Plaintiff-appellants Tanya Lacy and Richard Brooks were injured when the car in which they were passengers collided with a train operated by appellee CSX Transportation, Inc. (“CSX”), at a grade crossing in St. Albans, West Virginia in January 1995. Plaintiffs brought actions against both CSX and the driver of the car, Cacoe Sullivan, in the Circuit Court of Kanawha County. At the close of a two-week trial, which was bifurcated on the issues of liability and damages, the jury found both defendants negligent, but concluded in its special verdict that CSX’s negligence was not a proximate cause of the accident. Plaintiffs challenge the subsequent judgment entered in favor of CSX, arguing that (1) the lower court erred in refusing to instruct the jury on strict liability; (2) counsel for CSX was permitted to engage in improper argument with respect to the effect of West Virginia law concerning joint and several liability; and (3) the trial court erred in excluding a statement contained in a diagram prepared by a CSX employee following the collision indicating the location of one of the locomotives involved in the accident. We reverse, finding merit in the latter two contentions.

I.

BACKGROUND

Shortly after 11:00 p.m. on January 11, 1995, a car driven by Cacoe Sullivan left the Kroger parking lot in St. Albans, heading west on Third Avenue. Sullivan’s fiancee, Richard Brooks, was riding in the front passenger’s seat, while her mother, Tanya Lacy, was in the back seat with Sullivan’s and Brooks’s infant son. CSX’s railroad tracks, comprised of two main-line and two side tracks, run parallel to Third Avenue immediately to the south.

While traveling on Third Avenue, Sullivan’s car encountered a stop sign from where the occupants could see that the flashing lights and gates of the still-distant Fifth Street crossing were activated. Sullivan’s vehicle proceeded to the intersection of Third Avenue and Fifth Street (adjacent to the crossing), slowed but did not stop at a stop sign, made a left turn onto Fifth Street, went around one of the lowered gate arms onto the tracks, and was struck broadside by a westbound train traveling at 50 miles per hour.1 Brooks was apparently rendered paraplegic by the accident.

It was undisputed that from Sullivan’s view traveling on Third Avenue, a second, slower-*636moving “shifter” locomotive could be seen approaching the crossing from the west. There was, however, conflicting evidence regarding just how distant this locomotive was at the time of the accident. The testimony of the eastbound locomotive’s engineer, Calvin Bowen, placed it as close as 300 to 400 feet west of the Fifth Street crossing, traveling at fifteen to twenty miles per hour, when the car was struck by the westbound train. Plaintiffs proffered evidence in the form of a diagram prepared by a CSX accident investigator, G.A. Green (the “accident diagram”), indicating that the eastbound locomotive was further away, as far as two to three blocks to the west of the crossing; however, this evidence was excluded by the trial court.

The central issue at trial with respect to CSX was whether it was negligent in permitting both fast- and slow-moving locomotives to approach the Fifth Street crossing simultaneously on its main-line tracks. The crossing had an active warning system consisting of flashing-light signals and automatic gates. Plaintiffs asserted at trial that the ability of the crossing warning system to provide a “positive warning” of an approaching train was effectively neutralized by CSX’s practice of allowing slow-moving switching locomotives to use the main-line tracks. It was alleged that this practice frequently resulted in the extended activation of the crossing’s flashing lights and gate arms when no trains were in hazardous proximity. As a result, according to plaintiffs, CSX was not using the warning system in accordance with its design, and thus was not in compliance with 49 C.F.R. § 234.225 (1998).2

Several witnesses, including Sullivan, testified to their past experience of encountering extended activations because of slow-moving trains in the vicinity of the Fifth Street crossing. The former mayor of St. Albans, Edward Bassitt, indicated that he had previously discussed with CSX the problem of extended activations at the Fifth Street crossing as early as 1989. CSX employees also testified to the fact that drivers in the St. Albans area frequently ignored the crossing warning signals.

Plaintiffs’ expert in the area of grade-crossing safety, William Berg, Ph.D., testified that the fixed-distance circuitry installed on the main-line tracks at the Fifth Street crossing is designed to activate a warning whenever an approaching train is within 2,000 to 2,200 feet of the crossing, regardless of the train’s speed. Thus, while a train traveling at the maximum speed of sixty miles per hour would give a twenty-five second warning, the approach of a slower-moving locomotive could result in much longer warning times. Dr. Berg estimated, based in part upon information contained in the accident diagram, that the eastbound locomotive would have activated the warning system over forty seconds prior to the accident.

Dr. Berg further stated that the optimal warning time was twenty-five to thirty seconds, and that warning times in excess of forty seconds result in a dramatic increase in the number of people driving around deployed gates. He stressed the importance of giving motorists credible warnings, and the need to provide uniform warning times at crossings where there are significant disparities in train speeds.3 As one example of alternatives to CSX’s practices, Dr. Berg-pointed to so-called constant warning-time technology (“CWT”), which gives a consistent warning regardless of the speed of the approaching train. Other alternatives cited by Dr. Berg included relegating slow-moving *637trains to side tracks, where the fixed distance circuitry is specifically designed to accommodate the lower speeds,4 or keeping slower locomotives outside of the circuitry on the main-line tracks when faster trains are approaching.

In its ease, CSX presented the testimony of Gary Wolf, an expert in railway operations, and Joseph Blaschke, Ph.D., an expert in traffic engineering and highway design. Both of these witnesses rejected the contention that twenty-five to thirty seconds was an optimal warning time, and cited the absence of any federal regulation mandating maximum warning times.5 Each stated that CWT was intended primarily to improve vehicular flow at crossings, not to increase traffic safety. Dr. Blaschke also testified that CWT was indicated in situations involving both heavy vehicular traffic and heavy train activity, and that the Fifth Street crossing did not generate the level of vehicular traffic necessary to justify the installation of CWT. He gave the opinion that the characteristics of the crossing, including the existing presence of an active warning system with gates, and the excellent sight distance at the crossing, did not make Fifth Street a priority candidate for CWT.

After hearing the evidence, the jury deliberated until sending a note indicating that they were having difficulty reaching a unanimous verdict. The trial court then gave an AJZe%-type instruction.6 The jury subsequently rendered a special verdict regarding liability, finding CSX and Sullivan, as well as plaintiffs Tanya Lacy and Richard Brooks, negligent, but determining that Sullivan’s negligence was the sole proximate cause of the accident. The jury ascribed one percent negligence each to CSX, Lacy and Brooks, and ninety-seven percent to defendant Sullivan.7 The circuit court entered judgment in *638favor of CSX based upon the jury’s special verdict. Plaintiffs’ subsequent Motion for a New Trial and Judgment Notwithstanding the Verdict was denied by the trial court.

II.

DISCUSSION

A.

Joint and Several Liability

Plaintiffs first contend that the trial court erred in permitting counsel for CSX to argue the potential post-judgment effects of joint and several liability to the jury.8 We reverse on this issue, finding that the trial court abused its discretion by permitting counsel for CSX to speculate and otherwise mislead the jury regarding whether the railroad would ultimately be charged with paying the entire judgment if both CSX and defendant Sullivan were found at fault.

Prior to trial, plaintiffs filed a motion in limine “to exclude any questions, suggestions, comments, allegations, testimony or argument by the defendant, [CSX], as to the effect that West Virginia’s joint and several liability law may have upon [CSX].” The circuit court ruled on the motion after CSX proposed an instruction on the issue.9 CSX argued to the trial court that

the jury needs to understand the relationship between the parties and the effect of a finding of either one percent against CSX, because of the way the plaintiffs have plead[ed] this case and argued this case, because of the relationship between the parties and the cooperation that we’ve seen here, the jury needs to fully understand that.

This is no different tha[n] plaintiffs who every time we seem to try a case come in and tell the jury, well, if you find more than fifty percent at fault then this plaintiff is not going to get any money. It’s the same thing.

The trial court refused CSX’s proposed instruction, but nevertheless ruled that CSX could argue joint and several liability and “point out the intrigue.” Counsel for plaintiffs brought the issue to the court’s attention a second time immediately before closing arguments, citing Valentine v. Wheeling Elec. Co., 180 W.Va. 382, 376 S.E.2d 588 (1988), and asserting that such argument would be tantamount to “invit[ing] the jury to nullify the law of joint and several liability in West Virginia.” The trial court again ruled to permit CSX to argue joint and several liability.

Counsel for CSX stated the following during closing argument:

Let’s just stop for a minute and let’s talk about what this case is really about, what *639has been going on here for two weeks in this trial. Tanya Lacy, Richard Brooks, and Cacoe Sullivan are family. This is not a case where we have two plaintiffs suing two defendants. This is a case in which the family is trying to get money from the railroad. Tanya Lacy doesn’t want anything from her daughter.

They spent two weeks trying to convince you that CSX was at fault. They didn’t spend two weeks trying to convince you that Cacoe Sullivan was at fault. Why not? I’ll tell why not. If you go back into that jury room and return this verdict of shared responsibility that [plaintiffs’ counsel] wants, if you go back into that jury room and return a verdict that say[s] ... 99 percent Cacoe Sullivan’s fault, 1 percent CSX’s fault, guess what? Tanya Lacy and Richard Brooks can collect the entire judgment from CSX. They can also collect it from Cacoe Sullivan, if they wanted, but what are the odds a mother is going to actually ask her daughter.

So when you go back into that jury room and fill out this verdict form, any finding on the part of CSX, 1 percent, 10 percent, 50 percent, 100 percent, it’s the same thing. One percent is, in essence, telling CSX, you are completely and totally responsible for this accident.

So you have two choices when you go out on this verdict form. You can find that the responsibility for this accident was solely Cacoe Sullivan’s fault, or solely CSX’s fault, because any split and they’re going to come looking for us.

Counsel for Cacoe Sullivan objected to this argument at the time it was delivered, but was overruled by the trial court.

As a threshold matter, CSX obliquely suggests that appellate review of this issue is barred by plaintiffs’ failure to object during closing argument. This Court previously held in Syllabus point 1 of Winter v. Hinkle, 180 W.Va. 660, 379 S.E.2d 383 (1989), that “[a]n objection to an adverse ruling on a motion in limine to bar evidence at trial will preserve the point, even though no objection was made at the time the evidence was offered, unless there has been a significant change in the basis for admitting the evidence.” While the present case involves the arguments of counsel rather than the introduction of evidence, the underlying principle is equally applicable such that to preserve error with respect to closing arguments by an opponent, a party need not contemporaneously object where the party previously objected to the trial court’s in limine ruling permitting such argument, and the argument subsequently pursued by the opponent reasonably falls within the scope afforded by the court’s ruling. This conclusion is bolstered by West Virginia Trial Court Rule 23.04, which disfavors objections by counsel during closing arguments: “Counsel shall not be interrupted in argument by opposing counsel, except as may be necessary to bring to the court’s attention objection to any statement to the jury made by opposing counsel and to obtain a ruling on such objection.”

Plaintiffs were not, in this case, required to lodge an objection at the time counsel for CSX made the challenged remarks, since the trial court had already ruled in limine on plaintiffs’ objection to this line of argument. Consequently, the present issue has been preserved for review.

Reaching the merit of plaintiffs’ assertion of error, we note that this Court reviews rulings by a trial court concerning the appropriateness of argument by counsel before the jury for an abuse of discretion. “[A] trial court has broad discretion in controlling argument before the jury,” Dawson v. Casey, 178 W.Va. 717, 721, 364 S.E.2d 43, 47 (1987) (per curiam) (citation omitted), and such discretion “will not be interfered with by the appellate court, unless it appears that the rights of the complaining party have been prejudiced, or that manifest injustice resulted therefrom,” Syl. pt. 3, State v. Boggs, 103 W.Va. 641, 138 S.E. 321 (1927). See also Syl. pt. 2, State v. Bennett, 183 W.Va. 570, 396 S.E.2d 751 (1990). As we noted in Syllabus point 8 of Mackey v. Irisari, 191 W.Va. 355, 445 S.E.2d 742 (1994), “ ‘[g]reat latitude is allowed counsel in argument of cases, but counsel must keep within the evidence, not make statements calculated to inflame, prejudice or mislead the jury, nor permit or encourage witnesses to make re*640marks which would have a tendency to inflame, prejudice or mislead the jury.’ ” (Quoting Syl. pt. 2, State v. Kennedy, 162 W.Va. 244, 249 S.E.2d 188 (1978)).

While our ultimate focus is therefore on whether the trial court abused its discretion by permitting argument regarding the potential post-judgment effect of joint and several liability, such inquiry requires us to first determine whether it is generally appropriate to inform the jury about this doctrine.

This Court previously touched upon the issue of whether it is permissible to instruct a jury on joint and several liability in Valentine v. Wheeling Elec. Co., 180 W.Va. 382, 376 S.E.2d 588 (1988). In Valentine, the trial court gave an instruction identical to that proposed by CSX in the present case. See note 9, supra. On appeal, the defendant attempted to justify the trial court’s action by citing to Syllabus point 2 of Adkins v. Whitten, 171 W.Va. 106, 297 S.E.2d 881 (1982), which requires a court to instruct the jury on the doctrine of modified comparative negligence when requested. The Court rejected the applicability of Adkins, observing that the proposed

instruction ... did not deal with measuring the negligence of the appellant against that of the alleged tortfeasors. Rather, it dealt with apportionment of damages among alleged tortfeasors. Consequently, the duty of a trial court imposed by Adkins to instruct the jury as to the effect of comparative negligence did not exist in this case.

Valentine, 180 W.Va. at 386, 376 S.E.2d at 592 (emphasis in original). After concluding that the proffered instruction “was not appropriate in this case,” id., the Court went on to find that any error in instructing the jury on joint and several liability was harmless because the jury found neither defendant negligent. While Valentine certainly suggests that it is error to inform the jury of the doctrine of joint and several liability, the Court in that case did not attempt to forge a broad rule concerning this issue.

There are divergent views concerning the appropriateness of informing the jury of the effects of joint and several liability. Some jurisdictions, employing the same rationale used to permit instruction and argument on the workings of modified comparative negligence, sanction informing juries about joint and several liability because, in their estimation, juries are likely to respond to such information by being more conscientious about assigning responsibility to defendants. See DeCelles v. State ex rel. Dep’t of Highways, 243 Mont. 422, 425, 795 P.2d 419, 421 (1990); Luna v. Shockey Sheet Metal & Welding Co., 113 Idaho 193, 195-97, 743 P.2d 61, 64-65 (1987); Kaeo v. Davis, 68 Haw. 447, 460-61, 719 P.2d 387, 396 (1986). For example, in Luna, the Idaho Supreme Court stated that

the doctrine of joint and several liability, under which a defendant assessed a mere 1% negligence may be required to pay 100% of plaintiffs damages if, for some reason, the joint tortfeasor is unreachable through the judicial process, “poses a trap for the uninformed jury.” An informed jury will be much more likely to carefully examine the facts prior to reaching a verdict holding a defendant even 1% at fault, no matter how cosmetically appealing a partial allocation of fault might be.

Luna, 113 Idaho at 196, 743 P.2d at 64.

Other courts stress that consideration of joint and several liability is not relevant to determining any issue of fact. The Court of Appeals of South Carolina recently took this approach, where it held that it was not error for a trial court to refuse an instruction on joint and several liability “[b]ecause the doctrine has no bearing on the jury’s ultimate fact-finding role in determining the relative negligence of joint tortfeasors.” Fernanders v. Marks Constr. of S.C., Inc., 330 S.C. 470, 475, 499 S.E.2d 509, 510-11 (S.C.Ct.App. 1998); see also Dranzo v. Winterhalter, 395 Pa.Super. 578, 592, 577 A.2d 1349, 1356 (1990) (“the collectibility or uncollectibility of a judgment, the operation of joint and several liability, is simply not relevant to the jury’s consideration of whether the defendants were causally liable and in what percent”); Gehres v. City of Phoenix, 156 Ariz. 484, 486-87, 753 P.2d 174, 176-77 (Ariz.Ct.App. 1987).

Courts on both sides of the debate take credible positions; however, we perceive that *641resolution of this issue turns on practical considerations that have only been lightly touched upon.

In Adkins, the Court determined that a jury’s apportionment of fault would be more reliable if it were instructed on the operation of the doctrine of modified comparative fault:

From a practical standpoint it is apparent that a jury, given the type of verdict form mandated by Bradley [v. Appalachian Power Co., 163 W.Va. 332, 256 S.E.2d 879 (1979),] which requires a gross damage verdict and a finding of the plaintiffs percentage of negligence, may well surmise that the plaintiffs negligence may reduce his damage award. It seems to us that a jury’s deliberations should not be attended by such surmises but rather they should be openly informed as to the legal principles involved in our comparative negligence doctrine so that they may make a rational decision.

171 W.Va. at 109, 297 S.E.2d at 884. The instruction sanctioned in Adkins “enables the jury to understand the mechanics of the comparative contributory negligence rule,” King v. Kayak Mfg. Corp., 182 W.Va. 276, 279, 387 S.E.2d 511, 514 (1989), and thus eliminates the possibility that it will premise its factual findings upon an erroneous perception of the legal consequences of finding a plaintiff partially at fault. In determining that the utility of instructing the jury on comparative negligence outweighs other countervailing considerations, we rejected the notion that such information would inevitably result in bias on the part of the jury:

To argue that a jury once informed of the comparative negligence law might manipulate it in order to favor the plaintiff assumes a biased jury. Such an argument is premised on a theory that individual jurors will disregard their oaths to follow the court’s instructions as to the law. The same argument could as easily be made in regard to any instruction on any aspect of the law. We do not believe that jurors will disregard their obligations to apply the law objectively to the facts of the case.

Adkins, 171 W.Va. at 109, 297 S.E.2d at 884

The Court in Adkins also stressed that “under our jury trial system, it is incumbent on the court by way of instruction or charge to inform the jury as to the law applicable to the facts of the case. This should be the ease as to our law of comparative negligence.” Id. This requirement is echoed in W. Va. R. Civ. P. 49(a), which in the context of special verdicts requires a court to “give the jury such explanation and instruction concerning the matter thus submitted as may be necessary to enable the jury to make its findings upon each issue.” Professors Wright and Miller assert that the “preferable” approach under Rule 49(a) is to inform juries of the legal effect of their factual findings, noting (much as we did in Adkins) that to do otherwise “is likely to be unavailing, and there is always the danger that the jury will guess wrong about the law, and may shape its answers to the special verdicts, contrary to its actual beliefs, in a mistaken attempt to ensure the results it deems desirable.” 9A Charles A. Wright & Arthur R. Miller, Federal Practice and Procedure § 2509, at 197-98 (2d ed. 1995).

This Court obviously does not generally sanction “blindfolding” the jury regarding the legal effect of its factual findings, but nor do we endorse informing the jury of every conceivable consequence that may attach to such findings.

In addition to the superficial reasons stated in Valentine, we perceive an even more significant justification for rejecting application of Adkins in the context of joint and several liability: Any conclusion about how joint and several liability will ultimately affect a particular defendant is largely speculative. As the Superior Court of Pennsylvania pointed out in holding that is was proper for a trial court to refuse a jury instruction on joint and several liability, “neither the court nor the jury can say with assurance how much of the verdict rendered, if any, any one tortfeasor will in fact pay.” Dranzo, 395 Pa.Super. at 592, 577 A.2d at 1356; see also DeCelles, 243 Mont. at 429, 795 P.2d at 423 (Sheehy, J. dissenting) (“instruction [on joint and several liability] allowed the jury to speculate as to matters outside the evidence in this case”) (emphasis in original).

*642When a jury that has been instructed under Adkins considers the doctrine of comparative negligence in the context of apportioning fault, it is not required to speculate about the consequences of its verdict. Rather, the jury can easily comprehend what effect its findings will have on the litigants, without any need to consider evidence beyond that relevant to determination of the cause of action. The same cannot be said of the jury’s consideration of joint and several liability, where in most cases the ultimate financial impact of a jury verdict on individual defendants cannot be fully appreciated by anyone until long after judgment is rendered.

This Court has consistently rejected permitting counsel to base arguments before the jury upon mere speculation. In Syllabus point 2 of Jenrett v. Smith, 173 W.Va. 325, 315 S.E.2d 583 (1983), we noted that ‘“[tjhough wide latitude is accorded counsel in arguments before a jury, such arguments may not be founded on facts not before the jury, or inferences which must arise from facts not before the jury.’ ” (Quoting Syl. pt. 3, Crum v. Ward, 146 W.Va. 421, 122 S.E.2d 18 (1961)). Consequently, arguments that amount to “nothing more than speculation and conjecture ... [are] properly excluded _” Gardner v. CSX Transp., Inc., 201 W.Va. 490, 502, 498 S.E.2d 473, 485 (1997). Similarly, a court’s instructions should not prompt the jury to speculate as to facts that are not in evidence. Cf. Syl. pt. 1, Oates v. Continental Ins. Co., 137 W.Va. 501, 72 S.E.2d 886 (1952) (“A jury will not be permitted to base its findings of fact upon conjecture or speculation.”).

The line of argument pursued by CSX in the present case demonstrates how any consideration of the potential post-judgment effects of joint and several liability is likely to degenerate into conjecture about whether a particular defendant will ultimately bear a greater portion of the plaintiffs loss than is attributable to its fault. Counsel for CSX speculated that plaintiffs would be unwilling to collect any judgment against Cacoe Sullivan, and would instead resort to forcing CSX to pay the entire judgment. While such an outcome is perhaps a plausible inference given the unique familial relationship of these parties, there was nothing in evidence that otherwise directly supported such a contention.

CSX’s argument was, in any event, misleading to the extent that it implied that plaintiffs could ultimately control who would pay. This obviously ignores the fact that CSX would, if it were called upon by plaintiffs to satisfy the entire judgment, have a right of comparative contribution against Sullivan. W. Va.Code § 55-7-13 (1923); Syl. pt. 3, Sitzes v. Anchor Motor Freight, Inc., 169 W.Va. 698, 289 S.E.2d 679 (1982) (“As between joint tortfeasors, a right of comparative contribution exists inter se based upon their relative degrees of primary fault or negligence.”); Syl. pt. 3, Haynes v. City of Nitro, 161 W.Va. 230, 240 S.E.2d 544 (1977) (“In West Virginia one joint tort-feasor is entitled to contribution from another joint tort-feasor, except where the act is malum in se.”).

To inform the jury about the potential effects of joint and several liability without otherwise misleading it, trial courts could conceivably be required to instruct and/or permit evidence on such complex and often proscribed subjects as contribution, indemnity, bankruptcy, the effect of statutory and common-law immunities, the extent of defendants’ financial resources, and the existence of insurance coverage — just to name a few. Our discussion in Riggle v. Allied Chem. Corp., 180 W.Va. 561, 378 S.E.2d 282 (1989), illustrates the potentially unwieldy consequences that might flow from reading Adkins to require instruction on the post-judgment effects of a jury’s findings. In that case, the defendant-appellant, Griffith Brothers Contractors, claimed error with respect to the trial court’s refusal to instruct the jury on the effect of an agreement requiring it to indemnify its co-defendant, Allied Chemical, except where the latter was entirely at fault. Allied Chemical had entered into a “Mary Carter” settlement agreement prior to trial, which provided, among other things, that it would pay plaintiffs everything it recovered from its crossclaim under the indemnity agreement. (Allied Chemical had voluntarily limited its indemnity crossclaim to $500,-000 — the extent of Griffith Brothers’ insur-*643anee coverage.) The Court summarized Griffith Brothers’ argument as follows:

Appellant maintains that the court should have revealed the indemnity agreement to the jury. It argues that the jury should have known that if it assessed even one percent of the fault to Griffith Brothers, then the latter would have to pay the entire judgment. Appellant argues that because a court, on request, must instruct the jury on the effect of finding a plaintiff fifty percent negligent, the court should have instructed the jury on the effect of the indemnity agreement here.

Riggle, 180 W.Va. at 568, 378 S.E.2d at 289. The Court prefaced its discussion by noting that “we did not rule in Adkins that the jury should be instructed on how the court computes the amount of the plaintiffs’ recovery from the relative percentages of negligence and the amount of total damages.” Id. In concluding that the trial court did not abuse its discretion in refusing to instruct on the indemnity agreement, the Court in Riggle stated that “[t]he problem with instructing the jury on the indemnity agreement ... is that such an instruction would have been misleading without also instructing the jury on the settlement agreement and the insurance coverage of appellant. Comment to a jury concerning a party’s insurance coverage usually constitutes reversible error.” Id. (citation omitted). While the facts in Riggle may be unique, any attempt to analogously inform the jury about the legal effect of joint and several liability is likely to encounter similar obstacles.

We are not inclined to sanction forays into matters that invite speculation and conjecture on the part of the jury, and which do not suggest an easy stopping point with respect to the disclosures necessary to avoid misleading the trier of fact. Nor in the case of joint and several liability do we discern, as we did in Adkins, that juries are likely to harbor or otherwise act upon misconceptions regarding this doctrine. Accordingly, we hold that in a civil trial it is generally an abuse of discretion for the trial court to instruct the jury or permit argument by counsel regarding the operation of the doctrine of joint and several liability, where the purpose thereof is to communicate to the jury the potential post-judgment effect of their assignment of fault.

The circuit court in this case abused its discretion not only by permitting CSX to inform the jury about the possible legal effect of joint and several liability, but also by allowing it to go so • far as to effectively exhort the jury to absolve it of all liability on such basis. Counsel stated that “[o]ne percent is, in essence, telling CSX, you are completely and totally responsible for this accident,” a theme that was stressed a second time when the jury was told that it had a choice between two ultimate outcomes: “You can find that the responsibility for this accident was solely Cacoe Sullivan’s fault, or solely CSX’s fault, because any split and they’re going to come looking for us.” These statements gave the misleading impression that if CSX was found in any way at fault, it would invariably be left to pay the entire judgment. If, as we have repeatedly declared, “[t]his jurisdiction is committed to the concept of joint and several liability among tortfeasors,” Syl. pt. 2, in part, Sitzes, supra, a defendant cannot be permitted to argue against a finding of fault based upon misleading speculation about the possible ramifications of the doctrine’s application.

CSX argues that even if the trial court did abuse its discretion by permitting it to argue the effects of joint and several liability, such error was harmless because, notwithstanding its request for an opposite result, the jury found the railroad one percent negligent. Therefore, according to CSX, its closing argument had no prejudicial effect on the jury. We are not persuaded.

Under W. Va. R. Civ. P. 61,10 “[a] party is entitled to a new trial only if there is *644a reasonable probability that the jury’s verdict was affected or influenced by trial error.” Tennant v. Marion Health Care Found., Inc., 194 W.Va. 97, 111, 459 S.E.2d 374, 388 (1995). In making this determination, the reviewing court must ascertain whether it has “ ‘grave doubt about the likely effect of an error on the jury’s verdict’ ...; if a court does have grave doubt, then the error is harmful.” Skaggs v. Elk Run Coal Co., Inc., 198 W.Va. 51, 71, 479 S.E.2d 561, 581 (1996) (internal citation omitted).

The use of special verdicts or written interrogatories pursuant to W. Va. R. Civ. P. 49 generally aids appellate review, see 9 James W. Moore, Moore’s Federal Practice § 4.11[1][a], at 49-11 (3d ed. 1998); however, such devices are not particularly helpful in assessing the repercussions of broadly prejudicial remarks made before a jury. In this case, moreover, the interrogatory answer assigning one percent negligence to CSX was meaningless, since the trial court ultimately ruled that the railroad was not at fault based upon the jury’s finding regarding the absence of proximate cause.11 Regardless of how it reached its conclusion, the jury still found that CSX was not responsible for the accident. Because CSX’s argument concerning joint and several liability advocated such an ultimate outcome, we are left with grave doubts about the effect of such argument on the jury’s findings in this case. We are therefore precluded from finding that the trial court’s error was harmless.

B.

Admissibility of Accident Diagram

Plaintiffs also contend that the trial court erred in excluding a portion of a diagram prepared by a CSX employee in the course of investigating the accident, which includes a statement indicating the location of the slower-moving eastbound locomotive at time of the collision. They assert that such evidence was a record of a regularly conducted activity, and was thus admissible as an exception to hearsay under W. Va. R. Evid. 803(6). In support of the trial court’s exclusion of such evidence on hearsay grounds, CSX argues that its “investigator was not a witness to the accident, did not testify at trial, and based his conclusion on the position of the [eastbound] locomotive on what others told him in his investigation.” We conclude that the court below erred in interpreting Rule 803(6) to require plaintiffs to establish that the statement was within the personal knowledge of the railroad investigator who made the accident diagram.

The diagram, which was prepared in connection with an accident report mandated by state and federal law,12 provides a representation of the Fifth Street crossing indicating the relative movements of the car and train involved in the collision. The diagram also *645contains the following handwritten notation attached to a westward-pointing arrow: “ST. ALBANS SHIFTER 2 TO 3 BLOCKS WEST OF X-ING #2 MAIN TRACK.” Plaintiffs’ expert, Dr. Berg, in part relied upon this statement in forming his opinion regarding the cause of the accident. Also, plaintiffs apparently proffered this statement to deflect any contention that the driver, Cacoe Sullivan, was preoccupied with the eastbound locomotive due to its relative proximity to the crossing. CSX focused attention on the testimony of the engineer of the eastbound shifter, Calvin Bowen, indicating that it was only 300 to 400 feet west of the Fifth Street crossing when the car was struck by the westbound train. The railroad put considerable emphasis, both in its cross-examination of witnesses and during opening and closing statements, upon the shorter distance in establishing that the locomotive was in hazardous proximity to the crossing, and that Sullivan was attempting to “beat the eastbound train.”

Plaintiffs, at the conclusion of their case-in-chief, moved to admit the accident diagram in its entirety. CSX objected to the inclusion of the statement regarding the location of the eastbound shifter locomotive.13 The trial court subsequently ruled to admit the diagram as a record of a regularly conducted activity under W. Va. R. Evid. 803(6), but excluded the statement regarding the location of the eastbound locomotive. The court concluded that this statement was “third-hand hearsay” based upon the fact that the preparer of the diagram, G.A. Green, did not directly observe the accident in question, but instead apparently relied upon statements made by other eyewitnesses. It analogized to accident reports prepared by the police, stating that such evidence is often redacted to exclude hearsay statements. After the court made its ruling, plaintiffs elected not to put the remainder of the accident diagram into evidence.14

With few exceptions, this Court reviews evidentiary rulings made by a trial court for an abuse of discretion:

The West Virginia Rules of Evidence and the West Virginia Rules of Civil Procedure allocate significant discretion to the trial court in making evidentiary and procedural rulings. Thus, rulings on the admissibility of evidence ... are committed to the discretion of the trial court. Absent a few exceptions, this Court will review eviden-tiary and procedural rulings of the circuit court under an abuse of discretion standard.

Syl. pt. 1, McDougal v. McCammon, 193 W.Va. 229, 455 S.E.2d 788 (1995). We have previously held that “[t]he action of a trial court in admitting or excluding evidence in the exercise of its discretion will not be disturbed by the appellate court unless it appears that such action amounts to an abuse of discretion.” Syl. pt. 10, State v. Huffman, 141 W.Va. 55, 87 S.E.2d 541 (1955), overruled on other grounds, State ex rel. R.L. v. Bedell, 192 W.Va. 435, 452 S.E.2d 893; see also Syl. pt. 4, Riggle v. Allied Chem. Corp., 180 W.Va. 561, 378 S.E.2d 282 (1989). However, “[a]n interpretation of the West Virginia Rules of Evidence presents a question of law subject *646to de novo review.” Syl. pt. 1, Gentry v. Mangum, 195 W.Va. 512, 466 S.E.2d 171 (1995). Thus, “a trial court’s ruling on the admissibility of testimony is reviewed for an abuse of discretion, ‘but to the extent the [circuit] court’s ruling turns on an interpretation of a [West Virginia] Rule of Evidence our review is plenary.”’ State v. Sutphin, 195 W.Va. 551, 560, 466 S.E.2d 402, 411 (1995) (citations omitted); see also State v. Quinn, 200 W.Va. 432, 435, 490 S.E.2d 34, 37 (1997). Because the circuit court’s ruling was grounded upon a legal conclusion regarding the foundational requirements of Rule 803(6), we undertake de novo review to determine whether the court’s ruling was based upon a permissible interpretation of this rule.

Before evidence may be admitted under W. Va. R, Evid. 803(6),15 the proponent must demonstrate that such evidence is (1) a memorandum, report, record, or data compilation, in any form; (2) concerning acts, events, conditions, opinions or diagnoses; (3) made at or near the time of the matters set forth; (4) by, or from information transmitted by, a person with knowledge of those matters; (5) that the record was kept in the course of a regularly conducted activity; and (6) that it was made by the regularly conducted activity as a regular practice. See generally 2 Franklin D. Cleckley, Handbook on Evidence for West Virginia Lawyers § 8-3(B)(6), at 223 (3d ed. 1994).

Our focus here is on whether the trial court erred by excluding the statement in question on the basis of appellant’s asserted failure to satisfy the fourth foundational requirement of Rule 803(6), that the information set forth in the record be derived from a source with knowledge.

The requirement that statements contained in a record must be derived from sources with knowledge is a reflection of W. Va. R. Evid. 602.16 The Advisory Committee Notes to Fed.R.Evid. 803 make clear that “this rule ... [does not] dispense! ] with the requirement of first-hand knowledge.” Consequently, to be admissible under Rule 803(6), the matters set forth in a record must either be based upon the personal knowledge of the recorder, or the knowledge of the supplier of the recorded information, who together with those involved in transmitting it to the final recorder, must be acting routinely and under a duty of accuracy.17

*647Because of the complexities inherent in modern record-keeping processes, however, the knowledge element of Rule 803(6) is liberally construed. This Court has previously noted with respect to other hearsay exceptions contained in Rule 803 that “[t]he personal knowledge requirement, while not de minimis, is not meant to be a very difficult standard and may be satisfied if it is more likely than not that the evidence proves the percipiency of the declarant.” State v. Phillips, 194 W.Va. 569, 578, 461 S.E.2d 75, 84 (1995) (citing, inter alia, W. Va. R. Evid. 104(b)). Whether a statement contained in a record is based upon a source with knowledge “may appear from ... [the declarant’s statement] or be inferable from circumstances.” Advisory Committee Notes to Fed.R.Evid. 803.

McCormick discusses the inferential means by which the knowledge requirement of Rule 803(6) may be satisfied:

Direct proof that the maker of the statement had actual knowledge may be difficult, and it may even be impossible to prove specifically the identity of the informant with actual knowledge. Evidence that it was someone’s business duty in the organization’s routine to observe the matter will be prima facie sufficient to establish actual knowledge. This does not dispense with the need for personal knowledge, but permits it to be proved by evidence of practice and a reasonable assumption that general practice was followed with regard to a particular matter, or by other appropriate circumstances.

2 Kenneth S. Braun, et al., McCormick on Evidence § 290, at 275 (John W. Strong ed., 4th ed. 1992) (footnotes omitted); see also 5 Jack B. Weinstein & Margaret A. Berger, Weinstein’s Federal Evidence § 803.11[4], at 803-71 (Joseph M. McLaughlin ed., 2d ed. 1999) (“The name of the person whose firsthand knowledge was the basis of the entry need not be known so long as the regular practice was to get the information from such a person.”) (footnote omitted). At the time Rule 803(6) was incorporated into the Federal Rules of Evidence, the Senate Judiciary Committee stated its understanding that “[a] sufficient foundation ... will be laid if the party seeking to introduce the evidence is able to show that it was the regular practice of the activity to based such memorandums, reports, records, or data compilations upon a transmission from a person with knowledge ....” S.Rep. No. 1277, at 17 (1974), reprinted in 1974 U.S.C.C.A.N. 7051, 7063.

As numerous federal courts applying Rule 803(6) have concluded, “there is no requirement that the person whose first-hand knowledge was the basis of the [record] entry be identified, so long as it was the ... entity’s regular practice to get information from such a person.” Saks Int’l, Inc. v. M/V Export Champion, 817 F.2d 1011, 1013 (2d Cir. 1987); see also Munoz v. Strahm Farms, Inc., 69 F.3d 501, 503-4 (Fed.Cir. 1995); Baxter Healthcare Corp. v. Healthdyne, Inc., 944 F.2d 1573, 1577 (11th Cir. 1991), vacated and dismissed per stipulation, 956 F.2d 226 (1992); White Indus. v. Cessna Aircraft Co., 611 F.Supp. 1049, 1059-60 (D.Mo. 1985); United States v. Lieberman, 637 F.2d 95, 100 (2d Cir. 1980) (“direct proof of actual knowledge of the person making the record or providing the information is not required, and the requisite knowledge may be inferred from the fact that it was someone’s business to obtain such information”); United States v. Ahrens, 530 F.2d 781, 784 n. 6 (8th Cir. 1976) (Rule 803(6) “does not require personal knowledge of the maker of the record as a condition precedent to its admission into evidence”); Stone v. Morris, 546 F.2d 730, 738-39 (7th Cir. 1976).

Thus, in order to satisfy the knowledge requirement of Rule 803(6), the party seeking to admit such evidence may establish either (1) that the preparer of the record had knowledge of the matters reported; or (2) that the information reported was transmitted by a person with knowledge who was acting in the course of a regularly conducted activity; or (3) that it was a regular practice of the activity to rely upon communications from persons with knowledge. See In Re *648Japanese Electronic Prod. Antitrust Litig., 723 F.2d 238, 288 (3d Cir. 1983) (citing Zenith Radio Corp. v. Matsushita Elec. Indus. Co., 505 F.Supp. 1190, 1237 (E.D.Pa. 1980)), rev’d on other grounds sub nom., Matsushita Elec. Indus. Co., Ltd. v. Zenith Radio Corp., 475 U.S. 574, 106 S.Ct. 1348, 89 L.Ed.2d 538 (1986), abrogated on other grounds, Pfeiffer v. Marion Ctr. Area School Dist., 917 F.2d 779 (1990). We therefore hold that the circuit court erred as a matter of law in interpreting this rule as singularly requiring plaintiffs to prove that the preparer of the accident diagram, Mr. Green, had personal knowledge concerning the location of the eastbound shifter locomotive at the time of the collision.

We are handicapped in determining whether the trials court’s exclusion of this evidence is supportable on other grounds (i.e., whether plaintiffs otherwise satisfied the knowledge requirement of Rule 803(6)) by the fact that the parties stipulated that a custodian of the record in question was not required to testify. On this basis alone, we are inclined to excuse any failure by plaintiffs to demonstrate that the accident diagram was based upon a source with knowledge. See United States v. Saunders, 886 F.2d 56, 59 (4th Cir. 1989) (stipulation that police reports were “records that are maintained and kept as normal and customary business records” necessitated their admission at trial); Hard v. Stevens, 65 F.R.D. 637, 639-40 (E.D.Pa. 1975) (stipulation waiving necessity of calling custodian of records estopped party from challenging admission of hospital record based upon lack of trustworthiness). Nevertheless, we are satisfied that the knowledge requirement of Rule. 803(6) was met here, based upon circumstantial factors that give rise to an inference that the information contained in the accident diagram was derived from sources with knowledge.

While Rule 803(6) evidence is not self-authenticating, see State v. Fairchild, 171 W.Va. 137, 147, 298 S.E.2d 110, 120 (1982) (“in no instance may records of this kind prove themselves”); Daniel B. v. Ackerman, 190 W.Va. 1, 6, 435 S.E.2d 1, 6 (1993), this does not preclude admission where either the record itself or other circumstantial factors provide the necessary foundation. The “testimony of the custodian or other qualified witness is not a sine qua non of admissibility in the occasional case where the qualification as a business record can be met ... by circumstantial evidence, or by a combination of direct and circumstantial evidence.” Zenith Radio Corp. v. Matsushita Elec. Indus. Co., Ltd., 505 F.Supp. 1190, 1236 (E.D.Pa. 1980), modified on other grounds, 723 F.2d 238 (3d Cir. 1983); see also United States v. Franco, 874 F.2d 1136, 1140 (7th Cir. 1989); United States v. Kail, 804 F.2d 441, 449 (8th Cir. 1986); White Indus. v. Cessna Aircraft Co., 611 F.Supp. at 1060; see generally 5 Weinstein & Berger, supra, § 803.11[1], at 803-59 to -60. Thus, the foundation required by Rule 803(6) may be established by circumstantial evidence, or by a combination of direct and circumstantial evidence.

In this case, two factors create a strong inference that CSX had a regular practice of predicating its accident reports on first-hand information: First, plaintiffs’ expert, Dr. Berg, testified that it is standard procedure for railroad investigators to interview a locomotive crew following their involvement in an accident. Importantly, Rule 803(6) permits the foundational requirements to be “shown by the testimony of the custodian [of the record] or other qualified witness.” (Emphasis added.) “A qualified witness is not required ... to have personally participated in or observed the creation of the document ..., or know who actually recorded the information .... ” United States v. Franco, 874 F.2d at 1139 (citations and internal quotation marks omitted). Rather, a foundational witness “need only be someone with knowledge of the procedure governing the creation and maintenance of the records sought to be admitted.” United States v. Keplinger, 776 F.2d 678, 693 (7th Cir. 1985), cert. denied, 476 U.S. 1183, 106 S.Ct. 2919, 91 L.Ed.2d 548 (1986).18 Doctor Berg’s testimony regarding accident-investigation procedure provided circumstantial proof that the *649accident diagram was likely based upon information provided by the CSX employees involved in the accident.

The second circumstantial factor supporting admission of the accident diagram is the fact that the CSX personnel engaged in reporting the collision — from the locomotive crews involved in the accident to the employee responsible for preparing the accident diagram — were acting under a duty of accuracy. Railroads are required to accurately record and report information concerning grade-crossing accidents. 49 C.F.R. §§ 225.11, 225.19(b) (1998); W. Va.C.S.R. § 150-8-2 (1984). Federal regulations also mandate that railroads such as CSX adopt and comply with a written “internal control plan” pertaining to the reporting of accidents, and that they disseminate among their employees “[a] policy statement declaring the railroad’s commitment to complete and accurate reporting of all accidents, incidents, injuries ... arising from the operation of the railroad, [and] to full compliance with the letter and spirit of [the Federal Railroad Administration’s] accident reporting regulations.” 49 C.F.R. § 225.33(a)(1) (1998). The legal duty imposed upon CSX effectively creates a duty for its employees to reliably report accidents. Indeed, individual railroad employees are subject to possible criminal and/or civil penalties for causing the railroad to violate the mandated reporting requirements. 49 U.S.C. § 21311(a) (1994); 49 C.F.R. § 225.29 (1998).

A report or other record prepared by an organization19 in routine compliance with state and/or federal law is prima facie sufficient under Rule 803(6), where the duties imposed by such law give rise to an inference that it was a regular practice to base the report or record upon first-hand knowledge. Therefore, we hold that in light of circumstantial evidence demonstrating both a routine practice of obtaining information from railroad personnel involved in accidents, and an organization-wide duty to accurately report such mishaps, there was a sufficient foundation to warrant admission of the accident diagram in its entirety.

During oral argument CSX asserted that the trial court’s ruling was proper in that the statement contained within the diagram was untrustworthy. A trial court is entrusted with considerable discretion to exclude evidence that, although it satisfies the foundational requirements of Rule 803(6), otherwise lacks trustworthiness; however, it is not entirely clear that the court below was attempting to exercise its discretion in this manner. Moreover, a record of a regularly conducted activity that meets the foundational requirements of Rule 803(6) is presumptively trustworthy, and the burden to prove that the proffered evidence was generated under untrustworthy circumstances rests upon the party opposing its admission. See Syl. pt. 4, State v. Fairchild, 171 W.Va. 137, 298 S.E.2d 110 (1982) (“Records made routinely in the regular course of business ... are generally trustworthy and reliable, and therefore ought to be admissible when properly verified.”); cf. Syl. pt. 4, Hess v. Arbogast, 180 W.Va. 319, 376 S.E.2d 333 (1988) (“Under W. Va. R. Evid. 803(8)(C), the contents of a public report or document are ... assumed to be trustworthy, unless the opponent of the report establishes that the report is sufficiently untrustworthy.”).20

The presumption of trustworthiness is particularly robust where, as here, the record is adverse to the party who prepared it. Compare Yates v. Bair Transp., Inc., 249 F.Supp. 681, 690 (S.D.N.Y. 1965) (trustworthiness of medical records enhanced by fact that they *650were adverse to party on whose behalf they were prepared), with Palmer v. Hoffman, 318 U.S. 109, 114-15, 63 S.Ct. 477, 481-82, 87 L.Ed. 645, 650-51 (1943) (report of train accident prepared by the engineer two days after accident and favorable to railroad deemed inadmissible as made in contemplation of litigation).

CSX has not pointed to anything suggesting that its own report was prepared under questionable circumstances. The fact that the diagram in question was adverse to its originator unquestionably provides significant indicia of reliability. Also, the trustworthiness of this record is bolstered by the ultimate purpose for its preparation, which was to satisfy federal reporting requirements. See Lewis v. Baker, 526 F.2d 470, 473-74 (2d Cir. 1975) (fact that railroad was required to file accident report with the ICC gave report “sufficient indicia of trustworthiness” to be admissible as business record). We therefore fail to discern any basis upon which to conclude that the statement regarding the location of the eastbound shifter was untrustworthy.

Finally, CSX argues that any error related to the circuit court’s exclusion of this evidence was harmless in that it was cumulative of other evidence presented at trial. Our review of the record indicates, however, that the only other evidence at trial relating the position of the eastbound locomotive at the time of the accident was the testimony of the engineer, Calvin Bowen. Bowen’s statements on this issue were inconsistent, since at one point he indicated that his locomotive was 300 to 400 feet from the Fifth Street crossing when the accident occurred, while at another he stated that he was “just east of the Second Street crossing” (which can be construed as far away as three blocks from the site of the collision). Given this apparent contradiction in Bowen’s testimony, we find no merit in CSX’s assertion that the statement contained in the diagram was harmlessly cumulative of other evidence.

Thus, based on the importance of the evidence in question to support the testimony of plaintiffs’ expert, as well as to deflect CSX’s assertions regarding the proximity of the second locomotive at the time of the collision, we find that the trial court’s interpretation of Rule 803(6) as requiring the preparer of a record of a regularly conducted activity to have personal knowledge regarding its contents, and the concomitant refusal to admit the statement regarding the location of the eastbound shifter locomotive, was reversible error.21

III.

CONCLUSION

For the reasons stated, the judgment of the Circuit Court of Kanawha County is hereby reversed and remanded for a new trial consistent with this opinion.

Reversed and remanded.

dissenting:

(Filed July 12, 1999)

I must respectfully dissent with the majority’s conclusion that reversible error was committed, first by the trial court’s ruling on the admissibility of a statement contained in a railroad investigative report, and second, by counsel for CSX during closing argument in addressing the effect of joint and several liability.

The most serious and significant error in the majority’s opinion is its groundless determination that the trial court erred in making an evidentiary ruling concerning the admission of a portion of a document. Before proceeding to discuss the substantive problems with the majority position, I must first point out a flaw in the procedural path employed by the majority in its review of this issue. It is axiomatic that evidentiary rulings are reviewed under an abuse of discretion standard. See Syl. Pt. 1, McDougal v. McCammon, 193 W.Va. 229, 455 S.E.2d 788 (1995); see also West v. Wintergreen Partners, Inc., 908 F.2d 968 (4th Cir. 1990) (unpublished opinion) (stating that “determination of whether the proponent has met the foundation requirements of the business rec*651ords exception and whether the circumstances indicate a lack of trustworthiness is left to the sound discretion of the trial judge”). To circumvent the inherent limitations placed on an appellate court with regard to a trial court’s evidentiary ruling, the majority has constructed an argument that de novo review of the evidentiary ruling, as a whole, was permitted based on the trial court’s improper application of the business records hearsay exception. See Gentry v. Mangum, 195 W.Va. 512, 518, 466 S.E.2d 171, 177 (1995) (stating that “[a] party challenging a circuit court’s evidentiary rulings has an onerous burden because a reviewing court gives special deference to the evidentia-ry ruling's of a circuit court”). The majority cites this Court’s decision in State v. Sutphin, 195 W.Va. 551, 466 S.E.2d 402 (1995), to support its preferred approach of reviewing the trial court’s ruling on a de novo basis. Sutphin, however, does not depart from utilizing abuse of discretion as the reviewing standard for evidentiary rulings; it merely recognizes that an appellate court is permitted to review in plenary fashion, the limited aspect of the trial court’s interpretation of an evidentiary rule. Critically, the full review permitted of an evidentiary rule’s interpretation by the trial court does not alter the abuse of discretion standard that controls all evidentiary rulings. This is made clear by Sutphin’s pronouncement, which follows its recognition of plenary review for evidentiary interpretational issues, that “we will not disturb the evidentiary rulings absent an abuse of discretion by the trial court.” Id. at 560, 466 S.E.2d at 411. The majority completely misses this distinction and thus, its decision to completely dispense with the abuse of discretion standard of review is fatally flawed.

Not only is the majority mistaken with regard to the applicable standard of review and how the same is applied, but the majority errs in characterizing the trial court’s ruling on the evidentiary issue as being “grounded upon a legal conclusion regarding the foundational requirements of Rule 803(6).” The majority suggests that the sole basis for the trial court’s ruling was the lack of personal knowledge on the document preparer’s part. A review of the record demonstrates that this representation fails to accurately depict the full grounds for the trial court’s ruling. While the trial court did recognize that the document’s preparer lacked personal knowledge concerning the descriptive information about the location of the eastbound locomotive, the trial court’s stated basis for its ruling was the “third-hand hearsay” problem presented by the information and the inherent lack of reliability presented by such hearsay. The majority disingenuously omits reference to decisive comments made by the trial court concerning the basis for its ruling' on the railroad investigatory report and further omits reference to the inherent discretion that is built into Rule 803(6) of the West Virginia Rules of Evidence (hereinafter “Rule 803(6)”).

Most significantly, the majority circumscribes critical language in Rule 803(6) that is completely determinative of the evidentiary issue presented below. The business records hearsay exception permits the introduction of certain regularly conducted business activities “unless the source of information or the method or circumstances of preparation indicate lack of trustworthiness.” W.Va.R.Evid. 803(6), in part, (emphasis supplied); see also 2 Gregory P. Joseph, Stephen A. Saltzburg and Trial Evidence Comm. of the ABA Sec. of Litigation, Evidence in America, Rule 803 at 43 (1987) (observing that “[p]erhaps the most significant foundational requirement is trustworthiness” and “rule [803(6) ] contemplates that the judge will exclude any document if the source of the information contained in it or the method or circumstances of its preparation indicate a lack of trustworthiness”); see also Munoz v. Strahm Farms, Inc., 69 F.3d 501, 503 (Fed.Cir. 1995) (stating that “[r]eliability is the basis for admitting evidence under the business records exception”). The Second Circuit succinctly recognized in Saks Intern, Inc. v. M/V “Export Champion”, 817 F.2d 1011 (2nd Cir. 1987), that “[t]he principal precondition to admission of documents as business records pursuant to Fed.R.Evid. 803(6) is that the records have sufficient indicia of trustworthiness to be considered reliable.” Id. at 1013.

*652Ignoring the plain language used by the trial court with regard to its conclusion that the information at issue was not trustworthy, the majority states instead that “it is not entirely clear that the court below was attempting to exercise its discretion in this manner The record makes clear that the trial court’s ruling was indeed based on its determination that the handwritten train location details included on the railroad report were not reliable due to their “third-hand hearsay” origin. When pressed, the trial court explained its ruling: “Well, I’ve passed on it as to a lack of reliability. It’s hearsay, only hearsay, and no one knows what the basis for the claims adjustor [’s notation regarding train location was].” How much clearer does the court have to state its rationale? The trial court, as the majority recognizes, likened the redaction, which it required, to that of police reports where hearsay comments are typically removed before the reports are introduced into evidence.1 See United States v. Saunders, 886 F.2d 56, 58 (4th Cir. 1989) (holding that statements made to police officers by third parties under no business duty to report are not admissible under business records exception); Ramrattan v. Burger King Corp., 656 F.Supp. 522 (D.Md. 1987) (upholding trial court’s in limine ruling that portions of police report containing statements of witnesses were inadmissible because witnesses were not acting in regular course of business in making statements); United States v. Pazsint, 703 F.2d 420 (9th Cir. 1983) (ruling that tape-recorded calls made by witnesses to police reporting defendant’s actions were not business records since callers were not under business duty to report to police). The trial judge’s use of the police report analogy demonstrates that the court’s concern was in fact the hearsay problem presented by the descriptive information.

The crux of the evidentiary problem presented by the CSX investigator’s descriptive notation is the fact that such information is clearly hearsay and is being offered to prove the ultimate issue of liability.2 Since the document itself is hearsay, a multiple hearsay problem is presented by the document’s inclusion of information that originated from a third party. While the majority gives lip service to the proposition that Rule 803(6) does not solve the problem of hearsay within hearsay, it utterly fails to apply the very law which it cites concerning multiple hearsay to the facts presented by this ease. See Baxter Healthcare Corp. v. Healthdyne, Inc., 944 F.2d 1573, 1577 (11th Cir. 1991) (stating that “Rule 803(6) does not eliminate a double hearsay problem unless the informant’s statement also conforms to one of the exceptions to the rule against hearsay”), vacated, 956 F.2d 226 (11th Cir. 1992). To illustrate, the majority first cites the evidentiary assumption which underlies the admission of business records as requiring that “[e]ach participant in the chain that created the document — from the initial observer-reporter to the final entrant — must be acting in the course of the regularly conducted business ... The reason underlying the business records exception fails if any of the participants is outside the pattern of regular activity.” 5 Jack B. Weinstein & Margaret A. Berger, Weinstein’s Federal Evidence § 803.11[4] at 803-69 (2d ed. 1999) (emphasis supplied). Continuing its recitation of applicable law, the majority recognizes that “hearsay statements contained in police reports are inadmissible where the declarant has no duty to report” and even cites a case decided by the Tenth Circuit, acknowledging that “statements made to the [police] officer by third parties under no business duty to report may not” be admitted under Rule 803(6). United States v. Snyder, 787 F.2d 1429, 1434 (10th Cir.); cert. denied, 479 U.S. 836, 107 S.Ct. 134, 93 L.Ed.2d 78 (1986). When the facts of this case are applied to these well-established requirements for admission as a business record, it is clear that the necessary elements are not present.

*653Since the railroad employee who prepared the report in issue, Mr. 6.A. Green, was not present at the time of the accident, the only way he could have obtained the descriptive information regarding the location of the eastbound locomotive was from speaking with third parties. Obviously, those third parties were under no business duty to accurately report their observations. Therein lies the problem faced by the trial court — an unknown and consequently inherently unreliable declarant may have provided the critical descriptive information to Mr. Green. Without knowing anything about where that de-clarant was at the time of the accident and additional details which would aid in determining whether the information provided to Mr. Green was trustworthy, the trial court properly determined that the predicate basis for admitting records under the business records exception — reliability—was missing with regard to the descriptive information concerning the location of the eastbound locomotive. Had the trial court determined that the investigative report could not come in at all, then the majority would have properly had a foundational basis for denouncing the trial court’s reliance on lack of personal knowledge as a basis for excluding the documentary evidence. But that is not what occurred in this case. The trial court never made a ruling that the record in issue was not a business record or that it could not come in for lack of a proper foundation. The trial court recognized this railroad investigative report for the business record that it was. But the trial court went a step further, as it is required to do, and determined that a portion of the document could not be deemed reliable, as it was based on “third-hand hearsay.” See Meder v. Everest & Jennings, Inc., 637 F.2d 1182 (8th Cir. 1981) (holding that police report regarding cause of accident was inadmissible as business record because source of information was unknown as was information concerning when or under what circumstances information was obtained from source).

If the record provided an adequate basis for concluding that the unknown supplier of the crucial descriptive information was in fact a railroad employee who in turn was under a duty to accurately report the details of the accident, the analysis employed by the majority would be correct. The majority, however, reaches its conclusion that the source of the critical information was a railroad employee solely by inference. Since Appellant’s expert testified that railroads routinely question railroad employees when investigating accidents, the majority leaps to the conclusion that the source of the descriptive comments was necessarily a railroad employee. The weakness of this inference is further demonstrated by the fact that the majority requires as part of the elements necessary to satisfy Rule 803(6) “that the information reported was transmitted by a person with knowledge, who was acting in the course of a regularly conducted activity.” (emphasis supplied) Since the duty to accurately report, and hence the inference of reliability, arises from information that is prepared “in the course of a regularly conducted activity,” it is critical to the analysis employed by the majority that the source of the descriptive information relevant to the accident had to have been a railroad employee. As McCormick observes, “[i]f any person in the process is not acting in the regular course of the business, then an essential link in the trustworthiness chain fails, just as it does when the person feeding the information does not have firsthand knowledge.” McCormick on Evidence § 290 at 274 (4th ed. 1992). And, yet, the record is utterly devoid of any evidence which supports the majority’s assumption that the supplier of information to Mr. Green was a CSX employee. To proceed to the conclusion, as does the majority, that adequate guarantees of trustworthiness were present merely on the suggestion by Appellant’s expert as to the origin of the information is simply not prudent appellate review. Especially when the issue of whether the declarant was under a duty to accurately report the information was critical, as it is here, to the admission of the information under the provisions of Rule 803(6).

Just as the majority sought to reward the Appellants with another bite at the proverbial “apple” by finding reversible error in the mentioning of the effect of joint and several liability, so too does the majority find reversible error on grounds that routinely and *654properly prohibit a determination of reversible error. See Grogg v.. Missouri Pac. R.R., 841 F.2d 210 (8th Cir. 1988) (upholding trial court’s exclusion of railroad document stating that air hose was broken on date of brake accident on grounds that no evidence was presented that person who provided information recorded in railroad document was acting in regular course of business); City of Cleveland v. Cleveland Elec. Illuminating Co., 538 F.Supp. 1257 (N.D.Ohio 1980) (holding that reports prepared by City listing customers who canceled service and reasons for cancellation were not admissible as business records as customers were not under duty to City and comments were hearsay); Juneau Square Corp. v. First Wisconsin Nat’l Bank, 475 F.Supp. 451 (E.D.Wis. 1979) (upholding trial court’s ruling in antitrust lawsuit on inadmissibility of handwritten notations under Rule 803(6) based on absence of testimony concerning the source of the notations and evidence that such notations were made in course of regularly conducted business); State v. Vance, 633 S.W.2d 442 (Mo.Ct.App. 1982) (holding that trial court correctly refused to admit police accident report under business records exception where officer did not witness accident as such information was inadmissible hearsay); accord Kuhl v. Aetna Cas. & Sur. Co., 51 Md.App. 476, 443 A.2d 996, 1002 (1982), aff'd, 296 Md. 446, 463 A.2d 822 (1983); McCormick, § 324.1 at 368 (recognizing that with regard to police reports of accident investigations, primary statement prepared by police officer is admissible as business or public record, but statements of individuals made to officer, unless they meet another hearsay exception, do not come in under Rule 803(6)). Succinctly stated, the trial court’s ruling on the admission of the document in question was an ordinary evidentiary ruling, which is subject to reversal only for abuse of discretion. See Gentry, 195 W.Va. at 518, 466 S.E.2d at 177. The majority simply decided that it would have ruled differently had it been the trial court and thus, ignored the trial court’s right to make evidentiary rulings subject only to an abuse of discretion standard of review.

I am further concerned by the fact that the majority has used this case to carelessly insert a presumption of trustworthiness into the business records rule. Whereas this Court has previously recognized that business records are generally trustworthy, State v. Fairchild, 171 W.Va. 137, 298 S.E.2d 110 (1982), it is quite another thing to elevate business records to a level of presumed trustworthiness. This is especially of concern when you consider the presumption in light of the majority’s disinclination towards excluding obvious hearsay evidence from a business record — evidence that clearly would not be admissible if it were being offered independently. See Vance, 633 S.W.2d at 444 (discussing fact that police officer could not provide trial testimony concerning hearsay evidence contained in report sought to be admitted as a business record similarly proscribed introduction of hearsay evidence directly from report). The very concerns that Justice Douglas espoused in Palmer v. Hoffman, 318 U.S. 109, 63 S.Ct. 477, 87 L.Ed. 645 (1943), the seminal case which first addressed the issue of whether a statement given by the train’s engineer following a railroad accident could be introduced as a business record under the Federal Records Act,3 have come home to roost. In rejecting the admission of the engineer’s statement in Palmer,4 Justice Douglas expressed concern that “[r]egularity of preparation would become the test rather than the character of the records and their earmarks of reliability.” Id. at 114, 63 S.Ct. 477. Justice Douglas astutely anticipated that the very foundation of the business records exception could be eroded by virtue of merely requiring systematic record-keeping as the predicate basis *655for admission of business records rather than properly focusing on the more critical trustworthiness element:

Any business by installing a regular system for recording and preserving its version of accidents for which it was potentially liable could qualify those reports under the Act. The result would be that the Act would cover any system of recording events or occurrences provided it was ‘regular’ and though it had little or nothing to do with the management or operation of the business as such.... The probability of trustworthiness of records because they were routine reflections of the day to day operations of a business luould be forgotten as the basis of the rule.

318 U.S. at 113-14, 63 S.Ct. 477 (emphasis supplied); see also Bowman v. Kaufman, 387 F.2d 582, 587 (2nd Cir. 1967) (observing that liberal construction of federal business records act “does not mean that any particular business record may be admitted without carefully scrutiny of its reliability for the purpose for which it is offered as evidence”). Apparently, the day rued by Justice Douglas — when regularity of document preparation would supersede concerns for trustworthiness — has already arrived.

While I do not disagree with the majority’s determination that a jury should not be advised as to the effect that joint and several liability will have on its verdict. I must part ways with the majority’s determination that reversal is required in this case based on the closing argument comments. The better approach, and the one that is consistent with this Court’s previous rulings concerning this issue, is to follow the rationale we applied in Valentine v. Wheeling Electric Co., 180 W.Va. 382, 376 S.E.2d 588 (1988). When faced with the analogous issue of whether reversible error was committed by instructing the jury concerning the fact that the defendant electric company could be held responsible for the entire verdict even if other tortfeasors were found to have been negligent, this Court determined that although it was error, such error was harmless because the jury found no negligence as to any of the defendants. Id. at 386, 376 S.E.2d at 592.

Just as the error was deemed harmless in Valentine, any error that occurred in the instant case similarly fell into the “harmless error” category. Unlike the fact finders in Valentine, the jury in this case did return a finding of negligence. The verdict form itself, however, demonstrates that no reversible error occurred as a result of the closing argument discussion of joint and several liability. The substance of defendant CSX’s comments at trial which are presently at issue was that if the railroad was found to be even one percent negligent, it could be held responsible for the entire verdict. Significantly, the verdict form indicates that the jury did in fact assess one percent negligence against the railroad. Thus, it is clear that the closing argument comments made by CSX did not dissuade the jury from assessing negligence against CSX. Thus, the comment, although erroneous, obviously caused no harm to the plaintiffs.

Where the majority goes astray is to assume that because the jury determined that the negligence it assessed against CSX was not the proximate- cause of the accident, this secondary determination of proximate causation was necessarily affected by the joint and several liability comments. Only by proceeding down this analytical path of tying the joint and several liability discussion to the jury’s proximate causation determination could the majority conceivably reach its result-oriented determination that reversible error occurred. But, to conclude, as the majority does, that the discussion of joint and several liability had an impermissibly prejudicial affect on the jury’s determination regarding proximate causation not only requires the suspension of principles of logic, but also demonstrates a clear disdain for the jury’s ability to hear and decide such issues. Furthermore, such conclusion is clearly contradicted by the record in this case.5

Not only did the majority leave logic out of its approach, but rather than addressing the *656jury’s finding of one percent liability against CSX, the majority dispensed with that finding by characterizing it as “meaningless.” As a matter of sound appellate review, jury rulings should not be discarded in such a casual and callous fashion. The far better practice is to accept the jury’s rulings, especially, when as here, the Appellants failed to object to the jury verdict form below and similarly failed to raise the issue presented by the verdict form determinations as an assignment of error in their appeal to this Court. Thus, it is patently absurd for the majority to find reversible error arising from a jury verdict form from which no appeal has been taken.6 The majority simply chose to disregard a valid jury determination for the predilected purpose of finding reversible error.

Finally, I must express my heartfelt concern with regard to the increasing disregard that this Court is showing for well-established principles of appellate review. The ease with which the Court chooses to dispense with both legal precedent and valid jury determinations is both disheartening and alarming. Such an approach can have but one impact, and that is negative. Both the legal profession and the judicial system, as a whole, will suffer immeasurably from the result-oriented brand of justice that is currently being dispensed.