Plaintiffs, representatives for the decedent Chester W. Skinner, appeal from the Court of Appeals decision affirming the trial court’s grant of the defendant’s motion for summary disposition under MCR 2.116(C)(10). 195 Mich App 664; 491 NW2d 648 (1992). This Court granted plaintiffs’ application for leave to appeal, limited to the following questions:
(1) whether the trial court erroneously determined that there were no genuine issues of material fact and that as a matter of law defendants’ switch could not have caused decedent’s death, thus entitling defendants to a grant of summary disposition, and (2) whether the trial court erred in dismissing plaintiffs’ case without ruling on their claims regarding defendants’ failure to insulate the switch’s handle and defendants’ failure to warn. [443 Mich 884 (1993).]
We find that the plaintiffs failed to offer evidence from which reasonable minds could infer *157that the alleged defect caused the decedent’s death. Because the defendant’s summary disposition motion was properly supported, and the plaintiffs’ evidentiary response was not adequate, we hold that the trial court correctly granted the defendant’s motion. For the reasons stated by the Court of Appeals, we also hold that the trial court properly dismissed plaintiffs’ claims for failure to insulate and failure to warn.
I
We borrow from the Court of Appeals accurate summary of the relevant facts in this case:
Plaintiffs brought this products liability action against defendant, Square D Company, following the death of plaintiffs’ decedent, Chester W. Skinner. Mr. Skinner was electrocuted by his own homemade tumbling machine on which he had installed a switch manufactured by defendant.
[Mr. Skinner] was in the business of cleaning and finishing metal parts. To this end, [he] routinely used a homemade tumbling machine that he had designed and built himself. Essentially, the machine consisted of a large metal drum mounted on a frame. Rough metal parts were placed inside the drum along with a quantity of abrasive detergent. An electric motor then rotated the drum in one direction to wash the parts. After allowing the drum to rotate for a period of time, the operator would reverse the direction of the tumbler and the finished parts would be ejected from the drum.
Because of the way Mr. Skinner had designed the machine, reversing the direction of the drum’s rotation was a dangerous task. The motor that turned the drum was controlled by a switch manufactured by defendant. Mr. Skinner had connected the Square D switch to the motor by using three wires with insulated "alligator clips” on the ends. *158In order to reverse the direction of the machine, the operator was required to disconnect two of the alligator clips from the motor by hand and reverse them. For obvious reasons, it was important for the operator to make sure that the Square D switch was in the off position before disconnecting the wires from the motor.
On February 21, 1986, Mr. Skinner was in his shop, working in the room with the tumbling machines. Mrs. Skinner and two other women, Beulah McBride and Violet Whiting,, were in another room, racking parts. Suddenly, the women heard Mr. Skinner cry out. They ran into the room where Mr. Skinner was, and found him standing with his hands above his head, each hand grasping an alligator clip. Electric current was passing through Mr. Skinner’s body. Aware of what was happening, Mr. Skinner cried out to the women, "don’t touch me!” He then freed his left hand from the alligator clip and reached for the Square D switch. Mr. Skinner threw the switch into the off position, twisted, and fell over dead. [195 Mich App 665-666.]
Plaintiffs claim that the Square D switch was defectively designed because it had a large "phantom zone” that sometimes made the switch appear to be "off” when it was actually "on.” Plaintiffs assert that this defect proximately caused Mr. Skinner’s death.
The defendant responds that, even assuming the switch is defective, the plaintiffs’ evidence does not show that Mr. Skinner was misled by the switch when he was fatally electrocuted.1 Following numerous depositions and written discovery requests, the defendant filed a motion for summary disposi*159tion pursuant to MCR 2.116(0(10),2 effectively claiming that because plaintiffs are unable to establish a genuine issue of causation, it is entitled to judgment as a matter of law.
The trial court granted the defendant’s motion, ruling that the "Plaintiff has produced no evidence to show a specific fact issue to support his theory that the switch somehow caused Plaintiff’s decedent’s death.” The Court of Appeals affirmed, over the dissent of Judge Michael J. Kelly. The majority found that "at no time did plaintiffs advance a plausible theory regarding how the defective switch caused the accident.” 195 Mich App 668. The Court further concluded that there were insufficient facts to proceed on plaintiffs’ separate claims of failure to insulate and failure to warn.3
ii
It is well settled under Michigan law that a prima facie case for products liability requires proof of a causal connection between an established defect and injury. Mulholland v DEC Int'l Corp, 432 Mich 395, 415; 443 NW2d 340 (1989). While the plaintiff bears the burden of proof, the plaintiff is not required to produce evidence that positively eliminates every other potential cause. Rather, the plaintiff’s evidence is sufficient if it *160"establishes a logical sequence of cause and effect, notwithstanding the existence of other plausible theories, although other plausible theories may also have evidentiary support.” Id.
As indicated, the defendant in the instant products liability suit argued that the plaintiffs failed to establish a genuine issue of causation, entitling it to summary judgment.
The Michigan Court Rules provide a precise description of the respective burdens that litigants must bear when a motion for summary judgment is filed pursuant to MCR 2.116(0(10). Specifically, MCR 2.116(G)(4)4 mandates that the party seeking summary judgment must specify the issues for which it claims there is no genuine factual dispute. Provided the moving party’s motion is properly supported, MCR 2.116(G)(4) dictates that the opposing party must then respond with affidavits or other evidentiary materials that show the existence of a genuine issue for trial. If the opposing party does not so respond, the rule provides that "judgment, if appropriate, shall be entered against him or her.” MCR 2.116(G)(4). In a similar fashion, this Court has explained the burden of the nonmovant as follows:
Once a party is challenged as to the existence of the facts upon which he purports to build his case, *161the sum and substance of the summary judgment proceeding is that general allegations and notice pleading are not enough. Matters upon information and belief and alleged common knowledge are not enough. That party must come forward-with at least some evidentiary proof, some statement of specific fact upon which to base his case. If he fails, the motion for summary judgment is properly granted. [Durant v Stahlin, 375 Mich 628, 640; 135 NW2d 392 (1965). Emphasis added.][5]
III
A trial court tests the factual support of a plaintiff’s claim when it rules upon a motion for summary disposition filed under MCR 2.116(C)(10). Lichon v American Universal Ins Co, 435 Mich 408, 414; 459 NW2d 288 (1990). The court must consider the affidavits, pleadings, depositions, admissions, and documentary evidence submitted or filed in the action.6 The court is not permitted to assess credibility, or to determine facts on a motion for summary judgment. Zamler v Smith, 375 Mich 675, 678-679; 135 NW2d 349 (1965). Instead, the court’s task is to review the record evidence, and all reasonable inferences therefrom, and decide whether a genuine issue of any material fact exists to warrant a trial.7
*162To determine if a genuine issue of material fact exists, the test is "whether the kind of record which might be developed, giving the benefit of reasonable doubt to the opposing party, would leave open an issue upon which reasonable minds might differ.” Farm Bureau Mutual Ins Co v Stark, 437 Mich 175, 184-185; 468 NW2d 498 (1991).
Like the trial court’s inquiry, when an appellate court reviews a motion for summary disposition, it makes all legitimate inferences in favor of the nonmoving party. Moll v Abbott Laboratories, 444 Mich 1, 27, n 36; 506 NW2d 816 (1993).
Applying these rules, we find that the defendant fulfilled its burden under MCR 2.116(G)(4) when it effectively indicated in its motion that there was no genuine dispute regarding causation. In addition, we conclude that the plaintiffs failed to satisfy their duty because, for the reasons set forth below, their proofs did not establish a genuine issue of causation.
iv
So as to avoid any possible confusion, we make plain that the specific causation issue before the Court in this case is one of cause in fact, and not legal cause or "proximate cause.” That said, we also take this opportunity to clarify what is typically required to establish cause in fact.
Under Michigan products liability law, as part of its prima facie case, a plaintiff must show that the manufacturer’s negligence was the proximate cause of the plaintiff’s injuries. Brisboy v Fibreboard Corp, 429 Mich 540, 547; 418 NW2d 650 (1988) (emphasis added). We have previously explained that proving proximate cause actually entails proof of two separate elements: (1) cause in *163fact, and (2) legal cause, also known as "proximate cause.” Moning v Alfono, 400 Mich 425, 437; 254 NW2d 759 (1977).
The cause in fact element generally requires showing that "but for” the defendant’s actions, the plaintiff’s injury would not have occurred. Prosser & Keeton, Torts (5th ed), § 41, p 266. On the other hand, legal cause or "proximate cause” normally involves examining the foreseeability of consequences, and whether a defendant should be held legally responsible for such consequences. Moning at 439. See also Charles Reinhart Co v Winiemko, 444 Mich 579, 586, n 13; 513 NW2d 773 (1994). A plaintiff must adequately establish cause in fact in order for legal cause or "proximate cause” to become a relevant issue. We find that the plaintiffs here were unsuccessful in showing a genuine issue of factual causation. Accordingly, we need not and do not address legal cause or "proximate cause” in this case.
Because no one was present in the shop with the decedent immediately before the accident, the plaintiffs had to rely on circumstantial evidence to establish that the alleged defective switch was the cause in fact of the decedent’s death. This Court has repeatedly recognized that plaintiffs may utilize circumstantial proof to show the requisite causal link between a defect and an injury in products liability cases. Mulholland at 415; Holloway v General Motors Corp (On Rehearing), 403 Mich 614, 618; 271 NW2d 777 (1978); Schedlbauer v Chris-Craft Corp, 381 Mich 217, 223; 160 NW2d 889 (1968); Bronson v J L Hudson Co, 376 Mich 98; 135 NW2d 388 (1965).
While plaintiffs may show causation circumstantially, the mere happening of an unwitnessed mishap neither eliminates nor reduces a plaintiff’s duty to effectively demonstrate causation:
*164That there was no eyewitness to the accident does not always prevent the making of a possible issue of fact for the jury. But the burden of establishing proximate cause . . . always rests with the complaining party, and no presumption of it is created by the mere fact of an accident. [Howe v Michigan CR Co, 236 Mich 577, 583-584; 211 NW 111 (1926), cert den 274 US 738 (1927).]
To be adequate, a plaintiff’s circumstantial proof must facilitate reasonable inferences of causation, not mere speculation. In Kaminski v Grand Trunk W R Co, 347 Mich 417, 422; 79 NW2d 899 (1956), this Court highlighted the basic legal distinction between a reasonable inference and impermissible conjecture with regard to causal proof:
"As a theory of causation, a conjecture is simply an explanation consistent with known facts or conditions, but not deducible from them as a reasonable inference. There may be 2 or more plausible explanations as to how an event happened or what produced it; yet, if the evidence is without selective application to any 1 of them, they remain conjectures only. On the other hand, if there is evidence which points to any 1 theory of causation, indicating a logical sequence of cause and effect, then there is a juridical basis for such a determination, notwithstanding the existence of other plausible theories with or without support in the evidence.”
We want to make clear what it means to provide circumstantial evidence that permits a reasonable inference of causation. As Kaminski explains, at a minimum, a causation theory must have some basis in established fact. However, a basis in only slight evidence is not enough. Nor is it sufficient to submit a causation theory that, while factually supported, is, at best, just as possible as another theory. Rather, the plaintiff must present substan*165tial evidence from which a jury may conclude that more likely than not, but for the defendant’s conduct, the plaintiff’s injuries would not have occurred.8
We have consistently applied this threshold evidentiary standard to a plaintiff’s proof of factual causation in negligence cases:9
"The plaintiff must introduce evidence which affords a reasonable basis for the conclusion that it is more likely than not that the conduct of the defendant was a cause in fact of the result. A mere possibility of such causation is not enough; and when the matter remains one of pure speculation or conjecture, or the probabilities are at best evenly balanced, it becomes the duty of the court to direct a verdict for the defendant.” [Mulholland at 416, n 18, quoting Prosser & Keeton, Torts (5th ed), § 41, p 269.]
The mere possibility that a defendant’s negligence may have been the cause, either theoretical or conjectural, of an accident is not sufficient to establish a causal link between the two. [Jordan v *166Whiting Corp, 396 Mich 145, 151; 240 NW2d 468 (1976).]
There must be substantial evidence which forms a reasonable basis for the inference of negligence. There must be more than a mere possibility that unreasonable conduct of the defendant caused the injury. We cannot permit the jury to guess .... [Daigneau v Young, 349 Mich 632, 636; 85 NW2d 88 (1957). Citations omitted.]
Something more should be offered the jury than a situation which, by ingenious interpretation, suggests the mere possibility of defendant’s negligence being the cause of the injury. [Howe at 584.][10]
In light of the above pronouncements, we concur with the observation made in 57A Am Jur 2d, Negligence, § 461, p 442:
All that is necessary is that the proof amount to a reasonable likelihood of probability rather than a possibility. The evidence need not negate all other possible causes, but such evidence must exclude other reasonable hypotheses with a fair amount of certainty. Absolute certainty cannot be achieved in proving negligence circumstantially; but such proof may satisfy where the chain of circumstances leads to a conclusion which is more probable than any other hypothesis reflected by the evidence. However, if such evidence lends equal support to inconsistent conclusions or is *167equally consistent with contradictory hypotheses, negligence is not established.
Our case law effectively illustrates the degree of circumstantial proof required to enable jurors to reasonably infer the existence of a causal relationship between a defendant’s actions and a plaintiff’s injuries. For example, in Schedlbauer, a plaintiff boat owner suffered injuries when his pleasure boat exploded. The plaintiff filed suit against the boat manufacturer, claiming that a defectively porous fuel pump caused gasoline to leak into the boat’s bilge area, ultimately resulting in the explosion. The controlling question was whether the plaintiff could offer enough evidence to establish that the gasoline entered the bilge area via the allegedly defective fuel pump, thereby producing the blast.
The plaintiff relied upon circumstantial proof to verify his causation theory. The proofs included: the plaintiff’s testimony that, while the boat had operated without incident in the past, immediately before the explosion, the engine started to "run 'rough’ ”;11 the plaintiff’s expert testimony that if there had been a leak in the fuel pump, the engine would not turn off, but it would " 'run roughly’ ”;12 the expert’s opinion that the gasoline entered the bilge area through the fuel pump and caused the explosion; the expert’s view that if gasoline had entered the bilge area in any other way, the engine would have necessarily shut off automatically during the explosion; and the plaintiff’s testimony that during the explosion the motor continued to run until he manually turned it off.
In particular, this Court emphasized that the expert’s testimony "fairly indicate[d] a logical *168sequence of cause and effect’ ”13 between the alleged defect and resulting injury. The Court also acknowledged the defendant’s alternative causation theories, but found them to be less than probable. On the basis of the plaintiff’s circumstantial evidence, the Court concluded that a jury would have sufficient ground to infer that the defendant’s defective fuel pump caused the explosion and, in turn, the plaintiff’s injuries.
Kaminski also highlights the level of circumstantial evidence needed to adequately establish causation. A plaintiff factory worker incurred injuries when he was hit by a trailer that was parked near a railroad track. The plaintiff filed suit against the defendant railroad company, claiming that the conductor’s negligent failure to see the trailer caused the train to push the trailer into the plaintiff, injuring him. The collision occurred at night, and there were no eyewitnesses to the accident.
The plaintiff’s circumstantial proof showed that moments before the accident the plaintiff had stopped within six feet of the train track, the dark green trailer was parked near the track on the same side that the plaintiff had stopped, and the train was approaching the area where both the plaintiff and trailer were located. Nearer to the time of the collision, the evidence indicated that the train conductor heard a screeching noise coming from the side of the train where both the plaintiff and trailer were situated. At the time of the collision, the plaintiff, also having heard the same screeching noise, was suddenly struck by something he could not see. Immediately after the collision, the plaintiff found himself lying on top of the trailer. The evidence also indicated that, other *169than the moving train, there were no other objects in the area that could have pushed the trailer into the plaintiff.
The Court found that these circumstantial proofs could facilitate a jury in inferring a logical causal relationship between the defendant’s negligence and the plaintiff’s injuries. The Court mentioned the defendant’s contrary causation theories, but in the end dismissed them, explaining that while they were "possible,” they were "not probable.”
In contrast to these cases, the circumstantial evidence offered in Jordan did not afford jurors an opportunity to deduce a causal link between a defendant’s actions and an injury. In Jordan, the plaintiff’s decedent was electrocuted while repairing an overhead crane. The plaintiff brought suit against the manufacturer of the crane, asserting that the manufacturer had negligently failed to ground the crane’s electrical system, and that this negligence caused the decedent’s death. There were no eyewitnesses to the accident.
The plaintiff was unable to offer any evidence establishing where the decedent was located on the crane or what he may have been touching when he was electrocuted. The sole evidence presented consisted of testimony from the plaintiff’s expert regarding purely hypothetical situations. The expert was unable to give an opinion regarding whether the lack of grounding had any causal correlation with the decedent’s particular electrocution. On the basis of the lack of evidence, the Court concluded that plaintiff failed to adequately establish causation.
In Howe, the circumstantial causal proof was equally deficient. A brakeman for a railroad company was getting off a waycar in the dark when he fell to his death from the bridge on which the car *170had stopped. The brakeman’s representative brought suit against the railroad company, claiming that the railroad’s negligent failure to provide more space between the track and the edge of the bridge resulted in the decedent’s death. The misadventure occurred on a dark and rainy night, and no one observed the actual fall.
The circumstantial evidence established that the train had stopped on the bridge, and, pursuant to his duties, the decedent prepared to get off the train to set up flagman signals. The decedent was last seen buttoning his coat by the rear door of the car with two lanterns at his feet. The decedent’s representative argued that it was reasonable to infer that the decedent thereafter alighted from the train, and that the insufficient landing space caused the decedent to fall off the bridge. The Court regarded this causation theory as unlikely without additional facts and circumstances for support, reasoning:
This is one possible explanation of the manner decedent came to his death. It is by no means the only one. No eye saw him after he left the car. No one even knows from which side of the car he left. Is it not just as possible that he stumbled or slipped from the platform or steps of the car and fell into the river? If he did, the space afforded him for walking between the car and the edge of the bridge had nothing to do with it. One theory is as reasonable as the other. Additional ones might be and have been advanced, but the jury should not be permitted to conjecture that he fell from one cause and not from another. [Id. at 583-584.]
V
Unlike the circumstantial proofs submitted in Sehedlbauer and Kaminski, and more like those *171offered in Jordan and Howe, the plaintiffs’ circumstantial evidence here did not afford a reliable basis from which reasonable minds could infer that more probably than not, but for the defect in the Square D switch, Mr. Skinner would not have been electrocuted.
Plaintiffs’ causation theory is that the faulty switch caused the decedent to be confused about whether the machine was on or off; that this confusion induced him to touch the live wires; and that this contact resulted in his electrocution. To support this theory, the plaintiffs offered testimony from co-workers who stated that, in the past, the decedent was always a careful worker, and always turned the power off before he manually reversed the alligator clips. On the basis of this testimony, the plaintiffs urge that the switch must have confused the decedent regarding whether the machine was on or off on the day that he touched the wires and that this confusion led to his electrocution. The lower courts explicitly exposed the critical problem with this theory.
If the machine had been operating, the decedent could not have been confused by the switch in the way that the plaintiffs contend. The plaintiffs do not dispute that when the tumbler is running, it makes a considerable amount of noise, and that the drum moves in a circular motion. The decedent could not have been confused regarding the power’s status because either the noise or the visual appearance of movement would have affirmatively cued him regarding whether the power was on or off — regardless of what the switch may have indicated.
In view of this factually established phenomenon, the plaintiffs had to offer an alternative sequence of cause and effect, whereby the decedent was confused by the switch, but the machine was *172not in operation. The plaintiffs proposed an hypothesis under which Mr. Skinner thought that the machine was off because of a faulty indication from the switch, the power was actually on, but the machine was not running because the wires were not attached to the alligator clips. Under this hypothetical situation, the following antecedent events would necessarily had to have transpired: the machine had to have been turned off, the wires then unhooked, and, the machine thereafter turned back on. This scenario is fatally flawed in two respects.
First, the plaintiffs failed to produce any evidence from which a jury could reasonably conclude that the wires were unhooked when Mr. Skinner began using the machine just before the accident. In fact, the only record evidence pertaining to how the wires and clips were maintained indicated that they would probably have been connected.14 Second, the plaintiffs did not offer any proof from which it rationally could be inferred how the machine would have been turned back on after the wires had been unhooked.15
Of course, the plaintiffs’ offered scenario is a possibility. However, so are countless others. As explained above, causation theories that are mere possibilities or, at most, equally as probable as *173other theories do not justify denying defendant’s motion for summary judgment.16
Plaintiffs further argue that the testimony of their experts proved factual causation. In Jordan, this Court taught that expert opinion based upon only hypothetical situations is not enough to demonstrate a legitimate causal connection between a defect and injury.17 Moreover, in Mulholland at 411, we stated that "there must be facts in evidence to support the opinion testimony of an expert.”18
Plaintiffs’ expert testimony did not sufficiently establish causation. Plaintiffs’ experts maintained that the switch was defective, and that the defect was the proximate cause of the decedent’s death. *174The experts’ causation theories were deficient, however, because each lacked a basis in established fact. Specifically, each expert either assumed, or was asked to assume, that (somehow) the wires were unhooked, and that the power was on when Mr. Skinner began working on the machine. Because the experts’ conclusions regarding causation are premised on mere suppositions, they did not establish an authentic issue of causation.
Taking the record evidence in the light most favorable to the plaintiff, we conclude that the record facts do not manifest a genuine issue of factual causation. The offered evidence only established that an accident took place. Michigan law does not permit us to infer causation simply because a tragedy occurred in the vicinity of a defective product. The plaintiffs were required to set forth specific facts that would support a reasonable inference of a logical sequence of cause and effect. Instead, the plaintiffs posited a causation theory premised on mere conjecture and possibilities.19
We recognize that motions for summary judgment implicate considerations of the jury’s role to decide questions of material fact. At the same time, however, litigants do not have any right to submit an evidentiary record to the jury that would allow the jury to do nothing more than guess.
We are also aware that courts should exercise restraint when deciding to dismiss a suit for lack of evidence. Holloway at 622.20 Nonetheless, in this *175case we are confident that the plaintiffs were unable to establish an essential element of proof. The decedent’s tragic death remains a matter of speculation.21
Because the plaintiff did not show a genuine issue of material fact for trial as required under MCR 2.116(G)(4), we find that the lower courts did not err in granting the defendant’s motion for summary judgment.
VI
For the reasons stated by the Court of Appeals, we also find that the trial court did not err in granting the defendant’s motion for summary judgment on the plaintiffs’ claims of failure to warn and failure to insulate.
Accordingly, we affirm the judgment of the Court of Appeals.
Brickley, Boyle, Riley, and Mallett, JJ., concurred with Cavanagh, C.J.The defendant does not concede that the switch is defective. Rather, for purposes of this litigation, the defendant deems the alleged defectiveness to be an irrelevant issue in view of plaintiffs’ inability to show causation.
MCR 2.116(0(10) provides:
(C) Grounds. The motion [for summary disposition] may be based on one or more of these grounds, and must specify the grounds on which it is based:
(10) Except as to the amount of damages, there is no genuine issue as to any material fact, and the moving party is entitled to judgment or partial judgment as a matter of law.
Judge Kelly did not express any disagreement with the majority’s disposition of these latter two issues.
MCR 2.116(G)(4) provides:
(4) A motion under subrule (C)(10) must specifically identify the issues as to which the moving party believes there is no genuine issue as to any material fact. When a motion under subrule (C)(10) is made and supported as provided in this rule, an adverse party may not rest upon the mere allegations or denials of his or her pleading, but must, by affidavits or as otherwise provided in this rule, set forth specific facts showing that there is a genuine issue for trial. If the adverse party does not so respond, judgment, if appropriate, shall be entered against him or her.
While there was no majority opinion in Durant, we recently reaffirmed this principle in McCart v Thompson, Inc, 437 Mich 109, 115; 469 NW2d 284 (1991) (" 'When properly challenged, plaintiff must establish that he has a case on the law and that there are some evidentiary proofs to support his allegations as to any material fact’ ”) (emphasis added).
Specifically, MCR 2.116(G)(5) requires that the trial court examine "[t]he affidavits, together with the pleadings, depositions, admissions, and documentary evidence . . . submitted by the parties” when reviewing a motion filed under MCR 2.116(0(10).
As we noted in Durant at 646-647, "there is a great difference between an inquiry to determine whether or not there is an issue of fact and a trial to decide a disputed issue of fact.”
However, where several factors combine to produce an injury, and where any one of them, operating alone, would have been sufficient to cause the harm, a plaintiff may establish factual causation by showing that the defendant’s actions, more likely than not, were a "substantial factor” in producing a plaintiff’s injuries. Brisboy at 547. See also Prosser & Keeton, Torts (5th ed), § 41, p 266.
While some of these cases involve a motion for a directed verdict, the test regarding the sufficiency of causal proof is essentially the same in the contexts of summary judgments and directed verdicts, namely, whether reasonable minds, taking the evidence in a light most favorable to the nonmovant, could reach different conclusions regarding a material fact. Accord Anderson v Liberty Lobby, Inc, 477 US 242; 106 S Ct 2505; 91 L Ed 2d 202 (1986) ("'The primary difference between the two motions is procedural; summary judgment motions are usually made before trial and decided on documentary evidence, while directed verdict motions are made at trial and decided on the evidence that has been admitted.’ In essence, though, the inquiry under each is the same: whether the evidence presents a sufficient disagreement to require submission to a jury or whether it is so one-sided that one party must prevail as a matter of law”) id. at 251-252 (citations omitted).
The United States Supreme Court, in addressing summary judgment motions under the parallel PR Civ P 56(c), has also explicitly recognized that the nonmoving party must offer evidence of a character which demonstrates a need for trial:
If the evidence is merely colorable, or is not significantly probative, summary judgment may be granted. [Anderson, n 9 supra at 249-250. Citations omitted.]
Schedlbauer at 228.
Id. at 226.
Id. at 224.
In his deposition, Carl Jacobs, a co-worker of the decedent, described the normal status of the wires:
Q. So, when you would come in in the morning, to start your shift, how would you go about starting up everything?
A. Well, usually what I have to do is go to the off and on box and I have to turn it on, and I have to — when I turned it on, the wires are automatically hooked up, and I would see which way the barrel is running .... [Emphasis added.]
Plaintiffs claim that the switch was defective because it gave the appearance that the power was off when it was really on. They have never suggested that the switch was flawed because it sometimes made the power automatically go back on after it had been turned off.
We believe the Court of Appeals dissent was incorrect with regard to two presumptions — one in fact, the other in law. First, it assumed that there was evidence that the wires were unhooked, and, second, it assumed that a possible theory of causation was sufficient to withstand a motion for summary judgment:
[I]t was clearly established by experts for both parties that the power could be on with the switch in an ambiguous position, the tumbler not turning, and the motor not running, because the reversible wires were disconnected. ... It does not take a journeyman electrician, let alone a distinguished professor, to make sense of that possible occurrence. [195 Mich App 672-673. Emphasis added.]
In Jordan, because the experts’ testimony did not corroborate with any established facts, we ruled that a directed verdict was appropriate:
The sole evidence presented consists of expert testimony that a hypothetical person making simultaneous contact with a live conductor and metal would have a ”70 or 80 per cent chance of extracting himself from that connection” if the crane were grounded. The mere possibility that a defendant’s negligence may have been the cause, either theoretical or conjectural, of an accident is not sufficient to establish a causal link between the two. [Jordan at 150-151. Citation omitted.]
The expert in Mulholland had sufficient evidentiary support for his opinion, which, in particular, was premised on the expert’s firsthand perceptions of the incidents in question.
Moreover, plaintiffs’ attempts to refute the defendant’s offered theories of causation have no effect. Under MCR 2.116(G)(4), the plaintiffs must come forward with a set of evidentiary facts establishing their own genuine theory of causation.
"It is true that where an injury occurs that cannot be accounted for and where the occasion of it rests wholly in *175conjecture, the case may fail for want of proof. . . . But such cases are rare, and that rule should never be so extended as to result in a failure of justice, or in denying an injured person a right of action where there is room for balancing the probabilities, and for drawing reasonable inferences better supported upon one side than upon the other.” [Id., quoting Schoepper v Hancock Chemical Co, 113 Mich 582, 586; 71 NW 1081 (1897).]
Language from our decision in Howe seems particularly fitting in this regard:
"I cannot see in any of this evidence anything which would enable a jury to do more than speculate; that would enable a jury to fall back upon any proof as to what caused the accident. ... It impresses me as being one of the unfortunate, but inexplicable happenings, which is not susceptible of proof, and in the absence of proof it seems to me that the plaintiff’s action must fail.” [Howe at 585. Citations omitted.]