OPINION

BIRD, C.J.

What quantum of evidence must the plaintiff in a products liability action produce initially in order to establish a prima facie case of liability under Barker v. Lull Engineering Co. (1978) 20 Cal.3d 413 [ 143 Cal.Rptr. 225, 573 P.2d 443, 96 A.L.R.3d 1]?

I.

On August 21, 1973, plaintiff, Florence L. Campbell, was injured while riding a bus manufactured by General Motors Corporation (General Motors) and owned and operated by the City and County of San Francisco (City). Following the accident, plaintiff filed suit against both General Motors and the City.

In her complaint, plaintiff alleged that the City had been negligent in its operation of the bus and that this negligence was a proximate cause of her injuries. Specifically, plaintiff claimed that the City bus driver negligently caused the bus to "suddenly lurch, jerk, jolt and abruptly stop, throwing Plaintiff violently from her seat, against interior parts of said [bus] and to the floor. . . ."

As regards General Motors, the complaint alleged that the bus was defective in design in that plaintiff's seat lacked "handrails or guardrails within reasonable proximity" and that this defect was a proximate cause of her injuries. Accordingly, plaintiff claimed that General Motors was strictly liable for the damages she had sustained.

Because plaintiff settled her claim against the City, the trial was limited to the claim against General Motors. Plaintiff supported that claim with her own testimony and with photographs of the bus. The evidence presented during her case-in-chief may be summarized as follows.

Plaintiff was 62 years old at the time of the accident. When she first boarded the bus, she sat in a double seat but later moved to the first forward-facing single seat on the right side of the bus. Attached at shoulder level above the back of each forward-facing seat was a horizontal metal "grab bar." There was no such bar in front of plaintiff's seat because she faced not the rear of another forward-facing seat but rather the side of a lateral-facing double seat. The only handrail directly in front of plaintiff was the metal armrest attached — at about waist level — to the side of this seat. A vertical metal pole, extending to the ceiling, was attached to the aisle end of the "grab bar" behind plaintiff. A similar pole was attached to the "grab bar" of every alternate seat behind the first forward-facing seat, and to the floor midway along the front of the lateral-facing double seats.

Plaintiff knew that the bus would make a right turn from Market Street onto Eighth Street. As the bus entered this intersection, the bus driver "turned very sharply" and "as he turned [the bus] kept gaining speed." Plaintiff was propelled from her seat. She did not have time to put her foot out to brace herself. She did reach with both arms for something to hold on to but "[t]here was nothing there to grab."

Plaintiff was thrown to the floor on the opposite side of the bus, falling on her left hip. When she hit the floor, "it was like an explosion and I couldn't get my breath and I thought I was dying." Plaintiff was hospitalized for approximately 18 days and used crutches for several months thereafter. Plaintiff testified that she had serious continuing medical problems as a result of the accident.

Plaintiff's husband was questioned on the stand about any changes in his wife's physical condition which he had observed since the accident. He stated that plaintiff previously had been able to move without difficulty, but after the accident, quite the opposite was true.

At the conclusion of plaintiff's case-in-chief on the question of liability, General Motors moved for a judgment of nonsuit. (See Code Civ. Proc., § 581c) The company contended that plaintiff had failed to introduce evidence sufficient to establish that the bus was defective in design or that the alleged defect was a proximate cause of plaintiff's injuries.

The trial court granted the motion, dismissed the jury, and entered judgment for General Motors. Plaintiff appeals.

II.

(1) A motion for nonsuit is a procedural device which allows a defendant to challenge the sufficiency of plaintiff's evidence to submit the case to the jury. (See generally, James Hazard, Civil Procedure (2d ed. 1977) § 7.4, p. 236.) Because a grant of the motion serves to take a case from the jury's consideration, courts traditionally have taken a very restrictive view of the circumstances under which nonsuit is proper. The rule is that a trial court may not grant a defendant's motion for nonsuit if plaintiff's evidence would support a jury verdict in plaintiff's favor. ( Ewing v. Cloverleaf Bowl (1978) 20 Cal.3d 389, 395 [ 143 Cal.Rptr. 13, 572 P.2d 1155]; Pike v. Frank G. Hough Co. (1970) 2 Cal.3d 465, 469 [ 85 Cal.Rptr. 629, 467 P.2d 229]; Elmore v. American Motors Corp. (1969) 70 Cal.2d 578, 583 [ 75 Cal.Rptr. 652, 451 P.2d 84].)

(2) In determining whether plaintiff's evidence is sufficient, the court may not weigh the evidence or consider the credibility of witnesses. Instead, the evidence most favorable to plaintiff must be accepted as true and conflicting evidence must be disregarded. The court must give "to the plaintiff['s] evidence all the value to which it is legally entitled, . . . indulging every legitimate inference which may be drawn from the evidence in plaintiff['s] favor. . . ." ( Elmore v. American Motors Corp., supra, 70 Cal.2d at p. 583.)

Thus, the major issue in the present case is whether, if all legitimate inferences favorable to plaintiff are made, the evidence is sufficient to support her claim that her injuries were proximately caused by a design defect in the General Motors bus. (See Greenman v. Yuba Power Products, Inc. (1963) 59 Cal.2d 57, 62 [ 27 Cal.Rptr. 697, 377 P.2d 897; 13 A.L.R.3d 1049].) This question necessarily involves the standards for establishing a prima facie case of design defect set forth by this court in Barker v. Lull Engineering Co., supra, 20 Cal.3d 413.

(3) In Barker, two alternative tests were established for determining whether a product is defectively designed. Under the first test, "a product may be found defective in design if the plaintiff demonstrates that the product failed to perform as safely as an ordinary consumer would expect when used in an intended or reasonably foreseeable manner." ( Id., at p. 429.) This standard reflects a warranty analysis and is based on the theory that when a manufacturer places a product on the market, a representation is impliedly made that the product is safe for the tasks it was designed to accomplish. ( Greenman v. Yuba Power Products, Inc., supra, 59 Cal.2d at p. 64.)

The second Barker test involves an evaluation of the design itself. "[A] product may be found defective in design, even if it satisfies ordinary consumer expectations, if through hindsight the jury determines that the product's design embodies `excessive preventable danger,' or, in other words, if the jury finds that the risk of danger inherent in the challenged design outweighs the benefits of such design." ( Barker, supra, 20 Cal.3d at p. 430.)

The policies underlying the doctrine of strict products liability mandate shifting the burden of proof under the second test. "The allocation of [this] burden is particularly significant in this context inasmuch as this court's product liability decisions . . . have repeatedly emphasized that one of the principal purposes behind the strict product liability doctrine is to relieve an injured plaintiff of many of the onerous evidentiary burdens inherent in a negligence cause of action. Because most of the evidentiary matters which may be relevant to the determination of the adequacy of a product's design under the `risk-benefit' standard — e.g., the feasibility and cost of alternative designs — are similar to issues typically presented in a negligent design case and involve technical matters peculiarly within the knowledge of the manufacturer, we conclude that once the plaintiff makes a prima facie showing that the injury was proximately caused by the product's design, the burden should appropriately shift to the defendant to prove, in light of the relevant factors, that the product is not defective. Moreover, inasmuch as this conclusion flows from our determination that the fundamental public policies embraced in Greenman dictate that a manufacturer who seeks to escape liability for an injury proximately caused by its product's design on a risk-benefit theory should bear the burden of persuading the trier of fact that its product should not be judged defective, the defendant's burden is one affecting the burden of proof, rather than simply the burden of producing evidence. [Citations.]" ( Barker, supra, 20 Cal.3d at pp. 431-432.)

Under the second prong of Barker, the plaintiff must establish a prima facie case of causation. That is, evidence must be adduced which would permit a jury to find that a design feature of the product was a proximate cause of plaintiff's injury. Once plaintiff has made such a showing, the burden of proof shifts to respondent under the second Barker test. Therefore, in this case, if plaintiff's evidence was sufficient to establish a causative link between the design of the bus and the injury she sustained, the trial court erred in granting the motion for nonsuit.

In many products liability cases, the plaintiff points to an alleged malfunction of the product as the cause of injury. (See, e.g., Cronin v. J.B.E. Olson Corp. (1972) 8 Cal.3d 121, 124 [ 104 Cal.Rptr. 433, 501 P.2d 1153]; Greenman v. Yuba Power Products, Inc., supra, 59 Cal. 2d at p. 60.) In other situations, it is the absence of adequate warnings or directions which is asserted to be the cause of the injury. (See, e.g., Cavers v. Cushman Motor Sales, Inc. (1979) 95 Cal.App.3d 338, 342 [ 157 Cal.Rptr. 142]; Midgley v. S.S. Kresge Co. (1976) 55 Cal.App.3d 67, 74 [ 127 Cal.Rptr. 217].) In still other contexts, the plaintiff seeks to establish causation on the basis of the manufacturer's failure to provide a particular safety device. (See, e.g., Henderson v. Harnischfeger Corp. (1974) 12 Cal.3d 663, 667 [ 117 Cal.Rptr. 1, 527 P.2d 353]; Luque v. McLean (1972) 8 Cal.3d 136, 140 [ 104 Cal.Rptr. 443, 501 P.2d 1163]; Haft v. Lone Palm Hotel (1970) 3 Cal.3d 756, 762-763 [ 91 Cal.Rptr. 745, 478 P.2d 465]; McNeil v. Yellow Cab Co. (1978) 85 Cal.App.3d 116, 118 [ 147 Cal.Rptr. 733].)

The present case falls into the final category. (4a) Plaintiff claims that General Motors' failure to place a guardrail or handrail within her reach proximately caused the injury. Therefore, the bus was defective in design because it lacked a particular safety device that would have prevented the accident.

Unless very unusual circumstances exist, this type of claim presents a factual issue which can only be resolved by the trier of fact. "In the ordinary case the question becomes one of what would have happened if [the product had been] otherwise. This is of course incapable of mathematical proof, and a certain element of guesswork is always involved. Proof of the relation of cause and effect can never be more than `the projection of our habit of expecting certain consequents to follow certain antecedents merely because we have observed those sequences on previous occasions.' When a child is drowned in a swimming pool, no one can say with certainty that a lifeguard would have saved him; but the experience of the community is that with guards present people are commonly saved, and this affords a sufficient basis for the conclusion that it is more likely than not that the absence of the guard played a significant part in the drowning. Such questions are peculiarly for the jury. Whether proper construction of a building would have withstood an earthquake, whether reasonable police precautions would have prevented a boy from shooting the plaintiff in the eye with an airgun, whether a broken flange would have made an electric car leave the rails in the absence of excessive speed, whether a collision would have occurred if the defendant had not partially obstructed the highway, and many similar questions, cannot be decided as a matter of law." (Prosser, Proximate Cause in California (1950) 38 Cal.L.Rev. 369, 382-383, fns. omitted, italics added.) (5) The plaintiff in a strict liability action is not required to disprove every possible alternative explanation of the injury in order to have the case submitted to the jury. "It is not incumbent upon a plaintiff to show that an inference in his favor is the only one that may be reasonably drawn from the evidence; he need only show that the material fact to be proved may logically and reasonably be inferred from the circumstantial evidence. ( Estate of Rowley 257 Cal.App.2d 324, 334-335 [ 65 Cal.Rptr. 139], and cases cited therein; Witkin, Cal. Evidence (2d ed.) [§ 1133,] pp. 1050-1051.) . . . The mere fact that other inferences adverse to plaintiff might be drawn does not render the inference favorable to plaintiff too conjectural or speculative for consideration [by the jury]." ( Dimond v. Caterpillar Tractor Co. (1976) 65 Cal.App.3d 173, 182 [ 134 Cal.Rptr. 895]; see also Blank v. Coffin (1942) 20 Cal.2d 457, 460-461 [ 126 P.2d 868] (opn. by Traynor, J.) ["If a jury can reasonably infer from these primary facts that the material fact exists, the party has introduced sufficient evidence to entitle him to have the jury decide the issue."]; Greco v. Bucciconi Engineering Company (W.D.Pa. 1967) 283 F. Supp. 978, 984, affd. (3d Cir. 1969) 407 F.2d 87; Witkin, Cal. Evidence, supra, § 1133, pp. 1050-1051.)

It is particularly appropriate that the jury be allowed to determine the inference to be drawn when the evidence indicates that a safety device, designed to prevent the very injury that occurred, was not present. To take the case from the jury simply because the plaintiff could not prove to a certainty that the device would have prevented the accident would enable the manufacturer to prevail on the basis of its failure to provide the safeguard. (See Haft v. Lone Palm Hotel, supra, 3 Cal.3d 756, 772.) Such a rule would provide a disincentive to improve the safety features of a product and thereby interfere with one of the major policy goals of strict liability. (See Escola v. Coca Cola Bottling Co. (1944) 24 Cal.2d 453, 462 [ 150 P.2d 436] (conc. opn. of Traynor, J.).)

Furthermore, "[w]hether an inference should be drawn may be properly influenced `by a policy which makes the action favored or disfavored.' (Witkin, Cal. Evidence (2d ed.) p. 1051.) `[T]he paramount policy to be promoted by the rule [of strict liability] is the protection of otherwise defenseless victims of manufacturing defects and the spreading throughout society of the cost of compensating them.' ( Price v. Shell Oil Co., supra, 2 Cal.3d 245, 251 [ 85 Cal.Rptr. 178, 466 P.2d 722].) To deny to plaintiff the benefit of the inference of proximate cause would frustrate that policy." ( Dimond v. Caterpillar Tractor Co., supra, 65 Cal.App.3d at p. 183.)

(4b) Under these principles, the evidence introduced by plaintiff in the present case was sufficient to withstand the motion for nonsuit. Plaintiff testified that she was injured when thrown from her seat to the floor on the opposite side of the bus. She further testified that before falling she reached out with both arms for something to hold on to, but nothing was there. Given plaintiff's position at the time the bus turned, a jury could reasonably infer from the evidence that a handrail or guardrail within her reach would have prevented the accident. Although this fact may not be capable of mathematical proof, it is nevertheless a reasonable inference that may be drawn from the evidence.

Dimond, supra, 65 Cal.App.3d 173, presents an analogous situation. There, the plaintiff was discovered unconscious and lying face down four to five feet behind his towmotor, a forklift designed to carry large rolls of paper. A stack of paper rolls was nearby. One roll was lying across plaintiff's shoulders, and another one was on the opposite side of the towmotor. There was a dent in the overhead protective cage of the machine, and the towmotor was in neutral gear with the engine running.

Plaintiff suffered retrograde amnesia so he could not remember the events leading up to his being found sprawled near the towmotor. However, he recalled having read a warning that the overhead cage offered no protection against "heavy or capacity loads falling from any height." He also recalled fearing that a falling object could strike the towmotor's exposed propane tank and cause an explosion. Further, he testified that he always turned off the engine when he dismounted the towmotor. ( Id., at p. 178.)

Plaintiff claimed that the towmotor was defective because of the exposed location of the tank and because the warning failed to provide the operator with any criteria for determining the strength of the overhead cage. His theory of causation was that his fear of injury from one or both of these defects had caused him to flee the towmotor when rolls from the nearby stack began falling on it. ( Id., at p. 179.)

The trial court granted a motion for nonsuit on the ground that the inferences necessary to establish his theory — i.e., that he was in the towmotor when the rolls began to fall and that he fled because of his fear of the machine's defects — were too speculative and conjectural. ( Ibid.) The Court of Appeal reversed. It found that the jury could reasonably infer that plaintiff fled from the towmotor when the rolls began to fall because (1) the location of the machine was consistent with that premise; (2) plaintiff's supervisor had instructed him to move certain rolls of paper located near the scene of the accident; and (3) plaintiff ordinarily turned off the engine when not using the machine. ( Id., at pp. 181-182.)

The Court of Appeal noted that plaintiff could not provide direct evidence of a causal connection between the alleged defects and his injury. However, it stated that "the law recognizes that in a products liability case proof of [defect and proximate cause] by direct evidence is frequently impossible; a plaintiff may, therefore, satisfy his burden of proving defect and causation by circumstantial evidence. ( Elmore v. American Motors Corp., supra, 70 Cal.2d 578, 583-584; Vandermark v. Ford Motor Co. 61 Cal.2d 256, 260. . . .)" ( Dimond, supra, 65 Cal.App.3d at p. 183.) In light of the general policy favoring the victims of defective products and the fact that plaintiff produced all the evidence he could on the issue of causation, the Dimond court found that the case should have been submitted to the jury.

Lewis v. American Hoist Derrick Co. (1971) 20 Cal.App.3d 570 [ 97 Cal.Rptr. 798] is similar to Dimond. In Lewis, a crane boom collapsed and fell when a "pendant," a wire rope with swaged fittings which ran along the side of the boom, snapped. According to the manufacturer's specifications, the pendant should have been one and one-quarter inches in diameter. However, the diameter was only one and one-eighth inches.

The defendant appealed a verdict for plaintiff and claimed that "there [was] no evidence to show a causal relationship between the use of 1 1/8 [inch] rope and the accident, or, put another way, no evidence to show that use of 1 1/4 [inch] wire rope would have prevented the accident." ( Lewis, supra, 20 Cal.App.3d at p. 580.) Thus, there was insufficient evidence as a matter of law to support plaintiff's judgment.

Despite the absence of any direct evidence on the issue of causation, the Court of Appeal rejected defendant's claim. It pointed to testimony that crane operators ordinarily gauge the safety of a particular operation by the crane's "tipping capacity," the point at which the load will cause the crane to tip. As a general rule, a crane will tip before its boom will collapse and fall. When the pendant broke, the crane was stable. This circumstantial evidence supported the inference that use of the improper-sized rope proximately caused the accident. ( Id., at p. 581.)

In the case before this court, there was sufficient evidence from which the jury could infer that the lack of a restraining pole or rail within plaintiff's reach was a proximate cause of the injury. Plaintiff's seat, unlike the other forward-facing seats, did not have a metal "grab bar" at shoulder level in front of it. There was also no vertical pole in front of the seat. As she was being propelled from her seat, plaintiff reached for some object to grab on to, but found none. From this evidence, the jury could reasonably conclude that if a bar or pole had been present, plaintiff would not have been thrown to the other side of the bus and injured.

This question was not one that required expert testimony. "Although courts have not always used the same language, the decisive consideration in determining the [necessity] of expert opinion evidence is whether the subject of inquiry is one of such common knowledge that [persons] of ordinary education could reach a conclusion as intelligently as the witness or whether, on the other hand, the matter is sufficiently beyond common experience that the opinion of an expert [is required]." ( People v. Cole (1956) 47 Cal.2d 99, 103 [ 301 P.2d 854, 56 A.L.R.2d 1435]; see Evid. Code, § 801.)

Thus, in McNeil v. Yellow Cab Co., supra, 85 Cal.App.3d 116, the court held that expert testimony was not required to establish that the absence of a seat belt in the rear seat of a taxicab was a proximate cause of injury. "Plaintiff testified that he was not wearing a seat belt because he could not find any, and that the impact of the collision threw him from the right rear seat of the taxicab to its left front area where his head and arm struck some objects with the result that, presumably among other injuries, he broke an arm. It seems clear, simply as a matter of common sense, that the absence of seat belt restraint under these circumstances proximately caused at least some, if not all, of plaintiff's claimed injuries. The subject of inquiry, namely, the question whether the absence of seat belt restraint under these circumstances constituted a proximate cause of plaintiff's claimed injuries, was one of such common knowledge that persons of ordinary education could reach an intelligent answer." ( Id., at p. 118; see also Sappenfield v. Main St. etc. R.R. Co. (1891) 91 Cal. 48, 60-61 [27 P. 590] [expert testimony not required to determine the safety of a pin in the draw-head of a carriage].)

On the other hand, the court in Truman v. Vargas (1969) 275 Cal.App.2d 976 [ 80 Cal.Rptr. 373], held that expert testimony was required to establish what injuries would have been prevented by a seatbelt where the injured party was in a car that was traveling about 40 miles per hour and was brought to a sudden standstill when it crashed into another vehicle. ( Id., at pp. 982-983.) "Projected forward by the great force caused by the collision what would have been the effects upon his body if he had been wearing the seat belt? The nonexpert could only guess." ( Id., at p. 983.)

Unlike the situation in Truman v. Vargas, expert testimony is not required in the present case to determine whether a rail or bar would have prevented the incident. The jury would not have been required to distinguish arbitrarily among injuries in terms of cause. The need for a "grab bar" or pole to steady oneself when a bus turns a sharp corner is a matter within the common experience of lay jurors. Contrast this with the determination as to whether a seatbelt is helpful when a person is in a vehicle, which is moving at 40 miles per hour, and is involved in a serious collision.

The above analysis makes clear that plaintiff produced sufficient evidence to establish a prima facie case of causation. The evidence justified a conclusion that a design feature of the bus — the absence of a restraining bar or pole — was a substantial factor in causing the injury. Under the second prong of Barker, plaintiff discharged her evidentiary burden. At this point, the burden of proof should have shifted to defendant to offer evidence relevant to a risk-benefit evaluation of the design of the bus. Of course, defendant also could have attempted to refute plaintiff's theory of causation by producing evidence (including expert testimony, if desired) that a bar or pole would not have prevented the accident. However, the trial court's grant of the motion for nonsuit prematurely and erroneously withdrew these issues from the jury.

"Courts have been sensitive to problems faced by consumers or users of defective products in proving defect and proximate cause." ( Dimond v. Caterpillar Tractor Co., supra, 65 Cal.App.3d at p. 183.) Where the plaintiff in a strict liability action introduces evidence that she was injured while using the product in an intended or reasonably foreseeable manner and an attempt to avoid injury was frustrated by the absence of a particular safety device, the jury should be given the opportunity to determine whether the facts necessary to impose liability have been established. This is particularly true where the injury which occurred is precisely the sort which the safety device is designed to prevent.

The policies behind the rule of strict products liability favor jury resolution whenever the evidence can be interpreted to support plaintiff's position. (See Wade, On Product "Design Defects" and Their Actionability (1980) 33 Vand.L.Rev. 551, 573.) In the present case, a jury could have found that the design of the bus was a proximate cause of plaintiff's injury. Therefore, the trial court erred in granting the motion for nonsuit.

Alternatively, if a plaintiff proceeds under the first prong of Barker, in addition to establishing a prima facie case regarding causation, the plaintiff must also produce evidence that the product failed to satisfy ordinary consumer expectations as to safety. (See Barker, supra, 20 Cal.3d at p. 430.)

(6) Here, plaintiff presented sufficient evidence to have the case submitted to the jury on this theory as well. Not only did she testify about the accident (her use of the product), but she also introduced photographic evidence of the design features of the bus. This evidence was sufficient to establish the objective conditions of the product. The other essential aspect of this test involves the jurors' own sense of whether the product meets ordinary expectations as to its safety under the circumstances presented by the evidence. Since public transportation is a matter of common experience, no expert testimony was required to enable the jury to reach a decision on this part of the Barker inquiry.

Indeed, it is difficult to conceive what testimony an "expert" could provide. The thrust of the first Barker test is that the product must meet the safety expectations of the general public as represented by the ordinary consumer, not the industry or a government agency. "[O]ne can hardly imagine what credentials a witness must possess before he can be certified as an expert on the issue of ordinary consumer expectations." (Schwartz, Foreward: Understanding Products Liability (1979) 67 Cal.L.Rev. 435, 480.)

The quantum of proof necessary to establish a prima facie case of design defect under the first prong of Barker cannot be reduced to an easy formula. However, if the product is one within the common experience of ordinary consumers, it is generally sufficient if the plaintiff provides evidence concerning (1) his or her use of the product; (2) the circumstances surrounding the injury; and (3) the objective features of the product which are relevant to an evaluation of its safety. That evidence was provided in this case. (7) (See fn. 7.) Therefore, appellant was entitled to a jury determination concerning whether the bus satisfied ordinary consumer expectations.

III.

The judgment of the trial court is reversed.

Mosk, J., Richardson, J., Kaus, J., Broussard, J., Reynoso, J., and Racanelli, J., concurred.