Unknown Party v. Arizona Board of Regents

Opinion

Case 2:18-cv-01623-DWL Document 250 Filed 11/18/22 Page 1 of 46

1 WO 2 3 4 5 8 9 Unknown Party, No. CV-18-01623-PHX-DWL 10 Plaintiff, ORDER 11 v. 12 Arizona Board of Regents, et al., 13 Defendants. 14 15 Pending before the Court are a pair of related motions filed by Defendant Arizona 16 Board of Regents (“ABOR”): (1) a motion to exclude the opinions and testimony of Dr. 17 Lance Kaufman (Doc. 159); and (2) a motion for reconsideration of the Court’s August 30, 18 2022 summary judgment ruling (Doc. 227). For the following reasons, both motions are 19 denied. 21 The parties are familiar with the background details of this case, so only a brief 22 recap is necessary here. 23 In April 2016, Plaintiff John Doe, then a student-athlete at Arizona State University 24 (“ASU”), and another man had a three-way sexual encounter with ASU student Jane Roe 25 at an off-campus party. Roe immediately reported the incident to the police, who declined 26 to pursue criminal charges against Doe after reviewing videotape footage of the incident 27 (which the other male participant had surreptitiously created). Several months later, Roe 28 reported the incident to ASU, claiming she had been too intoxicated to consent. Doe was Case 2:18-cv-01623-DWL Document 250 Filed 11/18/22 Page 2 of 46

1 suspended by ASU and, after an investigation, expelled for violating various provisions of 2 the ASU Student Code of Conduct, including provisions related to sexual misconduct. 3 In this action, Doe initially asserted an array of claims against an array of 4 defendants, but his sole remaining claim is a Title IX claim against ABOR. Additionally, 5 as this case was proceeding, Doe prevailed in a separate state-court action in which he 6 sought review of ASU’s expulsion decision. There, the Arizona Court of Appeals held that 7 the sexual misconduct findings against Doe were “not supported by substantial evidence” 8 and thus vacated the expulsion order. Doe v. Ariz. Bd. of Regents, 2019 WL 7174525, *9 9 (Ariz. Ct. App. 2019). 10 Following the close of discovery, ABOR filed a motion for summary judgment on 11 Doe’s Title IX claim (Doc. 155) and both sides filed motions to exclude or limit expert 12 testimony, including ABOR’s pending motion regarding Dr. Kaufman (Docs. 158, 159, 13 161, 164, 165). 14 The Court first addressed ABOR’s summary judgment motion, concluding in an 15 August 30, 2022 order that summary judgment should be denied because “Doe’s proffered 16 evidence could lead a reasonable juror not only to conclude that his disciplinary proceeding 17 was marred by an array of procedural irregularities, but also that ASU’s disciplinary 18 process generates statistical anomalies that raise an inference of gender bias.” (Doc. 209 19 at 60.) Although this conclusion was premised on a finding that “Dr. Kaufman’s statistical 20 evidence is relevant and creates a genuine issue of material fact about whether ASU’s Title 21 IX disciplinary process was infected by gender bias,” the Court acknowledged that ABOR 22 had raised various challenges to the admissibility of Dr. Kaufman’s statistical evidence in 23 one of its pending exclusion motions. (Id. at 52-54.) Accordingly, during oral argument, 24 the Court clarified that “I plan to address the Daubert motions in the near future. And if 25 for whatever reason I were to exclude Kaufman’s opinions, at that point [ABOR] could 26 come back and seek reconsideration of the summary judgment [ruling].” (Doc. 228 at 47.) 27 On August 30, 2022, the Court also issued a tentative ruling addressing the parties’ 28 motions to exclude expert testimony. (Doc. 210.) It included a section addressing ABOR’s

-2- Case 2:18-cv-01623-DWL Document 250 Filed 11/18/22 Page 3 of 46

1 motion to exclude Dr. Kaufman. (Id. at 15-25.) However, before oral argument, ABOR 2 filed the pending motion for reconsideration of the August 30, 2022 summary judgment 3 order. (Doc. 227.) As discussed in more detail below, the reconsideration motion includes 4 various arguments regarding the admissibility and evidentiary value of Dr. Kaufman’s 5 statistical evidence. (Id. at 2-10.) Accordingly, in the final version of the order addressing 6 the parties’ motions to exclude expert testimony, which was issued on September 27, 2022, 7 the Court explained that it would “defer ruling on the motion to exclude Dr. Kaufman until 8 it reviews Doe’s response to the motion for reconsideration . . . [which] will also serve as 9 a supplemental brief regarding the motion to exclude Dr. Kaufman.” (Doc. 223 at 1 n.1.) 10 On October 13, 2022, Doe filed his response/supplemental brief. (Doc. 241.) 11 On October 25, 2022, ABOR filed a reply. (Doc. 247.)1 13 I. ABOR’s Motion To Exclude Dr. Kaufman (Doc. 159) 14 A. Legal Standard 15 “The party offering expert testimony has the burden of establishing its 16 admissibility.” Bldg. Indus. Ass’n of Wash. v. Wash. State Bldg. Code Council, 683 F.3d 17 1144, 1154 (9th Cir. 2012). Rule 702 of the Federal Rules of Evidence governs the 18 admissibility of expert testimony. It provides: 19 A witness who is qualified as an expert by knowledge, skill, experience, training, or education may testify in the form of an opinion or otherwise if: 20 (a) the expert’s scientific, technical, or other specialized 21 knowledge will help the trier of fact to understand the evidence 22 or to determine a fact in issue;

23 (b) the testimony is based on sufficient facts or data; (c) the testimony is the product of reliable principles and methods; 24 and 25 (d) the expert has reliably applied the principles and methods to 26 the facts of the case. 27 1 ABOR’s request for oral argument on its motion for reconsideration is denied 28 because the issues are fully briefed and have been addressed in extensive detail in earlier orders and oral arguments. See LRCiv 7.2(f).

-3- Case 2:18-cv-01623-DWL Document 250 Filed 11/18/22 Page 4 of 46

1 As for the threshold requirement that an expert witness be qualified “by knowledge, 2 skill, experience, training, or education,” “Rule 702 contemplates a broad conception of 3 expert qualifications.” Hangarter v. Provident Life & Acc. Ins. Co., 373 F.3d 998, 1015 4 (9th Cir. 2004) (internal quotation marks and emphasis omitted). Years of relevant 5 experience can establish the necessary “minimal foundation.” Id. at 1015-16. “Disputes 6 as to the strength of [an expert’s] credentials . . . go to the weight, not the admissibility, of 7 his testimony.” Kennedy v. Collagen Corp., 161 F.3d 1226, 1231 (9th Cir. 1998) (first 8 alteration in original) (internal quotation marks omitted). 9 A district court’s decision to admit or exclude expert testimony is guided by a two- 10 part test that focuses on the opinion’s relevance and reliability. Daubert v. Merrell Dow 11 Pharm., Inc., 509 U.S. 579, 589 (1993). “The inquiry envisioned by Rule 702 is . . . a 12 flexible one.” Id. at 594. “The focus, of course, must be solely on principles and 13 methodology, not on the conclusions that they generate.” Id. at 595. 14 Evidence is relevant if it has “any tendency to make the existence of any fact that is 15 of consequence to the determination of the action more probable or less probable than it 16 would be without the evidence.’” Id. at 587 (quoting Fed. R. Evid. 401). “The Rule’s basic 17 standard of relevance thus is a liberal one.” Id. 18 The basic standard of reliability is similarly broad. “Shaky but admissible evidence 19 is to be attacked by cross examination, contrary evidence, and attention to the burden of 20 proof, not exclusion.” Primiano v. Cook, 598 F.3d 558, 564 (9th Cir. 2010). “Basically, 21 the judge is supposed to screen the jury from unreliable nonsense opinions, but not exclude 22 opinions merely because they are impeachable.” Alaska Rent-A-Car, Inc. v. Avis Budget 23 Grp., Inc., 738 F.3d 960, 969 (9th Cir. 2013). See also Fed. R. Evid. 702, advisory 24 committee’s note to 2000 amendment (“[P]roponents do not have to demonstrate to the 25 judge by a preponderance of the evidence that the assessments of their experts are correct, 26 they only have to demonstrate by a preponderance of evidence that their opinions are 27 reliable. . . . The evidentiary requirement of reliability is lower than the merits standard of 28 correctness.”) (alteration in original) (internal quotation marks omitted).

-4- Case 2:18-cv-01623-DWL Document 250 Filed 11/18/22 Page 5 of 46

1 Nevertheless, courts serve an important “gatekeeper” role when it comes to 2 screening expert testimony. Gen. Elec. Co. v. Joiner, 522 U.S. 136, 142 (1997). “Unlike 3 an ordinary witness, an expert is permitted wide latitude to offer opinions, including those 4 that are not based on firsthand knowledge or observation.” Daubert, 509 U.S. at 592 5 (citation omitted). “Presumably, this relaxation of the usual requirement of firsthand 6 knowledge . . . is premised on an assumption that the expert’s opinion will have a reliable 7 basis in the knowledge and experience of his discipline.” Id. This “general ‘gatekeeping’ 8 obligation . . . applies not only to testimony based on ‘scientific’ knowledge, but also to 9 testimony based on ‘technical’ and ‘other specialized’ knowledge.” Kumho Tire Co., Ltd. 10 v. Carmichael, 526 U.S. 137, 141 (1999). 11 The Court has “broad discretion,” both in deciding whether the evidence is reliable 12 and in deciding how to test for reliability. United States v. Hankey, 203 F.3d 1160, 1168 13 (9th Cir. 2000). In Daubert, the Supreme Court listed various factors that might apply, 14 including whether the expert’s technique or theory (1) can be tested; (2) has been peer 15 reviewed or published; (3) has a known or potential basis for error; and (4) is generally 16 accepted in the pertinent scientific community. 509 U.S. at 593-94. However, 17 “[t]he Daubert factors were not intended to be exhaustive nor to apply in every case.” 18 Hankey, 203 F.3d at 1168. In particular, “[t]he Daubert factors . . . simply are not 19 applicable to [testimony] whose reliability depends heavily on the knowledge and 20 experience of the expert, rather than the methodology or theory behind it.” Id. at 1169. See 21 also Fed. R. Evid. 702, advisory committee’s note to 2000 amendment (“Some types of 22 expert testimony will be more objectively verifiable, and subject to the expectations of 23 falsifiability, peer review, and publication, than others. Some types of expert testimony 24 will not rely on anything like a scientific method, and so will have to be evaluated by 25 reference to other standard principles attendant to the particular area of expertise.”). The 26 bottom line is that “[t]he trial judge in all cases of proffered expert testimony must find that 27 it is properly grounded, well-reasoned, and not speculative before it can be admitted. The 28 expert’s testimony must be grounded in an accepted body of learning or experience in the

-5- Case 2:18-cv-01623-DWL Document 250 Filed 11/18/22 Page 6 of 46

1 expert’s field, and the expert must explain how the conclusion is so grounded.” See Fed. 2 R. Evid. 702, advisory committee’s note to 2000 amendment. 3 B. The Kaufman Report 4 Dr. Kaufman possesses a Ph.D. in economics, has been recognized as an “expert in 5 the field of economics,” and has “over 15 years of experience performing economic 6 research and modeling.” (Doc. 159-1 at 22.) Dr. Kaufman has provided expert witness 7 testimony on “forecasting, production cost modeling, and labor costs,” and his hourly rates 8 range from $350 to $500 per hour. (Id. at 22.) 9 Dr. Kaufman’s opinions in this case can be divided in two categories: (1) opinions 10 relating to the enforcement of ASU’s student code, or “whether ASU’s sanctions are gender 11 neutral”; and (2) opinions relating to Doe’s alleged economic loss, or “the economic harm 12 resulting from [ABOR’s] unlawful expulsion of Doe.” (Id. at 5.) 13 1. Student Code Enforcement Disparity 14 Dr. Kaufman explained in his report that ABOR produced certain “summary 15 records,” which he refers to as “SCC Charge Data,” during discovery in this case that 16 contain information about ASU’s history of student code enforcement concerning alcohol 17 (“F-15”) and sexual misconduct (“F-23”) violations. (Id. at 6.) Dr. Kaufman stated that 18 the information was categorized by: (1) incident ID, (2) gender of respondent, (3) incident 19 type, (4) date and time of incident, (5) date of report, (6) “SCC Charge Code,” (7) Dean of 20 Students’ decision, (8) sanctions, (9) whether the decision was appealed, (10) hearing 21 board’s recommendation, and (11) University Vice President’s decision. (Id. at 6-7.) For 22 F-23 violations, the SCC Charge Data also noted (1) respondent’s student athlete status, 23 (2) gender of counter party, (3) charge investigator, (4) date of decision, and (5) date of 24 review of decision. (Id. at 7.) 25 At the outset of his report, Dr. Kaufman noted that the “SCC Charge Data provide 26 several different types of sanctions. These sanctions can be grouped and ranked in order 27 of severity. I rank warnings as the mildest sanction, expulsion and suspension as severe 28 sanctions, and all other sanctions as mild sanctions.” (Id.) Having developed this

-6- Case 2:18-cv-01623-DWL Document 250 Filed 11/18/22 Page 7 of 46

1 convention, Dr. Kaufman found that of 161 male F-23 respondents, 3 received no sanction, 2 79 received a mild sanction, 27 were suspended, and 52 were expelled. (Id.) Of 10 female 3 F-23 respondents, 1 received no sanction, 9 received a mild sanction, none received a 4 suspension, and none were expelled. (Id.) Dr. Kaufman concluded that “males tend to 5 receive more severe sanctions than females” and described his “statistical analysis of the 6 disparity between male and female Respondents.” (Id.) As a preface, Dr. Kaufman 7 explained: 8 In cases concerning the disparate impact on protected classes, outcomes for the protected class are compared to the rates for other individuals to 9 determine if the difference is statistically significant. 10 A disparity is considered statistically significant if it would occur so rarely 11 in a nondiscriminatory situation that we can rule out that it occurred by chance. . . . [T]he courts typically require a demonstration . . . that the 12 disparity is large enough that it would occur by chance less than 5 percent of 13 the time, or in less than one in 20 non-discriminatory events. To assess the evidence of gender bias in sanctions, respondents are classified 14 by two criteria, whether they are a member of the protected class or not, and 15 whether they were severely sanctioned or not. . . . If the disparity in sanction rates between two groups is not sufficiently large, or if the direction of the 16 disparity varies substantially across subgroups, the disparity will not be 17 statistically significant. 18 (Id. at 8.) 19 Dr. Kaufman also described the “Chi-Square Analysis,” which posits that, if males 20 are a certain proportion of all respondents, a gender-neutral process will lead to a similar 21 proportion of males actually being punished, as compared to the overall pool of students 22 punished. (Id. at 9.) Dr. Kaufman explained that this test “gives us a way of determining 23 what proportion is ‘high enough’ to conclude that sanctions were not gender-neutral.” (Id.) 24 Separately, Dr. Kaufman discussed the “Mantel Haenszel Peer-Group Analysis,” which is 25 a method of peer-group analysis that informs Dr. Kaufman’s use of respondent charges to 26 create peer groups with the same combination of charges, controlling for the type and 27 number of violations. (Id. at 9-10.) 28 Dr. Kaufman performed four analyses. (Id. at 10.) First, he compared the Dean of

-7- Case 2:18-cv-01623-DWL Document 250 Filed 11/18/22 Page 8 of 46

1 Students’ decisions for male and female respondents in cases involving F-15 and F-23 2 charges and determined that males are found in violation more often than females. (Id.) 3 Second, he compared the sanctions imposed against male and female F-23 respondents and 4 found that males receive severe sanctions at a higher rate than females. (Id.) Third, he 5 compared the severity of sanctions for F-15 only charges between males and females and 6 found that males are given more severe sanctions. (Id. at 10, 12.) Fourth, he compared the 7 sanctions imposed in F-23 cases involving male and female complainants and found that 8 incidents involving male complainants (unlike female complainants) never lead to severe 9 sanctions. (Id. at 10.) 10 Based on these analyses, Dr. Kaufman formed the following three opinions: 11 (1) male respondents are found in violation of F-15 and F-23 charges more often than 12 female respondents, at a rate that is statistically significant; (2) male F-23 respondents are 13 sanctioned with expulsion and suspension at a higher rate than female F-23 respondents, at 14 a rate that is statistically significant; and (3) F-23 charges involving male complainants 15 result in expulsion and suspension at lower rates than F-23 charges involving female 16 complainants. (Id. at 14.) Dr. Kaufman noted that “[w]here possible my analysis controls 17 for peer-groups by grouping respondents according to charge or combination of charges” 18 but acknowledged that “even within combination of charges there may be non-gender 19 factors that affect the severity of sanctions.” (Id.) 20 2. Economic Loss 21 Assuming that ABOR’s “actions interrupted Doe’s education and competitive 22 sports career [which] modified Doe’s career trajectory and diminished his lifetime earning 23 capacity,” Dr. Kaufman estimated “the economic impact by comparing Doe’s economic 24 situation in a hypothetical (‘But-for’) world absent [ABOR’s] unlawful actions with Doe’s 25 current expected economic situation (‘Actual’).” (Id. at 15.) Dr. Kaufman based his 26 calculations on some of the opinions (e.g., Doe was an elite freestyle wrestler, and coaching 27 at an NCAA Division I program requires elite competitive wrestling experience) of Owen, 28 who is another of Doe’s experts, and on Doe’s testimony that he planned to become a

-8- Case 2:18-cv-01623-DWL Document 250 Filed 11/18/22 Page 9 of 46

1 college wrestling coach after his competitive career. (Id. at 15-16.) 2 Based on those assumptions, and accounting for various other sources of income, 3 Dr. Kaufman concluded that Doe’s economic loss could range from $631,490 in a scenario 4 in which Doe became a Division II or III coach to $5,543,579 in a scenario in which Doe 5 became a Division I coach. (Id. at 20.) 6 C. Student Code Enforcement Disparity Opinions 7 In its motion to exclude and motion for reconsideration, ABOR raises an array of 8 arguments related to Dr. Kaufman’s student code enforcement disparity opinions. Below, 9 the Court has grouped those arguments into four broad categories. 10 It should be noted that many of the arguments related to Dr. Kaufman in ABOR’s 11 reconsideration motion seem to address both the relevance/admissibility of Dr. Kaufman’s 12 opinions and the evidentiary weight (if any) those opinions should have been afforded for 13 purposes of evaluating ABOR’s entitlement to summary judgment. Accordingly, in this 14 part of the order, the Court addresses ABOR’s arguments to the extent they are intended 15 as relevance/admissibility objections (because such objections go to whether Dr. 16 Kaufman’s opinions should be excluded, as requested in ABOR’s motion at Doc. 159). In 17 Part II of this order, the Court addresses the related but sometimes distinct question of 18 evidentiary weight (which goes to ABOR’s reconsideration request at Doc. 227). 19 1. F-23 Opinions—Failure To Consider Underlying Facts 20 a. ABOR’s Arguments 21 Two of Dr. Kaufman’s opinions are intended to establish the existence of gender- 22 based disparities in the sanctions imposed against ASU students found responsible for 23 violations of section F-23 (i.e., “sexual misconduct”) of the Code of Student Conduct. 24 Those opinions are that (1) male F-23 respondents are sanctioned with expulsion and 25 suspension at a statistically significant higher rate than female F-23 respondents; and (2) 26 F-23 charges involving male complainants result in expulsion and suspension at lower rates 27 than F-23 charges involving female complainants. (Doc. 159-1 at 14.) 28 Most of ABOR’s challenges focus on those opinions. (Doc. 159 at 5-11; Doc. 227

-9- Case 2:18-cv-01623-DWL Document 250 Filed 11/18/22 Page 10 of 46

1 at 2-10.) ABOR’s overarching position is that because “sexual misconduct” is defined 2 under section F-23 as encompassing a wide range of conduct,2 yet Dr. Kaufman made no 3 effort to control for the specific types of sexual misconduct at issue in the cases on which 4 his F-23 statistics and opinions are based, it follows that those statistics and opinions are 5 irredeemably flawed. (See, e.g., Doc. 159 at 6 [“In view of the wide range of actions that 6 fall within the definition of ‘sexual misconduct,’ a valid (reliable) comparison of sanctions 7 given to male and female students found responsible for violating F-23 of the Code cannot 8 be made without specific information as to the type of conduct at issue in each instance.”].) 9 In its motion to exclude, ABOR characterizes this flaw both as a problem of 10 reliability, because the analysis was not based on sufficient facts and data, and as a problem 11 of relevance, because only comparisons of “similarly situated” individuals are relevant in 12 a gender discrimination case. (Doc. 159 at 2-7; Doc. 196 at 1-7.) ABOR also contends 13 that it provided the necessary factual information to Doe during the discovery process, in 14 the form of a “Title IX Case Tracking Report” that “detailed the particular allegations 15 against students alleged to have violated [F-23] and other Code violations,” but Doe 16 inexplicably “failed to provide it to Dr. Kaufman for analysis.” (Doc. 159 at 7-8.) 17 Meanwhile, in its motion for reconsideration, ABOR elaborates on these arguments by 18 advancing what the Court perceives to be seven related reasons why Dr. Kaufman’s F-23 19 opinions should be considered flawed: 20 (1) Because the Ninth Circuit has recognized that statistical evidence is 21 “inherently slippery” and can be exaggerated and distorted, it was manifestly erroneous for 22 the Court to rely on Dr. Kaufman’s opinions as part of the basis for denying summary 23 judgment. (Doc. 227 at 2-3; Doc. 247 at 1.) 24 2 Under section F-23, “sexual misconduct” is defined as follows: “a. Sexual violence 25 and other non-consensual sexual contact—actual or attempted physical sexual acts perpetrated against a person by force or without consent; or b. Sexual harassment— 26 unwelcome conduct of a sexual nature that is sufficiently severe or pervasive as to create an intimidating, hostile, or offensive environment; or c. Other unwanted or non-consensual 27 sexual conduct including but not limited to indecent exposure, sexual exploitation or voyeurism, or non-consensual photographing or audio-recording or video-recording or 28 another in a state of full or partial undress or while engaged in sexual activity, or publishing or disseminating such materials.” (Doc. 189-2 at 16-17.)

- 10 - Case 2:18-cv-01623-DWL Document 250 Filed 11/18/22 Page 11 of 46

1 (2) Dr. Kaufman’s acknowledgement that non-gender factors may affect the 2 severity of sanctions is an “admission that renders [his] opinions unreliable and 3 inadmissible to establish intentional gender bias under binding Ninth Circuit precedent,” 4 as recognized in Gay v. Waiters’ and Dairy Lunchmen’s Union, Local No. 30, 694 F.2d 5 531 (9th Cir. 1982), Aragon v. Republic Silver State Disposal, Inc., 292 F.3d 654 (9th Cir. 6 2002), Rose v. Wells Fargo & Co., 902 F.2d 1417 (9th Cir. 1990), and I.V. v. Wenatchee 7 School District No. 246, 342 F. Supp. 3d 1083 (E.D. Wash. 2018). (Doc. 227 at 3-4; Doc. 8 247 at 2, 5.)3 9 (3) To properly exercise its “role as gatekeeper,” the Court must determine that 10 the comparators at issue in Dr. Kaufman’s statistics are “similarly situated” to Doe, as the 11 Ninth Circuit did in Freyd v. University of Oregon, 990 F.3d 1211 (9th Cir. 2021), and it 12 is an “abdicat[ion]” of that gatekeeping role to characterize the dispute over similarity as 13 an argument going to weight rather than admissibility. (Doc. 227 at 4-6.) 14 (4) Because Dr. Kaufman had access to the information necessary to determine 15 whether the alleged comparators were similar (i.e., the Title IX Case Tracking Report) but 16 failed to utilize it, “[t]his renders his analysis unreliable and insufficient” under Pottenger 17 v. Potlach Corp., 329 F.3d 740 (9th Cir. 2003). (Doc. 227 at 6; Doc. 247 at 2-8.) 18 (5) The Court’s determination in the summary judgment order that it was 19 reasonable for Dr. Kaufman to treat all F-23 violations as comparable because ASU itself 20 has decided that all F-23 violations are unified by a population of similarly situated 21 respondents in some material respect (Doc. 209 at 59-60) was factually incorrect, violates 22 the principle of party presentation (because Doe never made that argument), and would 23 “lead to arbitrary variations in the application and enforcement of Title IX” and 24 “incentivize universities to ensure the unnecessary expense of redrafting their codes of 25 conduct.” (Doc. 227 at 7-8.) 26 3 As an example of a non-gender factor that may affect the severity of sanctions, 27 ABOR points to “the ASU Procedures themselves,” which “provide that, in determining the appropriate sanction to impose, the Dean considers ‘any mitigating or aggravating 28 factors, including any prior violations of the Student Code of Conduct.’” (Doc. 227 at 6- 7.)

- 11 - Case 2:18-cv-01623-DWL Document 250 Filed 11/18/22 Page 12 of 46

1 (6) ABOR was not required to affirmatively demonstrate the factual dissimilarity 2 of the F-23 incidents underlying Dr. Kaufman’s statistics, and any argument to the contrary 3 would amount to impermissible burden-shifting. (Doc. 227 at 8-9.) 4 (7) A new exhibit prepared by ABOR, which is enclosed as Exhibit 3 to its 5 reconsideration motion, shows that the “great majority” of female F-23 respondents 6 discussed in Dr. Kaufman’s statistics “were alleged to have engaged in either sexual 7 harassment . . . or incident exposure” and that the one female alleged to have engaged in 8 non-consensual sexual conduct was still “not comparable to the circumstances here,” and 9 thus “[n]one of the females on which Kaufman relied are valid comparators, and 10 Kaufman’s calculation is irrelevant.” (Doc. 227 at 9-10; Doc. 247 at 8.) 11 b. Analysis 12 Dr. Kaufman’s opinions related to gender-based disparities in F-23 enforcement, 13 although impeachable in various respects, are sufficiently reliable and relevant to qualify 14 for admission at trial. 15 As an initial matter, and contrary to ABOR’s assertion that “Kaufman’s analysis is 16 not evidence of gender bias” (Doc. 247 at 1), the Ninth Circuit has suggested that 17 enforcement-disparity statistics such as Dr. Kaufman’s can support an inference of gender 18 bias in a Title IX case. Schwake v. Ariz. Bd. of Regents, 967 F.3d 940, 949 (9th Cir. 2020) 19 (plaintiff’s “allegations of a pattern of gender-based decisionmaking against male 20 respondents in sexual misconduct disciplinary proceedings” were “relevant” and bolstered 21 plausibility of Title IX claim); Doe v. Regents of the University of California, 23 F.4th 930, 22 938 (9th Cir. 2022) (“Doe alleges that the respondents in Title IX complaints that UCLA 23 decided to pursue from July 2016 to June 2018 were overwhelmingly male (citing specific 24 statistics for each of those years) . . . . Doe also alleges that the University ‘has never 25 suspended a female for two years based upon these same circumstances’ . . . . As we noted 26 in Schwake, these are precisely the type of non-conclusory, relevant factual allegations that 27 the district court may not freely ignore.”). ABOR makes much of the fact that Schwake 28 and Doe were decided at the motion-to-dismiss stage (Doc. 227 at 3; Doc. 247 at 2 n.1),

- 12 - Case 2:18-cv-01623-DWL Document 250 Filed 11/18/22 Page 13 of 46

1 but it doesn’t follow that they are inapplicable here. ABOR’s motion to exclude questions 2 whether Dr. Kaufman’s challenged opinions should be admitted under Rule 702, and one 3 of the admissibility considerations under Rule 702—which ABOR expressly disputes in its 4 motion papers—is relevance. The Court interprets Schwake and Doe as recognizing that 5 statistical evidence establishing a pattern of gender-based decisionmaking against male 6 respondents in sexual misconduct disciplinary proceedings is, in general, relevant in a Title 7 IX case such as this one. Accordingly, and without prejudging ABOR’s more specific 8 arguments as to why the manner in which Dr. Kaufman conducted his statistical analysis 9 was flawed, the Court agrees with Doe that the challenged opinions satisfy Rule 702’s 10 “basic standard of relevance,” which “is a liberal one.” Daubert, 509 U.S. at 587. 11 These considerations also foreclose ABOR’s first reconsideration argument, which 12 is that courts must pay careful attention before admitting statistical evidence due to its 13 “inherently slippery nature.” As noted, Schwake and Doe expressly contemplate the use 14 of statistical evidence such as Dr. Kaufman’s to prove bias in a Title IX case. So does Gay, 15 which is one of the cases on which ABOR relies heavily in its reconsideration motion. In 16 Gay, the plaintiffs brought claims alleging “discrimination on the basis of race in the hiring, 17 promotion, and transfer of black male waiters and applicants by . . . several well-known 18 San Francisco hotels and restaurants.” 694 F.3d at 534. Following a bench trial, the district 19 court ruled in the defendants’ favor. Id. On appeal, the plaintiffs limited their challenge 20 to the district court’s rejection of their claims under 42 U.S.C. § 1981. Id. at 536. As an 21 initial matter, the Ninth Circuit clarified that such claims “require proof of intentional racial 22 discrimination” and are analyzed similarly to disparate treatment claims under Title VII. 23 Id. at 537-38.4 On the merits, the court noted that that plaintiffs “were unable to produce 24 any ‘smoking gun’ evidence of overt discrimination” but emphasized that “[t]his failure 25 was not fatal to their case . . . since it is settled that a prima facie showing of disparate 26 treatment may be made without any direct proof of discriminatory motivation.” Id. at 546. 27 4 As ABOR notes, Title IX claims such as Doe’s are analyzed like Title VII disparate 28 treatment claims. (Doc. 227 at 6 nn.5-6.) Thus, the analysis of the Title VII disparate treatment claim in Gay is instructive here.

- 13 - Case 2:18-cv-01623-DWL Document 250 Filed 11/18/22 Page 14 of 46

1 The court also emphasized “that proof of the four McDonnell Douglas criteria is not the 2 only way to establish a prima facie case of disparate treatment” and that, “if reliable, 3 generalized statistical data is relevant and admissible at the prima facie case stage of a 4 disparate treatment employment discrimination lawsuit.” Id. at 550. The court added: 5 “This is true whether or not the statistical data introduced is sufficient of itself to establish 6 a prima facie case. Since the question is one of inferring discriminatory intent, the district 7 court should make a sensitive inquiry into such circumstantial and direct evidence of intent 8 as may be available. . . . [A]ll evidence, both direct and circumstantial, statistical and 9 nonstatistical, relevant to that question should be assessed on a cumulative basis.” Id. 10 (citation omitted). Later, the court elaborated: “[T]he best prima facie case utilizing 11 statistical data, one allowing the strongest inference of intentional discrimination outside 12 the McDonnell Douglas framework, is that in which the plaintiff’s statistical proof is 13 ‘bolstered’ by other circumstantial evidence of discrimination bringing the cold numbers 14 convincingly to life.” Id. at 553 (citation omitted). Applying these standards, the court 15 stated that the district court had properly considered the plaintiffs’ statistical evidence, 16 notwithstanding “the hotels’ attempts to impeach the waiters’ statistical evidence by 17 demonstrating a lack of qualified blacks in the general labor pool and a relative lack of 18 blacks in the ‘applicant flow,’” but had also properly concluded that the statistical evidence 19 was not, alone, sufficient to establish a prima facie case of discriminatory treatment 20 because “the degree of disparate impact shown by the waiters’ statistical data was 21 insufficient . . . . Although statistical data may, in a proper case, be sufficient alone to raise 22 a prima facie case, the statistics must be considerably more stark than those involved here.” 23 Id. at 552 (emphasis added). In support of this conclusion, the court noted that “[i]n order 24 to establish a prima facie case of disparate treatment based solely on statistical evidence, 25 the plaintiff must produce statistics showing a clear pattern, unexplainable on grounds other 26 than race.” Id. (citation omitted and emphasis added). Finally, although the plaintiffs 27 attempted to identify other circumstantial evidence, apart from their statistical evidence, to 28 support their claim, the court concluded that “[t]his circumstantial evidence . . . does not

- 14 - Case 2:18-cv-01623-DWL Document 250 Filed 11/18/22 Page 15 of 46

1 appreciably bolster the waiters’ weak statistical proof.” Id. at 554. 2 Gay tends to undermine, rather than support, ABOR’s position for two reasons. 3 First, contrary to ABOR’s first reconsideration argument (i.e., statistical evidence in a 4 discrimination case should be viewed with disfavor due to its slippery nature), Gay 5 recognizes that “reliable, generalized statistical data is relevant and admissible at the prima 6 facie case stage of a disparate treatment employment discrimination lawsuit,” “whether or 7 not the statistical data introduced is sufficient of itself to establish a prima facie case.” Id. 8 at 550. Indeed, Gay touts the potential value of statistical evidence in this context, noting 9 that when it is combined with “other circumstantial evidence of discrimination,” this 10 combination can “bring[] the cold numbers convincingly to life.” Id. at 553. Second, 11 contrary to ABOR’s second reconsideration argument (i.e., if the proponent of statistical 12 evidence acknowledges that the disparities may be explained by non-discriminatory 13 factors, the evidence is irrelevant and inadmissible), Gay merely holds that when a plaintiff 14 seeks to prove disparate treatment based “solely” on statistical evidence or on statistical 15 evidence “alone,” the evidence is legally insufficient if the statistical disparities may be 16 explained by reasons other than discrimination. Id. at 552. But here, Doe does not rely 17 solely on statistical evidence to prove his claim—as discussed in more detail in Part II 18 below, he seeks to rely on the combined weight of Dr. Kaufman’s statistical evidence and 19 evidence of various irregularities. At any rate, the passage from Gay that ABOR block- 20 quotes in its motion (Doc. 227 at 3-4) and re-quotes in its reply (Doc. 247 at 5) is not a 21 statement about the relevance and admissibility of statistical evidence—it is a statement 22 about how to conduct a sufficiency-of-the-evidence review in an intentional discrimination 23 case where (unlike here) the plaintiff seeks to rely “solely” on statistical evidence but 24 cannot eliminate other possible explanations for statistical disparities. The passage does 25 not suggest that such evidence would be irrelevant or inadmissible, and indeed other 26 portions of Gay suggest that the district court properly considered the plaintiffs’ statistical 27 evidence notwithstanding the fact that it was arguably impeachable and was ultimately 28 deemed insufficient to support the plaintiffs’ claims. Id. at 551-52 (noting that “[t]he

- 15 - Case 2:18-cv-01623-DWL Document 250 Filed 11/18/22 Page 16 of 46

1 district judge undertook an exhaustive analysis of the statistical evidence introduced by the 2 waiters” and “rejected the hotels’ attempts to impeach the waiters’ statistical evidence by 3 demonstrating a lack of qualified blacks in the general labor pool and a relative lack of 4 blacks in the ‘applicant flow’”). 5 Aragon, which is another case on which ABOR relies (Doc. 247 at 3), reinforces 6 these points. In Aragon, a white employee of a waste disposal company brought a Title 7 VII race discrimination claim premised on the allegation that he and other “white or white- 8 looking” employees were terminated, while similarly situated black employees were not, 9 and the reasons given for his termination were pretextual. 292 F.3d at 657-58. The 10 plaintiff’s sole evidence of pretext was that “three of the four [employees] singled out for 11 lay off that night were white.” Id. at 663. See also id. (“Aragon’s statistics . . . stand 12 precariously unsupported by other probative evidence of race discrimination.”) (cleaned 13 up). Although the Ninth Circuit recognized in a footnote that this statistical evidence was 14 relevant,5 it held that the evidence was insufficient to create a triable issue of fact as to 15 pretext for two reasons: (1) “the sample size is so small” that is fails to establish a “stark 16 pattern”; and (2) “nor does it account for possible nondiscriminatory variables, such as job 17 performance.” Id. (citations omitted). 18 Like Gay, Aragon undermines ABOR’s broad relevance objection to Dr. Kaufman’s 19 statistical evidence as well as ABOR’s first and second reconsideration arguments. 20 Although ABOR cites those cases for the proposition that “Ninth Circuit precedent” 21 requires statistical evidence to “show a clear pattern, unexplainable on grounds other than 22 gender,” and “eliminate non-discriminatory explanations” (Doc. 247 at 2), this supposed 23 admissibility requirement does not exist—ABOR creates it by lifting out-of-context 24 passages from Aragon and Gay that address how to conduct a sufficiency-of-the-evidence 25 review in an intentional discrimination case where (unlike here) the plaintiff seeks to rely 26 solely on statistical evidence. Additionally, whereas the statistical evidence in Aragon was 27 5 Aragon, 292 F.3d at 663 n.6 (“[S]tatistical evidence of a company’s general hiring 28 patterns, although relevant, carries less probative weight than it does in a disparate impact case.”) (emphasis added).

- 16 - Case 2:18-cv-01623-DWL Document 250 Filed 11/18/22 Page 17 of 46

1 based on an inadequate sample size, here Dr. Kaufman opines that the enforcement 2 disparities he describes are statistically significant.6 3 Yet another case cited by ABOR, Rose, undermines rather than supports ABOR’s 4 position on these points for the same reasons. Rose was an age discrimination case where 5 the plaintiff submitted “a statistical analysis of Wells Fargo’s employment decisions 6 [suggesting] that the single most important factor in predicting retention or termination 7 . . . was age.” 902 F.2d at 1423. The Ninth Circuit found “these statistics insufficient in 8 themselves to create a triable issue of fact of intent to discriminate” because “to show a 9 prima facie case of disparate treatment based solely on statistics [plaintiffs] must show a 10 stark pattern of discrimination unexplainable on grounds other than age.” Id. (citation 11 omitted and emphases added). Again, this is a statement about sufficiency of the evidence 12 in a discrimination case premised solely on statistical evidence, not a statement about the 13 admissibility and relevance of that evidence (let alone the admissibility and relevance of 14 that evidence where, as here, it simply forms one component of the plaintiff’s proof).7 15 6 In a footnote in its reconsideration motion, ABOR suggests that Dr. Kaufman’s 16 sample sizes are too small. (Doc. 227 at 9 n.11.) Not only is this undeveloped argument forfeited—to the extent ABOR wished to seek exclusion on this ground, it needed to do so 17 in its Daubert motion—but it also fails on the merits. Dr. Kaufman opines that at least some of the results he generated are statistically significant. Freyd, 990 F.3d 1225-26 18 (“Our prior cases stated a general principle about the reliability of small data sets, but it does not establish a firm rule about denominators. And although there must be some floor 19 for the sample size a party must evaluate in order to reach statistical significance, this is not an appropriate case in which to set such a floor; at least not on this record, where the 20 expert witnesses themselves disagree about sample size’s relevance. . . . The number of data points surely goes to the probative value of [plaintiff’s] evidence. But that is a matter 21 for the experts to debate and the jury to resolve.”) (citations omitted). 7 For similar reasons, the out-of-circuit decisions cited in ABOR’s reply in support of 22 its exclusion motion (Doc. 196 at 4 n.3) do not require exclusion here. In Haidak v. University of Massachusetts-Amherst, 933 F.3d 56 (1st Cir. 2019), the First Circuit held 23 that the Title IX plaintiff’s statistical evidence was alone insufficient to survive summary judgment due to his expert’s “fail[ure] to address an array of alternative explanations” for 24 the gender-based disparity in enforcement outcomes in assault cases, including the possibility that the “male students . . . committed more serious assaults.” Id. at 75. Haidak, 25 like the Ninth Circuit cases discussed above, is not a case about admissibility under Rule 702 and does not address whether statistical-disparity evidence may be used in conjunction 26 with other circumstantial evidence of discriminatory intent (as Doe seeks to use it here) to survive summary judgment. Meanwhile, in Doe v. University of Denver, 952 F.3d 1182 27 (10th Cir. 2020), the statistical evidence did not address disparate enforcement outcomes but whether the school’s “decisions regarding the initiation of sexual-misconduct 28 proceedings were motivated by considerations of gender.” Id. at 1195 (emphasis added). Additionally, the court was only addressing the legal sufficiency of the statistical evidence

- 17 - Case 2:18-cv-01623-DWL Document 250 Filed 11/18/22 Page 18 of 46

1 Indeed, elsewhere in the opinion, the Ninth Circuit suggested that the statistical evidence 2 was relevant. Id. (“[E]ven if the statistical evidence and the evidence of replacement six 3 months following the merger arguably support an inference of discrimination with respect 4 to Reed, we find that evidence insufficient to support a jury verdict on the ultimate question 5 of discrimination.”).8 6 Having addressed some of ABOR’s broader objections to Dr. Kaufman’s F-23 7 opinions, the Court now turns to Dr. Kaufman’s failure to consider the factual 8 circumstances underlying each F-23 incident. ABOR’s overarching contention—which it 9 identified as both a reliability and relevance problem in its motion to exclude and 10 developed in more depth in its third and fifth reconsideration arguments—is that because 11 “sexual misconduct” is defined so broadly under F-23, it is meaningless to compare the 12 outcomes in F-23 enforcement proceedings without more information about the specific 13 conduct giving rise to each proceeding. 14 Although this argument is not without force, it does not require the exclusion of Dr. 15 Kaufman’s challenged opinions. ABOR may be correct that Dr. Kaufman’s statistics 16 for purposes of surviving summary judgment, not the relevance and admissibility of the evidence under Rule 702. 17 8 ABOR’s reliance on I.V. v. Wenatchee School District No. 246, 342 F. Supp. 3d 18 1083 (E.D. Wash. 2018), is even more misplaced. There, the plaintiff (IV) was a middle- school student who had been the victim of a series of bullying incidents over a period of 19 years committed by a fellow student (YAF). Id. at 1087-91. In an ensuing Title IX action against the school district, the plaintiff’s expert sought to offer “an opinion as to the 20 motivation behind YAF’s conduct (i.e., whether it was ‘because of’ sex).” Id. at 1091. This expert “opine[d] that the bullying IV experienced was ‘unquestionably sexually 21 driven.’” Id. at 1092. The district court excluded this opinion on a variety of grounds, including that (1) it was “not based on adequate information” because the expert, among 22 other things, “did not know whether YAF pushed, kicked, or hit other kids in the class” and thus did “not consider the full constellation of facts, as is required for the issue at 23 hand”; and (2) the expert “failed to set down an principles connecting the data and her conclusion” and instead simply offered “her own ipse dixit.” Id. at 1092-93. Here, in 24 contrast, Dr. Kaufman does not purport to offer the opinion that the decisionmakers in Doe’s case were, in fact, subjectively motivated by gender bias—an opinion that would 25 impermissible for a host of reasons, as recognized by the I.V. court. Instead, Dr. Kaufman merely seeks to offer opinions about the statistical significance of certain data he compiled, 26 while leaving it to the factfinder to decide whether an inference of discriminatory intent should be drawn from those statistics. (Doc. 159-1 at 6 [“Males receive severe sanctions 27 of expulsion and suspension at a higher rate than females charged with the same violations. This different is statistically significant.”].) Nothing in I.V. suggests that the use of 28 statistics in this fashion is improper, and the cases discussed elsewhere in this order suggest this is a permissible use.

- 18 - Case 2:18-cv-01623-DWL Document 250 Filed 11/18/22 Page 19 of 46

1 would have been more useful had he considered the factual circumstances giving rise to 2 the F-23 cases he analyzed, but statistical evidence is not rendered irrelevant and 3 inadmissible simply because other, even better statistics are available. Cf. J.SH Sec. Indus. 4 D.C.S., Ltd. v. Bartech Sys. Int’l, Inc., 2010 WL 11575538, *5 (D. Nev. 2010) (“Prior 5 dealings involving refrigerators would be more relevant, but it cannot be said that JSH’s 6 prior dealings with safe orders from LVS are irrelevant.”). Thus, while again 7 acknowledging that it would have been even more relevant for Dr. Kaufman to analyze the 8 outcomes of F-23 cases involving the specific type of “sexual misconduct” that Doe was 9 accused of (i.e., sex by force and/or without consent), it remains relevant that gender-based 10 disparities exist in ASU’s enforcement of “sexual misconduct” cases more generally. In 11 Austin v. University of Oregon, 925 F.3d 1133 (9th Cir. 2019), the Ninth Circuit suggested 12 that female students “accused of sexual misconduct” would qualify as “similarly situated 13 students” who had “been accused of comparable misconduct” in a case involving a Title 14 IX claim brought by a male student. Id. at 1138. As discussed in the summary judgment 15 order, although this passage from Austin is not dispositive here (because the Ninth Circuit 16 was not addressing the question of admissibility under Rule 702), it still “supports the 17 conclusion that Dr. Kaufman’s statistics qualify, at a minimum, as relevant evidence that a 18 reasonable juror could construe as supporting Doe’s Title IX claim.” (Doc. 155 at 55-57.) 19 As also discussed in the summary judgment order, “this is not a situation where Dr. 20 Kaufman made up his own definition of ‘sexual misconduct’ in an attempt to create some 21 sort of cherry-picked statistic. . . . [S]ection F-23’s definition of ‘sexual misconduct’ is a 22 definition both created and used by ASU, which suggests that ASU has already decided it 23 unifies a population of similarly situated respondents in some material respect.” (Id. at 59.) 24 Given this backdrop, the Court is satisfied that Dr. Kaufman’s challenged opinions 25 fall within Rule 702’s broad conception of relevance. Female students found responsible 26 for committing “sexual misconduct” in violation of F-23 are similar enough to Doe, who 27 was also found responsible for committing “sexual misconduct” in violation of F-23, that 28 a comparison of the two sets of cases (Dr. Kaufman’s first challenged F-23 opinion) is

- 19 - Case 2:18-cv-01623-DWL Document 250 Filed 11/18/22 Page 20 of 46

1 relevant here. Similarly, male students found responsible for committing “sexual 2 misconduct” in violation of F-23 against a male complainant are similar enough to Doe, 3 who was found responsible for committing “sexual misconduct” in violation of F-23 4 against a female complainant, that a comparison of the two sets of cases (Dr. Kaufman’s 5 second challenged F-23 opinion) is also relevant here. 6 In the summary judgment order, the Court reached the same conclusion, holding 7 that “[t]o the extent ABOR argues [Dr. Kaufman’s first F-23] statistic is misleading 8 because it involves a comparison of episodes of ‘sexual misconduct’ that may be factually 9 dissimilar, and thus does not involve a comparison of similarly situated male and female 10 students, this is an argument as to the weight of the evidence that ABOR can pursue at trial 11 through cross-examination.” (Doc. 155 at 57.) In its third reconsideration argument, 12 ABOR contends that “[t]his approach violates the Court’s role as gatekeeper” and cites 13 Freyd as a case supporting its position. (Doc. 227 at 4-6.) Accordingly, before addressing 14 the merits of ABOR’s argument, it is helpful to summarize Freyd. 15 There, a female professor at the University of Oregon brought claims under the 16 Equal Pay Act, Title VII, and Title IX after discovering that “the University pays [her] 17 several thousand dollars less per year than it does four of [her] male colleagues, despite 18 their being of equal rank and seniority.” 990 F.3d at 1214. The district court granted 19 summary judgment in the university’s favor on all of Freyd’s claims but the Ninth Circuit 20 reversed in part. As for the Equal Pay Act claim, the analysis turned on whether Freyd and 21 the four male professors held “substantially equal” jobs, such that the male professors could 22 be considered relevant comparators. The district court held “that Freyd and her 23 comparators’ jobs are rendered unequal by the differences in the research that they do, 24 centers that they run, and funding that they obtain” but the Ninth Circuit reversed, holding 25 that “the evidence here is not so one-sided as to mandate this conclusion as a matter of law” 26 and that “[w]e do not have the tools to resolve the despite without intruding on the civil 27 jury’s function. . . . [A] reasonable jury could find that Freyd and her comparators did 28 substantially equal work. Accordingly, the district court’s grant of summary judgment on

- 20 - Case 2:18-cv-01623-DWL Document 250 Filed 11/18/22 Page 21 of 46

1 this claim was in error.” Id. at 1220-23. Next, as for the Title VII disparate impact claim, 2 the Ninth Circuit again reversed, holding that the district court erred by finding “as a matter 3 of law that Freyd’s statistical evidence was insufficient to sustain a claim of disparate 4 impact” and that “[w]e do not think that we can resolve this dispute among the experts.” 5 Id. at 1225. In contrast, the Ninth Circuit affirmed the district court’s grant of summary 6 judgment on Freyd’s Title VII disparate treatment claim. Id. at 1228-29. As for that claim, 7 Freyd’s theory was that “similarly situated individuals outside her protected class were 8 treated more favorably than her.” Id. The Ninth Circuit held that Freyd’s evidence was 9 insufficient with respect to this theory because “Freyd’s comparators engaged in retention 10 negotiations with the University and were granted substantial salary increases as a result,” 11 whereas “Freyd has never engaged in retention negotiations,” and thus Freyd had “not 12 presented evidence sufficient to establish a prima facie case under the McDonnell Douglas 13 framework because she has not shown that similarly situated individuals outside of her 14 protected class were treated more favorably than her.” Id. Finally, the court affirmed the 15 district court’s grant of summary judgment on Freyd’s Title IX claim for the same reason, 16 holding that “[a]s with her Title VII disparate treatment claim, because Freyd has presented 17 no evidence of intentional discrimination, there is no genuine issue of material fact here.” 18 Id. at 1229. 19 In the Court’s estimation, Freyd does not undermine (and, arguably, supports) the 20 conclusion in the summary judgment order that the parties’ dispute over whether the 21 comparators selected by Dr. Kaufman were sufficiently similar to Doe is a question of fact 22 that should be decided by the jury at trial. Indeed, in Freyd, the Ninth Circuit reversed the 23 district court’s conclusion that the male professors were too dissimilar to qualify as relevant 24 comparators for purposes of the plaintiff’s Equal Pay Act claim, emphasizing that because 25 “the evidence here is not so one-sided as to mandate this conclusion as a matter of law,” 26 the district court had “intrud[ed] on the civil jury’s function” by concluding otherwise. Id. 27 at 1220-23. This approach is consistent with Court’s approach here. See also Alaska Rent- 28 A-Car, 738 F.3d at 968-70 (affirming district court’s admission of expert testimony, where

- 21 - Case 2:18-cv-01623-DWL Document 250 Filed 11/18/22 Page 22 of 46

1 the expert “compared Avis’s . . . experience with [Alamo’s] experience after Cerberus 2 bought Alamo out of bankruptcy” and the opposing party objected on the ground “that 3 differences between Alamo and Budget . . . made Alamo an invalid comparison,” because 4 the challenge to “using Alamo as the comparator” was a challenge that “go[es] to the 5 weight of the testimony and its credibility, not its admissibility”). Finally, although it is 6 true that Freyd affirmed the district court’s grant of summary judgment on the plaintiff’s 7 intentional discrimination claims under Title VII and Title IX, on the ground that the 8 plaintiff’s comparators were not similarly situated for purposes of those claims, the Court 9 agrees with Doe (Doc. 241 at 6) that this portion of the decision turned on fact-specific 10 considerations that shed little light on the dispute here. Even accepting that professors who 11 engaged in retention negotiations are dissimilar to professors who never engaged in 12 retention negotiations, it doesn’t follow that ASU students accused of “sexual misconduct” 13 as that term is defined in F-23 are too dissimilar from each other to provide a relevant basis 14 for comparison.9 The Court also notes that the passages from Freyd on which ABOR relies, 15 like the passages from Gay, Aragon, and Rose discussed above, do not address the 16 relevance and admissibility of statistical evidence but the sufficiency of that evidence for 17 purposes of surviving summary judgment. This further undermines ABOR’s reliance on 18 Freyd as a basis for seeking the exclusion of Dr. Kaufman’s statistics on relevance grounds 19 under Rule 702. 20 Nor is there any merit to ABOR’s fifth reconsideration argument, which attacks the 21 portion of the summary judgment order noting that Dr. Kaufman did not make up his own 22 23 9 For similar reasons, ABOR’s reliance on Moran v. Selig, 447 F.3d 748 (9th Cir. 2006) is misplaced. There, the Ninth Circuit concluded that the members of a class of 24 “virtually all Caucasian former MLB players who played in the Major Leagues for less than four years between 1947 and 1979 and were accordingly denied MLB pension and 25 medical benefits” were not “similarly situated to . . . the former Negro League players who also played in the Major Leagues between 1947-79 but receive[d] medical and 26 supplemental income benefits under the Negro League Plans” for the “basic” reason that “in order to qualify for the benefits under the Negro League Plans, the players to whom 27 [plaintiffs] seek to equate themselves must have played in the Negro Leagues—a qualification that [plaintiffs] indisputably lack.” Id. at 751, 755-56. This fact-specific 28 ruling sheds little light on the comparability of ASU students accused of sexual misconduct in violation of F-23.

- 22 - Case 2:18-cv-01623-DWL Document 250 Filed 11/18/22 Page 23 of 46

1 definition of “sexual misconduct” but instead utilized the definition created and utilized by 2 ASU. To the extent ABOR accuses the Court of “creat[ing] this argument on behalf of” 3 Doe, who “never made this argument” himself (Doc. 227 at 7), this accusation is belied by 4 the record. In response to the motion to exclude, Doe argued as follows: “ABOR has not 5 presented any evidence that the respondents reflected in the F-23 spreadsheets are not 6 similarly situated. . . . The data produced by ABOR categorizes all F-23 charges into a 7 single category of sexual misconduct. ABOR does not and cannot provide any support for 8 its allegation that the various conduct constituting ‘sexual misconduct’ under its own Code 9 is somehow ‘less’ serious than other harms.” (Doc. 188 at 6.) Elsewhere, Doe argued: 10 “ABOR argues that for purposes of a gender discrimination claim, ‘similarly situated’ 11 means that the misconduct in which the comparators are alleged to have engaged must be 12 of ‘comparable seriousness.’ This threshold is met—all violated F-23, the same sexual 13 misconduct violation under the Code.” (Id. at 8-9.) The analysis in the summary judgment 14 order reflected the Court’s acceptance and agreement with these arguments, albeit while 15 using slightly different phrasing. It is disappointing that ABOR would accuse the Court of 16 stepping outside the judicial role and acting as Doe’s advocate under these circumstances. 17 Next, as for ABOR’s contention that “[t]he Court also manifestly errs because the 18 [summary judgment order] does not cite evidence supporting its supposition that ASU 19 ‘decided’ all F-23 violations were comparable” (Doc. 227 at 7), the Court disagrees for the 20 reasons stated in Doe’s response: “[T]hat ASU grouped all F-23 conduct together by 21 definition in its own student handbook is a fact established by evidence presented by 22 ABOR [including the handbook].” (Doc. 241 at 7.)10 As Doe further notes, “ABOR does 23 not present any evidence supporting its conclusory argument that [the subsections within 24 F-23] are grouped by severity. . . . ABOR still does not endeavor to rank the seriousness 25 of the acts of sexual misconduct along these lines. Nor can it. Each subsection includes a 26 10 See also Freyd, 990 F.3d at 1222 (“The dissent’s conclusion that two faculty 27 members in the same department cannot be compared is inconsistent with the fact that the University’s own administrators regularly make these comparisons for purposes of setting 28 salaries. The question is not whether faculty members can be compared, but how they compare, and the latter comparison is one fraught with judgment, not law.”).

- 23 - Case 2:18-cv-01623-DWL Document 250 Filed 11/18/22 Page 24 of 46

1 range of behavior. . . . For instance, persistent sexual harassment that creates a hostile 2 environment (subsection B) could certainly rise to or past the level of seriousness of non- 3 consensual sexual touch in subsection A.” (Id. at 8-9.) Finally, as for ABOR’s contention 4 that the Court’s approach may “incentivize universities to endure the unnecessary expense 5 of redrafting their codes of conduct” (Doc. 227 at 8 n.9), such public policy arguments are 6 unhelpful when evaluating questions of relevance and reliability under Rule 702. 7 Having determined that Dr. Kaufman’s F-23 opinions are relevant despite his failure 8 to consider the specific factual circumstances giving rise to the underlying proceedings, the 9 Court next considers whether Dr. Kaufman followed a reliable methodology when forming 10 those opinions. ABOR’s overarching argument on that issue, which it frames as a failure 11 to consider “sufficient facts and data” in its motion to exclude and further elaborates upon 12 in its fourth reconsideration argument, is Dr. Kaufman’s approach was unreliable because 13 he had access to data setting forth the factual circumstances of the underlying incidents 14 (i.e., the Title IX Case Tracking Report) and simply failed to consider it. ABOR cites 15 Pottenger as its best case supporting exclusion in this circumstance. 16 This argument presents a fairly close call. Although it is not clear that Dr. 17 Kaufman’s failure to consider the Title IX Case Tracking Report poses a reliability 18 problem, as opposed to some other type of flaw,11 ABOR is correct that, under Pottenger, 19 an expert’s failure to consider a particularly critical set of data can undermine the 20 evidentiary value of the expert’s statistical analysis. In Pottenger, a 60-year-old executive 21 at a paper company asserted a claim of age discrimination after he was terminated and 22 replaced by a 43-year-old. 329 F.3d at 744-45. In support of this claim, Pottenger offered 23 11 The concept of reliability seems like a poor fit here because there was no need for 24 Dr. Kaufman to consider the Title IX Case Tracking Report when formulating his challenged opinions, which simply track the outcomes in cases involving F-23 violations. 25 Because those opinions don’t purport to control for the specific circumstances of the underlying incidents (beyond the fact that each incident involved conduct that fell within 26 F-23’s broad definition of “sexual misconduct”), it follows that the methodology Dr. Kaufman used when forming those opinions (i.e., considering the SCC Charge Data) was 27 reliable. At bottom, ABOR’s criticism is not that Dr. Kaufman used a flawed methodology when addressing the particular questions he chose to address. Instead, the criticism is that 28 he addressed the wrong questions. That seems like an argument about relevance, not reliability.

- 24 - Case 2:18-cv-01623-DWL Document 250 Filed 11/18/22 Page 25 of 46

1 statistical evidence suggesting that “the company’s June 2000 reduction in force [RIF] 2 disproportionately affected older employees.” Id. at 747. Nevertheless, the district court 3 granted summary judgment in the employer’s favor and the Ninth Circuit affirmed, holding 4 that the statistical evidence was insufficient to create a genuine issue of fact as to whether 5 the employer’s proffered reason for terminating Pottenger (i.e., “he was not prepared to 6 make the tough decisions necessary to turn around the Idaho Pulp and Paper Division”) 7 was pretextual. Id. at 745, 747-48. Although the court recognized that a plaintiff asserting 8 a discrimination claim may, in general, “use statistics to show an intent to discriminate,” it 9 held that Pottenger’s statistical analysis, which “t[ook] into account only two variables— 10 the employee’s age at the time of the RIF and whether the employee was terminated,” 11 should be “treated skeptically” because it “fail[ed] to account for other relevant variables.” 12 Id. at 747-48. In particular, the court noted that “the variable most likely to have offered a 13 legally appropriate explanation of why certain employees were selected for lay-off [was] 14 job performance.” Id. at 748. The court further noted that Pottenger’s expert was offered 15 access to job-performance data before conducting his analysis but “specifically acquiesced 16 in the suggestion that obtaining data about individual employees’ performance reviews was 17 unnecessary.” Id. As a result, the court concluded: “If Pottenger had had access to only 18 two variables, we would be presented with a different case. But here, where Pottenger had 19 or had access to additional relevant data and chose not to use it, we conclude that 20 Pottenger’s statistical analysis is insufficient to raise a triable issue of fact regarding 21 pretext.” Id. 22 It is understandable why ABOR would view Pottenger as a case supporting its 23 position. From ABOR’s perspective, Dr. Kaufman is just like Pottenger’s expert in that he 24 had access to the variable that was most likely to explain any differences in F-23 25 enforcement (i.e., the specific factual circumstances underlying each F-23 violation, which 26 were set forth in the Title IX Case Tracking Report) yet chose not to consider that data and 27 instead relied on other, less probative data (i.e., the details set forth in the SCC Tracking 28 Data, which were limited to the gender of the respondent/complainant, the outcome of the

- 25 - Case 2:18-cv-01623-DWL Document 250 Filed 11/18/22 Page 26 of 46

1 proceeding, and the bare fact that the respondent was found to have committed an F-23 2 violation). Nevertheless, the Court discerns two reasons why Pottenger does not do all of 3 the work that ABOR wants it to do here. 4 First, Pottenger is not a case about relevance and admissibility under Rule 702— 5 the Ninth Circuit’s holding was that Pottenger’s statistical evidence was not alone 6 sufficient to create a triable issue of fact on the issue of pretext due to his expert’s failure 7 to consider the most important variable. But it doesn’t necessarily follow, from the fact 8 that a particular piece of evidence is alone insufficient to avoid summary judgment, that 9 the piece of evidence is also irrelevant and inadmissible. Thus, to the extent ABOR seeks 10 to rely on Pottenger as a basis for excluding Dr. Kaufman’s statistical analysis (as 11 contrasted with contesting the weight of that statistical analysis for purposes of evaluating 12 ABOR’s entitlement to summary judgment, which is an issue addressed in Part II below), 13 that reliance is misplaced.12 14 Second, there are important differences between Pottenger and this case when it 15 comes to the expert’s explanation for not considering the additional data. As noted, in 16 Pottenger, the expert “specifically acquiesced in the suggestion” that job performance data 17 was “unnecessary.” 329 F.3d at 748. Here, in contrast, Dr. Kaufman stated in his report 18 that he attempted to obtain “a sample of case files to develop additional controls, but 19 [ABOR] declined to provide such data.” (Doc. 159-1 at 14.)13 Additionally, in a 20 declaration prepared several months later (in response to ABOR’s contention that he had 21 12 During oral argument on the Daubert motions, which occurred shortly after ABOR filed its reconsideration motion and cited Pottenger for the first time, the Court suggested 22 that Pottenger addresses the admissibility of statistical evidence. (Doc. 243 at 94.) Having now had a chance to more carefully review Pottenger, the Court concludes (as stated 23 above) that Pottenger is a case about evidentiary weight, not admissibility. 13 24 Here, Dr. Kaufman appears to be referring to a discovery dispute that arose in mid- 2021 over whether ABOR’s Rule 30(b)(6) designee needed “to become knowledgeable 25 about the details of hundreds, if not thousands, of individual discipline proceedings involving ASU students over a period of many years.” (Doc. 143 at 7.) The Court ruled 26 in ABOR’s favor as to this dispute because “although the general subject area Doe wishes to explore . . . which Doe describes in his response as ‘how ASU has treated others in the 27 student disciplinary system, specifically those involving claims of sexual assault’ . . . is relevant and ought to be fair game for examination, the verbiage he chose to use when 28 formulating the deposition topics is vague, overbroad, and would place overly burdensome and disproportionate preparation demands on ABOR.” (Id. at 9.)

- 26 - Case 2:18-cv-01623-DWL Document 250 Filed 11/18/22 Page 27 of 46

1 overlooked key details), Dr. Kaufman stated that he had reviewed the Title IX Case 2 Tracking Report before drafting his report but declined to rely on it for two reasons: (1) 3 although the Title IX Case Tracking Report “include[s] high level descriptions of the 4 charges,” it “does not categorize the charges by relative seriousness” and contains a “level 5 of detail” that is “unreliable and incomplete”; and (2) the Title IX Case Tracking Report 6 does not cover the full range of years covered by the SCC Tracking Data14 and the cases in 7 the Title IX Case Tracking Report are not “randomly sampled from the full incidents 8 reflected on the [SCC Tracking Data].” (Doc. 188-2 ¶¶ 4-6.)15 Given these details, Doe 9 argues that “[t]he data in the [Title IX Case Tracking Report] is a far cry from the reliable, 10 complete, ranked, and usable data of Pottenger” and that “[i]n rejecting the use of the [Title 11 IX Case Tracking Report] in favor of the complete data set sworn by ABOR to be accurate 12 in its interrogatory responses, Kaufman preserved the reliability [of] his analysis.” (Doc. 13 241 at 7.) 14 The Court agrees with Doe that these details distinguish this case from Pottenger. 15 Dr. Kaufman did not ignore the Title IX Case Tracking Report but instead made a 16 determination, after reviewing it, that it had various flaws that rendered it unhelpful in 17 developing additional controls for his analysis. Although ABOR vigorously disputes Dr. 18 Kaufman’s determination on this point (and, the Court suspects, will be able to use the Title 19 14 Dr. Kaufman stated in his declaration that the Title IX Case Tracking Report “only present[s] incidents of sexual misconduct reported to ASU from 2016-17.” (Doc. 188-2 at 20 6.) Doe acknowledges that Dr. Kaufman was mistaken on this point—the Title IX Case Tracking Report actually covers the years 2013 through 2017. (Doc. 241 at 6 n.5.) 21 Nevertheless, it is undisputed that the Title IX Case Tracking Report does not cover the full range of F-23 incidents summarized in the SCC Tracking Data, which cover the years 22 2012 through 2017. (Doc. 247 at 3 [“[T]he Report did not exist prior to the 2013-14 academic year.”].) 23 15 At times, Doe has suggested that the Title IX Case Tracking Report was produced 24 in a confusing and/or indecipherable manner during discovery, such that ABOR should bear any blame for Dr. Kaufman’s failure to consider it. (Doc. 243 at 87-89; Doc. 246 at 25 3 [“[T]he documents were not produced in a searchable format, yet ABOR’s counsel claimed the opposite during Daubert oral arguments. Presenting the documents in 26 searchable format would constitute a misrepresentation that would make the data within the TIX Spreadsheets appear more accessible than it truly is.”].) In response, ABOR has 27 provided a detailed timeline of its production efforts (Doc. 247 at 3-5) as well as a flash drive containing the Title IX Case Tracking Report itself (Doc. 248), which the Court has 28 now reviewed. Based on those submissions, the Court rejects any claim that ABOR is responsible for Dr. Kaufman’s decision not to consider the Title IX Case Tracking Report.

- 27 - Case 2:18-cv-01623-DWL Document 250 Filed 11/18/22 Page 28 of 46

1 IX Case Tracking Report to great effect when cross-examining Dr. Kaufman), the 2 appropriate way to resolve this dispute is to allow the parties to present their respective 3 positions to the jury at trial. Put in the parlance of Rule 702, Dr. Kaufman’s basis for 4 declining to rely on the Title IX Case Tracking Report may be “impeachable,” but it does 5 not veer into the realm of “unreliable nonsense.” Alaska Rent-A-Car, 738 F.3d at 969. 6 And as noted elsewhere in this order, “[s]haky but admissible evidence is to be attacked by 7 cross examination, contrary evidence, and attention to the burden of proof, not exclusion.” 8 Primiano, 598 F.3d at 564. 9 This leaves ABOR’s sixth and seventh reconsideration arguments, which the Court 10 views as a matched pair. The former is that it would amount to impermissible burden- 11 flipping to fault ABOR for failing to perform its own analysis of the specific factual 12 circumstances of the underlying F-23 incidents because Doe bears the burden of 13 establishing the admissibility of his expert’s opinions. The latter is that ABOR has, at any 14 rate, now conducted the very analysis that it shouldn’t have been required to perform, 15 which shows that the incidents in which female students were found responsible for F-23 16 violations involved much different (and less serious) forms of “sexual misconduct” than 17 the incident at issue in this case. 18 These arguments fail to establish that Dr. Kaufman’s F-23 opinions should be 19 excluded. As for the former, ABOR is of course correct that it does not bear the burden of 20 proof in this context, but this observation does nothing to detract from the Court’s 21 conclusion, as discussed in the paragraphs above, the Doe has met his burden of 22 establishing the relevance and reliability of the challenged opinions. As for the latter, the 23 initial problem is that ABOR did not conduct or disclose its new evidence during the 24 discovery process or in support of its motion to exclude—instead, it presented this evidence 25 for the first time as an attachment to its motion for reconsideration. This approach is 26 impermissible for a host of reasons. Kona Enterprises, Inc. v. Estate of Bishop, 229 F.3d 27 877, 891 (9th Cir. 2000) (“[A] motion for reconsideration . . . may not be used to raise 28 arguments or present evidence for the first time when they could reasonably have been

- 28 - Case 2:18-cv-01623-DWL Document 250 Filed 11/18/22 Page 29 of 46

1 raised earlier in the litigation.”) (citation and internal quotation marks omitted). More 2 important, even if this new evidence had been timely disclosed to Doe and timely provided 3 to the Court, it would not necessitate the exclusion of Dr. Kaufman’s opinions. Instead, it 4 simply constitutes evidence that ABOR may present at trial in an attempt to impeach Dr. 5 Kaufman and/or persuade the jury that Dr. Kaufman’s opinions do not support the 6 inferences that Doe hopes the jury will draw from them. 7 2. F-23 Opinions—Failure To Control For Other Charges 8 Although most of ABOR’s challenges to Dr. Kaufman’s F-23 opinions turn on his 9 failure to consider the factual details of the underlying incidents, ABOR also raises a 10 distinct challenge in its motion to exclude—that “Dr. Kaufman’s methodology is unreliable 11 because he does not account for the fact that many of the respondents charged with [F-23] 12 violations were also charged with other violations of the Code.” (Doc. 159 at 9.) ABOR 13 acknowledges that Dr. Kaufman purported to control for this variable by relying on the 14 Mantel Haenszel peer group test but argues that, because Dr. Kaufman admits this test does 15 not eliminate the possibility that “non-gender factors [may] effect the severity of the 16 sanctions,” it follows that his methodology is unreliable. (Id. at 9-10.) In response, Doe 17 argues that “[t]he Mantel Haenszel test[] is the most widely used method of testing the 18 statistical significance of disparities,” cites other cases approving the use of this test, and 19 argues that ABOR’s reliability challenge is therefore unavailing. (Doc. 188 at 9-10.) 20 Exclusion is not warranted on this basis. Dr. Kaufman sought to control for the 21 existence of other charged violations by applying the Mantel Haenszel test. To the extent 22 ABOR’s position is that the application of this test is insufficient because it does not 23 eliminate the possibility of non-discriminatory explanations for the statistical patterns, this 24 argument fails for the reasons set forth above. Although Gay, Aragon, and Rose recognize 25 that statistical evidence that does not eliminate the possibility of non-discriminatory 26 explanations for the disparity may be legally insufficient to create a triable issue of fact in 27 a discrimination case where the plaintiff seeks to rely solely on that statistical evidence, 28 they do not hold that such statistical evidence is inadmissible and irrelevant. Additionally,

- 29 - Case 2:18-cv-01623-DWL Document 250 Filed 11/18/22 Page 30 of 46

1 and as discussed in more detail in Part II below, Doe does not rely solely on statistical 2 evidence in this case. 3 3. Final F-23 Opinion—Failure To Consider Available Data 4 ABOR next raises a challenge that is specific to Dr. Kaufman’s final F-23 opinion 5 (which, as noted above, compares the severity of sanctions imposed in F-23 cases involving 6 male complainants with the sanctions imposed in F-23 cases involving female 7 complainants). ABOR contends this statistic is unreliable because, although Dr. Kaufman 8 stated that most of the data available to him did not identify a complainant’s gender, 9 information was available to him that provided this information. (Doc. 159 at 10.) 10 Specifically, ABOR points to Exhibit 5, row 250 of the Title IX Case Tracking Report. 11 (Id.) 12 In the tentative order issue before oral argument, the Court stated that this argument 13 was based on a false premise because row 250 of the cited document (Doc. 159-4 at 20- 14 22) does not identify the complainant’s gender. (Doc. 210 at 23.) In response, ABOR 15 clarified during oral argument that its brief had a typo and that the reference should have 16 been to row 251 of the cited document. (Doc. 243 at 82.) ABOR explained that this row 17 reveals that the underlying incident was “an F-23 incident involving a male complainant 18 and the sanction was expulsion” and thus Dr. Kaufman’s opinion that “incidents involving 19 male complainants never result in severe sanctions” is erroneous. (Id.) 20 The Court appreciates this clarification but still concludes that ABOR’s criticisms 21 are fodder for impeachment, not exclusion. Dr. Kaufman’s final challenged opinion is that 22 “F-23 charges involving male complainants result in expulsion and suspension at a lower 23 rate than charges involving female complainants.” (Doc. 159-1 at 14.) The support for 24 this opinion is set forth in Table 5, which states that of the 11 F-23 cases involving a male 25 complainant, zero cases resulted in a severe sanction (0%), whereas of the 98 F-23 cases 26 involving a female complainant, 28 cases resulted in a severe sanction (29%). (Id.) If, per 27 row 251, there was actually one F-23 case with a male complainant that resulted in a severe 28 sanction (rather than zero), this would not undermine Dr. Kaufman’s overall opinion that

- 30 - Case 2:18-cv-01623-DWL Document 250 Filed 11/18/22 Page 31 of 46

1 severe sanctions are imposed less frequently in F-23 cases involving male complainants 2 than in F-23 cases involving female complainants.16 3 More broadly, and as discussed elsewhere in this order, to the extent ABOR believes 4 the information set forth in the Title IX Case Tracking Report is useful in impeaching Dr. 5 Kaufman or otherwise undermining the inferences that Doe intends to ask the jury to draw 6 from Dr. Kaufman’s statistics and opinions, ABOR is free to make its case at trial. The 7 potential for such impeachment does not undermine the Court’s determination that Dr. 8 Kaufman’s challenged opinions are sufficiently relevant and reliable to qualify for 9 admission under Rule 702. 10 4. Alcohol-Related Opinions 11 Although most of ABOR’s challenges concern Dr. Kaufman’s opinions related to 12 disparities in F-23 proceedings, ABOR’s final challenge concerns Dr. Kaufman’s opinion 13 regarding disparities in F-15 proceedings.17 As noted, Dr. Kaufman’s opinion on that issue 14 is that “[m]ale respondents are found in violation of F-15 . . . charges more often than 15 female respondents.” (Doc. 159-1 at 14.) In its motion to exclude, ABOR argues this 16 opinion is “unreliable and irrelevant” because (1) even though Dr. Kaufman attempted to 17 “control for the ‘type’ of violation by only including matters in which [F-15] was the sole 18 alleged violation,” this attempt was unsuccessful because Dr. Kaufman failed to consider 19 the underlying factual circumstances of each incident and thus failed to “account for the 20 wide variety of conduct that could constitute an [F-15] violation”; and (2) alternatively, by 21 attempting to limit his opinion to disparities in proceedings involving only F-15 violations, 22 “Dr. Kaufman ensured his analysis has no relevance” because “Doe was not solely charged 23 with an [F-15] violation” and thus “[t]here is no issue in this lawsuit about whether the 24 16 ABOR may be correct that, in light of this clarification, the title appearing above Table 5, which is “Male Complainants Do Not Result In Severe Sanctions” (Doc. 159-1 at 25 14), is inaccurate. But as discussed above, the opinion that Dr. Kaufman discloses in his report is a relative comparison of the rates of severe sanctions imposed in F-23 cases 26 involving male and female complainants. 27 17 Section F-15 provides that the definition of “Prohibited Conduct” includes: “Violation of the Board or university rules or applicable laws governing alcohol, including 28 consumption, distribution, unauthorized sale, or possession of alcoholic beverages.” (Doc. 155-2 at 55-56.)

- 31 - Case 2:18-cv-01623-DWL Document 250 Filed 11/18/22 Page 32 of 46

1 sanctions imposed on students facing solely [F-15] violations are harsher for females than 2 they are for males.” (Doc. 159 at 11-12.) In response, Doe asserts that Dr. Kaufman’s 3 opinion about F-15 violations is reliable, for similar reasons as his opinions regarding F- 4 23 violations, and relevant, because a juror can infer the presence of gender bias from the 5 fact that men are sanctioned for alcohol violations at a higher rate than women. (Doc. 188 6 at 12-13.) Doe also contends that “the most granular data available on F-15 violations is 7 that presented in the F-15 Spreadsheet created and produced by ABOR.” (Id.) 8 Again, ABOR demands too much at the motion-to-exclude stage. If true, the 9 conclusion that ASU sanctions men accused of alcohol violations more often than women 10 accused of alcohol violations is relevant, both because (1) Doe was accused of an alcohol 11 violation and (2) any statistical evidence showing that ASU is biased against men in 12 disciplinary proceedings is at least some evidence supporting Doe’s gender bias claim. The 13 fact that Doe was accused of multiple violations, and that ASU’s record of enforcing 14 alcohol violations may not be as probative of bias as ASU’s record of enforcing sexual 15 misconduct, is fertile territory for cross-examination, but it does not provide a basis for 16 exclusion. Finally, as for Dr. Kaufman’s failure to account for the specific factual 17 circumstances of each underlying alcohol violation, not only does this criticism merely go 18 to whether Dr. Kaufman’s opinion is impeachable (as opposed to inadmissible) for the 19 reasons stated above with respect to Dr. Kaufman’s F-23 opinions, but this criticism does 20 not implicate Pottenger in the same way because there is no evidence that these details are 21 tracked by ASU or were provided to Dr. Kaufman, only for him to ignore them. 22 D. Economic Loss Opinions 23 1. The Parties’ Arguments 24 ABOR argues that Dr. Kaufman’s assumptions about Doe’s career progression are 25 unfounded and unreliable because there is no evidence in the record to support those 26 assumptions. (Doc. 159 at 12-14.) Relatedly, ABOR points to Dr. Kaufman’s assertions 27 that Doe “was a Division One wrestler on his way to earning the title of All-American at 28 nationals” and that Doe’s career plan was “consistent with the career paths of other college

- 32 - Case 2:18-cv-01623-DWL Document 250 Filed 11/18/22 Page 33 of 46

1 wrestling coaches” and argues that Dr. Kaufman is not qualified to opine on those issues. 2 (Id. at 13 n. 5.) ABOR also asserts that Dr. Kaufman’s assumptions about projected salary 3 data are inaccurate and inapplicable to the facts in this case. (Id. at 15-17.) 4 Doe responds that “there is no requirement that Doe unequivocally prove that he 5 would have achieved any of these positions before an expert can opine on his lost earnings 6 . . . . The standard for assessing earning capacity is whether the plaintiff had a ‘reasonable 7 probability’ of achieving that career path. . . . Dr. Kaufman is permitted to make 8 assumptions about Doe’s career path so long as there is a reasonable probability that those 9 paths would have . . . or are likely to occur.” (Doc. 188 at 14-15.) Doe also argues that 10 Dr. Kaufman’s sources for his salary projection are reliable and, to the extent there are 11 factual disputes to be had, those issues are either scrivener’s errors or may be addressed 12 during cross-examination. (Id. at 15-17.) 13 ABOR replies that Doe has mischaracterized its argument: ABOR argues that “Dr. 14 Kaufman has no expertise or foundational facts to support his assumptions as to Doe’s 15 expected career path.” (Doc. 196 at 9.) ABOR also objects to Dr. Kaufman’s use of 16 median, rather than entry-level, salary figures and contends that the inflated salary 17 information caused Dr. Kaufman’s opinion to be irrelevant. (Id. at 10-11.) 18 2. Analysis 19 ABOR’s challenges to Dr. Kaufman’s economic loss opinions lack merit. ABOR’s 20 first objection is that Dr. Kaufman’s assumptions about the trajectory of Doe’s career are 21 unfounded and inaccurate. As for foundation, there is evidence in the record to support the 22 assumption that Doe had a reasonable probability of achieving a wide range of professional 23 outcomes within the wrestling profession. Although the Court has excluded Owen from 24 testifying that Doe would have been one of the nation’s best wrestlers (Doc. 233 at 13-16), 25 the fact that Doe won several wrestling competitions will still be before the jury. 26 Additionally, both Doe’s and ABOR’s experts agree that Doe could have become a 27 Division I coach and simply disagree on where, exactly, Doe’s ceiling would be. (Doc. 28 163-1 at 4, Doc. 165-1 at 10.)

- 33 - Case 2:18-cv-01623-DWL Document 250 Filed 11/18/22 Page 34 of 46

1 As for accuracy, it is irrelevant—for threshold admissibility purposes—that Dr. 2 Kaufman’s assumptions may be proved inaccurate at trial. All that matters is that they 3 have some basis in the record. An expert may, in appropriate circumstances, rely on 4 assumptions when formulating opinions. Fed. R. Evid. 702, advisory committee notes to 5 2000 amendments (“The language ‘facts or data’ is broad enough to allow an expert to rely 6 on hypothetical facts that are supported by the evidence.”). Disagreement with an expert’s 7 assumptions does not, in general, provide a basis for excluding the expert’s testimony. See, 8 e.g., Marsteller v. MD Helicopter Inc., 2018 WL 3023284, *2 (D. Ariz. 2018) (“The 9 challenges to Equals’ opinions and the weaknesses in his assumptions are issues to be 10 explored on cross-examination.”). To be sure, “nothing in either Daubert or the Federal 11 Rules of Evidence requires a district court to admit opinion evidence that is connected to 12 existing data only by the ipse dixit of the expert. A court may conclude that there is simply 13 too great an analytical gap between the data and the opinion proffered.” Joiner, 522 U.S. 14 at 146. Nevertheless, “[s]haky but admissible evidence is to be attacked by cross 15 examination, contrary evidence, and attention to the burden of proof, not exclusion.” 16 Primiano, 598 F.3d at 564. “Basically, the judge is supposed to screen the jury from 17 unreliable nonsense opinions, but not exclude opinions merely because they are 18 impeachable.” Alaska Rent-A-Car, Inc., 738 F.3d at 969. See also Fed. R. Evid. 702 19 advisory committee’s note to 2000 amendment (“[P]roponents do not have to demonstrate 20 to the judge by a preponderance of the evidence that the assessments of their experts are 21 correct, they only have to demonstrate by a preponderance of evidence that their opinions 22 are reliable . . . . The evidentiary requirement of reliability is lower than the merits standard 23 of correctness.” (alteration in original) (internal quotation marks omitted)). 24 Here, Dr. Kaufman takes core assumptions that have a reasonable basis in the record 25 and stretches them to opposing extremes to establish a range of possible outcomes. If 26 ABOR can discredit these assumptions at trial, that will also invalidate the opinion. See 27 generally United States v. Crabbe, 556 F. Supp. 2d 1217, 1224 (D. Colo. 2008) (“An expert 28 witness may often ‘assume’ a fact for purposes of applying the methodology. . . . If the

- 34 - Case 2:18-cv-01623-DWL Document 250 Filed 11/18/22 Page 35 of 46

1 assumption is in error, the opinion may be entirely invalidated.”). But because the 2 assumptions about Doe’s future employability do not veer into “unreliable nonsense,” and 3 ABOR does not dispute Dr. Kaufman’s algebraic methodology of converting a range of 4 assumptions into a range of potential economic loss, exclusion is not warranted.18 5 As for Dr. Kaufman’s salary assumptions, largely the same analysis applies. ABOR 6 objects to Dr. Kaufman’s use of median wages that do not account for school size, stature, 7 or a coach’s experience, arguing that “Dr. Kaufman shows no connection between the 8 salary he selected and the facts in this case.” (Doc. 159 at 15-16.) Not so. Dr. Kaufman’s 9 choice of salary data and analysis thereof may not meet ABOR’s standards for 10 persuasiveness, but it has a foundation in the record. ABOR also points to actual errors in 11 Dr. Kaufman’s data, but Doe replies that these are scrivener’s errors which have since been 12 corrected and do not meaningfully affect Dr. Kaufman’s overall analysis. Scrivener’s 13 errors that have already been corrected are not grounds for exclusion. 14 … 15 … 16 … 17 18 During oral argument, ABOR placed particular emphasis on Dr. Kaufman’s 18 assumption that Doe might pursue a coaching position at a Division II or Division III college. (Doc. 243 at 83-86.) However, ABOR’s experts made the same assumption for 19 purposes of their opinions and calculations. (Doc. 165-1 at 10 [Borrelli report: “In my opinion, Doe’s likely career path (if he actually pursues it), both in the but-for and actual 20 worlds, is limited to coaching for small DI schools or schools in DII or DIII (and would have been limited to small DI schools or schools in DII or DIII even if Doe had not been 21 expelled from ASU). In my opinion, the disciplinary proceedings against Doe delayed, but did not preclude, his ability to obtain employment as a DI wrestling coach at one of the 22 smaller DI schools.”]; Doc. 165-1 at 154 [Duncan report: “[A]fter graduating from ASU, it is reasonable to assume that Mr. Doe would have secured a position as a Division II/III 23 assistant coach, and would likely have remained there for approximately four years. Assuming success in that position, Mr. Doe could have then applied for a Division I 24 assistant coach position.”].) Additionally, another of Doe’s experts, Owen, opined that Doe could have pursued a career as a “collegiate” coach. (Doc. 158-1 at 4 [“I believe the 25 two years that Doe lost cost him many achievements and accolades that would have allowed him to increase his marketability (sponsorships and employment) and earning 26 power down the road, should he choose to make a living as an athlete and then as a coach at the collegiate . . . level.”].) The Court has now held that Owen may offer this particular 27 opinion at trial. (Doc. 233 at 8-9 [“Owen is qualified to offer opinions about a collegiate wrestling coach’s typical qualifications . . . [and] has adequately explained how his 28 experiences informed his opinions regarding collegiate coaching qualifications.”].) Accordingly, Dr. Kaufman’s assumption has a sufficient enough basis to avoid exclusion.

- 35 - Case 2:18-cv-01623-DWL Document 250 Filed 11/18/22 Page 36 of 46

1 II. ABOR’s Motion For Reconsideration (Doc. 227) 2 A. Legal Standard 3 “The Court will ordinarily deny a motion for reconsideration of an Order absent a 4 showing of manifest error or a showing of new facts or legal authority that could not have 5 been brought to its attention earlier with reasonable diligence.” LRCiv. 7.2(g)(1). 6 Reconsideration is an “extraordinary remedy” that is available only in “highly unusual 7 circumstances.” Kona Enters., Inc., 229 F.3d at 890 (internal quotation marks omitted). 8 “Motions for reconsideration are disfavored . . . and are not the place for parties to make 9 new arguments not raised in their original briefs. Nor is reconsideration to be used to ask 10 the Court to rethink what it has already thought.” Motorola, Inc. v. J.B. Rodgers 11 Mechanical Contractors, 215 F.R.D. 581, 582 (D. Ariz. 2003) (citations omitted). 12 B. Analysis 13 In the August 30, 2022 order, the Court concluded that ABOR’s motion for 14 summary judgment should be denied because “Doe’s proffered evidence could lead a 15 reasonable juror not only to conclude that his disciplinary proceeding was marred by an 16 array of procedural irregularities, but also that ASU’s disciplinary process generates 17 statistical anomalies that raise an inference of gender bias.” (Doc. 209 at 60.) ABOR 18 challenges both of these conclusions in its motion for reconsideration. In Part A, ABOR 19 identifies reasons why it was improper to rely on the statistical anomalies identified by Dr. 20 Kaufman. (Doc. 227 at 2-10.) In Part B, ABOR identifies reasons why it was improper to 21 rely on the alleged irregularities. (Id. at 10-17.) 22 1. Statistical Evidence 23 ABOR’s seven reconsideration arguments concerning Dr. Kaufman require little 24 additional discussion because they are largely addressed in Part I above. To the extent 25 those arguments are not offered as relevance/admissibility objections but as evidentiary- 26 weight objections, the Court is unpersuaded: (1) it was not manifestly erroneous to rely on 27 Dr. Kaufman’s opinions for purposes of evaluating ABOR’s entitlement to summary 28 judgment because the Ninth Circuit has repeatedly recognized that statistics such as Dr.

- 36 - Case 2:18-cv-01623-DWL Document 250 Filed 11/18/22 Page 37 of 46

1 Kaufman’s are useful and relevant (even if not dispositive) in cases involving allegations 2 of intentional discrimination, including Title IX cases such as this one (Schake; Doe; Gay); 3 (2) although Dr. Kaufman’s acknowledgement that the statistical disparities may be 4 explained by non-discriminatory factors suggests the statistics are alone insufficient to 5 create a triable issue of fact (Gay; Aragon; Rose), that principle does not require 6 reconsideration here because Doe does not rely solely on Dr. Kaufman’s statistics; (3) the 7 parties’ dispute over whether the comparators chosen by Dr. Kaufman are similarly situated 8 to Doe is a factual dispute to be resolved by the jury at trial, not a basis for disregarding 9 Dr. Kaufman’s statistics for summary judgment purposes (Freyd; Alaska Rent-A-Car); 10 (4) although Dr. Kaufman’s failure to consider the factual details set forth in the Title IX 11 Case Tracking Report may render his F-23 analysis impeachable, the circumstances here 12 are sufficiently distinguishable from Pottenger that his statistics retain evidentiary value at 13 summary judgment; (5) there was no error, manifest or otherwise, in the Court’s 14 determination that it was reasonable for Dr. Kaufman to treat all F-23 violations as 15 comparable in part because ASU has decided that all F-23 violations are unified by a 16 population of similarly situated respondents in some material respect; (6) ABOR’s burden- 17 shifting arguments have no bearing on the summary judgment analysis; and (7) ABOR’s 18 new evidence does not provide a basis for reconsideration because ABOR did not provide 19 it in support of its summary judgment motion or in support of its exclusion motion, and 20 instead untimely provided it for the first time as an exhibit to its reconsideration request. 21 2. Roe’s December 11, 2016 Statement 22 This leaves ABOR’s reconsideration arguments concerning irregularities. The first 23 procedural irregularity identified in the summary judgment order was as follows: “[A] 24 reasonable juror could conclude that it was procedurally irregular for Davis not to share 25 Roe’s December 11, 2016 comments with Doe before submitting her final investigative 26 report to the Committee and that it was also procedurally irregular for Davis to assure Doe’s 27 counsel that Roe ‘did not offer any new evidence’ during the December 11, 2016 interview. 28 The former was procedurally irregular because, under § C(5) of the Procedures, ‘[b]efore

- 37 - Case 2:18-cv-01623-DWL Document 250 Filed 11/18/22 Page 38 of 46

1 concluding the investigation, and upon request, the Dean of Students will provide the 2 parties with an opportunity to respond to all investigative materials.’ The latter was 3 procedurally irregular because, for the reasons discussed above, a reasonable juror could 4 conclude that Roe did, in fact, provide important new evidence during the December 11, 5 2016 interview, by changing her account on several key points.” (Doc. 209 at 38-39.) 6 ABOR now seeks reconsideration of this analysis for four reasons: (1) the Court stepped 7 outside its judicial role by “go[ing] to lengths never attempted by Doe . . . to try to identify 8 ‘evidence’ appearing to create a question of fact” on whether Roe’s December 11, 2016 9 interview contained new information; (2) Roe’s December 11, 2016 interview did not, in 10 fact, contain any new information; (3) even assuming the interview contained new 11 information, “Davis did not violate any Procedure in stating that Roe ‘did not offer any 12 new evidence’”; and (4) even assuming there was some procedural violation, it was 13 harmless. (Doc. 227 at 10-14.) 14 ABOR is not entitled to reconsideration based on these arguments. First, ABOR’s 15 accusation of judicial overreach is, like its earlier accusation on this point, both 16 unsupported by the record and disappointing. In its summary judgment motion, ABOR 17 argued that, to the extent Doe sought to rely on “Davis’s statement to Doe that one of Ms. 18 Roe’s written submissions [from December 11, 2016] did not contain new evidence” as a 19 procedural irregularity, this reliance was misplaced because “[a]t best, this assertion 20 reflects a disagreement between Doe and Ms. Davis as to what constitutes ‘new evidence’” 21 and alternatively because “Doe received the opportunity to respond to all information 22 provided by Ms. Roe and did provide lengthy responses, including on December 21, 2016.” 23 (Doc. 155 at 10.) In response, Doe argued that one of the “egregious irregularities in the 24 disciplinary proceedings” occurred when “[t]he University tried to keep Doe from 25 responding to new evidence and ignored the response after receiving it.” (Doc. 189 at 11, 26 14.) Doe elaborated: “Parties have the right to ‘respond to all investigative materials.’ 27 Here, Davis refused to grant Doe that right until he hired an attorney who secured the right. 28 The University claimed no substantive new information was provided, but Doe’s attorney

- 38 - Case 2:18-cv-01623-DWL Document 250 Filed 11/18/22 Page 39 of 46

1 identified substantive, new information presented by Doe, and detail[ed] the egregious 2 flaws, inconsistencies, baseless conclusions, and legal infirmities with the draft report. . . . 3 A reasonable jury can infer gender bias based on the University’s failure to give the male 4 accused an opportunity to review and comment on evidence and granting the female 5 accuser that right . . . .” (Id. at 14-15.) Doe also included, as an exhibit to his response, a 6 copy of the December 21, 2016 letter from his attorney that purported to identify the new 7 information provided by Roe during the December 11, 2016 interview. (Doc. 189-2 at 52- 8 60.) Among other things, this letter stated that Roe provided new information when she 9 “[d]enie[d] being aware of video/photo being taken” and identified earlier statements in 10 which Roe purportedly made conflicting claims on that issue. (Id. at 54.) 11 The parties’ dispute over whether Roe’s December 11, 2016 interview contained 12 new substantive information—and whether ASU engaged in procedural impropriety by not 13 giving Doe a fair opportunity to address that new information—was also the subject of 14 extensive discussion during oral argument. After ABOR’s counsel argued that “Doe does 15 not describe in his papers the supposedly new information in Roe’s comments, or what 16 makes Miss Davis’s statement a misrepresentation at all” (Doc. 228 at 21), the Court posed 17 the following question to Doe’s counsel: “What’s your best piece of new substantive 18 information?” (Id. at 74.) Doe’s counsel responded: “I have a list. But the best one is, 19 that was the first time she said that she didn’t know she was being filmed. . . . [J]uxtapose 20 that with her [earlier] statements to the police, that not only did I know I was being filmed, 21 but that I stopped it, and I had the capacity to stop it, is evidence that she wasn’t 22 incapacitated . . . . So it’s substantive evidence that was never presented and it was the 23 first time that she presented it.” (Id. at 74-75.) In reply, ABOR’s counsel stated: “[Doe’s 24 counsel] said that the biggest inconsistency was [Roe’s] statement about not knowing she 25 was being filmed. There’s nothing in the record where she ever said that she knew she was 26 being filmed and stopped it. The evidence of what she actually said is she saw a flash. 27 That’s been extrapolated—now video was taken. She didn’t know. But once you see a 28 video is taken, all of a sudden she was stopping them from filming. That’s just not in the

- 39 - Case 2:18-cv-01623-DWL Document 250 Filed 11/18/22 Page 40 of 46

1 record.” (Id. at 82.) 2 Given this backdrop, the Court included, in the summary judgment order, a detailed 3 description of the various statements that Roe made about her awareness of being 4 filmed/videotaped, which included quotations from various documents in the summary 5 judgment record. (Doc. 209 at 38 [“Roe stated on December 11, 2016 that ‘she was not 6 aware that a photo was taken.’ However, during her April 2016 interview with the Tempe 7 Police Department, Roe reported that ‘she saw a camera flash and realized that [Witness 8 1] was taking pictures of her’ . . . . Similarly, during Roe’s initial meeting with Davis on 9 September 19, 2016, Roe stated that ‘[d]uring the encounter, she stated to see flashes going 10 off. She asked what it was and was told not to worry about it. She then responded that she 11 did not want to be recorded.’ And during her subsequent meeting with Davis on October 12 27, 2016, Roe stated that ‘the camera was brought out after she said no.’”].) Based on this 13 summary, the Court concluded that “a reasonable juror could conclude that Roe did, in fact, 14 provide important new evidence during the December 11, 2016 interview, by changing her 15 account on several key points.” (Id. at 39.) It is difficult to see how ABOR could view 16 this as an example of “assum[ing] Doe’s burden,” “act[ing] as a party’s lawyer,” and 17 deciding “to venture outside the adversarial issues presented to the Court by the parties.” 18 (Doc. 227 at 11, 13, cleaned up.) The Court resolved a factual dispute that both parties 19 expressly raised in their summary judgment papers and expressly raised again during oral 20 argument. Analyzing an expressly disputed issue is not judicial advocacy, but the normal 21 discharge of the judicial role. 22 To the extent ABOR’s claim of judicial advocacy concerns the Court’s 23 identification in the summary judgment order of two other examples of new information 24 provided by Roe during the December 11, 2016 interview—first, Roe’s statement that she 25 was too intoxicated to be aware of what was happening or to remember what happened; 26 and second, Roe’s statement that she did not say stop because she was too intoxicated (Doc. 27 209 at 37-38)—these examples were also identified in Doe’s counsel’s letter (Doc. 189-2 28 at 52-55, 59), which Doe cross-referenced in his summary judgment response (Doc. 189 at

- 40 - Case 2:18-cv-01623-DWL Document 250 Filed 11/18/22 Page 41 of 46

1 14). Accordingly, the inclusion of these examples in the summary judgment order was not, 2 as ABOR argues, an improper attempt to identify evidence “for the first time . . . which 3 Doe did not cite in response to ABOR’s Motion” (Doc. 227 at 11), but the Court’s best 4 effort to fully and fairly address each side’s arguments and proffered evidence. 5 ABOR has also failed to demonstrate that the Court’s resolution of this dispute was, 6 on the merits, manifestly erroneous. A reasonable juror could find that Roe made 7 statements during her December 11, 2016 interview that were inconsistent with her 8 previous statements. Although ABOR attempts in its reconsideration motion to show how 9 all of Roe’s statements can be reconciled with each other (Doc. 227 at 11-12), accepting 10 these arguments would, at a minimum, require the Court to resolve various inferences in 11 ABOR’s favor (which is impermissible at summary judgment). 12 Because a reasonable juror could find that Roe’s statements were inconsistent, a 13 reasonable juror could also find that it was procedurally irregular for Davis to assure Doe’s 14 counsel that Roe “did not offer any new evidence” during the December 11, 2016 15 interview. The Court fully agrees with Doe on this point: “[W]hile ABOR does not need 16 to use any specific language when conveying to a party that no new information was 17 received from the complainant, ASU cannot lie and say no new evidence was offered.” 18 (Doc. 241 at 14.) Without belaboring the issue, Roe’s account of the sexual encounter was 19 perhaps the most important evidence in the proceeding against Doe. Accordingly, evidence 20 bearing on Roe’s credibility—such as whether her description of the encounter varied over 21 time, even if those descriptions were not flatly inconsistent with each other—was critical. 22 Cf. Doe v. Univ. of Conn., 2020 WL 406356, *4 (D. Conn. 2020) (noting that “in any case 23 involving a ‘he said/she said’ dispute, . . . evidence bearing on credibility is critical”). See 24 generally Jencks v. United States, 353 U.S. 657, 667 (1957) (“Every experienced trial judge 25 and trial lawyer knows the value for impeaching purposes of statements of the witness 26 recording the events before time dulls treacherous memory. Flat contradiction between the 27 witness’ testimony and the version of the events given in his reports is not the only test of 28 inconsistency. The omission from the reports of facts related at the trial, or a contrast in

- 41 - Case 2:18-cv-01623-DWL Document 250 Filed 11/18/22 Page 42 of 46

1 emphasis upon the same facts, even a different order of treatment, are also relevant to the 2 cross-examining process of testing the credibility of a witness’ trial testimony.”). The 3 Court thus finds no error, manifest or otherwise, in its determination in the summary 4 judgment order that a reasonable juror could find that it was procedurally irregular for the 5 lead investigator in a sexual misconduct investigation to falsely assure the respondent that 6 the complainant had not offered any new evidence during her last pre-hearing interview 7 and then provide a summary of the last interview to the decisionmakers without first giving 8 the respondent an opportunity to review it.19 9 Finally, ABOR is not entitled to reconsideration based on its contention that any 10 procedural irregularity can be disregarded on harmlessness grounds. First, and as noted in 11 the summary judgment order, this argument is forfeited because it is a new theory that 12 ABOR did not properly raise in the summary judgment motion. (Doc. 209 at 43.) The 13 words “causation” and “harmless” do not appear anywhere in that motion (Doc. 155), 14 which argued that ABOR was entitled to summary judgment because there was no evidence 15 of external pressure to engage in gender bias (id. at 4-5), no evidence of comments 16 reflecting gender bias (id. at 5-7), no evidence of any procedural irregularities (id. at 7-13), 17 and no admissible expert evidence to support Doe’s claim (id. at 13-17). The argument 18 that no procedural irregularity occurred is very different from the argument that any 19 procedural irregularity was harmless. Second, even if the argument weren’t forfeited, Doe 20 has identified various reasons why a reasonable juror could view the irregularity 21 concerning Roe’s December 11, 2016 statement as harmful. (Doc. 241 at 14-15.) Third, 22 and more broadly, the Court is not persuaded that a procedural irregularity in a Title IX 23 19 In the summary judgment order, the Court concluded that a reasonable juror could 24 find procedural impropriety not only for the reasons stated above, but also because Davis’s approach violated § C(5) of the Code. In the motion for reconsideration, ABOR argues 25 that it was manifestly erroneous to conclude that § C(5) was implicated here. (Doc. 227 at 13.) The Court is not necessarily persuaded. True, one of Davis’s December 20, 2016 26 emails included the statement “I am happy to have you come in to view her responses” (Doc. 189-2 at 63), but Davis had already sent the final report to the Dean for review by 27 the time she made this offer (id. at 64). Nevertheless, regardless of any error in the interpretation of § C(5), the summary judgment order’s overall determination that a 28 reasonable juror could find procedural irregularity as to Davis’s treatment of Roe’s December 11, 2016 statement was not manifestly erroneous.

- 42 - Case 2:18-cv-01623-DWL Document 250 Filed 11/18/22 Page 43 of 46

1 case becomes irrelevant if the plaintiff cannot draw a direct causal link between that 2 irregularity and the outcome of the underlying disciplinary proceeding. Here, Doe seeks 3 to rely on an array of circumstantial evidence, including statistical evidence of disparities, 4 various procedural irregularities, and the fact that his expulsion was ultimately overturned 5 based on insufficient evidence, in an attempt to prove the existence of gender bias. As 6 noted in Gay, “[s]ince the question is one of inferring discriminatory intent, the district 7 court should make a sensitive inquiry into such circumstantial and direct evidence of intent 8 as may be available. . . . [A]ll evidence, both direct and circumstantial, statistical and 9 nonstatistical, relevant to that question should be assessed on a cumulative basis.” 694 10 F.2d at 550. Disregarding individual procedural irregularities on harmlessness grounds 11 would, at least under the circumstances of this case, violate this directive to conduct a 12 sensitive inquiry that considers the cumulative impact of all of Doe’s evidence. 13 3. Failure To Disclose “Sex By Force” Theory 14 The second procedural irregularity identified in the summary judgment order was 15 as follows: “A reasonable juror could conclude that the UHB’s decision to find Doe 16 responsible under a sex-by-force theory constituted a procedural irregularity. . . . [T]he 17 summary of the charges provided at the outset of the letter [sent by Dr. Hunter on 18 September 21, 2016] only accused Doe of engaging in sexual acts ‘without her consent’ 19 and did not make any separate mention of sex by force. Similarly, during later meetings 20 with Doe, Davis made comments that can be reasonably construed as suggesting the 21 sexual-misconduct investigation was only focused on incapacitation and the lack of 22 consent, not on the presence of force.” (Doc. 209 at 42-43.) ABOR now seeks 23 reconsideration of this analysis for four reasons: (1) the “undisputed evidence” shows that 24 ASU properly disclosed the sex-by-force theory; (2) alternatively, there is no requirement 25 under the Procedures to provide “a detailed description of each theory of sexual 26 misconduct”; (3) finding a disclosure violation under these circumstances would create an 27 “inherent contradiction” with the Court’s relevance and reliability determinations 28 concerning Dr. Kaufman; and (4) any irregularity was harmless. (Doc. 227 at 14-16.)

- 43 - Case 2:18-cv-01623-DWL Document 250 Filed 11/18/22 Page 44 of 46

1 ABOR is not entitled to reconsideration based on these arguments. The first and 2 second arguments are improper because they essentially repeat the same arguments that 3 ABOR made (and the Court rejected) during the summary judgment process. Motorola, 4 215 F.R.D. at 582 (“Nor is reconsideration to be used to ask the Court to rethink what it 5 has already thought.”). At any rate, even after considering ABOR’s latest recitation of 6 these arguments, the Court stands by its earlier conclusion. As Doe correctly notes in his 7 response, even though it is undisputed that “Doe was ‘charged’ with ‘Sexual Misconduct’ 8 under F-23 . . . the Dean of Students was required to explain those charges, and she did: by 9 explaining that Doe was being accused of engaging in sex with a person who was 10 incapacitated. . . . [T]here is sufficient evidence for a jury to find that the decision to find 11 Doe responsible for sex by force without notice constituted a procedural irregularity.” 12 (Doc. 241 at 16-17; Doc. 209 at 42-43.) 13 Nor does the Court find any inherent contradiction between this conclusion and the 14 analysis regarding Dr. Kaufman. The issue here concerns notice and compliance with 15 procedural regulations—specifically, whether it is procedurally irregular to inform a 16 student that he is being charged with one specific form of “sexual misconduct” under F- 17 23, only to find him responsible for committing an entirely different form of “sexual 18 misconduct” under F-23. The issue as to Dr. Kaufman was whether a statistical analysis 19 of all “sexual misconduct” cases under F-23, irrespective of the underlying facts, would 20 fall within Rule 702’s broad conception of relevance. Although both analyses touch, in a 21 broad sense, on the same subject matter, they turn on very different considerations. 22 Finally, ABOR’s harmlessness arguments as to the sex-by-force issue fail for the 23 same reasons as its harmlessness arguments regarding Roe’s December 11, 2016 24 statements—they are forfeited, factually unsupported for the reasons stated in Doe’s 25 response (Doc. 241 at 17), and legally misplaced. 26 4. Arizona Court Of Appeals Decision 27 The final irregularity identified in the summary judgment order was as follows: 28 “[T]he fact of reversal [by the Arizona Court of Appeals] helps support Doe’s contention

- 44 - Case 2:18-cv-01623-DWL Document 250 Filed 11/18/22 Page 45 of 46

1 that his disciplinary proceeding was infected by irregularities that may, in concert with 2 other evidence, give rise to an inference of gender bias. The clearest support for this 3 conclusion comes from the Ninth Circuit’s recent Doe decision, which identified ‘the state 4 court’s ruling . . . in the writ proceeding . . . that the evidence did not support the Regents’ 5 findings’ as one of the ‘procedural irregularities’ that could ‘support an inference of gender 6 bias, particularly when considered in combination with allegations of other specific 7 instances of bias and background indicia of sex discrimination.’ That is essentially what 8 happened here—although Dr. Hightower, the UHB, and Dr. Rund all determined that Doe 9 had committed sexual misconduct in violation of section F-23, the appellate court held that 10 no reasonable person could have reached that conclusion. Although the Court agrees with 11 ABOR that the fact of reversal does not, in itself, mean that Doe must prevail on his Title 12 IX claim or even survive summary judgment—there must be other evidence from which a 13 reasonable juror could conclude this irregularity was indicative of gender bias—it would 14 be reasonable for a juror to find irregularity based on the reversal.” (Doc. 209 at 45, 15 citations omitted). ABOR now seeks reconsideration of this analysis for two reasons: (1) 16 Doe never disclosed his intent to rely on the Arizona Court of Appeals’ decision as 17 evidence of a procedural irregularity; (2) on the merits, the decision does not qualify as 18 evidence of an irregularity. (Doc. 227 at 16-17.) 19 ABOR is not entitled to reconsideration based on these arguments. First, these are 20 essentially the same arguments that ABOR made (and the Court rejected) during the 21 summary judgment process. Motorola, 215 F.R.D. at 582. Second, the Court agrees with 22 Doe (Doc 241 at 18-19) that these arguments fail on the merits—Doe disclosed his intent 23 to rely on the decision (Doc. 155-5 at 100-02) and a reasonable juror could find an 24 irregularity based on the reversal, as recognized by the Ninth Circuit in Doe. 25 … 26 … 27 … 28 …

- 45 - Case 2:18-cv-01623-DWL Document 250 Filed 11/18/22 Page 46 of 46

1 Accordingly, 2 IT IS ORDERED that: 3 1. ABOR’s motion to exclude Dr. Kaufman (Doc. 159) is denied. 4 2. ABOR’s motion for reconsideration (Doc. 227) is denied. 5 Dated this 18th day of November, 2022. 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28

- 46 -