MARBLEY, D. J., delivered the opinion of the court, in which COLE, J., joined. BATCHELDER, J. (pp. 318-32), delivered a separate dissenting opinion.
MARBLEY, District Judge.Appellant James Smithers was convicted of bank robbery in violation of 18 U.S.C. § 2113(a). Smithers now appeals various aspects of his trial, including the district court’s exclusion of the testimony of an eyewitness identification expert, the limitation of Smithers’s wife’s testimony, and the district court’s response to questions posed by the jury after it began deliberating. For the following reasons, we REVERSE the conviction below and REMAND this case for a new trial pursuant to the law set forth herein.
I.
On the morning of November 12, 1996, a man walked into the Monroe Bank and Trust in Terence, Michigan, and presented bank teller Teresa Marino a note. The note read, “I have a gun. Give me your large bills.” Ms. Marino complied with the demand by turning over the money from her teller drawer. The robber asked for more money, and Ms. Marino unlocked her other drawer and gave him three packs of large bills totaling $3,400. When the robber repeated his demand for more money, Ms. Marino told him that was all she had, and he ran from the bank. The entire incident lasted about two minutes.
Two other witnesses observed the robbery. The first, Debra White, was also working as-a teller at the bank on November 12, 1996. She was sitting at a desk behind Ms. Marino when she noticed an unfamiliar customer standing at Ms. Mari-no’s teller station. Ms., White looked away for a moment and when she looked back, the man grabbed the money and walked quickly out of the bank. Ms. White asked Ms. Marino if she had been robbed. Learning that she had, Ms. White yelled that they had been robbed and went to lock the bank doors. While doing so, she observed the robber getting into the passenger side of a car parked in the parking lot.
Timothy Wilson, the second witness, was a bank customer who walked into the bank at the same time as the robber. The robber held the door open for him as they both entered the building. Mr. Wilson saw the robber go straight to the teller and then leave the bank quickly.
Investigators from the Monroe County Sheriffs Department spoke to the witnesses that day. Ms. Marino, who was approximately three feet from the robber, described him as a white male in his late twenties wearing a Nike jacket, baseball cap and sunglasses, over 6’ 2” tall, 180-185 pounds, with long bushy dark hair, a mous-tache and a thin beard. Ms. White described the robber as taller than average, with squinty eyes and wearing a bulky striped jacket. Ms. White described the *309car as a two-toned brown and black, late 1970’s Monte Carlo, with a cream colored landau roof and an Ohio license plate. Mr. Wilson recalled the robber as a very tall man, with a moustache and partial beard, wearing a baseball cap, dark sunglasses and a winter jacket.
The next day, officers of the Toledo Police Department noticed a vehicle fitting the description of the car used in the robbery at an apartment complex in Toledo. Monroe County Detective Thomas Redmond drove Ms. White to the vehicle, a 1976 Oldsmobile Cutlass, which she identified as the car used in the robbery. The car was registered to James Smithers.
Officers then went to Smithers’s home, where his wife, Josette Smithers, informed them that he was at his parents’ house. The officers searched Smithers’s apartment but found no incriminating evidence. They located Smithers at his parents’ home, and he accompanied the police to his apartment. Smithers told the officers that he bought the vehicle from his brother-in-law, Steve Dallas, who still retained a set of keys to the car. Smithers also stated that on the morning of November 12, 1996, he had noticed his rear license plate was missing, so he had moved his front plate to the rear. He also claimed to have noticed gas missing from the car on other mornings; later, Smithers said that there was a hole in the gas tank. Smithers consented to a search of the car, which produced no incriminating evidence. Smithers voluntarily went to the sheriffs department where he provided handwriting samples and was photographed and fingerprinted. When photographing him, Detective Redmond noted Smithers’s height as 6’ 6 /£”.
Detective Redmond prepared a photo spread of six photographs, including a photo of Smithers. On November 14, 1996, Detective Redmond showed the photo spread to Ms. Marino, Mr. Wilson and Ms. White. Ms. Marino and Mr. Wilson could not identify the robber from the photo spread. Ms. White picked out Mr. Smith-ers. Immediately after her identification, Ms. White told Ms. Marino that she had been able to identify the robber from the photo spread.
Smithers’s handwriting exemplars were submitted to the FBI laboratory for analysis. The results were inconclusive. The demand note was submitted to the Michigan State Police Laboratory for fingerprint analysis. The analysis produced one identifiable print. The government claims the print was inconclusive; Smithers claims the analysis showed that the print did not belong to him.
Peter Smith, an FBI examiner who specializes in analyzing exhibits in photographic form, performed a height analysis of the robber depicted in the bank videotape. Mr. Smith concluded that the robber measured approximately 6’ 5”. Mr. Smith also conducted a comparative analysis of the robber in the bank photos with a photograph of Smithers. He could neither positively identify nor eliminate Smithers as the bank robber.
On June 16, 1997, a grand jury returned an indictment charging Smithers with one count of bank robbery in violation of 18 U.S.C. § 2113(a).
On December 18, 1997, Smithers filed a ten-page motion in limine to determine the admissibility of certain expert testimony regarding eyewitness testimony. The district court commenced Smithers’s jury trial on January 14, 1998. After the jury was empaneled, the district court heard argument on Smithers’s motion in limine, and denied the motion, noting that everything an expert would have to say about eyewitness identification was within the jury’s “common knowledge” The court stated that it would give an instruction on eyewitness testimony. Smithers’s attorney requested permission to make a written proffer, which the court allowed.
The government presented its case, including eyewitness testimony from Ms. Marino, Ms. White and Mr. Wilson. Despite their prior inability to identify Smith-ers from a photo spread, Ms. Marino and *310Mr. Wilson identified Smithers as the robber in court. Ms. Marino and Ms. White testified that they did not notice that the robber had any distinguishing features. The government rested on January 16, 1998.
Smithers filed his renewed motion in limine and offer of proof, regarding expert testimony, on eyewitness identification on January 20, 1998. This proffer described the anticipated testimony of Dr. Solomon Fulero, a proposed expert on eyewitness identification. It noted that Dr. Fulero would “educate the jury about the general factors that may affect eyewitness accuracy,” including the specific the issues of: (1) “detail salience” (the fact that eyewitnesses tend to focus on unusual characteristics of people they observe); (2) the relationship between the time that has passed since observing the event and the accuracy of recalling it; (3) the effect of post-identification events on memory; (4) the fact that when one person both prepares and administers a photo spread, the likelihood of. misidentification increases; (5) the “conformity effect” (the fact that witnesses’ memories are altered by talking about the event with each other after it occurs); and (6) the relationship between a witness’s confidence in her recollection and its accuracy. Regarding the issue of detail salience, the proffer stated that “[h]ad Mr. Smithers been the robber, the eyewitnesses would have observed and been able to recall the large scar on Mr. Smithers’ [sic] neck.”
After hearing oral argument on the Defendant’s renewed motion, the district court ruled that it would exclude the expert testimony:
[primarily because it’s late in the day. It should have been done much earlier. On the other hand, I think you’ve got a very good, if there’s a conviction, I think you’ve made an excellent record that I’ve abused my discretion in failing to allow it, and I think there’s a certain — I prefer to see it that way.
The court also opined that Dr. Fulero’s testimony was “not a scientifically valid opinion,” “a jury can fully understand that its [sic] got an obligation to be somewhat skeptical of eyewitness testimony,” and “admission of Dr. Fulero’s testimony is in this case is almost tantamount to the Court declaring the defendant not guilty as a matter of law.... [A]bsent the eyewitness testimony I don’t think there’s enough here to go to the jury.” Finally, the district court remarked, “I’m also interested in seeing what a jury will do absent that expert testimony. It makes it a more interesting case. I recognize it’s the defendant’s fate that’s at stake, but you can always argue for a new trial if he’s convicted.”
After this ruling, Smithers presented a few witnesses, including his wife, who attempted to establish an alibi defense. Ms. Smithers testified that Smithers was sleeping in their house from 3:00 a.m. to 11:30 a.m. the morning of November 12, and that as a light sleeper she would have heard her husband leave the apartment. Ms. Smithers also spoke about her husband’s appearance, maintaining that Smithers weighed 245 pounds in November of 1996, is 6’ 8” tall and has a four-inch long scar on the right front side of his neck.
The case was submitted to the jury on January 21, 1998. The next day, the jury returned a verdict of guilty. The district court sentenced Smithers on June 4, 1998, to a forty-one month term of imprisonment. Smithers timely filed a notice of appeal on June 8, 1998. Smithers now appeals various aspects of his trial, only one of which we address today: the exclusion of Dr. Fulero as an eyewitness expert.
II.
Generally, a trial court’s evidentiary determinations are reviewed for an abuse of discretion. See United States v. Moore, 954 F.2d 379, 381 (6th Cir.1992).
*311Smithers argues that the district court’s denial of his motion to introduce testimony by an identification expert warrants the reversal of his conviction. The crucial element of the government’s case was eyewitness identification of the defendant and his car, Smithers argues, and Dr. Fulero’s testimony involved a proper subject that would have been helpful to the jury in evaluating this issue. Smithers, therefore, contends that the decision to exclude this expert’s testimony, to indulge the district judge in his rather eccentric courtroom experiment, was improper. The government counters that the district court’s decision was well within its discretion. The district court properly excluded Dr. Fulero’s testimony, the prosecution argues, based upon its lack of scientific validity, invasion of the jury’s province, possibility of confusion and the tardiness of Smithers’s proffer.
Courts’ treatments of expert testimony regarding eyewitness identification has experienced a dramatic transformation in the past twenty years and is still in a state of flux. Beginning in the early 1970’s, defense attorneys began to bring expert testimony into the courtroom. Then, courts were uniformly skeptical about admitting such testimony, elaborating a host of reasons why eyewitness experts should not be allowed to testify. In the first case to address the issue, United States v. Amaral, 488 F.2d 1148 (9th Cir.1973), the Ninth Circuit held that the district court did not err in excluding expert testimony regarding eyewitness identification because cross-examination was sufficient to reveal any weaknesses in the identifications. After that decision, a series of cases rejected similar evidence for a variety of reasons. See, e.g., United States v. Purham, 725 F.2d 450, 454 (8th Cir.1984) (finding the question is within the expertise of jurors); United States v. Thevis, 665 F.2d 616, 641 (5th Cir.1982) (reasoning that identification was adequately addressed through cross-examination); United States v. Sims, 617 F.2d 1371, 1375 (9th Cir.1980) (finding no general acceptance in scientific community); United States v. Posher, 590 F.2d 381, 383 (1st Cir.1979) (ruling that the testimony would be prejudicial).
This trend shifted with a series of decisions in the 1980’s, with the emerging view that expert testimony may be offered, in certain circumstances, on the subject of the psychological factors which influence the memory process. See, e.g., United States v. Moore, 786 F.2d 1308, 1313 (5th Cir.1986) (finding that “[i]n a case in which the sole testimony is casual eyewitness identification, expert testimony regarding the accuracy of that identification is admissible and properly may be encouraged ... ”); United States v. Downing, 753 F.2d 1224, 1232 (3d Cir.1985) (reasoning that “expert testimony on eyewitness perception and memory [should] be admitted at least in some circumstances”); United States v. Smith, 736 F.2d 1103, 1107 (6th Cir.1984) (“The day may have arrived, therefore, when Dr. Fulero’s testimony can be said to conform to a generally accepted explanatory theory.”). State court decisions also reflect this trend. See, e.g., State v. Buell, 22 Ohio St.3d 124, 489 N.E.2d 795 (1986) (overruling per se rule and holding expert testimony admissible to inform jury about factors generally affecting memory process). Indeed, several courts have held that it is an abuse of discretion to exclude such expert testimony. See, e.g., United States v. Stevens, 935 F.2d 1380, 1400-01 (3d Cir.1991) (reversing and remanding for new trial); Smith, 736 F.2d at 1107 (holding error harmless in light of other inculpatory evidence); Downing, 753 F.2d at 1232 (holding error harmless in light of other inculpatory evidence); State v. Chapple, 135 Ariz. 281, 660 P.2d 1208 (1983) (reversing and remanding for new trial). This jurisprudential trend is not surprising in light of modern scientific studies which show that, while juries rely heavily on eyewitness tes-*312timon'y, it can be untrustworthy under certain circumstances.1
Recognizing the dichotomy between eyewitness errors and jurors’ reliance on eyewitness testimony, this Circuit has held that expert testimony on the subject of eyewitness identification is admissible. In United States v. Smith, 736 F.2d 1103 (6th Cir.1984), this Court held that a trial court abused its discretion in excluding such an expert. In Smith, the defendant sought to introduce the testimony of psychologist Solomon Fulero — the same expert Smith-ers attempted to introduce at his trial — as an expert in the field of eyewitness identification to shed light upon an eyewitness’s testimony. The lower court excluded the testimony, finding that it was inadmissible pursuant to Federal Rule of Evidence 403. On appeal, this Court applied the four prong test for expert testimony articulated in United States v. Green, 548 F.2d 1261 (6th Cir.1977): (1) that the witness, a qualified expert, (2) was testifying to a proper subject, (3) which conformed to a generally accepted explanatory theory, and (4) the probative value of the testimony outweighed its prejudicial effect.
Applying that standard, the Court noted that the offered testimony would have *313been based on “a hypothetical factual situation identical” to the facts of the case and would have explained: (1) that a witness who does not identify the defendant in a first line-up may “unconsciously transfer” his visualization of the defendant to a second line-up and thereby incorrectly identify the defendant the second time; (2) that studies demonstrate the inherent unreliability of cross-racial identifications; and (3) that an encounter during a stressful situation decreases the eyewitness’s ability to perceive and remember and decreases the probability of an accurate identification. See Smith, 736 F.2d at 1105-06. The Smith Court held that expert testimony on the reliability of eyewitness identification met the “helpfulness” test of Federal Rule of Evidence 702 and therefore had been excluded improperly at trial. The Court explained that “[i]n reviewing a 403 balancing, the court must look at the evidence in the light most favorable to the proponent, maximizing its probative value and minimizing its prejudicial effect,” id. at 1107, and concluded that “[s]ueh testimony might have been relevant to the exact facts before the court and not only might have assisted the jury, but might have refuted their otherwise common assumptions about the reliability of eyewitness identification.” Id. at 1106. Further, the Smith Court expressed its acceptance of psychological studies as a scientifically sound and proper subject of expert testimony, noting, “[t]he science of eyewitness perception has achieved the level of exactness, methodology and reliability of any psychological research.” Id. at 1107.
Smith’s conviction was nonetheless affirmed on the ground that any error by the district court in excluding the proffered testimony was harmless. The Smith Court noted that the government had not only presented three witnesses who identified the defendant as the perpetrator, but that the defendant’s palm print was recovered at the scene of the crime, thus “wholly discrediting the defendant’s alibi” defense. Id. at 1107-08. Because there was other significant inculpatory evidence, the trial court’s error was deemed harmless, and the defendant’s conviction was affirmed.
Smithers now argues that the proper standard for the admission of eyewitness expert testimony is that set out in Smith. We disagree. The significance of Smith in terms of evaluating expert testimony is questionable after the landmark decision of Daubert v. Merrell Dow Pharmaceuticals, 509 U.S. 579, 113 S.Ct. 2786, 125 L.Ed.2d 469 (1993). In Daubert, the Supreme Court articulated the test that trial courts must use in determining whether scientific evidence and testimony is admissible. According to Daubert, a district court must “ensure that any and all scientific testimony or evidence admitted is not only relevant, but reliable.” Id. at 589, 113 S.Ct. 2786. Daubert thus requires trial courts to perform a two-step inquiry. First, the court must determine whether the expert’s testimony reflects “scientific knowledge,” that is, the court must make “a preliminary assessment of whether the reasoning or methodology underlying the testimony is scientifically valid and of whether that reasoning or methodology properly can be applied to the facts in issue.” Id. at 592-93, 113 S.Ct. 2786. Second, the court must ensure that the proposed expert testimony is relevant to the task at hand and will serve to aid the trier of fact. See id. The Supreme Court referred to this second prong as the “fit” requirement. See id.
Citing the concurring opinion of Justice Scalia’s in Kumho Tire Co. v. Carmichael, 526 U.S. 137, 119 S.Ct. 1167, 143 L.Ed.2d 238 (1999), the Dissent proclaims that Daubert is not “holy writ” to evaluate proffered experts under Rule 702. While it is true that the Daubert factors “do not constitute a ‘definitive checklist or test ...,’” Kumho Tire, 119 S.Ct. at 1175 (citing Daubert, 509 U.S. at 593, 113 S.Ct. 2786), the Supreme Court did conclude that “[a] trial court should consider the specific factors identified in Daubert where they are reasonable measures of the reliability of *314expert testimony.” Id. at 1176. The Court also stressed:
We conclude that Daubert’s general principles apply to the expert matters described in Rule 702. The Rule, in respect to all such matters, “establishes a standard of evidentiary reliability.” ... It “requires a valid ... connection to the pertinent inquiry as a precondition to admissibility.” ... And where such testimony’s factual basis, data, principles, methods, or them application are called sufficiently into question ... the trial judge must determine whether the testimony has “a reliable basis in the knowledge and experience of [the relevant] discipline.”
Id. at 1175 (citations omitted) (emphasis added).
The Supreme Court in Kumho indicated that the standards set forth in Daubert, depending on the “particular circumstances of the particular case,” id., should be flexibly applied. Contrary to the Dissent, the Supreme Court’s reasoning does not indicate that Daubert should be abandoned totally. This Court finds that in the case sub judice, given the expert and the testimony that was proffered, the standards of Daubert should have been applied.2
While it is true that several postDaubert eyewitness identification cases have found that the exclusion of the testimony was not an abuse of discretion, see,e.g., United States v. Hall, 165 F.3d 1095 (7th Cir.1999); United States v. Smith, 156 F.3d 1046 (10th Cir.1998); United States v. Smith, 122 F.3d 1355 (11th Cir.1997); United States v. Kime, 99 F.3d 870 (8th Cir.1996); United States v. Brien, 59 F.3d 274 (1st Cir.1995); United States v. Rincon, 28 F.3d 921 (9th Cir.1994), the lesson from these cases is not that expert testimony on eyewitness identification is never appropriate; rather, the cases indicate that courts must consider whether the testimony would be helpful or confusing to the jury. The cases also discuss whether this type of testimony touched on the “ultimate issue” in the case and therefore usurped the jury’s role; whether there was other evidence against the defendant; and whether the jury could more properly evaluate the reliability of eyewitness testimony through cross-examination. In light of these cases, we believe that the district court should have performed its analysis under the rule of Daubert, rather than, as Smithers argues, that of Smith. In any event, the trial court did not analyze the admissibility of the expert testimony in this case under either of these cases.
We find that the district court abused its discretion in excluding Dr. Fulero’s testimony, without first conducting a hearing pursuant to Daubert. There are several bases for this conclusion. As a threshold consideration, we address the district court’s “experiment” comment. The district court explained that it was interested in seeing what a jury would do absent the expert testimony because it would make the trial “more interesting.” The district court stated: *315This comment is gamesmanship at its worst and reveals a troubling disregard for this Defendant’s rights, relegating those rights to mere abstractions. The district court’s reasoning that it could indulge in this experiment because Smithers could “always appeal” ironically turned this trial into a laboratory experiment where the judge felt free to play with evidentiary variables at the cost of the Defendant’s rights. Basing an evidentiary decision on personal curiosity rather than on applicable case law and the rules of evidence is a patent abuse of discretion.
*314I’m also interested in seeing what a jury will do absent that expert testimony. It makes it a 'more interesting case. I recognize it’s the defendant’s fate that’s at stake, but you can always argue for a new trial if he’s convicted.
*315We do not, however, base our decision on the district court’s “experiment” comment alone. Even without this comment, the district court erred in its evidentiary analysis by failing to apply the Daubert test to the proposed expert testimony. Although the decision of whether to admit a witness’s testimony is left to the sound discretion of the trial court, a trial court cannot make an arbitrary decision. When a defendant’s liberty is at stake, it is incumbent upon the trial court to apply the correct law, follow the appropriate decision-making steps and articulate the bases upon which its decision rests. Here, the district court should have applied the analytical principles set forth in Daubert, but it did not.
Under Daubert, a trial court should consider: (1) whether the reasoning or methodology underlying the expert’s testimony is scientifically valid; and (2) whether that reasoning or methodology properly could be applied to the facts at issue to aid the trier of fact. The district court, in neglecting to undertake a Daubert analysis, failed to take these factors into consideration. Indeed, the district court did not make any determination as to this expert’s scientific reasoning or methodology. We find that if the district court had given this issue proper consideration, it may have deemed Dr. Fulero’s testimony scientifically valid.
Following Kumho Tire, 119 S.Ct. at 1176, we next consider the way the district court may have examined the Daubert factors in the present case. Tellingly, this Court has already accredited Dr. Fulero’s science and methodology. In Smith, this Court not only noted the jurisprudential movement toward admitting psychological studies of eyewitness experts in general, but praised the qualifications and scientific methods of this- same expert witness, Dr. Fulero. In addition, the district court could have concluded that this testimony— describing psychological factors such as detail salience, the conformity effect, the dynamics of photo identifications and the confidence-accuracy relationships — could have been applied to the facts at issue in this case. Information about the effects of detail salience would bear on the witnesses’ failure to notice Smithers’s conspicuous scar; evidence about the conformity effect would apply to Ms. Marino’s and Mr. Wilson’s ability to identify Smithers only after they had spoken with Ms. White; the suggestibility of photo identifications created and administered by a single person would apply to the procedures that Detective Redmond used; and explaining the lack of correlation between confidence and accuracy would bear upon the credibility of all of the eyewitnesses. Had the district court conducted a proper evaluation of this testimony, we believe it may have found that Dr. Fulero’s testimony met the first requirement of the Daubert test.
The trial court should have next considered whether the proposed expert testimony was relevant to the task at hand and would aid the trier of fact. The district court did, to some extent, discuss this second Daubert prong (even if it did not explicitly note that it was doing so), by stating that “a jury can fully understand” its “obligation to be somewhat skeptical of eyewitness testimony.” This point addresses whether the testimony would “aid the trier of fact.” The court’s statement, however, is simply wrong, and the district court, on remand, should reconsider this factor. As noted above, jurors tend to be unduly receptive to, rather than skeptical *316of, eyewitness testimony. Further, accepting the district court’s analysis that all jurors are aware of their obligation to be skeptical would lead to absurd results: expert testimony on eyewitness identification would never be admissible. As demonstrated by abundant ease law, this is not the conclusion that has been reached by courts addressing this issue. Today, there is no question that many aspects of perception and memory are not within the common experience of most jurors, and in fact, many factors that affect memory are counter-intuitive. In Smith we recognized the expediency of expert testimony to address these complex issues and to inform jurors fully of the issues they must decide.
The Dissent counters by arguing that eyewitness identification experts are not necessary because cross-examination and jury instructions should be the tools- used in a trial to discredit and flush-out eyewitness testimony. Unfortunately, the Dissent’s homage to trial procedures does not extend to expert witness testimony. The same argument can be made for the admission of expert testimony: cross-examination and jury instructions can be used to question the qualifications of the proffered expert, undermine the basis of the expert’s theories, explain the limits of social science’s validation studies and pick apart research methods. The only reason given by the Dissent for why cross-examination and jury instructions can serve these goals for eyewitness testimony, but not for expert testimony, is that the jury may take the expert’s testimony as “scientifically irrefutable truth.” The Dissent’s selective faith in the collective intelligence, common sense and decision-making ability of the jury is disheartening, and is also inconsistent with the Dissent’s deference to the jury on other matters, including judging the credibility of eyewitness identifications.
Further, based on the comment that Smithers’s proffer of Dr. Fulero’s testimony was “too late in the day,” the Dissent crafts a legal basis for the district court’s exclusion based on Federal Rule of Evidence 403. The Dissent concludes that Rule 403 permits the exclusion of relevant evidence based on “delay.” Fed.R.Evid. 403. The Dissent misquotes and misconstrues the meaning of “delay” in Rule 403. Not all delay authorizes the exclusion of relevant evidence — only “undue delay.” Moreover, the term “delay” does not connote delay in the submission of motions or proffers; rather, it encompasses the prolonging of the length of the trial, and can be read properly in conjunction with the other exclusionary factors: “waste of time, or needless presentation of cumulative evidence.” See, e.g., John McShain, Inc. v. Cessna Aircraft Co., 563 F.2d 632 (3d Cir.1977); United States v. International Bus. Mach., 87 F.R.D. 411 (S.D.N.Y.1980); SCM Corp. v. Xerox Corp., 77 F.R.D. 10 (D.Conn.1977).
“Delay” is a consideration of efficiency and is not readily distinguishable from “waste of time.” 22 Charles A. Wright & Kenneth W. Graham Jr., Federal Practice And Procedure: Evidence § 5218, at 296 (1978); see also Christopher B. Mueller & Laird C. Kirkpatricx, Evidence § 4.5 (1995) (concluding that “undue delay, waste of time or needless presentation of cumulative evidence” are concerns for the “concessions to the shortness of life,” “the limited resources of the judicial system,” and the presentation of cumulative evidence) (footnote omitted). “Delay” in Rule 403 does not mean “filed late” as the Dissent concludes.
Furthermore, the eases cited by the Dissent to support the contention that the basis for the district court’s exclusion of Dr. Fulero’s testimony was a consideration of “delay” under Rule 403 do not explicitly cite to the Rule nor do they mention delay as a factor. See United States v. Curry, 977 F.2d 1042, 1052 (7th Cir.1992); United States v. Dowling, 855 F.2d 114, 118 (3d Cir.1988).
The exclusion of Dr. Fulero’s testimony because the evidence was presented “late in the day,” contrary to the Dis*317sent’s assertion, was not a proper basis for exclusion. First, the Defendant filed Ms ten-page motion in limine requesting a ruling on this issue a full month before trial. At the beginning of trial, Smithers renewed his motion orally. A week later, he submitted an additional seven-page brief on the subject. Thus, it is impossible to say that either the court or the government did not have adequate notice of the issue. Second, “a criminal defendant’s relevant evidence may generally not be excluded on the basis of a discovery sanction. The defendant’s Sixth Amendment right to an effective defense will usually outweigh the interest served by pretrial discovery orders.” United States v. Collins, No. 87-5077, 1988 WL 4434, at *2 (6th Cir. Jan.25, 1988). Given the importance of eyewitness testimony in this case, the district court should not have excluded Dr. Fulero’s testimony based on its supposed tardiness.3
Finally, we find that the trial court’s error was not harmless. The complexion of the proceedings likely would have changed had the district court conducted a Daubert hearing and determined that Dr. Fulero’s testimony was admissible. And, as the Dissent properly points out, expert testimony should be admitted in the precise situation presented to the trial court in this case — that is, when there is no other inculpatory evidence presented against the Defendant with the exception of a small number of eyewitness identifications. See Smith, 736 F.2d at 1107; Moore, 786 F.2d at 1313; Downing, 753 F.2d at 1226.4 Here, eyewitness testimony was the crucial, if not the sole basis for Smithers’s conviction. The district court in this case concluded that “[a]dmission of Dr. Fulero’s testimony is in this case is almost tantamount to the Court declaring the defendant not guilty as a matter of law .... absent the eyewitness testimony I don’t think there’s enough here to go to the jury.” The lower court did not seem to realize that eyewitness expert testimony is most appropriate in such situations.5
*318The district court should have conducted a hearing under Daubert and analyzed the evidence to determine whether Dr. Fule-ro’s proffered testimony reflects scientific knowledge, and whether the testimony was relevant and would have aided the trier of fact. Based on its failure to perform the correct legal analysis — the Daubert analysis — as well as its “experiment” rationale for excluding the testimony, we find that the district court abused its discretion. We therefore REVERSE Smithers’s conviction and REMAND this case for proceedings in accordance with this decision.6
. A plethora of recent studies show that the accuracy of an eyewitness identification depends on how the event is observed, retained and recalled. See generally Roger V. Handberg, Expert Testimony on Eyewitness Identification: A New Pair of Glasses for the Jury, 32 Am. Crim. L. Rev. 1013, 1018-22 (1995). Memory and perception may be affected by factors such as:
(1) the retention interval, which concerns the rate at which a person’s memory declines over time; (2) the assimilation factor, which concerns a witness’s incorporation of information gained subsequent to an event into his or her memory of that event; and (3) the confidence-accuracy relationship, which concerns the correlation between a witness’s confidence in his or her memory and the accuracy of that memory. Other relevant factors include: (4) stress; (5) the violence of the situation; (6) the selectivity of perception; (7) expectancy; (8) the effect of repeated viewings; (9) and the cross-racial aspects of identification, that is where the eyewitness and the actor in the situation are of different racial groups.
Alan K. Stetler, Particular Subjects of Expert and Opinion Evidence, 31A Am Jdr. Expert § 371 (1989). Accordingly, "a jury should consider several factors in judging the accuracy of an eyewitness identificátion. Social science data suggests, however, that jurors are unaware of several scientific principles affecting eyewitness identifications.” Handberg, supra, at 1022. In fact, because many of the factors affecting eyewitness impressions are counter-intuitive, many jurors’ assumptions about how memories are created are actively wrong. See Downing, 753 F.2d at 1231 (finding that "[[¡actors bearing on eyewitness identification may be known only to some jurors, or may be imperfectly understood by many, or may be contrary to the intuitive beliefs of most”) (citations omitted).
This ignorance can lead to devastating results. One study has estimated that half of all wrongful convictions result from false identifications. See Elizabeth F. Loftus, Ten Years in the Life of an Expert Witness, 10 Law & Hum. Behav. 241, 243 (1986) (citing a 1983 Ohio State University doctoral dissertation). And ”[i]t has been estimated that more than 4,250 Americans per year are wrongfully convicted due to sincere, yet woefully inaccurate eyewitness identifications.” Andre A. Moenssens et al., Scientific Evidence in Civil and Criminal Cases § 19.15, at 1171-72 (4th ed.1995) (citing United States v. Wade, 388 U.S. 218, 87 S.Ct. 1926, 18 L.Ed.2d 1149 (1967)). A principal cause of such convictions is "the fact that, in general, juries are unduly receptive to identification evidence and are not sufficiently aware of its dangers.” Patrick M. Wall, EyeWitness Identification in Criminal Cases 19 (1965). Many jurists agree that eyewitness identifications are the most devastating and persuasive evidence in criminal trials. See, e.g., Watkins v. Sowders, 449 U.S. 341, 352, 101 S.Ct. 654, 66 L.Ed.2d 549 (1981) (stating that "[tjhere is almost nothing more convincing than a live human being who takes the stand, points a finger at the defendant, and says ‘That’s the one!’ ”) (Brennan, J., dissenting) (citations omitted); Manson v. Brathwaite, 432 U.S. 98, 120, 97 S.Ct. 2243, 53 L.Ed.2d 140 (1977) (stating that “juries unfortunately are often unduly receptive to [identification] evidence”) (Marshall, J., dissenting); Hon. D. Duff McKee, Challenge to Eyewitness Identification Through Expert Testimony, 35 Am. Jur POF 3d 1, § 1 (1996) ("Eyewitness testimony may be the least reliable, and yet the most compelling.”). Jurors tend, to overestimate the accuracy of eyewitness identifications because they often do not know the factors they should consider when analyzing this testimony. See Handberg, supra, at 1022.
. The Dissent finds that the Supreme Court’s Daubert decision is:
[n]ot "holy writ” that the district court must invoke by name in order to pass our scrutiny.
Instead, the Dissent suggests that the district court should have instead relied on a pre-Daubert Third Circuit precedent, United States v. Stevens, 935 F.2d 1380 (3d Cir.1991), and United States v. Downing, 753 F.2d 1224 (3d Cir.1985), as the standard for outlining the steps that Smithers should have followed in making his proffer to the Court. Apparently, the Third Circuit has provided what the Dissent characterizes as “holy writ,” notwithstanding the fact that Steves and Downing are pre-Daubert authority and that the proffer of testimony that these Third Circuit cases require does not meet Daubert's standard for determining whether scientific evidence is admissible.
. The government argues additionally that Smithers’s proffer demonstrates that the expert testimony would have invaded the jury's province. Specifically, the government points to the sentence in the proffer which states, "Had Mr. Smithers been the robber, the eyewitnesses would have observed and been able to recall the large scar on Mr. Smithers's neck.” We agree with the government that this was poorly chosen wording, and that no expert may testify as to what witness did or did not see. In a case heavily dependent upon eyewitness identification, such testimony could usurp the jury's function and produce an improper comment on the ultimate issue to be decided in the case. The district court, however, did not even mention, much less base its decision on the language in this sentence. Even if it had, the proper solution would have been to excise the inappropriate portion of the proffer rather than to exclude all of the testimony, the remainder of which dealt only with the psychological factors which may have impacted the perception and memory of the witnesses in this case. This evidence would have been both relevant and helpful to the jury.
. As one commentator has indicated:
[tjhere are some indications of a compromise position that would be more favorably inclined toward [eyewitness identification] testimony when specific factors of need arise. Where identification rests on testimony by someone who knew the defendant well and was in a good position to see the crime, or where the identification seems strongly established for other reasons (like physical evidence connecting defendant to the crime), there is little reason to admit such testimony. Where identity is a crucial and closely contested issue, however, and where critical testimony is given by people who did not know the perpetrator and had only a short time to see him or were limited or distracted by other factors, expert testimony seems more clearly warranted.
Christopher B. Mueller & Laird C. Kirkpatrick, Evidence, § 6.37, at 601 (1995).
.Presumably, the district court was trying to express that, the expert testimony would be unduly prejudicial. This conclusion is flawed. First, as the Smith Court noted, "in reviewing a 403 balancing [in a criminal case], the court must look at the evidence in the light most favorable to the proponent, maximizing its probative value and minimizing its prejudicial effect.” 736 F.2d at 1107. The district court did not apply this standard here. . Second, it appears the trial court thought the expert nature of the testimony would unduly impress the jury; this is an improper factor upon which to exclude expert *318testimony, for if this were the test, no expert could ever testify. The court erred in concluding that merely because testimony is given by an expert, it must be excluded.
. Smithers also appealed his conviction on two other grounds: (1) the district court’s exclusion of a portion of the testimony of Smithers’s wife on relevancy grounds, and (2) the district court’s response to questions posed by the jury after it began deliberating. Because we have remanded this case for a new trial based on the district court’s failure to conduct a Daubert test before excluding the eyewitness expert’s testimony, these additional issues are moot.