbery, second degree assault, and attempted theft. Mr. Mack was found guilty of attempted robbery, first degree assault, and attempted theft. Both defendants were found not guilty of the handgun charges.
On April 15, 2003, the court held a sentencing hearing related to Mr. Mack, merged the assault and attempted theft convictions, and sentenced him to six years imprisonment for the attempted robbery conviction, with no suspension of the time to be served. The court held a sentencing hearing for Mr. Smith on April 21, 2003, also merged the assault and attempted theft convictions, and sentenced him to six years [477]*477imprisonment, with all but two years suspended, followed by three years probation for the attempted robbery conviction.
Following their convictions and sentencing, both Mr. Smith and Mr. Mack noted a joint appeal to the Court of Special Appeals, raising the same issues argued herein. The Court of Special Appeals addressed the jury instruction issue and held that the trial court did not abuse its discretion by giving a general instruction on witness identification or in declining to give a special jury instruction identifying race as a factor in the case and the problems with cross-racial identification. See James Smith and Jason Mack v. State of Maryland, 158 Md.App. 673, 704, 857 A.2d 1198, 1216 (2004). In reaching its conclusion, the court determined that there was no evidence presented during the trial to show that race played a part in the identification of the defendants or that cross-racial identification was an issue.
In addressing the denial of the opportunity to argue the difficulty of cross-racial identifications in closing, the intermediate appellate court concluded that the trial court was correct in limiting closing argument to the evidence adduced. According to the opinion, both defense counsel and the State were free to argue matters of “common knowledge” and draw any reasonable inferences from the evidence. Also, defense counsel had been permitted the opportunity to refer to the race of the defendants and the victim and argue that the victim’s identification of the defendants was flawed, but there was no evidence to suggest that race was a factor in the victim’s identification of the defendants.
The defendants filed a petition for a writ of certiorari in this Court to consider the following questions:
I. Whether the trial judge erred in refusing to instruct the jury on cross-racial identification.
II. Whether the trial judge erred in precluding defense counsel from discussing in their closing arguments the difficulties of cross-racial identification.
We granted the petition and issued the writ of certiorari. James Smith and Jason Mack v. State, 384 Md. 448, 863 A.2d [478]*478997 (2004). Because we hold that under the circumstances of this case, the trial judge erred in precluding the defendants from discussing cross-racial identification in their closing arguments and reverse the defendants’ convictions, we do not reach the jury instruction question.
Discussion
On appeal before this Court, the defendants argue that the trial court abused its discretion by barring defense counsel from arguing the difficulty of cross-racial identification in closing argument. The defendants contend that they should have been allowed to refer to matters of “common knowledge” that eyewitnesses have problems identifying individuals who are not of their own race or ethnicity and to argue reasonable inferences from the evidence adduced during the trial. Because the defendants in this case were black and the eyewitness/victim who identified the defendants was white and had opined on her ability to identify people’s faces, the defendants argue that the trial court should have permitted them to refer to cross-racial identification in closing argument.
Conversely, the State’s argument mirrors the analysis of the Court of Special Appeals that the trial court did not err in limiting defense counsel’s closing argument because there was no evidence to suggest that race was a factor in the victim’s identification of the defendants. The State also maintains that cross-racial identification is not a matter of common knowledge because of conflicting scientific research on the subject. As such, the State argues that the trial court properly exercised its discretion in precluding defense counsel from mentioning cross-racial identification in closing argument.
Cross-Racial Identification
In general, a cross-racial identification occurs when an eyewitness of one race is asked to identify a particular individual of another race. See John P. Rutledge, They All Look Alike: The Inaccuracy of Cross-Racial Identifications, 28 Am. J.Crim. L. 207, 211 (Spring 2001). Over the past half-century, a growing body of scientific laboratory and field research [479]*479concerning eyewitness cross-racial identifications has emerged suggesting that some witnesses are better able to identify members of their own race, but are significantly impaired when attempting to identify individuals of another race or ethnicity. Id. at 211. This “phenomenon” has been termed the “own-race” or “other-race” effect, or “own-race” bias. See Gary L. Wells & Elizabeth A. Olson, The Other-Race Effect in Eyewitness Identification: What Do We Do About It?, 7 Psychol. Pub. Pol’y & Law 230, 230 (March 2001).
The majority of research on the cross-race effect has been conducted in the laboratory setting where subjects have been tested in a controlled environment using what has been termed “facial recognition paradigms” to evaluate the existence of the cross-race effect in various individuals. See Siegfried Ludwig Sporer, The Cross-Race Effect, Beyond Recognition of Faces in the Laboratory, 7 Psychol. Pub. Pol’y & Law 170, 173 (March 2001). In 1984, Sheri Lynn Johnson, then an Assistant Professor of Law at Cornell University, defined a facial recognition paradigm for conducting laboratory studies regarding the ability to identify members of other races:
In the typical laboratory experiment in face recognition, subjects view photographs of a number of faces that are later randomly mixed with a new set of faces. Usually the length of observation time is carefully controlled. The subject then is asked to select the “old” laces from among the “new” faces. Each subject’s performance is measured by plotting “hits” against “false alarms,” and compiling the scores statistically into a single measure of observer sensitivity. In studies investigating the own-race effect, the performance of the subjects is aggregated by race and then each racial group’s accuracy is measured on same-race and other-race photos. Differences in the aggregated scores are then tested for statistical significance.
Shell Lynn Johnson, Cross-Racial Identification Errors in Criminal Cases, 69 Cornell L.Rev. 934, 938 (June 1984) (footnotes omitted).
[480]*480Numerous of these studies have shown that the own-race effect is strongest when white participants attempt to recognize black faces. See e.g., P. Barkowitz & John Brigham, Recognition of Faces: Own Race Bias, Incentive, and Time Delay, 12 J. Applied Soc. Psychol. 255, 261 (1982); E.J. Chance, A.G. Goldstein & L. McBride, Differential Experience and Recognition Memory for Faces, 97 J. Soc. Psychol. 243, 249, 250-51 (1975); T.S. Luce, The Role of Experience in Inter-Racial Recognition, 1 Personality & Soc. Psychol Bull. 39, 40 (1974); Roy Malpass, Racial Bias in Eyewitness Identification, 1 Personality & Soc. Psychol. Bull. 42, 43 (1974). In one study, white subjects misidentified black faces two-to-three times more often than they misidentified white faces. See Roy S. Malpass & Jerome Kravitz, Recognition for Faces of Own and Other Race, 13 J. Personality & Soc. Psychology 330, 330-34 (1969).6
Scientists conducting laboratory studies on the cross-race effect have disagreed on whether cross-racial impairment affects all races. At least four studies have found that black eyewitnesses do not experience cross-racial impairment in that they do not have difficulties in recognizing individuals from other races. See Barkowitz et al., supra, 12 J. Applied Soc. Psychol, at 261; Chance et al., supra, 97 J. Soc. Psychol, at [481]*481249, 250-51; J.F. Cross, J. Cross & J. Daly, Sex, Race, Age and Beauty As Factors In Recognition of Faces, 10 Perception & Psychophysics 393, 394 (1971); Malpass, supra, 1 Personality & Soc. Psychol. Bull, at 43. Conversely, other studies have reported that black witnesses experience some degree of own-race effect and are significantly less able to identify white laces than faces of their own race. See R.E. Galper, Functional Race Membership and Recognition of Faces, 37 Perceptual & Motor Skills 455, 458 (1973); Luce, supra, 1 Personality & Soc. Psychol Bull, at 40.7
Only three published field studies have investigated the cross-race effect. See Sporer, supra, 7 Psychol. Pub. Pol’y & Law at 176. In one field study, black and white subjects posing as customers visited a series of convenience stores browsing for a lew minutes and then went to the register to pay. See John C. Brigham et al., Accuracy of Eyewitness Identifications in a Field Setting, 42 J. Personality & Soc. Psychology 673, 681 (1982). Researchers would then ask the convenience store clerks, some black and others white, to identify the “customers” from a photo array. Id. The study found some evidence of the cross-race effect in white clerks identifying black customers. Id. According to the study, the overall accuracy rate for all participants was only 34.2%, with black participants being more accurate 69.2% than whites 39.9%. Id. Specifically, white clerks misidentified white cus[482]*482tomers 34.9% of the time and misidentified black customers 54.8% of the time. Id.
A second field study conducted in 1988 was modeled after the Brigham study and included Hispanic participants in addition to black and white clerks and customers. See Stephanie J. Platz & Harmon M. Hosch, Cross-Racial/Ethnic Eyewitness Identification: A Field Study, 18 J. Applied Soc. Psychology 972, 977-78 (1988). Overall, the study found evidence of the cross-race effect. Id. White clerks correctly identified white customers 53.2% of the time, which was significantly higher as compared to the identification of black (40.4%) or Hispanic (34.0%) customers. The sample of black participants was too small to reveal any statistically significant evidence of own-race bias. Id.8
Most recently, researchers conducted a field study in South Africa and England in which black and white participants were asked to view individuals (also black and white) in a lineup and then were asked to identify photos of individuals from the lineup. See Daniel Wright, Catherine Boyd & Cohn Tredoux, Eyewitness Identification, A Field Study of Own-Race Bias in South Africa and England, 7 Psychol. Pub. Pol’y & Law 119 (2001). The researchers found a cross-race effect in both black and white participants; however, other researchers have noted that the study’s findings are difficult to compare to previous studies because of the procedures used to compile the data. See Sporer, supra, 7 Psychol. Pub. Pol’y & Law at 177.
Overall, there is strong consensus among researchers conducting both laboratory and field studies on cross-racial identification that some witnesses are more likely to misidentify members of other races than their own. See Wells & Olson, supra, 7 Psychol. Pub. Pol’y & Law at 230 (stating that “it is [483]*483reasonable to conclude that there is internal validity to the studies showing the other-race effect”). Although many scientists and researchers conducting these studies agree that some witnesses exhibit own-race bias, they disagree on the extent to which such bias affects eyewitness identification due to the variations in the statistical data showing a cross-race effect. See Sporer, supra, 7 Psychol. Pub. Pol’y & Law at 177; Deborah Bartolomey, Cross-Racial Identification Testimony and What Not To Do About It, 7 Psychol. Pub. Pol’y & Law 247, 249 (March 2001).
In analyzing the results of the laboratory and field studies, researchers and other professionals have advanced theories to explain the existence of the cross-race effect. One initial theory was that individuals with prejudicial attitudes toward members outside of their race were more likely to exhibit own-race bias. See Christian A. Meissner & John C. Brigham, Thirty Years of Investigating the Own-Race Bias in Memory for Faces A Meta-Analytic Review, 7 Psychol. Pub. Pol’y & Law 3, 7 (2001). Early research indicated that racial attitudes appeared to influence the degree of stereotyping assigned to other-race members and that people with less prejudiced racial attitudes were better at recognizing people outside of their race. Id. This explanation, however, has not garnered wide support in the scientific community because researchers have failed to find a correlation between racial attitudes and the own-race effect. Id.
A second theory called the “ethnocentric” or “physiognomic homogeneity” explanation hypothesizes that people exhibit own-race bias because members of a particular race have similar characteristics making it difficult to differentiate among the members. Johnson, supra, 69 Cornell L.Rev. at 945. This theory also has been criticized as ignoring evidence of anthropological studies of human faces finding no difference in perceived similarities between own and other-race faces. Id. at 943-44.
A number of researchers have posited a third theory, the interracial contact theory, to explain the existence of own-race [484]*484bias, stating that the quality or number of interracial contacts may play a role in the extent of own-race bias demonstrated by a particular individual. Meissner et al., supra, 7 Psychol. Pub. Pol’y & Law at 8. Researchers have proposed that increased contact with other-race individuals may increase a person’s ability to recall features of other-race faces by:
(a) reducing the likelihood of stereotypic responses and increasing the likelihood that individuals may look for more individuating information, (b) influencing individuals’ motivation to accurately recognize other-race persons through associated social rewards and punishments, or (c) reducing the perceived complexity of unfamiliar other-race faces.
Id. The validity of the interracial contact theory is unsettled and has not been widely accepted as a cause for own-race bias. Id.
Another point of contention among researchers and scholars regarding the legitimacy of the cross-race effect is whether the results of both the laboratory and field studies are applicable to “real-world” situations where a victim or witness confronts an assailant and then later must attempt to identify that person. See Wells & Olson, supra, 7 Psychol. Pub. Pol’y & Law at 230. Proponents of the studies argue that there is “no particular reason to think that the other-race effect ... does not apply [to] eyewitnesses in actual criminal cases,” id., whereas others contend that studies on cross-racial identification “bear little resemblance to real-life crimes.” Bartolomey, supra, 7 Psychol. Pub. Pol’y & Law at 249.
There have not been studies on whether the cross-race effect is impacted by traumatic events. Id. One suggested alternative is that researchers conduct archival studies in which information from real criminal cases is accessed through court records and analyzed for evidence of the cross-race effect in eyewitnesses. See Wright et al., supra, 7 Psychol. Pub. Pol’y & Law at 120-21. This alternative, however, has not been utilized by researchers studying the own-race effect on the ground that it is difficult to draw conclusions from the information because researchers cannot control the particular [485]*485situations and it would be difficult to know if a witness actually had been mistaken in his or her identification. Id.
Because there is debate concerning the results of the studies on cross-racial identification, researchers and other professionals have questioned whether courts should recognize the own-race effect in cases involving cross-racial identification of a defendant by a victim or witness. See Wells & Olson, supra, 7 Psychol. Pub. Pol’y & Law at 230; Rutledge, supra, 28 Am.J.Crim. L. at 211, compare Bartolomey, supra, 7 Psychol. Pub. Pol’y & Law at 230.
Some courts, however, when addressing the issue of a cross-racial identification of a criminal defendant, without crediting the studies themselves, have permitted jury instruction or closing argument on cross-racial identification. See e.g., People v. Carrieri, 4 Misc.3d 307, 777 N.Y.S.2d 627, 629 (N.Y.App. 2004) (permitting closing argument on cross-racial identification and stating that the quality of cross-racial witness identification fell within the ambit of jurors’ general knowledge and life experience); State v. Wiggins, 74 Conn.App. 703, 813 A.2d 1056, 1059 (2003) (stating that “closing argument may be employed to demonstrate the problems that might arise as a result of cross-racial identification”); State v. Cromedy, 158 N.J. 112, 727 A.2d 457, 467-68 (1999) (allowing jury instruction on cross-racial identification and stating that there is a “widely held commonsense view that members of one race have greater difficulty in accurately identifying members of a different race”); People v. Sanders, 11 Cal.4th 475, 46 Cal.Rptr.2d 751, 905 P.2d 420, 435 (1995) (holding that defendant had not been prejudiced by trial court’s refusal of expert testimony on eyewitness identification because defense counsel had been allowed to argue the problems of cross-racial identification in closing argument); State v. Cunningham, 863 S.W.2d 914, 923 (Mo.App.1993) (noting that counsel may discuss the problems with cross-racial identification in closing argument); State v. Patterson, 103 N.C.App. 195, 405 S.E.2d 200, 207 (1991) (upholding defendant’s conviction based upon eyewitness identification, in part, because counsel had been allowed to argue [486]*486the difficulties of cross-racial identification in closing argument).
Closing Argument
In this instance, the gist of the issue before us is whether the trial court should have allowed defense counsel to argue the difficulties of cross-racial identification in closing argument. “It is well settled that a criminal defendant’s Sixth Amendment right to counsel guarantees, in part, an opportunity for counsel to present closing argument at the close of the evidence.” Holmes v. State, 333 Md. 652, 658-59, 637 A.2d 113, 116 (1994), citing Herring v. New York, 422 U.S. 853, 858-65, 95 S.Ct. 2550, 2553-57, 45 L.Ed.2d 593, 598-602 (1975); Cherry v. State, 305 Md. 631, 636-47, 506 A.2d 228, 230-36 (1986); Spence v. State, 296 Md. 416, 419, 463 A.2d 808-09 (1983); Yopps v. State, 228 Md. 204, 208, 178 A.2d 879, 882 (1962). The United States Supreme Court in Herring v. New York, 422 U.S. 853, 95 S.Ct. 2550, 45 L.Ed.2d 593 (1975), discussed the importance of closing arguments:
It can hardly be questioned that closing argument serves to sharpen and clarify the issues for resolution by the trier of fact in a criminal case. For it is only after all the evidence is in that counsel for the parties are in a position to present their respective versions of the case as a whole. Only then can they argue the inferences to be drawn from all the testimony, and point out the weaknesses of their adversaries’ positions. And for the defense, closing argument is the last clear chance to persuade the trier of fact that there may be reasonable doubt of the defendant’s guilt. The very premise of our adversary system of criminal justice is that partisan advocacy on both sides of the case will best promote the ultimate objective that the guilty be convicted and the innocent go free.
Id. at 862, 95 S.Ct. at 2555, 45 L.Ed.2d at 600 (internal citations omitted).
In Wilhelm v. State, 272 Md. 404, 326 A.2d 707 (1974), this Court discussed the scope of permissible closing argument:
[487]*487As to summation, it is, as a general rule, within the range of legitimate argument for counsel to state and discuss the evidence and all reasonable and legitimate inferences which may be drawn from the facts in evidence; and such comment or argument is afforded a wide range. Counsel is free to use the testimony more favorable to his side of the argument to the jury, and the evidence may be examined, collated, sifted and treated in his own way.... Generally, counsel has the right to make any comment or argument that is warranted by the evidence proved or inferences therefrom; the prosecuting attorney is as free to comment legitimately and to speak fully, although harshly, on the accused’s action and conduct if the evidence supports his comments, as is accused’s counsel to comment on the nature of the evidence and the character of witnesses which the prosecution produces.
Id. at 412-13, 326 A.2d at 714. Judge O’Donnell, wilting for this Court continued to explain:
While arguments of counsel are required to be confined to the issues in the cases on trial, the evidence and fair and reasonable deductions therefrom, and to arguments to opposing counsel, generally speaking, liberal freedom of speech should be allowed. There are no hard-and-fast limitations within which the argument of earnest counsel must be confined — no well-defined bounds beyond which the eloquence of an advocate shall not soar. [Counsel] may discuss the facts proved or admitted in the pleadings, assess the conduct of the parties, and attack the credibility of witnesses. [Counsel] may indulge in oratorical conceit or flourish and in illustrations and metaphorical allusions.
Id. at 412-13, 326 A.2d at 714 (internal citations omitted). We also have held that in closing argument “[j]urors may be reminded of what everyone else knows, and they may act upon and take notice of those facts which are of such general notoriety as to be matters of common knowledge.” Wilhelm, 272 Md. at 438, 326 A.2d at 728.
[488]*488Thus, during closing argument, counsel may “state and discuss the evidence and all reasonable and legitimate inferences which may be drawn from the facts in evidence,” Henry v. State, 324 Md. 204, 230, 596 A.2d 1024, 1037 (1991), in addition to argue matters of common knowledge, see Wilhelm, 272 Md. at 438, 326 A.2d at 728. “Subject to the trial court’s discretion, both the State’s Attorney and defense counsel are given wide latitude in the conduct of closing argument, including the right to explain or to attack all the evidence in the case.” Trimble v. State, 300 Md. 387, 405, 478 A.2d 1143 (1984), cert denied, 469 U.S. 1230, 105 S.Ct. 1231, 84 L.Ed.2d 368 (1985). Closing argument, however, is not without limitation, in that the court should not permit counsel to state and comment upon facts not in evidence or to state what he or she would have proven. Wilhelm, 272 Md. at 414-15, 326 A.2d at 714-15. What exceeds the limits of permissible comment or argument by counsel depends on the facts of each case.
The case sub judice involves the victim’s eyewitness cross-racial identification of the defendants, which was the sole piece of significant evidence. During her testimony, Ms. Crandall described the physical characteristics of the defendants and opined on her ability to identify them. She stated that she was “extremely good with faces,” looked for distinctive features, and was “obsessed with people’s postures,” as the basis for identifying the defendants. Ms. Crandall further qualified her experience in recognizing people by stating that she was a teacher and had been studying art and painting people since childhood.
Defense counsel clearly was entitled to challenge Ms. Crandall’s “educated” identification of the defendants by arguing to the jury that her identification should not be accorded the weight that she credited to her own ability to identify them. At this juncture the extent to which own-race bias affects eyewitness identification is unclear based on the available studies addressing this issue, so that we cannot state with certainty that difficulty in cross-racial identification is an established matter of common knowledge. Here, however, the victim’s identification of the defendants was anchored in her [489]*489enhanced ability to identify faces. Under these circumstances, defense counsel should have been allowed to argue the difficulties of cross-racial identification in closing argument.
Therefore, we conclude that the trial court erred in refusing to allow defense counsel to comment on cross-racial identification in closing argument. As such, the verdicts of guilt as to both defendants cannot stand and any sentence apportioned thereto must be vacated. We further hold that Mr. Mack’s sentence of six years imprisonment for attempted robbery and Mr. Smith’s sentence of six years imprisonment, with all but two years suspended, followed by three years of probation for attempted robbery, are reversed.
JUDGMENT OF THE COURT OF SPECIAL APPEALS REVERSED AND CASE REMANDED TO THAT COURT WITH DIRECTIONS TO REVERSE THE JUDGMENT OF THE CIRCUIT COURT FOR BALTIMORE CITY AND TO REMAND THE CASE TO THAT COURT FOR A NEW TRIAL. COSTS IN THIS COURT AND IN THE COURT OF SPECIAL APPEALS TO BE PAID BY THE MAYOR AND CITY COUNCIL OF BALTIMORE.