In re Soc. Media Adolescent Addiction Litig., No. 4:22-md-03047-YGR (N.D. Cal. 2026)

and Dkt. No. 2292, [“Section 230 Mtn.”]. [p. 2] 1 The school district plaintiffs sue defendants on a theory that the design of defendants’ 2 platforms engenders compulsive and problematic use by students, causing disruption with which 3 the districts are forced to cope. At the motion to dismiss stage, defendants argued that the districts’ 4 negligence and nuisance claims were barred by Section 230 of the Communications Decency Act 5 of 1996, which provides certain protections to online platforms, and the First Amendment. The 6 Court explained that the SDs may proceed on a “core theory of injury [which] focuses on the 7 impact of compulsive use itself, irrespective of third-party content, defendants’ protected 8 publishing activity and defendants’ protected first-party speech.” (Mot. to Dismiss Order, (“MTD 9 Order”), Dkt. No. 1267 at 26.) Now at the summary judgment stage, the school districts adduce 10 evidence to support this theory, including the expert reports challenged here by defendants. 11 Plaintiffs’ GC Experts put forth opinions to show that defendants negligently designed their 12 platforms in ways that cause harmful use.3 Plaintiffs’ SD Expert opinions, in turn, seek to show 13 that each district experienced harm due to their students’ compulsive social media use. 14 II. LEGAL STANDARD 15 Federal Rule of Evidence 702 provides: 16 A witness who is qualified as an expert by knowledge, skill,

defendants also challenge. [p. 3] 1 WL 5604443, at *6 (E.D.Cal. 2015). Experts may also rely on “reports that would otherwise be 2 hearsay” for the basis of their opinion. Fannie Mae v. LaRuffa, 702 F. App’x 505, 507 (9th Cir. 3 2017). Additionally, evidence that “could be presented in an admissible form at trial” may be 4 considered in opposition to summary judgment. Fraser v. Goodale, 342 F.3d 1032, 1037 (9th Cir. 5 2003). As a result, “[h]earsay objections are often premature at summary judgment when asserted 6 by a moving party.” Moreno, 2015 WL 5604443, at *5. “[A] party opposing a motion for summary 7 judgment seeks a trial, not a verdict, and it stands to reason that if evidence may be converted to 8 admissible form for trial, it should not be excluded at summary judgment.” Id. 9 III. MOTION TO EXCLUDE SCHOOL DISTRICT EXPERTS 10 Defendants move to exclude the opinions of six SD experts4 and fourteen GC experts5. In 11 this Order, the Court addresses defendants’ arguments with respect to each SD expert although the 12 noted challenges (i.e. methodological, Section 230 and 1st Amendment) contain significant overlap 13 among the various experts. 14 A. ROBERT KLEIN 15 Dr. Klein describes his assignment as one: 16 to design, execute, and analyze a market research survey to measure

causation experts, excepting Dr. Lauren Hale and Dr. Bradley Zicherman who are AG experts only. [p. 4] 1 market research. (Id. ¶¶ 5-8.) By way of overview, Klein prepared and implemented a survey that 2 asked teachers to report the number of minutes, during a typical day, diverted from teaching due to 3 “social media (e.g., Facebook, Instagram, Snapchat, TikTok, YouTube, etc.)” at various intervals 4 from 2014 to the 2024-2025 school year. (Id. ¶¶ 10, 12.) Klein then issued reports as to each 5 bellwether school district that estimates, using the survey results, the time that teachers in each 6 district spent dealing with social media instead of teaching the scheduled curriculum. Id. These 7 estimates form the basis for part of the school district plaintiffs’ “lost time” damages: that is, the 8 opportunity cost in dollars of which the school districts were deprived of their employees’ value.6 9 Defendants do not challenge Klein’s qualifications to conduct surveys. Instead, defendants 10 level three categories of objections to exclude his opinions in their entirety: (1) a failure to account 11 for this Court’s MTD order requiring that liability be based on certain actionable features of 12 defendants’ platforms; (2) retrospective reporting; and (3) lack of representativeness. 13 The Court addresses each. 14 1. Failure to Account for This Court’s MTD Order 15 Defendants assert that the questions used in the survey (1) ask about “social media” 16 generally rather than specifically, i.e. to each of defendants’ particular platforms and (2) do not 17 exclude from consideration third-party content or the barred features. As a result, defendants 18 contend, Klein’s report is not relevant and could not help a jury. 19 As to the first, having reviewed the survey, the Court confirms that it does in fact name 20 each defendant’s platforms but then adds the words “e.g.,” and “etc.,” and provides a general 21 category title (“social media”). Defendants assert that these additions invite consideration of other 22 social media platforms. Plaintiffs respond that Dr. Klein phrased the question as he did to maintain 23 consistency with the survey’s other questions and thus avoid revealing the purpose of the survey, 24 which could bias responses.

other experts relying on witness affidavits. [p. 5] 1 “[P]urported flaws in survey design . . . will usually go to the weight a jury accords the 2 survey, not whether the jury can be shown it in the first place.” Johnson v. Nissan N. Am., Inc., 3 2022 WL 2869528, at *8 (N.D.Cal. 2022). Defendants offer no reason why Klein’s choice of 4 phrasing “renders the survey unreliable as a matter of law, or why [it] [wa]s so wrong that the jury 5 cannot be trusted to evaluate the merits of [their] objections.” Id. This objection is overruled. 6 As to the second, and as described in the Court’s Order Denying Motion for Summary 7 Judgment as to Breathitt school district (Dkt. No. 2728, “Breathitt MSJ Order”), plaintiffs may 8 proceed on their theory that defendants’ negligent design of their platforms causes students’ 9 compulsive social media use which, in turn, results in costs to the school district. Dr. Klein’s 10 evidence speaks to the effects of students’ compulsive use of social media, which is critical to 11 plaintiffs’ theory; that it does not tease out causation as to each actionable feature does not bar 12 admissibility. See Primiano v. Cook, 598 F.3d 558, 565 (9th Cir. 2010) (“Reliable expert testimony 13 need only be relevant, and need not establish every element that the plaintiff must prove, in order 14 to be admissible”); In re Toyota Motor Corp. Unintended Acceleration Mktg., Sales Pracs., & 15 Prods. Liab. Litig., 978 F.Supp. 2d 1053, 1066–67 (C.D.Cal. 2013) (“[N]o single expert provides a 16 self-sufficient opinion that [defendants’ conduct] in fact caused the [plaintiff’s harm]. This is not 17 dispositive.”). 18 Returning to defendants’ “relevance” objection, the Court does not find that the two 19 challenges vitiate the probative value of the survey in its entirety. That defendants may cross 20 examine on these issues is valid, but the issue is one of weight, not admissibility. The motion on 21 this ground is denied. 22 2. Retrospective Reporting 23 Defendants next argue that Klein’s method of asking respondents to recall how many 24 minutes they spent dealing with social media years ago is unreliable, as people cannot recall such 25 details with precision. Plaintiffs respond that retrospective surveys are standard and accepted in 26 research literature outside of litigation, and that the Klein survey also included instructions not to 27 guess and a “don’t know” answer to avoid recall bias. [p. 6] 1 Defendants’ cited authority fails to show that the Ninth Circuit prohibits admitting 2 retrospective surveys. Indeed, one of defendants’ cited cases acknowledges that survey respondents 3 “might reliably remember” certain retrospective events. In re: Autozone, Inc., 2016 WL 4208200, 4 at *19 (N.D.Cal. 2016). Where the survey respondents, as with the teachers here, cannot be 5 expected to have kept records of the relevant subject matter over many years, retrospective “survey 6 evidence is relevant to give an approximation of damages.” Alcantar v. Hobart Serv., 2013 WL 7 156530, at *2 (C.D.Cal. 2013) (unreasonable to expect employees to keep records of every meal 8 break provided). That is the purpose for which plaintiffs offer the survey here. As above, these 9 disputed methodological issues go to weight, not admissibility. This objection is overruled. 10 3. Lack of Representativeness and Hearsay Objections 11 Defendants argue that only a fraction of the districts’ teachers responded to Klein’s surveys 12 and that Klein has no way of knowing whether those who responded were representative of the full 13 population of teachers in the districts. The Third Edition of the Reference Manual on Scientific 14 Evidence describes high survey response rates as “desirable,” but nevertheless notes that 15 “surprisingly comparable results have been obtained in many surveys with varying response rates, 16 suggesting that surveys may achieve reasonable estimates even with relatively low response rates.” 17 Reference Manual on Scientific Evidence (3d ed.) (2011) at 384-85. Regardless, “any problems 18 with the response rate affect the weight, and not the admissibility of the study.” Alcantar, 2013 WL 19 156530, at *4. The Court declines to grant the motion on this ground. 20 Defendants further assert that Klein’s survey failed to sufficiently account for “non-21 response bias,” which they claim is contrary to the Third Edition of the Reference Manual for 22 Scientific Evidence and grounds for exclusion. In general, critiques regarding such “technical 23 inadequacies,” including survey methodology, “bear on the weight of the survey, not its 24 admissibility." Fortune Dynamic, Inc. v. Victoria's Secret Stores Brand Management, Inc., 618 25 F.3d 1025, 1036 (9th Cir. 2010); see also Clicks Billiards, Inc. v. Sixshooters, Inc., 251 F.3d 1252, 26 1263–64 (9th Cir. 2001) (same); Prudential Ins. Co. of America v. Gibraltar Financial Corp. of 27 California, 694 F.2d 1150, 1156 (9th Cir. 1982) (same). “On the other hand, substantial [p. 7] 1 deficiencies in the design or execution of a survey of individuals is grounds for its complete 2 exclusion.” In re: Autozone, Inc., 2016 WL 4208200, at *16 (N.D.Cal. 2016) (internal quotations 3 omitted). To that end, defendants cite several cases excluding survey results where the expert failed 4 to address non-response bias, along with other significant methodological flaws. See Jimenez v. 5 Allstate Insurance Company, 2019 WL 13088814, at *18 (C.D.Cal. 2019) (expert’s failure to 6 persuade the court that survey responses were representative of class “weigh[ed] in favor of 7 granting” Rule 702 motion); Autozone, 2016 WL 4208200, at *18-20 (granting class decertification 8 motion where problems with the survey were “fundamental,” including that the expert “did not 9 adequately account for the possibility of nonresponse bias.”) 10 The Court finds these non-binding cases distinguishable and not persuasive here. In 11 Autozone, the court’s concerns about non-response bias were driven by a particular “type of non-12 response called selected non-response,” that is, that “some of the persons called actively choose not 13 to respond for particular reasons.” 2016 WL 4208200, at *18. Hundreds of respondents actively 14 refused to participate in the Autozone plaintiffs’ survey, such that “refusals outnumbered surveys 15 responded to by almost two-thirds,” an unexplained red flag that raised concerns of systemic errors. 16 Id. The Autozone court was also concerned with “self-interest bias,” the risk that survey results 17 were biased by the incentive for class members to respond in ways to maximize their monetary 18 award. The survey respondents here, in contrast, are not class members or plaintiffs and lack that 19 incentive. Likewise, Jimenez involved concerns not present here, including that 86% of 20 respondents—who were also potential class members—were aware of the lawsuit, and that the 21 expert appeared to have inappropriately combined responses to several questions in making his 22 tabulations. 2019 WL 13088814, at *23-24. 23 As the Third Edition of the Reference Manual on Scientific Evidence notes, “the key is 24 whether nonresponse is associated with systematic differences in response that cannot be 25 adequately modeled or assessed.” Reference Manual on Scientific Evidence (3d ed.) (2011) at 385. 26 The Court does not view the risks of “systematic differences in response” here as particularly high 27 or concerning. As discussed above, the concerns about self-interest bias are not present. Klein also [p. 8] 1 took various steps to ensure reliability, including conducting the survey in a “double-blind” manner 2 to avoid bias, instructing respondents not to guess to reduce the likelihood of non-meaningful 3 responses, and pretesting the survey to ensure teachers reliably understood the questions. (Dkt. No. 4 2407-8, Yeates Decl., Ex. 6 ¶¶ 16, 39, 17). To the extent that defendants contend more is required, 5 such critiques are appropriate for cross-examination. As above, the Court concludes that the 6 methodological issues raised by defendants here are not grounds for exclusion.7 7 B. BRYCE WARD 8 Dr. Ward received his Ph.D. in economics from Harvard University and has worked for 9 decades in the fields of economic and statistical analysis. (Dkt. No. 2407-35, Yeates Decl., Ex. 33 10 ¶ 2.) Dr. Ward describes his assignment as one to “calculate the amount of money districts 11 expended while teachers, counselors, and/or administrators addressed issues caused by social 12 media in the school environment.” (Id. ¶ 5.) By way of overview, Dr. Ward proceeded by 13 multiplying estimates of “diverted time” due to social media—relying on Dr. Klein’s teacher 14 survey and witness affidavits—by the “compensation or cost of time for the relevant staff in the 15 district” to calculate the districts’ “past losses.” (Id. ¶¶ 13, 34.) His “analysis of past damages for 16 diversion thus equals the amount of time diverted times the cost of that time.” (Id. ¶ 12.) 17 1. Lost Time Damages Barred 18 Defendants maintain that the law of each bellwether plaintiff’s state exclude from damages 19 any lost time unaccompanied by an added financial expenditure. As applied here, they assert that 20 because Dr. Ward calculated damages based on the value of teacher and administrator time 21 diverted from other tasks, and not involving new costs to plaintiffs (i.e., no additional staff hired), 22 the damages’ calculations are irrelevant. Defendants maintain that because the school districts 23

at trial to admissible form. Supra Section II. [p. 9] 1 would have paid the same salaries regardless, the districts’ “lost time” damages are not recoverable 2 and thus the opinion should be excluded as irrelevant. 3 Contrary to defendants’ position, none of the school districts are barred under state law 4 from recovering for their employees’ lost and diverted time.8 See, e.g., Lurry v. PharMerica Corp., 5 2024 WL 2965642, at 3 (W.D.Ky. 2024) (“lost time and opportunity costs” recoverable under 6 Kentucky law); Aaron's, Inc. v. MDC Grp., Inc., 2010 WL 11505919, at *7 (N.D.Ga. 2010) (under 7 Georgia law, damages based on employee time “diverted from their normal job duties” not barred 8 even where employee salaries were “a cost that [plaintiff] would have incurred regardless”); 9 Comdyne I, Inc. v. Corbin, 908 F.2d 1142, 1146 (3d Cir. 1990) (“the value of the time [plaintiffs’] 10 employees expended in attempting to reduce or avert the damage” from defendants’ conduct 11 recoverable under New Jersey law); Charleston Lumber Co. v. Miller Hous. Corp., 458 S.E.2d 431, 12 437 (S.C.Ct.App. 1995) (“lost employee time” a compen-sable damage under South Carolina law); 13 Under Armour, Inc. v. Ziger/Snead, LLP, 232 Md. App. 548, 557 (2017) (“diverted employee 14 time” a compensable loss in Maryland); PivotHealth Holdings LLC v. Horton, 2025 WL 1865788, 15 at *2 (D.Ariz. July 7, 2025) (employee time spent responding to defendants’ conduct sufficiently 16 concrete damages under Arizona law); see also State v. Rouse, 647 N.W.2d 286, 289 (Wis.Ct.App. 17 2002) (collecting cases and noting that “[m]ost courts that have considered the issue have 18 concluded that a plaintiff can recover the value of employees’ lost services as damages”); Convoy 19 Co. v. Sperry Rand Corp., 672 F.2d 781, 785 (9th Cir. 1982) (same). 20 Accordingly, the Court declines to exclude Dr. Ward’s opinions on the basis that lost 21 employee time is not a cognizable form of damages. The motion on this ground is denied. 22 2. Opinions Based on Unreliable Inputs 23 Defendants argue that because Dr. Ward unquestioningly accepted inadmissible inputs (Dr. 24 Klein’s survey and school district employee affidavits) to his calculations, his opinions must be 25 excluded. As discussed above, the Court finds that the Klein survey is admissible. Supra at III(A).

same. [p. 10] 1 “An expert is permitted to rely on the opinion of another expert,” moreover, without needing to 2 independently validate it. Int'l Swimming League, Ltd. v. World Aquatics, 2025 WL 3257200, at *8 3 (N.D.Cal. 2025). Likewise, and as discussed in the Breathitt MSJ Order, defendants have not 4 carried their burden to demonstrate that plaintiffs’ employee affidavits are barred from forming the 5 basis of an expert’s opinions.9 As such, defendants’ contentions “as to the reliability of the facts 6 and opinions [Dr. Ward] relied upon . . . are questions of weight and credibility for a jury, not 7 questions of reliability for admissibility.” Int’l Swimming League, 2025 WL 3257200 at *8. 8 C. JEFFREY MEYERS 9 Mr. Meyers describes his assignment as one to calculate each school district’s “past 10 damages relating to certain out-of-pocket costs attributable to the alleged improper actions of 11 Defendants.” (Dkt. No. 2407-65, Yeates Decl., Ex. 63 ¶ 23.) Meyers received his Bachelor of 12 Science degree in 2003 from Louisiana State University and a Master of Science degree in 13 Mathematics in 2004 from the University of New Orleans. (Id., Appx. A). Meyers was designated a 14 certified Master Analyst in Financial Forensics by The National Association of Certified Valuators 15 and Analysts in 2010. (Id. ¶ 4.) Meyers has also served as a financial expert in dozens of litigation 16 matters, including in the Deepwater Horizon MDL for which he calculated “lost profits and 17 economic damages.” (Id. ¶ 7.) Defendants do not challenge Meyers’ qualifications. 18 Rather, defendants argue that neither Meyers nor the affiants upon which he relied tied 19 those expenditures specifically to defendants’ actionable conduct but simply to “social media” in 20 general. Meyers issued reports for each school district regarding the value of those direct damages, 21 which he calculated by multiplying the districts’ relevant expenditures by the portion of those 22 expenditures attributable to student social media use (the “Allocation Percents”). (Id. ¶ 23.) Meyers 23 relied on affidavits from school district administrators, which provided those Allocation Percents 24 as to the vendor costs. (Id. ¶ 21.)

be converted at trial to admissible form. Supra Section II. [p. 11] 1 As discussed above and in the Breathitt MSJ Order, plaintiffs may proceed on their theory

directly to this theory. (Dkt. No. 2407-82, Yeates Decl., Ex. 80 ¶ 35) (research finds social media is [p. 12] 1 a “significant source of distraction for students”); (id. ¶ 36) (average student use of social media 2 during school day is “problematic”); (id. ¶¶ 38-39) (excessive phone use during class time driven 3 by social media platforms including YouTube and Snapchat); (id. ¶¶ 77-83) (describing burden of 4 compulsive social media use on school resources). As already discussed, “[r]eliable expert 5 testimony need only be relevant, and need not establish every element that the plaintiff must prove, 6 in order to be admissible.” Primiano, 598 F.3d at 565 (9th Cir. 2010). 7 Accordingly, the Court concludes Section 230 and the First Amendment do not render 8 Hoover’s opinions irrelevant and unhelpful to a jury. 9 2. Specific Causation Opinions Not Based on Sufficient Facts or Data 10 Defendants seek to exclude all of Hoover’s opinions on the basis of a lack of data. More 11 specifically, defendants contend Hoover (a) lacks data on (i) how much time each districts’ 12 students spend on social media; and (ii) how many students in the districts are addicted to social 13 media, depressed, bullied, or suffered diminished social skills; and (b) relies on unreliable or 14 irrelevant sources such as plaintiffs’ fact sheets, “School Health Assessment and Performance 15 Evaluation” profiles (an assessment tool), and “key informant” interviews with teachers and 16 administrators of each district. As a result, defendants argue, Hoover’s causation opinions are not 17 based on independent data but instead regurgitate plaintiffs’ theory. Defendants analogize Hoover’s 18 “key informant” interviews to Therasense, Inc. v. Becton, Dickinson & Co., which decried that 19 “[o]ne of the worst abuses in civil litigation is the attempted spoon-feeding of client-prepared and 20 lawyer-orchestrated ‘facts’ to a hired expert who then ‘relies’ on the information to express an 21 opinion.” 2008 WL 2323856 (N.D.Cal. 2008). 22 The Court disagrees that Therasense is an apt analogy. Hoover explains that her opinions 23 rest on a multiplicity of inputs, including her own research on school mental health systems, review 24 of the scientific literature, graduate-level training in research and medicine, professional experience

as a director of the National Center for School Mental Health, interactions with educational [p. 13] 1 social media affected the districts is hardly surprising, much less nefarious. Hoover’s key 2 informant interviews, moreover, describe student compulsive use of social media and resulting 3 distraction and anxiety—evidence that is also offered by affidavits and deposition testimony of the 4 school districts’ teachers and administrators. Cf. Therasense, 2008 WL 2323856 at *2. (“Coming 5 from the expert alone. . . any opinion based on such untested and partisan foundation is not based 6 on sufficient facts and data within the meaning of Rule 702”) (emphasis supplied). Hoover is not 7 barred on hearsay grounds, either, from considering those interviews when forming her opinion. 8 See Fannie Mae v. LaRuffa, 702 F. App'x 505, 507 (9th Cir. 2017) (experts may rely on hearsay for 9 the basis of their opinions). The Court concludes that Hoover’s specific causation opinions are 10 based on sufficient facts and data for the purposes of Rule 702. 11 3. Abatement Plan Not Appropriate Remedy 12 Defendants claim that Hoover’s strategic plan is irrelevant and must be excluded because it 13 cannot support a claim for abatement. The Court agrees that Hoover’s plan cannot be used for 14 abatement purposes; however, because the plan supports future damages, her opinions are 15 admissible for that purpose. See Breathitt MSJ Order at 22-23.

a reliable basis in the knowledge and experience of the relevant discipline.” Primiano, 598 F.3d at 565. Here, Hoover’s strategic plan is validly connected to the task at hand—that is, her plan describes a series of interventions focused on student mental health and well-being in relation to social media. (See Dkt. No. 2407-82, Yeates Decl., Ex. 80 ¶¶ 23-36). Hoover applied her [p. 14] 1 knowledge and experience of the relevant discipline, through repeated citation to scientific 2 literature and drawing upon her professional qualifications as a licensed psychologist and director 3 of a center for school mental health policy. She hinges the 15 years on the effective lifecycle of a 4 student’s full school experience as well as with reference to “evidence from large-scale education 5 reform and mental health implementation science [which] indicates that sustainable system change 6 takes 10-15 years.” (Hoover Rep. ¶ 112.) Defendants’ disagreement with this metric concerns the 7 weight of her opinion and does not bar admissibility. See Primiano, 598 F.3d at 564 (“Shaky but 8 admissible evidence is to be attacked by cross examination, contrary evidence, and attention to the 9 burden of proof, not exclusion.”). 10 E. DOUGLAS LESLIE 11 Dr. Leslie describes his assignment as one to: “calculat[e] the present value of future costs 12 associated with implementation of the plan . . . recommended by Dr. [Sharon] Hoover” for each 13 school district. (Dkt. No. 2407-14, Yeates Decl., Ex. 96 ¶ 12.) Leslie received a Bachelor of 14 Science degree from the University of Virginia in 1990 and Ph.D. in Economics from Yale 15 University in 1998. (Id. ¶ 4.) Leslie has been a professor at the Pennsylvania State University for 16 over 15 years and holds appointments in the Department of Public Health Sciences and Department 17 of Psychiatry & Behavioral Health at the Penn State College of Medicine. (Id.) Leslie has 18 published hundreds of articles and sits on the editorial board of, among other academic journals, 19 the Journal of Mental Health Policy and Economics. (Id. ¶ 5.) Dr. Leslie’s opinion is limited to 20 costing out Dr. Hoover’s strategic plan. 21 Defendants argue that Hoover’s opinions should be excluded, which in turn requires 22 Leslie’s opinion to be excluded as well. As noted above, however, Hoover’s plan is admissible 23 under Rule 702. In general, “[a]n expert is permitted to rely on the opinion of another expert” and 24 is not required to independently validate it. Int’l Swimming League, 2025 WL 3257200 at *8. 25 Defendants may raise criticisms of Hoover’s plans, and Leslie’s calculations of those plans’ costs, 26 upon cross-examination. In re Apple iPhone Antitrust Litig., 2024 WL 5701895 at *11-13 [p. 15] 1 (N.D.Cal. 2024) (“Objections to an expert’s inputs may be raised on cross-examination.”) The 2 motion on this ground is denied. 3 F. BRIAN OSBORNE 4 Uniquely among the SD experts, Dr. Osborne offers an opinion based on his “experience in 5 public education, including service as a teacher, principal, superintendent, and consultant” and as a 6 “professor of practice,” rather than a traditional academic researcher, to offer what he describes as 7 an “experience-based account of what is happening in schools.” (Dkt. No. 2407-110, Yeates Decl., 8 Ex. 108 ¶¶ 2, 20.) Osborne holds an Ed.D. in Administration, Planning, and Social Policy from the 9 Harvard Graduate School of Education and an M.A. in Administration, Planning, and Social Policy 10 from Harvard University. (Id. ¶¶ 11, 12.) Osborne has over 25 years of experience in public 11 education, including as a teacher and superintendent of schools in New York and New Jersey. (Id. 12 ¶ 6.) Osborne currently instructs and consults with educators across New Jersey and Pennsylvania, 13 spending approximately 40-50 days a year in schools observing classroom instruction. (Id. ¶ 13.) 14 The Court addresses defendants’ five challenges. 15 1. Failure to Account for This Court’s MTD Order 16 As with the other expert opinions, defendants object that Osborne fails to offer an opinion 17 as to defendants’ specific platforms or the actionable features and opines only on social media in 18 general. For the same reasons previously articulated, the objection is overruled.

practical experience in a particular field, not just through academic training”). Osborne’s opinions are based not only on discussions with teachers or administrators from plaintiffs’ school districts but from his considerable work in the industry including his understanding of national trends [p. 16] 1 regarding social media in schools. That some of this may be hearsay is not a bar to his opinions, 2 although it would be a bar if he attempted to regurgitate the hearsay at trial. Supra Section II. As 3 such, Osborne is not required to exclude from consideration, in forming his opinions, conversations 4 he has had with other educators throughout his years of practice. That Osborne’s opinions are not 5 based on interviews or data from the six bellwether plaintiff school districts does not render them 6 inadmissible, either. Id. at 1430 (“Nor is the expert required to be personally familiar with the facts 7 or data of the particular case about which he is testifying; an expert opinion may be based on any 8 type of evidence commonly used by experts in the field.”) As a result, his opinions are not barred 9 based on a lack of reliable methodology or on hearsay grounds. 10 3. AI-Generated Citations 11 Defendants contend that Osborne admitted at deposition that articles upon which he 12 purportedly relied do not exist, and that other incorrect citations were miscites generated by AI. 13 Plaintiffs respond that none of the academic articles were actually false, but rather that citations 14 were merely incorrectly formatted based on use of an AI citation tool, which was subsequently 15 corrected. The Court declines to exclude Osborne’s opinions on this basis. Defendants may, 16 however, raise the issue on cross-examination.

excluded if proffered at trial in the manner identified in the report. Other portions of Osborne’s [p. 18] 1 report likewise contain statements that stray far from expertise and into argument. Id. ¶¶ 88, 89, 90, 2 101, 103, 107, 108, 110. 3 Three, and to be explicit, an expert can rely on hearsay to form his opinion but may not 4 regurgitate it. For example, Osborne repeats information and quotes from transcripts of depositions 5 of several school administrators in this litigation. (Yeates Decl., Ex. 96 ¶¶ 56-60, 62, 71, 80.) At 6 trial, Osborne may be permitted to opine that factual information which is independently offered 7 into evidence is consistent with or supports his opinions. He cannot, however, simply restate 8 portions of deposition transcripts as fact. Fed.R.Evid. 703. Similarly, Osborne cannot serve as a 9 conduit for facts and data from external studies and reports, as several portions of his report 10 attempt to do. (See, e.g., id. ¶¶ 61 (describing news stories about viral challenges), 69 (relaying the 11 results of academic studies about social media), 72 (same).) 12 Four, an expert’s testimony must also be reasonably based in his field of knowledge. 13 Fed.R.Evid. 703. Osborne states that students “fear being recorded, mocked, or turned into a 14 meme,” but is not a psychologist nor claims unique insight into the minds and thought processes of 15 students. Id. ¶ 79. Because he is not able to testify to those students’ motivations, such statements 16 could not be offered as testimony. 17 IV. CONCLUSION 18 Defendants’ motion is DENIED. 19 This terminates Dkt. No. 2291 in Case No. 22-md-03047. 20 IT IS SO ORDERED. 21 22