FILED
                                                                    United States Court of Appeals
                                       PUBLISH                              Tenth Circuit

                      UNITED STATES COURT OF APPEALS                       March 9, 2020
                                                                       Christopher M. Wolpert
                             FOR THE TENTH CIRCUIT                         Clerk of Court
                         _________________________________

 JOHN DOE,

       Plaintiff - Appellant,

 v.                                                         No. 18-1162

 UNIVERSITY OF DENVER;
 UNIVERSITY OF DENVER BOARD OF
 TRUSTEES; REBECCA CHOPP,
 individually and as agent for University of
 Denver; KRISTIN OLSON, individually
 and as agent for University of Denver;
 JEAN MCALLISTER, individually and as
 agent for University of Denver;
 KATHRYNE GROVE, individually and as
 agent for University of Denver; ERIC
 BUTLER, individually and as agent for
 University of Denver,

       Defendants - Appellees.
                      _________________________________

           APPEAL FROM THE UNITED STATES DISTRICT COURT
                   FOR THE DISTRICT OF COLORADO
                     (D.C. No. 1:16-CV-00152-PAB-STV)
                    _________________________________

Philip A. Byler of Nesenoff & Miltenberg, LLP, New York, New York (Andrew T.
Miltenberg, Stuart Bernstein, Tara J. Davis, and Jeffrey Berkowitz of Nesenoff &
Miltenberg, LLP, New York, New York, and Michael J. Mirabella and Patricia Mellen of
Michael J. Mirabella, P.C., Denver, Colorado, with him on the briefs), for Plaintiff -
Appellant.

Jim Goh (E. Rayner Mangum with him on the brief), Constangy, Brooks, Smith &
Prophete, LLP, Denver, Colorado, for Defendants - Appellees.
                       _________________________________
Before BACHARACH, McKAY, and CARSON, Circuit Judges.
                 _________________________________

McKAY, Circuit Judge.
                         _________________________________


       This appeal involves the fairness of sexual-misconduct disciplinary proceedings at

colleges and universities. In the district court, Plaintiff John Doe asserted that the

disciplinary proceeding brought against him by Defendants, the University of Denver

(“DU”) along with several of its employees, violated his rights under the Fourteenth

Amendment’s Due Process Clause and under Title IX. The court granted summary

judgment to Defendants on the Fourteenth Amendment claim because Plaintiff had failed

to show that DU was a state actor. The court also granted Defendants summary judgment

on the Title IX claim, concluding that Plaintiff had adduced insufficient evidence of

gender bias. 1

       I.        Fourteenth Amendment Due Process Claim

       We turn first to Plaintiff’s due process claim. DU is a private school, and thus

its actions are not normally subject to constitutional due process requirements. See

Brentwood Acad. v. Tenn. Secondary Sch. Athletic Ass’n, 

, 295 (2001)

(“[S]tate action [is] subject to Fourteenth Amendment scrutiny[,] and private conduct

(however exceptionable) . . . is not.”); Browns v. Mitchell, 

, 594 (10th


       1
        Plaintiff’s complaint also asserted several state-law claims and sought
declaratory relief. After disposing of the substantive federal-law claims, the court
declined to retain jurisdiction over the state-law claims, dismissed them and the
request for declaratory relief without prejudice, and closed the case.
                                             2
Cir. 1969) (“It is axiomatic that the due process provisions of the Fourteenth

Amendment proscribe state action only and do not reach acts of private persons

unless they are acting under color of state law.” (internal quotation marks omitted)).

As Plaintiff’s briefing suggests, his claim is cognizable only if DU may be deemed a

state actor for purposes of constitutional due process. See Brentwood 

(outlining tests used to determine whether state action should be

attributed to nominally private entities). Thus, at summary judgment, Plaintiff had

the burden to produce evidence demonstrating that DU should be deemed a state

actor. See Gallagher v. Neil Young Freedom Concert, 

, 1450, 1455–56

(10th Cir. 1995) (affirming summary judgment in favor of defendants in part because

plaintiffs failed to produce evidence to satisfy state-action tests).

       There are two constitutional sources of due process rights, the Fifth

Amendment and the Fourteenth Amendment. Plaintiffs pursuing procedural due

process claims based on actions by the federal government must proceed under the

Fifth Amendment, while plaintiffs bringing such claims based on actions by state

governments must proceed under the Fourteenth Amendment. See Koessel v.

Sublette Cty. Sheriff’s Dep’t, 

, 748 n.2 (10th Cir. 2013).

       Plaintiff has eschewed any reliance on the Fifth Amendment. Plaintiff

expressly relied only on the Fourteenth Amendment in his complaint and district

court briefing, and he continues to do so on appeal even in the face of both the

district court’s suggestion and DU’s assertion that the Fourteenth Amendment is

                                             3
inapposite for a due process claim based exclusively on the federal government’s

activities. (See Appellant’s Opening Br. at 53–54 (arguing that, although the district

court suggested “federal government activity is irrelevant to the 14th Amendment[,] .

. . the 14th Amendment . . . appl[ies]”).) Plaintiff is the master of his complaint. See

Bledsoe v. Vanderbilt, 

, 1119 (10th Cir. 2019). We are satisfied that

Plaintiff intended to bring this claim under the Fourteenth Amendment, and that is

how we will assess it. See In re Storer, 

, 1129 & n.6 (6th Cir. 1995)

(declining to assess claims under Fourteenth Amendment Due Process Clause where

plaintiffs clearly intended to rely only on Fifth Amendment Due Process Clause). 2


      2
         Plaintiffs might fail to reference the correct constitutional amendment
through mere inadvertence. Or, they might do so simply because they mistakenly
believe they need only show that a defendant’s actions should be attributed to
government in the generic sense, without distinguishing between federal and state
government. So, we have sometimes winked at a plaintiff’s reliance on the incorrect
amendment as an inconsequential mistake when the error appears to be the product of
inadvertence and where the distinction would be immaterial to the analysis, see Ward
v. Anderson, 

, 932 n.3 (10th Cir. 2007); see also Greene v. Impson, 530
F. App’x 777, 779 n.3 (10th Cir. 2013); Sawyer v. Burke, 504 F. App’x 671, 673–74
(10th Cir. 2012), and district courts within this circuit have done the same, see Sigg
v. Dist. Ct., No. 06-2436-KHV, 

, at *5 n.9 (D. Kan. Mar. 23, 2007);
Thunder v. Gunja, No. Civ.A03CV01575REBOES, 

, at *9 (D.
Colo. Aug. 11, 2005), adopted by 

(D. Colo. Sept. 27, 2005).
Other courts of appeal have done so as well. See Kell v. Smith, 743 F. App’x 292,
295–96 (11th Cir. 2018); Martial-Emanuel v. Holder, 523 F. App’x 345, 349 n.1 (6th
Cir. 2013); Collins v. Univ. of N.H., 

, 12 n.1 (1st Cir. 2011); High v.
Angelone, 

(table), 

, at *3 (9th Cir. 1999); Bieregu v.
Reno, 

, 1454 (3d Cir. 1995), abrogated on other grounds by Lewis v.
Casey, 

(1996), as recognized by Oliver v. Fauver, 

, 178
(3d Cir. 1997); United States v. Couch, 

, 79–83 & n.2 (5th Cir. 1990).
And, of course, excusing a mistaken reference to the wrong amendment is especially
appropriate when, unlike here, the plaintiff is proceeding pro se and cannot be
                                              4
      In support of his claim that DU was a state actor, Plaintiff relied solely on

evidence of the federal government’s involvement in DU’s affairs. Specifically,

Plaintiff pointed to (1) DU’s compliance with guidance from the Department of

Education’s Office for Civil Rights regarding Title IX’s requirements that was

contained in a 2011 Dear Colleague Letter (“DCL”), 3 which, Plaintiff asserts,

pressured DU to amend its policies in ways that were biased against male students

accused of sexual misconduct; and (2) the threatened loss of federal funding if DU

failed to conform to the DCL’s guidance. We have previously held, however, that

evidence regarding the federal government’s involvement with a private school or its

decision to discipline students has no bearing on whether the school is a state actor

under the Fourteenth Amendment, which is concerned only with the actions of state




expected to identify the specific legal source of his claim with the precision of a
trained lawyer. See Firstenberg v. City of Santa Fe, 

, 1024 (10th Cir.
2012). However, the error might also be the result of a calculated decision. For
instance, plaintiffs might avoid reliance on the Fifth Amendment due to the
limitations on such claims. See generally Big Cats of Serenity Springs, Inc. v.
Rhodes, 

, 858–64 (10th Cir. 2016); Peoples v. CCA Detentions Ctrs.,

, 1096–1108 (10th Cir. 2005).
       Here, we cannot construe Plaintiff’s claim as if brought under the Fifth
Amendment. Plaintiff is represented by capable attorneys, and his choice to eschew
reliance on the Fifth Amendment cannot be chalked up to mere inadvertence.
      3
        As explained below, the DCL “ushered in a more rigorous approach to
campus sexual misconduct allegations” by providing guidance that encouraged
schools to take tougher stances on students accused of sexual misconduct. Doe v.
Purdue Univ., 

, 668 (7th Cir. 2019). By informing schools that funding
depended on compliance with OCR’s guidance, the DCL was viewed as pressuring
schools to adhere to its guidance or else lose federal funding. See 

                                            5
governments. See 

(“Inasmuch as . . . 42 U.S.C. § 1983[ 4] is

concerned only with state action and does not concern itself with federal action[,] we

lay to one side as entirely irrelevant any evidence concerning the participation of the

federal government in the affairs of the University. And so it is state action with

which we are here concerned and more particularly . . . whether the State of Colorado

. . . [should be viewed as involved in] the challenged disciplinary proceeding.”

(citation and quotation marks omitted)). 5 Thus, Plaintiff failed to adduce any

relevant evidence to show that DU is a state actor for purposes of his Fourteenth

Amendment claim.

      In sum, although we agree with the district court that Plaintiff failed to

demonstrate that DU was a state actor for purposes of his Fourteenth Amendment due

process claim, we reach this conclusion on somewhat different grounds, namely that

Plaintiff failed to adduce any evidence of a state’s involvement in the disciplinary

proceeding he challenges. See Richison v. Ernest Grp., Inc., 

, 1130

(10th Cir. 2014) (“[W]e may affirm on any basis supported by the record . . . .”).




      4
         For purposes of determining whether a private entity may be held liable as a
state actor, the state-action requirement of the Fourteenth Amendment and the under-
color-of-state-law requirement of § 1983 are identical. See Neil Young Freedom

–47.
      5
        Other circuits have readily employed this same principle without hesitation.
See Musso v. Suriano, 

, 61 n.4 (7th Cir. 1978); Berrios v. Int’l Am. Univ.,

, 1332 n.5 (1st Cir. 1976); Weise v. Syracuse Univ., 

, 404
(2d Cir. 1975); Blackburn v. Fisk Univ., 

, 123 (6th Cir. 1971).
                                          6
Accordingly, we will affirm the court’s decision to grant Defendants summary

judgment on the due process claim.

      II.    Title IX Claim

      We now turn to Plaintiff’s Title IX claim, which requires some background.

Plaintiff is a male who enrolled as a freshman at DU in 2014. In October 2014,

Plaintiff had a sexual encounter with Jane Doe, a female freshman, in his dorm room.

Six months later, in April 2015, Jane’s boyfriend reported the encounter as an alleged

sexual assault to a DU resident director. The resident director then spoke with Jane,

who repeated the allegations and later filed with DU’s Office of Equal Opportunity a

complaint of non-consensual sexual contact.

      Under DU’s policies, a student’s non-consensual sexual contact with another is

a policy violation. Prohibited sexual contact includes contact by “coercion,” which

the policy defines as “unreasonable and persistent pressure to compel another

individual to initiate or continue sexual activity against an individual’s will,” such as

“continued pressure” after “someone makes clear that they do not want to engage in

sexual contact.” (Appellant’s App. at A139.)

      Two of the named Defendants, Kathryne Grove, OEO’s director, and Eric

Butler, an OEO investigator, investigated Jane’s allegations. The investigators

separately interviewed Plaintiff and Jane twice in May and June 2015, allowing each

of them to offer corrections to their own summary statements, which the investigators

had drafted for them based on their respective interviews, and allowing Plaintiff to

                                            7
submit a list of witnesses for the investigators to interview. The investigators also

interviewed other witnesses—Plaintiff’s roommate, a mutual acquaintance who was

present in the dorm room before the encounter took place, Jane’s boyfriend, and the

resident director who first received the allegations. In late June, the investigators

issued a preliminary report to Plaintiff and Jane, allowing them to offer any further

corrections to their own statements. The preliminary report, which did not make any

findings or conclusions, offered Plaintiff the first opportunity to see Jane’s

allegations against him.

       In mid-July 2015, the investigators issued their final report, which depicted a

he-said-she-said situation. After summarizing witness interviews, the investigators

“f[ound] it more likely than not that [Plaintiff]’s actions . . . resulted in non-

consensual sexual contact with [Jane] by means of coercion in violation of [DU’s]

policies.” (Appellant’s App. at A159.) No hearing was held. Pursuant to its

procedures, DU convened an outcome council to review the case and determine a

sanction. The outcome council decided to permanently dismiss Plaintiff from DU.

Plaintiff submitted an internal appeal challenging the investigation process, but it was

denied.

       In his complaint, Plaintiff claimed the disciplinary proceedings DU subjected

him to violated Title IX. The district court granted Defendants summary judgment

on the claim, concluding Plaintiff had failed to adduce evidence showing DU’s

actions were motivated by gender bias.

                                             8
      “We review the district court’s summary-judgment order de novo, applying the

same standard that the district court is to apply.” Singh v. Cordle, 

,

1037 (10th Cir. 2019). “Summary judgment is appropriate ‘if the movant shows that

there is no genuine dispute as to any material fact and the movant is entitled to

judgment as a matter of law.’” Evans v. Sandy City, 

, 852 (10th Cir.

2019) (quoting Fed. R. Civ. P. 56(a)). “A fact is material if, under the governing law,

it could have an effect on the outcome of the lawsuit. A dispute over a material fact

is genuine if a rational jury could find in favor of the nonmoving party on the

evidence presented.” Jones v. Norton, 

, 573 (10th Cir. 2015) (internal

quotation marks omitted). “In reviewing a motion for summary judgment, we review

the facts and all reasonable inferences those facts support[] in the light most

favorable to the nonmoving party.” 

(internal quotation marks

omitted).

      Title IX provides that “[n]o person in the United States shall, on the basis of

[gender], be excluded from participation in, be denied the benefits of, or be subjected

to discrimination under any education program or activity receiving Federal financial

assistance.” 20 U.S.C. § 1681(a). Title IX is “enforceable through an implied

private right of action.” Gebser v. Lago Vista Indep. Sch. Dist., 

, 281

(1998); see also Jackson v. Birmingham Bd. of Educ., 

, 183 (2005)

(explaining the Supreme Court “ha[s] consistently interpreted Title IX’s private cause

of action broadly to encompass diverse forms of intentional [gender]

                                           9
discrimination”). Generally, to succeed on a claim under Title IX, “a plaintiff must

show: (1) that he or she was excluded from participation in, denied the benefits of,

or subjected to discrimination in an educational program; (2) that the program

receives federal assistance; and (3) that the exclusion from the program was on the

basis of [gender].” Seamons v. Snow, 

, 1232 (10th Cir. 1996). Here,

there is no dispute that DU offers an educational program receiving federal assistance

or that Plaintiff was excluded from participating in that program. Thus, the only

issue is whether Plaintiff adduced sufficient evidence to raise a genuine dispute that

he was excluded from DU on the basis of gender.

      The district court concluded that Plaintiff had failed to adduce sufficient

evidence to raise a genuine dispute that gender was a motivating factor in DU’s

decision to expel him. The court recounted the litany of evidentiary arguments

Plaintiff raised in opposition to summary judgment but in the end concluded that

most of Plaintiff’s evidence was aimed at demonstrating that DU was biased in favor

of sexual-misconduct complainants and against sexual-misconduct respondents. In

the court’s view, assuming Plaintiff had created a genuine dispute that DU’s process

is biased against respondents, it was not reasonable to infer from this, without

additional evidence, that DU’s process is biased against males. The court found the

remainder of Plaintiff’s evidence similarly unavailing, concluding that none of it

raised a genuine dispute that DU’s decision was motivated by gender bias.



                                          10
      On appeal, Plaintiff argues that the district court erred both in refusing to

consider all of the evidence he presented and in concluding that his evidence was

insufficient to support an inference that DU’s decision to expel him was motivated by

gender bias. We consider these arguments in turn.

             A. Exclusion of Evidence

      In support of his opposition to summary judgment, Plaintiff submitted the

expert report of law professor Aya Gruber. In her report, Prof. Gruber opines that

Plaintiff’s disciplinary proceeding was marked by numerous deficiencies that give

rise to an appearance of bias based on gender stereotypes. The court declined to

consider Prof. Gruber’s report in its assessment of Plaintiff’s Title IX claim for two

reasons. First, the court pointed out that Plaintiff cited the report in his opposition

only three times, and never in support of his Title IX claim. Second, although the

court acknowledged that the report highlights alleged deficiencies in the disciplinary

proceedings against Plaintiff, the court concluded the report was not material to the

question before it—whether DU’s decisions were motivated by gender bias. On

appeal, Plaintiff argues the court erred by failing to consider Prof. Gruber’s report in

support of his Title IX claim. 6 He contends the court should have considered the


      6
         After filing their motion for summary judgment, Defendants filed a motion to
exclude Prof. Gruber’s expert testimony pursuant to Fed. R. Evid. 702. Because the
district court concluded it would not consider Prof. Gruber’s report on other grounds,
the court declined to resolve DU’s motion to exclude her expert testimony under Rule
702 and denied the motion as moot. On appeal, some of Plaintiff’s argument is
directed at showing that Prof. Gruber was qualified and that her testimony should
                                           11
report because it directly addresses the issue the district court said it did not, namely

whether gender bias was a motivating factor in DU’s decision to expel him.

       “We review a district court’s decisions excluding evidence at the summary

judgment stage only for an abuse of discretion.” LifeWise Master Funding v.

Telebank, 

, 927 (10th Cir. 2004). “Under this standard, we will not

disturb the district court’s decision unless we have a definite and firm conviction that

the lower court made a clear error of judgment or exceeded the bounds of permissible

choice in the circumstances.” 

marks omitted). Under Rule 56,

“[a] party asserting that a fact . . . is genuinely disputed must support the assertion by

. . . citing to particular parts of materials in the record.” Fed. R. Civ. P. 56 (c)(1).

Where a report or other material is “made part of the record” but the party “fail[s] to

cite to the ‘particular parts’ of the record that support[] [a particular] argument,” the

district court is “under no obligation to parse through the record to find the uncited

materials.” Unal v. Los Alamos Pub. Sch., 638 F. App’x 729, 742 (10th Cir. 2016);

see also Cross v. Home Depot, 

, 1290 (10th Cir. 2004) (“[I]t is the

responding party’s burden to ensure that the factual dispute is portrayed with

particularity, without depending on the trial court to conduct its own search of the

record.” (ellipsis and internal quotation marks omitted)); Mitchell v. City of Moore,




have been admitted under Rule 702. Because the district court never decided those
issues, we do not address them.
                                       12


, 1199 (10th Cir. 2000) (“The district court was not obligated to comb

the record in order to make [the non-movant]’s arguments for him.”).

      The district court did not abuse its discretion in declining to consider Prof.

Gruber’s report for purposes of the Title IX claim. Even assuming the court

misapprehended the contents of Prof. Gruber’s report, the court properly declined to

consider it in addressing the Title IX claim because Plaintiff failed to cite the report

in his summary-judgment arguments regarding that claim. Instead, he only cited the

report (1) in his statement of facts to dispute DU’s assertion that its investigators

understood the preponderance-of-the-evidence standard and (2) in his arguments

regarding his state-law claims to assert, based on the investigators’ allegedly one-

sided credibility assessments, that there remained a genuine dispute whether the

investigation was thorough, impartial, and fair enough to satisfy DU’s contractual

obligations. Plaintiff neither cited the report nor discussed the investigators’

understanding of the preponderance standard or their credibility assessments in his

arguments regarding his Title IX claim. In other words, Plaintiff did not meet his

burden to cite the particular part of the record he now claims should have been

considered to support his Title IX argument. See Unal, 638 F. App’x at 742. Under

these circumstances, we cannot fault the district court for declining to parse through

the record in order to conjure up arguments from the record that Plaintiff might have




                                           13
made on his own, and its decision to refrain from doing so was no abuse of

discretion. 7

                B. Evidence of gender bias

       On appeal, Plaintiff argues that several categories of evidence he adduced in

the district court were sufficient to create a genuine dispute regarding whether gender

was a motivating factor in the proceeding DU brought against him. We evaluate each

category in turn.

       First, as other plaintiffs have in recent years, Plaintiff sets the stage for his

Title IX claim by shining a spotlight on the 2011 Dear Colleague Letter, which

“ushered in a more rigorous approach to sexual misconduct allegations,” Doe v.


       7
         On appeal, Defendants assert that Plaintiff adduced no direct evidence of
gender bias and that, even if we concluded the district court erred by refusing to
consider Prof. Gruber’s report, we should nonetheless affirm the dismissal of the
Title IX claim because a party opposing summary judgment cannot rely solely on an
expert report to create a genuine dispute on a material issue. This argument is
debatable. See KSR Int’l Co. v. Teleflex Inc., 

, 427 (2007) (explaining
that “expert testimony . . . may resolve or keep open certain questions of fact” at
summary-judgment stage); Talkington v. Atria Reclamelucifers Fabrieken BV, 

, 264 (4th Cir. 1998) (concluding that expert testimony on causation
element, standing alone, is sufficient to support jury verdict). Defendants also argue
that Prof. Gruber’s report is unsworn and is thus not competent summary judgment
evidence. This argument is likewise debatable. Prof. Gruber signed and dated her
report and later signed and dated a declaration, sworn under penalty of perjury,
stating that the report, which she attached, was a true and correct copy. These
actions might satisfy the requirements of 28 U.S.C. § 1746, which would render Prof.
Gruber’s report competent for summary judgment purposes. See Fed. R. Civ. P.
56(c)(4) advisory committee’s notes to 2010 Amendments. Ultimately, however, we
decline to address these arguments because we conclude the district court properly
refused to consider Prof. Gruber’s report, regardless of whether it was competent
summary judgment evidence.
                                            14
Purdue Univ., 

, 668 (7th Cir. 2019); accord Menaker v. Hofstra Univ.,

, 26 (2d Cir. 2019), by providing guidance to schools receiving federal

funding regarding Title IX’s requirements as they relate to sexual assault. Moreover,

Plaintiff presents evidence specific to DU’s response to the DCL, which included

“engag[ing] national experts” to “evaluat[e] its processes”; “[c]reat[ing] a team of

administrators to address concerns”; establishing positions for a Title IX coordinator

and a second Title IX investigator at the school; altering its investigative model;

engaging in several sexual-assault awareness campaigns on campus; and “review[ing

and] revis[ing]” its “support and resources for victims,” methods of “handl[ing]

expressions of concern,” and “prevention efforts.” (Appellant’s App. at A507.)

Plaintiff further presents evidence that DU’s training materials warned employees

that they “need to take [compliance with Title IX] very seriously” because it “is the

focus of OCR right now,” emphasizing that (1) the Department of Education could

“cut off federal funding/initiate proceedings to do so”; (2) OCR could commence

“compliance review,” which “is very time consuming, creates extremely negative

publicity for the school, and is very thorough”; and (3) “individual employees” could

be “personally sued in a civil lawsuit by student[s]” if they failed to comply with

Title IX or “possibly . . . be held personally liable” if they were “aware of sexual

harassment of student[s] and show[ed] ‘deliberate indifference’” to it. (Id. at A510–

11.) Plaintiff then contends that the DCL and the pressure DU felt to comply with its

guidance give rise to an inference of gender bias.

                                           15
      The majority of other courts to have considered this issue have concluded that,

although evidence of the DCL and external pressure placed on the school to conform

with its guidance may provide the plaintiff with “a story about why [the school]

might have been motivated to discriminate against males accused of sexual assault,”

such evidence is insufficient in itself to support any inference that the school’s

actions in a particular case were motivated at least in part by gender bias. Purdue

; see also, e.g., Doe v. Baum, 

, 586 (6th Cir.

2018) (“Of course, all of this external pressure alone is not enough to state a claim

that the university acted with bias in this particular case. Rather, it provides a

backdrop that, when combined with other circumstantial evidence of bias in Doe’s

specific proceeding, gives rise to a plausible claim.”). We agree. The DCL is

gender-neutral on its face, see Neal v. Colo. State Univ., No. 16-cv-873-RM-CBS,

, at *11 (D. Colo. Feb. 16, 2017), and evidence that a school felt

pressured to conform with its guidance cannot alone satisfy Title IX’s fundamental

requirement that the challenged action be “on the basis of [gender],” 20 U.S.C. §

1681(a). Thus, Plaintiff’s evidence of the DCL and the pressure DU felt to comply

with its guidance cannot support his summary judgment burden unless combined with

a “particularized ‘something more,’” Doe v. Columbia Coll. Chi., 

, 856

(7th Cir. 2019), that would indicate that DU’s decision in his particular case was

based on his gender. And, as explained below, we conclude that Plaintiff has not

made this particularized showing here.

                                           16
       Second, Plaintiff points to statistical evidence showing an overwhelming

disparity in the gender makeup of sexual-assault complainants and sexual-assault

respondents at DU. Specifically, between 2011 and 2016, nearly all complainants

(35 out of 36) were female, and all respondents (36 out of 36) were either listed as

male or could be presumed to be male based on the nature of the complaint. Plaintiff

does not explain how this disparity amounts to gender bias on the part of DU, except

to say that DU should have expected that its implementation of the DCL’s guidance

would disproportionately affect men because the DCL was intended to address a

perceived epidemic of male sexual assault against women. But, on its face, the DCL

says no such thing, and Plaintiff points to no evidence suggesting that DU changed

its policies in light of this statistical disparity or in order to combat sexual assault

perpetrated specifically by men against women. At best, then, the statistical disparity

can only create a genuine dispute to the extent it generates a reasonable inference that

DU’s decision to expel Plaintiff was motivated by considerations of gender.

Plaintiff’s argument thus reduces down to an inferential proposition: a factfinder can

reasonably infer from the fact that sexual-assault respondents are overwhelmingly

male that a school’s decision to initiate proceedings against respondents is motivated

by the fact that they are male.

       Assessing what inferences may reasonably be drawn from the statistical

disparity in the gender makeup of sexual-assault complainants and respondents is one

of the more perplexing aspects of addressing Title IX challenges to sexual-

                                             17
misconduct disciplinary proceedings. 8 See Doe v. Univ of Colo. ex rel. Bd. of

Regents of Univ. of Colo., 

, 1075–76 (D. Colo. 2017)

(cautioning against accepting or rejecting inferences in similar context without

reflection). The courts that have engaged in this analysis have generally concluded

that statistical disparities in the gender makeup of complainants and respondents can

readily be explained by “an array of alternative” nondiscriminatory possibilities,

potentially “reflect[ing], for example, that male students on average . . . committed

more serious assaults,” that sexual-assault victims are likelier to be women, or that

female victims are likelier than male victims to report sexual assaults. Haidak v.

Univ. of Mass.-Amherst, 

, 75 (1st Cir. 2019); see also Doe v. Trs. of Bos.

Coll., 

, 92 (1st Cir. 2018); Doe v. Cummins, 662 F. App’x 437, 453–54




       8
         Under similar anti-discrimination statutes, statistical disparities of this nature
are often used to prove a disparate-impact theory of liability, which does not require
proof of intentional discrimination. Some courts of appeals, however, have held or
suggested that a disparate-impact theory of liability is not cognizable under Title IX.
See Fort v. Dallas Indep. Sch. Dist., 

(table), 

at *3 n.3
(5th Cir. 1996) (noting circuit split). Although noting that “there has been some
question whether Title IX prohibits disparate impact discrimination,” we have
suggested that a Title IX disparate-impact claim might be viable, Mabry v. State Bd.
of Cmty. Colls. & Occupational Educ., 

, 316 n.6, 318 (10th Cir. 1987),
but we have never directly addressed the issue. This appeal does not present an
occasion to do so, as Plaintiff disclaims any reliance on a disparate-impact theory of
liability. But, aside from proving disparate impact, “proper evidence of a statistical
disparity may [also] generate an inference of intentional discrimination” if it
“‘tend[s] to show that there was a causal connection between the outcome of the
disciplinary proceedings and gender bias.’” Haidak v. Univ. of Mass.-Amherst, 

, 75 (1st Cir. 2019) (brackets omitted) (quoting Doe v. Trs. of Bos. Coll., 

, 91 (1st Cir. 2018)).
                                            18
(6th Cir. 2016); Doe v. Univ. of Cincinnati, 

, 607–08 (S.D. Ohio

2016). When the statistical evidence does nothing to eliminate these obvious,

alternative explanations for the disparity, an inference that the disparity arises from

gender bias on the part of the school is not reasonable. See 

;

Bos. 

; Cummins, 662 F. App’x at 453–54; Univ. of 

–08.

      We agree with this analysis. A factfinder could not reasonably infer from bare

evidence of statistical disparity in the gender makeup of sexual-assault complainants

and respondents that the school’s decision to initiate proceedings against respondents

is motivated by their gender. This is so because, at least in the discrimination

context, the extent to which a discriminatory motive may be reasonably inferred from

evidence of statistical disparity often depends on the evidence’s ability to eliminate

obvious nondiscriminatory explanations for the disparity. See Luster v. Vilsack, 

, 1094 (10th Cir. 2011) (“In order to be probative of discrimination,

statistical evidence must eliminate nondiscriminatory explanations for the disparity.”

(internal quotation marks omitted)); Turner v. Pub. Serv. Co. of Colo., 

, 1147 (10th Cir. 2009) (“In order for statistical evidence to create an inference

of discrimination, the statistics must show a significant disparity and eliminate

nondiscriminatory explanations for the disparity.” (brackets omitted) (quoting Fallis

v. Kerr-McGee Corp., 

, 746 (10th Cir. 1991)); Schulte v. Potter, 218 F.

App’x 703, 714 (10th Cir. 2007) (explaining that, where evidence “wholly fail[s] to

                                           19
eliminate nondiscriminatory explanations for” disparate treatment, “[i]t would be

unreasonable to draw an inference of” intentional discrimination (internal quotation

marks omitted)).

       This principle is especially applicable here. In employment discrimination

cases, the nondiscriminatory explanations for statistical disparity that prevent an

inference of discriminatory intent often involve the employer’s own hiring or

promotion criteria. See, e.g., 

. One might suspect that the

principle requiring a plaintiff to negate nondiscriminatory explanations of statistical

disparity would be at its weakest where the defendant controls the putative

nondiscriminatory causes of disparate treatment. In Title IX challenges to sexual-

misconduct proceedings, however, the putative nondiscriminatory causes of

disparity—the gender makeup of sexual-assault perpetrators, victims, and reporters—

are almost completely beyond the control of the school. See Univ. of Colo., 255 F.

Supp. 3d at 1078 (“[T]he University is not responsible for the gender makeup of

those who are accused by other students of sexual misconduct.” (internal quotation

marks omitted); accord Cummins, 662 F. App’x at 454. We think then that the

principle would be at its strongest in this context. 9


       9
         Further, statistical disparity by itself does little to inform the factfinder of
whether the school was motivated by gender with respect to the particular proceeding
brought against the plaintiff. See 

(“Even if one could infer
from the data that another decision maker issued higher penalties based on [gender],
that inference says little about whether the decision maker in this case brought to
bear any bias on the basis of [gender].”); 

(“Turner’s
                                              20
      Here, Plaintiff’s statistical evidence does not create a reasonable inference that

DU’s decisions regarding the initiation of sexual-misconduct proceedings were

motivated by considerations of gender. His statistical evidence does nothing to

eliminate the nondiscriminatory explanations for the disparity identified above, and

thus it would be unreasonable for a factfinder to infer from the statistical disparity

alone that DU decides to initiate proceedings against respondents based on their

gender. Something more is needed to show the disparity results from gender bias

rather than nondiscriminatory, exogenous factors—something like an affidavit from a

knowledgeable person stating the school exhibits a pattern of prosecuting complaints

against male but not female students, see Doe v. Miami Univ., 

, 593 (6th

Cir. 2018), or a statement from school officials touting such statistics in response to

public criticism of the school’s previous handling of female students’ sexual-assault

allegations, see Doe v. Geo. Wash. Univ., 

, 12–13 (D.D.C. 2018).

Plaintiff points to no additional evidence of this kind.

      Third, Plaintiff points to evidence of DU’s alleged bias against respondents in

sexual-misconduct proceedings. Specifically, Plaintiff presented evidence that DU’s




statistic regarding the gender imbalance of the . . . workforce . . . does not, without
additional evidence, suggest that Turner herself experienced discrimination. The
numbers fail to provide any information regarding whether the decision not to hire
Turner, and that decision alone, involved discrimination on the basis of [gender].”
(internal quotation marks omitted)).
                                          21
training materials 10 referred to sexual-misconduct complainants as “survivors” and

directed staff to “[e]mpower the survivor” and “[c]ommunicate that you believe the

survivor.” (Appellant’s App. at A519.) 11 Plaintiff also presented evidence that,

when the investigation against him began, DU provided a list of resources to him and

Jane to help them navigate the Title IX process. Plaintiff asserts these resources were

complainant-specific and thus supported the needs of complainants but not

respondents. For instance, Ms. Grove testified at her deposition that one resource on

the list, the Center for Advocacy and Prevention and Empowerment, did not “support

men who were accused of sexual assault.” (Id. at A351.) Of course, as the district

court noted, there is no evidence in the record that CAPE would support women


      10
         Plaintiff also points to Prof. Gruber’s report, which asserts that the training
received by the two investigators assigned to Plaintiff’s case, Mr. Butler and Ms.
Grove, was suffused with stereotypical assumptions about men and women, leading
them to investigate Jane’s allegations in a gender-biased way. We have already
concluded that the district court did not err by declining to consider Prof. Gruber’s
report for purposes of Plaintiff’s Title IX claim. We do not consider the report here
either.
      11
          In this same vein, Plaintiff points out that DU’s Title IX Coordinator,
Defendant Jean McAllister, referred to complainants as “victims” and “survivors”
during her deposition and acknowledged approaching interviews of complainants
with the belief that they are “survivor[s]” and that their “report[s] [are] legitimate.”
(Appellant’s App. at A431–32.) Citing Mallory v. Ohio Univ., 76 F. App’x 634, 640
(6th Cir. 2002), the district court concluded that any indication that Ms. McAllister
had a bias against males was irrelevant because Plaintiff had failed to raise a genuine
dispute that Ms. McAllister, who began her position with DU as the proceeding
against Plaintiff neared its end, had any meaningful involvement in the proceeding.
Plaintiff does not contest this conclusion on appeal, so his argument regarding Ms.
McAllister’s bias is waived. See Talley v. Time, Inc., 

, 906 n.28 (10th
Cir. 2019).
                                             22
accused of sexual assault either, and Ms. Grove testified that other resources on the

list would provide support to men accused of sexual assault. Additional testimony

from Ms. McAllister that Plaintiff himself points to—that she would like to develop

named support programs for respondents in the same way DU has developed named

support programs for complainants—highlights that Plaintiff’s argument is based on

the relative disparity between resources for complainants and resources for

respondents.

      Whether factfinders may reasonably infer anti-male bias from evidence of a

school’s anti-respondent bias is another thorny issue that often arises in Title IX

challenges to sexual-misconduct disciplinary proceedings. See Norris v. Univ. of

Colo., 

, 1014–15 (D. Colo. 2019); Univ. of Colo., 

at 1075–76. Most courts to have addressed the issue have concluded that evidence

of a school’s anti-respondent bias does not create a reasonable inference of anti-male

bias. See Cummins, 662 F. App’x at 453; Doe v. Rider Univ., No. 3:16-cv-4882-

BRM-DEA, 

, at *10 (D.N.J. Jan. 17, 2018); Doe v. Colgate Univ.,

No. 5:15-cv-1069 (LEK/DEP), 

, at *11 (N.D.N.Y Oct. 31, 2017);

Doe v. Columbia Coll. Chi., 

, 956–57 (N.D. Ill. 2017); Ruff v. Bd.

of Regents of Univ. of N.M., 

, 1302 (D.N.M. 2017); Univ. of

Colo., 

at 1079; Doe v. Univ. of St. Thomas, 

, 991

(D. Minn. 2017); Doe v. Univ. of Mass., No. 14-30143-MGM, 

, at

*8 (D. Mass. July 14, 2015); Haley v. Va. Commonwealth Univ., 

,

                                           23
579 (E.D. Va. 1996). They reason that evidence of a school’s anti-respondent bias

does not permit a reasonable inference of an anti-male bias because both males and

females can be respondents. See, e.g., Cummins, 662 F. App’x at 453 (“[A]

disciplinary system that is biased in favor of alleged victims and against those

accused of misconduct . . . does not equate to gender bias because sexual-assault

victims can be both male and female.”).

       We agree. We have relied on the same rationale in the employment

discrimination context and have held that, on its own, evidence of an employer’s

discriminatory treatment of a group to which both genders can belong does not give

rise to an inference of gender discrimination. See Adamson v. Multi Cmty.

Diversified Servs., Inc., 

, 1148–49 (10th Cir. 2008) (“‘Familial status’

is not a classification based on [gender] any more than is being a ‘sibling’ or

‘relative’ generally. It is, by definition, gender neutral. . . . Assertions that an

employer discriminated against an individual on the basis of his or her ‘familial

status’ alone state no cognizable cause of action under Title VII.”). 12 The reasoning


       12
         Other courts have also employed this rationale in employment
discrimination cases involving employer policies that might be understood to
discriminate against a group that may include both men and women, such as
employees who suffer from infertility or employees who are new parents. See, e.g.,
Saks v. Franklin Covey Co., 

, 347 (2d Cir. 2003) (“Because male and
female employees . . . are equally disadvantaged by the [policy], we conclude that the
Plan does not discriminate on the basis of [gender].”); Piantanida v. Wyman Ctr.,
Inc., 

, 342 (8th Cir. 1997) (explaining that “[a]n employer’s
discrimination . . . based on a gender-neutral status potentially possessible by all
employees, including men and women,” is not cognizable); cf. Hall v. Nalco Co., 534
                                           24
applies equally well in the Title IX context. See Gossett v. Oklahoma ex rel. Bd. of

Regents for Langston Univ., 

, 1176 (10th Cir. 2001) (“Courts have

generally assessed Title IX discrimination claims under the same legal analysis as

Title VII claims.”). Classification as a sexual-misconduct respondent is not a

classification based on gender. It is gender-neutral because both men and women can

be respondents. Accordingly, by itself, evidence of a school’s anti-respondent bias

does not permit a reasonable inference of discrimination based on gender.

      Here, Plaintiff’s evidence demonstrates at most that DU had an anti-

respondent or pro-complainant bias, which is insufficient to create an inference of

anti-male bias. A number of courts have determined that references to complainants

as “victims” or “survivors” or language suggesting a pro-victim viewpoint exhibits at

most a bias in favor of complainants qua complainants and against respondents qua

respondents. See Bos. 

; Doe v. Quinnipiac Univ., 

643, 661 n.6 (D. Conn. 2019); Rider Univ., 

, at *10; Colgate

Univ., 

, at *14–15; Columbia Coll. 

.

Plaintiff’s reliance on similar pro-victim language in DU’s training materials at most

demonstrates an anti-respondent bias. Likewise, the relative lack of support

resources DU offers to respondents compared to the resources it offers complainants




F.3d 644, 646–49 (7th Cir. 2008) (implying that policy affecting group that includes
both male and female employees is not cognizable).
                                        25
demonstrates at most a bias against respondents. However, this evidence of anti-

respondent bias does not raise an inference of discrimination based on gender. 13

      Fourth, Plaintiff argues the investigators exhibited bias by finding Plaintiff

responsible for non-consensual sexual contact despite evidence supporting his

version of the events. In Plaintiff’s view, the evidence before the investigators so

clearly favored a finding that Plaintiff’s and Jane’s sexual encounter was consensual

that the investigators’ finding to the contrary creates an inference of bias in their

decision.

      For support, Plaintiff relies on Doe v. Columbia University, 

(2d

Cir. 2016). In Columbia University, the Second Circuit reviewed an order dismissing

a complaint pursuant to Fed. R. Civ. P. 12(b)(6). According to the allegations in the

complaint, the student complainant was “an altogether willing participant” in the

underlying sexual encounter; “‘no evidence was presented’” to the school’s tribunal

to support the complainant’s claim that sexual activity was coerced; and the tribunal

“chose to accept [the complainant’s] unsupported accusatory version” of events and

“declined even to explore the testimony of [the] [p]laintiff’s witnesses.” Columbia


      13
          Plaintiff’s only response to this analysis has been to argue that his evidence
of DU’s anti-respondent bias amounts to evidence of an anti-male bias because the
statistical evidence shows that respondents are overwhelmingly male. But we have
already determined that Plaintiff’s statistical evidence is insufficient because it fails
to eliminate non-gender-based explanations for the disparity. Indeed, other courts
have viewed a school’s bias in favor of complainants as one of the legitimate, non-
gender-based explanations for the disparity that bare statistical evidence fails to
eliminate. See, e.g., Univ. of Colo., 

at 1079.
                                             26


. These allegations, which the court was obligated to “accept in

the light most favorable to [the] [p]laintiff,” gave “plausible support to the

proposition that the[ tribunal’s members] were motivated by bias” because, “[w]hen

the evidence substantially favors one party’s version of a disputed matter, but an

evaluator forms a conclusion in favor of the other side (without an apparent reason

based in the evidence), it is plausible to infer . . . that the evaluator has been

influenced by bias.” 

does not aid Plaintiff’s cause. We assume, without

deciding, that the chief proposition from Columbia University Plaintiff relies on—

that an inference of bias arises when an evaluator’s decision in favor of one side

lacks an apparent, evidence-based reason, and the evidence substantially favors the

other side—is correct. But that proposition has no application here. Simply put,

DU’s investigators were not faced with a situation in which the evidence

substantially favored Plaintiff. Unlike in Columbia University, there was evidence

presented in favor of Jane’s claim that the sexual encounter was not consensual, and

it cannot be said that the investigators lacked an evidence-based reason for reaching

their decision. Thus, it would not be plausible or reasonable to infer merely from the

investigators’ weighing of the evidence that they were biased.

       Further, even if we agreed with Plaintiff that the evidence before the

investigators was so one-sided in Plaintiff’s favor that their decision in favor of Jane

could reasonably give rise to an inference of bias, this would still fall short of

                                            27
demonstrating bias based on gender. Columbia University itself acknowledges that

an evaluator’s decision at odds with the great weight of evidence “support[s] [an]

inference of bias” but “not necessarily” a “bias on account of [gender].” 

Circuit instead concluded that additional allegations in the complaint gave

“ample plausible support to a bias with respect to [gender],” namely “substantial

criticism” from “the student body and in the public media, accusing the [school] of

not taking seriously complaints of female students alleging sexual assault by male

students” as well as an allegation that the school “was cognizant of, and sensitive to,

these criticisms.” 

Thus, in our view, the allegations at issue in

Columbia University reflect gender-biased public pressure accompanied by

procedural irregularity in the proceeding at issue. Here, however, Plaintiff has

adduced only evidence of gender-neutral public pressure. So, even if we were to

accept the inference of bias he presses, he has failed to adduce the additional

evidence needed to demonstrate bias on account of gender. 14

      Fifth, Plaintiff argues that the severity of the sanction he received—

expulsion—resulted from DU’s anti-male bias. Under DU’s policies, the outcome

council is to consider a number of factors to determine an appropriate sanction for a

student found responsible for violating DU’s sexual-misconduct policy, including (1)




      14
         Plaintiff again resorts to Prof. Gruber’s report to argue that the investigation
was pockmarked by procedural deficiencies that disfavored Plaintiff. We again do
not consider Prof. Gruber’s report.
                                            28
the “nature and severity of the act,” (2) the “number of complainants,” (3) the “prior

student conduct history of the respondent,” (4) the outcome council’s “assessment of

the effect . . . the act or policy violation has on the complainant, community[,] and

University environment,” and (5) the “complainant[’s] and community[’s] safety.”

(Appellant’s App. at A153 (capitalization standardized).) Plaintiff also elicited

deposition testimony from Defendant Kristin Olson, a member of DU’s outcome

council in Plaintiff’s proceeding, that, in her experience, the respondent was expelled

in every case where investigators found non-consensual sexual conduct involving

penetration. Plaintiff also points to DU’s records confirming that, for the 14 non-

consensual sexual contact cases between 2013 and 2016 that resulted in dismissal or

rescission of an admission offer, each case involved a female complainant, a male

respondent, and allegations of penetration.

      Plaintiff contends that DU, in derogation of its own policies, expels males

found responsible for non-consensual sexual contact involving penetration regardless

of the circumstances. For instance, in his case, Plaintiff points out that several of the

factors the outcome council was required to consider weighed in his favor: the

allegations did not involve physical violence or lead to a criminal investigation; only

one complainant accused him of misconduct; he had no prior record of student

conduct issues; and the facts that Plaintiff and Jane met socially after the incident and

that DU did not impose an interim suspension on him after the complaint was filed

suggest he posed no threat to Jane’s or the community’s safety. He argues that the

                                           29
outcome council simply ignored these factors and imposed expulsion without

considering them. The severity of the sanction, Plaintiff asserts, gives rise to an

inference of bias on account of gender, as it reflects a belief that males need to be

sanctioned severely for sexual misconduct.

      A factfinder could not reasonably infer from this evidence that the severity of

the sanction DU imposed was motivated by Plaintiff’s gender. First, Plaintiff ignores

the fact that DU’s policies, in addition to laying out factors for the outcome council

to consider, also expressly state that, “[i]n general[,] violations of the non-consensual

sexual contact provision” of the policy “typically result in a dismissal.” (Id. at

A154.) Moreover, much of Plaintiff’s argument again relies to some degree on

evidence of a statistical disparity between the numbers of men and women expelled

from DU for engaging in non-consensual sexual contact involving penetration.

However, for evidence of this nature to raise an inference of gender bias, it must

eliminate obvious, nondiscriminatory explanations for the disparity. Again, Plaintiff

has not eliminated the obvious, nondiscriminatory explanation that DU, as expressed

in its own policy, has legitimate interests in expelling students—regardless of their

gender—who engage in non-consensual sexual contact, and, though not expressed in

its policies, DU might have even greater interests in doing so when that contact

involves penetration. In short, something more is needed to show that the cited

expulsions resulted from the fact the respondents were male rather than the fact they

were found responsible for sexual misconduct, but Plaintiff has failed to adduce it.

                                           30
      To the extent Plaintiff contends that the outcome council ignored the factors it

was required to consider in his proceeding, that contention is not borne out by the

record. Plaintiff points to no evidence showing that the outcome council failed to

consider the factors. In fact, the only evidence in the record on this point—Ms.

Olson’s deposition testimony—strongly suggests the outcome council did consider

those factors when contemplating the sanction it would impose on Plaintiff. Her

testimony also strongly suggests the outcome council concluded that the nature and

severity of the contact (non-consensual penetration) and the threat Plaintiff posed to

the community (as he did not consider himself responsible and was thus unlikely to

rehabilitate) outweighed any of the factors that might be in his favor. The outcome

council’s letter to Plaintiff notifying him of its decision specifically referenced these

two factors, explaining that its decision to expel him was “due to the nature and

severity of [Plaintiff]’s actions and in an effort to protect the community.”

(Appellant’s App. at A163.) We have no call to review the outcome council’s

consideration of these sanctioning factors, for, where the evidence regarding

sanctioning factors is not clearly one-sided, the mere fact that Plaintiff or this court

might have considered the factors differently or imposed a less severe sanction does

not create a reasonable inference of bias, let alone bias based on gender. See Doe v.

Colgate Univ., 760 F. App’x 22, 33 (2d Cir. 2019); cf. Davis ex rel. LaShonda D. v.

Monroe Cty. Bd. of Educ., 

, 648 (1999) (“[C]ourts should refrain from

second-guessing the disciplinary decisions made by school administrators.”).

                                            31
      Further, even if we agreed that DU’s pattern of sanctions exhibited some bias

against students who, like Plaintiff, are found responsible for non-consensual sexual

contact involving penetration, this would not amount to a bias on account of gender.

This is so because both men and women can engage in non-consensual sexual

contact, and, for both men and women, that contact can involve penetration. 15 As we

have explained above, evidence of a school’s discriminatory treatment of a group that

can include both men and women does not create a reasonable inference of gender

discrimination. See 

–49. DU’s treatment of students

found responsible for non-consensual sexual contact involving penetration is gender-

neutral because both men and women can be included in that group. Thus, DU’s

alleged bias against that group does not permit a reasonable inference of bias based

on gender.

      Sixth, Plaintiff argues that DU encouraged the filing of sexual-misconduct

complaints specifically against males. For support, Plaintiff cites his own deposition

testimony in which he stated that DU placed “numerous posters all around the

school” to encourage the reporting of sexual misconduct and recalled seeing one

“poster that said [‘]if you regret it, it was rape.[’]” (Appellant’s App. at A425–26.)

In Plaintiff’s view, this kind of encouragement was intended to increase the number

of sexual-misconduct complaints in a way that targeted males.


      15
         Ms. Olson’s deposition testimony makes clear that DU considers penetration
to include oral, anal, or vaginal penetration with a penis, digit, or foreign object.
                                            32
      To the extent Plaintiff contends that an inference of anti-male bias arises from

DU’s attempts to encourage sexual-misconduct reporting generally, we find any such

argument unpersuasive. At most, encouragement of this nature might possibly be

construed as exhibiting a bias against potential respondents because it increases the

likelihood that potential respondents will be subjected to investigation and possibly

sanctioned if found responsible. But both men and women can be potential

respondents, and therefore any bias against them would not be bias on account of

gender.

      As for the specific poster Plaintiff recalls, the poster’s language—“if you

regret it, it was rape”—viewed in Plaintiff’s favor, can reasonably be interpreted to

encourage the reporting of sexual misconduct committed specifically by men against

women. Although in modern usage “rape” can refer generally to “forced, non-

consenting, or illegal sexual intercourse with another person” or “sexual violation or

assault,” regardless of the gender of the perpetrator or victim, 16 the term

“[o]riginally” and still “chiefly” can refer to “the act or crime, committed by a man,

of forcing a woman to have sexual intercourse with him against her will.” Rape,

Oxford English Dictionary (3d ed. 2008) (emphasis added). Thus, viewed in

Plaintiff’s favor, the poster could be understood to have been directed at women who




      16
         See also Rape, American Heritage Dictionary of the English Language (5th
ed. 2011) (defining the term without reference to the gender of either perpetrator or
victim).
                                          33
had sexual encounters with men, and it encouraged them to view and report

encounters with men they regretted as instances of sexual misconduct by equating

regret, which typically is not viewed as an indication of misconduct, with rape,

perhaps the most serious form of misconduct.

      Even viewed in Plaintiff’s favor, however, the poster does not create a genuine

dispute that DU was motivated by considerations of gender in Plaintiff’s proceeding.

For one thing, there is no evidence suggesting that the poster Plaintiff recalls was

sponsored or approved by DU or that its message otherwise can be attributed to DU

generally or to any of the decisionmakers in his proceeding specifically. 17 See

Hysten v. Burlington N. Santa Fe Ry. Co., 415 F. App’x 897, 911 (10th Cir. 2011)

(“[E]vidence of discrimination in the decision-making process must be distinguished

from ‘stray remarks in the workplace, statements by nondecisionmakers, or

statements by decisionmakers unrelated to the decisional process.’” (quoting




      17
         In other contexts, when a plaintiff’s claim hinges to some degree on a
message contained in a poster, flyer, or the like, courts have often looked for
indications that the message can be attributed to the defendant. See, e.g., Child
Evangelism Fellowship of N.J., Inc. v. Stafford Twp. Sch. Dist., 

, 525
(3d Cir. 2004); Munoz-Feliciano v. Monroe-Woodbury Cent. Sch. Dist., No. 13-CV-
4340 (CS), 

, at *5 (S.D.N.Y Mar. 25, 2015); DeCarolis v.
Presbyterian Med. Ctr., No. 11-cv-1422, 

, at *6 (E.D. Pa. Aug.
20, 2012); Chacas v. City of Ely, 

, 1209 (D. Nev. 2009). And in
the context of challenges to sexual-misconduct disciplinary proceedings, courts have
emphasized that circumstantial evidence of gender bias on the part of non-
decisionmakers is largely irrelevant. See, e.g., 

. We have
made the same point in employment discrimination cases. See, e.g., 

.
                                           34
Clearwater v. Indep. Sch. Dist. No. 166, 

, 1126 (8th Cir. 2000))).

Thus, the poster’s connection to DU’s motivations in pursuing sexual-misconduct

allegations generally, not to mention its motivations in regard to Plaintiff’s

proceeding particularly, is tenuous at best. And, even if the poster could be

attributed somehow to DU or the decisionmakers at issue, it amounts to nothing more

than “an isolated and ambiguous comment” that “is generally considered too abstract

to support an inference of discrimination.” 

. Beyond this,

we think a single reference in Plaintiff’s own deposition testimony to an

ambiguously-worded poster with nothing connecting it to DU, the relevant

decisionmakers, or Plaintiff’s proceeding amounts to nothing more than a scintilla of

evidence that is insufficient to withstand summary judgment. See 

.

      In sum, we conclude the district court did not err in determining that Plaintiff

failed to adduce sufficient evidence to create a genuine dispute that DU was

motivated by considerations of gender in the proceeding it brought against him. The

only potential evidence of bias on account of gender Plaintiff presented was his

recollection of the if-you-regret-it-it-was-rape poster, which is simply too thin a nail

to hang a claim of gender bias on. Aside from the poster, Plaintiff’s evidence

demonstrates, at most, only that (1) as is almost certainly the case at nearly every

school, the overwhelming majority of sexual-misconduct respondents are men, and

(2) DU’s policies and procedures exhibit an anti-respondent bias. As we have

                                           35
explained, neither the statistical disparity in the gender makeup of respondents nor

evidence of an anti-respondent bias can create a reasonable inference of bias on

account of gender.

      We are not unmindful that the combination of this statistical disparity and

overt anti-respondent bias—a combination not unlikely to recur with some frequency

at other schools—raises palpable concerns that schools might be making a distinction

without a real difference and that stereotypes and prejudices against a class protected

by Title IX (males) are beginning to infect the enforcement of sexual-misconduct

policies under the auspices of presumptions regarding an unprotected class

(respondents). See generally Univ. of Colo., 

at 1075–76. 18



      18
         This concern is only heightened when there is not only evidence that the
school exhibits an anti-respondent bias generally but also colorable evidence that the
school employed that bias in the sexual-misconduct proceeding at issue. Here, for
instance, there is colorable evidence that the investigators:

   • refused to follow leads that were potentially exculpatory;
   • disbelieved Plaintiff from the outset due to the “innate motive” respondents
     have to lie about wrongdoing (Suppl. App. at 61), while failing to consider
     obvious motives Jane might have to lie about the extent to which she initiated
     or invited the sexual encounter with Plaintiff, such as her new boyfriend’s
     insistence that she report the incident as well as his presence at her initial
     reporting and subsequent interviews;
   • selectively determined which post-encounter evidence they would consider
     relevant (e.g., considering Jane’s allegation that Plaintiff offered her Aderall
     after the encounter in assessing Plaintiff’s credibility but not considering
     Jane’s inconsistent statements on whether the two saw each other after the
     encounter in assessing her credibility);
   • allowed Jane’s boyfriend to act both as Jane’s support person who was present
     at her interviews and as a fact witness who provided information in the
                                          36
       proceeding to corroborate Jane’s story and to impeach the testimony of
       witnesses who contradicted her story, in violation of DU’s policies;
   •   selectively viewed Jane as “heavily intoxicated,” implicitly rejecting
       Plaintiff’s and his roommate’s statements that Jane exhibited no indication of
       intoxication in order to support a finding that Plaintiff coerced Jane into sex
       (Suppl. App. at 58) but then accepting Plaintiff’s and his roommate’s
       statement in order to find that Jane’s intoxication had little effect on her ability
       to accurately recollect the encounter that night;
   •   faulted Plaintiff for making corrections to his summary statement and used it
       to attack his credibility, despite expressly inviting Plaintiff to make such
       corrections and apparently violating DU’s informal policy allowing
       interviewees to correct summary statements in order to accurately reflect their
       testimony;
   •   emphasized inconsistencies in Plaintiff’s and his roommate’s story while
       disregarding numerous inconsistencies in the versions of the story told by Jane
       and her friend;
   •   suggested Plaintiff’s failure to recollect details was indicative of deception and
       guilt while suggesting Jane’s failure to recollect details was the result of
       intoxication;
   •   viewed Plaintiff’s roommate’s statements corroborating Plaintiff’s story as
       tainted by Plaintiff’s and his roommate’s prior conferral regarding the events
       of that night, while not applying this same logic to the statements of Jane’s
       friend who corroborated Jane’s story, even though Jane called her friend
       specifically to relate to him “her portrayal of the night” and to tell him “that it
       was rape” (Appellant’s App. at A229);
   •   attacked Plaintiff’s and his roommate’s credibility on the grounds they seemed
       overly eager to offer consistent denials of any on-campus alcohol use, without
       applying the same logic to the vague and inconsistent stories provided by Jane
       and her friend regarding their own on-campus alcohol use, even though DU
       offers amnesty to complainants who admit to on-campus drug and alcohol use,
       but not to respondents.

       A few procedural irregularities in this vein are not necessarily uncommon or
even all that troubling. After all, sexual-misconduct investigations and proceedings
will not be perfect. But an accumulation of irregularities all disfavoring the
respondent becomes deeply troubling because benign, stochastic explanations for the
errors become implausible. Instead, it looks more like a railroading. Patterns of
procedural irregularities like this become even more troubling when, as in the case of
DU’s investigative model, the investigators committing such errors are also the
finders of fact on the ultimate issue of whether the alleged sexual misconduct
                                           37
Nevertheless, these concerns do not alter the obligation of a Title IX plaintiff

opposing summary judgment to adduce evidence from which a reasonable factfinder

could infer that the school’s proceeding was motivated by considerations of gender.

We will therefore affirm the grant of summary judgment to Defendants.

                                     *     *      *

        We AFFIRM the district court’s summary judgment order dismissing with

prejudice Plaintiff’s Fourteenth Amendment Due Process and Title IX claims and

dismissing without prejudice his state-law claims and his claim for declaratory

relief. 19

        Judge BACHARACH joins the opinion except for footnote 18.




occurred. Indeed, permitting, or even encouraging, an investigator who also acts as
inquisitor, judge, and jury to harbor an anti-respondent bias is repugnant to basic
notions of due process and substantial justice. However, as deeply troubling as this
kind of bias may be, it is simply not proscribed by Title IX, which only prohibits
discrimination “on the basis of [gender].” 20 U.S.C. § 1681(a).
        19
          Plaintiff argues that the district court should have retained supplemental
jurisdiction over his state-law claims because it erred in dismissing his federal-law
claims. Because we conclude that the court properly dismissed the federal-law
claims, we see no error in the court’s decision to decline supplemental jurisdiction
over the state-law claims.
                                              38