concurring, with separate mandate.
In Howerton v. Arai Helmet, Ltd., this Court examined and explained the standard for “ruling on the admissibility of expert testimony” in North Carolina. 358 N.C. 440, 455, 597 S.E.2d 674, 684 (2004). We acknowledged, on the one hand, that “trial courts must decide preliminary questions concerning the qualifications of experts to testify or the admissibility of expert testimony,” and we reaffirmed that such decisions will generally be reviewed on appeal for abuse of discretion. Id. at 458, 597 S.E.2d at 686. We emphasized, on the other hand, that the trial court’s preliminary assessment should not “go so far as to require the expert’s testimony to be proven conclusively reliable or indisputably valid before it can be admitted into evidence.” Id. *150at 460, 597 S.E.2d at 687. Evidence may be “ ‘shaky but admissible,’ ” and it is the role of the jury to make any final determination regarding the weight to be afforded to the evidence. Id. at 460-61, 597 S.E.2d at 687-88 (quoting Daubert v. Merrell Dow Pharm., Inc., 509 U.S. 579, 596 (1993)).
This Court took great care in Howerton to distinguish our approach to expert qualification and admissibility of expert testimony from the federal court procedures described in Daubert v. Merrell Dow Pharmaceuticals, Inc., 509 U.S. 579 (1993). Howerton, 358 N.C. at 469, 597 S.E.2d at 692-93. We stated that “application of the North Carolina approach is decidedly less mechanistic and rigorous than the ‘exacting standards of reliability’ demanded by the federal approach.” Id. at 464, 597 S.E.2d at 690 (quoting Weisgram v. Marley Co., 528 U.S. 440, 455 (2000)). Our concern was that “trial courts asserting sweeping pre-trial ‘gatekeeping’ authority under Daubert may unnecessarily encroach upon the constitutionally-mandated function of the jury to decide issues of fact and to assess the weight of the evidence.” Id. at 468, 597 S.E.2d at 692.
In the context of medical malpractice cases, our General Assembly has expressed a similar sentiment regarding the jury’s function in weighing expert testimony. See N.C.G.S. § 90-21.12 (2007). Assuming expert testimony is properly qualified and placed before the trier of fact, section 90-21.12 reserves a role for the jury in determining whether an expert is sufficiently familiar with the prevailing standard of medical care in the community. See id. Under the statute, “the trier of the facts” must be “satisfied by the greater weight of the evidence that the care of [the] health care provider was not in accordance with the standards of practice among members of the same health care profession with similar training and experience situated in the same or similar communities.” Id.
In the instant case, the record before this Court appears to present a close question as to whether plaintiffs’ proffered expert, Dr. Elliott, was sufficiently familiar with the standard of care in Goldsboro. Dr. Elliott’s deposition testimony tended not to support the admission of his testimony at trial. For instance, he did not know the designation of Wayne Memorial Hospital (in which plaintiffs’ daughter was born) or the number of labor and delivery suites it had. He demonstrated little familiarity with Goldsboro or Wayne County beyond a basic estimate of population and general location within the state. He testified that most of his obstetrics career was spent in Phoenix, a metro area he believed had more than twenty times the *151number of obstetricians than Goldsboro and a population exceeding that of Goldsboro by over four million people. Dr. Elliott himself had never performed the Zavanelli maneuver, nor had he ever observed it performed during his twenty-four years of practice in Phoenix. Moreover, at several points during his deposition, he appeared to be applying a national standard of care rather than the “same or similar community” standard required by our General Assembly in section 90-21.12. See § 90-21.12.
Dr. Elliott’s affidavit, on the other hand, indicated that he had researched and was knowledgeable about the standard of care in Goldsboro. For example, Dr. Elliott stated that after reviewing various materials, he was familiar with “the training, education and experience of Dr. Peter Roethling,” “the size of the population [of Goldsboro], the level of care available at the hospital, the facilities and the number of health care providers for obstetrics,” and “the prevailing standard of care for handling shoulder dystocia in the same or similar community to Goldsboro.”
Our statutes and case law do not require an expert to have actually practiced in the community in which the alleged malpractice occurred, or even to have practiced in a similar community. See § 90-21.12; see also N.C.G.S. § 8C-1, Rule 702(b) (2007) (indicating that an expert in a medical malpractice case need not be licensed in North Carolina so long as the expert is licensed in some other state). In this regard, I agree with Justice Hudson’s opinion that our law does not “prescribe any particular method by which a medical doctor must become ‘familiar’ with a given community.” Book or Internet research may be a perfectly acceptable method of educating oneself regarding the standard of medical care applicable in a particular community. See, e.g., Coffman v. Roberson, 153 N.C. App. 618, 624, 571 S.E.2d 255, 259 (2002) (holding medical expert demonstrated sufficient familiarity with applicable standard of care when that familiarity was gained in part from “Internet research about the size of the hospital, the training program, and the AHEC (Area Health Education Center) program”), disc. rev. denied, 356 N.C. 668, 577 S.E.2d 111 (2003).
Although the trial court appropriately considered, both Dr. Elliott’s deposition testimony and his affidavit in determining whether to admit his expert opinion at trial, these discovery materials did not adequately convey a complete picture of Dr. Elliott’s qualifications or the reliability of his proposed testimony. Defend*152ants’ deposition of plaintiffs’ proposed expert suggested a lack of relevant knowledge about Goldsboro, while the expert’s affidavit asserted his familiarity without explaining what materials he reviewed or the way in which those materials influenced his determination of the applicable standard of medical care. Moreover, the trial court based its decision to exclude Dr. Elliott primarily on a paper record, considering the video deposition transcript, the affidavit, and brief oral argument by counsel. Thus, the trial court was in no better position than this Court to review the record and to assess Dr. Elliott’s qualifications and the reliability of his proposed testimony. See In re Greene, 306 N.C. 376, 380, 297 S.E.2d 379, 382 (1982) (explaining that “[t]his Court, unlike a trial court, is ill-equipped to resolve disputed questions of fact” because we “do not hear live testimony of sworn witnesses and are required to rely exclusively upon written records”).
When the proffered expert’s familiarity with the relevant standard of care is unclear from the paper record, our trial courts should consider requiring the production of the expert for purposes of voir dire examination. In such situations, particularly when the admissibility decision may be outcome-determinative, the expense of voir dire examination and its possible inconvenience to the parties and the expert are justified in order to ensure a fair and just adjudication. Voir dire examination provides the trial court with the opportunity to explore the foundation of the expert’s familiarity with the community, the method by which the expert arrived at his conclusion regarding the applicable standard of care, and the link between this method and the expert’s ultimate opinion. Moreover, unlike the non-adversarial discovery process, counsel for both parties may participate equally in a voir dire hearing and help elicit all information relevant to the expert’s qualifications and the admissibility of the proposed testimony.
Perhaps most importantly, voir dire examination provides the trial court with an informed basis to guide the exercise of its discretion. It is precisely because the trial court “ ‘has the advantage of seeing and hearing the witnesses’ ” that the trial court’s discretionary decision is entitled to deference on appeal. State v. Lasiter, 361 N.C. 299, 305, 643 S.E.2d 909, 912 (2007) (quoting State v. Little, 270 N.C. 234, 240, 154 S.E.2d 61, 66 (1967)) (explaining further that the trial court’s firsthand observations of jury voir dire enable it to “ ‘gain a “feel” of the case which a cold record denies to a reviewing court’ ” (quoting Little, 270 N.C. at 240, 154 S.E.2d at 66)).
*153I do not suggest that voir dire examination is necessary in every case in which opposing counsel challenges a proffered expert’s qualifications or proposed testimony. In light of the emphasis Howerton places on the jury’s role in evaluating expert testimony, however, voir dire examination may be prudent in close cases. In Howerton, this Court expressed concern with “the case-dispositive nature of Daubert proceedings, whereby parties in civil actions may use pretrial motions to exclude expert testimony under Daubert to bootstrap motions for summary judgment that otherwise would not likely succeed.” Howerton, 358 N.C. at 467, 597 S.E.2d at 691 (stating further: “[A] party may use a [pre-trial] hearing to exclude an opponent’s expert testimony on an essential element of the cause of action. With no other means of proving that element of the claim, the non-moving party would inevitably perish in the ensuing motion for summary judgment.” Id. at 468, 597 S.E.2d at 692.).
The same concern is implicated in the instant case, in which defendants sought and received summary judgment immediately after the trial court’s exclusion of plaintiffs’ tendered expert. At the end of counsels’ arguments, following discussion about Dr. Elliott’s deposition testimony and affidavit, plaintiffs’ counsel noted to the trial court that “[t]his is not the cross-examination of Dr. Elliott at a voir dire [examination].” As counsel’s remark implies, here, and in similar cases, the voir dire procedure provides a more reliable assessment mechanism than discovery depositions or conclusory affidavits, protecting the jury from unreliable expert testimony yet preserving the jury’s role in weighing the credibility of expert testimony when appropriate.
For the foregoing reasons, this case is reversed and remanded to the Court of Appeals for further remand to the trial court with instructions to conduct a voir dire examination of plaintiffs’ proffered expert and, based on this evidentiary foundation, to determine the admissibility of the proposed expert testimony. See Marks v. United States, 430 U.S. 188 (1977).
Justice EDMUNDS concurs in this opinion.