Opinion

Defendant appealed from a substantial jury verdict awarded against it in this product liability action; plaintiffs cross-appealed from the trial court’s reduction of the compensatory portion of the award. The Court of Appeal overturned the judgment in its entirety and ordered a new trial on the sole ground of juror misconduct. We granted a hearing primarily to clarify (1) under what circumstances juror inattentiveness during trial proceedings will constitute misconduct requiring a new trial, and (2) what type of evidence may be introduced to establish or rebut claims of juror misconduct. Because the Court of Appeal resolved the juror misconduct issue, albeit incorrectly, it did not reach defendant’s remaining assertions of error. As will appear, we conclude that none of defendant’s contentions has merit.

One evening in July 1970, James Hasson, then a 19-year-old college freshman, borrowed his father’s 1966 Lincoln Continental to take some visiting friends on a tour of portions of the Los Angeles area. He drove his friends to the top of Mount Olympus Drive to see the view. As the car descended, its brakes failed. James’ efforts to slow the car by using the emergency brake and by throwing the transmission into reverse proved unavailing, and the vehicle careened down the steep, curving street, eventually crashing into a fountain at the base of the hill. Although the four passengers escaped serious injury, James did not. He suffered a severely fractured skull which caused extensive brain damage and abruptly ended his pursuit of a college education and projected *397medical career. In addition, he has encountered profound psychological problems and total, permanent physical disability.

James and his father filed suit in 1971 against Ford Motor Company (Ford), the manufacturer of the automobile, and against other defendants for damages sustained as a result of the accident. The trial court submitted the case to the jury on strict liability and negligence theories, and the jury returned a verdict of $1,123,840 against Ford. On a prior appeal, we reversed that judgment because the judge erred in failing to instruct the jury on the defense of contributory negligence, although we found the evidence sufficient to support a verdict against Ford. (Hasson v. Ford Motor Co. (1977) 19 Cal.3d 530 [138 Cal.Rptr. 705, 564 P.2d 857, 99 A.L.R.3d 158].)

The case was retried in 1978 with Ford the sole defendant and James’ negligence no longer a significant issue. Again, the plaintiffs relied on theories of strict liability and negligence. They attempted to prove that the accident was the result of brake failure which occurred when during operation of the vehicle the brake fluid heated up enough to cause it to vaporize. Although the fluid in Hasson’s Continental had a boiling point of 555°F when installed at the factory, it had a boiling point of 304° or less when tested after the accident. The reduced boiling point corresponded to a vaporization temperature of only 275°F to 280° F. The reason for the drastic reduction in boiling point—and consequently in the temperature at which brake failure could occur—was that the fluid had a hygroscopic quality; that is, it tended to absorb water vapor. As more moisture was absorbed into the brake fluid, its boiling point became lower. Hasson’s experts testified that Ford was aware of the danger of brake failure due to heat-induced fluid vaporization; they expressed the opinion that Ford should have increased the safety of the brake system by measures such as warning dealers and owners to periodically replace used fluid with new fluid having a higher boiling and vaporization point. Alternatively, plaintiffs’ experts testified that Ford could have installed a dual master cylinder at minimal cost to prevent complete brake failure in the event of fluid vaporization.

The necessity of proving this highly technical theory of liability caused the retrial to be lengthy and complex. It lasted nearly 3 months, required the calling of 50 witnesses, and generated a reporter’s transcript of almost 6,000 pages. Hasson and Ford produced experts who testified in excruciating detail about the design of the brake system installed in 1965 and 1966 Lincoln Continentals, the scientific properties *398of brake fluid, and measures Ford could or should have taken to alleviate the danger of brake failure. Furthermore, there was extensive proof of James’ catastrophic injuries and his years of medical history since the accident.

The jury found Ford to be negligent and strictly liable in tort; it awarded plaintiffs $7,570,719 in compensatory damages and $4,000,000 in punitive damages. After the ensuing judgment, Ford moved for a new trial; it asserted numerous grounds therefor, including several varieties of juror misconduct. The court ruled that the compensatory damages award was excessive and compelled plaintiffs to consent to a reduction of the award to a total of $9,247,719 in order to avoid a new trial. (See Code Civ. Proc., § 662.5, subd. (b).) The other grounds for a new trial were rejected, and judgment was entered for the reduced amount.

I.

Ford mounts a detailed challenge to the sufficiency of the evidence to support each of the findings of the jury, including the existence of negligence or a defect in the brakes on the accident vehicle, causation, and grounds for punitive damages. Ford has skillfully attempted to persuade us that the jury should have accepted its version of the facts. Unfortunately, that effort is largely misdirected.

In 1977, when this court unanimously overturned the first jury verdict against Ford, the majority opinion by Justice Richardson summarized the accepted principles governing appellate review of a jury’s factual determinations: “We do not reweigh the evidence on appeal, but rather determine whether, after resolving all conflicts favorably to the prevailing party [citations], and according prevailing parties the benefit of all reasonable inferences [citation], there is substantial evidence to support the judgment.” (Hasson v. Ford Motor Co., supra, 19 Cal.3d 530, 544.) The evidence, viewed in light of these principles, was found to be amply sufficient “to support a determination that fluid vaporization was a proximate cause of the accident.” {Id., at p. 545.) After a second trial and a second unfavorable jury verdict, Ford’s main argument for reversal is an augmented version of the sufficiency claims we previously rejected. Ford’s prolix briefs summarize virtually all the evidence adduced at trial and point out its strengths and weaknesses. However, that showing is largely irrelevant to the issue on appeal: whether the evidence in plaintiffs’ favor provides a sufficient basis for the jury’s findings. Ford’s elaborate factual presentation is but an attempt to re-*399argue on appeal those factual issues decided adversely to it at the trial level, contrary to established precepts of appellate review. As such, it is doomed to fail.

The primary theory advanced by plaintiffs at trial was that the design of the disc brake system installed on 1966 Lincoln Continental automobiles was defective because it could potentially generate enough heat during normal operations to cause the brake fluid to vaporize, resulting in total loss of braking capability. Ford introduced the disc brake system on the 1965 Lincoln Continentals, the first time that an American automobile manufacturer had offered disc brakes as standard equipment on a domestic model. The new braking system was considered a breakthrough because it was believed to provide more predictable and dependable braking than the drum brakes then in general use. However, it had one important disadvantage: disc brakes tend to generate tremendous amounts of heat during use. Hydraulically operated brakes rely for their effectiveness on the principle that brake fluid is incompressible, so that an application of pressure to the pedal results in an instantaneous transfer of force to all four wheels, actuating the wheel cylinders which press the brake linings against a revolving rotor, slowing the forward progress of the vehicle. The heat produced by friction between the rotor and the lining must be dissipated into the surrounding atmosphere and the other components of the brake system. In the disc brake system, the wheel cylinders are located very close to the point of contact between the lining and rotor. Therefore, the fluid tends to heat up during application of the brakes. When the fluid reaches a certain temperature, it instantaneously vaporizes and becomes compressible, so that the driver is able to depress the brake pedal all the way to the floorboard without encountering any resistance—and without achieving any braking power.

Fluid vaporization is an insidious cause of brake failure: its symptoms disappear and full pedal returns as soon as the fluid cools down by a few degrees. Thus, an inspection conducted shortly after total brake failure might disclose no indication that fluid boil had occurred.

Shortly after releasing its 1965 Lincoln Continentals, Ford began to receive numerous complaints of brake loss attributable to fluid boil. As a result, all the 1965 models were recalled in an attempt to alleviate the problem. The main modifications were the installation of a dust shield designed to increase the flow of air across the brakes and the replacement of the brake fluid with one having a much higher “dry” boiling *400point. The “dry” boiling point is the temperature at which newly installed fluid will come to a boil. During use, brake fluid tends to absorb moisture, lowering its boiling point considerably. For purposes of brake failure, the significant factor is the fluid’s vaporization point, which is somewhat lower than its boiling point. The fluid originally installed on the 1965 Lincoln Continentals had a dry boiling point of 375°F; the replacement fluid had a dry boiling point of 550°F. Unfortunately, the 550°F fluid tended to absorb water vapor at a higher rate; after a few years of use, its actual boiling point was no higher than that of used 375°F fluid.

Ford argues that the jury could not reasonably have found that the disc brake system on the accident vehicle was defective, but the evidence is to the contrary. It was established that the vaporization temperature of the 550°F fluid lowered dangerously in use. The evidence further indicated two possible ways Ford could have alleviated the danger of brake loss: (1) by warning dealers and users that the brake fluid should be periodically replaced with fresh fluid having a higher boiling and vaporization point; and (2) by installing as factory equipment a dual master cylinder or by recalling the cars and retrofitting them with the dual master cylinder.

Periodic replacement of the brake fluid would have substantially reduced the danger of fluid vaporization. The dual master cylinder would have essentially created two separate braking systems, one for the front wheels and one for the rear wheels. In the event that fluid vaporization did occur, the dual master cylinder would enable the alternate system to continue functioning, thus preventing total failure. Ford installed dual master cylinders on its 1967 Lincoln Continentals, indicating that the system was available well before the accident in question occurred.

As an alternative to finding the system to be defective, the jury could have found that Ford was negligent: Ford was aware of the danger of brake failure posed by the disc brake system, yet did not take adequate measures to eliminate the danger.

With respect to the issue of causation, Ford claims the evidence conclusively established that fluid boil could not occur in normal usage. Thus, it reasons, either the brake failure on the accident vehicle had a different cause; or James Hasson abused the brakes by “dragging” them, i.e., driving with his right foot on the accelerator and his left foot *401resting on the brake pedal.1 However, James testified unequivocally that he was not dragging his brakes on the date of the accident. Further, there is ample evidence consistent with the theory that fluid boil caused the accident, even though the car was being operated in a normal manner. Justice Richardson’s analysis of this issue after the first trial remains accurate: “The record included evidence that air temperatures were warm on the day of the accident, which would tend to diminish the cooling effect of ventilation of the brakes. The driving pattern was stop-and-go over hilly terrain, meaning frequent application of the brakes, plus the additional buildup, or soakup of heat which occurs when already warm brakes are allowed to stand momentarily without ventilation. Plaintiffs’ experts pointed to characteristics of disc brakes in general, as well as specific features of the 1966 Lincoln’s brake system design in particular, which they believed would contribute to the buildup of heat under such conditions. Moreover, the symptoms described by the passengers and other witnesses—the apparent sudden, complete loss of pedal pressure (supported by the absence of skid marks) and return of pedal within 45 minutes after the accident (confirmed by investigating officers)—were entirely consistent with plaintiffs’ theory of the accident....

“Ford elected not to dispute much of this evidence, suggesting rather, that the entirety of the evidence, including the results of its own tests, was more consistent with the probability of driver error as the sole cause of the accident.

“The jury, of course, was not compelled to accept Ford’s view simply because more than one inference could reasonably be drawn from the record. So long as the foundation for the opinions of plaintiffs’ experts was sufficient, as we think it was, the jury was entitled to consider those opinions in forming its own conclusions. It was the function of the trier of fact to weigh all the evidence and to draw any reasonable inferences it found warranted.

“We think the inferences here drawn were reasonable. That the evidence might also have supported Ford’s version of the accident is irrele*402vant on appeal. We therefore hold that there was sufficient evidence to support a determination that fluid vaporization was a proximate cause of the accident.” (Hasson v. Ford Motor Co., supra, 19 Cal.3d at p. 545.)

Ford additionally claims that the evidence at trial was not sufficient to support the jury’s punitive damages award. Once again, Ford draws our attention to evidence it deems favorable to its position and asks that we upset the verdict because of the strength of such evidence. As we have previously stated, Ford has a difficult hurdle to overcome: It must convince us of the absence of substantial evidence on which the jury could have based its verdict; a mere conflict of evidence will not suffice. (E.g., Crawford v. Southern Pacific Co. (1935) 3 Cal.2d 427, 429 [45 Cal.Rptr. 183].)

Punitive damages are appropriate if “the defendant has been guilty of oppression, fraud, or malice. ...” (Civ. Code, § 3294. )2 “[A] conscious disregard of the safety of others may constitute malice within the meaning of section 3294 of the Civil Code. In order to justify an award of punitive damages on this basis, the plaintiff must establish that the defendant was aware of the probable dangerous consequences of his conduct, and that he wilfully and deliberately failed to avoid those consequences.” (Taylor v. Superior Court (1979) 24 Cal.3d 890, 895-896 [157 Cal.Rptr. 693, 598 P.2d 854].)

Mindful of the limited scope of appellate review, we now examine plaintiffs’ evidence to determine its sufficiency. Plaintiffs’ showing emphasized heavily the testimony of Harley Copp, a former Ford employee for 30 years who held numerous high level engineering and management positions. Copp testified, inter alia, that although Ford knew of the fluid boil problem with its Continentals from dealer and customer complaints, it deliberately failed to warn dealers or owners of avail*403able remedial steps because it was protecting the Continental’s reputation among consumers. He further testified that Ford deliberately failed to run adequate tests to accurately define the nature of the brake loss problem and deliberately failed to install a dual master cylinder on the 1966 Continental as original equipment or on recall.

At one point, plaintiffs’ counsel directed the following question to Copp: “Was there anything in the owner’s manual to indicate that . .. if there was any kind of a fluid boil, that there would be no brakes at all?” Copp responded: “No. The tags ... on the steering wheel, and in the owner’s manual .. . [don’t] say anything about a potential brake failure.”

When asked: “In your opinion was there a conscious disregard of safety on the part of Ford with respect to not putting a dual master cylinder on the 1966 Lincoln Continental?” He answered: “Yes.”

Dr. John Albert Fellows, a scientist and consultant, testified that Ford management had “adopted a policy of advertising that the Lincoln was free [from] the need of service for at least a good portion of its components ... and that they were opposed to abandoning that policy in public recognition.”

Despite this evidence, Ford now asks us to set aside the jury verdict because of asserted inconsistencies and conflicts in testimony favorable to plaintiffs. The jury, however, was responsible for judging the credibility of the witnesses; it would be wholly improper for us to usurp that function by reweighing the evidence. We hold that substantial evidence supports the award of damages.

II.

Ford argues that the trial court erred in admitting evidence of prerecall brake failures in 1965 models. This contention is easily resolved. At trial, Ford contended that the 1966 brake system was substantially different because of design modification instituted pursuant to the recall campaign and maintained on 1966 models: The 1966 system’s fluid had a higher dry boiling point; Ford also installed a vented dust shield and changed the brake lining. Plaintiffs countered with expert testimony suggesting that the changes were insignificant and, in the case of the vented dust shield, completely ineffective. The trial court plainly had a reasonable basis for admitting evidence of the numerous *404failures occurring in 1965 models for the purpose of showing the nearly identical 1966 models to be similarly defective. Plaintiffs were not required to prove that the 1965 system was exactly the same as the 1966 system. “Identical conditions will rarely be found. Substantial similarity is normally sufficient.” (Jensen v. Southern Pacific Co. (1954) 129 Cal.App.2d 67, 74 [276 P.2d 703].) This determination “is primarily the function of the trial judge.” (Ibid.)

The trial court also admitted into evidence letters sent to Ford and testimony describing incidents of brake failure in 1965 and 1966 Lincoln Continentals. One letter informed Ford that a certain private toll road had been closed to Lincoln Continentals as a result of reports of brake failures occurring with Lincolns using the road. A second letter complained of a brake failure—impliedly due to fluid boil—occurring in a postrecall 1965 Lincoln Continental. Two Continental owners related instances of brake failure. The evidence was offered as proof that Ford had notice that the fluid boil problem persisted after the brake system was modified by the addition of different brake fluid and the vented dust shield. Ford argues that the trial judge abused his discretion by admitting the evidence because the circumstances surrounding the reported brake failures were not similar enough to those surrounding the failure which caused Hasson’s accident. (See Ault v. International Harvester Co. (1974) 13 Cal.3d 113, 121-122 [117 Cal.Rptr. 812, 528 P.2d 1148, 74 A.L.R.3d 986]; Kopfinger v. Grand Central Pub. Market (1964) 60 Cal.2d 852, 861 [37 Cal.Rptr. 65, 389 P.2d 529].)

When evidence is offered to show only that defendant had notice of a dangerous condition, the requirement of similarity of circumstances is relaxed: “‘all that is required ... is that the previous injury should be such as to attract the defendant’s attention to the dangerous situation. . ..’” (Laird v. T. W. Mather Inc. (1958) 51 Cal.2d 210, 220 [331 P.2d 617].)

It does not appear that the evidence was improperly admitted; there were sufficient facts from which the jury could have justifiably inferred that these postrecall failures were the result of fluid boil. All of the incidents were characterized by the sudden loss of all pedal and brake function after a period of continuous hard use. In several of the incidents, the evidence showed that full pedal returned within a brief period after total failure, a clear symptom of fluid boil. Although the trial judge might justifiably have excluded some of the evidence on the ground that *405its potential for prejudice outweighed its probative value (see Evid. Code, § 352), he did not abuse his discretion by admitting it.

III.

Ford raises several assertions of error concerning the trial court’s rulings on requested jury instructions.

Plaintiffs’ theory at trial was that the accident occurred because of a defectively designed brake system which allowed the brake fluid to overheat and vaporize, resulting in a complete loss of braking power. Ford, in contrast, theorized that the accident was caused by a booster hose that was improperly installed by a mechanic when the car was serviced, so that it later became disconnected and caused brake loss. Ford requested and was denied an instruction that the disconnected booster hose was a superseding cause of the accident. (See generally Phillips v. G. L. Truman Excavation Co. (1961) 55 Cal.2d 801, 806 [13 Cal.Rptr. 401, 362 P.2d 33].)

In Self v. General Motors Corp. (1974) 42 Cal.App.3d 1 [116 Cal.Rptr. 575], cited by Ford, the Court of Appeal reversed a verdict for plaintiff because the trial judge incorrectly denied defendant’s request for an instruction on superseding causation. In Self plaintiff’s car burst into flames after being hit from behind. Plaintiff argued that the fire resulted from a design defect, the unsafe location of the fuel tank. General Motors, however, contended that the impact of the collision was so great that even a properly located fuel tank would have caught fire. The Court of Appeal held that it was error not to instruct the jury that the harm caused by the defective tank placement could have been superseded by the sheer force of the impact. Self is factually distinguishable: Here, a disconnected booster hose would not have caused a complete brake loss; plaintiff would have only lost the “power assist” braking capability. He would not have experienced the total brake failure to which he testified: “[T]here was no resistance whatsoever and the brake pedal went straight to the floor.. .. ” Therefore, the hose problem could not have been a superseding cause; it was at most a concurrent cause of the accident, and the jury was instructed on the theory of concurrent causation. The trial court acted correctly in refusing the proffered instruction.

Ford maintains that the trial court erred by giving plaintiff’s nondelegable duty instruction: “The manufacturer of a completed prod*406uct cannot delegate to anyone its duty to have its product delivered to the ultimate user free from dangerous defects.” Ford maintains that the instruction misstates the holding of the case from which it derives. That opinion used the phrase “ultimate purchaser” rather than “ultimate user.” (Vandermark v. Ford Motor Co. (1964) 61 Cal.2d 256, 261 [37 Cal.Rptr. 896, 391 P.2d 168].) Ford also claims the instruction was “thoroughly misleading” (see Henderson v. Harnischfeger Corp. (1974) 12 Cal.3d 663, 670 [117 Cal.Rptr. 1, 527 P.2d 353]) because it suggests that Ford would be liable for defective maintenance. Although the instruction is not a verbatim quotation from Vandermark, it is an accurate statement of the law. In Vandermark, we noted that “‘[A] manufacturer is strictly liable in tort when an article he places on the market ... proves to have a defect that causes injury to a human being.’” (Vandermark, supra, 61 Cal.2d at p. 261, quoting 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 duty runs to all who are injured by a defective product, not just ultimate purchasers. (Elmore v. American Motors Corp. (1969) 70 Cal.2d 578, 586 [75 Cal.Rptr. 652, 451 P.2d 84].) Moreover, the jury was explicitly instructed that Ford was liable only for manufacturing defects that existed when the car left Ford’s possession.

Ford also contends that the trial court incorrectly instructed the jury on the existence of a manufacturing defect because no substantial evidence had been advanced to support the instruction. The only possible manufacturing defect in the particular Lincoln Continental owned by Hasson would have been a defectively installed booster hose. Ford argued at trial that if the hose was improperly connected, faulty maintenance at the dealership was responsible, not incorrect factory installation. Ford’s theory was based on the testimony of the car’s former owner that he “had all new hoses replaced under the hood.” Further, a Ford employee testified that the marks on the booster hose removed from the accident vehicle indicated that it was a replacement hose, not an item of original equipment. On the other hand, there was evidence that the brake booster hose in question was designed to last for the life of the car so that it would not normally be replaced routinely. Therefore, it might reasonably be inferred that, despite the employee’s testimony, the booster hose had not been replaced. Alternatively, the evidence supported the inference that if replacement had occurred, it was necessitated by defective factory installation of the original hose. Although the evidence of a manufacturing defect was not strong, the jury might reasonably have believed plaintiffs’ version of the facts.

*407Ford insists that it was prejudicial error for the court to have instructed the jury that the standards of the Society of Automotive Engineers (SAE) were only “minimal.” A substantial amount of evidence was introduced at trial about government and industry standards for automotive products. The jury was instructed that “[standards concerning component parts of braking systems of automobiles promulgated by the [SAE] are only minimal in nature and do not establish the standard of care for a reasonable manufacturing company under the circumstances of this case.” Ford asserts that it was prejudicial error for the judge to characterize the standards as minimal without any probative facts in evidence on this subject. Ford insists that the jury was invited to erroneously conclude that the SAE did not observe very high standards and, therefore, neither did Ford. But Ford misunderstands the instruction. Objectively viewed, the instruction means only that compliance with industry standards does not always insulate a manufacturer from negligence liability. “[W]hen the manufacturer or supplier knows of, or has reason to know of, greater dangers [despite compliance with regulations] its duty . .. may not be fulfilled.” (Stevens v. Parke Davis & Co. (1973) 9 Cal.3d 51, 65 [107 Cal.Rptr. 45, 507 P.2d 653, 94 A.L.R.3d 1059].)

Ford requested an instruction that custom in the industry “is relevant and ought to be considered, but is not necessarily controlling on the question whether or not [the defendant] exercised ordinary care. ... ” Ford maintains that the trial court erred in refusing the requested instruction. However, Ford was unable to show that any custom or practice had developed regarding industry design, manufacture, or maintenance of disc brake systems. This was largely because the disc brake system was new; Ford was the first American car manufacturer to introduce it as standard equipment. The system was introduced in 1965, one year before plaintiff's car was manufactured. Other American car manufacturers marketed disc brakes in 1965, but only as optional equipment. Ford's reliance on the custom and practice of other manufacturers regarding drum brakes is inapposite because the two systems are fundamentally different. The judge correctly refused to give the instruction.

IV.

Ford asserts that reversal is necessary because of a number of instances of juror misconduct.

*408We may easily dispose of the contention that a retrial is necessary because two jurors concealed bias against Ford when questioned on voir dire. Ford points out that no juror responded when counsel for Ford floated this question to an assembled group of potential jurors: “I believe Mr. Harney [counsel for plaintiffs] asked you if you had been involved in litigation arising out of automobile accidents. Are there any of you who have been involved in lawsuits for any other reason?” One of the jurors present when that question was propounded had been a defendant in several lawsuits brought by large corporate creditors. Another juror remained silent when he was among a group of potential jurors who were asked whether any of them had “dealt with brain injuries”; the juror did not volunteer the fact that his son had died as a result of brain damage sustained in an automobile accident. Not surprisingly, Ford cites no authorities to support its claim that these facts establish misconduct. It is difficult to see how either of these incidents involving failure to affirmatively respond to such generalized inquiries asked of a group of jurors can be thought to amount to concealment of bias. (Cf. Weathers v. Kaiser Foundation Hospitals (1971) 5 Cal.3d 98, 110 [95 Cal.Rptr. 516, 485 P.2d 1132].) Moreover, Ford presented no evidence of actual bias other than the jurors’ silence on voir dire; and the trial court, in denying a new trial on this ground, impliedly determined that there was insufficient proof of concealed bias.3 We see no reason to disturb that finding.

Similarly unpersuasive are Ford’s claims of misconduct due to one juror’s nighttime legal studies during trial and the alleged reading of prejudicial newspaper articles. During the final three weeks of trial, one juror attended night classes in paralegal studies. The subject of one class was the law of products liability. Ford interviewed the lecturer and obtained a declaration stating that he discussed a case in which a jury awarded a large amount of damages to an individual badly injured when the gas tank on his Ford Pinto exploded in flames after a collision. Ford was subjected to punitive damages because, in order to save money, it had consciously decided to abstain from modifying the Pinto in the manner necessary to make it more safe. Plaintiffs’ counsel obtained counterdeclarations from the lecturer and the *409juror in question pointing out that the class was designed for beginners and did not cover in detail the subject of products liability.4

Ford argues that the juror’s paralegal studies amounted to the improper reception of evidence concerning the subject of the trial (see, e.g., Smith v. Covell (1980) 100 Cal.App.3d 947, 952-953 [161 Cal.Rptr. 377]), implying that the juror purposely sought out extrajudicial opinion concerning the issues at trial. We review the record differently: The juror’s decision to undertake paralegal studies during trial appears to have been wholely coincidental. If she intended to solicit improper evidence, she certainly undertook a circuitous route toward that objective. The lecturer’s declaration, viewed objectively, indicates merely that a juror inadvertantly attended a single class where the subject of an arguably related piece of litigation was mentioned in passing. The juror’s actions were not misconduct.

Ford also charges that some of the jurors were exposed to prejudicial newspaper articles which discussed litigation concerning Ford Pinto automobiles. One juror declared that an alternate juror brought in an article about a Pinto accident in which three teenage girls were killed; she further stated that some jurors “read and discussed” the article. The juror also declared: “On another occasion during the trial, I observed that some jurors were reading a newspaper article brought into the jury room by Alternate Juror Rash. They were reading and discussing an article on the lawsuits and accidents concerning the Pinto automobile. During this discussion, Mrs. Davis said that there must be something to Hasson’s case if Ford is paying for all these Pinto accidents.” A second affidavit stated: “During the middle part of the trial, I saw some jurors in the jury room reading and discussing an article in a newspaper concerning the problems with the Pinto gas tank.” The second article discussed a case in which a child orphaned in a Pinto crash received a settlement for $600,000.

Plaintiffs’ counsel solicited contrary declarations. Alternate Juror Rash, the one said to have provided the inflammatory articles, stated that “I did not present to any juror in the Hasson case any newspaper article concerning the Ford Pinto automobile, nor did I engage in any *410discussions or conversations concerning the Ford Pinto automobile.” He also denied discussing any other lawsuits or verdicts against Ford. Eleven jurors, including juror Davis, declared that “I did not see Alternate Juror Rash present or allude to any newspaper article concerning the Ford Pinto automobiles, nor did I hear any discussion concerning the Ford Pinto automobile.” Juror Davis specifically denied making the statement that “there must be something to Hasson’s case. ..

It does not appear that Ford met its burden of establishing misconduct due to the improper reception of evidence. Although the two affidavits it presented constitute a prima facie showing of misconduct, they are directly rebutted in all important respects by a number of counter-declarations. The trial court correctly declined to settle this “battle of the juror declarations” in Ford’s favor by granting a new trial.

In support of its claim of juror misconduct due to inattentiveness at trial, Ford presented three juror declarations stating that one fellow juror was observed reading a novel entitled “A Night in Byzantium” during trial proceedings. Two of the declarations said that this activity took place “while witnesses and evidence were being presented.”5 The declarations did not specify which side was presenting evidence during the novel-reading, nor did they cite specific dates; they stated variously that the juror read the novel “over approximately a one-month period,” “[o]n many occasions,” and “intermittently over a period of many days.” Two of the declarations further noted that certain jurors had worked crossword puzzles at unspecified dates and for unspecified periods of time “while evidence and testimony were being presented.” Four of the identified jurors, however, signed counterdeclarations containing this statement: “I specifically deny that I did not pay attention to the testimony of witnesses and evidence being presented during the trial or that I was reading extraneous material or doing crossword puzzles in any manner or to any extent, whereby I was not able to pay close attention to the testimony of each and every witness and the presentation of all evidence in open court. I specifically state that I did pay attention to all testimony and evidence presented during the trial herein.” The counterdeclaration of a fifth accused juror did not contain the above disclaimer. None of the counterdeclarations denied engaging in the alleged activities during trial; they sought to show only that no activities had diverted their attention from the trial proceedings.

*411Ford contends that the jurors’ activities during trial constitute serious misconduct requiring reversal of the judgment below.6 We agree with the basic premise that a jury’s failure to pay attention to the evidence presented at trial is a form of misconduct which will justify the granting of a new trial if shown to be prejudicial to the losing party. (See Code Civ. Proc., § 657, subd. 2.) The duty to listen carefully during the presentation of evidence at trial is among the most elementary of a juror’s obligations. Each juror should attempt to follow the trial proceedings and to evaluate the strengths and weaknesses of the evidence and arguments adduced by each side so that the jury’s ultimate determinations of the factual issues presented to it may be based on the strongest foundation possible. Were the rule otherwise, litigants could be deprived of the complete, thoughtful consideration of the merits of their cases to which they are constitutionally entitled. (U.S. Const., 6th & 7th Amends.; Cal. Const., art. I, § 16.)

Although implicitly recognizing that juror inattentiveness may constitute misconduct, courts have exhibited an understandable reluctance to overturn jury verdicts on the ground of inattentiveness during trial. In fact, not a single case has been brought to our attention which granted a new trial on that ground. Many of the reported cases involve contradicted allegations that one or more jurors slept through part of a trial. Perhaps recognizing the soporific effect of many trials when viewed from a layman’s perspective, these cases uniformly decline to order a new trial in the absence of convincing proof that the jurors were actually asleep during material portions of the trial. (People v. Lee Yick (1922) 189 Cal. 599, 609-610 [209 P. 538]; People v. Ung Sing (1915) 171 Cal. 83, 88-89 [151 P. 1145]; Callegari v. Maurer (1935) 4 Cal.App.2d 178, 184 [40 P.2d 883]; People v. Roselle (1912) 20 Cal.App. 420, 423-424 [129 P. 477]; State v. Cuevas (Iowa 1979) 281 N.W.2d *412627, 632; State v. Pace (Utah 1974) 527 P.2d 658, 659; Maxwell v. State (1946) 32 Ala.App. 487 [27 So.2d 804, 806]; Powell v. Louisville & N. R. Co. (1916) 172 Ky. 285 [189 S.W. 213, 214-215]; Continental Casualty Co. v. Semple (Ky.App. 1908) 112 S.W. 1122, 1123.)

A number of decisions have considered claims of juror intoxication when presented with evidence that jurors imbibed alcoholic beverages prior to hearing evidence or engaging in delibérations. The decisions have generally rejected claims of misconduct if satisfied that the consumption of liquor was not likely to have affected the indulgent jurors’ capacity to competently perform their duties. (E.g., People v. Leary (1895) 105 Cal. 486, 491-496 [39 P. 24]; People v. Deegan (1881) 88 Cal. 602, 604-607 [26 P. 500]; People v. Manson (1976) 61 Cal.App.3d 102, 215 [132 Cal.Rptr. 265]; but cf. People v. Lee Chuck (1889) 78 Cal. 317, 330-339 [20 P. 719].)

A few other cases have rejected allegations of misconduct based upon the apparently inattentive demeanor of jurors during trial proceedings. In State v. Williams (Mo.App. 1978) 577 S.W.2d 59, 62, a juror was observed reading a newspaper during the giving of testimony. The trial judge had the paper taken away. The appellate court upheld the judge’s decision not to declare a mistrial, noting that the complaining party had shown no demonstrable prejudice. In Ferman v. Estwing Manufacturing Company (1975) 31 Ill.App.3d 229 [334 N.E.2d 171, 174-175], the appellate court overturned an order granting a new trial because a juror had appeared bored and inattentive during the trial. The court held that the party seeking a new trial must affirmatively establish prejudice resulting from juror inattention. Finally, in Wofford v. State (Okla.Crim.App. 1972) 494 P.2d 672, 674-675, the court found no error in the trial judge’s refusal to dismiss a juror who yawned and cleaned his fingernails during the giving of instructions.

Turning to the facts of the present case, it appears that Ford has made a prima facie showing of improper conduct by certain jurors. No evidence contradicted the declarations to the effect that some jurors engaged in distracting activities during the presentation of evidence at trial. It may reasonably be argued that the participating jurors did not at all times devote their full attention to the proceedings before them.

Plaintiffs rely on the counterdeclarations to rebut the inference that some jurors were inattentive during the trial. Those counterdeclarations *413in essence deny that the jurors’ diverting activities prevented them from carefully listening to all the evidence put before them. Ford persuasively responds that Evidence Code section 1150, subdivision (a), renders the counterdeclarations inadmissible. That section provides: “Upon any inquiry as to the validity of a verdict, any otherwise admissible evidence may be received as to statements made, or conduct, conditions, or events occurring, either within or without the jury room, of such a character as is likely to have influenced the verdict improperly. No evidence is admissible to show the effect of such statement, conduct, condition, or event upon a juror either in influencing him to assent to or dissent from the verdict or concerning the mental processes by which it was determined.”

The primary authority interpreting this section is People v. Hutchinson (1969) 71 Cal.2d 342 [78 Cal.Rptr. 196, 455 P.2d 132], in which we declared the rule as follows: “[Section 1150, subdivision (a), draws a] distinction between proof of overt acts, objectively ascertainable, and proof of the subjective reasoning process of the individual juror, which can be neither corroborated nor disproved.. .. ” We noted that Evidence Code section 1150 limits impeachment evidence to “proof of overt conduct, conditions, events, and statements.... This limitation prevents one juror from upsetting a verdict of the whole jury by impugning his own or his fellow jurors’ mental processes or reasons for assent or dissent. The only improper influences that may be proved under section 1150 to impeach a verdict, therefore, are those open to sight, hearing, and the other senses and thus subject to corroboration.” (Id., at pp. 349-350.)

Ford points out that the counterdeclarations relate to the subjective mental processes of the jurors: i.e., whether they were in fact able to pay full attention to the matters before them. Plaintiffs take the position that the counterdeclarations should be admissible to disprove the fact of misconduct. Each of these contentions has some logical support. On the one hand, the counterdeclarations plainly are an attempt to directly prove the subjective state of mind of individual jurors; therefore, they appear to run afoul of the rule of Hutchinson. On the other hand, the counterdeclarations do not relate to the deliberative mental processes of the jurors, but only to the issue whether the jurors physically paid attention to the evidence presented at trial.

In one previous case, we considered direct, subjective proof of the state of mind of a juror in rejecting allegations of misconduct due to *414sleepiness. In People v. Ung Sing, supra, 171 Cal. 83, 88-89, the defendant sought a new trial; he presented an affidavit alleging that one juror was asleep during some testimony. This court upheld the trial judge’s denial of a new trial, relying on the accused juror’s counteraffidavit stating that he was awake and heard all of the testimony.

Plaintiffs also place reliance on People v. Deegan, supra, 88 Cal. 602, 604-607. There the court—citing counteraffidavits of other jurors and persons present in the courtroom who did not perceive the juror to be intoxicated—rejected a claim of misconduct based on the drinking of alcohol by a juror prior to entering the courtroom. Although plaintiffs place substantial reliance on Deegan, we view that case as standing only for the proposition that when objective, circumstantial proof of a juror’s ability to deliberate is offered to show misconduct, that proof may be rebutted by similar objective proof to the contrary. (Cf. People v. Stokes (1894) 103 Cal. 193, 196-197 [37 P. 207].) Ung Sing is the only case cited which allowed direct, subjective proof of a juror’s state of mind; that authority, however, was decided many years before the enactment of section 1150 and our explanatory decision in Hutchinson. It no longer accurately reflects the law in this state.

The rule of Hutchinson serves a number of important policy goals: It excludes unreliable proof of jurors’ thought processes-and thereby preserves the stability of verdicts.7 It deters the harassment of jurors by losing counsel eager to discover defects in the jurors’ attentive and deliberative mental processes. It reduces the risk of postverdict jury tampering. Finally, it assures the privacy of jury deliberations by foreclosing intrusive inquiry into the sanctity of jurors’ thought processes.

We therefore decline to obfuscate the clear line drawn in Hutchinson between proof of objectively ascertainable facts and proof of the subjective mental processes of jurors. The counterdeclarations fall into the latter category and should not have been considered by the trial court in *415ruling on the motion for a new trial.8 The allegations contained in Ford’s declarations therefore remain unrebutted. It must be concluded that by failing to fulfill their duty of attentiveness, the jurors committed misconduct.

This conclusion does not end our discussion, however, because a new trial is required only if it can be established that Ford was somehow prejudiced by the jurors’ inattentiveness. Prejudice exists if it is reasonably probable that a result more favorable to the complaining party would have been achieved in the absence of the misdconduct.

On these facts, there is but the flimsiest evidence of actual prejudice to Ford. Only if we can infer from the bare fact of the jurors’ diverting activities that they had prejudged the outcome of the case and closed their minds to further consideration of the evidence can it be said that actual prejudice occurred. Such an inference of partiality would be patently unwarranted on this record.

*416Nevertheless, Ford urges that we should presume prejudice from the fact of inattentiveness alone. In People v. Honeycutt (1977) 20 Cal.3d 150, 156 [141 Cal.Rptr. 698, 570 P.2d 1050], we stated: “It is well settled that a presumption of prejudice arises from any juror misconduct. ... However, the presumption may be rebutted by proof that no prejudice actually resulted.” (See also People v. Pierce (1979) 24 Cal.3d 199, 205-209 [155 Cal.Rptr. 657, 595 P.2d 91].)9 The presumption of prejudice is an evidentiary aid to those parties who are able to establish serious misconduct of a type likely to have had an effect on the verdict or which deprived the complaining party of thorough consideration of his case, yet who are unable to establish by a preponderance of the evidence that actual prejudice occurred. The law thus recognizes the substantial barrier to proof of prejudice which Evidence Code section 1150 erects, and it seeks to lower that barrier somewhat.

Plaintiffs argue that the presumption of prejudice should not apply in civil cases. It is true that the presumption developed in criminal cases. But regardless of the rule’s origin, civil litigants, like criminal defendants, have a constitutionally protected right to the complete consideration of their case by an impartial panel of jurors. (U.S. Const., 7th Amend.; Cal. Const., art. I, § 16; Andrews v. County of Orange (1982) 130 Cal.App.3d 944, 953 [182 Cal.Rptr. 176].) People v. Honeycutt, supra, 20 Cal.3d 150, 156, footnote 3, relied in part on civil cases applying a rebuttable presumption of prejudice. (See also Smith v. Covell (1980) 100 Cal.App.3d 947, 953-954 [161 Cal.Rptr. 377].) Code of Civil Procedure section 475 does not compel a different result. That section states in pertinent part: “There shall be no presumption *417that error is prejudicial, or that injury was done if error is shown.” We long ago rejected a rigid interpretation of section 475 in San Jose Ranch Co. v. San Jose Land & Water Co. (1899) 126 Cal. 322, 324-325 [58 P. 824], Furthermore, parallel provisions in the California Constitution and the Penal Code have not prevented us from applying the presumption in criminal cases. (See Cal. Const., art. VI, § 13; Pen. Code, §§ 1258, 1404.) No principled distinction can be drawn between civil and criminal cases for purposes of the presumption of prejudice arising from juror misconduct.

However, the presumption is not conclusive; it may be rebutted by an affirmative evidentiary showing that prejudice does not exist or by a reviewing court’s examination of the entire record to determine whether there is a reasonable probability of actual harm to the complaining party resulting from the misconduct. (In re Winchester (1960) 53 Cal.2d 528, 535 [2 Cal.Rptr. 296, 348 P.2d 904]; People v. Phillips (1981) 122 Cal.App.3d 69, 81 [175 Cal.Rptr. 703]; People v. Bullwinkle (1980) 105 Cal.App.3d 82, 91-92 [164 Cal.Rptr. 163]; Smith v. Coveil, supra, 100 Cal.App.3d 947, 953-954; People v. Martinez (1978) 82 Cal.App.3d 1, 20-25 [147 Cal.Rptr. 208],)10 Some of the factors to be considered when determining whether the presumption is rebutted are the strength of the evidence that misconduct occurred, the nature and seriousness of the misconduct, and the probability that actual prejudice may have ensued.

Here the jurors engaged in essentially neutral, albeit distracting, activities at unspecified times during the presentation of evidence. There was overwhelming proof of liability against Ford and no substantial likelihood that actual prejudice may have resulted from the jurors’ activities. It was not clear what type of evidence was being presented while the misconduct occurred or even which side’s case was being presented. In sum, the showing of misconduct is rebutted by an examination of the record which reveals no substantial likelihood that Ford was given anything less than a full and fair consideration of its case by an impartial jury. The instances of misconduct demonstrated here do not *418rise to the level of evidence “of such a character as is likely to have influenced the verdict improperly.” (Evid. Code, § 1150, subd. (a).) The trial court so found in its denial of a motion for new trial.

We take this opportunity to emphasize our unwillingness to allow the impeachment of jury verdicts on a bare showing that some jurors failed to conform their conduct to the ideal standard of utmost diligence in the performance of their duties. Even the most diligent juror may reach the end of his attention span at some point during a trial and allow his mind to wander temporarily from the matter at hand. We do not condone such conduct and trust that trial courts will be alert and take appropriate action if it occurs. But we recognize that this is especially likely to occur in such a complex and lengthy trial as the case at bar. Retrials are to be avoided unless necessitated by a more substantial dereliction of jurors’ duties than was evident in this case. “Society has a manifest interest in avoiding needless retrials: they cause hardship to the litigants, delay the administration of justice, and result in social and economic waste.” (Mercer v. Perez (1968) 68 Cal.2d 104, 113 [65 Cal.Rptr. 315, 436 P.2d 315].) This plaintiff was seriously and permanently injured in 1970. He has prevailed in two lengthy jury trials, but for twelve years has received no recovery. Justice will not be served by a second reversal, yet another lengthy trial, to be followed in all likelihood by further appeals.

V.

Finally, Ford urges us to overturn the jury’s compensatory award on the ground that it is excessive as a matter of law. Plaintiffs’ expert projected the special damages as follows:

Past Medical Expenses $ 70,719.00

Future Medical Expenses $ 333,000.00

Past Attendant Care $ 77,000.00

Future Attendant Care $ 789,000.00

Future Earnings Losses $2,350,000.00

Total $3,619,000.00

The jury ultimately awarded a total of $7,500,000 in compensatory damages; the trial court remitted $1,650,000 of the award; and the compensatory portion of the ultimate judgment was $5,850,000.

*419Of course, we may overturn the award of damages only if the award is excessive as a matter of law or if after reviewing the record favorably to the judgment, we conclude that the award is so grossly disproportionate to the harm suffered as to raise the presumption that it resulted from passion or prejudice. (Bertero v. National General Corp. (1974) 13 Cal.3d 43, 64 [118 Cal.Rptr. 184, 529 P.2d 608, 65 A.L.R. 3d 878]; Schroeder v. Auto Driveway Co. (1974) 11 Cal.3d 908, 919 [114 Cal.Rptr. 622, 523 P.2d 662].) Although the trial court may weigh the evidence and grant a new trial or order a remittitur if it finds the jury’s award to be against the weight of the evidence, we are not so empowered.

Accepting Ford’s assumption that the jury awarded the full amount projected by plaintiffs’ expert and that the remainder of the award was for pain and suffering, there is some arguable merit to Ford’s claim that the jury’s award was excessive. The claims for future medical expenses and future attendant care may be somewhat exaggerated. Further, the claim for future earnings losses is based on the speculative assumption that James Hasson would fulfill his lifelong dream of becoming a medical doctor. The only tangible support for that assumption was the testimony of a college professor that James was “capable” of completing the necessary schooling, but James’ scholastic history made that possibility dubious. At the time of the accident, he had completed only one year of college, earning less than a “B” average. Moreover, his high school grades and Scholastic Aptitude Test scores were unspectacular. On the other hand, it was reasonable to assume that James would have completed college and accordingly had a future earnings capacity with a present value of $868,000 or more. Furthermore, the relevant figure for purposes of reviewing the excessiveness of damages is the total reflected in the postremittitur judgment. The trial court reduced the compensatory award by $1,650,000. In so doing, it brought the total amount of damages within reasonable limits and rendered it nonexcessive.

VI.

Plaintiffs have cross-appealed from the trial court’s order reducing the amount of their compensatory award. The procedural history of the order is somewhat complicated: After the entry of a judgment against it, Ford moved for a new trial on numerous grounds. The court heard defendant’s motion on December 1, 1978, indicating at the conclusion of argument that it intended to grant a conditional new trial on *420the ground of insufficiency of the evidence to support the compensatory award. (Code Civ. Proc., § 662.5.) The new trial was to concern the issue of damages only, and it would be avoided if plaintiffs consented to a reduction of the award by $1,650,000. However, the minute order erroneously stated that a conditional new trial was to be granted “on all issues.” Subsequently, on December 11, plaintiffs’ counsel sought to correct the error by way of a letter to the trial judge which suggested language for a new order conforming to the oral directions given by the judge at the conclusion of the new trial hearing. The judge adopted counsel’s wording verbatim and entered the new order on December 12, nunc pro tunc as of December 1.

Plaintiffs now contend in their cross-appeal that both orders are invalid because neither contains an adequate explanation of the trial judge’s reasons for ordering the conditional new trial. Of course, the requirement of a written specification of reasons for granting a new trial is well established. (Code Civ. Proc., § 657; Mercer v. Perez (1968) 68 Cal.2d 104 [65 Cal.Rptr. 315, 436 P.2d 315]; Scala v. Jerry Witt & Sons, Inc. (1970) 3 Cal.3d 359, 365 [90 Cal.Rptr. 592, 475 P.2d 864]; Stevens v. Parke, Davis & Co. (1973) 9 Cal.3d 51, 59-63 [107 Cal.Rptr. 45, 507 P.2d 653, 94 A.L.R. 4th 1059]; La Manna v. Steward (1975) 13 Cal.3d 413, 417-425 [118 Cal.Rptr. 761, 530 P.2d 1073].) The rule serves the dual purposes of “encquraging careful deliberation by the trial court before ruling on a motion for new trial, and of making a record sufficiently precise to permit meaningful appellate review.” (Scala v. Jerry Witt & Sons, Inc., supra, 3 Cal.3d at p. 363; see also Mercer v. Perez, supra, 68 Cal.2d at pp. 112-113.) The requirement applies equally to grants of conditional new trials. (Neal v. Farmers Ins. Exchange (1978) 21 Cal.3d 910, 930-931 [148 Cal.Rptr. 389, 582 P.2d 980].) It is plain that neither of the minute orders satisfied the requirement of a written specification of reasons. The first referred only to the “insufficiency of the evidence to justify the verdict”; the second granted a new trial “on the ground of excessive damages.” Although the trial court’s remarks upon granting the conditional new trial were more detailed, they obviously were not sufficient because they were oral, not written. (See Stevens v. Parke, Davis & Co., supra, 9 Cal.3d at pp. 62-63.)

Nevertheless, plaintiffs are estopped to complain of the trial court’s error because they participated in its commission. (See Johnson v. Rich (1957) 150 Cal.App.2d 740, 747 [310 P.2d 980].) It would be anomalous to allow plaintiffs to base their appeal solely on the ground of the *421deficiency of an order which their counsel drafted. Under the unusual circumstances of this case—plaintiffs’ counsel, who lost the motion for a new trial, drafted the adverse ruling said to be deficient—the trial court’s order may stand even though it contains no written statement of reasons. Counsel cannot escape the effect of such invited error by pointing out that the trial judge had an opportunity to enter a specification of reasons separately from the order.

The judgment is affirmed in all respects.

Bird, C. J., Newman, J., Broussard, J., Reynoso, J., and Brown (Gerald), J.,* concurred.

I respectfully dissent.

The case presents an important issue involving the integrity of our jury system, namely, whether a verdict may stand despite proof that sitting jurors were permitted, during the presentation of evidence, to read books or work crossword puzzles. Through the sworn declarations of three named jurors (one of whom voted for plaintiff, one of whom voted for defendant, and one of whom did not vote) the following record was established:

Juror A declared that “over approximately one month ... [juror D] was reading ... a book in the jury box while witnesses and evidence were being presented,” and that three other jurors (L, G and V) “over a period of several weeks during the trial were doing crossword puzzles in the jury box while witnesses and evidence were being presented”; juror G declared, “During the course of the trial I observed [juror L] in the jury box, had open written material or material with figures, not having to do with the trial, which she was working on or doing something with while testimony and evidence were being presented. On many occasions during the trial I saw [juror D] reading a book in the jury box while evidence and witnesses were being presented”; juror W declared that “During the course of trial I saw jurors [L, G and V] doing crossword puzzles in the jury box while witnesses and evidence were being presented. I observed that [juror D] while sitting in the jury box during court sessions was reading a book. Her reading continued intermittently over a period of many days.”

*422The foregoing sworn declarations from three of the sitting jurors involved conduct of five of the twelve jurors. One of the jurors charged with having worked the crossword puzzles did not deny that she had done so. The other four, in identical language, denied that “I was reading extraneous material or doing crossword puzzles in any manner or to any extent, whereby I was not able to pay close attention to the testimony.” (Italics added.) The emphasized language is significant, containing an implicit acknowledgement that the misconduct occured. Although the jurors asserted that the misconduct did not prevent them from following the testimony, this claim of extenuation is inadmissible under Evidence Code section 1150, subdivision (a).

The majority has frankly conceded that defendant “has made a prima facie showing of improper conduct by certain jurors. No evidence contradicted the declarations to the effect that some jurors engaged in distracting activities during the presentation of evidence at trial.” (Ante, p. 412.) The majority adds, further, that “It must be concluded that by failing to fulfill their duty of attentiveness, the jurors committed misconduct.” (Ante, p. 415.)

Fully acknowledging this misconduct, however, the majority nonetheless insists that there was “no substantial likelihood that actual prejudice may have resulted from the jurors’ activities. It was not clear what type of evidence was being presented while the misconduct occurred or even which side’s case was being presented.” (Ante, p. 417.) With due respect, I think the majority errs. It does not matter what kind of evidence was being offered or who presented it during these periods of improper inattention. The majority of this court held only five years ago that, whether in a civil or criminal case, “It is well settled that a presumption of prejudice arises from any jury misconduct. In an early case we said: ‘For, when misconduct of jurors is shown, it is presumed to be injurious to defendant, unless the contrary appears .... [1Í] Juror misconduct has occurred in several forms requiring reversal when prejudice is presumed in the absence of evidence to rebut the presumption.’” (People v. Honeycutt (1977) 20 Cal.3d 150, 156 [141 Cal.Rptr. 698, 570 P.2d 1050], italics added.)

The present majority ignores this long established presumption of prejudice by purporting to rebut the presumption because defendant has failed to show actual prejudice! However, as the majority itself has observed, the presumption of prejudice was intended specifically to assist those litigants “who are unable to establish by a preponderance of *423the evidence that actual prejudice occurred.” (Ante, p. 416, italics added.) Thus, the majority casts the burden of showing a “substantial likelihood” of actual prejudice upon the very party whose inability to prove such prejudice created the presumption in its favor. This reasoning cannot be the law and it surely has not been our previous position. For example, quite recently in a criminal context, People v. Pierce (1979) 24 Cal.3d 199 [155 Cal.Rptr. 657, 595 P.2d 91], we said “jury misconduct raises a presumption of prejudice; and unless the prosecution rebuts that presumption by proof that no prejudice actually resulted, the defendant is entitled to a new trial. [Citations.]” (P. 207, italics added.) Similarly, in the case before us when jury misconduct is established, the burden is upon the plaintiff to demonstrate that no prejudice resulted from the misconduct. It is not the task of defendant, who has the benefit of the presumption, to show prejudice.

Nor is the misconduct trivial or inconsequential. A defendant’s right to a fair jury trial in civil litigation is of both federal and state constitutional significance. (Byram v. Superior Court (1977) 74 Cal.App.3d 648, 654 [141 Cal.Rptr. 604]; Clemens v. Regents of University of California (1971) 20 Cal.App.3d 356, 360 [97 Cal.Rptr. 589].) We should not countenance such a complete erosion of a constitutional command.

The fact, of course, if it be a fact, that the evidence against defendant on the issue of liability was, in the majority’s words, “overwhelming,” does not detract one whit from defendant’s right to the jurors’ careful independent evaluation of the damage aspect of the case. There was certainly no “overwhelming proof” of plaintiff’s entitlement to $11,570,719, the amount of the jury’s verdict, which the trial court itself voluntarily reduced. In my view, this is an exceedingly large verdict, and the jurors’ admitted inattention to the flow of the evidence may very well have occurred during the presentation of the damage phase of the case. We do not know. Moreover, my conclusion is not changed by defendant’s inability to identify and match the particular periods of the jurors’ distraction with the specific evidentiary presentation by one party or the other. That, of course, is not a critical point because oral or documentary evidence favorable to a defendant may be received during a plaintiff's presentation, and vice versa.

How, in fairness, is it possible for defendant which did not know of the misconduct, nor did anyone else outside of the jury box apparently, *424to prove that the jury’s inattention injured it, either as to the liability or damage issues in this case? The jury’s misconduct here was real, it was substantial and it is admitted. It is not an answer to say that because no one saw the misconduct, not judge, counsel, bailiffs or anyone else, therefore it must not have occurred. The record beyond doubt establishes that in fact it did occur ánd the majority freely acknowledges that it did. The fact that the jury misconduct may have been surreptitious does not dilute the force of the majority’s conclusion that, “by failing to fulfill their duty of attentiveness, the jurors committed misconduct.” (Ante, p. 415.) This misconduct was pervasive, involving five of the twelve jurors including the “forewoman.” It continued over an extended period of time, variously described as “approximately a one-month period,” or “over a period of several weeks,” or “on many occasions,” or “intermittently over a period of many days.” It occurred “while witnesses and evidence were being presented.” (Italics added.) The misconduct was not the momentary dozing of a single juror in an isolated incident. Rather, it involved almost half the jury in frequent, prolonged, intentional mental activity of a type that was diverting and that required thought and contemplation. I respectfully suggest that there are very few jurors, or anyone else to my knowledge, who can simultaneously read a book or work a crossword puzzle while following attentively the testimony in a courtroom. Such activities, in my opinion, were wholly incompatible with a juror’s duties and, with full respect to my esteemed colleagues, we delude ourselves if we think otherwise. The misconduct poisoned the verdict.

Thus, I am unable to square the degree of admitted jury misconduct in this record with what I have always believed was the sworn duty of a juror to “well and truly try the matter in issue.” (Code Civ. Proc., § 604.) This duty surely entails giving undivided attention to the evidence and court proceedings whether the trial lasts three hours, three weeks or three months. Litigants are entitled to no less.

Accordingly, I concur in the conclusion of the unanimous Court of Appeal herein that “A crossword-puzzle working juror attempting to ascertain the proper word has a closed mind, or at minimum, an interrupted attention span. Similarly a novel-reading juror cannot concentrate on both the flow of the plot and the flow of the testimony. Such inattention implies prejudgment of the case which is misconduct. [1i] Nothing admissible appears in the record herein to rebut the presumption of prejudice which arises from such juror misconduct. The inescap*425able conclusion is that the parties did not have 12 unbiased, impartial jurors.”

Believing that we should not approve as a standard for California litigants the jury conduct in this case, I would reverse the judgment.

The petition of appellant Ford Motor Company for a rehearing was denied November 15, 1982. Newman, J. and Kaus, J., did not participate therein. Richardson, J., was of the opinion that the petition should, be granted.