J. Endres v. Ne. Ohio Med. Univ.

938 F.3d 281

• Court: Court of Appeals for the Sixth Circuit
• Decided: August 30, 2019
• Docket: 18-3825
• Status: Published

Opinion

RECOMMENDED FOR FULL-TEXT PUBLICATION Pursuant to Sixth Circuit I.O.P. 32.1(b) File Name: 19a0222p.06

UNITED STATES COURT OF APPEALS FOR THE SIXTH CIRCUIT

J. ENDRES, ┐ Plaintiff-Appellant, │ │ │ v. │ > No. 18-3825 │ NORTHEAST OHIO MEDICAL UNIVERSITY; NORTHEAST │ OHIO MEDICAL UNIVERSITY BOARD OF TRUSTEES; │ SANDRA EMERICK; JAY A. GERSHEN, │ Defendants-Appellees. │ ┘

Appeal from the United States District Court for the Northern District of Ohio at Akron. No. 5:17-cv-02408—Sara E. Lioi, District Judge.

Argued: May 2, 2019

Decided and Filed: August 30, 2019

Before: MERRITT, KETHLEDGE, and NALBANDIAN, Circuit Judges.

_________________

COUNSEL

ARGUED: John S. Marshall, MARSHALL AND FORMAN LLC, Columbus, Ohio, for Appellant. Todd R. Marti, OFFICE OF THE OHIO ATTORNEY GENERAL, Columbus, Ohio, for Appellees. ON BRIEF: John S. Marshall, MARSHALL AND FORMAN LLC, Columbus, Ohio, for Appellant. Todd R. Marti, Anthony J. Farris, OFFICE OF THE OHIO ATTORNEY GENERAL, Columbus, Ohio, for Appellees. No. 18-3825 Endres v. Ne. Ohio Med. Univ., et al. Page 2

OPINION _________________

NALBANDIAN, Circuit Judge. Northeast Ohio Medical University (“NEOMED”) dismissed Julian Endres, a medical student, for cheating on a test. Endres denies cheating, but his guilt or innocence is not relevant to this appeal. Instead, this appeal presents two questions. The first concerns when Endres should have brought the claims that he brings here. And the second concerns what process the school should have afforded him before dismissing him. After his dismissal, Endres sued NEOMED, its president and board of trustees, and a NEOMED administrator, Sandra Emerick, alleging violations of his right to procedural due process, along with the Americans with Disabilities Act and the Rehabilitation Act of 1973. But before considering the substance of those claims, the district court dismissed Endres’s complaint as untimely. The district court also held that even if Endres’s due process claim were timely, Emerick is entitled to qualified immunity. Endres appeals that decision.

We hold that Endres’s claims are timely and REVERSE the district court’s dismissal of his complaint. Endres has also alleged facts which, taken as true, establish several violations of his procedural due process rights. But because the contours of those rights were not clearly established, we AFFIRM the district court’s grant of qualified immunity to Emerick, which immunizes her from damages though not from injunctive relief.

I.

Julian Endres has Attention Deficit Hyperactivity Disorder (“ADHD”). Since his diagnosis at age six, Endres has taken medication to treat that condition, beginning with Ritalin, and he seemed to manage well. Endres graduated from high school as valedictorian and enrolled in the accelerated B.S./M.D. program at the Northeast Ohio Medical University (“NEOMED”), a public university. After completing his undergraduate studies in two years, graduating magna cum laude, Endres began medical school.

When Endres started his second semester of medical school, he experienced a new side effect from Ritalin. Endres alleges that around March 2015, he felt that the medicine was over- No. 18-3825 Endres v. Ne. Ohio Med. Univ., et al. Page 3

sedating him and making him lethargic—so he stopped taking it altogether. By that time, Endres had passed fourteen of the fifteen required classes to complete his first year of medical school, but he still had one class left. No longer on Ritalin, Endres failed that class. So NEOMED made Endres repeat the entire first-year curriculum during the following academic year, including the fourteen classes he had passed.

Before he returned to NEOMED that fall, Endres consulted with his physicians to find a medicine that would treat his ADHD without inducing the unwanted side-effects, and in August 2015, Endres switched to Strattera. Endres alleges that Strattera helped him concentrate without making him feel drowsy but that it was not a magic bullet. Unlike Ritalin, Strattera did not suppress his fidgeting. Even so, he stuck with the new medicine.

The Test. Back at NEOMED, Endres re-enrolled in the mandatory first-year curriculum, including Human Development and Structure (“HDS”), taught by Professor Hans Thewissen. It made no difference that Endres passed HDS the prior year because he had to retake all first-year classes.

Tests at NEOMED are a serious undertaking. Students take their tests on NEOMED’s laptops, using a special software program that allows them to zoom in and out and manipulate the images on the screen. Endres alleges that NEOMED’s laptops are set at the lowest brightness level and that students cannot make the screen any brighter. And while students are taking tests, NEOMED officials are watching—both with their own eyes and with a camera that records students taking examinations. Students knew that NEOMED was filming them. Indeed, Endres had worked part-time as an assistant, helping NEOMED professors set up the camera to record class lectures.

This case involves the 25-question test Endres took on September 28, 2015, as part of the HDS course. Endres sat directly in front of the camera. And as it turned out, the camera focused on Endres throughout much of the test. Footage from the test reveals Endres fidgeting and repeatedly glancing toward his right—and in the direction of his seatmate’s (“Student B”) laptop. Endres concedes that his eye movements and sideways glances might look suspicious to someone watching the video with no context, but he offers an alternate explanation. According No. 18-3825 Endres v. Ne. Ohio Med. Univ., et al. Page 4

to Endres, Student B repeatedly used the software’s zoom feature, which created brief flashes of light that caused Endres to look repeatedly to the source. In any event, Endres contends that it was physically impossible to see any legible content on the laptop to his right, both because he was sitting about five feet away, and because all laptops were set to the lowest brightness level.

When the test ended, one of the proctors completed an irregularity report, noting that Endres appeared to look repeatedly at the laptop to his right—but that perhaps he was just nervous. That report made its way to Sandra Emerick, NEOMED’s chief officer of student affairs. And so NEOMED’s investigation of Endres began.

The Investigation and Referral to CAPP. Endres alleges that Emerick began a three-day investigation into his conduct during the September 28 test. Emerick reviewed the video recording of the test and obtained Endres’s and Student B’s answer keys, which revealed that 84% of their answers were identical. Student B and Endres answered the same six questions incorrectly, the same fifteen questions correctly, and the same four questions differently. In the end, Endres scored a 72 while Student B scored a 64. The class mean was a 56.5. Emerick then sent her findings to several university administrators and professors, including Thewissen, writing that the video footage “clearly” showed Endres “viewing peer’s computer screen repeatedly during the exam,” and noting in boldface that “84% of the tests were identical.” (R. 23, First Am. Compl. at ¶ 47.)

NEOMED’s student handbook outlines various procedures to address allegations of student misconduct, including cheating, and gives NEOMED administrators discretion over which procedures to use. Under the handbook, the chief officer of student affairs works with several administrators to decide where to refer misconduct allegations. The administrators have three basic options—they may refer the allegations to: (1) a formal disciplinary body; (2) the Committee on Academic and Professional Progress (“CAPP”); or (3) a less formal process, such as counseling or mediation. When administrators refer a matter to the formal disciplinary body, the student has a right to an adversarial hearing—where he may learn the evidence against him, cross-examine witnesses, and present evidence in his defense. No. 18-3825 Endres v. Ne. Ohio Med. Univ., et al. Page 5

Emerick referred the matter to CAPP, which reviews issues related to academic performance and professionalism. CAPP accepted the referral and later sent Endres a letter notifying him that he had to attend a hearing on October 21, where a panel would discuss the “professionalism concern,” as well as Endres’s “overall educational performance.” (Id. at ¶ 80.) The letter instructed Endres to review his student file and to complete a student interview form with Emerick, which Emerick would then add to his student file and present to CAPP at the hearing.

Emerick Meets with Endres. After referring the matter to CAPP, Emerick called Endres for a brief meeting on October 7, marking the first time that he learned of the allegations against him. Endres alleges that he was “completely stunned” to learn that he stood accused of cheating. (Id. at ¶ 72.) And when Emerick showed Endres footage of the test, Endres alleges he was “shell-shocked to see how ‘bad’ his fidgeting and side-glances caused by distractions looked.” (Id.) Endres informed Emerick that he had recently switched from Ritalin to Strattera, which did not suppress his fidgeting, so Emerick requested that Endres provide documentation from his doctor about the medication change. Less than a week later, Endres’s doctor, Dr. Caroline Ramos, sent Emerick a letter describing Endres’s ADHD, along with medical records documenting Endres’s medication adjustment between April and August 2015. Ramos stated that she viewed the video recording and found that Endres’s behavior during the test mimicked the fidgeting that Endres had described at an April 2015 appointment. Ramos also suggested that because Endres had recently started the Strattera, NEOMED should make certain accommodations for him, especially so that no one would misinterpret his fidgeting as academic misconduct.

After the initial meeting, Endres emailed Emerick about his case. Endres explained to Emerick that the position of the laptops prevented a test-taker from viewing any meaningful information on a nearby laptop. So Endres requested that CAPP recreate the testing conditions through a “field test,” which would prove that it was impossible to cheat. Endres alleges that Emerick never responded to that request. In the same email, however, Endres requested information and evidence on his case, and Emerick responded to that request, explaining that his No. 18-3825 Endres v. Ne. Ohio Med. Univ., et al. Page 6

official student file contained all information from the investigation that she would present to CAPP.

On October 14, Endres visited the student enrollment office to review his file. Endres alleges that he found a folder containing all of his records since he applied to NEOMED as a high-school student—and in no particular order. According to Endres, his file contained only two records related to the misconduct allegation: (1) an email from Emerick, referring the matter to CAPP; and (2) Endres’s and Student B’s answer keys to the September 28 test. Endres alleges that the proctor’s irregularity report was missing from his file, prompting him to email Emerick and request to view the report. Endres returned to the student enrollment office the next day to view the report, which was then in the file.

Endres met once more with Emerick before the CAPP hearing. He alleges that Emerick did not discuss any information about his case or describe to him the evidence that she would present before CAPP. Endres then submitted his student interview form, a personal statement, a statement from Dr. Ramos, and his medical records, which CAPP could review during the hearing.

The First CAPP Hearing. CAPP conducted the first hearing on October 21. CAPP procedures allow the accused student to speak to his unsatisfactory performance and respond to questions from committee members but do not allow the accused student in the room while the committee reviews documents about his case. So the meeting began with Emerick in the room, presenting the case against Endres, while Endres waited down the hall. After Emerick finished her presentation, the panel allowed Endres into the room, and he presented his case. Endres explained to the panel that it would be physically impossible to commit the alleged misconduct and described the field test he conducted with three randomly selected NEOMED students. According to Endres, all three students agreed that it was impossible to view any information on the laptop immediately to their right—even if they turned their heads and stared directly at the laptop. Endres also explained his sideways glances, telling the panel about his recent medication change and that Strattera did not suppress his fidgeting or movements. And Endres noted that the flashes of light from Student B’s computer repeatedly distracted him during the test. As Endres explained, NEOMED’s laptop software allows students to zoom in on pictures and No. 18-3825 Endres v. Ne. Ohio Med. Univ., et al. Page 7

diagrams, causing a color change on the laptop screen—and for Endres, a flash appearing in the corner of his eyes. This prompted him to turn repeatedly to his right, whenever his seatmate adjusted the content on the laptop. Finally, Endres discussed the exam answers themselves. Although Endres did not have the answer keys for the entire class, he explained that the percentage of identical answers that two students might share simply turns on their scores and the number of correctly or incorrectly answered questions. According to Endres, the normal distribution range of similarity for two students who scored a 64 and a 72 (Student B’s score and Endres’s score, respectively) would be between 64% and 92%. Thus, Endres explained that the fact that he and Student B answered 84% of the questions identically was unremarkable.

After Endres completed his 30-minute presentation, the panel questioned him about why he did not request an accommodation before taking the test. Endres explained that before September 28, he had not taken any test while on Strattera and that he had never needed any accommodations while on Ritalin. Endres alleges that Emerick remained in the room throughout his presentation and the panel’s questions. But according to Endres, the panel did not inform Endres about any part of Emerick’s presentation, nor did the panel ask Endres any questions related to that presentation.

Endres left the room, and the panel began its deliberations with Emerick still present. Emerick called Endres the next day, telling him that the panel found him responsible for the misconduct. Endres later received an email informing him of the consequences of CAPP’s decision. He could: (1) withdraw within four days, in which case his transcript would state that he had withdrawn; (2) accept the decision for dismissal, in which case his transcript would state that he had been dismissed; or (3) request an appeal of CAPP’s decision if there is “significant and compelling new information” unavailable at the hearing or if he had evidence that there was a defect or irregularity in the proceedings. (Id. at ¶ 118.)

Endres and His Parents Meet with Emerick. Two days after the hearing, Endres and his parents met with Emerick. Emerick discussed the evidence that she presented during the hearing and that the committee relied on in reaching its decision. One key piece of evidence was a statistical analysis from Professor Thewissen (“Thewissen Analysis”), which claimed that there was a 0.000036% chance that two students would have six identical wrong answers. Endres No. 18-3825 Endres v. Ne. Ohio Med. Univ., et al. Page 8

alleges that this was the first time he learned about the Thewissen Analysis and that the analysis was not present when he reviewed his student file before the hearing. The other key piece of evidence was Emerick’s verbal interpretation of the video recording from the test. Emerick relayed to Endres and his parents that the panel did not believe that his ADHD caused his eye movements and that he should have used his right hand to shield his eyes from the occasional flashes of light.

Endres Files Appeal, Requests Additional Evidence. NEOMED’s student handbook provides students with a limited right of appeal. The Executive Review Committee (“ERC”) considers appeals—but only if the student presents “significant and compelling new information that was not available for presentation [to CAPP], or evidence of a defect or irregularity in the CAPP [ ] Committee proceeding.” (R. 25, Mot. to Dismiss Appx. at PageID #691.) The student must file an appeal (NEOMED calls it a “Petition for Executive Review”) within four working days after learning of CAPP’s decision, and if he fails to file an appeal in that time, CAPP’s “decision becomes final.” (Id.) Moreover, CAPP’s decision becomes final if the ERC declines to accept the appeal. (Id.)

After the meeting with Emerick, Endres requested additional evidence related to the exam, which he planned to introduce on appeal. In particular, Endres sought: (1) the student performance report for the entire HDS class, containing each individual score and answer key; (2) the Thewissen Analysis; and (3) timestamp reports from the laptop software, showing precisely when Student B and Endres answered each exam question. Although NEOMED first declined to provide Endres with this evidence, it ultimately turned over those records on the eve of Endres’s deadline to appeal.

Hours later, Endres filed his appeal with the ERC and submitted new evidence, including: (1) hundreds of test comparisons based on the actual data from the HDS class, which Endres alleges undermined the Thewissen Analysis and the fact that his answers and Student B’s answers were 84% similar; (2) signatures from HDS students, who agreed that it was impossible to view content on a nearby laptop; and (3) a psychological evaluation from Dr. Andrew Trauben, a psychiatrist specializing in ADHD, who reviewed the video of the test and met with Endres to evaluate his symptoms. Endres tried to submit additional evidence after filing his No. 18-3825 Endres v. Ne. Ohio Med. Univ., et al. Page 9

appeal, including his own statistical analysis that allegedly debunked the Thewissen Analysis. Hours before the ERC hearing, however, Emerick emailed Endres to notify him that the ERC would not accept his report and that he could not distribute copies of that report to ERC members.

The ERC Hearing. The ERC conducted a hearing on November 2, using the same procedures as CAPP. This meant that Emerick presented her case while Endres waited outside the room until it was his turn to speak. Endres then entered the room, and although he could not distribute copies of his statistical analysis, he described that analysis—as well as flaws in Emerick’s statistical comparison and the Thewissen Analysis. As Endres explained, Emerick’s statistical comparison—that he and Student B had answered 84% of the questions identically— proved no misconduct. Of the 160 HDS students who took the exam, more than 80 students had 85% similar answers with at least one other student in the class. And more than two-thirds of the class had six or more identical wrong answers with at least one other student in the class. According to Endres, that data refuted the Thewissen Analysis’s claim that the probability two students would mistakenly answer the same six questions was 0.000036%.

The next day, Endres received a letter from the ERC granting him a rehearing before CAPP but restricting the information he could present. While Endres could introduce Trauben’s report, he could not submit any evidence contesting the Thewissen Analysis.

The Second CAPP Hearing. Knowing that CAPP would not consider his statistical analysis, Endres sent test data to Thewissen and asked him to defend his findings given the full class data. Thewissen declined to review the data, and NEOMED’s legal counsel later informed Endres that Thewissen was not open to making any statement about potential errors in his analysis. That said, NEOMED’s legal counsel told Endres that the CAPP committee had agreed to allow Trauben to speak at the hearing—and that the committee members would have some questions for the physician.

Several hours before the second CAPP hearing on November 18, NEOMED’s legal counsel issued a limiting instruction to the CAPP: No. 18-3825 Endres v. Ne. Ohio Med. Univ., et al. Page 10

Please be advised that during our deliberations today, the Committee should accord no weight to any statement made or documents presented during the initial hearing in [Endres’s] CAPP case that recounted any statistical analysis performed on his test answers. Consistent with disregarding any statements made or documents presented about any statistical analysis, the committee is to likewise accord no weight to any documents set forth by [Endres] throughout his appeal that include any statistical analysis of his test answers. While these documents were included in the record received by the Committee in advance of today’s meeting, they should not be considered during our review of this matter today, or in making a decision.

(R. 23, First Am. Compl. at ¶ 157.) This meant that Endres could not discuss the Thewissen Analysis or distribute any material about that analysis at the hearing. Nor could Endres introduce recorded video from a field test, which he alleges proved that it was physically impossible to view content on a neighboring laptop.

The second CAPP hearing proceeded much like the first hearing. Emerick presented her case, after which Endres entered the room to present his version of events. Endres alleges that CAPP only allowed him to speak generally and that he could not present, distribute, or discuss any material that would have exonerated him, such as a video recording from a field test he conducted. According to Endres, the recording would have proven that it was physically impossible for any student to view content on a neighboring laptop. CAPP then called Trauben, who reiterated that Endres’s behavior during the test was a manifestation of his ADHD. Endres alleges that the committee asked Trauben some questions about his credentials and qualifications to write the report he submitted on behalf of Endres. The next day—November 19, 2015— Endres learned that the committee had voted to dismiss him.

The Emerick Memo. Several days after the second CAPP hearing, Endres obtained a hard copy of his student file, which he alleges contained information he had not reviewed before that hearing. Endres reviewed the file on November 11, when it was supposed to include all evidence and information that Emerick would present at the second CAPP hearing. But when he reviewed the file after that hearing, it contained new evidence: a memo from Emerick (“Emerick Memo”), dated November 2. No. 18-3825 Endres v. Ne. Ohio Med. Univ., et al. Page 11

Endres alleges that the Emerick Memo inaccurately summarizes Emerick’s conversations with two of his physicians, Ramos and Trauben. Indeed, Ramos and Trauben submitted affidavits stating that the Emerick Memo did not faithfully memorialize their conversations with Emerick and that its content was either false, misleading, or taken out of context. For example, the Emerick Memo states that Trauben “understood that he needed a quick turn around on documentation in order to submit a CAPP appeal letter.” (Id. at ¶ 171.) But Trauben explained in his affidavit that “[t]hat sentence appears to suggest that my analysis of Mr. [Endres], or the opinion I rendered, was compromised because of the timeframe in which it was completed. That is not so, and it is not something I ever conveyed to [Defendant Emerick.]” (Id.) Trauben concluded that Emerick “did not faithfully memorialize the contents of our telephone call.” (Id. at ¶ 173.)

Ramos recounted similar inaccuracies and misrepresentations in the Emerick Memo. The Emerick Memo states that “Dr. Ramos only met with [J. Endres] three times in total” and that Ramos did not meet with Endres before submitting her October 13th letter documenting her treatment of Endres. (Id. at ¶ 174.) Ramos objects to both statements, noting that she met with Endres three times within a four-month period in 2015, including an August meeting about a month before Endres took the HDS test. Ramos explained that she met with Endres more often than she does with most of her patients. The Emerick Memo also alleges that Ramos “simply prepared the letter based upon the materials [Plaintiff’s] parents sent her via jump drive regarding [Plaintiff’s] case.” (Id. at ¶ 176.) Ramos described that statement as “false,” noting that it “suggests that the views I expressed in my letter were somehow biased by information I received from Mr. [Endres’s] parents.” (Id.) Finally, Endres accuses Emerick of falsely characterizing his feelings toward his physicians and his medical care. Emerick allegedly told Ramos that Endres blamed her for the side effects he experienced on Strattera. Yet Endres alleges that he never told Emerick that Ramos was responsible and maintains that he—not his physicians—wanted to try a different medicine. Endres hypothesizes that Emerick mischaracterized his feelings to elicit a response from Ramos that would damage his credibility before CAPP. No. 18-3825 Endres v. Ne. Ohio Med. Univ., et al. Page 12

Endres filed his two-count complaint in federal court on November 16, 2017, naming NEOMED, NEOMED president Jay Gershen, NEOMED’s board of trustees, and Emerick as defendants (“Defendants”). Endres’s first count, under 42 U.S.C. § 1983, alleges that Gershen, Emerick, and the board of trustees violated his constitutional rights by depriving him of his property and liberty interests in enrollment at NEOMED without procedural due process. His second count, under the Americans with Disabilities Act (“ADA”) and the Rehabilitation Act of 1973, alleges that NEOMED effectively concluded that his ADHD precluded him from completing his medical studies—even though NEOMED could have made reasonable accommodations to ensure that his ADHD-related symptoms would not create the appearance of academic misconduct.

Defendants moved to dismiss the complaint on statute of limitation grounds, contending that Endres’s two-year window to file his complaint started in October 2015, when CAPP first concluded that Endres should be dismissed. Emerick also claimed qualified immunity. The district court ruled for the Defendants, holding that the two-year statute of limitations barred Endres’s suit and that in any case, Emerick is entitled to qualified immunity. Endres appeals that decision.

II.

We review de novo the district court’s dismissal of a complaint on statute of limitations grounds. Banks v. City of Whitehall, 344 F.3d 550, 553 (6th Cir. 2003). Neither § 1983, the ADA, nor the Rehabilitation Act provides a statute of limitations, so courts must borrow one from the most analogous state cause of action. McCormick v. Miami Univ., 693 F.3d 654, 663 (6th Cir. 2012). This court has applied Ohio’s two-year statute of limitations governing personal injury actions to Ohio cases arising under § 1983, the ADA, and the Rehabilitation Act. Banks, 344 F.3d at 553 (applying two-year period to § 1983 suit); McCormick, 693 F.3d at 663 (applying two-year period to ADA suit); Bishop v. Children’s Ctr. for Dev. Enrichment, 618 F.3d 533, 536 (6th Cir. 2010) (applying two-year period to Rehabilitation Act suit). This appeal, however, is not about the length of the limitations period, which both parties agree is two years. Rather, it turns on when the limitations period starts to run, and federal law holds the answer to that question. See Wallace v. Kato, 549 U.S. 384, 388 (2007). Under federal law, the No. 18-3825 Endres v. Ne. Ohio Med. Univ., et al. Page 13

statute of limitations period begins “when the plaintiff knows or has reason to know of the injury which is the basis of his action.” Sevier v. Turner, 742 F.2d 262, 273 (6th Cir. 1984) (citations omitted). Therein lies the heart of this dispute.

Endres and Defendants agree that the injury triggering this suit is Endres’s dismissal from NEOMED. But the disputed question is when Endres learned that he had no future at NEOMED. According to Endres, the limitations period began on November 19, 2015, when he learned the outcome of the second CAPP hearing. If Endres is correct, then his claim is timely because he filed his complaint on November 16, 2017, within the two-year statute of limitations. But Defendants say the limitations period began on October 22, 2015, when Endres learned the outcome of the first CAPP hearing, and if they are correct, Endres’s claim is time-barred.

Both parties point us to Delaware State College v. Ricks, in which the Supreme Court held that a college professor’s employment discrimination claim was time-barred. 449 U.S. 250, 254 (1980). The case warrants close study. After the plaintiff, Ricks, had been teaching at Delaware State College for about three years, the college’s tenure committee recommended in February 1973 that he not receive tenure. Id. at 252. That committee agreed to reconsider its decision in a year, and it did so in February 1974, adhering to its original recommendation. Id. Next, the faculty senate voted to approve the committee’s recommendation, and in March 1974, the board of trustees voted to deny Ricks tenure. Id. Ricks then filed a grievance with the board’s educational policy committee, and in May 1974, that committee took the grievance under consideration. Id. Although the college denied him tenure, Ricks was not immediately jobless. The college had a policy of providing junior faculty members who had been denied tenure with a one-year terminal contract, and Ricks received one of those contracts on June 26, 1974. Id. at 253–54. The board informed Ricks on September 12, 1974, that it had denied his grievance, so Ricks left the college when his contract expired on June 30, 1975.

Ricks turned to federal court for relief, filing his employment discrimination claim on September 9, 1977. Id. at 254. Up against a statute of limitations defense, Ricks argued that the statute of limitations period commenced on June 30, 1975, when his contract with the college ended. The Court rejected that argument, noting that the “termination of employment at Delaware State is a delayed, but inevitable, consequence of the denial of tenure.” Id. at 257–58. No. 18-3825 Endres v. Ne. Ohio Med. Univ., et al. Page 14

According to the Court, the limitations period began when the college communicated to Ricks its final, official decision to deny him tenure. Id. at 259. That happened when the college offered Ricks a one-year terminal contract on June 26, 1974, by which point “the tenure committee had twice recommended that Ricks not receive tenure; the Faculty Senate had voted to support the tenure committee’s recommendation; and the Board of Trustees formally had voted to deny Ricks tenure.” Ricks, 499 U.S. at 262. In short, the college “had established its official position—and made that position apparent to Ricks.” Id.

Defendants contend that CAPP’s first decision to dismiss Endres commenced the statute of limitations, calling CAPP’s October 22 letter an “unequivocal” signal that Endres’s “time at NEOMED had come to an end.” (Appellee Br. at 24.) According to Defendants, Endres “must have known” that the first CAPP decision “constituted a rejection of his disability-based assertions,” thus placing him on notice of his impending dismissal and triggering the limitations period. (Id.) But for that argument to hold any weight, we would have to ignore the existence of NEOMED’s appeals procedure. To be sure, NEOMED students are not entitled to an appeal as a matter of course. Only if a student sets forth “significant and compelling new information that was not available for presentation to the CAPP [ ] Committee, or evidence of a defect or irregularity in the CAPP [ ] Committee proceeding” will the ERC consider the appeal. (R. 25, Mot. to Dismiss Appx. at PageID #690.) But Endres alleged precisely that—both that Emerick presented evidence, including the Thewissen Analysis, that he never had a chance to review before the first CAPP hearing (a procedural defect) and that his statistical analysis undermined the accuracy of the Thewissen Analysis (significant and compelling new information that he could have not presented until learning that the Thewissen Analysis existed). Far from “establish[ing] its official position—and ma[king] that position apparent to [Endres],” the ERC remanded the matter to CAPP, giving Endres a second chance to defend himself against allegations of misconduct. Ricks, 499 U.S. at 262. NEOMED did not reach a final decision— and communicate that decision to Endres—until it informed him on November 19 that the second CAPP panel voted to dismiss him, at which point Endres had exhausted the appeals process. No. 18-3825 Endres v. Ne. Ohio Med. Univ., et al. Page 15

Citing Ricks, Defendants also contend that NEOMED’s appeals process is analogous to a collateral attack on CAPP’s decision, in which case Endres’s appeal would not have tolled the statute of limitations. In Ricks, the E.E.O.C. filed an amicus brief, arguing that the statute of limitations did not start until the college notified Ricks that it denied the grievance he filed after the board of trustees voted to deny him tenure. Id. at 260. The Court rejected that argument, drawing a critical distinction between collateral and direct review of a hiring decision. Although a meritorious grievance may well change the result of an adverse hiring decision, a grievance is a “remedy for a prior decision, not an opportunity to influence that decision before it is made.” Id. at 261 (emphasis in original). Thus, a pending grievance or “some other method of collateral review of an employment decision, does not toll the running of the limitations period.” Id. (citing Elec. Workers v. Robbins & Myers, Inc., 429 U.S. 229 (1976)). But Ricks also suggests that the statute of limitations will not begin until the institution conveys its final decision to the plaintiff. There, the Court did not select February 1973—when the college’s tenure committee first recommended denying Ricks tenure—as the accrual date for Ricks’s claim. Not until the tenure committee reevaluated that decision one year later, the faculty senate approved that decision, the board of trustees voted to deny Ricks tenure, and the college extended Ricks a one- year terminal contract did Ricks’s claim accrue.1

Although Defendants are correct that a collateral challenge to a final decision will not toll the statute of limitations, their argument misses the mark because Endres’s appeal was not collateral. A collateral challenge necessarily implies the existence of a final decision, but NEOMED’s own procedures tell us that the appeals process postpones the finality of a CAPP decision. For one, the student has four working days to appeal an adverse CAPP decision to the ERC. Failure to file an appeal within that timeline amounts to a “waiver of the right to appeal, and the decision becomes final.” (R. 25, Mot. to Dismiss Appx. at PageID #691.) That of course implies that CAPP’s decision does not become final until—at the earliest—four working days

1In a footnote, the Court hinted that the accrual date may have occurred in March 1974, when the college’s board of trustees voted to deny Ricks tenure, even though the district court held that Ricks’s claim accrued when the college extended him a one-year terminal contract in June 1974. Ricks, 449 U.S. at 262 n.17. Ricks filed his grievance against the college after the board’s vote—but before the college offered the contract. In any event, the Court concluded that even if the claim accrued in June 1974, Ricks’s claim was still untimely, noting that “[w]e cannot say that [the district court’s] decision was erroneous.” Id. at 262. No. 18-3825 Endres v. Ne. Ohio Med. Univ., et al. Page 16

after its issuance. Second, NEOMED’s procedures explain that if the ERC refuses to grant the student’s appeal, the CAPP decision becomes final. Again, that implies that the CAPP decision is not final while an appeal request is pending before the ERC or after the ERC grants the appeal and remands the matter to CAPP for reconsideration, as it did here. And when the ERC remands the matter for reconsideration, it effectively gives the accused student another shot at proving his innocence. That is, the student gets another “opportunity to influence [CAPP’s] decision before it is made.” Ricks, 449 U.S. at 261 (emphasis in original).

Defendants’ arguments fare no better under this court’s precedents. Defendants draw our attention to Janikowski v. Bendix Corp., in which we held that the plaintiff’s repeated requests for relief from an adverse decision did not toll the statute of limitations period for his employment discrimination claim. 823 F.2d 945, 948 (6th Cir. 1987). We find that comparison inapt. In Janikowski, the plaintiff’s employer announced a workforce reduction plan and informed the plaintiff on September 4, 1980, that his position would end on September 30, 1981. Id. at 946. Following up in November 1980, the plaintiff’s employer sent him a memo, reminding him that his position would no longer exist on September 30, 1981, and warning him that if he did not find a different position within the company, his employment with the company would also end. Id. The employer gave the plaintiff time off to interview, but as September 1981 approached, the plaintiff had still not found employment, either within the company or elsewhere. Id. So the plaintiff negotiated to postpone his termination, and the employer agreed to retain him on a month-to-month basis—but ultimately terminated him on November 30, 1981. Id.

We agreed with the district court that the three-year statute of limitations barred the plaintiff’s suit, which he filed more than three years after his claim accrued. We found that the accrual date was September 4, 1980, when the plaintiff received “notice of termination, not when his employment actually cease[d].” Id. True, the employer accommodated the plaintiff, even after the September 30, 1981, deadline had passed, but there was never any question that the employer would terminate the plaintiff’s position. Although the plaintiff sought more time to “avoid the consequences of the termination,” he knew on September 4, 1980, that his position would not exist in about a year. Accordingly, we repeated the Supreme Court’s remark that in a No. 18-3825 Endres v. Ne. Ohio Med. Univ., et al. Page 17

statute of limitations analysis, “the proper focus is on the time of the discriminatory act, not the point at which the consequences of the act become painful.” Id. at 947 (quoting Chardon v. Fernandez, 454 U.S. 6, 8 (1981)) (emphasis in original).

Under Janikowski, Endres could not argue (nor does he) that his claim accrued when NEOMED made him clear out his dorm room and leave campus, even if the consequences of NEOMED’s decision were most painful then. Rather, our statute of limitations analysis focuses on the alleged wrong giving rise to the claim—here, NEOMED’s dismissal decision. Though it was always a foregone conclusion that the employer in Janikowski would terminate the employee’s position, Endres’s fate remained uncertain until the second CAPP panel voted to dismiss him. Only then—by which point he had exhausted the appeals process—was there no doubt that Endres would have to leave NEOMED.

As a last resort, Defendants argue it makes no difference that NEOMED provides an appeals process because federal law controls when a claim accrues and the NEOMED handbook, of course, is not federal law. Thus, Defendants claim that NEOMED’s procedures do not inform our analysis. In support of that claim, Defendants cite Haeberle v. University of Louisville, in which the plaintiff argued that because the University of Louisville is a “creature of the state of Kentucky, it is therefore subject to the interpretations of the Kentucky Supreme Court and the Kentucky legislature with respect to the finality of state entity decisions.” 90 F. App’x 895, 900 (6th Cir. 2004) (internal quotation marks omitted). We rejected that view, explaining that federal law governs when the limitations period begins. Haeberle, however, is inapposite. Although the plaintiff in Haeberle asked this court to apply Kentucky law—not federal law—to determine when the statute of limitations commences, Endres makes no such request. Indeed, Endres does not challenge our principle that the statute of limitations starts “when the plaintiff knows or has reason to know of the injury which is the basis of his action.” Sevier, 742 F.2d at 273. That principle—one of federal law—does not keep us from considering how NEOMED’s policies may have informed Endres about his injury. That is not to say that the policies themselves have independent legal significance or that we are substituting the policies in place of our law. Rather, the policies help us determine when Endres knew that his time at NEOMED had ended. No. 18-3825 Endres v. Ne. Ohio Med. Univ., et al. Page 18

They reveal that until CAPP’s decision became final, Endres did not know for certain that he had no future at NEOMED.

For these reasons, we conclude that the statute of limitations did not start until November 19, 2015, when Endres learned that the second CAPP panel issued a final, non-appealable decision recommending his dismissal. Having filed his complaint within two years of that date, Endres has satisfied the statute of limitations.

III.

Because Endres’s suit is timely, his claim under the ADA and the Rehabilitation Act may go forward. But Endres also brings a claim under 42 U.S.C. § 1983, alleging that Sandra Emerick violated his Fourteenth Amendment rights by depriving him of his liberty and property interests in continued enrollment at NEOMED without procedural due process.2 Emerick, a state official named in her official and individual capacity, argued below that she is entitled to qualified immunity on Endres’s § 1983 claim. Although the district court held that Endres’s entire suit was time-barred, it also held in the alternative that Emerick is entitled to qualified immunity. Endres appeals that decision, which we review de novo. O’Brien v. City of Grand Rapids, 23 F.3d 990, 998 (6th Cir. 1994).

In any qualified immunity analysis, we must answer two questions: (1) was there a constitutional violation; and (2) if so, was the right clearly established when the violation occurred? Everson v. Leis, 556 F.3d 484, 494 (6th Cir. 2009). We consider each question in turn.

A.

At the outset, Endres must show a constitutionally protected life, liberty, or property interest to prevail on his procedural due process claim. Endres alleges both a property and liberty interest in his continued enrollment at NEOMED, and our case law supports—at the very

2Endres sued NEOMED president Jay Gershen and NEOMED’s board of trustees in their official capacities and Emerick in her official and individual capacity. As the district court noted, state defendants acting in their official capacities are not “persons” under § 1983 with respect to claims for monetary damages. See, e.g., Gean v. Hattaway, 330 F.3d 758, 766 (6th Cir. 2003). Thus, Endres cannot seek money damages under § 1983 from Gershen or NEOMED’s board of trustees. No. 18-3825 Endres v. Ne. Ohio Med. Univ., et al. Page 19

least—his alleged property interest. In Doe v. University of Cincinnati, we explained that a university’s decision to suspend the student for two years “clearly implicates a protected property interest” in the student’s continued enrollment at the university. 872 F.3d 393, 399 (6th Cir. 2017) (internal quotation marks omitted). If a two-year suspension is sufficient, then surely permanent dismissal is sufficient, too—and Emerick does not argue otherwise. Indeed, we have repeatedly held that a state-university student facing a significant disciplinary decision, such as expulsion, is entitled to “minimum due process protections.” Id.; Flaim v. Med. Coll. of Ohio, 418 F.3d 629, 633 (6th Cir. 2005).

So how much process is enough? To be sure, universities are not courts, and an accused student cannot expect the same procedural safeguards as a criminal defendant. Flaim, 418 F.3d at 635 n.1. Still, our cases tell us there must be some process; how much depends on the basis of the dismissal. The Supreme Court has recognized a critical distinction between dismissals for disciplinary misconduct and dismissals for academic underperformance. Compare Goss v. Lopez, 419 U.S. 565 (1975), with Bd. of Curators of Univ. of Mo. v. Horowitz, 435 U.S. 78 (1978). When a school imposes a serious sanction like dismissal to address a student’s disciplinary misconduct, the student is entitled to more robust process, including a hearing to “present his side of the story.” Goss, 419 U.S. at 581. In Goss, one of the named plaintiffs, a public high school student, allegedly participated in a lunchroom melee that destroyed some school property, leaving high school administrators to resolve a binary, factual question: did the student engage in that misconduct? Id. at 569 n.4. The administrators answered in the affirmative and suspended the student without conducting a hearing, which, according to the Court, violated the student’s procedural due process rights. The Court held that when a student accused of misconduct faces “total exclusion from the educational process for more than a trivial period,” he is entitled to an “opportunity to present his side of the story.” Id. at 576, 581. Even if the evidence looks stacked against the student, “things are not always as they seem to be.” Id. at 584. Thus, hearings “provide a meaningful hedge against erroneous action” by, at the very least, allowing the student to alert administrators to factual disputes and breaks in causal chains. Id. at 583. No. 18-3825 Endres v. Ne. Ohio Med. Univ., et al. Page 20

Academic inquiries, not always so cut-and-dry, are a different matter. In Horowitz, the Court held that a medical school did not violate a student’s due process rights when it dismissed her without a hearing because the basis for that dismissal was academic. 435 U.S. at 89–90. In that case, the medical school faculty reported concerns with the student’s clinical performance, not to mention that the student’s attendance at clinical sessions was “erratic” and that the student “lacked a critical concern for personal hygiene.” Id. at 81. The medical school did not summarily dismiss the student upon hearing these complaints but instead allowed her to spend time with seven physicians, who would evaluate her performance and recommend whether she should graduate on time. Id. Of those seven, two recommended that the medical school dismiss the student immediately; three recommended that the school postpone her graduation. Id. The medical school ultimately dismissed the student, id. at 82, so she sued, alleging that the medical school violated her due process rights. Id. at 91.

The Court, however, found no constitutional violation. Unlike the factual question in Goss—i.e., did the student engage in misconduct—the question that the medical school administrators had to resolve in Horowitz was far from objective. As the Court explained, measuring a student’s academic performance “by its nature [is] more subjective and evaluative than the typical factual questions presented in the average disciplinary decision” and often requires “an expert evaluation of cumulative information.” Id. at 90. It is also less contentious. Although a student with poor grades may face the same consequences (e.g., dismissal) as the student accused of disciplinary misconduct, the educational process itself “is not by nature adversar[ial]” and therefore not as “readily adapted to the procedural tools of judicial or administrative decisionmaking.” Id. Finally, in an academic context, the decision-makers are not drawn by lot to find facts, like we ask laymen to do as members of a jury. Rather, they exercise their “historic judgment” as educators to evaluate the student, and their decision is entitled to substantial deference. Id.

In the end, the Court declined to “formalize the academic dismissal process by requiring a hearing.” Id. So long as the university’s evaluation and ultimate decision is “careful and deliberate,” the university meets the Due Process Clause’s requirements. Id. at 85. The upshot for students facing dismissal on academic grounds is that “the due process afforded is minimal.” No. 18-3825 Endres v. Ne. Ohio Med. Univ., et al. Page 21

Yoder v. Univ. of Louisville, 526 F. App’x 537, 549 (6th Cir. 2013) (citing Horowitz, 435 U.S. at 85; Ku v. Tennessee, 322 F.3d 431, 436 (6th Cir. 2003)). The university must provide the student with notice of his unsatisfactory academic performance and deliver a “careful and deliberate” decision on the student’s fate. Ku, 322 F.3d at 436 (citing Horowitz, 435 U.S. at 85– 86). But the university need not provide a hearing.

Whether Endres can show a constitutional violation turns on the basis for his dismissal. Emerick claims that the dismissal was for academic reasons, meaning that NEOMED did not even need to provide Endres with a hearing. Indeed, Emerick contends that Endres “received at least as much process as was required for an academic dismissal.” (Appellee Br. at 32.) Endres, by contrast, alleges that allegations of cheating fall squarely on the disciplinary side of the disciplinary-academic divide, triggering more robust procedures under the Due Process Clause.

As Emerick acknowledges, the boundary between disciplinary misconduct and unsatisfactory performance can be hazy, but she offers two signs of when a dismissal rests on academic grounds: (1) the student’s conduct relates to the student’s prospects for success in his chosen field; and (2) the university labels the student’s conduct as “academic” in its standing procedures. We consider each in turn.

Emerick is correct that grades, by themselves, do not dictate a student’s prospects for success in his chosen field and that the university may consider more than just grades to assess a student’s academic standing. Following Horowitz, this court has endorsed a broad view of what constitutes academic conduct. In Al-Dabagh v. Case Western Reserve University, we did not second-guess the medical school’s decision to treat a student’s lack of professionalism as an academic matter. 777 F.3d 355, 357 (6th Cir. 2015). Although he performed well academically, the student arrived habitually late to important meetings, displayed a poor bedside manner around patients, and harassed two female students at a school dance. Id. at 357–58. That conduct prompted the medical school to contemplate dismissing the student, although the school ultimately settled on some lesser sanctions. Id. at 358. But the shoe finally dropped after a North Carolina court convicted the student for drunk driving, at which point the medical school No. 18-3825 Endres v. Ne. Ohio Med. Univ., et al. Page 22

convened a committee and refused to certify him for graduation. 3 Id. The student sued, and although he did not allege any due process deprivation, we said that his “dismissal on professionalism grounds amounts to a deference-receiving academic judgment.” Id. at 359. Noting that professionalism “has been a part of the doctor’s role since at least ancient Greece,” we explained that academic evaluations “may permissibly extend beyond raw grades and other objective criteria.” Id. at 360 (internal alterations, quotation marks, and citations omitted).

Medical schools have no interest in credentialing students with dim prospects in the medical profession, be that because of low aptitude, poor hygiene, or drunkenness. And to the extent that the medical profession demands certain traits in physicians, medical schools have good reason to ensure that their students possess those traits. Even so, it cannot be the case that because the alleged misconduct somehow relates to a professional trait, the medical school need only treat the matter as academic and provide the student with minimal process. If that were so, the medical school could reasonably construe all types of misconduct as a sign of the student’s lack-of-professionalism and thus avoid providing the student with the heighted procedures that the Due Process Clause may demand. To give an example, a medical school could treat allegations of physical assault as academic, given that violent medical students might be likely to disobey the commandment that as physicians, they first do no harm. By treating those allegations as academic rather than disciplinary, the university could dismiss the accused student without providing him the opportunity to contest facts or present evidence of his innocence, even though this is exactly the type of situation that Goss says warrants a hearing. Such a broad conception of academic conduct would swallow all forms of conduct, sweeping Goss into oblivion and obviating the need to ever hold a disciplinary hearing. For these reasons, the relationship between the conduct and the student’s future professional success cannot be what distinguishes academic and disciplinary conduct.

Emerick points to a second sign that the conduct falls on the academic side of the disciplinary-academic divide: the label that the medical school affixes to the alleged misconduct.

3After the student sued, the district court ordered the medical school to grant his diploma. Al-Dabagh, 777 F.3d at 358–59. The medical school did so—but then appealed to this court, and we reversed the district court’s order. Id. at 361. No. 18-3825 Endres v. Ne. Ohio Med. Univ., et al. Page 23

According to Emerick, the student handbook defines cheating as a form of academic misconduct, which supports the inference that the basis for Endres’s dismissal was academic. Endres, however, correctly identifies two problems with that inference—one specific to NEOMED’s polices, the other more universal. First, NEOMED’s student handbook does not state that cheating is necessarily academic. The student handbook categorizes cheating as a form of “academic misconduct,” which itself is a type of “student misconduct.” (R. 25, Mot. to Dismiss Appx. at PageID #618.) But NEOMED administrators have discretion over how to adjudicate allegations of student misconduct. As the student handbook explains, NEOMED’s chief officer of student affairs—here, Emerick—must consult with other administrators and decide whether the allegations are “best resolved through the formal disciplinary process . . . or by way of a referral to Committee on Academic and Professional Progress (CAPP) or some other mechanism such as counseling or mediation.” (Id. at PageID #613.) That is, NEOMED—acting through Emerick—gets to decide whether allegations of cheating merit a disciplinary hearing or some other type of adjudication with less rigorous procedures. Nothing in the student handbook, however, suggests that cheating necessarily falls on the academic side of the disciplinary- academic divide.

The second problem with Emerick’s argument is more general. Whether the university describes conduct as academic or disciplinary does not dictate what process the Constitution demands. As a default, it is the Constitution, not a university handbook, that establishes what process is due: the Constitution sets “the floor or lowest level of procedures acceptable.” Flaim, 418 F.3d at 636; see also Heyne v. Metro. Nashville Pub. Schs., 655 F.3d 566, 569 (6th Cir. 2011). Of course, parties may agree by contract to waive their constitutional rights, such as when a student agrees to warrantless searches of his dormitory room as a condition of living in a university building. See, e.g., Medlock v. Trs. of Indiana Univ., 738 F.3d 867, 872 (7th Cir. 2013). Such contracts, however, require “clear and unmistakable language” for us to infer that one party has agreed to waive his constitutional rights—language not present in NEOMED’s student handbook. Morrison v. Warren, 375 F.3d 468, 474 (6th Cir. 2004).

There is a more straightforward way to tell when the university’s decision is disciplinary or academic. A decision is disciplinary when the university engages in first-level factfinding to No. 18-3825 Endres v. Ne. Ohio Med. Univ., et al. Page 24

resolve a disputed, objective question about the student’s conduct—and the outcome of that inquiry could lead to the student’s dismissal or a long suspension. Our method of distinguishing academic from disciplinary decisions is not something of our own creation. Rather, it tracks the distinction that the Supreme Court has drawn through Goss and Horowitz—and that courts have followed. See, e.g., Fenje v. Feld, 398 F.3d 620, 625 (7th Cir. 2005) (describing disciplinary dismissals as “being more objective in nature and not dependent upon the analytical expertise of professional academicians”); Pugel v. Bd. of Tr. of Univ. of Ill., 378 F.3d 659, 663 (7th Cir. 2004) (“Courts addressing graduate student dismissals on charges of academic dishonesty traditionally have relied upon Goss.”).

We underscore the critical difference between engaging in first-level factfinding and drawing subjective conclusions from established facts. Al-Dabagh draws out this distinction. In that case, the university refused to certify the student for graduation and effectively dismissed him for lack of professionalism—an academic decision—when it learned of his North Carolina conviction for drunk driving. Al-Dabagh, 777 F.3d at 358–59. Although drunk driving might seem like inherently disciplinary conduct, the university was not adjudicating whether the student had, in fact, driven while intoxicated. Rather, the university had to decide, given the court’s decision, whether the student possessed the necessary traits to succeed in the medical profession. That is a purely subjective inquiry, analogous to the university’s decision in Horowitz, that does not warrant a hearing under the Due Process Clause.

The same goes when the university evaluates a student’s academic performance. It is true that this type of evaluation “also involves [questions] of objectively determinable fact—e.g., whether the student gave certain answers on an examination.” Horowitz, 435 U.S. at 95 n.5 (Powell, J., concurring). But the purpose of the inquiry is not to answer an objective question about the student’s conduct. Rather, the university’s “critical decision requires a subjective, expert evaluation as to whether [the student’s] performance satisfies some predetermined standard of academic competence,” which itself “is set by a similarly expert judgment.” Id. This contrasts with a disciplinary inquiry, which “requires a factual determination as to whether the conduct took place or not” and thus warrants more rigorous protections under the Due Process Clause. Id. No. 18-3825 Endres v. Ne. Ohio Med. Univ., et al. Page 25

What does this all mean for Endres? Emerick referred the cheating allegations to CAPP, which then resolved a disputed, objective question: whether Endres cheated on the September 28 test. In answering that question, CAPP engaged in first-level factfinding and ultimately dismissed Endres, no doubt a serious sanction. The basis for Endres’s dismissal was disciplinary, calling for more rigorous procedures under the Due Process Clause.

The next question is whether the process Endres received was sufficient. When a university student faces a serious sanction like dismissal over allegations of disciplinary misconduct, he is entitled to a “fundamentally fair hearing.” Flaim, 418 F.3d at 635 n.1 (citation omitted). Endres received a hearing. But his allegations, which we must take as true at this stage, reveal that hearing was far from fair. For one, the student has a “right to be present for all significant portions of the hearing,” provided the hearing is live. Flaim, 418 F.3d at 635. And even when the hearing is not live, the university must “provide the accused with the opportunity to ‘respond, explain, and defend.’” Id. (quoting Gorman v. Univ. of R.I., 837 F.2d 7, 13 (1st Cir. 1988)). Endres, however, alleges he was not allowed in the room while Emerick presented her case to the CAPP panels. That alone establishes a due process violation, but Endres’s allegations do not end there.

Doe also says that the university must provide the student with “an explanation of the evidence” against him, but Endres’s allegations show that NEOMED repeatedly failed on this front. 872 F.3d at 399–400. According to Endres, Emerick presented a statistical analysis to the first CAPP panel showing that the odds he and Student B would have six identical wrong answers was 0.000036%. But that evidence—the Thewissen Analysis—was not present when Endres reviewed his student file before the hearing. Moreover, NEOMED did not allow Endres to listen in when Emerick presented the Thewissen Analysis and the rest of the case against him. Without knowing that the Thewissen Analysis even existed, Endres could not have responded to what seemed like incontrovertible proof that he was a cheater.

Even if the second CAPP hearing cured the defects of the first hearing, the second hearing was not flawless. Again, NEOMED excluded Endres from the room while Emerick presented her case. And once more, NEOMED failed to inform Endres of the key evidence against him. After the second CAPP panel dismissed him, Endres reviewed his student file and No. 18-3825 Endres v. Ne. Ohio Med. Univ., et al. Page 26

discovered a memo Emerick had drafted, which summarized her conversations with two of his physicians. Endres not only alleges that Emerick discussed the memo as she presented her case to the second CAPP committee—but also that its content was false. Both physicians provided sworn affidavits in which they disputed Emerick’s account of their conversations, but those affidavits were of no use once the second CAPP panel voted to dismiss Endres, marking the end of the appeals process. Without knowledge of the memo’s existence, Endres could not “present his side of the story.” Doe, 872 F.3d at 400 (citation omitted).

B.

Endres has alleged more than enough to establish a due process violation, but that does not end the matter. Because Emerick has claimed qualified immunity, Endres must also show that the constitutional rights Emerick violated were clearly established when the violation occurred. See DiLuzio v. Vill. of Yorkville, 796 F.3d 604, 608–09 (6th Cir. 2015). “For a right to be clearly established, the contours of the right must be sufficiently clear that a reasonable officer would understand that what he is doing violates that right.” McGlone v. Bell, 681 F.3d 718, 735 (6th Cir. 2012).

Emerick alleges that the law defining Endres’s due process rights was not clearly established, and on this front, she is correct. To be sure, the Supreme Court’s decisions in Goss and Horowitz make clear that a student facing a serious sanction for disciplinary misconduct is entitled to a fair hearing, but neither those cases nor our own decisions have articulated a bright- line rule to distinguish academic from disciplinary matters. Moreover, clearly established law “must be ‘particularized’ to the facts of the case,” yet no case from the Supreme Court or this court has held that cheating is a disciplinary matter warranting more robust procedures under the Due Process Clause. White v. Pauly, 137 S. Ct. 548, 552 (2017) (per curiam) (quoting Anderson v. Creighton, 483 U.S. 635, 640 (1987)). And because no precedent clearly established that Endres was even entitled to a hearing, it follows that his right to be present at the hearing and to hear the evidence against him was not clearly established, either. We therefore hold that Emerick is entitled to qualified immunity. No. 18-3825 Endres v. Ne. Ohio Med. Univ., et al. Page 27

We note, however, that qualified immunity “‘only immunizes defendants from monetary damages’—not injunctive or declaratory relief.” Kanuszewski v. Mich. Dep’t of Health and Human Servs., 927 F.3d 396, 417–18 (6th Cir. 2019) (quoting Williams v. Com. of Ky., 24 F.3d 1526, 1541 (6th Cir. 1994)); see also Ward v. Polite, 667 F.3d 727, 742 (6th Cir. 2012). Thus, our ruling shields Emerick from monetary damages. But the qualified immunity doctrine does not preclude Endres from continuing to pursue the injunctive and declaratory relief that he has also requested in his § 1983 claim.

IV.

We REVERSE the district court’s dismissal of Endres’s complaint and AFFIRM the district court’s alternative holding that Emerick is immune from monetary damages on Endres’s § 1983 claim.