UNITED STATES DISTRICT COURT DISTRICT OF CONNECTICUT --------------------------------------------------------------- x MARK PATANE, et al., : : Plaintiffs, : MEMORANDUM & : ORDER -against- : : 3:17-CV-1381 (VDO) NESTLÉ WATERS NORTH AMERICA, INC., : : Defendant. : --------------------------------------------------------------- x VERNON D. OLIVER, United States District Judge: Before the Court are Defendant’s motions to preclude, including: (1) the motion to preclude the proffered testimony and opinions of Plaintiffs’ proposed damages expert, Dr. Phillip Johnson, who claims to have used hedonic regression analysis to estimate a differential between the prices consumers paid for Poland Spring (“PS”) bottled water products and the prices they hypothetically would have paid if those bottled water products had been sold with a substitute “purified water” label,1 and (2) the motion to preclude the proffered testimony and opinions of Plaintiffs’ proposed survey expert, Dr. J. Michael Dennis, who designed a consumer preference survey and conjoint survey.2 After considering the parties’ filings,3 and for the reasons discussed below, the Court denies Defendant’s motions to preclude. I. BACKGROUND In this putative class action, the plaintiffs allege that bottled water they purchased from the defendant was not “spring water” as defined by the standards of identity as incorporated
1 ECF No. 735. 2 ECF No. 738. 3 ECF Nos. 735, 736, 737, 750, 751 (Johnson); ECF Nos. 738, 739, 747, 748, 752, 753, 754 (Dennis). into the law of various states. Put more simply, they allege that Nestlé’s “Poland Spring Water” (“PSW”) is not actually “spring water,” despite branding and advertising claiming as much. This action has seen two motions to dismiss and three motions for summary judgment.4 The
third summary judgment motion, the most recent dispositive motion, was resolved by the Court (Meyer, J.) on December 30, 2024. Following transfer of the matter to the undersigned, this Court denied the parties’ cross-motions for reconsideration of Judge Meyer’s December 2024 ruling granting in part and denying in part Defendant’s motion for summary judgment.5 II. LEGAL STANDARD Federal Rule of Evidence 702 provides that a qualified expert “may testify in the form of an opinion” if the proponent of the expert demonstrates to the Court that “(a) the expert’s scientific, technical, or other specialized knowledge will help the trier of fact to understand the
evidence or to determine a fact in issue; (b) the testimony is based on sufficient facts or data; (c) the testimony is the product of reliable principles and methods; and (d) the expert’s opinion reflects a reliable application of the principles and methods to the facts of the case.” Fed. R. Evid. 702. District courts perform a “gatekeeping” function in deciding whether an expert’s testimony is admissible under Rule 702, Daubert v. Merrell Dow Pharms., Inc., 509 U.S. 579,
597 (1993), one that operates within the “liberal standard of admissibility” embodied by Rule 702. See Nimely v. City of New York, 414 F.3d 381, 395–96 (2d Cir. 2005). In defining the
4 See Patane v. Nestlé Waters N. Am., Inc., 314 F. Supp. 3d 375 (D. Conn. 2018) (Patane I); Patane v. Nestlé Waters N. Am., Inc., 369 F. Supp. 3d 382 (D. Conn. 2019); Patane v. Nestlé Waters N. Am., Inc., 478 F. Supp. 3d 318 (D. Conn. 2020); Patane v. Nestlé Waters N. Am., Inc., 583 F. Supp. 3d 341 (D. Conn. 2022); Patane v. Nestlé Waters N. Am., Inc., 761 F. Supp. 3d 424 (D. Conn. 2024) (Patane V). 5 Patane v. Nestle Waters N. Am., Inc., 786 F. Supp. 3d 474 (D. Conn. 2025) (Patane VI) gatekeeping role of the Court, the Second Circuit has distilled Rule 702’s requirements into three broad criteria: (1) qualifications, (2) reliability, and (3) relevance and assistance to the trier of fact. Id. at 396–97. Put another way, “Rule 702 governs the admissibility of expert
testimony: the expert must be qualified, his testimony must be helpful, and his conclusions must derive from reliable principles and methods.” Structured Asset Sales, LLC v. Sheeran, 120 F.4th 1066, 1077 (2d Cir. 2024). Whether the witness is “qualified by knowledge, skill, experience, training, or education to render his or her opinions as an expert” is a “threshold matter” that courts consider before analyzing the relevance and reliability of the testimony itself. Vale v. United States, 673 F. App’x 114, 116 (2d Cir. 2016) (citing Nimely, 414 F.3d at 396 n.11). A witness is qualified
where he or she has “superior knowledge, education, experience, or skill with the subject matter of the proffered testimony.” United States v. Tin Yat Chin, 371 F.3d 31, 40 (2d Cir. 2004). However, the Second Circuit has indicated that an expert’s knowledge need not be perfectly tailored to the facts of the case. See Stagl v. Delta Air Lines, Inc., 117 F.3d 76, 81– 82 (2d Cir. 1997). “If an expert has educational and experiential qualifications in a general field closely related to the subject matter in question, the court will not exclude the testimony solely on the ground that the witness lacks expertise in the specialized areas that are directly
pertinent.” Tardif v. City of New York, 344 F. Supp. 3d 579, 598 (S.D.N.Y. 2018). If an expert meets the threshold requirement of qualification, the Court then must determine whether the expert’s testimony itself is reliable. In Daubert, the Supreme Court identified several factors that may be considered in assessing reliability: (1) whether a theory or technique “can be (and has been) tested,” (2) “whether the theory or technique has been subjected to peer review and publication,” (3) a technique’s “known or potential rate of error,” and “the existence and maintenance of standards controlling the technique’s operation’” and (4) whether a particular technique or theory has gained “general acceptance” in the relevant scientific community.
Amorgianos v. Nat’l R.R. Passenger Corp., 303 F.3d 256, 265 (2d Cir. 2002) (quoting Daubert, 509 U.S. at 593–94) (internal quotation marks and citations omitted). These factors, however, do not constitute a “definitive checklist or test.” Kumho Tire Co., Ltd. v. Carmichael, 526 U.S. 137, 150 (1999). Instead, the inquiry is a flexible one and must be “tied to the facts of a particular case” with attention to “the nature of the issue, the expert’s particular expertise, and the subject of his testimony.” Id.; see also Nicholas v. Bratton, 376 F. Supp. 3d 232, 290 (S.D.N.Y. 2019) (stating that “[w]here a proposed expert witness bases his testimony on practical experience rather than scientific analysis . . . courts recognize that [e]xperts of all kinds tie observations to conclusions through . . . general truths derived from . . . specialized experience”) (internal quotation marks and citations omitted). Next, in addition to ensuring that expert testimony is reliable, the Court must decide whether the expert’s testimony is relevant, i.e., whether it will “help the trier of fact.” In re
Mirena IUD Prod. Liab. Litig., 169 F. Supp. 3d 396, 413 (S.D.N.Y. 2016). Like other forms of evidence, expert testimony is relevant if it has “any tendency to make the existence of any fact that is of consequence to the determination of the action more probable or less probable than it would be without the evidence.” Fed. R. Evid. 401. “Once the thresholds of reliability and relevance are met, the testimony is admissible. Thereafter, any purported weakness in an expert’s methodology or conclusion goes to the degree of credibility to be accorded to the evidence, not to the question of its admissibility.” Royal Ins. Co. of Am. v. Joseph Daniel
Const. Inc., 208 F. Supp. 2d 423, 426 (S.D.N.Y. 2002) (citing Ambrosini v. Labarraque, 101 F.3d 129, 133-35 (D.C. Cir. 1996)). III. DISCUSSION A. Qualifications The first factor in the Court’s inquiry involves the proposed expert witness’s qualifications. A witness may be classified as an expert based upon personal experience alone. Frederick v. Deco Salon Furniture, Inc., No. 16-CV-60, 2018 WL 2750319, at *3 (D. Conn.
Mar. 27, 2018). “When considering an expert’s ‘practical experience and educational background as criteria for qualification, the only matter the court should be concerned with is whether the expert’s knowledge of the subject is such that his opinion will likely assist the trier of fact in arriving at the truth.’” SLSJ, LLC v. Kleban, No. 14-CV-390 (CSH), 2017 WL 4329732, at *4 (D. Conn. Sept. 29, 2017) (quoting Valentin v. New York City, No. 94-CV- 3911, 1997 WL 33323099, at *14 (E.D.N.Y. Sept. 9, 1997)).
1. Dr. Johnson Defendant does not appear to challenge Dr. Johnson’s qualifications, nor could it. Dr. Johnson holds a bachelor’s degree in economics from California State University, Northridge and graduate degrees in economics from the University of California, Los Angeles.6 Dr. Johnson has over twenty-five years of experience in economic analysis, including serving as an Assistant Professor at Instituto Tecnológico Autónomo de Mexico and as an Economist and Managing Director at Econ One Research, Inc.7 Dr. Johnson has analyzed “overcharges, damages, and class certification in the food products industry, services industry, various high
technology industries, and the finance industry.”8 And federal courts have permitted Dr.
6 ECF No. 750-1 at 74. 7 Id. 8 ECF No. 736-1 at 5. Johnson to offer expert opinions on issues related to damages and class certification. See In re: Keurig Green Mountain Single-Serve Coffee Antitrust Litigation (“Keurig”), No. 14-MD- 2542, 2025 WL 354671, at *38 (S.D.N.Y. Jan. 30, 2025) (denying motion to exclude Dr.
Johnson’s testimony); see also Palmer v. Cognizant Tech. Sols. Corp., No. 17-CV-6848, 2023 WL 4155403, at *11 (C.D. Cal. June 1, 2023) (same). Considering Dr. Johnson’s professional experience, the Court finds that he has sufficient “knowledge, skill, experience, training, or education” and is qualified to testify about the hedonic regression analysis and the differential between the prices consumers paid for products sold as “spring water.” Fed R. Evid. 702. 2. Dr. Dennis While Defendant challenges Dr. Dennis’s qualifications, the Court finds that Dr. Dennis is qualified. Dr. Dennis holds a Doctor of Philosophy in political science from the
University of Chicago,9 and has worked as a survey research expert for more than twenty years, authoring more than sixty articles, conference and seminar papers, or book chapters.10 Dr. Dennis is a Senior Vice President at NORC, which is a survey research organization associated with the University of Chicago.11 Dr. Dennis has designed and conducted surveys for more than twenty-five years, including consumer surveys that have been accepted by state and federal courts,12 including, like here, price premium studies using conjoint methodology in the context of litigations and consumer perception surveys.13 And federal courts have accepted Dr.
Dennis’s expert consumer surveys and denied motions to exclude his testimony and opinions.
9 ECF No. 748-3 at 2. 10 ECF No. 739-1 ¶ 14. 11 Id. ¶ 13. 12 Id. ¶ 6. 13 Id. ¶¶ 7, 8. See Newton v. R.C. Bigelow, Inc., No. 22-CV-5660, 2025 WL 994721, at *9 (E.D.N.Y. Feb. 14, 2025), R&R adopted, 2025 WL 965690 (E.D.N.Y. Mar. 31, 2025); see also McMorrow v. Mondelez Int’l, Inc., No. 17-CV-2327, 2021 WL 859137, at *9 (S.D. Cal. Mar. 8, 2021).
Considering Dr. Dennis’s professional experience, the Court finds that he has sufficient “knowledge, skill, experience, training, or education” and is qualified to testify on consumer perception and preferences, including the results of his conjoint survey and price premium analysis in this case. Fed R. Evid. 702. B. Reliability The second factor in the Court’s inquiry involves reliability. “In deciding whether a step in an expert’s analysis is unreliable, the district court should undertake a rigorous examination of the facts on which the expert relies, the method by which the expert draws an
opinion from those facts, and how the expert applies the facts and methods to the case at hand.” Amorgianos, 303 F.3d at 267. At a minimum, “[a]n expert opinion requires some explanation as to how the expert came to his conclusion and what methodologies or evidence substantiate that conclusion.” Riegel v. Medtronic, Inc., 451 F.3d 104, 127 (2d Cir. 2006), aff’d on other grounds, 552 U.S. 312 (2008). Finally, this Court has “considerable leeway in deciding” how it will go about determining the reliability of proffered expert testimony. Kumho, 526 U.S. at
152. 1. Dr. Johnson The Court finds that Defendant has not sufficiently established that any step in Dr. Johnson’s analysis is unreliable. As relevant here, Dr. Johnson conducted a hedonic regression analysis to estimate the alleged price premium.14 “In a hedonic regression, the economist attempts to consistently estimate the relationship between prices and product attributes in a differentiated product
market. The regression coefficients are commonly referred to as implicit (or hedonic) prices, which can be interpreted as the effect on the market price of increasing a particular product attribute while holding the other attributes fixed.” Patrick Bajari et al., A Rational Expectations Approach to Hedonic Price Regressions with Time-Varying Unobserved Product Attributes: The Price of Pollution, 102 Am. Econ. Rev. 1898, 1898 (2012). Dr. Johnson’s regression included water type, product size, state of sale, retailer, brand, container, and business quarter as variables. Dr. Johnson purportedly measured the market price premium that Nestlé was able
to obtain on Poland Spring Water by using a regression model and transactional records of actual sales of bottled water which were obtained from (1) Nestlé, (2) the data analytics and market research company Information Resources, Inc. (“IRI”) and (3) nonparty retailers.15 Using this methodology, he calculated that “Nestlé obtained a price premium of 8.5%” due to its “spring water” representation and estimated that overall class damages totaled $1.453 billion.16 a. Defendant’s Allegations of a Fraud on the Market Theory
Defendant argues that Dr. Johnson’s opinions are inadmissible to establish any injury or damages because his opinions involve a fraud on the market theory that is incapable of
14 ECF No. 750-1 ¶¶ 13, 98. 15 Id. 16 Id. ¶ 13. measuring actual economic injury caused by the allegedly deceptive PS “spring water” statement.17 But the record does not provide any direct support for Defendant’s contention that Dr.
Johnson’s price premium damages model necessarily involves a fraud on the market theory, nor do the plaintiffs advance such a theory. Typically found in the securities litigation context, a fraud on the market theory enables a plaintiff to “establish reliance simply by virtue of the defendant’s public dissemination of misleading information” in an efficient market, where “the market at large internalized the misrepresentation to such an extent that all plaintiffs can be said to have relied on it.” McLaughlin v. Am. Tobacco Co., 522 F.3d 215, 224 (2d Cir. 2008); see also In re Petrobras Sec., 862 F.3d 250, 275–76 (2d Cir. 2017) (describing theory in the
securities litigation context). Defendant does not rely on its own experts to support its position that fraud on the market is involved in this case. Instead, Defendant’s contention that Dr. Johnson’s damages model includes this theory principally relies on a stand-alone district court case. But, unlike here, that case involved a stipulation that the fraud on the market theory applied to a price premium analysis. See In re POM Wonderful LLC, No. 10-ML-2199, 2014 WL 1225184, at * 3 (C.D. Cal. Mar. 25, 2014) (explaining that the parties appeared to agree that the damages model depends upon a fraud on the market theory); see also Saavedra v. Eli
Lilly & Co., No. 12-CV-9366, 2015 WL 9916598, at *3 (C.D. Cal. July 21, 2015) (relying on POM Wonderful). In applying the parties’ stipulation, the POM Wonderful court itself stated it was unaware “of any authority applying a fraud on the market theory to a consumer action.”
17 ECF No. 735-1 at 29–31. 2014 WL 1225184, at *4. There is thus a sharp distinction here, where there is no such stipulation. b. Defendant’s Allegations of Dr. Johnson’s Use of an Average Price Differential Defendant next argues that Dr. Johnson’s use of an average price differential is unreliable on three grounds. These arguments hone in specifically on Dr. Johnson’s opinion that spring water brands are priced 8.48% more than if they were sold as purified water because that price premium is a single, unweighted average price differential between the spring water
and purified water products of nine private-label/value sellers.18 First, Defendant challenges Dr. Johnson’s methodology as a valid way to estimate a PS “spring water” price premium, arguing that value brands are not comparable to mainstream and premium waters. Given the limitations of using averages, Defendant argues that Dr. Johnson’s methodology cannot be used to determine a spring-water price premium for any specific brand’s label statement, such as Poland Spring.19 Second, Defendant argues that Dr. Johnson’s methodology is not a legally
permissible way of proving classwide injury under Tyson Foods, Inc. v. Bouaphakeo, 577 U.S. 442 (2016).20 Third, Defendant argues that there is no support for Dr. Johnson’s opinion that his average retail price differential percentage between spring waters and purified waters applies to Defendant’s direct sales of PS to home & office (H&O) customers.21 As an initial matter, the Court finds that Dr. Johnson’s use of an average price differential rests on scientifically established methods. Dr. Johnson’s derivation of the price
18 ECF No. 735-1 at 31–34. 19 Id. 20 ECF No. 735-1 at 34–38. 21 ECF No. 735-1 at 38–40. premium in percentage terms (8.48%) rests on the usage of a hedonic regression. Hedonic regression is “a tool that purports to measure the value of various product attributes in order to demonstrate the existence of, and to isolate the amount of, a price premium” associated with a
particular attribute. Kurtz v. Kimberly-Clark Corp., 321 F.R.D. 482, 523–24 (E.D.N.Y. 2017). Law review articles have explored whether the statistical technique of multiple regression could be appropriately used to disentangle damages, and the academic verdict is generally favorable.22 Hedonic regression models are routinely accepted to prove price premiums of specific brands in consumer class actions. See, e.g., Allegra v. Luxottica Retail N. Am., 341 F.R.D. 373, 458–59, n.64 (E.D.N.Y. 2022); see also Kurtz, 321 F.R.D. at 523. The Court also finds that Dr. Johnson’s use of an average price premium in percentage terms, rather than a
fixed dollar amount, is an accepted approach for estimating price impacts, such as overcharges.23 Defendant’s argument regarding the unreliability of Dr. Johnson’s use of averages is inconsistent with the Supreme Court’s recognition that “[c]alculations need not be exact” in a model purporting to measure damages attributable to the theory injury of accepted for class-action treatment. Comcast Corp. v. Behrend, 596 U.S. 27, 35 (2013); see also Oliver v. Am. Express Co., No. 19-CV-566, 2024 WL 100848, at *22 (E.D.N.Y. Jan. 9, 2024) (rejecting contention that an expert’s use of averages is fatally flawed) (citing Waggoner v.
Barclays PLC, 875 F.3d 79, 106 (2d Cir. 2017)). Defendant’s attempt to invoke Tyson Foods, Inc. v. Bouaphakeo, to support exclusion is unavailing. The Tyson Foods Court, recognizing that its decision would “reach too far were
22 Finklestein, Michael O. and Hans Levenbach, “Regression Estimates of Price Fixing Damages,” Law and Contemporary Problems, Vol. 46, No. 4, 1983, p. 146. 23 Id. at 150, 156 (describing experts use of price premium in percentage terms). it to establish general rules governing the use of statistical evidence, or so-called representative evidence, in all class-action cases,” explained that a permissible method of proving classwide damages is “by showing that each class member could have relied on that sample to establish
liability if he or she had brought an individual action.” 577 U.S. at 455. The Tyson Court also recognized that “in many cases, a representative sample is the only practicable means to collect and present relevant data establishing a defendant's liability.” Id. (cleaned up). In affirming a district court’s decision to admit evidence of the average time the plaintiffs spent donning uniforms, the Court found that “the experiences of a subset of employees can be probative as to the experiences of all of them” and “probative as to the time actually worked by each employee.” Id. at 459.
While the Tyson Court recognized that a party could challenge an experts’ methodology under Daubert by showing that the representative evidence is “statistically inadequate or based on implausible assumptions,” id., Defendant fails to clear that hurdle. Defendant’s expert, Dr. Ugone, criticizes Dr. Johnson’s model for its failure to consider the variability of pricing across brands, which purportedly results in an “extremes-masking average” that is not representative of any bottled water brand.24 But these concerns about variability and representativeness do not meet the threshold to exclude expert testimony: that “there is simply too great an analytical
gap between the data and the opinion proffered.” Gen. Elec. Co. v. Joiner, 522 U.S. 136, 146 (1997). “[G]aps or inconsistencies in the reasoning leading to [an expert’s] opinion” frequently “go to the weight of the evidence, not to its admissibility.” Restivo v. Hessemann, 846 F.3d 547, 577 (2d Cir. 2017) (cleaned up). Nor does Defendant show that Dr. Johnson’s damages
24 ECF No. 735-1 at 35–37. model is statistically inadequate or based on implausible assumptions, and thus, Defendant’s arguments are fit for the factfinder. Kurtz, 818 F. App’x at 62 (“Defendants’ central contention is that Weir’s analysis either does not or cannot establish a price premium because of issues
such as an incomplete dataset, flawed parameters of the regression, or business considerations not captured by the model. A factfinder may ultimately agree. But if that is the case, then the class claims will fail as a unit.”). Defendant also takes issue with Dr. Johnson extending the same price premium from retail market to the H&O purchasers, arguing that the failure to ground his opinion in any evidence or economic principles leaves only his ipse dixit that the retail price differential applies to PS sale in the H&O market. While “[t]rained experts commonly extrapolate from
existing data . . . nothing in either Daubert or the Federal Rules of Evidence requires a district court to admit opinion evidence that is connected to existing data only by the ipse dixit of the expert.” General Electric Co., 522 U.S. at 146. The Second Circuit has recognized, however, that the “adversary system provides the necessary tools for challenging reliable, albeit debatable, expert testimony” and thus, a district court “should only exclude the evidence if the flaw is large enough that the expert lacks ‘good grounds’ for his or her conclusions.” Amorgianos, 303 F.3d at 267. Here, there are “good grounds” for Dr. Johnson to conclude,
based on his hedonic regression, that Defendant exacts the same price premium from H&O purchasers as from its other customers. Dr. Johnson purported to test the robustness of his approach through sensitivity analyses and concluded that separate H&O data, if available, would likely yield similar results given the shared market dynamics.25 Dr. Johnson used the
25 ECF No. 750-1 ¶¶ 95–97. same regression and included an interaction variable of the Poland Spring indicator with an indicator for Nestle H&O data.26 Any flaws associated with this determination is appropriately raised before the jury.
c. Defendant’s Allegations that Dr. Johnson’s Model Includes Sales That Were Not Made or Are Not Actionable and Failed to Consider Critical Determinants of Bottled Water Prices Defendant also challenges the sufficiency of the data in Dr. Johnson’s model on multiple grounds. First, Defendant argues that Dr. Johnson’s damages opinions are unreliable because they are overstated due to inclusion of sales of PS bottles that were not made to consumers or which are not actionable following the Court’s summary judgment ruling.27 Second, Defendant argues that the damages model produced unreliable results by failing to consider critical determinants of bottled water prices.28 The first argument regarding the sufficiency of the data—that Plaintiff’s use of wholesales sales data included bottles that were never sold due to promotions, breakage, discount giveaways, or other factors—however, fails to warrant exclusion.29 As Dr. Johnson explained, he purportedly accounted for discounts and promotions in calculating the average prices by using net sales and quantities where the data was available.30 Nor is Dr. Johnson’s method rendered unreliable post-summary judgment. Federal Rule of Civil Procedure 26(e) allows Dr. Johnson to update his damages calculation to refine existing calculations without introducing any new theories or methodologies because of subsequent information that was
26 Id. 27 ECF No. 735-1 at 40–42. 28 Id. at 42–43. 29 Id. at 41. 30 ECF No. 750-1 ¶¶ 72, 76, 142. previously unknown or unavailable, such as the Court’s conclusion of law. Anthem, Inc. v. Express Scripts, Inc., 660 F. Supp. 3d 169, 185 (S.D.N.Y. 2023) (granting leave to submit a “narrowly-tailored supplemental expert report” where “the request was made due to the
Court’s conclusions of law” and “not because of an initial careless deficiency”). The second argument, too, does not warrant exclusion. Defendant argues that Dr. Johnson’s model failed to control for the influence of, and completely ignored, the major factors of taste, brand/reputation/heritage, and market segment that influence demand for, and therefore, the price of bottled water. But “the selection of the variables to include in a regression analysis is normally a question that goes to the probative weight of the analysis rather than to its admissibility.” Ploss v. Kraft Foods Grp., Inc., 637 F. Supp. 3d 561, 576
(N.D. Ill. 2022) (citing Bazemore v. Friday, 478 U.S. 385, 400 (1986) (Brennan, J., concurring in part)). While a factfinder may give Dr. Johnson’s testimony less weight due to the purported failures to consider certain factors that influence the demand for bottled water, neither science nor law mandate exclusion of Dr. Johnson’s exclusion here. d. Defendant’s Allegations of Dr. Johnson’s Improper Use of Dr. Dennis’ Price Differentials Finally, Defendant argues that Dr. Johnson’s opinions utilizing Dr. Dennis’s price differentials should be excluded because he failed to scrutinize Dr. Dennis’s analysis in any way. 31 But courts “regularly permit experts to offer opinions based on materials they have not personally prepared, such as other expert opinions and reports, so long as the ultimate opinion is their own.” Keurig, 2025 WL 354671, at *31 (collecting cases). There is no reason here to
conclude differently.
31 ECF No. 735-1 at 44–45. 2. Dr. Dennis The Court finds that Defendant has not sufficiently established that either Dr. Dennis’s conjoint analysis or consumer perception survey is unreliable. a. Defendant’s Allegations that the Conjoint Analysis is Unreliable Defendant argues that Dr. Dennis’s conjoint survey analysis should be excluded on four grounds, that: (1) the conjoint survey is not suitable for ascertaining market values of bottled water attributes,32 (2) the conjoint survey cannot reliably measure any damages because the methodology failed to account for supply-side factors,33 (3) the conjoint survey placed zero
value on taste and other attributes important to consumers and asked respondents to consider unrealistic choice sets for non-existent bottled water products,34 (4) the conjoint survey resulted in an ipse dixit single price differential percentage that defies basic economic principles and is clearly wrong,35 and (5) Dr. Dennis’s application of a retail market price differential to the H&O market ignores fundamental economic differences between the two markets.36
To measure the alleged price premium, Dr. Dennis conducted a conjoint survey.37 Conjoint analysis “ask[s] respondents to choose from or rate hypothetical profiles that combine multiple attributes,” such as (in this case) brand, price, product characteristics, and labeling choices.38 See, e.g., Jens Hainmueller et al., Causal Inference in Conjoint Analysis:
32 ECF No. 738-1 at 21–24. 33 Id. at 24–26. 34 Id. at 26–29. 35 Id. at 29–32. 36 Id. at 32–35. 37 ECF No. 739-1 at 10 ¶¶ 27-28. 38 Id. at 34–35. Understanding Multidimensional Choices via Stated Preference Experiments, 22 Pol. Analysis 1, 2 (2014). Survey respondents typically repeat this selection task across several hypothetical choice sets with the attributes varied. So for example, in one selection task, Dr. Dennis might
ask a consumer to select between an Aquafina bottle labeled “drinking water” and priced at $7.99 and a Poland Spring bottle labeled “spring water” and priced at $6.99; in the next, the same consumer might be asked to select between a Crystal Geyser bottle labeled “well water” and a Poland Spring bottle labeled “purified water.”39 Dr. Dennis ran these sort of choices by 1,278 respondents, who each completed 12 choice tasks.40 He then calculated the relevant importance of the various attributes he measured and determined how much lower a non-spring water PSW product (i.e., “drinking water” or
“well water”) would have to be priced in order to sell at parity with a spring water PSW product.41 Using this methodology, Dr. Dennis concluded that Nestlé’s spring water claim permitted the company to charge a little over 10% more for their product than it could have if labeled PSW “drinking water.”42 The issue with Defendant’s arguments to exclude Dr. Dennis’s conjoint survey analysis is that it squarely challenges the well-established principal in the Second Circuit that “[e]rrors in survey methodology generally ‘go only to the weight of the evidence—subject, of course,
to Rule 403’s more general prohibition against evidence that is less probative than prejudicial or confusing.’” Bustamante v. KIND, LLC, 100 F.4th 419, 427 (2d. Cir. 2024) (quoting
39 Id. at 39 ¶ 115. Respondents were also given the option of rejecting all of the choices. Id. at 40 ¶ 116. 40 Id. at 33 ¶ 106; id. at 40 ¶ 119. 41 Id. at 42-43 ¶¶ 124, 128. 42 See, e.g., id. at 43 ¶ 129. Schering Corp. v. Pfizer Inc., 189 F.3d 218, 228 (2d Cir. 1999)). The sweeping contention that Dr. Dennis’s conjoint survey analysis must be excluded is belied by case law showing that conjoint analysis is a reliable method of measuring the price premium attributable to alleged
label misrepresentations. See, e.g., Kurtz, 321 F.R.D. at 551; see also Cadena v. American Honda Motor Co., Inc., No. 18-CV-40007, 2024 WL 4005097 at *6 (C.D. Cal. July 2, 2024) (finding defendants’ arguments that Dr. Dennis failed to account for supply-side considerations “go to the weight, not the admissibility, of the proposed survey”). That Defendant disagrees with the application of Dr. Dennis’s conjoint analysis, largely based on its own expert testimony, goes to show that this case involves a battle of the experts, the credibility of which is for the jury to decide.
Lastly, the challenge to Plaintiffs’ application of Dr. Dennis’s price premium to both retail and H&O sales mostly rests on the contention that Dr. Dennis lacks relevant data. While “a flawed universe minimizes the probative value of a survey and may lead to skewed results,” Louis Vuitton Malletier S.A. v. Sunny Merch. Corp., 97 F. Supp. 3d 485, 510 (S.D.N.Y. 2015), “issues with representativeness are the kind of errors in survey methodology that usually go to the weight of the evidence and do not warrant wholesale exclusion.” Keurig, 2025 WL 354671, at *26. Therefore, this contention does not show that the survey is unreliable.
b. Defendant’s Allegations that the Consumer Perception Survey is Unreliable As to the consumer perception survey, Defendant first argues that the failures to define “spring water,” “well-water,” and “purified/filtered” undermine the reliability of the surveys.43
43 ECF No. 738-1 at 36–40. To test Plaintiff's theory that Defendant’s labels consisting of “100% Natural Spring Water” and “Natural Spring Water” misleads consumers, Dr. Dennis conducted a consumer perception survey. The survey included two substantive questions. The first question, designed to isolate the impact of the of the “Spring Water” representation on consumers’ understanding of the type of water in the products, used an experimental design whereby respondents were randomly assigned to one of two conditions, the “Spring Water” or “Well Water” condition, and shown one of the following labels:“
“Spring Water” “Well Water”
panmirnete ree oy ae) nD) eo) Ue) AD) a Eb] so) od NET ae NET Tat
Then, after respondents viewed the label, both groups were asked, “Based on the product label, what is your understanding of the type of water in this bottled water product?” and were prompted to choose the following options: “Purified/Filtered water,” “Spring water,” “Well water,” “Something else” or “Don’t know.” The second question, called the “Rank Order Preference Question,” asked respondents to rank the types of bottled waters respondents prefer to purchase.** There were two parts to the question: first, “Which type of bottled water do you most prefer to purchase?” among
Id. 4 80. 46 477.
“Purified/Filtered water,” “Spring water,” “Well water,” “Something else” or “Don’t know/not sure”; and a second follow-up prompt that showed respondents the remaining, unselected water types and asked, “Of these two types of bottled water, which type do you most prefer to purchase? Please select one.”47 Dr. Dennis’s expert survey evidence shows that reasonable
consumers understand that the ingredient contained in bottled water labeled as “100% Natural Spring Water” or “Natural Spring Water” is spring water rather than purified, well, or some other type of water.48 Cases from this Circuit reflect the necessity of precluding testimony related to surveys when the flaws are so egregious as to undermine the probative value of a survey—e.g., when there is an “obvious leading question” that “suggested its own answer,” Universal City Studios,
Inc. v. Nintendo Co., Ltd., 746 F.2d 112, 118 (2d Cir. 1984), or when there are leading questions that “invite guessing by those who did not get any clear message at all,” Procter & Gamble Co. v. Ultreo, Inc., 574 F. Supp. 2d 339, 352 (S.D.N.Y. 2008). Courts exclude surveys that use questions with undefined, ambiguous terms. Malletier v. Dooney & Bourke, Inc. (“Vuitton I”), 525 F. Supp. 2d 558, 604 (S.D.N.Y. 2007) (finding confusion survey inadmissible where the survey’s second main question used a “highly ambiguous” term); de Lacour v. Colgate- Palmolive Co., No. 16-CV-8364, 2024 WL 36820, at *5 (S.D.N.Y. Jan. 3,
2024) (finding report and testimony inadmissible where “no meaningful conclusion can be drawn from respondents’ answers” due to the failure to provide respondents with adequate
47 Id. ¶ 86. 48 ECF No. 692-1 at 210 ¶¶ 15-23; ECF No. 692-1 at 152 ¶ 31 (stating on basis of consumer surveys that “[m]y first key finding, therefore, is that the reasonable consumer understands the ‘100% Natural Spring Water’ claim to be communicating that the at-issue Products actually contain ‘Spring Water’ and not ‘Purified/Filtered Water’ or ‘Well Water.’”). definitions of certain terms); Edmondson v. RCI Hosp. Holdings, Inc., No. 16-CV-2242, 2020 WL 1503452, at *8 (S.D.N.Y. Mar. 30, 2020) (finding undefined terms undermined the reliability of the responses to survey questions). But this case is distinguishable and does not
demand exclusion on this ground. It would be too wide a stretch for this Court to conclude that the failure to define “spring water,” “well-water,” or “purified/filtered,” which are terms that jurors will have necessarily dealt with throughout their lives, would undermine the probative value of the survey. The simplicity of the terms being challenged here is materially different from the terms found ambiguous in Vuitton I (“some business relationship”), de Lacour (“artificial” and “natural”), and Edmondson (“lifestyle” and “events”). The fact that Dr. Dennis’s survey was excluded in the In re KIND/Bustamante litigation
does not bear on whether the survey in this case must be excluded, as it is readily distinguishable. In that case, in support of the plaintiff’s theory that the phrase “All Natural” that appeared on the labels KIND products was deceptive, Dr. Dennis designed a survey where the first question “ask[ed] only about one potential definition of ‘All Natural’ – the definition that plaintiffs selected for this case – and only allow[ed] survey participants to select from finite choices agreeing, disagreeing, or not having an expectation about this definition.” In re KIND LLC "Healthy & All Natural" Litig., 627 F. Supp. 3d 269, 288 (S.D.N.Y. 2022). The
district court concluded that the survey improperly directed respondents to an answer “to validate plaintiffs’ theory.” Id. at 288. Then, in the second question, Dr. Dennis asked consumers whether they would expect the product to be made using certain identified, but undefined, “chemicals,” which the district court found was an attempt to “to manipulate consumers into selecting the answer that plaintiffs preferred.” Id. at 289. Here, after careful consideration of the consumer perception survey in this case, the Court concludes that the survey questions are neither inappropriately leading nor manipulative. Unlike In re Kind, there is no controversial language that required further definition or risks tainting the survey’s results.
C. Relevance Finally, “in fulfilling this gatekeeping role, the trial court should look to the standards of Rule 401 in analyzing whether proffered expert testimony is relevant, i.e., whether it has any tendency to make the existence of any fact that is of consequence to the determination of the action more probable or less probable than it would be without the evidence.” Amorgianos, 303 F.3d at 265 (cleaned up). “Evidence need not be sufficient by itself to prove a fact in issue” in order to meet the “very low standard” of relevancy. United States v. Litvak, 889 F.3d 56, 68 (2d Cir. 2018) (quoting United States v. Abu-Jihaad, 630 F.3d 102, 132 (2d Cir. 2010)).
1. Defendant’s Allegations of Comcast Barring Dr. Johnson and Dennis’s Opinions As a preliminary matter, the Court first addresses whether the standard articulated in Comcast Corp. v. Behrend, 569 U.S. 27, 37 (2013) demands exclusion of Dr. Dennis and Dr. Johnson’s opinions for lack of relevance. Specifically, Defendant argues that the expert reports are not relevant because they do not isolate the price premium attributable to the allegedly deceptive interpretation of the PS “100% Natural Spring Water” label claim.49 In Comcast, the Supreme Court analyzed the propriety of class certification where the plaintiffs, subscribers to Comcast’s cable-television services, filed a class-action antitrust alleging four theories of antitrust impact. 569 U.S. at 31. The district court found that damages resulting from one of those theories could be calculated on a classwide basis and thus certified
49 ECF No. 735-1 at 18–29 (Johnson); ECF No. 738-1 at 18–21 (Dennis). the class, despite unrebutted testimony that the damages model “did not isolate damages resulting from any one theory of antitrust impact.” Id. at 32. Indeed, “the sole theory of liability that the district court determined was common in that antitrust action, overbuilder competition,
was a theory of liability that the plaintiffs’ model indisputably ‘failed to measure’ when determining the damages for that injury.” Sykes v. Mel S. Harris & Assocs. LLC, 780 F.3d 70, 88 (2d Cir. 2015) (cleaned up). In reversing the grant of class certification, the Comcast Court rejected the proposition that “any method of measurement [of damages] is acceptable so long as it can be applied classwide, no matter how arbitrary the measurements may be[,]” and instead held that “a model purporting to serve as evidence of damages in [a] class action must measure only those damages attributable to [a] theory” of injury advanced by the plaintiffs.
Comcast, 596 U.S. at 35–36. The Second Circuit has also clarified that, following Comcast, “[a]ll that is required at class certification is that the plaintiffs must be able to show that their damages stemmed from the defendant’s actions that created the legal liability.” Sykes, 780 F.3d at 88 (cleaned up). But Defendant raises many of the same arguments under Comcast here that it raised during summary judgment. A district court judge has already scrutinized Defendant’s arguments that Dr. Dennis and Johnson’s expert reports measured the wrong alleged
misrepresentation or failed to demonstrate a price premium derived from the alleged fraud, and rejected those arguments.50 In repacking the same arguments in its Daubert motions, Defendant has not provided a considered explanation as to why the earlier rulings in this case were wrong. Nor has Defendant justified an inconsistent result where these expert reports
50 Patane V, 761 F. Supp. 3d at 449–52. should be excluded when Judge Meyer found that these expert reports satisfied Comcast and created genuine disputes of fact sufficient to deny summary judgment. This Court has already declined to reconsider Judge Meyer’s well-reasoned summary judgment opinion,51 and will
reaffirm the reasoning here to conclude that Dr. Dennis and Johnson’s expert reports satisfies Comcast. Plaintiff has sufficiently shown that Dr. Johnson’s hedonic regression analysis and Dr. Dennis’s conjoint analysis purport to isolate the impact on Poland Spring pricing directly attributable to the label’s “spring water” representation and thus, they serve as common evidence of plaintiffs’ theory and do not run afoul of Comcast. 2. Defendant’s Allegations of Dr. Johnson’s Irrelevant Testimony on Statutory Damages Defendant argues that Dr. Johnson’s opinions interpreting statutes providing for statutory damages is irrelevant and will unfairly prejudice the jury.52 It is well established “that although an expert may opine on an issue of fact within the [factfinder’s] province, he may not give testimony stating ultimate legal conclusions based on those facts.” United States v.
Bilzerian, 926 F.2d 1285, 1294 (2d Cir. 1991). Here, Plaintiffs clarify that Dr. Johnson does not intend to opine on whether statutory damages are appropriate but would opine on statutory damages if they are available, depending on the outcome of claims administration and the class. Dr. Johnson’s opinions summing up the recoverable statutory damages should not be excluded under Daubert or Rule 702, as “[t]he mere use of hypotheticals does not usurp the [factfinder’s] function of applying the law to the facts of the case.” Id.; see also In re Scotts EZ Seed Litigation, No. 12-CV-4727, 2017 WL
51 Patane VI, 786 F. Supp. 3d at 487. 52 ECF No. 735-1 at 45–46. 3396433, *11 (S.D.N.Y. August 8, 2017) (finding damages calculations are admissible with respect to the calculation of statutory damages). But “[e]ven if admissible under Rule 702, expert testimony is still subject to exclusion
under Rule 403 if its probative value is substantially outweighed by the danger of unfair prejudice.’” United States v. Castillo, 924 F.2d 1227, 1232 n.9 (2d Cir. 1991) (cleaned up). The Court will exercise its discretion to consider that challenge on a more fulsome record at the motions in limine stage. 3. Defendant’s Allegations of Dr. Dennis’s Irrelevant Rank-Order Question Results As to Dr. Dennis, Defendant argues that his ranked preference question is not probative of materiality because the results conflate “preference” for a product with how consumers “value” the product in analyzing the results.53 But the Court finds that Dr. Dennis’s survey data is likely to help the trier of fact to determine a fact in issue, as required by Rule 702. In re Coffee, No. 21-CV-828, 2024 WL 4068778 (W.D. Mo. July 31, 2024), is illustrative. There,
the district court found Dr. Dennis’s opinion to be relevant even though additional proof would be needed to support the case that the defendants “are liable for labeling coffee canisters in a way that misrepresents how many cups of coffee a consumer can brew from the canisters’ contents.” Id. at *1, *5. Much like In re Coffee, Dr. Dennis’s resulting opinion here related to the rank-order question that consumers “prefer” the challenged product “can be used to support the conclusion that an inflated [label claim] would cause reasonable consumers to enter into a transaction.” 2024 WL 4068778 at *3, *4. Accordingly, the Court finds that Dr. Dennis’s
survey data/opinion is probative evidence of materiality.
53 ECF No. 738-1 at 40–43. IV. CONCLUSION For the reasons described above, it is hereby ordered that: (1) The motion to preclude the opinions and testimony of Dr. Johnson is denied, and (2) The motion to preclude the opinions and testimony of Dr. Dennis is denied.
SO ORDERED.
Hartford, Connecticut March 28, 2026
/s/Vernon D. Oliver VERNON D. OLIVER United States District Judge