Huang v. Ohio State University, No. 23-3469 (6th Cir. 2024)

Annotate this Case

Meng Huang, a former Ph.D. student at The Ohio State University (OSU), alleged that her advisor, Professor Giorgio Rizzoni, sexually harassed and assaulted her during her studies. Huang filed a lawsuit against OSU and Rizzoni, claiming Title VII quid pro quo sexual harassment and retaliation against OSU, and a due process violation against Rizzoni under 42 U.S.C. § 1983.

The United States District Court for the Southern District of Ohio granted summary judgment to OSU on Huang’s Title VII claims, concluding she was not an "employee" under Title VII until August 2017. The court also ruled that Huang’s retaliation claim failed because her first protected activity occurred after the alleged adverse actions. Huang’s § 1983 claim against Rizzoni proceeded to trial, where the court trifurcated the trial and excluded evidence of Rizzoni’s alleged manipulation and coercion. The jury found in favor of Rizzoni.

The United States Court of Appeals for the Sixth Circuit reviewed the case. The court found that the district court erred in determining Huang was not an employee under Title VII before August 2017, as her research work and the control Rizzoni exerted over her indicated an employment relationship. The court also held that Huang’s resistance to Rizzoni’s advances constituted protected activity under Title VII, and she presented sufficient evidence of adverse employment actions linked to her resistance.

The Sixth Circuit reversed the district court’s summary judgment on Huang’s Title VII claims, vacated the trial verdict in favor of Rizzoni on the § 1983 claim, and remanded for a new trial. The court emphasized that the district court’s exclusion of relevant evidence regarding Rizzoni’s power and manipulation was an abuse of discretion, which prejudiced Huang’s ability to present her case.

Download PDF

RECOMMENDED FOR PUBLICATION Pursuant to Sixth Circuit I.O.P. 32.1(b) File Name: 24a0203p.06 UNITED STATES COURT OF APPEALS FOR THE SIXTH CIRCUIT MENG HUANG, Plaintiff-Appellant, > v. THE OHIO STATE UNIVERSITY; GIORGIO RIZZONI, Defendants-Appellees. No. 23-3469 Appeal from the United States District Court for the Southern District of Ohio at Columbus. No. 2:19-cv-01976—James L. Graham, District Judge. Argued: March 21, 2024 Decided and Filed: August 28, 2024 Before: KETHLEDGE, READLER, and BLOOMEKATZ, Circuit Judges. _________________ COUNSEL ARGUED: Hugh T. McKeegan, OBERMAYER REBMANN MAXWELL & HIPPEL LLP, Pittsburgh, Pennsylvania, for Appellant. Jeffrey C. Gerish, PLUNKETT COONEY, Bloomfield Hills, Michigan, for Appellee. Alexandra Z. Brodsky, PUBLIC JUSTICE, Washington, D.C., for Amici Curiae. ON BRIEF: Bruce C. Fox, Andrew J. Horowitz, OBERMAYER REBMANN MAXWELL & HIPPEL, Pittsburgh, Pennsylvania, for Appellant. Jeffrey C. Gerish, Christina L. Corl, PLUNKETT COONEY, Bloomfield Hills, Michigan, for Appellee. Alexandra Z. Brodsky, PUBLIC JUSTICE, Washington, D.C., for Amici Curiae. BLOOMEKATZ, J., delivered the opinion of the court in which KETHLEDGE, J., joined in full. READLER, J. (pp. 37–48), delivered a separate opinion concurring in part and dissenting in part. No. 23-3469 Huang v. Ohio State Univ. Page 2 _________________ OPINION _________________ BLOOMEKATZ, Circuit Judge. Meng Huang, a former engineering student at The Ohio State University, alleges that her advisor, Professor Giorgio Rizzoni, sexually harassed and assaulted her while she pursued her Ph.D. She brought this action against OSU and Rizzoni. Relevant on appeal are the district court’s grant of summary judgment to OSU on Huang’s Title VII quid pro quo sexual harassment and retaliation claims, and a trial verdict in Rizzoni’s favor on Huang’s due process claim that he violated her right to bodily integrity. The district court mishandled Huang’s claims against both defendants. First, with respect to the claims against OSU, Huang raised a material dispute of fact as to whether she was an “employee” of OSU for Title VII purposes, even though she was also a student. The two roles are not mutually exclusive, as the district court mistakenly held. So, Huang’s sexual harassment claim under Title VII must go to a jury. Additionally, when Huang physically resisted Rizzoni’s advances by pushing him away, she “opposed” his improper touching even if she did not vocalize her objection. That is enough for “protected activity” under Title VII, so her retaliation claim also should have survived. Second, with respect to the claim against Rizzoni, the district court violated Huang’s substantial rights when it excluded multiple categories of evidence in service of a confusing trifurcation order. Deeming this evidence “irrelevant” to whether Rizzoni improperly touched Huang, the district court categorically excluded evidence of Rizzoni’s threats to her academic career after she refused his advances. It also prevented Huang from presenting evidence of her impeccable academic credentials and stellar performance during her Ph.D. program. She sought to introduce that evidence to rebut Rizzoni’s claims that she fabricated the sexual harassment allegations in revenge for his decision to eject her from the engineering program. Yet all this was relevant under basic rules of evidence. Given these errors, we reverse the grant of summary judgment, vacate the trial verdict, and remand the case to the district court. No. 23-3469 Huang v. Ohio State Univ. Page 3 BACKGROUND I. Factual History Meng Huang, a Chinese native, moved to Columbus, Ohio in August 2014 to begin a Ph.D. program at The Ohio State University (OSU). Giorgio Rizzoni, an Italian native, was Huang’s Ph.D. advisor. He is a tenured professor at OSU in the College of Engineering, and the director of the Center for Automotive Research (CAR). CAR is an interdisciplinary center that conducts research partnerships with automotive manufacturers and other external funders, including the United States government. Huang alleges that from the moment she first met Rizzoni and throughout her three and a half years as his advisee, he subjected her to unwanted touching of an increasingly sexual nature. She claims Rizzoni used his power as her Ph.D. advisor and director of CAR, as well as his distinguished position in their professional field, to coerce her to meet with him alone and subject her to harassment. She describes the constant threat that he could expel her from the program, rescind her pay, and ultimately cause her to lose her visa if she did not stay in his good graces. And she says that Rizzoni used that power to engage in a pattern of harassment whereby he praised her when she complied with his requests to meet alone and punished her when she resisted. But, after several years, Rizzoni grew so frustrated by her refusal to submit to his advances that Huang claims he rushed her Ph.D. candidacy exam, manipulated the examination committee so that it failed her, and then denied her the customary opportunity to retake it. Rizzoni tells an opposite story. He denies he ever touched Huang inappropriately and blames the breakdown in his relationship with Huang on her increasing refusal to meet with him or complete Ph.D.-level work. At trial, he testified he was the “victim” of Huang’s false allegations after she—because of her poor performance and unwillingness to follow his advice— failed her candidacy exam by a unanimous vote of the committee. Trial Tr. Vol. III, R. 219, PageID 8206. The legal claims arising from these allegations involve several fact-intensive inquiries. So, before analyzing them, we provide an extensive review of the record evidence. No. 23-3469 Huang v. Ohio State Univ. Page 4 We recognize that many of these facts are vigorously disputed, but at this stage we recite them in the light most favorable to Huang. Palma v. Johns, 27 F.4th 419, 423 (6th Cir. 2022). A. Huang’s Matriculation and First Year at OSU Before she matriculated to OSU, Huang received a master’s degree from Tongji University, a highly regarded university in Shanghai. She graduated at the top of her class. During her master’s program, Huang had an internship researching battery technology, and after graduation she worked as an engineer at Shanghai General Motors. Huang contacted Rizzoni in late 2013, when she was considering Ph.D. programs in the United States. Rizzoni encouraged Huang to apply. He described Huang’s school in China as “world class” and said that her master’s degree, plus her work experience, made her a “worthwhile” prospect for “continu[ing] the stream of work in battery electric modeling” that he had begun with previous Ph.D. students. Rizzoni Dep., R. 98-1, PageID 3620–21. He informed Huang that she could receive university funding, and expressed hope that Huang could meet with him in person when he traveled to Shanghai in the coming spring on unrelated business. Huang applied, and in March 2014, OSU admitted her to the Ph.D. program in its Department of Mechanical and Aerospace Engineering. Huang testified that Rizzoni first inappropriately touched her later that month, on March 29, 2014, when she met him at his hotel in Shanghai. She says he escorted her to his room, made her an espresso with a portable machine he traveled with, and sat uncomfortably close to her as she completed exam problems at his request on his laptop. As they discussed OSU, he grabbed her left hand with both of his as he told her that not only would her Ph.D. be completely free, but that he would give her “an extra bonus.” Trial Tr. Vol. I, R. 217, PageID 7833. Huang and Rizzoni then had dinner with the OSU China program director. At the dinner, Huang says Rizzoni grabbed her hand, turned to the director, and repeated “look at her, look at her,” while gesturing to Huang. Id., PageID 7838. Rizzoni does not dispute that they met at the hotel and had dinner, but denies he brought Huang to his hotel room or touched her inappropriately. The next month, Rizzoni emailed Huang to offer her a Graduate Research Associate (GRA) position with him at CAR, which would cover her full tuition, fees, and a stipend of $1,900 per month. In the offer letter, Rizzoni told Huang, “In recognition of your excellent No. 23-3469 Huang v. Ohio State Univ. Page 5 academic record and of you[r] experience, I will augment your [GRA] stipend by . . . 10% using discretionary funds under my control.” GRA Offer, R. 204, PageID 7648. Huang accepted the GRA position and enrolled in the Ph.D. program in the engineering department. In the summer before she arrived, Rizzoni informed Huang that she had been “selected to be funded as a United States Department of Energy . . . Graduate Fellow for the first two years of [her] studies (with the same stipend and benefits outlined in [the] original offer letter),” and stressed that this was a “prestigious appointment.” DOE Fellow Offer, R. 204, PageID 7658. While the offer letter said that she would not have any “specific” research assignments, it did assign her general research topic, mandate both regular group meetings and individual meetings with Rizzoni, and guarantee access to OSU laboratories for the completion of the assigned research topic. Id. Thus Huang had accepted two overlapping offers: (1) enrollment at OSU in the engineering department’s Ph.D. program; and (2) a Graduate Fellow position that would fund her tuition and provide her with a stipend and extra bonus. Under the prior GRA offer, OSU would have classified Huang as an “employee,” but under this Graduate Fellow position, it classified her as a “student.” Bons Aff., R. 12-2, PageID 147. Although OSU changed Huang’s classification for internal administrative purposes, nothing changed regarding Rizzoni’s expectations of Huang or her benefits.1 Indeed, Rizzoni called Huang a “graduate research associate” in his deposition. Rizzoni Dep., R. 99-1, PageID 3983. Huang began at OSU in August 2014, as planned. She took classes in the engineering department and conducted research at CAR. Rizzoni was her Ph.D. advisor. Upon her arrival, Rizzoni quickly assigned Huang to conduct research on an ongoing research project that was sponsored by the Ford Motor Company. This project—the Ford University Research Project (URP)—focused on modeling battery aging for electric vehicles, and continued research started by two of Rizzoni’s previous graduate students. With the URP, Ford paid OSU to conduct company research. Rizzoni had recruited Huang to work specifically on this project because of her background, and it doubled as her dissertation topic. 1 Rizzoni’s deposition testimony suggests the school changed her title so it could pay her stipend with funds it had received from the Department of Energy while abiding by its internal accounting rules. No. 23-3469 Huang v. Ohio State Univ. Page 6 Rizzoni actively supervised Huang’s research on the URP. CAR had its own building separate from other parts of the engineering department. And Rizzoni assigned Huang a desk near his office; he told his assistant “I would like her to be close to me.” Trial Tr. Vol I, R. 217, PageID 7865. Rizzoni kept in close contact with his graduate student researchers. He would meet with them regularly—often on the weekends—and sometimes drive them to CAR from their homes. Huang was no exception. Rizzoni required frequent one-on-one meetings with Huang to discuss her research and sometimes drove her to CAR. Huang alleges that during these meetings and while driving together, Rizzoni would grab and rub her hands, shoulders, and waist in an affectionate but highly inappropriate and uncomfortable manner. Huang also had to attend bi-weekly virtual meetings with Rizzoni and Ford engineers, and sometimes travel to Dearborn, Michigan to meet with engineers at Ford headquarters. She traveled to Dearborn twice with Rizzoni during her first year at OSU (in January and April 2015). On these trips, Huang says that Rizzoni would encourage her to sit next to him for presentations and kiss her goodbye (on the cheek) after he dropped her off at the end of the trip. On the second drive to Dearborn, Huang testified that Rizzoni wanted to demonstrate the massaging seats in his new car to her and another Ph.D. student who was seated in the rear. Huang says he grabbed her hand, shoved it under his hip, and said “Feel it, feel it, feel it. Meng, Feel the massage.” Trial Tr. Vol. II, R. 218, PageID 7912–13. At trial, the other graduate student who was in the car testified that Huang had been seated in the back of the car and that the incident with the massaging seats did not occur. Huang declined an invitation to travel with Rizzoni to Dearborn for a third time during the first semester of her second year. She told Rizzoni she felt the brief meeting would not be worth the trip, and called in from Columbus. Rizzoni chided Huang for missing the opportunity. B. Huang’s Second Year at OSU During the fall semester of 2015, Rizzoni asked Huang to have dinner with him at an Italian restaurant in Columbus. Rizzoni does not deny that the dinner occurred but does deny Huang’s version of the events: that Rizzoni ordered two glasses of wine and told Huang they could share tastes of each; that Rizzoni rubbed her thighs through her dress; that as she tried to push his hand away he told Huang, “Meng, I pay your stipend. You better listen to me. No. 23-3469 Huang v. Ohio State Univ. Page 7 Otherwise . . . we’re going to have a problem.” Trial Tr. Vol. II, R. 218, PageID 7917. Huang alleges that Rizzoni’s touching became more sexual after this dinner—during meetings he would grab her hands, touch her face, and hold her shoulder while rubbing her back and lap. Huang began to pull away from Rizzoni, which irritated him. For instance, when Huang said she wasn’t available to meet with Rizzoni on December 27, 2015, he told her, “Too bad . . . It looks as if we are going to miss out on this opportunity to have some quiet time and make progress.” Dec. 27, 2015 Email, R. 94-4, PageID 2386. And Rizzoni objected when he learned Huang had enrolled in academic courses not focused on engineering. He claimed these classes “took away from her research time” and meant “she wouldn’t meet with [him] to discuss her research.” Rizzoni Dep., R. 99-1, PageID 4030–31. He told her to deregister from them. In January 2016, Huang and Rizzoni met in his office on the weekend. According to Huang, during the meeting Rizzoni rolled his office chair up to her as she was demonstrating something on his computer, grabbed her hand, and squeezed her legs between his thighs. The arrangement of the furniture made it so Huang could not move away, and Rizzoni was “short,” so Huang’s leg was uncomfortably close to his groin. Huang Report, R. 98-3, PageID 3790. In the days following this incident, Rizzoni discovered that Huang had disobeyed his instruction to deregister from classes outside of her Ph.D. program and sent her an email conveying that such disobedience meant he wasn’t “getting [his] money’s worth.” It read: You are the most stubborn student I have ever come across. The stupidest thing a Ph.D. student can do is to ignore the advice of their advisor. If you cannot see that, then you do not belong in our Ph.D. program. . . . I am clearly not getting my money’s worth. Jan. 31, 2016 Email, R. 94-4, PageID 2374. Rizzoni started to ask Huang to review academic articles he received as a peer reviewer. Huang declined some of these assignments as beyond the scope of her “responsibility and capability as a second year [mechanical engineering] Ph.D. student” and Rizzoni criticized Huang for not doing this work. Huang Report, R. 98-3, PageID 3789. For the rest of the spring 2016 semester, Huang alleges Rizzoni routinely touched her inappropriately when they met in No. 23-3469 Huang v. Ohio State Univ. Page 8 person. During this time, Rizzoni criticized Huang for working in locations other than the CAR facility and not meeting with him in person as often as he believed she should have. C. Huang’s Third Year at OSU In Fall 2016, Huang returned to Columbus to begin her third year at OSU. That summer, Rizzoni had arranged for Huang to work at Ford in Dearborn and for OSU to pay her during this time. Upon her return, Rizzoni expressed multiple frustrations with Huang. He grew angry when Huang refused to grade homework for him, continued to register for classes that he believed distracted from her Ph.D. candidacy, and did not meet with him when he requested. As Huang describes, shortly after the semester began, she told Rizzoni she couldn’t meet with him because she had recently been involved in a car accident. Rizzoni emailed Huang in response: You have disappeared. . . . You have really not understood what it means to be a Ph.D. student, and how fortunate you are to have funding for your studies. . . . Of all the Ph.D. students I have advised in my career (~40), you are positively the worst Ph.D. student I have ever advised. Primarily, because you do not listen to advice. . . . You are a net negative burden on me, because you cost me money and give back nothing. Sept. 21, 2016 Email, R. 94-4, PageID 2375. Shortly after she received this email, Huang met with Valerie Hendrickson, a law student dating one of Rizzoni’s other Ph.D. students. Given the email, she expressed concern that her degree was at risk and was visibly anxious. Hendrickson recommended Huang seek mental health counseling. During a November 2016 meeting, Huang requested to visit her family in China during the upcoming winter break (between semesters); Rizzoni approved. Having forgotten, Rizzoni was upset when Huang informed him she couldn’t meet with him during break, and told her in an email: You should not have made plans to leave the country for three weeks without checking with me first. . . . Do not forget that I am the one who pays the bills. Dec 3, 2016 Email, R. 94-4, PageID 2376. After Huang told Rizzoni he’d already approved her visit, he admitted that he’d forgotten since he had “[t]oo many things going on[.]” Id. No. 23-3469 Huang v. Ohio State Univ. Page 9 Huang testified that upon her return, Rizzoni sexually harassed her twice, and both instances were more severe than the previous harassment. On Sunday, February 5, she met Rizzoni in his office and gifted him a red Lunar New Year decoration that she’d bought for him in China. Rizzoni asked her to put it up in his office window, and Huang testified that as she leaned over a cupboard to hang the paper cut-out, Rizzoni walked up behind her and rubbed her buttocks. Afterward, the two discussed Huang’s research, and she says Rizzoni appeared excited and had a satisfied smile on his face. Soon after that incident, Rizzoni had knee surgery. At his request, Huang met him at his house on a Sunday to discuss a report Huang had prepared. It was a few weeks into his recovery and, according to Huang, Rizzoni could walk freely around the house despite his knee brace. Huang recounts that when they were alone in the house, Rizzoni held her by the waist as he showed her around the first floor of his home. He then led her to his kitchen, maneuvered her into a corner, and grabbed her breast. Disturbed, Huang pushed Rizzoni’s hand off her and went to sit down on the living room couch. Rizzoni followed her and sat down on a footstool just across from her. Then he again reached toward Huang; as she sunk into the couch, he started touching her hair, fondling her breast, and eventually rubbing his penis. Huang says Rizzoni sat with an “excited weird and dirty expression on his face[,] and kept touching himself (his penis) with his left hand.” Huang Report, R. 98-3, PageID 3796. Huang says she felt “almost dead inside under such tremendous humiliation” and that there was nothing she could do to resist Rizzoni in his own home. Id. When the meeting ended, Huang was distraught. Hendrickson, “Got time to talk for a sec? Girls talk She went to her car and texted .” Trial Ex. 138, R. 204, PageID 7665. Hendrickson replied that she was sick and preferred to text. Huang didn’t feel comfortable texting about the incident and demurred, and eventually decided not to involve Hendrickson in the situation since she was dating another of Rizzoni’s Ph.D. students. Huang further testified that she did not report any of these incidents to university officials because she wanted to remain in the Ph.D. program, keep her visa, get her degree, and maintain a good relationship with Rizzoni given his prominence in the field. For the rest of the semester, although Huang alleges that she attempted to avoid in-person encounters with Rizzoni, he berated her when she avoided No. 23-3469 Huang v. Ohio State Univ. Page 10 him and praised her when she would meet with him—and often those meetings involved unwanted touching. That summer, Huang returned to Dearborn to work at Ford. She received high praise for her work. Dyche Anderson, one of her supervisors, testified that his impression of her was “very, very, very positive” since “she caught on to things that . . . only someone with a deep understanding would be able to,” and that he was “impressed” with her “work ethic,” “focus and commitment,” “seriousness of purpose,” and her “earnestness and honesty.” Anderson Dep., R. 76-1, PageID 675–76. Anderson told Rizzoni that after Huang gave a presentation to a group of Ford engineers, they wanted to hire her right away and were excited to continue their work with her. Huang alleges that Rizzoni inappropriately rubbed her when he came to Michigan that summer for an event with Ford. D. Huang’s Last Semester with Rizzoni and Ph.D. Candidacy Examination Huang returned to Columbus in August 2017 for what would be her final semester studying with Rizzoni. Two things happened on the day that semester began. First, Huang’s appointment at OSU changed from a Graduate Fellow to a Graduate Research Associate—the appointment Rizzoni had originally offered her in April 2014. Rizzoni had decided to terminate Huang’s 10% supplemental stipend as punishment for what he saw as “her inattention” to completing her Ph.D., but otherwise the GRA appointment was a change in title rather than responsibilities. Rizzoni Dep., R. 99-1, PageID 4029–30. Second, Huang agreed to meet Rizzoni in his office. During this meeting, she alleges Rizzoni walked up behind her and grabbed her breasts as she was discussing her research with him. As the semester continued, Huang prepared for her candidacy exam, which Rizzoni had scheduled for December. In September and October 2017, Rizzoni reviewed drafts of Huang’s written candidacy exam proposal and told Huang she was “off to a good start,” had “made a very big step forward,” and that her proposal was “viable.” Sept. 19 and Oct. 4, 2017 Emails, R. 94-4, PageID 2403–05. After consultation with Huang, Rizzoni assigned two professors to be on the committee, and encouraged Huang to have Anderson, her supervisor from Ford, sit on her committee as well. No. 23-3469 Huang v. Ohio State Univ. Page 11 This momentum shifted about a month later. On Wednesday, November 15, 2017, Huang had a Webex meeting with Rizzoni and the Ford URP team. Rather than attend the Webex at CAR with Rizzoni, as he preferred, Huang called into the meeting from another building on OSU’s main campus. As Huang presented her research to the Ford team, Rizzoni interrupted her to complain that she wasn’t in his office with him and that she wouldn’t meet with him on weekends. Huang recounts that being berated made her “so angry,” because she believed Rizzoni wanted to meet with her in person so that he could grope her. Huang Report, R. 98-3, PageID 3803–04. She became audibly agitated on the call and yelled at Rizzoni that she needed to take the meeting away from CAR because of a computer issue, and that she did not want to meet with him in his office on the weekend. Rizzoni responded, insisting he meet with Huang in person and told Huang that “if you do not want to finish your Ph.D., we can quickly terminate it!” Id. Anderson, who sat in on the meeting, recalled in deposition that Huang’s “outburst” was “completely out of character.” Anderson Dep., R. 76-1, PageID 714–15. After discussing the incident with his wife, Anderson emailed Rizzoni, reporting that his wife advised him Huang’s comments “might not be about not wanting to work on a Sunday, but rather to be alone in a man’s office with no one else in the building. She said it would have made her uncomfortable.” Nov. 16, 2017 Email, R. 94-5, PageID 2421. Days later, on November 17, 2017, Rizzoni informed Huang that he had switched out one of the professors on her exam committee for Dr. Marcelo Canova, a close friend of Rizzoni’s who is expected to succeed him as the director of CAR. He also informed her that Anderson would not be on the exam committee. When Huang objected, he replied, “That is what[] happens when you never meet with me and never reply to my emails . . . things change.” Nov. 17, 2017 Email, R. 94-4, PageID 2409 (ellipses in original). In the weeks leading up to the exam, Rizzoni privately communicated his displeasure and frustration with Huang to the professors who would be evaluating her. He emailed one professor her written exam materials, and added, “I will talk to you about this student by voice.” Nov. 22, 2017 Email, R. 99-2, PageID 4137. The day before the exam, Rizzoni individually emailed an identical message to the three other members of the committee, explaining that Huang had been No. 23-3469 Huang v. Ohio State Univ. Page 12 “a challenge,” and that “I would like this exam to be a real exam. The outcome will be whatever it needs to be.” Dec. 7, 2017 Emails, R. 99-2, PageID 4139 (Guezennec), 4141–42 (Kim), 4143 (Canova). Rizzoni replied to one professor’s acknowledgement of his email that “the message might be terminal[.]” Dec. 7, 2017 Reply, R. 99-2, PageID 4139. Huang took her oral candidacy exam on December 8, 2017. During the exam, Huang faced tough questions from the committee members. She claims that though she ably answered each question, many of them were unfair or impertinent, designed to make her look bad. Dr. Canova testified in his deposition that he was “uncomfortable” during the exam because Huang’s answers were “confrontational” and “dismissive,” and that overall she was “not behaving appropriately.” Canova Dep., R. 190-1, PageID 7322. The department’s practice is that following a candidacy exam, the committee members independently complete their evaluations of the student’s performance. For Huang, Canova explained that “we were not supposed to discuss together, but we ended up discussing,” and the committee unanimously determined she had failed. Id. PageID 7334. Though Huang had failed, the committee still faced a choice whether to allow Huang to retake her exam. The committee—again contrary to the school’s policy of independent evaluation—allowed Rizzoni to decide whether Huang could retake her exam. Rizzoni met with Huang the Monday following the exam and, because he believed she had a poor attitude during the meeting, notified Huang and the committee members that he would not give her the option to retake the exam. Evidence in the summary judgment record suggests that Huang’s examination process was unusual. A former CAR Ph.D. student testified that the candidacy exam is “akin to a presentation of the dissertation topic,” which “feels like an informal seminar rather than a real exam.” Zhang Aff., R. 114-8, PageID 6352. The student also noted that advisors only schedule candidacy exams when they believe their advisee is ready. Canova also mused that he would “never have allowed the exam to happen,” since it “could have been postponed.” Canova Dep, R. 190-1, PageID 7331. Multiple OSU personnel, including one of the members of the exam committee, testified that it is rare for a Ph.D. student to fail a Ph.D. candidacy exam, and rarer still for them not to be allowed to retake it. Even Rizzoni, responding to a school administrator, No. 23-3469 Huang v. Ohio State Univ. Page 13 agreed the situation was “unusual.” Dec. 14, 2017 Email, R. 99-2, PageID 4158. Indeed, Huang was the only one of Rizzoni’s students to ever fail their candidacy exam. After Rizzoni told Huang that she would not be allowed to retake the exam, he emailed Anderson at Ford to let him know she wouldn’t be allowed to participate in the Ford URP. Huang also called Anderson to let him know the news and told him that Rizzoni had conducted himself improperly towards her but that she couldn’t discuss it with anyone she knew at the school, since they operated under Rizzoni’s influence. Dec. 12, 2017 Email, R. 94-5, PageID 2422–23. Anderson, inferring sexual harassment from Huang’s insinuating comments, advised her to report the harassment to OSU. The next day, December 12, 2017, Huang reported to the chair of the engineering department that Rizzoni had sexually harassed her throughout her candidacy, and that she believed he had rigged her candidacy examination. Anderson also emailed OSU’s address for reporting sexual harassment given his conversation with Huang. Concerned that Huang would be improperly dismissed from the Ph.D. program and lose her visa, the engineering department’s chair assigned Huang a new advisor and a new GRA appointment for the spring semester. The university also removed Rizzoni from campus pending an investigation and instructed him to complete his work from home. Eventually, Huang successfully completed her Ph.D. under her new advisor. After the university concluded that there was insufficient evidence to find that Rizzoni violated its sexual misconduct policy, Rizzoni resumed his position. II. Procedural History Huang brought this action against both Rizzoni and OSU in August 2018. Relevant to this appeal are: (1) Huang’s claims against OSU for quid pro quo sexual harassment and retaliation in violation of Title VII, 42 U.S.C. §§ 2000e-2, e-3, and (2) Huang’s 42 U.S.C. § 1983 claim against Rizzoni for unwanted sexual touching under color of state law in violation of her Fourteenth Amendment due-process right to bodily integrity. All these claims survived the defendants’ motion to dismiss. No. 23-3469 Huang v. Ohio State Univ. Page 14 On summary judgment, the district court ruled in favor of OSU on both of Huang’s Title VII claims. For the quid pro quo claim, the district court’s decision turned on its view of when Huang was an “employee” for Title VII purposes. All the parties agreed that Huang was an employee in August 2017, when OSU changed her title from Graduate Fellow to GRA. But before that date, the district court adopted the university’s contention that she was solely a student, finding that any work Huang did “was in pursuit of her own educational goals.” Op., R. 143, PageID 6671. Because she was only a student during that time, she could not have suffered an “adverse employment action” as required for Title VII. Id., PageID 6672 (emphasis added). And even though Huang was terminated from the URP after she became a GRA (and hence an “employee” at that point), the court concluded that her work with Ford was also not actionable because it was “part of her dissertation research and not part of her employment relationship” with OSU. Id., PageID 6673. For the Title VII retaliation claim, the district court concluded that Huang’s “first statutorily protected activity was reporting Rizzoni’s alleged sexual harassment” in December 2017. Id., PageID 6690. Because all the alleged adverse actions occurred beforehand, they could not have been causally linked to her protected activity. Huang’s claim that Rizzoni improperly touched her proceeded to trial. Rizzoni moved to bifurcate the trial with one phase for liability and compensatory damages, and another for punitive damages. He then moved to exclude from the first proposed phase any evidence or testimony related to Rizzoni’s alleged “retaliation” against Huang. Mot. Limine, R. 160, PageID 6895–96. While Huang did not oppose separating punitive damages from the rest of the trial, she objected to dividing the trial to the extent it would prevent her from presenting evidence of Rizzoni’s power and influence, explaining that such evidence was “inextricably intertwined” to the claims of improper touching. Resp. Mot. Bifurcate & Mot. Limine, R. 168, PageID 7034–35. Disagreeing, the district court went further than Rizzoni requested. It trifurcated the trial into separate phases for liability, compensatory damages, and punitive damages, and excluded any evidence of Rizzoni’s “alleged manipulation, coercion, and influence” from the liability phase. Order, R. 176, PageID 7085. In its view, evidence of Rizzoni’s retaliation and power, and other actions outside of the alleged touching incidents, did not prove “whether [Rizzoni] inappropriately touched [Huang]” and was therefore “irrelevant to the issue of liability.” Order, R. 188, PageID 7295–96. And it policed this line, cutting off testimony about Rizzoni’s No. 23-3469 Huang v. Ohio State Univ. Page 15 retaliation, excluding evidence of Huang’s academic credentials and performance, and quashing subpoenas for the medical providers who treated Huang in the spring of 2018. On the final day of trial, just before closing arguments, the district court questioned the efficacy of its devised trial structure and recognized that the jury had heard “a lot about the other issues” that would “relate to the damages side of the case.” Trial Tr. Vol. IV, R. 220, PageID 8333–34. It asked the parties: “should we just hear it all” or “complete the stages” as originally ordered. Id., PageID 8334. Rizzoni’s counsel expressed that it was unclear how they would “start over” at that late stage and so they should “finish this phase.” Id. Huang’s counsel concurred. The jury returned a verdict for Rizzoni after the four-day trial. Huang timely appealed the grant of summary judgment to OSU on the Title VII claims. She also challenged the judgment in Rizzoni’s favor on the Fourteenth Amendment claim, arguing that she is entitled to a new trial because of the district court’s trial structure and corresponding evidentiary decisions. ANALYSIS I. Title VII Claims Against OSU We review de novo the district court’s grant of summary judgment to OSU. See Milczak v. Gen. Motors, LLC, 102 F.4th 772, 782 (6th Cir. 2024). OSU is entitled to summary judgment if “there is no genuine dispute as to any material fact” and therefore the university “is entitled to judgment as a matter of law.” Fed. R. Civ. P. 56(a). But OSU is not entitled to summary judgment if there is evidence in the record “on which the jury could reasonably find for the plaintiff.” Anderson v. Liberty Lobby, Inc., 477 U.S. 242, 252 (1986). Our role at this stage is not to make “[c]redibility determinations” or “weigh[]” the evidence.” Id. at 255. Although the defendants dispute many of the facts, at this stage we view the record evidence in the light most favorable to Huang and draw all reasonable inferences in her favor. See Milczak, 102 F.4th at 782. Our first task for the Title VII analysis is to determine whether Huang was an “employee” for Title VII purposes during the period of the alleged harassment. We then review Huang’s two Title VII claims separately: first, that Rizzoni committed quid pro quo sexual No. 23-3469 Huang v. Ohio State Univ. Page 16 harassment (for which OSU would be vicariously liable); second, that Rizzoni retaliated against her for opposing his harassment. A. “Employee” Status Title VII “protect[s] employees.” Shah v. Deaconess Hosp., 355 F.3d 496, 499 (6th Cir. 2004); see 42 U.S.C. §§ 2000e-2, 2000e-3. Huang contends that she counts as an OSU employee and is thus protected by Title VII. But the district court concluded that, as a matter of law, she was not an employee of OSU between August 2014 and August 2017—the period she was deemed a Graduate Fellow by the university. It held that she only became an employee once she became a GRA. The court was mistaken to make that determination at the summary judgment stage.2 Title VII defines “employee” as “an individual employed by an employer.” 42 U.S.C. § 2000e(f). Because that definition “is completely circular and explains nothing,” we apply the common law meaning of the term and have adopted an agency test to help us answer the legal question of whether an individual is an “employee” for Title VII purposes. Bryson v. Middlefield Volunteer Fire Dep’t, Inc., 656 F.3d 348, 352 (6th Cir. 2011) (quoting Nationwide Mut. Ins. Co. v. Darden, 503 U.S. 318, 323 (1992)). The “crux” of the test requires the court to consider the 2 OSU contends that Huang forfeited the argument that she was an employee during the time OSU classified her as a Graduate Fellow—August 2014 to August 2017. But its decision to point the finger at Huang is curious given that, in its motion for summary judgment, OSU mentioned Huang’s “student” status in at most a few sentences that asserted, rather than argued, Huang was not an “employee” while a Graduate Fellow. After all, OSU bore the responsibility of demonstrating an absence of a genuine dispute of material fact regarding Huang’s employment status. Carver v. Bunch, 946 F.2d 451, 454 (6th Cir. 1991). But rather than argue about the realities of Huang’s experience, OSU’s brief argument focused only on titles. Because a job title is not determinative of employee status, it does not constitute a material fact in this analysis. See Bryson v. Middlefield Volunteer Fire Dep’t, Inc., 656 F.3d 348, 355 (6th Cir. 2011) (explaining that we are to “consider and weigh all incidents of the relationship no matter how the parties characterize the relationship” (emphasis added)). And, crucially, OSU conceded that Huang became an employee when her title switched to “Graduate Research Associate,” yet it did not point to a single substantive difference from her duties as a “Fellow.” We recognized that Huang’s brief below incorrectly asserted she was a Graduate Research Associate when she was a Fellow, but because titles are immaterial to our analysis, that mistake does not constitute forfeiture. Moreover, her opposition brief to OSU’s motion for summary judgment clearly stated she was “employed” by Rizzoni, who was her “supervisor” at the school. Opp’n Summ. J., R. 114, PageID 6259, 6263. She also identified supporting record evidence, such as her duties to OSU on the Ford URP, the benefits she received for that work, and the control Rizzoni imposed on her time. Huang has therefore not forfeited this issue. Cash-Darling v. Recycling Equip., Inc., 62 F.4th 969, 975 (6th Cir. 2023). Additionally, the district court addressed the issue in granting OSU summary judgment. And “there can be no forfeiture ‘where the district court . . . addressed the merits of the issue.’” United States v. Clariot, 655 F.3d 550, 556 (6th Cir. 2011) (citation omitted). No. 23-3469 Huang v. Ohio State Univ. Page 17 scope of OSU and Rizzoni’s “right to control the manner and means by which” Huang accomplished the work of her Ph.D. program. Weary v. Cochran, 377 F.3d 522, 525 (6th Cir. 2004) (quoting Darden, 503 U.S. at 323). We do that by examining “all incidents of the relationship.” Bryson, 656 F.3d at 355. The factors of the agency test most relevant here look to the nature of Huang’s work, how much control OSU exercised over it, and the benefits she received. See Darden, 503 U.S. at 323 (listing factors). We ignore “how the parties characterize” the relationship. Bryson, 656 F.3d at 355. That OSU classified Huang as a graduate student, therefore, does not disqualify her from “employee” status under Title VII. It is plain that one’s status as a student is not preclusive of also being an employee. Consider a student who enrolls in a full load of courses and then works at the library restacking books to earn an hourly wage. This student might “control the manner and means by which” she completes her coursework, Darden, 503 U.S. at 323, but no one would dispute that her library job fell under Title VII. But a student’s academic and employment work can also overlap, and that does not remove them from the employment protections of Title VII either. Consider a medical resident who is enrolled in an accredited academic training program and also teaches through supervised work in clinical settings. The resident learns and completes a program necessary for board certification, but in doing so works by seeing patients and treating ailments. Recognizing this, the Third Circuit held that a resident can be a hospital’s “‘employee’ notwithstanding any other status the law may or may not have reposed on her (for example, a ‘student’).” Mercy Catholic, 850 F.3d at 559. The resident could thus be both a student and an employee with respect to the same conduct (i.e., seeing patients). Id. The district court, by contrast, did not recognize that Huang’s research and other work at OSU could be simultaneously an academic and employment activity. Instead, it attempted to sort her responsibilities into one category or the other, akin to the student who works a side job at the library. See Op., R. 143, PageID 6669. And though it acknowledged that that the commonlaw agency doctrine applied, it only looked to her Graduate Fellow offer letter—not the actual “incidents of the relationship” between Huang and OSU—to determine whether she was a student or an employee. Bryson, 656 F.3d at 355. With this constrained view, it determined that No. 23-3469 Huang v. Ohio State Univ. Page 18 the letter “outline[d] a purely academic relationship,” that Huang’s stipend was akin to “an academic scholarship,” and that she “was not obligated to perform any work or service for OSU.” Op., R. 143, PageID 6670–71. The court thus concluded that Huang was not an employee of OSU for Title VII purposes when a Graduate Fellow, but, curiously, that she was when OSU classified her as a GRA. Under the agency test, these conclusions are incompatible, since OSU’s control over Huang’s work (via Rizzoni) remained the same even as her title changed. Instead of the district court’s approach, we apply the common law agency test by holistically evaluating Huang’s relationship with OSU. Rizzoni served as Huang’s academic advisor and the supervisor of her research on the URP. To determine whether this relationship made Huang an “employee under the common law of agency,” the Supreme Court has instructed that our task is to evaluate OSU’s “right to control the manner and means by which the product is accomplished.” Darden, 503 U.S. at 323 (citation omitted). To do so, the Supreme Court articulated thirteen factors: [1] the hiring party’s right to control the manner and means by which the product is accomplished. Among the other factors relevant to this inquiry are [2] the skill required; [3] the source of the instrumentalities and tools; [4] the location of the work; [5] the duration of the relationship between the parties; [6] whether the hiring party has the right to assign additional projects to the hired party; [7] the extent of the hired party’s discretion over when and how long to work; [8] the method of payment; [9] the hired party’s role in hiring and paying assistants; [10] whether the work is part of the regular business of the hiring party; [11] whether the hiring party is in business; [12] the provision of employee benefits; [13] and the tax treatment of the hired party. Id. at 323–24 (citation omitted). Although the “original purpose of the factors” was to distinguish between employees and independent contractors, Marie v. Am. Red Cross, 771 F.3d 344, 352 (6th Cir. 2014), we have nonetheless applied this test “to determine whether an employment relationship exists under different statutes.” Bryson, 56 F.3d at 352. Yet we have also recognized that these factors do not fit neatly in every context. See Marie, 771 F.3d at 352 (noting many of the factors do not apply to the volunteer context). Moreover, we have observed that some factors may be more or less relevant depending on the context, and that no factor is dispositive. See id. at 352–54; Bryson, 56 F.3d at 352–54. No. 23-3469 Huang v. Ohio State Univ. Page 19 Here, we examine the nature of Huang’s work, the degree of control that OSU exerted in those endeavors, and Huang’s compensation. See Darden, 503 U.S. at 323. Construing the record evidence in the light most favorable to Huang, as we must, she has created a material question of fact that she was an “employee” entitled to Title VII’s protections. First, we ask what type of work Huang engaged in as she progressed through her Ph.D. at OSU during the time she was a Graduate Fellow, August 2014 to May 2017. This factor is important because, as we’ve explained, if a plaintiff’s “work is part of the regular business of the hiring party,” that suggests she is a common-law employee. Marie, 771 F.3d at 352, 359 (quoting Darden, 503 U.S. at 323). Huang’s main engagement at OSU was her work at CAR researching battery aging technology for Ford as part of the URP. As Rizzoni testified, he recruited Huang specifically to continue ongoing work on the URP, noting that her experience qualified her for that work, akin to how one would hire an applicant for a job. Rizzoni testified that recruiting students like Huang to continue preexisting research workstreams (like the URP) was “routine” and “how we do things.” Rizzoni Dep., R. 98-1, PageID 3620. This was part of CAR and OSU’s “regular business.” Darden, 503 U.S. at 324. And OSU received significant economic benefits from Huang’s work—through CAR, OSU received grants and funding from industry leaders like Ford and from the government, building CAR’s reputation as a world-class research center. That said, Huang took classes and had to pass a candidacy exam. And Rizzoni served as Huang’s advisor not just to supervise her research with Ford, but to guide her to obtain her Ph.D. degree. These aspects of Huang’s relationship with Rizzoni (and OSU) indicate an academic relationship. A reasonable review of the record, however, shows that the bulk of Huang’s relationship with Rizzoni was focused on URP research and meetings—part of OSU’s regular business—suggesting that she is properly considered an employee. See Marie, 771 F.3d at 359. Second, we consider the extent to which OSU controlled Huang’s research work. While the extent of control is not “dispositive,” we have emphasized that a defendant’s “ability to control job performance and employment” can be an important factor indicating an agency relationship. Marie, 771 F.3d at 356 (citing Janette v. Am. Fid. Grp., Ltd., 298 F. App’x 467, 472 (6th Cir. 2008)) (cleaned up). Huang has supplied evidence that, if accepted by a jury, No. 23-3469 Huang v. Ohio State Univ. Page 20 demonstrates that Rizzoni controlled her work along various dimensions. Since Rizzoni disputes Huang’s testimony regarding his control of her work, we cannot resolve her employee status as a matter of law at this stage. See Simpson v. Ernst & Young, 100 F.3d 436, 439 (6th Cir. 1996). Recall that shortly after Huang matriculated to OSU, Rizzoni assigned Huang to work on the URP and for it to be her dissertation topic—as she and Rizzoni describe, this wasn’t her choice, it was driven by CAR’s needs. Then, Rizzoni largely controlled when and where she worked. Cf. id. at 356 (citation omitted) (emphasizing the importance of an “employer’s ability to control job performance”). While Huang did not have a traditional nine-to-five schedule, she did have regular bi-weekly meetings with Ford, and frequent one-on-one meetings with Rizzoni. He dictated those meetings, including making her meet with him on weekends and chastising her if she did not meet him as often as he wanted. Based on Huang’s testimony, a jury could also find that Rizzoni controlled her vacation schedule, even when her vacation coincided with university holidays. He admonished Huang when she left the country to visit her family during OSU’s winter break without getting his permission (apparently forgetting that she had). According to Huang, he also dictated where they met—mainly at CAR, but also at his home, and sometimes they traveled to Dearborn for meetings. The degree of control he wielded over the location of her work is reflected in his infuriation for her intransigence when she did not participate in the meetings at CAR with him but did so virtually. He also offered (repeatedly) to drive Huang to the CAR facility to ensure that she did her work there. Because Rizzoni dictated Huang’s research topic based on the university’s needs and set the times and location of her work, a jury could conclude that Rizzoni exercised the type of control over Huang that made her an OSU employee. Third, we consider Huang’s compensation. The parties do not dispute that on top of a full-tuition scholarship, OSU paid Huang a stipend—presumably for her living expenses—and a discretionary 10% bonus in recognition of her academic and professional background. It’s true that Huang also received educational benefits, fulfilled requirements for her academic program, and eventually graduated with her Ph.D. Often, students receive stipends without having to perform any work in exchange. But other times, as with medical residents, students’ academic pursuits overlap with work that mimics employment. Here, Huang’s funding was tied to her No. 23-3469 Huang v. Ohio State Univ. Page 21 research at CAR, not just her enrollment in the Ph.D. program at OSU, suggesting an employment relationship. OSU argues that Huang’s compensation was not employment related. It notes that Huang didn’t have an employment contract with the university, and that neither tax contributions nor benefit deductions were withheld from her paycheck. See Darden, 503 U.S. at 323–24 (listing “the method of payment[,] . . . the provision of employee benefits[,] and the tax treatment of the hired party” as relevant factors). OSU has not identified evidence to support these assertions, so we do not consider them undisputed in our analysis. See Fed. R. Civ. P. 56(c)(1). If at trial OSU musters support for these assertions, the district court may consider them as a few ingredients among many in the common-law agency test, in which “no one factor [is] decisive.” Bryson, 656 F.3d at 354 (quoting Darden, 503 U.S. at 324). More importantly, even if OSU did not have a written employment contract with Huang, by dictating the hours and expectations of a job, Rizzoni used his control over Huang’s tuition, stipend, bonus, visa, and career success to de facto set the terms and conditions of her work. When she didn’t abide those terms, he would threaten to “terminate [their] relationship,” and remind her not to “forget that [he was] the one who pays the bills.” Sept. 21, 2016 Email, R. 944, PageID 2375; Dec 3, 2016 Email, R. 94-4, PageID 2376. We have observed that the “ability to terminate a non-compliant employee” and thus cut off “the source of income upon which [an employee] depends” is an employer’s “greatest source of control.” Marie, 771 F.3d at 357–58. Rizzoni—and by extension OSU—had that power over Huang. Considering the totality of these circumstances, Huang has amply met her burden to create a genuine dispute of material fact as to whether she was an OSU “employee” under Title VII when she suffered Rizzoni’s purported harassment, from August 2014 to December 2017. At trial, the jury should make factual findings regarding Rizzoni’s disputed control over where and when Huang worked. Once any genuine dispute is settled by the jury, the district court should decide as a matter of law whether Huang was an OSU employee, consistent with this opinion’s reasoning. No. 23-3469 B. Huang v. Ohio State Univ. Page 22 Quid Pro Quo Sexual Harassment Claim Having determined that Huang could be considered an employee under Title VII, we can now evaluate whether her sexual harassment claim should survive summary judgment. Viewing the evidence in the light most favorable to Huang, she has established disputes of material fact that preclude summary judgment on this claim. Huang asserts that she suffered quid pro quo harassment, which occurs when “an employer demand[s] sexual favors from an employee in return for a job benefit.” Burlington Indus., Inc. v. Ellerth, 524 U.S. 742, 752 (1998). To succeed on this claim, Huang must show that: (1) she was a member of a protected class; (2) Rizzoni subjected her to unwanted sexual advances; (3) these unwanted advances were because of her sex; (4) “submission to the unwelcomed advances was an express or implied condition for receiving job benefits,” or that Huang’s refusal to submit to those advances caused Rizzoni to take adverse employment action against her; and (5) OSU is vicariously liable for Rizzoni’s actions. Bowman v. Shawnee State Univ., 220 F.3d 456, 461, 461 n.5 (6th Cir. 2000) (citing Kauffman v. Allied Signal, Inc., 970 F.2d 178, 186 (6th Cir. 1992)). OSU does not dispute that Huang’s testimony suffices to establish the first three factors at summary judgment, nor that it can be held vicariously liable for Rizzoni’s sexual harassment.3 We focus our analysis on the fourth prong. Huang argues that submitting to Rizzoni’s unwanted advancements was an implied condition of staying in the Ph.D. program and researching at CAR, and that when she ultimately refused to submit, he took adverse employment action against her by revoking her 10% bonus, ending her role with the Ford URP, and terminating her GRA position. OSU largely argues that, as the district court held, Huang could not experience any adverse employment actions because she was not an “employee.” We have already determined that Huang raised a material dispute as 3 The university does not dispute that Rizzoni was Huang’s supervisor at OSU. Nor does it dispute that if Huang proves that Rizzoni took an adverse action against her because she refused to submit to his sexual advances, or imposed submission to his sexual advances as an “implied condition,” OSU is vicariously liable for his sexual harassment. See Ellerth, 524 U.S. at 762 (“[A] tangible employment action taken by the supervisor becomes for Title VII purposes the act of the employer.”). OSU raised an affirmative defense to liability before the district court based on the Faragher/Ellerth doctrine. But it has not asserted that defense on appeal, and it does not apply when an employer has taken an adverse action. See id. No. 23-3469 Huang v. Ohio State Univ. Page 23 to whether she was an employee of OSU the entire time Rizzoni was her Ph.D. advisor, so these adverse actions cannot be dismissed as being purely academic related. Construing the record in Huang’s favor, she shows both that submitting to the unwanted advances was an “implied condition” of her job, and that when she refused, she suffered adverse employment action. As detailed above, Huang has presented evidence that when she withdrew from Rizzoni, resisted his sexual advances, or avoided him, he would threaten to terminate her research and eliminate or reduce her pay. He even chastised her in front of potential employers when she refused to take meetings with him alone, in person, on a Sunday (and instead did them remotely). A jury could reasonably conclude that keeping him “happy” was an implied condition of Huang’s job. Moreover, it is undisputed that Rizzoni cut off her 10% discretionary stipend bonus in August 2017, terminated her position on the Ford URP in December 2017 following the candidacy exam, and ended her GRA. A decrease in pay is a quintessential adverse employment action. Dye v. Racing Comm’n, 702 F.3d 286, 303 (6th Cir. 2012); see also Howington v. Quality Rest. Concepts, LLC, 298 F. App’x 436, 442 (6th Cir. 2008) (loss of tip income). So is a termination. See Cicero v. Borg-Warner Auto., Inc., 280 F.3d 579, 584 (6th Cir. 2002). Thus, a jury could find that Huang suffered an adverse employment action. Did Rizzoni take these adverse actions because Huang did not submit to his advances? OSU does not dispute that Huang has raised a material question that he did. Huang’s testimony and the emails from Rizzoni (e.g., where he threatened to cut off her funding if she did not make him “happy”) suffice to create a fact question for the jury on causation. Cf. Moran v. Al Basit LLC, 788 F.3d 201, 205 (6th Cir. 2015) (explaining that the plaintiff’s coherent and detailed testimony defeated a motion for summary judgment). Accordingly, OSU is not entitled to summary judgment on Huang’s quid pro quo sexual harassment claim. C. Retaliation Claim Huang’s second Title VII claim against OSU is for retaliation. We evaluate Title VII retaliation claims with the well-established McDonnell Douglas burden shifting test. See Redlin v. Grosse Pointe Pub. Sch. Sys., 921 F.3d 599, 613 (6th Cir. 2019). Under that framework, Huang must first make a prima facie case by showing (1) she engaged in a Title VII protected No. 23-3469 Huang v. Ohio State Univ. Page 24 activity, (2) her employer knew about it, and (3) her employer took an adverse employment action against her because of her protected activity. Wyatt v. Nissan N. Am., Inc., 999 F.3d 400, 419 (6th Cir. 2021). If she does, OSU would need to assert a nondiscriminatory reason for its actions, and Huang would have to show pretext. Redlin, 921 F.3d at 613–14. But because the parties do not discuss the “nondiscriminatory reason” and “pretext” steps of the McDonnell Douglas burden shifting framework, we limit our analysis to the prima facie stage. For that, we need to determine only whether Huang engaged in “protected activity” when she resisted Rizzoni’s sexual advances. If she did, then she has made a prima facie showing—we have already rejected OSU’s arguments that Huang did not suffer an adverse employment action, and OSU does not dispute that there is a fact question as to whether Rizzoni took the adverse actions against Huang because she resisted his advances. We ask, then, what counts as “protected activity” that can give rise to a retaliation claim? Title VII prohibits an employer from discriminating against an employee who “has opposed” discriminatory conduct prohibited by Title VII, like sexual harassment. 42 U.S.C. § 2000e-3(a). To “oppose” means “to resist or antagonize; to contend against; to confront; resist; withstand,” and “to be hostile or adverse to, as in opinion.” Crawford v. Metro. Gov’t of Nashville & Davidson Cnty., 555 U.S. 271, 276 (2009) (cleaned up). The district court held that Huang did not engage in “protected activity” until December 2017, when she filed an official complaint with the university. And because all the adverse actions occurred before that point, it could not have been causally linked to the complaint. But to count as “protected activity,” opposition does not need to be formalized, like filing an official grievance. Contrary to the district court’s opinion and OSU’s arguments, we do not look with such a narrow lens. Instead, to constitute “protected activity,” we have held it is enough that a plaintiff “resists” a supervisor’s sexual advances. E.E.O.C. v. New Breed Logistics, 783 F.3d 1057, 1067–68 (6th Cir. 2015). Huang testified she resisted Rizzoni’s sexual advances by pushing his hands off her or moving away from him whenever he touched her inappropriately, and by trying to “keep a distance from him” to minimize his opportunities to harass her. Huang Dep., R. 102-1, PageID 4691–93. These actions exemplify the type of activities that Title VII protects. Though Huang did not expressly tell Rizzoni “no,” or to “leave me alone,” a jury could find that No. 23-3469 Huang v. Ohio State Univ. Page 25 it would’ve been obvious to him that his overtures were unwanted. Huang testified that she did not file formal complaints against Rizzoni because he held power over her pay, visa, research, and degree. Given this testimony, a jury could find she opposed Rizzoni’s sexual advances just by pushing him away. OSU resists this conclusion, citing an employment law treatise to support its argument that “simply ‘rejecting sexual advances does not constitute a protected activity.’” Appellee Br. at 45 (quoting Retaliation, Labor and Employment Law: Compliance and Litigation § 4:34 (3d ed.)). That may be the case in the Fifth Circuit, to which the treatise cites for that proposition. See Labor and Employment Law § 4:34 (citing Lemaire v. Louisiana Dep’t of Transp. & Dev., 480 F.3d 383 (5th Cir. 2007)). But that is not the law here. See New Breed Logistics, 783 F.3d at 1067–68. Nor is it the law in the Eighth Circuit. See Ogden v. Wax Works, Inc., 214 F.3d 999, 1007 (8th Cir. 2000). Other courts to address the issue have also concluded that resisting a supervisor’s sexual advances is protected activity under Title VII or its analogues. See Vill. of Tequesta v. Luscavich, 240 So. 3d 733, 742–43 (Fla. Dist. Ct. App. 2018) (rejecting LeMaire in a Title VII case before a state court); see also Mihalik v. Credit Agricole Cheuvreux N. Am., Inc., 715 F.3d 102, 115–16 (2d Cir. 2013) (concluding that resisting sexual advances is protected activity under the New York City Title VII analogue); but see id. at 115 n.12 (expressing no opinion on whether Title VII or a state-law analogue protects such activity). We have recognized a circuit split on this issue, and we take the majority view. See New Breed Logistics, 783 F.3d at 1068 (noting that “only one of our sister circuits” has concluded that resisting sexual advances is not protected activity under Title VII). Rejecting sexual advances—using words or actions—is within the scope of conduct Title VII protects. Because Huang has established a prima facie case of Title VII retaliation, we remand for the district court to consider the next steps of the McDonnell Douglas framework. II. Fourteenth Amendment Claim Huang’s § 1983 claim against Rizzoni proceeded to trial, and the jury found for Rizzoni. On appeal, Huang challenges the district court’s decisions to try the issue of liability separately from the issues of compensatory and punitive damages, and then to exclude from the liability phase evidence of Rizzoni’s “alleged manipulation, coercion, and influence.” Order, R. 176, No. 23-3469 Huang v. Ohio State Univ. Page 26 PageID 7085. We consider Huang’s challenge to the trifurcation order and the district court’s evidentiary decisions in tandem, as the issues are intertwined. As we read it, Huang does not dispute the district court’s decision to trifurcate the trial per se, she just disputes it to the extent that the district court excluded relevant evidence from the liability phase. The district court established those evidentiary constraints in the trifurcation order (and in its pretrial order, order quashing subpoenas, and order denying rehearing), and then consistently enforced its prior rulings by refusing to allow Huang to introduce evidence that would violate them. Regardless of how it’s framed, our review regarding the trifurcation of the trial and the exclusion of evidence are both for an abuse of discretion. Martin v. Heidman, 106 F.3d 1308, 1311 (6th Cir. 1997); In re Air Crash Disaster, 86 F.3d 498, 526 (6th Cir. 1996). The district court abused its discretion, and because its narrow view of relevancy permeated the trial, Huang is entitled to a new trial. Before analyzing this issue, we pause to consider and reject Rizzoni’s argument that Huang waived her challenge to the trial structure. After the close of evidence, while it was discussing the jury instructions, the district court noted that “we’ve heard a lot of evidence that would relate to the damage side of the case. So should we just hear it all or does it make sense to complete the stages as I originally intended?” Trial Tr. Vol. IV, R. 220, PageID 8334. Huang’s counsel agreed with Rizzoni’s that it would not make sense to “start over” so late in the trial, just before closing arguments. Id. Contrary to Rizzoni’s argument, this response did not undo Huang’s numerous pretrial objections to the district court’s trial structure and attendant evidentiary rulings. A party need only timely object to an evidentiary decision on the record once. Fed. R. Evid. 103(a). “Once the court rules definitively on the record—either before or at trial—a party need not renew an objection or offer of proof to preserve a claim of error for appeal.” Fed. R. Evid. 103(b). Given all her objections, Huang did not “fail to make [her] opposition known to the district court in a timely fashion.” 9A Fed. Prac. & Proc. Civ. § 2392 (3d ed.). The district court, after the close of evidence, appeared to question its own contorted trial structure, stating that it “must confess” that it was “[its] idea to structure this trial” as such, and it was unsure if it had been “fair and efficient or not.” Trial Tr. Vol. IV, R. 220, PageID 8333–34. In this context, it was unclear exactly what the district court meant when it said “should we just hear it all[?]” or how it imagined it would have proceeded. Id. At that point, No. 23-3469 Huang v. Ohio State Univ. Page 27 how would the parties undo the harm? Agreeing that it was too late at that point to “start over” does not equate to waiver. And it would not serve the purposes of our waiver doctrine to require Huang to accede to such an ambiguous question after the evidence had already closed. See Scottsdale Ins. Co. v. Flowers, 513 F.3d 546, 552 (6th Cir. 2008) A. The Trial Structure and Evidentiary Rulings While a district court has latitude over the trial structure, that latitude comes with guiderails. A district court may trifurcate a trial “[f]or convenience, to avoid prejudice, or to expedite and economize” the proceedings. Fed. R. Civ. P. 42(b). In dividing a trial into separate phases, a district court needs to consider “the potential prejudice to the parties, the possible confusion of the jurors, and the resulting convenience and economy.” Martin, 106 F.3d at 1311 (citing In re Beverly Hills Fire Litig., 695 F.2d 207, 216 (6th Cir. 1982)). We have recognized that bifurcation can be appropriate, for example, “when the evidence pertinent to the two issues is wholly unrelated and the evidence relevant to the damages issue could have a prejudicial impact upon the jury’s liability determination.” Helminmski v. Ayerst Labs., 766 F.2d 208, 212 (6th Cir. 1985) (cleaned up). On the other hand, we have warned of the “danger that bifurcation may deprive plaintiffs of their legitimate right to place before the jury the circumstances and atmosphere of the entire cause of action which they have brought,” artificially separating the wrongdoing “from the reality of injury.” Beverly Hills, 695 F.2d at 217. In other words, Rule 42(b) recognizes that sometimes separating a trial into phases can be helpful, but if it means the story has “missing chapters,” it sometimes crosses a line. Old Chief v. United States, 519 U.S. 172, 189 (1997). When bifurcation leads to the exclusion of categories of relevant evidence, it certainly crosses this line. Relevant evidence is, of course, not just facts that “carry a party’s evidentiary burden” by proving an element of a claim. Dortch v. Fowler, 588 F.3d 396, 401 (6th Cir. 2009). The standard for relevance is capacious—evidence is relevant if it has “any tendency” to make a “fact of consequence” to the case “more probable or less probable than it would be without the evidence.” Id. at 400 (quoting Fed. R. Evid. 401). All it needs to do is “advance the ball” by showing that it is more (or less) likely that the plaintiff’s theory of the defendant’s liability is true. Id. No. 23-3469 Huang v. Ohio State Univ. Page 28 Here, the district court carved up the trial in a manner that excluded categories of relevant evidence as to whether Rizzoni improperly touched Huang. In its view, Rizzoni’s “alleged manipulation, coercion, and influence with respect to [Huang’s] Ph.D. program and candidacy exam” was “irrelevant” to the liability phase of trial. Order, R. 188, PageID 7295–96. That’s because, it explained, “adverse actions, retaliation, coercion, power dynamics and pretext are not built into the elements of the claims and defenses.” Id., PageID 7298–99. Though the court recognized that “a jury could infer that [Rizzoni’s] displeasure stemmed from [Huang’s] refusal to submit to his sexual advances,” it explained that this inference would only be permissible if Huang “first prove[d] that the sexual advances took place in the form of unwanted touching.” Id. The district court thus confined much of Huang’s evidence to her accounts of Rizzoni’s “unwanted sexual touching.” Id. These decisions, as Huang argues, rested on “a faulty conception of relevance.” Appellant Br. at 29. Rizzoni’s alleged “manipulation, coercion, and influence” over Huang constitute circumstantial evidence that he touched her sexually. They provide an explanation of her behavior, and his, regarding the touching. And even though retaliatory actions may not, on their own, prove that the triggering incident occurred, they can still be probative. Indeed, all sorts of effects can be circumstantial evidence of a possible cause. See 2 Wigmore on Evidence § 436 (4th ed. 1985) (“[I]n general, the inference from an effect to the existence or operation of a cause is usually so proper as to be unquestionable[.]”). And we frequently permit plaintiffs to prove their case using backward-looking inferences. Cf. Mickey v. Zeidler Tool & Die Co., 516 F.3d 516, 525 (6th Cir. 2008) (evidence of temporally proximate adverse action is sufficient to create an inference of causation). We focus on two specific ways that the district court’s trifurcation and evidentiary orders hamstrung Huang’s ability to present her theory of liability to the jury: (1) it limited circumstantial evidence of Rizzoni’s harassment; and (2) it limited Huang’s ability to counter Rizzoni’s evidence suggesting she fabricated the assault allegations in retaliation for failing the candidacy exam.4 4 Huang argues that the district court made categorical errors in its trifurcation orders and then erred in excluding specific testimony at trial. Because the challenge to the specific evidentiary rulings at trial stems from the same issue as the categorical errors, we address them together. Similarly, we do not address every single evidentiary No. 23-3469 Huang v. Ohio State Univ. Page 29 Circumstantial Evidence of Harassment. Huang’s theory of the case was that Rizzoni’s sexual harassment followed a pattern, praising her when she submitted to his demands and advances and punishing her when she didn’t. For that, she needed to show the jury that Rizzoni vacillated between praise and harsh criticism corresponding with her testimony of his improper touching. That includes evidence aside from the touching itself, such as evidence of both compliments and retaliation. But the district court excluded much of this circumstantial evidence. For example, the court prevented Huang from admitting most of Rizzoni’s emails where he threatened her or complained about her to others. These often came on the heels of her refusal to submit to his alleged touching. In one email, Rizzoni complained to a colleague that Huang did not think he was “next to God, and act accordingly.” Sept. 21, 2016 Email, R. 11410, PageID 6359. When Huang crossed Rizzoni on the “next to God” statement, the district court struck the question from the record and told Huang’s counsel he’d cause a mistrial if he asked about it again. Trial Tr. Vol. II, R. 218, PageID 8104–05. In other emails never seen by the jury, Rizzoni told her that she did “not belong in” the Ph.D. program if she did not take his “advice.” Jan. 31, 2016 Email, R. 94-4, PageID 2374. Another key component of Huang’s theory was that Rizzoni wielded power over her— that he could use his position as her Ph.D. advisor and director of CAR to get her to meet with him alone so he could sexually abuse her. She sought to prove that his power over her research, pay, visa, academic success, and future career made her susceptible to his abuse and hesitant to report him. For instance, she sought to establish that Rizzoni made it impossible for her to avoid him (and his alleged sexual abuse) by demanding she meet with him in person—alone and on the weekends—and threatening her when she did not. When Huang later began consistently avoiding Rizzoni, he grew so frustrated that he wielded his power to manipulate her candidacy exam, revoke her funding, and expel her from OSU. Huang explained this to the district court in opposing trifurcation, arguing that: “[e]vidence of Rizzoni’s power and influence at Ohio State University is inextricably intertwined to the facts at issue in this case, including as to why Rizzoni behaved the way that Plaintiff alleges he did toward her, [and] why he believed he could get away with itb.” Resp. Mot. Bifurcate, R. 168, PageID 7034. But the district court said it was ruling, but instead address them collectively as they too stem from the district court’s narrow construction of relevancy. No. 23-3469 Huang v. Ohio State Univ. Page 30 “irrelevant.” Order, R. 188, PageID 7295. It wasn’t. We are “firmly convinced” that it was a “mistake” for the district court to categorically exclude circumstantial evidence which lent credence towards a critical element of Huang’s story: the way Rizzoni committed his alleged sexual abuse. Kovacevich v. Kent State Univ., 224 F.3d 806, 832 (6th Cir. 2000).5 Relatedly, because the district court did not view evidence of the effect of improper touching as probative of whether it happened, it quashed subpoenas for two of Huang’s medical providers at OSU. Because Huang sought treatment from these providers after Rizzoni terminated her from the URP, the district court reasoned that the providers could not have “personal knowledge” of any relevant facts. Order, R. 183, PageID 7257 (granting Mot. Quash); see also Order, R. 188, PageID 7300–01 & n.1 (explaining, in order denying reconsideration, that testimony would be admissible under hearsay exception for statements made for purposes of medical treatment only if Huang first proved Rizzoni inappropriately touched her). This too was an abuse of discretion, as the medical providers could have corroborated Huang’s testimony about Rizzoni’s unwanted touching. Hill v. Homeward Residential, Inc., 799 F.3d 544, 552 (6th Cir. 2015) (review of a decision to quash a subpoena on relevancy grounds is for abuse of discretion); Field v. Trigg Cnty. Hosp., Inc., 386 F.3d 729, 735 (6th Cir. 2004) (“[S]tatements made by an individual to physicians for purposes of diagnosis or treatment are considered exceptionally trustworthy because the declarant has a strong motive to tell the truth in order to receive proper care.”). Evidence Regarding Huang’s Academic Abilities and Candidacy Exam. The district court also ruled that evidence pertaining to Huang’s academic performance and capabilities was irrelevant. That was particularly harmful to Huang because it undermined her ability to counter Rizzoni’s defense at trial—that she was a poor student who was stubborn and wouldn’t take his 5 Evidence of Rizzoni’s power and influence was also relevant to whether Rizzoni acted under color of state law, which was an issue the jury had to decide during the liability phase. The court instructed the jury it could find Rizzoni’s actions were under color of state law if he did them “in the course of performing an actual or apparent duty of his office, or if the conduct is such that the actor could not have behaved as he did without the authority of his office.” Jury Instructions, R. 203, PageID 7636. Evidence that Rizzoni wielded his control of Huang’s Ph.D. funding as a threat against Huang so she would do what he wanted was therefore relevant to whether his alleged abuse was under color of state law. No. 23-3469 Huang v. Ohio State Univ. Page 31 advice, failed her candidacy exam, and then retaliated against him by fabricating sexual harassment allegations. Even assuming that evidence of Huang’s credentials and academic performance was not otherwise relevant, Rizzoni opened the door by making it central to his defense. See Francis v. Clark Equip. Co., 993 F.2d 545, 550 (6th Cir. 1993) (“[W]hen a party opens up a subject, there can be no objection if the opposing party introduces evidence on the same subject.”). Beginning with his opening statement, Rizzoni explained to the jury that Huang was lying to get “revenge” against Rizzoni after she “was deemed by [the Ph.D. candidacy] committee to have failed her exam.” Trial Tr. Vol. I, R. 217, PageID 7799–7800. Rizzoni developed this theory throughout the trial; he testified about the rigorous requirements of a Ph.D. and called two of his former students to testify that Rizzoni had high standards, that Huang “bull[ied]” Rizzoni, and that she didn’t understand the advisor-advisee relationship. Trial Tr. Vol. III, R. 219, PageID 8261–63. The jury learned that she failed the candidacy exam and was kicked out of the program. In his closing statement, Rizzoni’s counsel reemphasized that Huang’s case “adds up to revenge.” Trial Tr. Vol IV, R. 220, PageID 8372. Rizzoni made Huang’s academic performance central to the liability phase by staking his case on the theory that Huang was a bad Ph.D. student with a motive to lie about him. Yet the jury heard little of Huang’s evidence to the contrary. The district court prevented Huang from entering evidence of her academic capabilities in multiple respects. In general, the court ruled that it would not let the parties get “sidetracked in disputes about [Huang’s] academic studies.” Trial Tr. Vol. II, R. 218, PageID 7928. Rizzoni recruited Huang to CAR because of her impressive career in China, but the court excluded that testimony, deeming it her “childhood.” Tr. Vol. I, R. 217, PageID 7811–12. It did not allow Dr. Canova—Rizzoni’s friend and expected successor—to testify that Huang got As in both of his classes, which covered similar topics to her dissertation research. The court allowed Huang to play a portion of Anderson’s deposition testimony, but not a portion where he explained that his colleagues were so impressed with her final presentation, they wanted her to start working at Ford right away. And this was just a small portion of Anderson’s extensive and heartfelt praise of Huang, including his direct refutation of Rizzoni’s criticisms of her. Likewise, the district court prevented the jury from hearing from the then-chair of the engineering department, who No. 23-3469 Huang v. Ohio State Univ. Page 32 explained that he organized the effort to get Huang a new advisor so she could complete her studies and avoid deportation, that Huang’s new advisor had high standards, and that she “earned” the Ph.D. she eventually received. Subramaniam Dep., R. 95-1, PageID 2547–49. Huang also had significant testimony in the summary judgment record that Rizzoni’s handling of her candidacy exam was unusual and particularly harsh. Again, the district court kept it from the jury. In their deposition testimony, multiple OSU officials and professors testified about the normal custom for candidacy exams, and how Rizzoni did not follow it with Huang. Before the exam, when a Ph.D. student doesn’t seem ready, the standard course is to allow them more time to prepare. Rizzoni shared his “frustration” with Huang’s (reconstituted) committee that she was not ready, but instead of giving her more time, Rizzoni instructed the committee members to “be tough” on Huang and “make it a real exam.” Dec. 7, 2017 Emails, R. 99-2, PageID 4139, 4141–42, 4143. Similarly, OSU officials and professors explained in their depositions that for the engineering department, the custom is to allow a student who fails their candidacy exam to retake it. Even while acknowledging that Huang’s substantive exam performance was unimpressive, one of Huang’s committee members said it was unprecedented for Rizzoni to have denied Huang a chance to take the exam again. This evidence could have undermined Rizzoni’s theory that Huang deserved to fail and, upset that she did, had an axe to grind against him. It would also have allowed the jury to infer that Rizzoni personally targeted Huang by dealing with her in an unusual and vindictive manner, giving credence to her story that Rizzoni was on a campaign against her because she refused his advances. It was an abuse of discretion for the district court to deem it irrelevant. B. Prejudice Having determined that the district court abused its discretion in excluding various categories of evidence, we must determine whether these errors entitle Huang to a new trial. We may order a new trial only if Huang has shown that the errors affected her “substantial rights,” A.K. ex rel. Kocher v. Durham Sch. Servs., L.P., 969 F.3d 625, 630 (6th Cir. 2020) (citing Fed. R. Evid. 103(a); Fed. R. Civ. P. 61; 28 U.S.C. § 2111). If they did, a new trial is necessary to preserve “substantial justice.” Martin, 106 F.3d at 1311 (quoting Fed. R. Civ. P. 61). To evaluate whether Huang has made that showing, we undertake a fact-intensive inquiry, “consider No. 23-3469 Huang v. Ohio State Univ. Page 33 the combined effect” of the district court’s errors, Beck v. Haik, 377 F.3d 624, 645–46 (6th Cir. 2004) (cleaned up), overruled on other grounds by Adkins v. Wolever, 554 F.3d 650 (6th Cir. 2009), and ask whether they “potentially chang[ed] the outcome” of the trial, P.I. & I. Motor Express, Inc. v. RLI Ins. Co., 40 F.4th 398, 416 (6th Cir. 2022). Our inquiry is not, however, whether Huang would have prevailed at trial. If it were, ordering a new trial on appeal would be the same thing as ordering judgment for the appellant. Our “function” as a court of error is not to resolve fact questions—that task is “exclusively for the jury.” Kotteakos v. United States, 328 U.S. 750, 763 (1946). Instead, we examine what the jury did hear at trial and ask whether we can “say with fair assurance” that “the judgment was not substantially swayed by the error.” Id. at 765; Schrand v. Fed. Pac. Elec. Co., 851 F.2d 152, 157 (6th Cir. 1988); Beck, 377 F.3d at 635. By excluding evidence of Rizzoni’s “manipulation, coercion, and influence” over Huang’s Ph.D. program, the district court deprived Huang of her “substantial rights” at trial. The evidentiary error here was systemic. Before trial even began, the district court imposed a false boundary between evidence of touching and evidence of Rizzoni’s power and manipulation. That meant that even before opening statements or the swearing-in of the first witness, Huang’s counsel had to devise a trial strategy within an improperly narrowed lane. It’s hard to evaluate the scope of that impact on the trial—but our view of the summary judgment record, and the deposition testimony of many witnesses Huang could not call, and exhibits she could not enter, leave us with the firm conviction that her trial did not provide “substantial justice.” Like many incidents of sexual misconduct, most of Huang’s allegations against Rizzoni are about conduct that occurred when no one else was around to witness it. So, the main thing the jury had to decide was whether to believe Huang or Rizzoni. That means, in addition to their own testimony, corroborating evidence and credibility evidence was paramount. On Huang’s side of the scale, all the jury had to consider was Huang’s own testimony. The district court had excluded everything else she tried to present, including corroborating evidence from her treating medical providers. Simply put, if the jury didn’t believe Huang, she would lose. And even with her own testimony, she was limited (as described above) in placing her theory of Rizzoni’s pattern of harassment before the jury. No. 23-3469 Huang v. Ohio State Univ. Page 34 On the flipside, Rizzoni had his own testimony and the testimony of his corroborating witnesses. Although one of those witnesses was able to specifically deny Huang’s allegation that Rizzoni “massaged” her on a trip to Dearborn, his witnesses mainly served to damage Huang’s credibility and intelligence. Those witnesses bolstered Rizzoni’s testimony that Huang was a poor student, deserved to fail her exam, and lied about the sexual misconduct in revenge. Yet Huang was not allowed to even testify to her full academic credentials and to how Rizzoni manipulated her committee. And the district court prevented her from calling witnesses that would have corroborated her version of events: that Rizzoni took unprecedented steps to rush her candidacy exam, goaded her examination committee into pressing her to fail, and then sealed Huang’s fate by delivering the “terminal” message that—contrary to policy—she would be unable to retake the exam. Dec. 7, 2017 Email, R. 99-2, PageID 4139. Presented in full, Huang’s evidence (if credited) would have raised significant questions. Why would Huang—who had been an excellent student her entire life and performed so well at Ford—want to avoid meeting with Rizzoni in the first place? Why was Rizzoni so harsh in his criticisms of Huang and why did he use such threatening language? And why did Rizzoni take such drastic and unprecedented steps towards Huang’s candidacy exam? The jury may not have credited her evidence, and even if it did, we cannot definitively say how it would have answered these questions. But the task was still for the jury, and the district court should have allowed it to receive all the relevant evidence. Rizzoni argues that the district court’s exclusion of this evidence was harmless because it would have been cumulative of the evidence the jury heard. See Kocher, 969 F.3d at 632. As all parties acknowledge, some evidence of Rizzoni’s influence, threats, and coercion did get to the jury. See Appellant’s Br. at 26 (conceding that “some scraps of evidence concerning Rizzoni’s retaliation and the harm Huang suffered” went to the jury). But the excluded evidence cannot be dismissed as cumulative. For one, so much was excluded as a categorical matter pre-trial that we do not know how the parties would have strategized under an alternative trial structure. For another, the district court’s ad hoc admission of some of this evidence does not mean that Huang got to paint a full picture to the jury. That’s particularly so because the only testimony the district court allowed Huang to introduce was her own. As the Supreme Court has recognized, No. 23-3469 Huang v. Ohio State Univ. Page 35 if “a jury naturally would tend to discount” testimony like Huang’s “as self-serving,” it is neither cumulative nor harmless to prevent a jury from hearing corroborating testimony from disinterested witnesses—such testimony carries “much greater weight.” Skipper v. South Carolina, 476 U.S. 1, 8 (1986). The district court’s exclusion of corroborating witnesses was therefore particularly prejudicial. The medical providers may have corroborated her accounts of sexual touching. And while Rizzoni begrudgingly testified that Ford was happy with Huang’s work, Anderson’s testimony about Huang’s sophisticated technical capabilities may have raised questions about why Rizzoni would terminate her from the URP. Huang’s professors and engineering department officials could have testified to Rizzoni’s unusual behavior leading up to the exam, and his unprecedented decision not to let her retake it. Keeping this from the jury, especially the corroborating witnesses, deprived Huang a fair trial and leaves us without “a ‘fair assurance’ that the trial’s outcome was not altered by error.” Beck, 377 F.3d at 635–36 (quoting Schrand, 851 F.2d at 157).6 It is true that some of the testimony that Huang wished to introduce is unfavorable to her. But we cannot surmise that the excluded favorable and unfavorable testimony somehow balanced out, obviating the need for a new trial. That’s not how juries work—in a trial, each piece of evidence “has force beyond any linear scheme of reasoning[.]” Old Chief, 519 U.S. at 187. For instance, Huang wanted to admit Dr. Canova’s deposition testimony; he expressed how poorly Huang performed on her candidacy examination and that he independently concluded she had failed. But he also testified that Rizzoni’s actions before and after the exam were unprecedented, suggesting he set her up to fail. Huang’s counsel determined that overall, this testimony was favorable to Huang, despite the negative parts. We don’t know what a jury would have taken from this testimony. But our role is not to say definitively how a jury would have reacted to the swath of evidence excluded from the trial. Cf. Kotteakos, 328 U.S. at 763 (“[I]t is not the appellate court’s function . . . to speculate.”). Instead, we ask whether Huang has shown that the evidentiary decisions affected her substantial rights, such that she is entitled to a new trial. The answer is yes. 6 See Old Chief, 519 U.S. at 187 (“[A]s [the evidentiary] pieces come together a narrative gains momentum, with power not only to support conclusions but to sustain the willingness of jurors to draw the inferences, whatever they may be, necessary to reach an honest verdict.”). No. 23-3469 Huang v. Ohio State Univ. Page 36 CONCLUSION For the foregoing reasons, we (1) reverse the district court’s grant of summary judgment to OSU on Huang’s Title VII harassment and retaliation claims, (2) vacate the judgment in Rizzoni’s favor on the § 1983 claim, and (3) remand for a new trial and proceedings consistent with this opinion. No. 23-3469 Huang v. Ohio State Univ. Page 37 _____________________________ CONCURRENCE / DISSENT _____________________________ CHAD A. READLER, Circuit Judge, concurring in part and dissenting in part. With all its twists and turns, trial court litigation is rarely easy. That is true for the parties and the district court alike. So despite the best efforts of all involved, missteps can—and often do—happen. When an error occurs, we are tasked with determining whether it is so severe that the judgment below cannot stand. Not all errors are created equal, and not all require a new trial. Nor should they, given the considerable resources a remand often demands. By and large, Meng Huang has not shown the type of grave error that warrants a new trial. In critiquing the district court’s summary judgment decision on her quid-pro-quo discrimination claim, she makes arguments she failed to make to that court. That is no basis for awarding her relief here. As for the issues submitted to the jury, Giorgio Rizzoni’s alleged conduct towards Huang fairly gives one pause. But the jury heard that evidence and decided that Huang had not proven her claims. While a handful of evidentiary errors occurred along the way, Huang has not demonstrated that they affected the trial’s outcome. A. Start with the district court’s award of summary judgment to Ohio State on Huang’s claim for quid pro quo discrimination. In framing our analysis, it bears reminding that we are “a court of review, not of first view.” Byrd v. Haas, 17 F.4th 692, 700 (6th Cir. 2021) (citation omitted). Accordingly, we “review the case presented to the district court, rather than a better case fashioned after a district court’s unfavorable order.” Barner v. Pilkington N. Am., Inc., 399 F.3d 745, 749 (6th Cir. 2005) (citation omitted). As litigants have well been warned, the failure to properly present an argument to the district court typically results in the forfeiture of the argument on appeal. Bannister v. Knox Cnty. Bd. of Educ., 49 F.4th 1000, 1011–12 (6th Cir. 2022); Sheet Metal Workers’ Health & Welfare Fund of N.C. v. L. Off. of Michael A. DeMayo, LLP, 21 F.4th 350, 355 (6th Cir. 2021); Scottsdale Ins. Co. v. Flowers, 513 F.3d 546, 552 (6th Cir. 2008). Specifically, in the instance of a party opposing summary judgment, that party must respond to the movant’s dispositive arguments to avoid forfeiture of those arguments on appeal. No. 23-3469 Huang v. Ohio State Univ. Page 38 Adkins v. Marathon Petroleum Co., 105 F.4th 841, 854 (6th Cir. 2024). While that result may seem harsh, it has several justifications. One that warrants emphasis here is the interest in preserving the district court’s role in resolving issues in the first instance. It is deeply unfair to the district court for us to reverse that court on grounds never presented to it by an appellant. Sheet Metal Workers’ Health & Welfare Fund of N.C., 21 F.4th at 355. It also wastes judicial resources to proceed that way. Id. Before the district court, Huang forfeited the arguments she makes today in support of her quid-pro-quo Title VII claim. By way of background, Ohio State moved for summary judgment on that claim in part on the ground that all the illicit conduct supporting the claim occurred while Huang was a student, not an employee, meaning her suit could not sound in Title VII, which covers employer-employee relationships. Ohio State distinguished between the first several years of Huang’s Ph.D. program, ending in August 2017—when she was classified as a “Graduate Fellow”—and the final year of her program—when she became a “Graduate Research Associate.” During the first several years of her Ph.D. program, the University explained, Huang, as a Graduate Fellow, “received University credits toward her [Ohio State] degree” and completed two internships with Ford Motor Company. R.105 PageID 5471. Although she “received stipends from the University to offset her educational costs during this time,” Ohio State maintained that these stipends were not employment related. Id. Because all of the alleged quid-pro-quo sexual harassment and its corresponding adverse effects pertained to Huang’s status as a student before her final year of the program, Ohio State concluded, she could not maintain a claim under Title VII. Huang filed an opposition brief. Yet she was conspicuously silent on her employment status prior to August 2017. She stated that Rizzoni “admit[ted] that Huang was employed” as a Graduate Research Associate, and “he expected her to be at work during working hours.” And she noted that she received stipends from the University. In a footnote, she added that “[g]raduate students like Huang are protected by Title VII in certain circumstances.” This was all true as to Huang beginning August 2017. But she said nothing in response to Ohio State’s argument that she was a student before that time. If anything, she reaffirmed that conclusion. In other portions of her opposition brief, in fact, Huang suggested that her Graduate Research No. 23-3469 Huang v. Ohio State Univ. Page 39 Associate role, one she assumed in August 2017, was distinct from other portions of her Ph.D. studies. R.114 PageID 6250 (“Rizzoni essentially terminated Huang’s Ph.D. studies and her employment as a graduate research associate.” (emphasis added)). In its order granting summary judgment to Ohio State, the district court recognized what it understood to be the parties’ shared understanding of Huang’s employment status. “Ohio State,” the district court acknowledged, “conced[ed] that [Huang] was an employee in her status as a Graduate Research Associate, which began in late August 2017.” But as to Huang’s prior role as a Graduate Fellow, the district court, seeing no argument to the contrary, concluded that the Fellow position was academic in nature, not employment related. Huang, the court noted, “presented no evidence to the contrary.” Tellingly, she did not ask the district court to reconsider that decision, a tack she took at many other stages of the district court proceedings. While such a motion is neither required nor favored, it seems most apt, if ever, when a court asserts that a party did not make an argument the party believes it in fact made, given the dramatic consequences tied to a forfeiture. On appeal, Huang alludes to several facts that she believes demonstrate she was an employee prior to August 2017. But having failed to present those arguments to the district court, Huang has forfeited the points, preventing us from entertaining them now. Sheet Metal Workers’ Health & Welfare Fund of N.C., 21 F.4th at 357; Bannister, 49 F.4th at 1012. True, as the majority opinion notes, Huang’s summary judgment opposition brief stated that she was “employed” as a Graduate Research Associate and characterized Rizzoni as her “supervisor.” But those statements were entirely unresponsive to Ohio State’s motion. There, the University acknowledged that Huang was an employee when she became a Graduate Research Associate. The key question was whether Huang was an employee before then. And her opposition brief failed to discuss the issue. It merely asserted that she was, at some point, employed by the University. That loose statement in no way refuted Ohio State’s argument that it was entitled to summary judgment on Huang’s quid-pro-quo claim. Much the same is true for the majority opinion’s assertion that Huang’s opposition brief generally identified record evidence capable of demonstrating her employee status. That is a No. 23-3469 Huang v. Ohio State Univ. Page 40 generous reading of Huang’s submission. She never argued, as the majority opinion implies, that her “duties to [Ohio State] on the Ford URP” before August 2017 made her an employee. See Maj. Op. at 16 n.2. Nor did she argue that the “control Rizzoni imposed on her time” during that period similarly informed her employment status. Id. For support on these points, the majority opinion at most can point to occasional record cites, not argument. And only with the benefit of hindsight would one even begin to realize that these scattered citations in Huang’s summary judgment brief support the position she now asserts. At best, Huang averred to these matters in “a perfunctory manner, unaccompanied by some effort at developed argumentation.” Cockrun v. Berrien County, 101 F.4th 416, 418 (6th Cir. 2024) (citation omitted). That is insufficient to preserve the point. Remember, as the party opposing summary judgment, Huang was required to designate portions of the record with “enough specificity that the district court can readily identify the facts upon which the nonmoving party relies.” Guarino v. Brookfield Twp. Trs., 980 F.2d 399, 405 (6th Cir. 1992). When it instead stays “silent[,] its opportunity is waived and its case wagered.” Id. Otherwise, we penalize the district court for not resolving merits arguments never made to that court. With little to defend Huang’s litigating choices, the majority opinion attempts to turn the tables. It was Ohio State, the majority opinion maintains, who merely “asserted, rather than argued, [that] Huang was not an ‘employee’ while a Graduate Fellow.” Maj. Op. at 16 n.2. Setting aside the dim distinction between asserting and arguing, the fact remains that Ohio State’s summary judgment motion amply contested Huang’s alleged employment status. Yes, it noted the different titles Huang held during her time at the university. But Ohio State did not stop there. It added, for example, that Huang received stipends “to offset her educational costs,” evidence the majority opinion considers quite probative of Huang’s employment status during the period of alleged harassment. See id. at 20–21. Even the majority opinion concedes that this collection of points amounted to at least a “brief argument” as to Huang’s employment status. Id. at 16 n.2. In the end, the majority opinion simply takes issue with the perceived weakness of Ohio State’s assertion. That is far more than Huang ever did. And that reality should end the matter. An opposing party, remember, can forfeit counterarguments to good and bad contentions alike. Cf. Ohio State Univ. v. Redbubble, Inc., 989 F.3d 435, 445 (6th Cir. 2021). By failing to respond to Ohio State’s argument, Huang did just that. No. 23-3469 Huang v. Ohio State Univ. Page 41 Nor can I agree that this issue is preserved because the district court purportedly addressed the arguments Huang now advances. Two problems plague that assertion. One, even the majority opinion agrees that the district court did not tackle those issues. See Maj. Op. at 17– 18. As that opinion acknowledges, the district court never considered factors such as “how much control OSU exercised” over Huang and whether that degree of control established a principalagent relationship. Id. at 16–17. So it is quite a leap to characterize the district court as having taken up the arguments Huang makes now in the first instance. Two, this framing ignores the settled duties Huang disregarded in district court. Remember that at summary judgment, a “nonmoving party can forfeit an argument if,” like Huang, she “fail[s] to respond to the moving party’s arguments.” Adkins, 105 F.4th at 854; see Bennett v. Hurley Med. Ctr., 86 F.4th 314, 324 (6th Cir. 2023); Scottsdale, 513 F.3d at 551–54; Chicago Title Ins. Corp. v. Magnuson, 487 F.3d 985, 995 (6th Cir. 2007); Guarino, 980 F.2d at 405. That is what happened here. That makes this case far different from one in which a district court addresses the merits of arguments not put forth by the parties. Perhaps the district court could have conducted a more extensive analysis of the record to evaluate Huang’s employment status prior to August 2017. But it had no obligation to do so. See Murthy v. Missouri, 144 S. Ct. 1972, 1991 n.7 (2024); Wimbush v. Wyeth, 619 F.3d 632, 638 n.4 (6th Cir. 2010) (explaining that parties must “point to the evidence with specificity and particularity in the relevant brief”). In the end, it is incumbent on the parties to “advanc[e] the facts and arguments entitling them to relief.” United States v. Akridge, 62 F.4th 258, 264 (6th Cir. 2023) (citation omitted). Huang failed to do so. That she fashions a “better case” on appeal is no basis for undermining the district court’s judgment. Barner, 399 F.3d at 749. B. Huang likewise failed to demonstrate that she was prejudiced by any evidentiary errors at trial. No two trials look alike, and none are perfect. See United States v. Hasting, 461 U.S. 499, 508 (1983). But most stand up on appeal. So when does the exclusion of evidence affect the outcome of a trial so deeply as to require a re-do? In civil cases, the party seeking a new trial must show that a district court’s evidentiary errors affected her “substantial rights.” A. K. ex rel. Kocher v. Durham Sch. Servs., L.P., 969 F.3d 625, 630 (6th Cir. 2020). In most cases, that means the alleged error must have “made a difference in the trial’s outcome.” Id. at No. 23-3469 Huang v. Ohio State Univ. Page 42 632; see also Dortch v. Fowler, 588 F.3d 396, 400, 402 (6th Cir. 2009). If it is unclear whether the refusal to admit admissible evidence in a civil case would have affected the trial’s outcome, the verdict stands, as the appellant bears the burden to show prejudice. Kocher, 969 F.3d at 629. To attempt to climb that mountain, Huang points to her inability to introduce three categories of evidence that she believes support her efforts to prove that Rizzoni inappropriately touched her. 1. The first category of evidence concerns her academic performance, including the reasons she failed her Ph.D. candidacy exam. Huang accused Rizzoni of causing her to fail the exam after she resisted his advances by making the exam process excessively difficult. Accordingly, she says, evidence and testimony about her academic and professional capabilities could have shown that she was a capable student who did not deserve to fail. And if she could prove that her exam was rigged, Huang adds, it would tend to corroborate her story that Rizzoni retaliated against her after she resisted his advances. Her contention faces multiple hurdles. First, even if relevant, see Fed. R. Evid. 401, the evidence she highlights was excludable on other grounds. As a matter of law, there is no prejudice if the excluded evidence would have been inadmissible for another valid reason. See 11 Charles Alan Wright, Arthur R. Miller & Mary Kay Kane, Federal Practice & Procedure § 2885 (3d ed. 2024); In re Air Crash Disaster, 86 F.3d 498, 530 n.21 (6th Cir. 1996) (“Although the court did not exclude the evidence . . . under Rule 403 . . . that exclusion . . . may alternatively be affirmed under that Rule.”). Here, that reason is Federal Rule of Evidence 403. A district court “may exclude relevant evidence if its probative value is substantially outweighed by a danger of . . . unfair prejudice, confusing the issues, misleading the jury, undue delay, wasting time, or needlessly presenting cumulative evidence.” Fed. R. Evid. 403. District courts enjoy considerable latitude in making that assessment. See United States v. Howard, 621 F.3d 433, 457 (6th Cir. 2010). In this case, it seems all but certain that the district court would have excluded, under Rule 403, evidence and testimony on Huang’s academic record. In responding to an offer of proof by Huang’s counsel, the district court explained that it would be “very complicated” and “time-consuming” for a jury to determine why Huang failed her candidacy exam. R.218 PageID 7929. The court feared that a trial on this question would confuse the jurors and cause them to No. 23-3469 Huang v. Ohio State Univ. Page 43 conflate educational questions and legal ones. Id. PageID 7928. As Huang’s and Rizzoni’s academic and professional disagreements were long running, focusing on those issues, the district court worried, would be “a total distraction from the core issue” of whether “Rizzoni engage[d] in unwanted physical touching” of Huang. Id. PageID 7924–25. In other words, even if the reason Huang failed her candidacy exam was relevant to her allegations of inappropriate sexual contact, it was not an abuse of discretion to limit testimony on the issue in light of Rule 403. Consider the complexity of the parties’ academic relationship and the factors tied to evaluating a Ph.D. exam. Huang’s academic record spanned almost four years. During that time, she and Rizzoni often clashed over the classes Huang was taking, the work Rizzoni asked her to perform, and her availability to discuss research. Placing those disputes before the jury would have complicated the trial. The same is true for Huang’s candidacy exam. Huang was pursuing a Ph.D. in mechanical engineering. The exam involved written and oral components. It was administered by a committee consisting of five members of Ohio State’s faculty. And each committee member explained (often in technical language) their reasons for failing Huang. With all of this in mind, determining whether Huang deserved to fail the exam inevitably would have involved intricate questions about Huang’s competency in the material and the exam’s relative difficulty. Huang would have turned the jury’s attention away from her interactions with Rizzoni and toward complicated matters such as whether, during her exam, she failed to explain “that the SEI Layer is a result of the decomposition of the electrolyte solvent, which leads to loss of cyclable lithium.” R.105-8 PageID 5861. Broaching these issues ran a deep risk of confusing the jury and unduly extending the trial. And even if the jury successfully waded through Huang’s academic record, it is not clear that, in the end, it would have had a better sense of whether Rizzoni sexually assaulted her. Accordingly, excluding this evidence on relevance grounds was at most harmless error as the evidence failed Rule 403’s admissibility standards. See United States v. Seymour, 468 F.3d 378, 387 (6th Cir. 2006); Jones v. Wiseman, 838 F. App’x 942, 949–51 (6th Cir. 2020). Second, although this information was ultimately inadmissible, the jury eventually learned of the evidence during trial. In our circuit, the exclusion of relevant evidence is not No. 23-3469 Huang v. Ohio State Univ. Page 44 prejudicial when the facts proven by the excluded evidence were covered by “substantially similar” evidence at trial. M.J. ex rel. S.J. v. Akron City Sch. Dist. Bd. of Educ., 1 F.4th 436, 447–48 (6th Cir. 2021); In re Air Crash Disaster, 86 F.3d at 526. That was the case here. The jury heard both testimony and argument suggesting that Rizzoni caused Huang to fail the exam as punishment for refusing to give in to his advances. Huang testified that Rizzoni retaliated against her when he “changed [her] committee member for [her] Ph.D. candidacy exam and failed [her] and denied [her] the second chance without giving any solid reasons,” and then “cut [her] off of [her] research project and the collaboration with Ford.” R.217 PageID 7856. Huang added that Rizzoni’s harassment ended after he “took away the Ford project [she] had been working on for years, and kicked [her] out of the program,” which “jeopardize[d] [her] immigration status.” R.218 PageID 7966; see also id. at 8049–50 (similar). Huang also explained to the jury that Rizzoni’s professional criticism stemmed from her resistance to his advances. The jury likewise heard about Huang’s success at Ford as well as her ability to successfully complete her Ph.D. under a new advisor. Huang testified that she performed well on her assignments at Ford, “received a very good review and feedback,” and was invited to make a presentation to one of Ford’s chief engineers. Rizzoni himself later confirmed that Huang’s company supervisors were pleased with her work. And the parties stipulated that Huang successfully completed her Ph.D. after she reported Rizzoni’s conduct and was assigned a new advisor. Huang’s counsel drove the point home during closing arguments. Counsel explained that after years of resisting his advances, Rizzoni gave Huang “the final punishment” when he caused her to fail the exam. Yet despite Rizzoni’s efforts “to expel her from his program and destroy her,” Huang successfully completed her Ph.D. and secured a prestigious job. Accordingly, even if the district court was unjustified in excluding some items related to Huang’s academic record, Huang nonetheless presented much of this evidence to the jury. And, it bears adding, she did so without rebuttal. Even though Rizzoni’s theory of the case was that Huang accused him of misconduct in retaliation for her having failed the exam, he did not put on direct evidence that Huang deserved to fail because she was a poor student. R.219 No. 23-3469 Huang v. Ohio State Univ. Page 45 PageID 8297 (Rizzoni’s counsel stating at a sidebar near the end of trial that “we are not getting into the reasons for her failing the candidacy exam, and I’ve stayed very far away from that”). Had he done so, it would have been unlikely to help Huang. During discovery, the members of Huang’s candidacy committee submitted affidavits explaining that Huang failed the exam because she lacked “essential foundational knowledge” and failed to answer relatively simple questions. They also expressed concerns about Huang’s “incredible lack of curiosity” and her “argumentative” and “dismissive” attitude during the exam. Opening the trial to evidence of Huang’s academic record thus would not have elevated her chances of success, indeed it may have harmed her. For these reasons too, it was no more than harmless error to keep this issue from the jury. 2. Huang also sought to introduce circumstantial evidence that Rizzoni inappropriately touched her. For various reasons, exclusion of that evidence similarly fails to warrant a new trial. Start with an email that Rizzoni sent to Huang shortly after one alleged incident. In the email, Rizzoni called Huang stupid and stubborn. Huang claims the district court declined to let her discuss this email. Not so. On direct examination, Huang’s counsel read from the email, confirming with Huang that Rizzoni called her “stupid” and “stubborn” just days after an alleged assault. Counsel also displayed the email to the jury. By the time the district court eventually ordered counsel to move on and discontinue discussion of the email, Huang had already testified to its contents. Huang has not explained how further discussion would have affected the trial’s outcome. M.J., 1 F.4th at 447–48. So too for emails Rizzoni sent to Huang and Ford employees related to disputes over the classes Huang was taking, many of which Rizzoni thought were a waste of time. While the emails themselves were excluded, their substance was not. Huang testified about Rizzoni’s criticisms of her class schedule. She similarly testified that Rizzoni threatened her in person and by email when she did not give into his sexual demands. See R.218 PageID 7917, 7919, 7933. Huang twice confirmed that Rizzoni sent her emails characterizing her, among other derisive terms, as “positively the worst Ph.D. student that he has ever advised.” R. 218 PageID 7933, 7936. And, it is worth emphasizing, Rizzoni never disputed that he sent these emails. No. 23-3469 Huang v. Ohio State Univ. Page 46 True, direct evidence of these emails may have reenforced Huang’s argument that Rizzoni was unfairly and unnecessarily critical of her, especially in the weeks following his allegedly illicit conduct. But because Huang addressed Rizzoni’s criticisms, it is hard to say that the emails themselves were so probative that their introduction would have altered the trial’s outcome. See Dortch, 588 F.3d at 402 (deeming harmless the exclusion of evidence with “little probative value”). 3. That leaves the testimony of two medical providers from whom Huang sought psychiatric treatment in the months after she reported Rizzoni’s alleged conduct. This issue seemingly is closer than the rest. The evidence would have been relevant at trial to corroborate Huang’s story, and Rule 403 likely would have been satisfied as well. But its exclusion was largely due to missteps by Huang’s trial counsel, making it difficult to pin the omission on the district court’s relevancy determination. Turn to the record in district court. Rizzoni moved to quash the subpoenas issued to two of Huang’s medical providers at Ohio State. Rizzoni argued that the providers’ testimony would be irrelevant to the liability stage of the trial. Huang opposed the motion but failed to argue why the testimony would be relevant. Yes, Huang later explained the relevancy of that testimony in a motion for reconsideration. But the time to make those arguments had passed. Respecting the district court’s time and authority, a motion for reconsideration cannot be used “to raise arguments which could, and should, have been made” earlier. Leisure Caviar, LLC v. U.S. Fish & Wildlife Serv., 616 F.3d 612, 616 (6th Cir. 2010) (citation omitted). It follows that “[a]rguments raised for the first time in a motion for reconsideration are untimely and forfeited on appeal.” Evanston Ins. Co. v. Cogswell Props., LLC, 683 F.3d 684, 692 (6th Cir. 2012). Numerous other problems further undermine Huang’s argument. As a factual matter, the jury, in the end, was told that Huang sought psychiatric treatment after going public with her story. Huang testified that she sought “medical and therapy treatment” after being subjected to Rizzoni’s inappropriate touching, and Rizzoni’s counsel used records of Huang’s therapy appointments to refresh her memory and confirm that she sought medical treatment several times in 2019. That relegates any error to harmless territory. Likewise, as a practical matter, we do No. 23-3469 Huang v. Ohio State Univ. Page 47 not know how the providers would have testified, making it difficult to say that their testimony would have altered the trial’s outcome. And as a procedural matter, Huang’s subpoenas to the medical providers violated Federal Rule of Civil Procedure 45 (Subpoena). As Rizzoni explained in district court, Huang neither served the subpoenas on the providers in question nor gave the providers a reasonable time to comply with the subpoenas’ demands. R.182 PageID 7219–23 (explaining that none of the witnesses had been served because eight days before trial, “Plaintiff’s process server picked a random building on OSU’s campus and dropped off a stack of witness subpoenas to the receptionist at a customer service desk”). Accordingly, even if it were error to quash the subpoenas on relevancy grounds, the district court could have quashed them pursuant to Rule 45. See Hill v. Homeward Residential, Inc., 799 F.3d 544, 552–53 (6th Cir. 2015) (holding that it is not an abuse of discretion to quash a subpoena that does not comply with Rule 45); see also EEOC v. Ferrellgas, L.P., 97 F.4th 338, 350 n.7 (6th Cir. 2024) (explaining that we “can affirm a decision of the district court on any grounds supported by the record, even if different from those relied on by the district court”). The majority opinion views the evidence differently. That may be because it is viewing it through the wrong lens. The majority opinion starts off on the right foot by noting that Huang bears the burden of showing that any error below affected her substantial rights. See Maj. Op. at 32 (citing Kocher, 969 F.3d at 630). But it then proceeds to apply a materially different standard. According to the majority opinion, a new trial is warranted because it cannot “‘say with fair assurance’ that ‘the judgment was not substantially swayed by’” the exclusion of evidence from Huang’s trial. Id. at 33. In practice, that formulation simply amounts to considering whether there was a chance the error altered the trial. Perhaps that is a correct formulation of the standard for prejudice in criminal cases, where the government must show that any error did not affect the outcome of the trial. See, e.g., United States v. Agrawal, 97 F.4th 421, 429 (6th Cir. 2024) (explaining that, at a minimum, the government must provide a “fair assurance that the alleged error did not substantially sway the verdict” (cleaned up)). In other words, in a criminal trial, if the whole record shows that the alleged error may have affected the jury verdict, we must reverse. But in civil cases, demonstrating prejudice requires a greater No. 23-3469 Huang v. Ohio State Univ. Page 48 showing. In that setting, the appellant bears the burden of showing that the alleged error affected her substantial rights, meaning the error affected the outcome of the trial. Kocher, 969 F.3d at 629, 632; Dortch, 588 F.3d at 400, 402; see also Tompkin v. Philip Morris USA, Inc., 362 F.3d 882, 891 (6th Cir. 2004) (explaining that even in the wake of an evidentiary error, “a new trial will not be granted unless the evidence would have caused a different outcome at trial” (quoting Morales v. Am. Honda Motor Co., 151 F.3d 500, 514 (6th Cir. 1998))); Muldrow ex rel. Est. of Muldrow v. Re-Direct, Inc., 493 F.3d 160, 168 (D.C. Cir. 2007) (“[T]he error must have been prejudicial: It must have affected the outcome of the district court proceedings.” (quoting United States v. Olano, 507 U.S. 725, 734 (1993))). Where does the majority opinion look to derive its “fair assurance” standard? Our decision in Schrand v. Fed. Pac. Elec. Co., 851 F.2d 152, 157 (6th Cir. 1988). Schrand, in turn, conjured this standard from Kotteakos v. United States, 328 U.S. 750, 765 (1946). Yet Kotteakos has since been recast to support the outcome-determinative (rather than fair assurance) test. See Olano, 507 U.S. at 734 (citing Kotteakos, among other cases, to establish the outcomedeterminative test). Much to the same end, the majority opinion invokes our decision in Beck v. Haik, 377 F.3d 624 (6th Cir. 2004), as favoring the middling fair assurance test. But it fails to mention our more recent decision in Kocher, which expressly noted Beck’s inconsistency with the Supreme Court’s later guidance in Shinseki v. Sanders, 556 U.S. 396 (2009), and which instead opted to apply the outcome-determinative test. Kocher, 969 F.3d at 629–630, 632. Again, in a close case, “the tie goes to the verdict.” Id. at 630. It is not enough, therefore, to show a mere chance that the alleged error affected the trial. Instead, Huang must show that the trial’s outcome was actually affected. Is it possible that the excluded evidence affected the outcome of the trial? Perhaps. But has Huang carried her burden to demonstrate it in fact did? No. Said differently, one cannot definitively say that the omitted items altered the jury’s verdict. In short, Huang is not entitled to a new trial.

Some case metadata and case summaries were written with the help of AI, which can produce inaccuracies. You should read the full case before relying on it for legal research purposes.

This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.