dissenting.
Although I join in Parts I through V and VII through IX of the majority’s opinion, I believe that the majority has made a significant error regarding expert testimony, and I dissent on this point. Because I believe that this error is not harmless, I believe we should vacate the judgment and remand this case for a new trial.
In concluding that the District Court properly excluded the testimony of Dr. Gottheimer, the majority seriously misconceives the proper approach to the admission of expert opinion testimony under Federal Rule of Evidence 702. As an initial point, the majority overstates the degree to which we owe deference to the District Court’s decision in a case, such as this, tried to the court. The District Court’s decision deserves no additional deference simply because the court sat as both evidentiary gate-keeper and fact-finder. If anything, the practicalities of the matter suggest that the trial court in such a situation should be more reluctant than usual to exclude evidence, although I do not suggest that we should apply a correspondingly more stringent standard of review.
The District Court and the majority also make three important errors in analyzing the substantive requirements of Rule 702. First, the majority confuses the reliability of an expert witness — a matter for the jury — with the reliability of his or her methodology- — a matter initially for the trial judge — and therefore erroneously concludes that questions about an expert witness’s general credibility are a proper basis for excluding his or her testimony. The question for the judge under Rule 702 is not whether the witness is reliable but whether the methodology the expert uses in reaching his conclusions is reliable. As to this latter point, the witness’s general credibility is simply irrelevant; the relevant issues in determining the reliability of an expert’s principles and methods are of the sort set forth with respect to scientific testimony in Daubert v. Merrell Dow Pharms., Inc., 509 U.S. 579, 593-95, 113 S.Ct. 2786, 125 L.Ed.2d 469 (1993).
Second, the majority misconstrues the nature of the requirement of “fit” between the expert’s testimony and the facts at issue. The majority concludes that the District Court correctly found that no fit exists in this case because Dr. Gottheimer’s experience, in the area of property-casualty insurance, on which his testimony is based, is not in the specific area — life insurance — with which the facts in this case deal. But the majority’s focus on the connection between Dr. Gottheimer’s claimed basis for being an expert and the facts at issue, as opposed to the connection between the substance of his testimony and the facts, is irrelevant to the question of fit. Their concern about his background should more appropriately be directed at Dr. Gottheimer’s qualifications, not the fit between his testimony and the facts.
Third, the majority permits the District Court to set the qualifications bar for expert testimony too high. It approves the District Court’s rejection of the expert’s testimony simply because his qualifications are not of the “highest caliber.” This conclusion is inconsistent with our longstanding liberal approach to the matter of expert witness qualifications. Also, the connection between Dr. Gottheimer’s expertise and the issues in this case— which the majority discusses in the context of fit — are not too remote for him to qualify as an expert under Rule 702.
Finally, I think that the majority wrongly concludes that any error in the exclusion *162of Dr. Gottheimer’s testimony was harmless. While improper admission of evidence is usually harmless error in a bench trial, the improper exclusion of an expert witness who would have offered a party’s sole expert testimony on an element of its case ordinarily is not harmless. The fact that the District Court found some inconsistencies in Dr. Gottheimer’s voir dire testimony was not a sufficient basis for changing the ordinary rule. Such inconsistencies are not enough to convince me that it is highly probable that Dr. Got-theimer’s testimony, if admitted, would not have changed the outcome. This is especially true in light of the fact that the inconsistencies the District Court discusses were at worst minor.
I. Overview of Rule 702
Under Rule 702, “[i]f scientific, technical, or other specialized knowledge will assist the trier of fact to understand the evidence or to determine a fact in issue, a witness qualified as an expert by knowledge, skill, experience, training, or education, may testify thereto in the form of an opinion or otherwise.” Fed.R.Evid. 702. The Supreme Court has recognized that Rule 702, although it limits the scope of permissible evidence, is part of “the ‘liberal thrust’ of the Federal Rules and their ‘general approach of relaxing the traditional barriers to “opinion” testimony.’ ” Daubert, 509 U.S. at 588, 113 S.Ct. 2786 (quoting Beech Aircraft Corp. v. Rainey, 488 U.S. 153, 169, 109 S.Ct. 439, 102 L.Ed.2d 445 (1988)). We have recognized that this “principle of liberal admission of expert testimony is found in Rule 702 itself, in the advisory committee note to the rule, and in our case law.”1
To these ends, Rule 702 embodies three distinct substantive restrictions on the admission of expert testimony: qualifications, reliability and fit. See In re Paoli R.R. Yard PCB Litig. (“Paoli II”), 35 F.3d 717, 741-43 (3d Cir.1994). First, an expert witness must be qualified by virtue of specialized expertise. See Fed.R.Evid. 702 (permitting expert testimony of a witness “qualified as an expert by knowledge, skill, experience, training, or education”); Paoli II, 35 F.3d at 741. Second, “an expert’s testimony is admissible so long as the process or technique the expert used in formulating the opinion is reliable.” Paoli II, 35 F.3d at 742 (citing Daubert, 509 U.S. at 589-90, 113 S.Ct. 2786). Third, “[ejxpert testimony which does not relate to any issue in the case is not relevant and, ergo, non-helpful.” Daubert, 509 U.S. at 591, 113 S.Ct. 2786 (citations omitted).2
*163II. Standard of Review
As an initial matter, I think the majority errs in its analysis of the applicable standard of review and the extent to which we owe deference to the District Court’s decision. Of course, the decision whether to admit or exclude expert testimony is largely within the hands of the trial judge. We review such a decision for abuse of discretion. See General Elec. Co. v. Joiner, 522 U.S. 136, 118 S.Ct. 512, 517, 139 L.Ed.2d 508 (1997) (“We have held that abuse of discretion is the proper standard of review of a district court’s evidentiary rulings.” (citations omitted)); In re Paoli R.R. Yard PCB Litig. (‘Paoli I”), 916 F.2d 829, 856 & n. 33 (3d Cir.1990). The standard does not change when we are reviewing a comb’s decision to exclude, as opposed to admit, expert testimony, in spite of the liberal standard for the admission of such testimony. See Joiner, 118 S.Ct. at 517.
Even though we apply an abuse of discretion standard of review, however, “to the extent the district court’s ruling turns on an interpretation of a Federal Rule of Evidence our review is plenary.”3 Furthermore, although our review is highly deferential, it is not a complete bar to reversing a district court’s .decision even where the court does not commit purely legal error. See, e.g., Paoli I, 916 F.2d at 855-56 (rejecting as an abuse of discretion the trial court’s insistence on certain credentials as expert qualifications); 4 Wein-stein’s Federal Evidence § 702.02(2) n. 9 (2d ed. Nov.1998) (collecting cases).
The majority concludes that the fact that this case involved a bench trial requires additional deference to the District Court’s evidentiary exclusion decision. In reaching this conclusion, the majority relies on Goodman v. Highlands Insurance Go., 607 F.2d 665 (5th Cir.1979), which states that “a trial judge sitting without a jury is entitled to even greater latitude concerning the admission or exclusion of evidence.” Goodman, 607 F.2d at 668. The majority asserts that, after Joiner — in which the Supreme Court held that the same abuse of discretion standard of review applies to an evidentiary ruling on expert testimony regardless of whether the trial court admits or excludes the evidence — -the “Goodman” rule “has been given increased viability.” Maj. Op. at 156. For a variety of reasons, I cannot agree with the majority’s conclusion that our review is affected by the fact that this case was tried to the court.
As an initial matter, Goodman does not support the meaning the majority draws from it. The statement in Goodman upon which the majority relies refers not to the trial court’s decision on admissibility but to whether the trial court’s decision, if erroneous, was a harmless error. This reading is made manifest by an examination of the sentence following that quoted above:
In a non-jury case, the admission of incompetent evidence will not warrant reversal unless all of the competent evidence is insufficient to support the judgment, or unless it affirmatively appears that the incompetent evidence induced the court to make an essential finding *164which would ■ otherwise not have been made.
Goodman, 607 F.2d at 668 (citations omitted). This is simply a restatement of the familiar harmless error test for review of decisions admitting evidence in bench trials.4 It is, however, irrelevant to our determination of whether the district court’s decision to exclude expert testimony from evidence is reversible error.5
Furthermore, even assuming the majority correctly states the Fifth Circuit’s Goodman rule, I think that rule is not a proper one and would not follow it. The Federal Rules of Evidence apply with full force to bench trials. See Fed.R.Evid. 1101(b); 9 Charles Alan Wright & Arthur R. Miller, Federal Practice & Procedure § 2411, at 587 (2d ed. 1995) (“In theory, the Federal Rules of Evidence apply equally in court trials and jury trials.”). After all, a trial is a trial. To me, the proposition that we owe more deference to trial court decisions excluding evidence in bench trials is inconsistent with the Federal Rules and encourages sloppy district court decisionmaking.6
If anything, trial courts should be more chary of excluding evidence in bench trials than in jury trials. See Builders Steel Co. v. Commissioner, 179 F.2d 377, 379 (8th Cir.1950) (“[A] trial judge who, in the trial of a nonjury case, attempts to make strict rulings on the admissibility of evidence, can easily get his decision reversed by excluding evidence which is objected to, but which, on review, the appellate court believes should have been admitted.”), quoted in 9 Wright & Miller, supra, § 2411, at 587. The better course is to admit the evidence and then take factors that otherwise might affect its admissibility into consideration in determining its weight, rather than waste time debating *165the propriety of admitting the evidence.7 I believe the majority’s approach grants undue deference to trial court’s decisions excluding evidence in bench trials.
III. Rule 702 Requirements
I now turn to the substantive requirements of Rule 702. Under Rule 702, expert testimony is admissible only if it meets the requirements of Rule 702: qualifications, reliability and fit. The majority concludes that the District Court properly found that Dr. Gottheimer’s testimony met none of these requirements. I disagree, and discuss each of these factors, although not in the usual order.
A. Reliability
The majority begins with the indisputable premise that reliability is a key factor in determining the admissibility of expert testimony. “[Ujnder the Rules the trial judge must ensure that any and all scientific testimony or evidence admitted is not only relevant, but reliable.” Joiner, 118 S.Ct. at 517 (quoting Daubert, 509 U.S. at 589, 113 S.Ct. 2786). From this, the majority concludes that, as part of the reliability analysis, a trial court may consider the general credibility of a witness in determining whether his or her testimony is appropriately admitted as expert witness testimony. See Maj. Op. at 155 (“Thus in our view, the Court’s emphasis on reliability as well as on relevancy embraces within its standard the credibility of the witness proffering expert opinion.”). Accordingly, the majority holds that, since the District Court found that Dr. Gottheimer was not a credible witness it could properly exclude his expert testimony under Rule 702. The majority misconceives the fundamental nature of the reliability inquiry under Dau-bert. The reliability inquiry focuses not on the witness’s reliability in an evidentiary sense, but on the reliability of the methodology that the expert applies in arriving at an opinion.
[I]n order to qualify as “scientific knowledge,” an inference or assertion must be derived by the scientific method. Proposed testimony must be supported by appropriate validation — i.e., “good grounds,” based on what is known. In short, the requirement that an expert’s testimony pertain to “scientific knowledge” establishes a standard of eviden-tiary reliability.
Daubert, 509 U.S. at 590, 113 S.Ct. 2786 (footnote omitted). This inquiry focuses on the expert’s principles and methodology, not his results.
The inquiry envisioned by Rule 702 is, we emphasize, a flexible one. Its overarching subject is the scientific validity — and thus the evidentiary relevance and reliability — of the principles that underlie the proposed submission. The focus, of course, must be solely on principles and methodology, not on the conclusions that they generate.
Daubert, 509 U.S. at 594-95, 113 S.Ct. 2786. Under Rule 702, the trial court’s preliminary reliability analysis in making an admissibility determination must focus on the 'witness’s methods, not his or her testimony as a whole.8 A current proposed amendment to Rule 702 reempha*166sizes this focus on the reliability of the methodology, as opposed to the witness.9
Credibility plays no appropriate part in the analysis of the reliability of a proposed expert’s methodology. The Court in Dau-bert recognized a number of factors pertinent to the reliability inquiry, including testability, peer review or publication, potential rate of error, existence of standards and controls, and general acceptance. See 509 U.S. at 593-94, 113 S.Ct. 2786. We, as well as others, have suggested numerous additional factors relevant to reliability as set forth in the margin.10 The key point is that none of these factors requires consideration of the proffered expert’s credibility in general. “[Evaluating the reliability of scientific methodologies and data does not generally involve assessing the truthfulness of the expert witnesses.... ” Paoli II, 35 F.3d at 749.11 Of course, a particular witness can lie about whether one of the factors mentioned above is present. But the witness’s general credibility — i.e., credibility based on matters not directly related to these factors, such as bias or unrelated prior inconsistent statements — is not relevant to a trial court’s preliminary determination that these factors are present. These factors are relatively objective matters that the court can generally analyze independent of the witness’s testimony about them. For example, a court is fully capable of determining whether a theory or method is testable.
*167Furthermore, permitting the trial court to consider the general credibility of an expert witness in its analysis of the reliability of proposed expert opinion testimony improperly trenches on the province of the fact-finder. The fact-finder is ordinarily the arbiter of general credibility. Rule 702 is intended not to impinge on the authority of the fact-finder in making credibility determinations, but rather to ensure that the fact-finders’ ability to find facts independently is not overwhelmed by complex and authoritative-seeming expert testimony. See 29 Wright & Gold, supra, § 6262, at 179 (“If the trier of fact is unable or disinclined to question the expert’s opinion, it surrenders its central function to an expert whose testimony may be unreliable.”). In order to avoid this eventuality, Rule 702 provides for a preliminary inquiry into whether proposed expert testimony is the kind that is appropriately admitted into evidence, i.e., that the method is reliable and based on expertise. See id. at 183-84. But this inquiry cannot extend into matters that are the proper province of the fact-finder, such as general credibility. See id. at 184 (“Importantly, however, Rule 702 on its face creates no general power in the trial judge to exclude expert testimony on the grounds that it is unreliable.”).
This limitation on the power of the trial court to admit and exclude expert testimony rests on sound policy considerations. First, it is emblematic of the Rules’ generally liberal approach to the admissibility of evidence, discussed above. See supra Part I. Second, it reflects the fact that the power to evaluate witness’s credibility lies traditionally with the fact-finder, not the trial court making evidentiary rulings. See 29 Wright & Gold, stipra, § 6262, at 185 (“[T]he power to evaluate the credibility of witnesses and give testimony its proper weight primarily resides with the trier of fact.”).
Finally, it rests on the general assumption underlying the Federal Rules of Evidence as a whole: the fact-finder is best situated to determine the credibility of witnesses. See id. at 185 (“[J]uries generally have the ability to accurately weight and evaluate witness credibility.”). “Accordingly, the most common judicial response to attacks on the reliability of expert testimony is that such matters go to weight, not admissibility.” Id. at 185-86. The Court in Daubert recognized the importance of leaving such matters to the fact-finder, and not determining them on evi-dentiary grounds:
Respondent expresses apprehension that abandonment of “general acceptance” as the exclusive requirement for admission will result in a “free-for-all” in which befuddled juries are confounded by absurd and irrational pseudoscientific assertions. In this regard respondent seems to us to be overly pessimistic about the capabilities of the jury and of the adversary system generally. Vigorous cross-examination, presentation of contrary evidence, and careful instruction on the burden of proof are the traditional and appropriate means of attacking shaky but admissible evidence.
Daubert, 509 U.S. at 595-96, 113 S.Ct. 2786 (citation omitted; emphasis added).
Accordingly, I conclude that the District Court erred to the extent that it considered Dr. Gottheimer’s general credibility as a factor in determining whether the principles and methodology underlying his proposed expert testimony were reliable under Rule 702. This result is unaffected by the fact that the District Court — post hoc — went ahead and made credibility findings that arguably would have resulted in his rejection of Dr. Gottheimer’s testimony in its role as fact-finder. I deal with this aspect of the matter in my discussion of harmless error. See infra Part IV. Questions about Dr. Gottheimer’s credibility should have been left for substantive fact-finding, not evidentiary rulings. Since the District Court and Unisys suggest no other concerns about the reliability re*168quirement, I would conclude that Dr. Got-theimer’s proposed testimony satisfies it.
B. Fit
The majority also concludes that the District Court properly found that Dr. Gottheimer’s testimony did not fit with the question in issue at the trial. In particular, the majority concludes that Dr. Got-theimer’s experience and knowledge lay in a field not sufficiently connected with the question at issue to meet the requirement of fit. “Dr. Gottheimer’s alleged expertise, limited in any event to methods of investing with respect to property casualty insurance, did not fit with or meet the need of the District Court for expert testimony in life insurance investing.” Slip Op. at 21. The majority misconceives the requirement of fit under Rule 702.
The requirement of fit is essentially a relevance requirement. Under Rule 702, expert testimony is admissible if it “will assist the trier of fact.” Fed.R.Evid. 702. “This condition goes primarily to relevance.” Daubert, 509 U.S. at 591, 113 S.Ct. 2786. “[Ajdmission depends upon the ‘fit,’ i.e., upon a specific proffer showing that scientific research has established” some point relevant to the facts of the case. United States v. Downing, 753 F.2d 1224, 1226 (3d Cir.1985). We have further clarified this point:
An additional consideration under Rule 702 — and another aspect of relevancy' — -is whether expert testimony proffered in the case is sufficiently tied to the facts of the case that it will aid the jury in resolving a factual dispute. In this regard, we hold that a defendant who seeks the admission of expert testimony must make an on-the-record detailed proffer to the court, including an explanation of precisely how the expert’s testimony is relevant to the [factual issue] under consideration. The offer of proof should establish the presence of factors ... which have been found by researchers to [provide a basis for the proffered opinion]. Failure to make such a detailed proffer is sufficient grounds to exclude the expert’s testimony.
753 F.2d at 1242 (citations omitted); see also Lauria v. National R.R. Passenger Corp., 145 F.3d 593, 600 (3d Cir.1998); United States v. Velasquez, 64 F.3d 844, 850 (3d Cir.1995).
The majority errs in concluding that Dr. Gottheimer’s field of expertise has any relevance to the fit inquiry. As the foregoing discussion of the fit requirement shows, it is satisfied if the proffered expert opinion is relevant to a factual issue before the fact-finder. The expert’s field of expertise is irrelevant to an inquiry into the connection between the opinion itself and the issues in the case. A simple example shows this. In Lauria, a railroad worker slipped and was injured when he stepped on a loose railroad tie that was sitting between some tracks. “The primary issue [was] whether Amtrak was negligent in failing to remove a [railroad] tie from Lauria’s workplace.” Lauria, 145 F.3d at 600. An expert opinion “that the tie was a dangerous obstruction that should have been discovered and removed and, ... that Amtrak’s negligence made the workplace unsafe” clearly satisfies the requirement of fit. 145 F.3d at 600.
But the expertise of the proponent of the opinion is irrelevant to this inquiry. This is true even if the witness proposing to testify to the above opinion is a medical doctor who has no experience with train tracks. Under the majority’s reasoning, however, the doctor’s testimony would not meet the fit requirement. Clearly, a medical doctor with no experience working with train tracks has no expertise with regard to the placement of loose railroad ties. If Dr. Gottheimer’s testimony would not meet the fit requirement because his expertise lay in a different field of insurance than that in issue in the case, clearly the hypothetical doctor cannot provide testimony that meets the fit requirement. But the doctor’s testimony should be excluded because of his lack of qualifications, not *169because of a supposed lack of fit. The proposed expert’s expertise is simply irrelevant to determining whether the proffered expert opinion is relevant to issues in the case.
As this example shows, the majority’s and the District Court’s concerns about the connection between Dr. Gottheimer’s expertise and the issues in the case are actually relevant to the step of the Rule 702 inquiry which I consider infra: qualifications. Under Rule 702, a witness can offer an expert opinion if he or she is “qualified as an expert by knowledge, skill, experience, training, or education.” Fed. R.Evid. 702. The nature of a witness’s specific field of expertise is part of the expert’s background that is considered in determining whether a witness is qualified. See Paoli II, 35 F.3d at 741.12
I think a proper fit inquiry would show that Dr. Gottheimer’s testimony meets the fit requirement. Plaintiffs’ counsel said that Dr. Gottheimer would have testified that certain standard tests exist in the insurance industry to analyze the financial condition of insurance companies, and that he was familiar with those tests. He would have further testified that he performed these tests on Executive Life data, and concluded “that the tests set up certain red flags that should have caused a person familiar with the tests and performing the tests, to ask further questions about the solvency and the credit worthiness of Executive Life.” I think this is the kind of “on-the-record detailed proffer” that we required in Downing, “including an explanation of precisely how the expert’s testimony is relevant to” determining whether Unisys acted prudently. Downing, 753 F.2d at 1242. Accordingly, I would conclude that Dr. Gottheimer’s proposed testimony meets the fit requirement. The nature of Dr. Gottheimer’s expertise, and its connection with the factual issues in this case, are relevant only to the Rule 702 inquiry into qualifications, to which I now turn.
C. Qualifications
The majority also concludes that the District Court properly excluded Dr. Got-theimer’s testimony because he was not qualified as an expert with respect to the issues in this case. Specifically, the majority adopts the District Court’s conclusion that Dr. Gottheimer was not qualified because his “qualifications were less than stellar.” Slip Op. at 21. The District Court noted that “Dr. Gottheimer claims a doctoral degree from a correspondence school, an additional ground for my refusal to qualify him as an expert.” In re Unisys Sav. Plan Litig., No. 91-3067, 1997 WL 732473, at *26 (E.D.Pa. Nov. 24, 1997). In addition, as discussed above, see supra section III.B, the majority concludes that the District Court properly rejected Dr. Gottheimer’s testimony because his field of expertise was not sufficiently connected to the issues in the case. “Dr. Gottheimer’s experience in the insurance area was limited to property casualty insurance and not life insurance.” Unisys, 1997 WL 732473, at *22. Neither of these are appropriate grounds for excluding an expert witness for lack of qualifications.
The requirement that an expert witness be qualified is well established. A witness may only provide expert testimony to the extent that he or she is “qualified as an expert by knowledge, skill, experience, training, or education.” Fed.R.Evid. 702. *170The bases for concluding that a witness is qualified are flexible. “[IJnsistence on a certain kind of degree or background is inconsistent with our jurisprudence in this area.” Paoli I, 916 F.2d at 855; accord Waldorf v. Shuta, 142 F.3d 601, 626 (3d Cir.1998) (“[I]n considering the qualification of witnesses as experts, we stress that ordinarily an otherwise qualified witness is not disqualified merely because of a lack of academic framing.”); Paoli II, 35 F.3d at 741 (“We have eschewed imposing overly rigorous requirements of expertise and have been satisfied with more generalized qualifications.”). A particular educational background is unnecessary; practical experience is sufficient to conclude that a witness is qualified as an expert. See Lauria, 145 F.3d at 599 (citing American Tech. Resources v. United States, 893 F.2d 651, 656 (3d Cir.1990)). “Following this logic, it is an abuse of discretion to exclude testimony simply because the trial court does not deem the proposed expert to be the best qualified or because the proposed expert does not have the specialization that the court considers most appropriate.”13
Although, as always with evidentiary questions, we apply a deferential standard of review to a trial court’s determination of whether a proposed expert is qualified, we have on numerous occasions found that a district court abused its discretion in excluding a proffered expert because of his or her qualifications. Some of these cases are summarized in the margin.14
Supported by these cases, I believe that Dr. Gottheimer was in fact qualified to offer the expert testimony proffered, and that the District Court’s conclusion to the contrary was an abuse of discretion. The District Court found that Dr. Gottheimer was not qualified on two grounds: the nature of his educational credentials and the distinction between his experience in property-casualty insurance companies and the issues in the case involving life insurance companies. In light of the case law discussed in the margin above, both of *171these findings are inconsistent with the exercise of sound discretion.
The District Court concluded that Dr. Gottheimer could not be qualified because his doctorate was awarded by a correspondence school. If this were plaintiffs’ sole basis for claiming that Dr. Gottheimer was qualified, I would probably agree with the District Court. It was not, however. Rather, the record discloses numerous grounds on which to conclude that Dr. Gottheimer was qualified as an expert. Dr. Gottheimer’s resume demonstrates his expertise through three distinct areas: experience, education and teaching. First, he has worked for a dozen years as a consultant in the insurance industry, following thirty years of employment by various insurance companies. His consulting work has included analyses of both property-casualty and life insurance companies.15 Second, he has bachelor’s and master’s degrees in insurance-related fields, as well as a doctorate from a correspondence school.16 He also possesses several professional affiliations in insurance professionals’ organizations. Finally, Dr. Gottheimer has taught for over twenty-five years at the College of Insurance. The College of Insurance is an accredited, industry-sponsored school that offers classes in all aspects of insurance business. He is now on the fulltime faculty there, and has taught courses in a variety of fields, including insurance company management.
In light of these extensive qualifications, I have no doubt that Dr. Gottheimer was qualified and should have been permitted to testify as an expert under Rule 702. The District Court’s decision to the contrary was an abuse of discretion. As the majority points out, the District Court refused to qualify Dr. Gottheimer because his qualifications “were not of the highest caliber.” Maj. Op. at 156. But in light of our longstanding jurisprudence, this is not an appropriate basis for excluding a proffered expert witness. See Kannankeril v. Terminix Intl., Inc., 128 F.3d 802, 809 (3d Cir.1997); Holbrook, 80 F.3d at 782; Paoli II, 35 F.3d at 741. Accordingly, I must conclude that the District Court abused its discretion in excluding his testimony on this basis.
For all these reasons, I think that Dr. Gottheimer should have been permitted to testify as an expert witness under Rule 702. As discussed above, Dr. Gottheimer fully met all three of the Rule 702 requirements: qualifications, reliability and fit. The only remaining question is whether that error was reversible or harmless error.
*172IY. Harmless Error
In the closing paragraph of its discussion of the Rule 702 evidentiary issue, the majority concludes that, even if the District Court did commit an error in excluding Dr. Gottheimer’s testimony, the error was harmless. In reaching this conclusion, it relies on the District Court’s statement that it “could not find [Dr. Gottheimer] to be a credible witness given his evasiveness if not his propensity to state falsehoods.” 1997 WL 732473, at * 26. The District Court pointed to a few alleged inconsistencies in Dr. Gottheimer’s deposition and trial testimony in support of this conclusion. The majority concludes that, since the District Court would not have believed Dr. Gottheimer’s testimony, his testimony could not have been given any weight if admitted. Thus its exclusion did not have a substantial effect on the outcome and any error in excluding it was harmless. Once again, I must disagree.
Under the Federal Rules of Evidence, an evidentiary error to which a party has raised a proper objection is not a grounds for reversal “unless a substantial right of the party is affected.” Fed.R.Evid. 103(a); see also 28 U.S.C. § 2111; Fed.R.Civ.P. 61. An error is harmless, i.e., it does not affect a substantial right, only if “it is highly probable that the error did not contribute to the judgment.” Murray v. United of Omaha Life Ins. Co., 145 F.3d 143, 156 (3d Cir.1998) (citing McQueeney v. Wilmington Trust Co., 779 F.2d 916, 923-27 (3d Cir.1985)). Although, as discussed above, the improper admission of evidence is usually harmless error in a bench trial, see supra Part II, the erroneous exclusion of evidence in a bench trial can be reversible error just as in a jury trial. See 11 Charles Alan Wright, Arthur R. Miller & Mary Kay Kane, Federal Practice & Procedure § 2885, at 454 (2d ed. 1995) (“In nonjury cases the district court can commit reversible error by excluding evidence but it is almost impossible for it to do so by admitting evidence.” (footnote omitted)). Error is especially likely not to be harmless where the excluded expert was the only one a party offered to prove an essential element of its case.17
In this case, Dr. Gottheimer was the only expert witness plaintiffs offered to prove that Unisys acted imprudently. His proffered testimony, set forth in the margin, was strong.18 By excluding Dr. Got-theimer’s testimony, the District Court deprived plaintiffs of their best evidence that Unisys breached its duty of prudence, a key element of their case. In light of our conclusion in Lauria and Holbrook, I can*173not say that it is highly probable that the exclusion of Dr. Gottheimer did not affect the outcome of the trial. This is especially true in light of the other evidence admitted at trial, set forth in the margin.19 The District Court essentially decided all questions the evidence raised in favor of Unisys and concluded that Unisys acted prudently. Although I agree with the majority that, based on the admitted evidence, this conclusion was not clearly erroneous, this is to me an exceedingly close question. Accordingly, Dr. Gottheimer’s testimony, if admitted, stood a good chance of changing this balance and consequently changing the decision of the District Court.
That the District Court had questions about Dr. Gottheimer’s credibility should not affect our harmless error analysis. The District Court’s conclusion that it would not have found Dr. Gottheimer’s testimony credible, based only on his voir dire testimony, is not sufficient grounds for concluding that exclusion of his testimony was harmless. I believe that there is a reasonable chance that, if the District Court had given Dr. Gottheimer the opportunity to present his testimony in full, it would have found him to be a credible witness.
Our decisions in Lauria and Holbrook at least implicitly support this conclusion. In each of those cases, we reversed a district court decision excluding expert testimony on the grounds that the expert witness lacked the necessary qualifications. Such evidence goes to the weight, not the admissibility, of the expert testimony. See Kan-nankeril, 128 F.3d at 809. In neither Lauria nor Holbrook did the effect of the expert’s particular qualifications on the weight properly accorded to his testimony play any part in our harmless error analysis. See Lauria, 145 F.3d at 600 (discussing harmless error without mentioning the quality of the improperly excluded expert’s qualifications); Holbrook, 80 F.3d at 787 (same). Similarly, the impact of questions about Dr. Gottheimer’s credibility on the weight due his testimony should not play a part in our harmless error analysis in this case.
I find further support for this conclusion in an examination of what the District Court identified as “impeachment” of Dr. Gottheimer. The District Court stated that “Dr. Gottheimer was impeached no fewer than four times on the relatively straight forward questions on his qualifications.” 1997 WL 732473, at *26; see also 1997 WL 732473, at *21-*22. But this so-called “impeachment” involved at most minor inconsistencies. First, Dr. Gottheimer testified at trial that he could not recall having testified in court in a case involving a life insurance company, although he stated at his deposition three years earlier that twenty-four out of the twenty-five times he had testified in court involved property-casualty insurance companies. See 1997 WL 732473, at *21. Second, although at his deposition he could not recall having done consulting work involving life insurance company solvency, he testified that he had been retained once before his deposition and several times after his deposition to do such consulting. See 1997 WL 732473, at *21-*22. Third, *174while he testified at trial that there were “some differences and there are also some similarities in the way” one analyzes life insurance as opposed to property-casualty insurance company solvency, in his deposition he agreed with Unisys’s attorney that there were “fundamental differences.” See 1997 WL 732473, at *22.
I cannot see how these answers, under any reasonable reading, suggest that Dr. Gottheimer is a completely incredible witness. At most, they suggest that he could remember some things at his deposition that he could not at trial, and vice versa. Certainly, Dr. Gottheimer’s statements at trial were not directly inconsistent with those in his deposition. A comparison of the statements does not raise an inference that Dr. Gottheimer was being evasive, let alone lying. Thus, I cannot agree with the majority that the at worst minor inconsistencies in Dr. Gottheimer’s testimony make it highly probable that his testimony would not have affected the District Court’s judgment.
Accordingly, I do not think that the District Court’s erroneous exclusion of Dr. Gottheimer’s expert witness testimony was harmless. Therefore, I would remand this case for a new trial in which Dr. Gottheimer’s testimony could be presented; hence, I dissent.
Present: BECKER, Chief Judge, SLOVITER, STAPLETON, MANSMANN, GREENBERG, SCIRICA, NYGAARD, ALITO, ROTH, LEWIS, McKEE, RENDELL, WEIS,* and GARTH,* Circuit Judges.. Habecker v. Copperloy Corp., 893 F.2d 49, 51 (3d Cir.1990); accord Waldorf v. Shuta, 142 F.3d 601, 625 (3d Cir.1998) (citing the “policy of liberal admissibility of expert testimony”); Holbrook v. Lykes Bros. S.S. Co., 80 F.3d 111, 782 (3d Cir.1996) (citing "our liberal approach to admitting expert testimony”); United States v. Downing, 753 F.2d 1224, 1230 (3d Cir.1985) (citing “the liberal standard of admissibility mandated by Rule 702”); Knight v. Otis Elevator Co., 596 F.2d 84, 88 (3d Cir.1979).
. It is not a settled question in this Circuit whether the Daubert requirements apply to nonscientific testimony such as may be at issue here. See Lauria v. National R.R. Passenger Corp., 145 F.3d 593, 599 n. 7 (3d Cir.1998) (questioning, but not resolving, whether Daubert analysis should apply to testimony of expert in train track maintenance); United States v. Velasquez, 64 F.3d 844, 850 (3d Cir.1995) (questioning the propriety of applying Daubert to handwriting analysis, but applying it in an exercise of caution). This question is currently before the Supreme Court. See Carmichael v. Samyang Tire, Inc., 131 F.3d 1433 (11th Cir.1997), cert, granted sub nom. Kumho Tire Co. v. Carmichael,U.S. -, 118 S.Ct. 2339, 141 L.Ed.2d 711 (1998). The current proposed amendment to Rule 702 would apply a distillation of the Daubert analysis to all expert testimony. See Fed.R.Evid. 702 (proposed amendment 1998) (further limiting admissibility of expert testimony to where “(1) the testimony is sufficiently based upon reliable facts or data, (2) the testimony is the product of reliable principles and methods, and (3) the witness has applied the principles and methods reliably to the facts”).
The majority contends that I have erred by focusing on Daubert analysis in a case in which it might not apply. See Maj. Op. at 157. But even to the extent the majority *163turns out to be correct that a strict Daubert analysis does not apply, my conclusions would not change. First, I do not believe that the result would be any different if we were to apply the principles of Rule 702 sans Daubert. In fact, since Daubert imposes additional requirements for scientific testimony beyond the usual requirements for expert testimony, not applying Daubert would provide even greater reason to believe that Dr. Gottheimer's testimony should have been admitted. Thus, the majority’s criticism supports my contention. Second, even if the Supreme Court in Kumho decides that strict Daubert analysis should not apply to non-scientific testimony, I still think that the basic principles of reliability and fit would be relevant in determining the admissibility of expert testimony under Rule 702.
. DeLuca v. Merrell Dow Pharms., Inc., 911 F.2d 941, 944 (3d Cir.1990) (citing In re Japanese Elec. Prods. Antitrust Litig., 723 F.2d 238, 277 & n. 43 (3d Cir.1983), revd. on other grounds sub. nom. Matsushita Elec. Indus. Co. v. Zenith Radio Corp., 475 U.S. 574, 106 S.Ct. 1348, 89 L.Ed.2d 538 (1986)); accord Barker v. Deere & Co., 60 F.3d 158, 161 (3d Cir.1995).
. As we have stated:
[I]t is well settled that in a nonjury case, an appellate court will not reverse on the basis of an erroneous admission of evidence unless (1) there is insufficient evidence other than the challenged evidence to support the district court's conclusion, or (2) the district court is induced by the challenged evidence to make an essential finding that it would not have made otherwise.
United States v. Local 560, Intl. Bhd. of Teamsters, 780 F.2d 267, 278 (3d Cir.1985) (alteration in original) (quoting De Laval Turbine, Inc. v. West India Indus., Inc., 502 F.2d 259, 263-64 (3d Cir.1974)); accord 12 Rya W. Zo-bel, Moore's Federal Practice § 61.06(2) (3d ed.1998). This is a sensible rule based on the assumption that judges are more capable of ignoring prejudicial or irrelevant evidence than juries. See 1 Weinstein's, supra, § 103.41(4)(a) (“At one end of the scale is the nonjury trial in which the judge is often assumed, even in a criminal case, to have disregarded inadmissible evidence in arriving at a decision.”).
. Recent cases from the same court reveal that the quoted portion of Goodman in fact refers to harmless error analysis. See, e.g., Southern Pac. Transp. Co. v. Chabert, 973 F.2d 441, 448 (5th Cir.1992) (quoting Goodman and citing it in support of the harmless error standard).
.The majority suggests that I have failed to recognize the critical fact in this case: that the judge was the fact-finder as well as the Rule 702 gatekeeper. The majority is incorrect. Of course I recognize this fact, but think that it should make no difference in our analysis. The majority essentially contends that, once the trial judge in a bench trial makes up his or her mind during an in limine hearing that a witness is not credible, that decision is cast in concrete and the judge will close his or her ears to any further (trial) testimony from the witness. Concomitantly, the majority suggests that, with any witness, the court proceeding to a bench trial may exclude a prospective witness's testimony based not on its admissibility but on the witness’s credibility. See Maj. Op. at 157 (“We would be hard pressed to require a District Court judge sitting in a nonjury case who credibly and with reason found that he could not believe a witness to nevertheless hear the witness’s direct examination, cross-examination, and rebutlal examination in an extended trial when he knew that he would only reject it as unbelievable.”). I, to the contrary, think it would be preferable for the trial judge to listen to the witness and keep his or her mind open to the possibility that the entirety of the witness's trial testimony could change his or her view of the witness’s credibility. Listening, after all, is a major part of the judge's job.
. See Builders Steel, 179 F.2d at 379 (quoting Donnelly Garment Co. v. NLRB, 123 F.2d 215, 224 (8th Cir.1941)). In Donnelly Garment, the court noted that it is usually more efficient in a bench trial for the court to simply admit questionable evidence, and then take such questions into consideration in determining the weight it should be given. See 123 F.2d at 224.
. Even the broadest understandings of the Daubert reliability inquiry recognize that such reliability determinations are limited to the witness's methods and related matters. "The broadest reading of Daubert is that it applies to all reliability issues presented by all expert testimony. Under this interpretation, all reliability issues raised by an expert’s application, methodology, reasoning, or underlying theories are admissibility questions to be resolved by the gatekeeper-judge.” 29 Charles Alan Wright & Victor James Gold, Federal Practice & Procedure § 6266, at 290 (1997).
. The Judicial Conference of the United Stales's Standing Committee on Rules of Practice and Procedure is currently considering an amendment to Rule 702 proposed by its Advisory Committee on Rules of Evidence that would permit the admission of expert testimony if "(1) the testimony is sufficiently based upon reliable facts or data, (2) the testimony is the product of reliable principles and methods, and (3) the witness has applied the principles and methods reliably to the facts.” Fed.R.Evid. 702 (proposed amendment 1998). Under the proposed rule, the trial judge must determine whether the proposed testimony "is properly grounded, well-reasoned and not speculative before it can be admitted." Fed.R.Evid. 702 advisory committee's note (proposed amendment 1998). "If there is a well-accepted body of learning and experience in the expert's field, then the expert’s testimony must be grounded in that learning and experience to be reliable, and the expert must explain how tire conclusion is so grounded.” Id.
. See, e.g., Downing, 753 F.2d at 1238-39 (listing additional factors: novelty, specialized literature, and non-judicial uses of techniques); Joiner, 118 S.Ct. at 519 (whether the expert has unjustifiably extrapolated from an accepted premise to an unfounded conclusion); Sheehan v. Daily Racing Form, Inc., 104 F.3d 940, 942 (7th Cir.) (whether "the expert is being as careful as he would be in his regular professional work outside his paid litigation consulting”), cert. denied, 521 U.S. 1104, 117 S.Ct. 2480, 138 L.Ed.2d 989 (1997); Daubert v. Metrell Dow Pharms., Inc., 43 F.3d 1311, 1317 (9th Cir.1995) ("whether the experts are proposing to testify about mal-ters growing naturally and directly out of research they have conducted independent of the litigation, or whether they have developed their opinions expressly for purposes of testifying”); Claar v. Burlington N.R.R. Co., 29 F.3d 499, 502 (9th Cir.1994) (whether the expert has adequately accounted for obvious alternative explanations).
.Courts have held in numerous other cases that credibility is irrelevant to determining whether a proposed expert witness's testimony is admissible under Rule 702, and particularly whether it is based on reliable methodology. See, e.g., Breidor v. Sears, Roebuck & Co., 722 F.2d 1134, 1138-39 (3d Cir.1983) ("Where there is a logical basis for an expert’s opinion testimony, the credibility and weight of that testimony is to be determined by the jury, not the trial judge.”); see also Kannankeril v. Terminix Intl., Inc., 128 F.3d 802, 809 (3d Cir.1997) ("If the expert meets liberal minimum qualifications [under Rule 702], then the level of the expert’s expertise goes to credibility and weight, not admissibility.”). For example, expert witnesses cannot be excluded on the basis of bias. See, e.g., Marshall v. Perez Arzuaga, 828 F.2d 845, 851-52 (1st Cir.1987); Gideon v. Johns-Manville Sales Corp., 761 F.2d 1129, 1135-36 (5th Cir.1985). Similarly, factual errors in a witness’s testimony are not grounds for excluding the witness from testifying as an expert. See Paoli II, 35 F.3d at 753-54. Finally, general attacks on credibility based on a lack of personal knowledge are not a proper basis for excluding expert testimony. See Dixon v. International Hawester Co., 754 F.2d 573, 580 (5th Cir.1985).
. In fact, the case the majority cites in support of its conclusion that Dr. Gottheimer’s testimony does not meet the fit requirement involved an inquiry into the witness's qualifications, not the fit between his proposed testimony and the issues in that case. See Surace v. Caterpillar, Inc., 111 F.3d 1039 (3d Cir.1997). In Surace, the plaintiff offered the testimony of an electromechanical engineer concerning workers’ habituation to auditory warning devices. The district court excluded Brink’s testimony and we affirmed, noting that the expert's experience was limited to mechanical, as opposed to human, factors in design, and he therefore was not qualified to testify about the latter. See 111 F.3d at 1055-56.
. Holbrook, 80 F.3d at 782 (citing Paoli I, 916 F.2d at 856); accord Kannankeril, 128 F.3d at 809 (“Whether the appellants' expert might have done a better job is not the test.”); Paoli II, 35 F.3d at 741 (“[Ejxclusion was not the proper remedy ‘simply because the experts did not have the degree or training which the district court apparently thought would be most appropriate.’ ” (quoting Paoli I, 916 F.2d at 856)).
. See Lauria, 145 F.3d at 599 (“Slavin's twenty years of experience with track equipment, maintenance, and safety procedures qualified him as an expert who could testify as to Amtrak's responsibility to inspect and maintain the track in a safe condition,” even though he did not have particularized training other than that which anyone who had done such work for twenty years would have); Holbrook, 80 F.3d at 781-82 (reversing exclusion of treating physician's testimony as to whether plaintiff's cancer was mesothelioma; district court had reasoned that doctor was not qualified because he was not an oncologist; stating that trial court erroneously "restricted Dr. Carpenter’s testimony based on a requirement that the witness practice a particular specialty to testify concerning certain matters”); Paoli I, 916 F.2d at 856 (district court excluded witnesses who would have testified about gas chromatography tests and differential diagnoses, because they lacked degrees in chemistry and medicine respectively: "In light of the liberal Rule 702 expert qualification standard, we hold that the district court abused its discretion in excluding portions of [the experts'] testimony simply because the experts did not have the degree or training which the district court apparently thought would be most appropriate.” (footnote omitted)); Habecker, 893 F.2d at 52-53 (concluding that district court abused its discretion when it excluded expert testimony concerning connection between lack of operator restraints and plaintiff's injury, where plaintiff was injured when he was thrown from the cab of a forklift; district court's sole reason for finding expert was not qualified was because he lacked an engineering degree); Knight, 596 F.2d at 88 (finding error in district court’s exclusion of expert testimony concerning whether unguarded elevator control buttons were a design defect "because it believed that such expertise would require some background in the design and manufacture of elevators,” which proposed expert lacked; noting our "reluctance to require highly particularized, subspecialization on the part of experts”).
. The District Court held that Dr. Got-theimer was not qualified because his experience lay largely in the area of property-casualty insurance, not life insurance, noting that Dr. Gottheimer testified that there were "fundamental differences” between the two. See 1997 WL 732473, at *22. This conclusion contradicts our holdings in Knight and Holbrook, in which we reversed district courts’ exclusions of experts whose expertise the trial courts concluded was not sufficiently specialized. As we emphasized in those cases, we are reluctant "to require highly particularized, sub-specialization on the part of experts.” Knight, 596 F.2d at 88; accord Holbrook, 80 F.3d at 782. Any differences between the two areas “should go to the weight, and not the admissibility, of [the expert's] opinion.” Knight, 596 F.2d at 88.
. The District Court's focus on the nature of Dr. Gottheimer's doctorate and its consequent implicit dismissal of his other qualifications is also inconsistent with the exercise of sound discretion. If the district courts abused their discretion in Habecker, Paoli I and Lauria by insisting that the expert have a particular type of degree, the District Court in this case erred in insisting that the expert have not just a particular degree, but a degree from a particular kind of school. Cf. Lauria, 145 F.3d at 599; Paoli I, 916 F.2d at 856.
Furthermore, the District Court's reliance on Van Blargan v. Williams Hospitality Corp., 754 F.Supp. 246 (D.P.R.1991), in discounting Dr. Gottheimer’s degree is misplaced. In that case, the district court excluded an expert because, in addition to discounting his doctorate from a correspondence school, the court found that he had no other satisfactory qualifications. See Van Blargan, 754 F.Supp. at 248-49. Here, by contrast, Dr. Gottheimer has numerous other qualifications in addition to his doctorate.
. See Lauria, 145 F.3d at 600 ("Finally, we note that because Slavin was the only witness originally offered to prove Amtrak’s negligence with respect to the base tie, his exclusion from the trial did not constitute harmless error.”); Holbrook, 80 F.3d at 787 (finding that the error in excluding plaintiff's two doctors who were his only evidence regarding the type of cancer from which he suffered was not harmless); see also Habecker, 893 F.2d at 53 (finding that the error in excluding one of plaintiff's two expert witnesses was not harmless where defendant proffered three experts on the same point).
. In particular, plaintiffs’ attorney stated that, if he had been permitted to testify, Dr. Gottheimer’s testimony would have been as follows:
We have offered [Dr. Gottheimer’s] testimony to establish ... that in the insurance industry, there are some standard tools of tests that are performed in conducting an analysis of the financial condition of insurance companies, both life and health, and property and casualty, in terms of ratios that are generated from the annual statement, tests prescribed by the National Association of Insurance Commissioners, and the Best leverage and liquidity and profitability ratios.
The witness has performed these tests in his own capacity in the past. He has performed them with respect to Executive Life, based upon information that was available prior to the time of the three bids in question. He is able to interpret the tests. He is familiar with the tests. And his testimony would establish that the tests set up certain red flags that should have caused a person familiar with the tests and performing the tests to ask further questions about the solvency and the credit worthiness of Executive Life.
. The evidence on Unisys’s prudence admitted at trial included primarily the testimony of White and Level, two Unisys executives charged with responsibility for the Funds, and Becker, an advisor whose services White and Level engaged. The evidence before the District Court concerned several questions which, although the court decided them in Unisys's favor, raised serious questions about Unisys's prudence. (1) Whether Unisys conducted an adequate independent investigation into Becker’s recommendation? (2) Whether Unisys conducted an adequate investigation of its own after it ceased use of Becker's services? (3) Whether credit ratings were sufficient to prove that Unisys acted prudently? (4) Whether the Unisys trustees adequately considered and debated the advantages and disadvantages of the Executive Life GICs? (5) Whether the fact that the Executive Life GICs bore a higher interest rate was sufficient to suggest that they were not prudent investments? (6) Whether it was imprudent for Unisys not to maintain written investment guidelines?