Plaintiffs are two individuals who purchased Marlboro Light cigarettes in Oregon. Defendant Philip Morris is the company that manufactures, markets, and sells Marlboro Lights. Plaintiffs brought this action under Oregon’s Unlawful Trade Practices Act (UTPA),1 alleging that defendant misrepresented that Marlboro Lights would deliver less tar and nicotine than regular Marlboros and that, as a result of that misrepresentation, plaintiffs suffered economic losses. Plaintiffs did not bring the action to remedy only their own claimed losses, however. Rather, they moved to certify a class consisting of approximately 100,000 individuals who had purchased at least one pack of Marlboro Lights in Oregon over a 30-year period — from 1971 to 2001. The trial court denied plaintiffs’ motion after concluding that individual inquiries so predominated over common ones that a class action was not a superior means to adjudicate the putative class’s UTPA claim.
On appeal, in a divided en banc decision, a majority of the Court of Appeals disagreed with the trial court’s predominance assessment, concluding that the essential elements of the UTPA claim could be proved through evidence common to the class. Pearson v. Philip Morris, Inc., 257 Or App 106, 172, 306 P3d 665 (2013). The majority remanded to the trial court to reconsider whether, without the trial court’s predominance assessment, a class action was a superior means of litigating the class claims. Id. We allowed defendant’s petition for review. On review, the parties’ arguments frame several issues for our resolution, including the appropriate standards for determining whether common issues predominate for purposes of the class action certification decision, and what a private plaintiff in a UTPA case of this nature must prove.2 As we will explain, we conclude *91that the trial court properly denied class certification, and accordingly, we reverse the contrary decision of the Court of Appeals and remand to the trial court for further proceedings on the individual plaintiffs’ claims.3
I. BACKGROUND
A. Development and Labeling of Marlboro Lights
In the 1950s, governmental and health organizations began to publicize information about the link between lung disease and tar and nicotine in cigarette smoke, which in turn gave rise to increasing concerns among the public about the dangers of smoking cigarettes.4 In an effort to capitalize on those growing health concerns, cigarette manufacturers introduced new varieties of cigarettes that they advertised as delivering lower levels of tar and nicotine. Although the public health community generally supported the idea of offering smokers low tar and nicotine alternatives, no accepted or approved method for measuring the tar and nicotine yields of cigarettes existed. Thus, “low” and “lower” tar and nicotine claims by cigarette manufacturers could not be substantiated. The Federal Trade Commission (FTC), which regulates the cigarette manufacturing industry, *92therefore initially prohibited cigarette manufacturers from marketing their cigarettes based on low tar and nicotine claims.
Eventually, however, the FTC devised a standardized method for measuring tar and nicotine yields of cigarettes. The “FTC method” used a machine that captured and analyzed substances that were drawn into the machine as it “smoked” a cigarette. The machine regulated variables such as the placement of the cigarette in the machine, the volume of each “puff,” the frequency of puffs, and the portion of the cigarette smoked. In 1967, the FTC instructed cigarette manufacturers that they could represent their cigarettes as having lower tar than regulars if, and only if, the cigarette had a tar yield of 15 milligrams or less as measured by the FTC method.
The lowered tar and nicotine levels measured by the FTC method did not necessarily reflect reality for human smokers, however. The FTC was aware of that fact. Indeed, in hearings that the FTC held before adopting its testing method, the tobacco industry expressed concerns that, due to considerable variations in individual smoker behavior, the FTC’s method did not, and could not, measure the amount of tar and nicotine that smokers actually inhale. When the FTC adopted its mechanical test method, it issued a press release acknowledging the limitations of that testing method and in particular acknowledging that its test could not accurately gauge the amount of tar and nicotine that even an “average” smoker will draw from a cigarette. The FTC explained that it nevertheless was adopting its mechanical test, because it provided a “reasonable standardized method” of measuring tar and nicotine yields that was “capable of being presented to the public” in a “readily understandable” manner.
In 1971, after the FTC adopted its method of measuring tar and nicotine yields, defendant introduced Marlboro Lights to the market. At the time, Marlboro Lights tested below the 15 milligram tar-yield limit using the FTC method. Defendant therefore permissibly could—and did—label and advertise Marlboro Lights as “lowered tar and nicotine” cigarettes.
*93To accomplish the lowered yields as measured by the FTC method, defendant did not decrease or alter the tobacco content of Marlboro Lights. Instead, defendant perforated the cigarette filter with microscopic holes that allowed extra air to be drawn into the smoke passing through it, which diluted the smoke, thereby delivering less tar and nicotine. As already noted, the way that the FTC’s machine smoked a cigarette carefully controlled such variables as the placement of the cigarette in the machine, the number and volume of puffs taken, and the portion of the cigarette smoked. Under those controlled mechanical conditions, Marlboro Lights achieved the yields that permitted defendant, consistently with federal regulations, to claim that Marlboro Lights had lowered tar and nicotine levels. In actual practice, however, smokers could easily defeat the design. In particular, smokers could cover the air holes in the cigarette’s filter, which would produce a more concentrated smoke stream with greater amounts of tar and nicotine than with the holes uncovered. Smokers also could take more frequent puffs, hold the smoke in their lungs for a longer period of time, and smoke more of the cigarette itself.5 And smokers could smoke more cigarettes.
A principal reason why smokers might—either consciously or unconsciously—smoke Marlboro Lights in a way that defeats their design is to achieve a higher level of nicotine in their blood than the cigarettes would otherwise *94deliver. Nicotine is a stimulant, one to which smokers become addicted or habituated. The nature of nicotine significantly increases the probability that smokers will alter their behavior (i.e., by blocking air holes, taking deeper and more frequent puffs, smoking farther down the cigarette, or smoking more cigarettes) for either of two reasons. One is that smokers — even ones who have never smoked any other brand or variety of cigarette — will desire a higher stimulant effect than the “light” cigarettes would otherwise deliver (the phenomenon of altering smoking behavior for that reason is termed “titration”). The other reason arises for smokers who switch from regular-strength cigarettes to so-called “light” ones. Those individuals often have become habituated to a particular level of nicotine and, to satisfy their craving for that level, they alter how they smoke a lowered tar and nicotine cigarette (a phenomenon termed “compensation”). In either of those circumstances, the amount of tar and nicotine delivered to a smoker will be higher than the amount measured by the FTC method and potentially will be the same as the amount that a smoker would obtain by smoking a regular cigarette.
B. Plaintiffs’ UTPA Claims
That brings us to plaintiffs’ UTPA claims. Plaintiffs alleged, and maintained that they would prove, that from the time that defendant introduced Marlboro Lights to the market, defendant was well aware of the compensation and titration phenomena. According to plaintiffs, defendant understood the likelihood that many human smokers of Marlboro Lights would not obtain the benefit of “lowered tar and nicotine” that the labeling appeared to offer. In fact, plaintiffs intended to prove that defendant purposefully chose an “elastic” tar and nicotine delivery design — that is, one subject to manipulation by the smoker — to give Marlboro Lights smokers the illusion of lowered tar and nicotine while allowing them to obtain the higher levels of nicotine that they potentially craved. Defendant, for its part, was prepared to dispute many of plaintiffs’ assertions, including that defendant deliberately designed Marlboro Lights to pass the FTC’s method, while permitting smokers to readily defeat that design. What defendant did not dispute, however, is that it did not, until 1990, qualify its “lowered tar *95and nicotine” representation by explaining or warning that the amounts of tar and nicotine that Marlboro Lights deliver to a smoker could vary depending on how the cigarette is smoked. And when defendant began to give that disclaimer in 1990, it did so in its print advertising of Marlboro Lights; defendant did not provide that added information on the cigarette packages themselves.6
Based on the design and marketing of Marlboro Lights, the two named plaintiffs in this case brought this action seeking economic damages for themselves and an estimated class of 100,000 individuals who purchased Marlboro Lights in Oregon over a 30-year period. In their complaint, plaintiffs set forth two UTPA claims. Both claims alleged that defendant represented Marlboro Lights as having characteristics that they do not have, in violation of ORS 646.608(l)(e) (making such representations unlawful). The two claims, however, were premised on different factual theories.
Plaintiffs’ first claim asserted that, contrary to defendant’s “lowered tar and nicotine” representation, Marlboro Lights did not deliver lowered tar and nicotine to smokers, but instead delivered the same levels as regular cigarettes.7 In that claim, plaintiffs asserted that they and the class suffered economic losses because they paid for lowered tar and nicotine cigarettes that did not in fact deliver lower levels of tar and nicotine than regular cigarettes.
Plaintiffs’ second claim was not premised on the amount of tar and nicotine that Marlboro Lights were represented to and would deliver. Instead, it was based on what *96defendant allegedly represented about Marlboro Lights’ “inherent” design:
“Defendant *** represent[ed] that its *** ‘light’ cigarettes were inherently lower in tar and nicotine than defendant’s regular cigarettes, no matter how they were smoked. In fact, as defendant well knew, whether a smoker actually received lower tar and nicotine depended on several factors, such as whether the smoker covered ventilation holes in the cigarette, the number of puffs taken on each cigarette, and the amount of each cigarette smoked, none of which defendant disclosed to any plaintiff or class member. Defendant both affirmatively misrepresented that its ‘light’ cigarettes would inherently deliver low tar and nicotine, and failed to disclose that, in order to receive lower tar and nicotine, the smoker would have to smoke the ‘light’ cigarettes in a particular way.”
In their second claim, plaintiffs further alleged that they and the class members suffered ascertainable losses because they paid for cigarettes that they believed were inherently lower in tar and nicotine than defendant’s regular cigarettes— that is, lower in tar and nicotine no matter how they were smoked — but received cigarettes that would deliver lowered tar and nicotine only if smoked in a particular way. For both claims, plaintiffs requested “[e]conomic damages for purchase price refund or diminished value, in an amount to be proved at trial.”
In answer to the complaint, defendant denied most of plaintiffs’ allegations. Defendant also asserted numerous affirmative defenses, including that the claims of the named plaintiffs and putative class members were barred by the one-year statute of limitations for private UTPA claims. See ORS 646.638(6) (claim must be brought one year from discovery of unlawful conduct).
Shortly after defendant filed its answer, plaintiffs moved under ORCP 32 C(1) for class certification of the entire action.8 In the alternative, plaintiffs sought certification of *97“common issues” under ORCP 32 G — so-called, “issue class” certification.9 Defendant opposed both the class and issue class certification requests. Among other arguments that it advanced in opposition, defendant urged that several of the issues central to defendant’s liability on plaintiffs’ claims could not be tried based on evidence common to the class members as a whole, but instead would require individual inquiries. Relevant to plaintiffs’ first UTPA claim, defendant produced studies, expert opinion, and other evidence to show that smoking behavior — and the phenomena of titration and compensation- — -vary widely from one smoker to another, with the result that, for many smokers, Marlboro Lights in fact do deliver lower levels of tar and nicotine than regular cigarettes. Relevant to plaintiffs’ second UTPA claim, defendant likewise submitted studies, expert opinion, and even deposition testimony from one of the named plaintiffs to demonstrate that many consumers do not believe that lowered tar and nicotine cigarettes are healthier to smoke and purchase Marlboro Lights for reasons unrelated to their possible health benefits. Defendant argued that it was entitled to probe individual class members’ perceptions of the product and reasons for buying it to determine whether, as alleged, those individuals paid for Marlboro Lights based on a representation that they had a characteristic or quality that they in fact did not have.
Finally, as relevant to defendant’s statute of limitations defense, defendant pointed out that the plaintiff in a UTPA action has only one year from discovery of the unlawful practice to file a claim. ORS 646.638(6). As relevant both to that affirmative defense and to plaintiffs’ second UTPA claim, defendant produced evidence that, beginning in the 1970s and at times throughout the 30-year class period, health organizations and the lay press publicized the health risks of lowered tar and nicotine cigarettes. Those articles challenged the notion that “light” cigarettes were healthier to smoke than regular cigarettes, principally on the basis that smoker behavior often readily defeated the lower tar *98and nicotine designs of the cigarettes. Defendant also relied on defendant’s print advertising from 1990 forward, which contained express disclaimers to the effect that tar and nicotine delivery might vary depending on how the cigarette was smoked. Defendant argued that many of the 100,000 estimated class members who purchased Marlboro Lights in Oregon could have had actual knowledge of the express advertising disclaimers and public media reports well before July 2001 (one year before this action was filed), which would bar their claims. Defendant urged that litigation of its affirmative statute of limitations defense would therefore also require highly individualized inquiries and could not be litigated based on evidence common to the class as a whole.
At the hearing on plaintiffs’ motion for certification, in the face of defendant’s evidentiary submissions, plaintiffs pressed only their second claim — that defendant, by labeling Marlboro Lights as “lowered tar and nicotine” cigarettes, had represented that they were “inherently” lower in tar and nicotine than regular cigarettes (i.e., lower regardless of how they are smoked).10 Plaintiffs maintained that, to prove their UTPA claim, they did not have to show that they and the class members bought Marlboro Lights because defendant represented them to have “lowered tar and nicotine” — that is, plaintiffs did not have to show that purchasers relied on the representation. Rather, plaintiffs’ position was that it was enough that defendant represented Marlboro Lights to be inherently lower in tar and nicotine, that plaintiffs and the class members bought them (regardless of why they bought them), and that what plaintiffs and *99the class members bought lacked the characteristic of being inherently light. Alternatively, plaintiffs argued that defendant’s evidentiary submissions were flawed in various ways, and that the trial court should conclude that the representation of Marlboro Lights as inherently lower in tar and nicotine was a substantial factor in every class member’s decision to purchase them. Under that alternative theory, plaintiffs’ theory was that every class member had suffered an economic loss because they did not get what they believed they were buying.
C. The Trial Court and Court of Appeals Decisions
At the conclusion of the hearing on the class certification motion, the trial court took the matter under advisement, later issuing a written opinion to explain its denial of the motion. The trial court ultimately determined that “individual issues vastly predominate over the common issues of law and fact” and, because of that, a class action was not a superior means for resolving the putative class members’ individual claims. As earlier noted, the Court of Appeals disagreed with the trial court’s assessment that individual issues predominate over common ones, and reversed and remanded on that basis. Although we discuss more detailed aspects of the trial court’s and the Court of Appeals’ rationales in our analysis of the issues, an overview of their respective opinions provides context for the issues before us on review.
1. Trial Court’s Decision
In declining to certify the class, the trial court first examined the elements of plaintiffs’ UTPA claim. To recover damages in a private action based on an alleged unlawful trade practice, a plaintiff must suffer “an ascertainable loss of money or property * * * as a result of’ the alleged unlawful trade practice. ORS 646.638(1). The trial court determined that plaintiffs’ class claim depended on proof that, first, plaintiffs and the class members suffered an ascertainable loss of money or property, and second, their loss was caused by — that is, was “as a result of’ — the alleged unlawful trade practice, which in this case was the “lower tar and nicotine” representation. Both elements of plaintiffs’ claim, the trial court concluded, would require individualized inquiries of the class members.
*100With respect to ascertainable loss, the trial court rejected the plaintiffs’ argument that the mere fact that a product differs from how it is represented per se makes it less valuable. The court observed that some products, if not as represented, could in fact be more valuable than they would otherwise be (such as a stone represented to be cubic zirconia that is in fact a real diamond). With respect to Marlboro Lights, the evidence before the trial court was that they had always been priced the same as Marlboro regulars. Thus, Marlboro cigarettes cost the same with or without the feature of being lower in tar and nicotine. Given that evidence, the trial court concluded that it could not be inferred that purchasers overpaid for Marlboro Lights on the basis that they were not inherently lower in tar and nicotine. Although plaintiffs represented that they would be able to present expert testimony at trial that an “inherently” light cigarette had special economic value, thus making Marlboro Lights worth less because they did not in fact have that inherent characteristic, plaintiffs did not come forward with that expert evidence to support their motion for class certification. Left with only ordinary market inferences to draw and the lack of any price difference between Marlboro Lights and regulars, the trial court found no basis to infer that purchasers of Marlboro Lights, as a group, suffered an ascertainable loss of money based on the fact of their purchase, without more.
Without a viable theory of that kind, the trial court concluded that the issue of ascertainable loss would require extensive individualized inquiries. The court emphasized, and quoted from, defendant’s expert evidence that “the vast majority” of “light” cigarette smokers titrate or compensate “only partially” on a per-cigarette basis. On the basis of that evidence, defendant had urged, many putative class members got exactly what defendant represented — lower tar and nicotine. The trial court specifically noted that plaintiffs had not presented evidence of their own to refute defendant’s evidence on that point. On the record before it, the trial court found that the amount of tar and nicotine that Marlboro Lights delivered varied significantly based on smoker behavior. The trial court therefore concluded that, to establish an ascertainable loss from having purchased *101Marlboro Lights, “each plaintiff and each member of the proposed class must prove that the Marlboro Lights he or she purchased did not deliver lowered tar and nicotine to the person who smoked them.”
The trial court likewise concluded that individual inquiries would predominate in litigating whether any ascertainable loss was “as a result of’ defendant’s alleged misrepresentation. The trial court reasoned that, in this case, to establish that plaintiffs had suffered an ascertainable economic loss that was caused by (i.e., “as a result of’) defendant’s unlawful practice, plaintiffs had to show that the class members purchased Marlboro Lights in reliance on the perceived health benefits of a lowered tar and nicotine cigarette. Said another way, plaintiffs had to show that defendant’s marketing of Marlboro Lights as “lowered tar and nicotine” cigarettes was a substantial factor in each class member’s purchase decision. In the trial court’s view, defendants had produced extensive evidence that many individuals bought Marlboro Lights for reasons other than their perceived health benefits. Plaintiffs, in response, had not demonstrated to the trial court that they could prove, through evidence common to the class, that the alleged representation that Marlboro Lights were inherently lower in tar and nicotine was a substantial or motivating factor in every class member’s purchase decision. Consequently, the trial court concluded, plaintiffs had not demonstrated that causation/reliance could be litigated without the need for individualized inquiries into the class members’ individual reasons for their purchase.
The trial court further concluded that there were other “indisputably substantial issues unique to each class member”; the court pointed particularly to the individual inquiries required of the class members to litigate defendant’s statute of limitations defense. The trial court considered those other “unique to each class member” issues in its decision, but did not independently rest its ruling on them. Rather, the trial court specifically stated that its class certification decision was principally driven by its conclusion that the individual inquiries required to litigate the issues of ascertainable loss and causation/reliance “overwhelmingly” predominated over common issues. The trial court further *102stated that its conclusions on those two elements of plaintiffs’ claim (causation/reliance and ascertainable loss) were “separate and independent grounds for finding that individual questions predominate.” As the trial court explained: “I would reach the conclusion that individual questions predominate over common questions (to a degree that requires denial of class certification) even if my finding on one of those issues were found on appeal to be wrong.”
The trial court concluded its written opinion and order by denying plaintiffs’ request to certify an issue class. The trial court emphasized in that regard that plaintiffs had not specified what the “issue classes” might be, but instead merely had asserted that “[a] 11 common issues identified by plaintiffs in this motion are appropriate for class certification.” The trial court noted that plaintiffs had listed 17 issues of fact arising out of their own factual statement and 39 issues of law arising out of defendant’s affirmative defenses, all of which plaintiffs had asserted were issues “common” to the class. The trial court declined to certify one or more “issue classes,” concluding that plaintiffs’ nonspecific proposal had not demonstrated how issue certification would “eliminate the main obstacle” to resolution of the class claim, which was the “overwhelming predominance of individual issues.”
2. Court of Appeals’ Decision
Plaintiffs appealed to the Court of Appeals, challenging the trial court’s order denying their motion for class certification and issue class certification. For purposes of a UTPA claim, the majority held, a plaintiff suffers an ascertainable loss if he or she purchases a product that was represented to have a “feature of value” that it in fact does not have, which in turn renders the product less valuable. Pearson, 257 Or App at 136. The loss is, in that instance, “the value of that feature.” Id. at 137. The fact that Marlboro Lights and regulars were always priced the same did not defeat plaintiffs’ claim of ascertainable loss, according to the majority, because that was the wrong comparison. Id. at 138-39. The correct comparison was the value of the represented product versus the value of the received product. Id. at 139. The majority acknowledged that plaintiffs had produced *103no evidence of the value of an “inherently” light cigarette (i.e., one that delivered lower tar and nicotine than a regular cigarette, no matter how it was smoked). Id. at 138. The majority concluded, however, that “a jury could infer” that an inherently light cigarette would be more valuable than a “potentially light cigarette” (i.e., one that could deliver lower tar and nicotine than a regular cigarette if smoked in the same way as a regular). Id. The majority therefore determined that, contrary to the trial court’s ruling, ascertainable loss could be “litigated on a class-wide” basis, which weighed in favor of class certification. Id. at 139.
The Court of Appeals majority then turned to whether individual inquiries would be required to establish that the class members’ ascertainable losses were “as a result of’ the alleged unlawful representation. The majority agreed with the trial court that, to prove causation given the nature of their UTPA class claim, plaintiffs had to prove that the class members relied on defendant’s lowered tar and nicotine representation in purchasing Marlboro Lights. Id. at 143-46. Where the Court of Appeals majority parted company with the trial court was on whether plaintiffs had established that they could prove reliance through evidence common to the class as a whole. The majority reasoned that an inference of classwide reliance could be drawn from the uniform nature of defendant’s representations, defendant’s design and marketing of Marlboro Lights, and the fact that, in the studies and surveys that defendant had submitted, “many” persons who smoked light cigarettes believed that they were safer than regular cigarettes. That evidence “convince [d]” the majority that defendant’s representations were a substantial factor “in the vast majority” of the putative class members’ purchase decisions, and that plaintiffs therefore could prove reliance on the basis of evidence common to the class. Id. at 160.
Finally, although the Court of Appeals majority acknowledged that other issues — and, in particular, the affirmative statute of limitations defense — will require individual inquiries of class members, it concluded that those issues would arise “only after a jury has determined the central question of defendant’s liability to the class.” Id. at 167.
*104On balance, then, the Court of Appeals determined that common issues predominated. Id. at 166. Recognizing that whether a class action is a superior means for trial of the class claims is a decision ultimately committed to the trial court’s sound discretion, the court remanded the case to the trial court to reevaluate the superiority question in light of the majority’s decision on the predominance of common issues. Id. at 168-69. The majority directed the trial court on remand to also “revisit” its denial of issue class certification, given that the trial court’s predominance conclusion had been central to how it had exercised its discretion in that regard. Id. at 171-72.
Judge Duncan wrote a separate opinion concurring in part and dissenting in part, in which three other judges joined. Judge Duncan agreed with several aspects of the majority’s decision, including that ascertainable loss could be litigated through evidence common to the class and that causation in this context required proof of reliance. Id. at 173 (Duncan, J., concurring in part, dissenting in part). She disagreed, however, that reliance could be litigated on a classwide basis. Significantly, her reasoning differed from that of the trial court.
In particular, in Judge Duncan’s view, defendant’s representation that Marlboro Lights were “lowered tar and nicotine” was open to multiple interpretations on the purchasers’ part. Id. at 175. Judge Duncan observed that some purchasers may have understood from the “lowered tar and nicotine” representation that the cigarettes were “inherently” light — that is, “that either the contents or the design of the cigarettes made it impossible” to deliver the same amount of tar and nicotine as regular cigarettes. Id. On the other hand, she reasoned, other purchasers may have understood the representation to mean only that Marlboro Lights would deliver less tar and nicotine if smoked in the same way as regulars. Id. Because defendant’s representation was open to multiple interpretations, Judge Duncan believed that there were likely variations in whether defendant’s representations played a substantial role in the class members’ decisions to purchase Marlboro Lights. Id. She also observed that the class members’ individual understandings were *105likely to vary given the 3 0-year class period involved, during which the lay press publicized information that the amount of tar and nicotine that light cigarettes delivered depended on how they were smoked. Id. at 176. Given those likely variations in purchasers’ understanding, Judge Duncan disagreed with the majority that plaintiffs could prove that the class members, through common evidence, could establish that they made their purchases based on the same understanding of and uniform reliance on the alleged misrepresentation; rather, that issue, in Judge Duncan’s assessment, would have to be litigated “based on evidence specific to each class member.” Id. at 177. Judge Duncan, and the three other members of the court who joined her, therefore concluded that the trial court correctly denied class certification. Id. at 177-78.11
II. DISCUSSION
On review to this court, the parties renew many, if not most, of the arguments that they presented to the trial court and to the Court of Appeals. The overarching issue that we must resolve is whether, on the record before us, plaintiffs failed to demonstrate that issues common to their class UTPA claim predominate over individual issues. Resolution of that issue requires us to consider several subsidiary questions, some of which turn on the standards for class certification and others of which turn on the substantive law that governs plaintiffs’ UTPA claim. We begin our analysis by examining the standards that govern class certification in Oregon, as well as the respective roles of the trial court in making and an appellate court in reviewing the class certification determination. We then turn to plaintiffs’ UTPA class claim to resolve what issues would be pivotal in litigation of that claim and whether plaintiffs have demonstrated that, given those issues, their UTPA claim could be litigated on the basis of evidence common to the class.
*106A. Class Certification Generally
The standards that govern class certification are set out in ORCP 32. Under that rule, a class certification determination divides into two basic inquiries. First, the trial court must determine if the action meets five prerequisites: The class must be so numerous that simple joinder is impracticable (“numerosity”); there must be questions of law or fact common to the class (“commonality”); the named representatives’ claims must be typical of those of the class (“typicality”); the named representatives must be individuals who will adequately protect the interests of the class (“adequacy”); and prelitigation notice requirements must have been complied with (“notice”). ORCP 32 A(1) - (5). If any one of the five requirements is not satisfied, the case cannot go forward as a class action. ORCP 32 B.
If, however, all five requirements are met, the second basic inquiry comes into play: whether “a class action is superior to other available methods for the fair and efficient adjudication of the controversy.” ORCP 32 B. Again, the plaintiff must prevail on the superiority question before the action may be maintained as a class action. Id. The rule identifies eight factors “pertinent” to assessing superiority. The third factor is one that frequently drives class certification decisions, and did in this case: “[t]he extent to which questions of law or fact common to the members of the class predominate over any questions affecting only individual members.” ORCP 32 B(3).12 Neither the “predominance” factor nor any of the other seven, however, is controlling. Rather, the trial court has considerable discretion in weighing all of the factors that apply in a given case and determining if a class action will be a superior means of litigating the class *107claims.13 See generally Newman v. Tualatin Development Co. Inc., 287 Or 47, 51, 597 P2d 800 (1979) (trial court’s determination that action may proceed as a class action “is largely a decision of judicial administration *** [and, i]n making such decisions the trial court is customarily granted wide latitude”).
Establishing that the standards for class certification are satisfied under both ORCP 32 A and B is not a mere exercise in pleading. Rather, a plaintiff seeking class certification has the affirmative burden to demonstrate that the requirements of ORCP 32 are satisfied. Bernard v. First Nat’l Bank, 275 Or 145, 153, 550 P2d 1203 (1976).14 Thus, before a trial court may certify a class under ORCP 32, it must make an affirmative determination that the rule’s requirements for certification are satisfied. Although a class certification decision is not a trial of the merits, Newman, *108287 Or at 51, the issues that must be resolved for the class certification determination frequently overlap with the merits of a plaintiffs’ class claim, Coopers & Lybrand v. Livesay, 437 US 463, 469, 98 S Ct 2454, 57 L Ed 2d 351 (1978) (“the class determination generally involves considerations that are enmeshed in the factual and legal issues comprising the plaintiffs cause of action”).
As federal courts have observed of the parallel requirements of the federal class action rule, the essential objective of the class determination is to “formulate some prediction as to how specific issues will play out” at trial. Waste Mgmt. Holdings, Inc. v. Mowbray, 208 F3d 288, 298 (1st Cir 2000). To that end, a trial court must “probe behind the pleadings” to the extent necessary to resolve the class claims. General Telephone Co. v. Falcon, 457 US 147, 160, 102 S Ct 2364, 72 L Ed 2d 740 (1982). If a class certification decision could come out different ways, depending on how factual disputes are resolved, the answer is not — as it would be for summary judgment — that the class should be certified and the dispute be resolved at trial. Instead, the trial court must resolve the dispute for the limited purpose of the class certification decision. See generally Richard A. Nagareda, Class Certification in the Age of Aggregate Proof, 84 NYU L Rev 97, 100, 114 (2009) (trial court must resolve factual disputes, even if they go to merits of dispute, with no issue preclusive effect, if disputes bear on whether class certification standards are satisfied). Likewise, if the parties have competing views of the law that governs the class claim, a court must “stand ready to say what the law is” to the extent that class determination will come out differently depending on which view is correct. Id. at 164 (internal quotation marks omitted); see also Tardiff v. Knox County, 365 F3d 1, 4-5 (1st Cir 2004) (court must test disputed legal premises of claim at class certification stage if class action would be proper on one premise but not another).
For the superiority determination under ORCP 32 B, in particular, the factors to be weighed by the court are legal in nature; however, their application can require, and even pivot on, the resolution of disputed facts. For example, in Bernard, this court described one of those factors— predominance, i.e., whether common legal or factual questions *109predominate over individual ones — as ultimately a legal conclusion. 275 Or at 153. The court explained, however, that arriving at that conclusion may require factfinding by the trial court. By way of illustration, Bernard noted that if statistical evidence were put before a trial court bearing on the knowledge or state of mind of class members in entering a particular transaction, the conclusion to be drawn from that evidence would be factual in nature, and thus for the trial court to determine. Id. at 153-54. By analogizing the applicable scope of review to the one announced in Ball v. Gladden, 250 Or 485, 487-88, 443 P2d 621 (1968), Bernard conveyed that an appellate court must defer to the trial court’s resolution of any disputed facts, reviewing the record only to determine if no evidence supports the trial court’s express and implicit factual findings. The appellate court then determines if, on the basis of that resolution of any disputed factual issues, the trial court’s ultimate predominance determination was legally correct. Id. at 154. That is the task that this case presents for us.
B. The Predominance Standard
Against that general backdrop, we turn more specifically to the predominance factor set out in ORCP 32 B(3). For purposes of our review, plaintiffs in this case have met the five threshold prerequisites of ORCP 32 A (numerosity; commonality; typicality; adequacy; and notice). The trial court determined that those prerequisites were satisfied, and defendant does not challenge that determination. Because the trial court’s superiority assessment was driven by its predominance conclusion, 358 Or at 106, the question at this juncture is only whether, based on this record, plaintiffs established that common issues predominate over individual ones, contrary to the trial court’s conclusion. We therefore examine the predominance criterion in greater depth.
Although plaintiffs have established, as the “commonality” prerequisite of ORCP 32 A requires, that their claim entails issues of fact or law common to the class, satisfying that requirement is not the same as satisfying the predominance factor for purposes of assessing superiority. As federal courts have recognized of the parallel federal rule, “the predominance criterion is far more demanding” *110than the commonality requirement. Amchem Prods., Inc. v. Windsor, 521 US 591, 623-24, 117 S Ct 2231, 138 L Ed 2d 689 (1997); see also Benedict v. Altria Group, Inc., 241 FRD 668, 672 (D Kan 2007) (commonality requirement is generally subsumed within and superseded by the “far more demanding” predominance requirement for class certification). Commonality asks only if there are questions of law or fact common to the class. ORCP 32 A(2). It does not test how central the common questions are to the resolution of the action. Nor does it take into account the nature of the proof required to litigate those common issues.
The predominance inquiry, on the other hand, asks exactly those things — how central are the common questions, and will common proof resolve them? To test whether common issues of fact or law predominate over individual ones, the trial court must assess whether it is “likely” that the final determination of the action will require separate adjudications to resolve factual or legal questions regarding the individual class members and, if so, how many individual adjudications would be required. See Bernard, 275 Or at 157-62 (stating standard in the context of the issue of individual class member knowledge); accord Newman, 287 Or at 53-54 (relying on standard from Bernard). The predominance criterion requires the trial court to predict how the issues will play out at trial by considering whether the adjudication can be resolved with evidence common to the class (i.e., proof for one class member will be the same for all), or whether it will entail separate inquiries for the individual class members. As one foremost authority on class actions has observed of the predominance inquiry:
“What matters to class certification * * * is not the raising of common ‘questions’ — even in droves — but, rather, the capacity of a class-wide proceeding to generate common answers apt to drive the resolution of the litigation. Dissimilarities within the proposed class are what have the potential to impede the generation of common answers.”
Nagareda, 84 NYU L Rev at 132.
In effect, predominance asks: What do the individual class members have in common, what don’t they have in common, and how much will those similarities and *111dissimilarities matter in litigating the case? In practical terms, the inquiry is designed to determine if proof as to one class member will be proof as to all, or whether dissimilarities among the class members will require individualized inquiries. How the predominance inquiry is answered, then, is a key factor in the trial court’s discretionary assessment of whether a class action will be a fair and efficient means of litigating the case, and thus superior over other available means to resolve the controversy. See ORCP 32 B (certification requires court to find that class action is superior method for fair and efficient adjudication of controversy).
Our cases illustrate that understanding of and approach to the predominance inquiry. The lead case is Bernard, which involved a putative class action brought against a bank for damages based on the bank’s method of computing interest on loans. The plaintiffs’ breach of contract and assumpsit claims alleged that the bank failed to disclose to class members its method of computing loan interest, and the class members did not otherwise know of that computation method. Given that legal and factual context, the predominance inquiry depended on whether it was “‘likely’ that final determination of the action [would] require separate adjudications concerning the knowledge of ‘numerous’ plaintiffs” as to how the bank computed interest on the loans. Bernard, 275 Or at 149. The written contract for the loans had stated that interest would be calculated at a rate-of-percent “per annum,” without explaining which of three different common methods of calculating annual interest the bank would use. Id. at 147-48. The plaintiffs’ theory was that, because the bank had failed to disclose its method of calculation, it could be inferred that borrowers— uniformly, throughout the class — did not know what that method was. See id. at 162 (describing plaintiffs’ evidence and theory).
This court in Bernard concluded, however, that the inference that the plaintiffs relied on was insufficient to show that their knowledge (or lack thereof) could be resolved on the basis of evidence common to the class. The court reasoned that resolving what the class members knew required, in part, determining “the meaning attached by the parties to the words ‘per annum,”’ which in turn was a “matter of *112interpretation” by individual borrowers. Id. at 154-55. The court pointed to the general difficulty of establishing individual class members’ “state of mind” on a classwide basis. Id. at 156. Ultimately, the court declined to certify the class, concluding from the record that it appeared “probable that many claimants’ knowledge will legitimately be in issue and that separate adjudications of the claims of numerous members of the class will be required to dispose of the question of defendants’ liability.” Id. at 162.
Since Bernard, our cases have approached the predominance inquiry in that same way — viz., by asking what legal and factual issues must be resolved, and what the plaintiff and the defendant are entitled to prove and contest, to fairly and fully litigate the issues. The outcomes of the cases have depended on the particular claims involved and the record made on the class certification motion.
In Newman, for example, the question was whether the individual class members in purchasing townhomes had relied on a representation in a warranty brochure that the pipes were copper. The plaintiffs produced no direct evidence on the point, but instead relied on an inference that most or all class members had relied on the representation, which was one among many construction features described in the brochure. Analogizing to Bernard, the court determined that the inference on which the plaintiffs relied was too weak to establish that “every member of the class read, was aware of, and relied upon each of the representations in the brochure.” Newman, 287 Or at 54.
In contrast, in Derenco v. Benj. Franklin Fed. Sav. and Loan, 281 Or 533, 577 P2d 477 (1978), the record established that individual variations among the class members were unlikely. Derenco was similar to Bernard in that it involved how a bank invested certain funds that it held for borrowers; a significant issue for resolution was, again, whether borrowers knew of the bank’s practice. The class certification decision came out differently than in Bernard, however, because of Derenco’s different factual record. In Derenco, the class consisted of persons unfamiliar with banking practices who had not been informed, either by the terms of their contract or otherwise, how the bank invested *113the funds, and the banking industry had no uniform practice in that regard. Id. at 556-57. The court contrasted that record in Derenco with the record in Bernard and concluded, opposite of the conclusion in Bernard, that it was unlikely that any significant number of class members knew of the bank’s investment practice or, for that matter, that it even occurred to them to consider whether and how the bank invested the funds. Id. at 572.15
Finally, Hurt v. Midrex Division, 276 Or 925, 556 P2d 1337 (1976), like Derenco, resulted in class certification based on the predominance of common issues. There, a putative class of employees of an iron ore reducer brought an action for monetary damages, alleging that the defendant’s ore reduction plant emitted paint particulates that damaged the employees’ cars parked outside the plant. By way of defense, the defendant alleged that the employees were aware of the damage being done to their cars but had parked there anyway, and therefore had assumed the risk of where they chose to park or, alternatively, had contributed to and aggravated their own damages. Hurt, 276 Or at 927-28. The defendant urged that the class members’ awareness of the paint particulates would require individualized inquiries, but the court disagreed on the particular record before it. It was “implausible,” the court concluded, that the employees who parked on the defendant’s premises “would not have been equally aware of the problem at the time or shortly after” they first parked there. Id. at 928. In fact, the record showed that the problem of the paint damage to employee cars had been “of sufficient notoriety to be the subject of negotiation between the employees’ union and [the] defendants.” Id. The court viewed the potential for individual variances in the class members’ subjective awareness to *114be “theoretical” only. Id. at 929. A defendant could not, the court declared, defeat the predominance of common issues by “dreaming up a theoretical defense requiring individual inquiries for which there is little basis in fact.” Id. (quoting Bernard, 275 Or at 158).
Collectively, our cases demonstrate that whether common issues predominate in a particular case for purposes of class certification depends on a pragmatic assessment of how a case, if fairly and fully tried, is likely to be litigated. The point of asking whether common issues predominate is to predict the degree to which litigation of the controversy will require delving into individualized proof or, conversely, the degree to which the issues lend themselves to resolution through common proof — that is, proof for one individual class member will be proof for all. The inquiry looks not only to how a plaintiff can prove its prima facie case; it considers, as well, the nature of the plaintiffs claim more generally, the defenses to the claim, the legal and factual issues framed by the parties’ positions, and the record made on the disputed issues of fact. See Bernard, 275 Or at 159 (class action procedures not designed to deprive defendants of valuable procedural and substantive rights by preventing them from asserting what appear to be bona fide defenses; predominance inquiry requires consideration of likelihood that individual inquiries are necessary to permit defendant to litigate legitimate issues in defense). If the record suggests legitimate and legally material factual differences among the class members that a defendant is entitled to expose through individualized inquiries — what Professor Nagareda terms “fatal dissimilarities” among the class16 — the predominance inquiry must take those individualized inquiries into account.
*115C. The UTPA and Plaintiffs’ Specific Class Claim
With that discussion of the class certification standards in place as a foundation, we turn to an analysis of plaintiffs’ UTPA class claim. We first consider the nature of UTPA claims more generally. We then examine the specific claim that plaintiffs seek to pursue on behalf of the putative class. As we will discuss, with respect to the specific claim that plaintiffs seek to pursue, the parties’ positions frame both legal and factual disputes that bear directly on the class certification question of whether common issues predominate over individual issues. We work through those disputes, consistently with our role in resolving legal issues and the trial court’s role in resolving factual issues. We then resolve the ultimate issue before us — whether, given what plaintiffs must prove to establish their UTPA claim, plaintiffs carried their burden to show that, on issues apt to drive the resolution of the litigation, evidence common to the class will generate common answers for the individual members; or in contrast, whether a full and fair resolution of the controversy has a substantial — and not just theoretical— likelihood of requiring individual inquiries.
1. UTPA Claims Generally
Oregon’s UTPA, like those of many other jurisdictions, was enacted as a comprehensive statute for the protection of consumers from unlawful trade practices. See State ex rel Redden v. Discount Fabrics, 289 Or 375, 382, 615 P2d 1034 (1980) (discussing UTPA generally). The trade practices declared unlawful under the UTPA are extensive, too much so for description. See generally ORS 646.607 (setting forth unlawful trade practices involving unconscionable tactics and failure to deliver as promised or failure to refund for undelivered goods and services); ORS 646.608 (enumerating approximately 76 additional unlawful trade practices). But as relevant here, under ORS 646.608(1)(e), a person engages in an unlawful trade practice if, in the course of that person’s business, vocation, or occupation, the person “[Represents that goods * * * have sponsorship, approval, characteristics, ingredients, uses, benefits, quantities or qualities that [they] do not have.” For purposes of the UTPA, a “representation” includes “any assertion by words or conduct” and also “a failure to disclose a fact.” ORS 646.608(2).
*116The UTPA has both public and private enforcement provisions. ORS 646.632 (authorizing officials to bring action in name of state); ORS 646.638 (authorizing action by private person). A public official bringing an enforcement action may seek, among other possible relief, injunctions, imposition of statutory penalties, and loss of licenses and franchises. See Weigel v. Ron Tonkin Chevrolet Co., 298 Or 127, 134, 690 P2d 488 (1984) (citing provisions). A public action does not require proof that any consumer has suffered economic loss or other injury as a result of the unlawful practice. Discount Fabrics, 289 Or at 384.17 Nor does a statute of limitations apply to a public action.
A private action, in contrast, must be brought within one year after discovery of the unlawful conduct. ORS 646.638(6). And, unlike in an action a UTPA claim pursued by a public official, a plaintiff in a private action must prove injury in the form of an “ascertainable loss of money or property.” ORS 646.638(1). By way of relief, the injured consumer can recover actual damages or a statutorily set minimum damage amount of $200, “whichever is greater.” ORS 646.638(1). And although a private action may be brought as a class action, at the time of this action, class members were limited to actual damages; they could not recover the statutory minimum of $200 each. See ORCP *11732 K (2007) (class action cannot be maintained for minimum damages under ORS 646.638).
The nature of the “ascertainable loss” that the private plaintiff must prove distinguishes a UTPA claim from any other claims or remedies that a plaintiff might pursue through other actions. The requirement that the loss be “ascertainable” connotes generally that it is one “capable of being discovered, observed, or established.” Scott v. Western Int. Sales, Inc., 267 Or 512, 515, 517 P2d 661 (1973). Thus, the loss must be objectively verifiable, much as economic damages in civil actions must be. See ORS 31.710(2)(a) defining “economic damages” as “objectively verifiable monetary losses”). But unlike general economic damages in a civil action, the loss required for a UTPA claim must be specifically of “money or property, real or personal.” ORS 646.638(1). An ascertainable loss of some other kind—such as loss of physical ability due to a personal injury—is not cognizable in a UTPA claim. Likewise, noneconomic losses cognizable in a civil action—such as physical pain, emotional distress, or humiliation (ORS 31.710(2)(b))—will not satisfy a private UTPA plaintiffs burden. Finally, the ascertainable loss must be “a result of’ the unlawful trade practice. That is, the unlawful trade practice must have caused the ascertainable loss that the plaintiff suffered.
2. Plaintiffs Specific Class Claim
As earlier described, plaintiffs brought two UTPA claims, both of which alleged that defendant represented Marlboro Lights as having characteristics that they do not have, in violation of ORS 646.608(l)(e). The two claims differed, however, in the nature of the characteristic involved. The first claim asserted that Marlboro Lights promised lower tar and nicotine amounts and failed to deliver those lower amounts. The second claim was premised on a different promised characteristic: that defendant represented Marlboro Lights to be “inherently” lower in tar and nicotine—that is, lower no matter how they were smoked—when in fact the amount of tar and nicotine varied based on smoker behavior. As to the class, plaintiffs pursued only their second claim — the claim based on “inherent” lightness.
*118In pleading that claim, plaintiffs asserted that defendant made the misrepresentation both affirmatively and by omission:
“Defendant both affirmatively misrepresented that its ‘light’ cigarettes would inherently deliver low tar and nicotine and failed to disclose that, in order to receive lower tar and nicotine, the smoker would have to smoke the ‘light’ cigarettes in a particular way.”
As required for their private UTPA action, plaintiffs also asserted that they and the class members suffered ascertainable losses “as a direct result” of defendant’s misrepresentation “because they paid for cigarettes they believed were inherently lower in tar and nicotine than defendants’ regular cigarettes but received cigarettes that would deliver lowered tar and nicotine only if smoked in particular ways.” By way of relief, plaintiffs asked for “[e]conomic damages for purchase price refund or diminished value, in an amount to be proved at trial.”
In the trial court and in the Court of Appeals, as we earlier described, the dispute between the parties on the predominance factor centered on the elements of “ascertainable loss” and “causation” of that loss. The Court of Appeals (both the majority and the dissent) concluded that ascertainable loss could be resolved based on evidence common to the class, contrary to the trial court’s conclusion. Both the Court of Appeals majority and the dissent also agreed with the trial court that proof of causation would require proof of “reliance.” But the majority concluded, contrary to the trial court, that causation/reliance could be proved by proof common to the class, while the dissent agreed with the trial court (albeit for different reasons) that litigation of that issue would require numerous individualized inquiries of the class members.
To address those issues, it is helpful to distinguish expressly between two different and analytically distinct theories of ascertainable loss on which plaintiffs predicated their class claim. One theory was a loss based on “diminished value.” Specifically, plaintiffs’ claim was that they and the class members bought a product that was worth less than they paid for it, and their damages were the difference *119in the value of the product as represented versus the value of the product they actually received. Plaintiffs’ other theory of loss was that they and the class members bought Marlboro Lights believing defendant’s representation of inherent lightness, they did not get what they believed they were getting, and they therefore were entitled to a refund of their purchase price. Those distinctive theories present different concerns in terms of their legal viability and the nature of the proof required to adjudicate them. We therefore analyze them separately, beginning with plaintiffs’ diminished value theory of ascertainable loss.
a. Diminished Value
The premise of plaintiffs’ diminished value theory was that the characteristic of “inherent lightness” has economic value, so that plaintiffs and the class members suffered an economic loss at the moment of purchase because they bought a product that lacked that characteristic. In effect, they bought a product that was the equivalent of Marlboro regulars, rather than the Marlboro Lights that the product purported to be. At the hearing on their motion to certify the class, plaintiffs explained to the trial court that they were prepared to produce expert testimony at trial that an inherently lighter cigarette (i.e., one that delivered less tar and nicotine than a regular cigarette, no matter how it was smoked) had greater economic value than a regular one. As a result, their theory went, by purchasing cigarettes that were not inherently light, plaintiffs and the class members did not get full value for their purchase. In colloquy at the hearing on the motion for class certification, the trial court clarified the gravamen of plaintiffs’ diminished value theory by stating that “the heart” of that theory was that the class members suffered an economic loss without having “to ever open the pack” because the product they bought was not what it was labeled to be. Plaintiffs’ counsel replied, “That’s correct.” The trial court further clarified that the loss was not because, “when smoked, [Marlboro Lights] deliver less tar and nicotine” than regulars, to which plaintiffs’ counsel replied, “You are there.”
Defendant responded to that theory by producing evidence that Marlboro Lights had always been priced the *120same as its regular cigarettes. Defendant emphasized that the essence of plaintiffs’ class claim was that the Marlboro Lights were not substantially different from Marlboro regulars. Plaintiffs’ diminished value theory hinged on the idea that they and the class members therefore suffered out-of-pocket losses as measured by the difference in value between the product they were supposed to have gotten — Marlboro Lights — and the product they in fact got — the equivalent of Marlboro regulars. Defendant maintained that plaintiffs’ diminished value theory of loss was not viable given the evidence that the two products — Marlboro Lights and Marlboro regulars — had always been priced the same.
Plaintiffs’ position, however, was that such proof was beside the point. Plaintiffs, continuing to point to the expert evidence that they said they would produce at trial, maintained that the characteristic of “inherent” lightness had economic value in the market. In their view, regardless of what purchasers paid for Marlboro Lights, the product was worth less than it would have been if Marlboro Lights had been inherently light, as allegedly represented. Thus, under plaintiffs’ theory, they and the class members suffered an economic loss in the form of the diminished value of the product they bought as compared to the value the product would have had if it were as represented.
Plaintiffs did not, however, produce that expert testimony for the trial court. To the extent plaintiffs’ position was that, any time a product is not as represented, there is automatically an “ascertainable loss,” the trial court rejected its logical viability. “A simple example,” the trial court explained, “shows the fallacy of that proposition as a blanket assertion. If a consumer buys a stone advertised as a cubic zirconi[a], he or she has suffered no ascertainable loss if the stone turns out to be a diamond.” On the record before it — where there was no difference in price between the Marlboro Lights and regular Marlboro cigarettes — the trial court found the inference of diminished value that plaintiffs said could be drawn for the class as a whole to be untenable. Given plaintiffs’ failure to come forward with the expert evidence they said they could produce at trial, the trial court had no other evidence of plaintiffs’ theory of *121diminished value — other than plaintiffs’ flawed logic — to establish the legal viability of that claim.
The Court of Appeals disagreed, not on the basis that plaintiffs had promised to produce an expert at trial who would support their theory, but instead because it concluded that plaintiffs’ theory rationally supported their claim of diminished value. No expert testimony was required, in the Court of Appeals’ view. Rather, “a jury could infer that an inherently light cigarette would be more valuable than a potentially light cigarette.” Pearson, 257 Or App at 138 (emphasis added). Plaintiffs’ claim, the Court of Appeals explained, was that they had purchased a product that was represented to have a feature “of value” that it did not have. Id. at 137-38. The court thus distinguished the trial court’s cubic zirconia example, saying simply that it was “inapt,” because here “plaintiffs’ theory is that they received a product that was less valuable than it was represented to be.” Id. at 137 (emphasis in original).
In reasoning that an inference of diminished value was permissible on this record, the Court of Appeals relied on our decision in Scott v. Western International Surplus Sales, Inc., 267 Or 512, 517 P2d 661 (1973). There, the plaintiff had purchased a backpacking tent that came in packaging that pictured a tent with eaves and a window, and that expressly stated “Nylon Net Rear Window with ZIPPERED flap.” Id. at 514. The tent inside, however, did not have those features. Id. The evidence established only that the plaintiff had paid $38.86 for the tent; no evidence established the price of the same tent without eaves and a window. Id. at 516. On that state of the record, this court concluded that a jury could permissibly infer, first, that a tent with the features pictured on the packaging (eaves and a window) was worth $38.86 and, second, that a tent without those features was worth less than $38.86. 267 Or at 515-16.18 Analogizing to Scott, the Court of Appeals held that a *122reasonable inference could likewise be drawn in this case that the value of being “inherently light” made Marlboro Lights worth more than regular cigarettes. Pearson, 257 Or App at 135-37.
Although such an inference may be logically permissible based on common consumer experience or normal market assumptions for many goods and products, it does not hold up in this context. Here, unlike in Scott, the undisputed evidence affirmatively establishes that there is not — and never has been — a price difference between Marlboro Lights and regulars. When the price of goods is not different based on some represented quality (say, for example, color or flavor), it simply does not logically follow that the quality has greater economic worth.19 In terms of economic loss — which is the kind of loss required here — when there is no price difference for a good with a particular feature and the same good without it, a plaintiff has not paid any extra for the represented quality that the plaintiff did not receive. In other words, the plaintiff is not out of pocket any additional money based on the purchase; the plaintiff got the represented feature for no charge. To be sure, the purchaser may be disappointed in the product because it is not what the purchaser believed he or she was buying, and the purchaser may have suffered a loss in the form of buying something he or she would not have otherwise bought. But a loss in the form of the purchase price is a different theory, one we discuss separately. For a diminished value theory, the fact that the product costs the same with or without a represented characteristic defeats a logical inference that the product without the feature is worth less.
Two other factual aspects of this case compound the difficulty of the inference that the Court of Appeals said was permissible. First, even if we could accept plaintiffs’ unsupported premise — that light cigarettes are economically worth more than regulars — the fact that Marlboro Lights and *123regulars cost the same still defeats a reasonable inference that plaintiffs and the class members paid too much (which a diminished value theory requires). It is just as likely that purchasers of Marlboro regulars, whose product should have been worth less under plaintiffs’ theory, overpaid. Given the price parity between the two products, it is impossible to know which product (if either) is overpriced relative to the other.
The other factor complicating the inference that the Court of Appeals said could be drawn is that this product is one that is consumed, rather than sold on a secondary market. For promotional or other reasons, some goods with a represented quality or characteristic of value may initially be sold at a price that is the same as the good without that quality (such as a car represented to be a new model year, where the price set is the same as for the current model year). For purposes of resale, however, the good may be worth more if it has that quality, even though the purchase price was the same. If, then, the good was misrepresented and does not have that characteristic (for example, the car turns out to be the current year’s model, contrary to the representation), the purchaser suffers an economic loss, if not at the moment of purchase, at least when the purchaser later resells the good.20 But a good that is consumed or otherwise extinguished or depleted in its use is not subject to the same analysis. For goods of those kinds, under a diminished value theory, a purchaser can suffer an economic loss only in the form of having paid too much at the time of purchase. And *124when, as here, the good is the same price with or without a represented feature, no logical inference of economic loss arises.
It may be, as plaintiffs represented to the trial court, that through expert testimony, they had some specialized economic theory to present to establish on a classwide basis that each of the class members suffered a diminished value loss. In other cases involving light cigarettes and similar allegations of loss, the plaintiffs have attempted to rely on expert testimony to establish that the tobacco industry marketing drove up demand, thus inflating the purchase price of light cigarettes, so that the plaintiffs paid more than they otherwise would have if the truth about light cigarettes had been known. See, e.g., McLaughlin v. Am. Tobacco Co., 522 F3d 215, 226-27 (2d Cir 2008) (rejecting theory of economic loss akin to “fraud on the market theory”). So far, no court presented with a specialized economic theory of loss has found the theory viable. But in all events, plaintiffs in this case did not come forward with the evidence that they said they would produce at trial. We are therefore left with whatever logical inferences, based on ordinary market assumptions, that a jury could draw. Contrary to the Court of Appeals’ reasoning, where, as here, there is no difference in the price between a product with the represented feature and one without, plaintiffs’ theory of diminished value provides no logically viable theory on which classwide economic losses can be established.
b. Refund of Purchase Price
Plaintiffs’ alternative theory of economic loss was, as we have described, based on their and the class members’ alleged failure to receive what defendant’s representation led them to believe they were buying. Specifically, plaintiffs alleged that they suffered ascertainable loss “as a direct result” of defendant’s misrepresentation because they paid for cigarettes that “they believed were inherently lower in tar and nicotine than defendant’s regular cigarettes but received cigarettes that would deliver lowered tar and nicotine only if smoked in particular ways.” They sought a refund of their purchase price as a remedy.
*125After plaintiffs moved for certification of the class, and once the parties squared off in their written memo-randa and evidentiary submissions, both legal and factual issues pertinent to the certification decision emerged. One of their key legal disagreements was whether plaintiffs, to prevail on their class claim, would have to prove “reliance.” Specifically, the dispute was whether plaintiffs’ claim required proof that a substantial factor in each class member’s decision to purchase Marlboro Lights was the “lowered tar and nicotine” representation on their packaging. Resolution of that legal question was important to the class certification decision, because it would affect the likelihood that individual inquiries of the class members would be required to determine whether they had, in fact, relied on defendant’s representations. See, e.g., Newman, 287 Or at 54 (reliance of class members on information in sales brochure called for individual inquiries); Bernard, 275 Or at 154, 162 (individual issues predominated where class claim depended on determining individual meaning that class members attached to bank’s representation, as well as what class members knew about bank’s practices). As we earlier recounted, the trial court and Court of Appeals agreed that reliance on defendant’s alleged misrepresentation was required. The parties, before this court, again wage that battle. Although neither the parties nor the trial court or Court of Appeals explicitly analyzed reliance as an aspect of plaintiffs’ purchase price theory of loss, that theory is the appropriate context for the issue. We therefore discuss it at this juncture.21
The starting point is the statute. ORS 646.638(1) does not — at least by its terms — require reliance. It does, however, provide that a person pursuing a private action under the UTPA must suffer “an ascertainable loss of money or property, real or personal, as a result of” an unlawful trade practice. Id. The key phrase is “as a result of.” That *126phrase effectively requires that the unlawful trade practice cause the ascertainable loss on which a UTPA plaintiff relies. In several previous cases, we have examined whether the causation element of the statute equates with a requirement that a plaintiff prove reliance. Our answer has been: “It depends.” Whether reliance is required to establish causation turns on the nature of the unlawful trade practice and the ascertainable loss alleged. Discount Fabrics, 289 Or at 384; Sanders v. Francis, 277 Or 593, 598-99, 561 P2d 1003 (1977).
Although reliance is not, in and of itself, an element of a UTPA claim, it is a natural theory to establish the causation of the loss (i.e., the “injury” in a UTPA claim) for a purchaser seeking a refund based on having purchased a product believing it had a represented characteristic that it did not have. Causation is logically established if a purchaser shows that, without the misrepresentation, the purchaser would not have bought the product and thus should be entitled to a refund. But if the purchaser did not care whether the product had a character or quality as represented (or was not aware of the representation) and bought it for other reasons, then the purchaser’s expectations have not been frustrated. In that circumstance, the misrepresentation cannot be said to have “caused” the purchaser to suffer a loss in the form of the purchase price. As a function of logic, not statutory text, when the claimed loss is the purchase price, and when that loss must be “as a result of’ a misrepresentation, reliance is what “connects the dots” to provide the key causal link between the misrepresentation and the loss. See Poulos v. Caesars World, Inc., 379 F3d 654, 664-65 (2004) (so observing in civil RICO context, where action requires proof of injury as result of unlawful conduct).
Plaintiffs nevertheless argue that the class does not have to establish reliance because theirs was a “failure to disclose” theory, rather than one predicated only on an affirmative misrepresentation. They expressly pleaded, they point out, that defendant “both affirmatively misrepresented that its ‘light’ cigarettes would inherently deliver low tar and nicotine and failed to disclose that, in order to receive lower tar and nicotine, the smoker would have to *127smoke the ‘light’ cigarettes in a particular way.” Relying on Sanders, they urge that a UTPA action based on failure to disclose does not require reliance because it would be “artificial” to require a plaintiff to prove that it relied on undisclosed information. 277 Or at 598-99. In response, defendant characterizes this as a “half-truth” case, one that involves an alleged affirmative misrepresentation coupled with the alleged failure to disclose, and therefore one that requires proof of reliance.
Those arguments miss the mark. It is not the nature of the misrepresentation in this case that requires proof of reliance. It is the misrepresentation coupled with plaintiffs’ theory for having suffered a loss in the form of the purchase price because they did not get what they believed they were buying. In this case, reliance inheres in the combination. We decline the parties’ mutual invitations to reduce the analysis to an exercise in attaching labels. What a plaintiff must prove is that (1) the defendant committed an unlawful trade practice; (2) plaintiff suffered an ascertainable loss of money or property; and (3) plaintiffs injury (ascertainable loss) was the result of the unlawful trade practice. In other words, plaintiff must suffer a loss of money or property that was caused by the unlawful trade practice. Whether, to prove the requisite causation, a plaintiff must show reliance on the alleged unlawful trade practice depends on the conduct involved and the loss allegedly caused by it. Sanders, 277 Or at 598-99; see also Discount Fabrics, 289 Or at 384 (so characterizing the holding in Sanders). The answer requires reasoned analysis of the claim, not labeling. Here, in their complaint, although plaintiffs described defendant’s representation in hybrid terms — that is, part affirmative misrepresentation and part failure-to-disclose — that description did not change the causative link that their refund theory depended on. Under that theory, proof of reliance on the alleged misrepresentation was integral to plaintiffs’ class claim. See Benedict, 241 FRD at 679 (for light cigarette purchasers seeking to recover money on claim that they did not get what they thought they were paying for, reliance is required to show loss was caused by alleged misrepresentation).
*128D. Whether Common Issues Predominate
With our conclusion as to what plaintiffs were required to prove to establish their class claim, we turn to whether they carried their burden to show that the claim could be litigated through common proof and would not, instead, require a significant number of individual inquiries. As both the trial court and Court of Appeals concluded, the pivotal issue was whether plaintiffs could prove reliance through common evidence.
Plaintiffs argued to the trial court — and continue to argue on review — that they can prove classwide reliance circumstantially, by inference. Specifically, they urged that a factfinder could find that the representation made through the labeling of Marlboro Lights to all purchasers (i.e., “lowered tar and nicotine”) was made uniformly to all purchasers; that all purchasers would have had a common understanding of the representation (i.e., that the cigarettes were “inherently” lighter no matter how they were smoked); and that the representation went to the product’s “defining feature,” so that all purchasers would have naturally relied on it. From those circumstantial facts, plaintiffs argued, a fact-finder reasonably could infer that the class as a whole relied on defendant’s alleged misrepresentation.
Defendant responded to plaintiffs’ circumstantial basis for establishing classwide reliance by presenting opinion surveys, statistical data, expert opinion, and market analyses. Much of defendant’s evidence was based on actual smoker responses; all was offered to show that smokers have varying beliefs about whether lowered tar and nicotine cigarettes are healthier than regular cigarettes and often choose light cigarettes over regulars for reasons other than perceived health benefits, such as taste. Defendant also produced articles published by the lay press describing how smoker behavior can and often does defeat the lower yields of tar and nicotine that light cigarettes might otherwise deliver. Although plaintiffs took issue with what conclusions should be drawn from defendant’s evidence and what weight the trial court should give it, plaintiffs came forward with no similar direct evidence of their own of why people purchase or smoke light cigarettes; instead, they *129rested on the inference of reliance that they asserted could be drawn.
So framed by the parties’ respective positions, the disputed and central factual question was: Why do people buy light cigarettes? More fundamentally for purposes of the predominance inquiry, the disputed factual issue was: Do all (or nearly all) people buy them for the same reason and can that factual question be determined based on common evidence, or does deciding it require individual inquiries? The trial court expressly found that “it is not as self-evident as plaintiffs contend that every purchaser of Marlboro Lights was motivated substantially by health concerns and acted because he or she was misle[d] by [the] name Marlboro ‘Lights’ or the statement ‘lowered tar and nicotine.’”22 The trial court bolstered that finding by observing that evidence in the record showed “the irrationality of smoking and cigarette purchasing.” In the trial court’s view, plaintiffs had urged that reliance was “susceptible” to proof on a classwide basis, but had failed to present adequate common proof to establish that the 100,000 individuals who had purchased one or more packs of Marlboro Lights over a 30-year period had all done so for the same reason — i.e., because they read, understood, and believed that “lowered tar and nicotine” meant the cigarettes were “inherently” lighter and would deliver less tar and nicotine no matter how they were smoked.
In reaching the opposite conclusion, the Court of Appeals majority went through the record at some length— a record that consists principally of defendant’s evidence, given that plaintiffs effectively presented no evidence of their own beyond the circumstantial evidence of the representation made. See Pearson, 257 Or 161-65. The majority identified problems in defendant’s surveys, such as the fact that some did not “necessarily” capture information relevant to reliance, at least to the majority’s satisfaction. Id. at 163. With others, the majority declined to give them the “weight” that defendant urged they could bear. See, e.g., *130id. (citing example). For survey evidence where consumers answered that they selected light cigarettes for their taste, the majority found that the consumers’ own explanations for purchasing light cigarettes were impeached by tobacco industry research concluding that smokers are “remarkably insensitive to taste nuances in cigarettes.” Id. at 164-65. The majority, after considering the circumstantial inference that plaintiffs relied on, as against the surveys, data, and expert opinions produced by defendant, concluded:
“The uniform nature of defendant’s representations, defendant’s design and extensive marketing of the cigarettes, and studies and surveys that indicate that many[23] persons who smoked light cigarettes believed that they were safer than regular cigarettes convince us that defendant’s representations were a substantial factor in the vast majority of the putative class members’ purchases of at least one pack of Marlboro Lights.”
Id. at 160 (emphasis added).
In approaching the predominance inquiry as it did, the majority stepped out of its institutional role by substituting its judgment for that of the trial court in terms of the weight to be given to the competing evidence that the parties presented. In effect, the Court of Appeals majority took over for, and displaced, the trial court as factfinder. The majority did not conclude that, as a matter of law, the trial court could not find on this record, as the trial court did, that people have varying and often irrational reasons for purchasing the brand and type of cigarettes they do. Had the majority so reasoned, however correctly or incorrectly, that reasoning would at least have been consistent with its role as a reviewing court. Instead, the majority simply disagreed with the trial court on what factual conclusion to draw from the record.
The majority, having overstepped its role in reviewing the facts, then asked the wrong legal question. Looking to Strawn v. Farmers Ins. Co., 350 Or 336, 258 P3d 1199 *131(2011), the majority asked, as we had in Strawn, whether plaintiffs had produced evidence that, even though circumstantial, would permit an inference of classwide reliance. Pearson, 257 Or App at 158-60. The majority pointed out— correctly — that this court in Strawn declined to hold that, in a class action, reliance always must be established by separate evidence specific to each class member. The majority also noted — again correctly — that we concluded in Strawn that the evidence that the plaintiffs had produced in that case permitted a factual inference of classwide reliance. Id. at 159-60. The majority then compared the plaintiffs’ circumstantial evidence of reliance in Strawn to plaintiffs’ circumstantial evidence of reliance in this case, and concluded that an inference of reliance was as permissible in this case as it was in Strawn. Id. at 160.
Strawn, however, arose in a significantly different procedural posture; because of that different posture, it provides no particular guidance for this dispute. The issue in Strawn was whether the defendants were entitled to a directed verdict on the theory that plaintiffs, for their evidence to be legally sufficient, had to produce direct, rather than circumstantial, evidence of reliance. The issue was, in other words, whether the plaintiffs’ proof at trial was legally sufficient to create a jury question on whether the class as a whole had relied; we held that it was. 350 Or at 362. This court was not asked to review — and did not review— whether the class in Strawn had been properly certified; the trial court’s pretrial certification decision was not challenged before this court. 350 Or at 356 n 13 (on review, defendant’s challenges to class certification had dropped from the case). Thus, as the case came to this court, Strawn did not involve the predominance inquiry that this case involves — a pretrial prediction of whether, given the plaintiffs’ claims, the defendant’s defenses, and the record, it was likely that individual issues would predominate over common ones in the adjudication of the reliance element of the plaintiffs’ claim.24 *132The guidance it gave for assessing whether a class plaintiffs circumstantial evidence created a jury question on the issue of classwide reliance is inapposite to the class certification issue that this case presents.
The controlling question for the majority in this case, then, should not have been whether reliance was “susceptible” to classwide proof on the basis of the inference that plaintiffs relied on. See Pearson, 257 Or App at 158 (so framing the question). That framing of the question asks, at most, whether plaintiffs potentially could prove a prima facie case. As we have already discussed, the nature of the evidence required to prove a plaintiffs case is a consideration in the predominance inquiry, but it is only part of the equation. The trial court must consider what it will take to fully and fairly adjudicate the factual disputes between the parties, which requires the court also to consider the nature of the proof that the defendant is entitled to present to raise legitimate defenses and to otherwise challenge the factual premises of the plaintiffs’ theory.
In this case, we decline, for purposes of our analysis, to describe the extensive record in detail, as the Court of Appeals majority did. It was the trial court’s function to *133determine from that record whether, factually, the class members could have had different subjective reasons for their purchase decisions, some of which had nothing to do with whether they did or did not believe that Marlboro Lights were “inherently” lighter. It was likewise the trial court’s function to determine from the record whether, factually, the individual class members were likely to all have had the same or differing beliefs as to whether a cigarette represented to be lower in tar and nicotine was inherently healthier, no matter how it was smoked, than a regular cigarette.
As we have previously observed, sustaining a class action where the claim requires a large number of individual members to have the same subjective states of mind is difficult. Bernard, 275 Or at 156-57. For at least some commodities, the only logical explanation for a consumer’s purchase may be that the product has — or is represented to have — an essential quality, without which it would be worthless. See, e.g., Garner v. Healy, 184 FRD 598 (ND Ill 1999) (consumers purchased “car wax” and allegedly received worthless “non-wax” product). For products that are worthless without a particular represented characteristic or quality, a defendant who asserts that individual inquiries are needed to establish that the product was purchased for other reasons may be “dreaming up a theoretical defense requiring individual inquiries for which there is little basis in fact.” Bernard, 275 Or at 158. This is not that kind of case. Rather, this is a more typical consumer transaction, one that involves consumer choices that implicate states of mind, perceptions, beliefs, and conscious and subconscious motivations.25 Indeed, as *134the trial court expressly found, smoking is in many ways an irrational choice. When a consumer’s choice to engage in activity or buy a product involves irrational motivations, it is all but patent that individual inquiries will be required to determine why the individual members of a large class make the choices they make. See, e.g., Poulos, 379 F3d at 668 (no single, logical explanation for gambling exists; for different gamblers, activity may be an addiction, a form of escape, a casual endeavor, a hobby, a risk-taking money venture, or scores of other things).
Judge Duncan, writing for four members of the Court of Appeals, identified a further reason why the issue of reliance would entail individualized inquiries of the class members. As she explained, the very representation that plaintiffs relied on injected individualized issues as to whether the 100,000 estimated class members likely had a common understanding of the alleged misrepresentation on which plaintiffs relied:
“The representations at issue were that Marlboro Lights were ‘Lights’ and had ‘Lowered Tar and Nicotine.’ In my view, reliance on the part of all putative class members cannot be inferred from circumstantial evidence in this case because of the likely variations in both how putative class members understood defendant’s representations and, relatedly, whether the representations played a substantial role in their decisions to purchase Marlboro Lights.”
257 Or App at 175 (Duncan, J., concurring in part, dissenting in part).
As Judge Duncan intimated, the words “inherently light” do not appear in the representation that defendant made. Likewise, nothing in defendant’s representation said anything to the effect of “lower in tar and nicotine no matter how you smoke them.” Some purchasers, Judge Duncan explained, might infer that Marlboro Lights would deliver less tar and nicotine no matter how they were smoked. Id. Others might infer that, all other things being equal, Marlboro Lights would deliver less tar and nicotine than regular cigarettes only if smoked in the same way as regulars. Id. Compounding the problem, Judge Duncan observed, was that “the class period is long. It runs from 1971 to 2001. In *135that 30-year period, information about dilution filters and the phenomena of titration and compensation was increasingly available.” Id. at 176. The fact that plaintiffs’ claim of reliance depended on a purchaser’s subjective interpretation of the representation, coupled with the long class period over which information that Marlboro Lights were not inherently light was increasingly available to the public, led Judge Duncan to conclude that there was a “legitimate question” whether a considerable number of the class members knew, or were on notice, that Marlboro Lights were not inherently light and did not rely on that representation. Id. at 177. Judge Duncan’s conclusion that individual issues would predominate over common ones in litigating the reliance element of plaintiffs’ claim, although different from the trial court’s reasoning, is correct as well.26
Thus, we agree with both the trial court and Judge Duncan that plaintiffs failed to show that the reliance required to prove their refund theory of economic loss could be litigated through common evidence, rather than requiring individual inquiries of the class members. As the trial court found, and as the record supports, there was ample evidence that, among the putative class of 100,000 who purchased Marlboro Lights in Oregon over a 30-year period, the purchasers had varying beliefs as to whether lowered tar and nicotine cigarettes were “healthier” to smoke. The record also contains ample evidence that the putative class *136members were motivated to buy Marlboro Lights for a variety of reasons, some irrational, and others unrelated to whether Lights were “inherently” lower in tar and nicotine than regular cigarettes. And, as Judge Duncan reasoned, the representation of “inherent” lightness on which plaintiffs rely depends significantly on interpretation and what purchasers understood the “lowered tar and nicotine” label to imply, which added to the potential for the class members to have varying motivations for buying Marlboro Lights. For those reasons, the potential for fatal dissimilarities (to borrow again from Professor Nagareda) in terms of the class members’ reliance on the alleged misrepresentation was great.
As we have emphasized, a trial court’s role in deciding whether to certify a class is to make a preliminary forecast of how the adjudication of the issues at trial likely will play out. Unlike in Strawn, where the issue was the sufficiency of the plaintiffs’ evidence to survive a motion for directed verdict, the trial court does not decide class certification by assessing only whether an inference of classwide reliance permissibly can be drawn from evidence common to the class. The class certification decision also must account for how the defendant legitimately may seek to dispute the plaintiffs evidence. Here, the trial court found that, factually, there were likely to be material differences among the individual class members as to whether they subjectively believed that Marlboro Lights were inherently lower in tar and nicotine and purchased them in reliance on defendant’s lower tar and nicotine representation. On the basis of that showing by defendant, the trial court correctly determined that plaintiffs did not carry their burden to show that, on the element of reliance, common issues prevailed over individual ones. See Bernard, 275 Or at 162 (predominance standard not satisfied where record establishes that state of mind of individual class members will legitimately be in issue and will require separate adjudications of claims of numerous members of class).
E. Statute of Limitations
On review, defendant renews its argument that, in addition to the individual inquiries required to resolve *137the issues of ascertainable loss and causation of that loss (reliance), the litigation of defendant’s statute of limitations defense likewise would require highly individualized inquiries and could not be resolved on the basis of common proof. The trial court did not suggest that it would have denied class certification on that basis alone. It did, however, agree with defendant and considered the need for individual inquiries to resolve defendant’s statute of limitations defense a factor in its conclusion that individual issues predominated, which in turn weighed in the trial court’s discretionary assessment that a class action was not a superior means for litigating the claim in this case. The Court of Appeals, on the other hand, although it agreed that resolving defendant’s statute of limitations defense would require individual inquiries, concluded that it should not be a factor in assessing predominance. Pearson, 257 Or App 166-67. We therefore discuss, briefly, why the trial court was correct and the Court of Appeals was not.
As we have already described, under ORS 646.638(6), a private UTPA action must be brought within one year from the discovery of the unlawful trade practice on which it is based. Because the limitation period is tied to the plaintiffs “discovery” of the unlawful conduct, it runs in this case from when the plaintiffs and the members of the plaintiff class either actually knew or should have known that the representation that Marlboro Lights were lower in tar and nicotine was not true. See generally FDIC v. Smith, 328 Or 420, 428, 980 P2d 141 (1999) (“In general terms, a cause of action does not accrue under the discovery rule until the claim has been discovered or, in the exercise of reasonable care, should have been discovered.”); see also Sazenz v. Pittenger, 78 Or App 207, 211-12, 715 P2d 1126 (1986) (UTPA statute of limitations begins running when plaintiff knows or should have known of the allegedly unlawful conduct).
In class actions, the extent to which a statute of limitations defense is likely to entail highly individualized inquiries of class members depends on the nature of the claim and the specific facts involved. Here, no one disputes that individualized inquiries would be required to determine whether and when individual class members *138were aware, through publicized reports or otherwise, that Marlboro Lights either were not lighter at all or were lighter only if smoked in a particular way. Defendant asserts that individual inquiries of class members would be required to determine whether and when they became aware that the alleged representation was false; plaintiffs, the trial court, and the Court of Appeals agree. In effect, there is consensus that, given the estimated 100,000 members in the putative class, coupled with the significant 30-year time period involved and the publicized information during that period that “light” cigarettes were not necessarily healthier, the claims of many of the putative class members might be time barred.27
The Court of Appeals recognized that the statute of limitations issue would require individualized inquiries and proof; in its view, however, that fact did not affect the predominance inquiry, because those individualized questions would arise “only after a jury has determined the central question of defendant’s liability to the class.” Pearson, 257 Or App at 166-67. The Court of Appeals considered the statute of limitations defense as analogous to resolving damages, which it described as a part of a claim distinct from liability, *139and something to be resolved through individualized inquiries after a determination of liability to the class as a whole. Id. at 167 (so concluding; citing federal cases).
The Court of Appeals’ reasoning was flawed, however, and the court too quickly dismissed the problem. If the statute of limitations has run on individual class members’ claims, those claims are barred. Defendant is not liable on them. A statute of limitations defense is approached like other issues that go to the merits on liability. If the facts are undisputed, the defense can appropriately be resolved on summary judgment. See, e.g., Greene v. Legacy Emanuel Hospital, 335 Or 115, 129, 60 P3d 535 (2002) (no genuine issue of material fact existed as to when plaintiff discovered harm; summary judgment on statute of limitations correctly granted). But if disputed facts must be resolved to determine if a claim is time barred, those facts must be resolved by the finder of fact at trial. See, e.g., Keller v. Armstrong World Industries, Inc., 342 Or 23, 33, 147 P3d 1154 (2006) (genuine issue of material fact as to when plaintiff was aware that asbestos might be cause of pulmonary problems precluded summary judgment on statute of limitations ground).
For purposes of a class certification decision, when a statute of limitations defense is not just a theoretical or frivolous issue, but instead has a legitimate basis given the nature of the claim and the facts, a trial court should consider it along with other central issues in the case in the predominance inquiry. In this case, given the large putative class of an estimated 100,000 members, the lengthy 30-year time period for plaintiffs’ UTPA claim, and the significant opportunities throughout that time period for class members to have read or otherwise become aware of publicized information that light cigarettes were potentially no lower in tar and nicotine than regular cigarettes, the individualized inquiries required by defendant’s statute of limitations defense were properly considered in the trial court’s predominance determination.28
*140F. Issue Class Certification
The only remaining issue is plaintiffs’ argument that the trial court erred in denying their alternative request for certification of an issue class under ORCP 32 G. In denying plaintiffs’ request for issue certification, the trial court reasoned that the most fundamental issues in the case — that is, the issues most apt to drive the resolution of the class claim at trial — were likely to require individualized inquiries of the class members, so that individual issues still would predominate for any potential issue class. Because the Court of Appeals disagreed with the trial court’s central determination that individual issues predominate over common ones, it reversed the trial court’s denial of issue class certification, and instructed the trial court to reconsider issue class certification based on what the Court of Appeals deemed to be the correct predominance determination (assuming that the trial court did not certify the class on reconsideration). Pearson, 257 Or at 171-72.
*141We already have concluded that the trial court’s predominance determination was correct. That effectively moots the Court of Appeals’ dispositional rationale for reversing the trial court on the issue class certification issue. The question before us is whether the trial court erred in denying the motion for issue certification.
Plaintiffs’ approach to issue certification was of little aid to the trial court, and is of little aid to us. Plaintiffs did not — and still do not — identify any particular issue or group of issues as the objects of their motion. Instead, at trial and on review, they asserted summarily that “all common issues identified by plaintiff in this motion are appropriate for class certification.” The motion listed 17 “common issues of fact” and 39 “common questions of law,” many of which appear to be duplicative or, at least, overlapping. Plaintiffs do not attempt to demonstrate how those issues are adequate for certification under the applicable rule. And they have not explained how those more specific issues would avoid many of the individualized issues that we have identified. At the least, the need for individualized inquiries to resolve defendant’s statute of limitations defense would appear as likely to arise for the issue classes as for the class as a whole.
ORCP 32 G commits issue class certification, in significant part, to the discretion of the trial court. The particular claim or issue to be certified for class treatment must satisfy all prerequisites for class certification under ORCP 32 A(1) except numerosity—i.e., commonality, typicality, adequacy, and notice. Beyond that, the rule merely provides that, when “appropriate” under the general standards for class certification, the trial court “may” order a class action with respect to a particular claim or issue. See 358 Or at 97 n 9. In this case, given the broad discretion that the rule confers on the trial court and given the limitations of plaintiffs’ argument in favor of issue class certification, we conclude that the trial court did not abuse its discretion, and therefore did not err, in denying plaintiffs’ motion for issue certification.
The decision of the Court of Appeals on class certification and issue class certification is reversed. The trial court order denying class certification and issue class *142certification is affirmed. The case is remanded for further proceedings on the named plaintiffs’ individual claims.
The UTPA is codified at ORS 646.605 to 646.656. The specific provisions under which plaintiffs brought this action are cited and discussed later.
As we later discuss, as an alternative to class certification, plaintiffs also sought certification of an “issue class”' — that is, a class for purposes of resolving one or more elements of, but not the entire, UTPA claim. The trial court denied issue class certification, and the Court of Appeals remanded for reconsideration of that ruling as well. On review, both parties renew their arguments in that regard. We consider whether the trial court correctly declined to certify an issue class after first determining if it correctly denied full class certification.
Plaintiffs -unsuccessfully applied to the Court of Appeals for an interlocutory appeal of the order denying class certification under ORS 19.225. After the interlocutory appeal was denied, the trial court proceeded with the UTPA claims of the two named plaintiffs and granted summary judgment for defendant on the ground that plaintiffs’ UTPA claims were preempted by federal law. Plaintiffs appealed that judgment, challenging both the denial of the motion for class certification and the grant of summary judgment. While that appeal was pending, the United States Supreme Court decided Altria Group, Inc. v. Good, 555 US 70, 129 S Ct 538, 172 L Ed 2d 398 (2008), which held that federal law does not preempt state claims based on the false advertising of cigarettes. On appeal, defendant conceded that the grant of summary judgment on federal preemption grounds was error in light of Altria. The case therefore must be remanded to the trial court for further proceedings on the individual plaintiffs’ claims.
Because this case arises on a motion for class certification, the facts in the record have been developed for that specific purpose, and do not necessarily reflect the factual record that would be made at trial on either the class claims or the claims of the individual plaintiffs. The parties do not dispute many of the facts that we recite by way of general background. The parties do, however, disagree on certain other facts — and the inferences to be drawn from those facts— that the trial court considered in deciding the extent to which plaintiffs’ claims would entail common or individualized inquiries. We take up those disputes, and the respective roles of the trial and appellate courts in resolving contested facts of that kind, in our later analysis of class certification issues.
The trial court aptly captured the difference between more typical consumer products, where their contents dictate what a consumer ingests, and a product such as a cigarette, where consumer behavior significantly affects what is ingested:
“With a product such as milk or bread, one is concerned mainly with what the item contains. With regard to tar and nicotine from a cigarette, one’s main concern is not so much with what the unlit cigarette contains as what it yields when smoked. A cup of milk contains a certain amount of fat regardless of whether it is sipped or gulped. The fat content does not vary depending on whether the consumer uses a straw or drinks straight from the carton. The tar and nicotine yield of a cigarette, however, depends not only upon what is contained in the unlit tobacco in the column, but upon the way the cigarette is smoked. The yield can vary with the depth of puff taken, frequency of puffs and how far down the column the cigarette is smoked. If the cigarette has a filter with holes that allow air in to dilute the smoke as it is drawn into the smoker’s mouth, the yield will also be changed if the smoker covers some or all of the holes with his or her mouth or fingers.”
(Emphasis in original.)
To explain more fully: Starting in 1990, defendant added a statement to its Marlboro Lights print advertisements to the effect that tar and nicotine delivery might vary depending on how the cigarette was smoked. After this action and others elsewhere in the country were filed, defendant placed “onserts” on some Marlboro Lights packages that also provided that information. Defendant did so only briefly. Soon after that, defendant removed the “lowered tar and nicotine” statement from Marlboro Lights packages entirely.
In particular, plaintiffs alleged:
“Defendant engaged in an unlawful trade practice within the meaning of ORS 646.608(1)(e) by representing that its cigarettes * * * would deliver to plaintiff and other Marlboro Light smokers less tar and nicotine than defendant’s regular *** cigarettes. In fact, as defendant well knew, plaintiff and other class members would actually receive the same tar and nicotine from Marlboro Light cigarettes as from defendant’s [regular] cigarettes.”
OECP 32 C(1) provides, in relevant part:
“As soon as practicable after the commencement of an action brought as a class action, the court shall determine by order whether and with respect to what claims or issues it is to be so maintained and shall find the facts specially and state separately its conclusions thereon.”
ORCP 32 G provides: “When appropriate an action may be brought or ordered maintained as a class action with respect to particular claims or issues. Each subclass must separately satisfy all the requirements of this rule except for subsection A(1).”
Plaintiffs appeared to abandon their first claim for purposes of class certification in light of defendant’s evidence that the amount of tar and nicotine delivered to individual smokers varied significantly, and was frequently less than the amount that a smoker would obtain by smoking regular cigarettes. The same and similar evidence produced by defendant in this case has caused the overwhelming majority of courts throughout the country to reject class certification where the claim is that most or all smokers received the same amount of tar and nicotine as they would have by smoking regular cigarettes. See, e.g., Phillips v. Philip Morris Companies, Inc., 298 FRD 355, 362 (2014) (citing cases); In re Light Cigarettes Marketing Sales Practices Litigation, 271 FRD 402, 413 (2010) (same). Those courts have reasoned that the evidence of individual variations in the amount of tar and nicotine that “lights” smokers ingest is sufficient to demonstrate that the issue is not a common one, but instead requires individual inquiries. See, e.g., Phillips, 298 FRD at 364-65 (discussing evidence and concluding individual inquiries predominate).
Because the trial court, in denying issue class certification, had appeared to rely at least in part on its conclusion that ascertainable loss could not be tried on the basis of common evidence, the four members of the court who separately concurred and dissented agreed that the case should be remanded to the trial court to reevaluate that aspect of its decision. Id. at 178-79.
Other factors include whether a separate action on the class members’ claims will risk inconsistent adjudications or impair the ability of class members to protect their interests; whether individual members of the class have an interest in individually controlling the action on their claim; whether a class action will be unmanageable; and whether the class members’ claims are too small to justify the expense of litigating them on an individual basis. ORCP 32 B(1), (4), (7), (8). The trial court in this case found that some of the factors listed in the rule weighed in favor of class certification, especially the generally low value of the individual members’ claims and the high expense of litigation. But in the trial court’s judgment, the factors favoring class certification were vastly outweighed by the predominance factor, which favored denying certification.
Oregon’s original class action statute was modeled after Rule 23 of the Federal Rules of Civil Procedure, although some of its provisions differed. See generally Bernard v. First Nat’l Bank, 275 Or 145, 150-51, 550 P2d 1203 (1976) (describing history of Oregon class action statute). Under the original class action statute, which was later codified as ORCP 32 (1991), the predominance requirement was one of the threshold prerequisites for class certification; if it was not satisfied, the class action could not be maintained. Id. at 149-50 (quoting statute); see also ORCP 32 B (1991) (same provision as original statute). In 1992, the Council on Court Procedures substantially modified the rule. See generally Council on Court Procedures, Staff Comment, 1992, reprinted in Lisa A. Kloppenberg, Oregon Rules of Civil Procedure 1994 Handbook 88 (1993) (comment on 1992 amendments). The Council made predominance a factor in the superiority decision, rather than one that, as before, had precluded certification. Id. The Council specifically endorsed commentary accompanying proposed changes to the federal rule (which were not adopted), which emphasized that the predominance question remained “unquestionably important,” but should be weighed with other factors in the superiority assessment. Report and Recommendation of the Special Committee on Class Action Improvements, 110 FRD 195, 204 (1986). Thus, the Council in 1992 did not change the standard by which predominance is tested, but did turn it into a factor for the trial court’s discretionary assessment of superiority, rather than a prerequisite for class certification.
Although ORCP 32 does not expressly place the burden on the plaintiff, it does so implicitly. See ORCP 32 B (prerequisites for class certification must he “satisfied” and trial court must “find” that a class action is superior to other adjudication methods before action may be maintained as class action); ORCP 32 C(l) (trial court must decide class certification by order in which trial court “shall find the facts specially and state separately its conclusions thereon”). The same is true of the federal class action rule. Wal-Mart Stores, Inc. v. Dukes, 564 US 338, 131 S Ct 2541, 2551, 180 L Ed 2d 374 (2011) (party seeking to maintain a class action likewise “must affirmatively demonstrate his compliance” with the requirements of Rule 23).
In Derenco, the court more or less blatantly engaged in de novo review of the factual record, without saying so and without acknowledging the standard of review announced in Bernard. See Derenco, 281 Or at 572 (“We conclude from the evidence in this case * * *[i]” and “It is our conclusion that the proof here indicates ***[.]”). One explanation may be that Derenco involved an action for an accounting, which at common law was an equitable proceeding. See Carey v. Hays, 243 Or 73, 77, 409 P2d 899 (1966) (action for accounting at common law equitable in nature); see former ORS 19.125(3) (1975) (equity cases subject to de novo review on appeal). In all events, Derenco did not purport to change the standard of review announced in Bernard, and we would not adhere to it now if it had.
See, e.g., Nagareda, 84 NYU L Rev at 107, 131 (using term). Professor Nagareda also describes what he terms “fatal similarities” — characteristics common to the class that may defeat their substantive claims. See, e.g., id. (using term). Fatal similarities may defeat the class claim at the summary judgment stage or at some other procedural juncture, but they do not defeat class certification. Id. at 107. Hurt, 276 Or at 928-29, provides an example of a potentially fatal similarity — the common knowledge of all employees that, by parking in their employer’s parking lot, they would incur damage to their cars. That similarity aided the plaintiffs in achieving class certification; on the merits, however, that similarity stood to significantly advance the defense claims that the employees had assumed the risk of parking there and had contributed to their own damages.
That is not to imply that the public official cannot seek to make whole any consumers who may have been injured economically by a practice. Under ORS 646.636, a court’s remedial powers in a public enforcement action include “restoring] to any person in interest any moneys or property, real or personal, of which the person was deprived by means of any” unlawful trade practice. But a private loss of money or property is not an element of the public action. That makes sense because many of the trade practices made unlawful by the statute, although contrary to public policy because of their potential for economic injury, deception, and frustration of consumer expectations, would not necessarily or even likely result in actual or measurable loss of money or property. Examples include vague or false representations about where a product was made, ORS 646.608(1)(b); disparaging comments about a competitor’s product that are false or misleading, ORS 646.608(1)(h); selling goods door to door without making certain required disclosures, ORS 646.608(1)(n); making a false or misleading statement about a prize or contest, ORS 646.608(1)(p); or attempting to induce membership in a pyramid club, ORS 646.608(1)(r). For those and many other unlawful practices listed in the statute, enforcement through a public action, which within 10 days of filing can result in cessation of the unlawful practice through a defendant’s voluntary compliance agreement (ORS 646.632(2)), is often the most effective means of protecting consumers from the practices that the statute makes unlawful.
The plaintiff might have had a problem in Scott had the plaintiff sought damages in the amount of the actual diminished value of the tent that he received, given that state of the record. But this court pointed out in Scott that the plaintiff sought to recover the minimum statutory damages of $200. Id. at 515. Thus, the plaintiff could prevail on the basis of a showing of “some loss,” without showing the amount of the tent’s actual diminished value. Id. at 516.
For example, if cherry cola and regular cola have always cost the same, and through a bottling error, thousands of bottles labeled cherry cola are filled with regular cola and vice versa, it is difficult to see how the purchasers suffered a diminished value loss, although, to be sure, for purchasers who bought the cola based on flavor, they did not receive what they believed they were buying.
The car example is similar to the facts in Weigel, where the plaintiff purchased what was represented to he a new car that was in fact used. The plaintiff pursued a UTPA claim on the theory that he suffered an ascertainable loss in the form of the difference between the price paid and the actual market value of the car, which this court held a jury could infer was less than the price paid. Weigel, 298 Or at 134-37. In other words, as in Scott, the plaintiff suffered an out-of-pocket loss at the moment of purchase based on the price paid and the actual value of the item. Our example poses a harder problem, because the purchase price of the new model year car was unchanged from last year’s model. Again, the UTPA requires ascertainable loss in the form of money or property. It does not, at least textually, include a loss of “property value.” Whether, to prove an ascertainable loss for purposes of the UTPA, a purchaser can rely on diminished property value alone, without some out-of-pocket loss incurred through resale, is nothing we decide in this case. We offer the example to demonstrate the significant difference, however, between goods that are consumed and those that are possessory property with ongoing resale value.
The nature of the causal link required for plaintiffs’ diminished value theory of loss — that is, what would be required to show that the loss was “as a result of’ the alleged misrepresentation — would likewise bear on whether common or individual issues predominate for purposes of that theory. But because plaintiffs’ diminished value theory is not viable on the record before us, we do not decide whether that alternative theory of loss would or would not require reliance to establish the requisite causal connection for plaintiffs’ UTPA claim.
Plaintiffs resist any characterization of that statement by the trial court as a “factual finding,” insisting instead it is at most an observation about the force of plaintiffs’ contention. Suffice it to say, we disagree.
Note that the court said “many,” without attempting to quantify, and left undiscussed how many others held a contrary belief. Nor did the court purport to conclude that the number was sufficiently large as to render the number of consumers with a different belief de minimis.
Significantly, defendant asserted in Strawn that the trial court precluded it from presenting individualized evidence to refute the inference of classwide reliance on which the plaintiffs relied. In rejecting that argument, we specifically noted that the defendant was entitled to contest the plaintiffs’ proof of reliance, and had the opportunity to offer individualized evidence to that end; *132the defendant, however, had failed to take advantage of that opportunity. Id. at 349-50, 356 n 13. Because the defendant in Strawn ultimately did not make individualized inquiries of the class members, we do not know what evidence the defendant would have offered or the extent to which individualized issues would have revealed fatal dissimilarities in the class or undermined the efficiency of the class action that the court had certified.
Nor do we know whether the defendant in Strawn at the class certification stage, in the face of the inference of classwide reliance on which the plaintiffs relied, created legitimate doubt as to whether the class members had uniformly understood and relied on the representation at issue. In the face of such evidence, the fact that the plaintiffs’ evidence could support a rational inference of classwide reliance would not have been enough to carry the plaintiffs’ burden on the predominance inquiry. See Bernard, 275 Or at 159-60 (class certification not appropriate where a legitimate defense will require individual inquiries; class action is procedural device and does not erode the substantive rights of parties or deprive defendant of presenting factual and legal defenses). Rather, if the defendant in Strawn at the class certification stage legitimately drew the uniformity of the class members’ states of mind into doubt, the trial court should have considered that fact in assessing whether individual inquiries predominated over common ones. The important point is not what the answer should have been in Strawn if the defendant had challenged class certification pretrial on that basis; the important point is that the answer would not have been based on the sufficiency of the evidence analysis that our decision in Strawn procedurally involved.
The trial court was not alone in its conclusion that individuals who purchase and smoke Marlboro Lights do so for various reasons, and with a range of beliefs about whether lower tar and nicotine cigarettes are in fact healthier than regular cigarettes. Related putative class litigation has been brought in courts throughout the country over the past decade or so. The overwhelming majority of courts have similarly declined class certification. As did the trial court in this case, those courts have concluded that, to resolve why individuals purchase light cigarettes and what they know and believe about whether and under what circumstances those cigarettes will deliver less tar and nicotine, would require numerous individual inquiries of the putative class members. See Phillips, 298 FRD at 365-66 (so holding; citing representative cases); Lawrence v. Philip Morris USA, Inc., 164 NH 93, 99-101, 53 A3d 525 (2012) (reversing trial court certification of class; extensively discussing factual showing by defendant; concluding as a matter of law that knowledge and mental state of class members required individualized inquiries that predominate over common ones).
Worth emphasizing is that the problem that Judge Duncan identified is one of common reliance and causation among the class members, not whether there was a misrepresentation common to the class. Defendant does not argue — and the trial court did not conclude — that individual differences were likely to plague the question of whether defendant’s lower tar and nicotine labeling was a misrepresentation. Presumably, whether that was a misrepresentation is determined based on an objective standard of what a reasonable consumer would understand the representation to be; no party argues otherwise in this case, and that is not an issue. As an objective inquiry, it would be common to the class. Reliance, however, is necessarily subjective — it turns on what individual purchasers in fact believed and whether their beliefs motivated their purchases of Marlboro Lights. The fact that the misrepresentation on which plaintiffs rely was not express, but depended on inference and subjective interpretation, has bearing on the reliance issue, but not on whether defendant made a misrepresentation. Although defendant does not concede that it misrepresented its product, it does not dispute on review that whether it in fact there was a misrepresentation as alleged is an issue that can be and likely would be resolved for or against defendant based on proof common to the class.
Although she did so to make a different point, Judge Duncan canvassed briefly the information made public about the fact that the amount of tar and nicotine delivered by light cigarettes to smokers depended on how the cigarette was smoked. Pearson, 257 Or App at 176 (Duncan, J., concurring in part, dissenting in part) (discussing that evidence to demonstrate degree to which individual class members may not have understood lowered tar and nicotine representation to mean that Marlboro Lights were inherently light, no matter how they were smoked). Judge Duncan pointed out that information was publicized as early as 1976, five years after the beginning of the 30-year class period, and continued with other reports and articles throughout the class period. Id. Relying on much of the same evidence that defendant put into the record in this case, courts in similar cases in other jurisdictions have likewise concluded that the amount of publicity that occurred between the 1970s and 2001 (when defendant removed the lower tar and nicotine label from Marlboro Lights) made it highly likely that individual purchasers would vary significantly in their awareness of that publicity. See, e.g., Lawrence, 164 NH at 100-01 (describing publicity, including 1982 American Cancer Society and 1993 National Cancer Institute publications distributed nationally, the latter of which specifically advised smokers, when smoking light cigarettes, to not “smoke more cigarettes, inhale them more often or more deeply, or place your fingertips over the holes on the filters”); In re Light Cigarettes, 271 FRD at 412, 421 (noting public information disseminated through class period; concluding that statute of limitations defense would require individualized inquiries of putative class members because, as “a matter of law,” number of potential class members aware of publicity was not de mimimis).
The Court of Appeals had another layer to its reasoning, one that compounded the problem of the need for individual inquiries on the statute of limitations defense. On the issue of classwide reliance, the court concluded that, as long as class members relied “initially” on the alleged misrepresentation in *140purchasing Marlboro Lights, it would not matter if they later became aware (through the publicity or otherwise) that Marlboro Lights delivered lower tar and nicotine only if smoked in the same way as a regular cigarette. That information, the court reasoned, would mean only that class members could not recover for later purchases. Pearson, 257 Or App at 160. The court thus appears to have concluded that an inference of classwide reliance was possible at least for the class members’ initial purchases of Marlboro Lights. As we have explained, the permissible nature of that or any other inference of classwide reliance is not the correct test for class certification; the test requires consideration as well of what defendant would be entitled to show to defeat that inference.
Equally important, though, the more narrow inference on which the Court of Appeals relied would inject further individual inquiries into the case. The class members were defined as individuals who had purchased Marlboro Lights in Oregon during a 30-year period. They did not have to have been Oregon residents at any point during those 30 years or throughout the time they purchased Marlboro Lights. Some members of the putative class may have lived for many years in other states and later moved to Oregon. Others may have traveled or visited Oregon frequently, without ever having lived here. Defendant would be entitled to determine whether nonresident class members made their initial “reliance” purchases in Oregon or elsewhere. Conversely, even lifelong Oregon residents may have made their initial purchases while out of state (attending school or stationed elsewhere with the military, as only two examples). Narrowing the inference, as the Court of Appeals did, thus invited an additional round of individual inquiries as to where the class members made their initial purchases. And it also placed added importance on when those initial “reliance” purchases were made, because the farther back in the 30-year class period each class member began purchasing Marlboro Lights, the more likely that their initial “reliance” purchases would be barred by the one-year statute of limitations.