concurring in part and dissenting in part.
Although I agree with much of the majority’s opinion, I respectfully dissent from its conclusion that the district court: (1) improperly denied Conti the opportunity to depose AAM CEO Richard Dauch and (2) erred in granting summary judgment for AAM on Conti’s gender-discrimination claims.
I.
As the majority acknowledges, a district court holds broad discretion over discovery decisions, and “[a]n order denying further discovery will be grounds for reversal only if it was an abuse of discretion resulting in substantial prejudice.” Bush v. Dictaphone Corp., 161 F.3d 363, 367 (6th Cir.1998). Because I believe that the district court acted well within its discretion in denying Conti an opportunity to depose Dauch, I would affirm the court’s discovery order.
The record fails to corroborate an argument that Dauch played a significant deci-sionmaking role either in Conti’s alleged 2003 demotion or the 2004 transfer to the managerial position in the New Model Launch division. Conti contends that defense counsel questioned her at length about Dauch, asking “scores of questions” over forty-eight pages of her deposition. But Conti overstates the scope and value of these questions. Defense counsel merely elicited that Dauch met and hired Conti in January 1997 and that Conti’s late father had a professional relationship with Dauch. Conti does mention a meeting *921with Dauch during her first year at AAM, which she initiated when AAM required her to switch managerial positions, but Conti explicitly disavows that employment transfer as the basis for her current lawsuit. Neither Conti’s briefing nor the majority’s analysis attach Conti’s meeting with Dauch to the adverse employment actions that prompt her current litigation.
The majority concedes that “the record supports AAM’s claim that Conti never reported directly to Dauch,” but maintains that the record fails to support a “broader suggestion that Dauch has not been involved in decisions relating to Conti’s employment.” Maj. Op. at 904. Specifically, the majority points out that Conti’s demotion occurred after she informed AAM’s in-house counsel that one of her superiors, Roy Langenbach, instructed her not to hire two female applicants who Conti judged to be “very capable candidates” because “one was a mother [and another] one was overweight.” Although the majority claims that “the record suggests that Dauch was aware of this information,” the only record evidence the majority points to is Daueh’s acknowledgment — in answering Conti’s interrogatories- — -that he “retained trial counsel reports to AAM staff counsel [and that] AAM staff counsel [sent him] periodic updates on litigation matters.” But in the same set of responses, Dauch also stated that he did not read any witness deposition transcripts related to the case and that he had no knowledge through any other source as to what prior witnesses had said on the record. Unlike the majority, I would not hold that from these brief comments, “it seems to follow that [Dauch] would have been aware of Conti’s decision to testify against the company’s interests.” Maj. Op. at 904-05.
The majority also holds that “contrary to AAM’s portrayal of Dauch as an aloof, hands-off CEO, significant record evidence attests to Dauch’s personal involvement in and control over all aspects of his company.” Maj. Op. at 905. In support, the majority points to a business article that characterized Dauch as “run[ning] his empire like a front-line general, relying on a mixture of fear and respect to craft an unlikely success story out of a business the skeptics long ago were ready to write off.” But the deposition testimony that the majority suggests “confirms” that description is insubstantial. Even though Earhart alluded to some trepidation in writing a performance review about Conti due to Dauch’s relationship with Conti’s father, he also conceded that he didn’t know whether Dauch would ever see the performance review in the normal course of business. And to the extent that Conti cites record evidence suggesting that Dauch defines the company’s goals and objectives, determines bonus structure, and approves final promotion decisions, these actions represent typical CEO responsibilities and have no real bearing on decisions about Conti.
Although the majority comments that the record evidence, taken cumulatively, “casts serious doubt on the notion that Dauch was not aware of and played no role in any decision relevant to this action,” Maj. Op. at 906, it bears repeating that our review is guided by a different metric. The question before us is not simply whether further discovery might provide a more thorough record; our analysis asks whether the district court abused its discretion to the point of causing substantial prejudice. See Bush, 161 F.3d at 367.
The majority’s analysis virtually ignores the prejudice aspect of not deposing Dauch, much less the “substantial” prejudice required to reverse an order denying discovery. See id. at 367. In his answers to Conti’s interrogatories, Dauch emphasized that the decision to demote Conti arose “without [his] involvement,” and in *922his sworn affidavit, Dauch testified that he had “no personal knowledge of any alleged gender discrimination against Ms. Conti,” that he had “never been Ms. Conti’s direct or indirect supervisor,” that he had “no independent or unique knowledge of employment decisions made with respect to Ms. Conti’s employment with AAM,” and that deposing him would be “unduly burdensome, harassing, and annoying because [he would] be distracted from [his] duties and responsibilities.” See Thomas v. Int’l Bus. Machs., 48 F.3d 478, 483 (10th Cir.1995) (accepting the affidavit of IBM’s Chairman as support for his lack of direct knowledge of the facts in dispute). The district court acted well within its “broad discretion” when it denied a deposition for a CEO who “had no knowledge as to facts pertinent to [Conti’s] action.” Lewelling v. Farmers Ins. of Columbus, Inc., 879 F.2d 212, 218 (6th Cir.1989). I would affirm the district court’s discovery decision.
II.
Reviewing de novo, I would also affirm the district court’s grant of summary judgment for AAM on Conti’s gender-discrimination claims. See Tucker v. Tennessee, 539 F.3d 526, 531 (6th Cir.2008) (applying de-novo review).
A.
The majority claims that Conti provided sufficient evidence — both direct and circumstantial — to support her ELCRA gender-discrimination claims. First, Conti argues that the record contains direct evidence of AAM’s gender discrimination. Namely, that (1) Dauch authored a book in which he expressed admiration for Japanese and German cultures, where, “[i]f you are a woman ... you will not get equal consideration;” (2) co-Director Larry Earhart allegedly told her that she could not be promoted because she didn’t “have a penis;” (3) one of Conti’s superiors, Roy Langenbach, allegedly instructed her not to hire overweight women or mothers; and (4) statistical evidence illustrates that AAM has yet to hire a female Executive Director, Plant Manager, or Vice President, and that only a small percentage of AAM’s lower-level executives are female.
But these claims do not constitute evidence that “requires the conclusion that unlawful discrimination was at least a motivating factor in the employer’s actions.” Jacklyn v. Schering-Plough Healthcare Prods. Sales Corp., 176 F.3d 921, 926 (6th Cir.1999). As discussed above, Conti does not point to evidence suggesting that Dauch acted as a decision-maker with respect to her career. Moreover, the relevant passages in Dauch’s Passion for Manufacturing merely describe the “geography, demographics, and cultures of Germany and Japan,” and even label the enforced homogeny as “intolerance.” These statements bear little relevance to Conti’s situation in specific and fail to substantiate discriminatory management practices. As for Earhart’s alleged comment that Conti did not receive a 2001 promotion because she “did not have a penis,” the record citation provided by both Conti and the majority contains no support for (or even reference to) this statement. Langen-bach’s comment about overweight women or mothers likewise does not rise to the level of direct discrimination because it does not require the conclusion that gender discrimination “was at least a motivating factor.” See Jacklyn, 176 F.3d at 926; Cushman-Lagerstrom v. Citizens Ins. Co. of Am., 72 Fed.Appx. 322, 331-32 (6th Cir.2003) (affirming summary judgment where “disparaging remarks” about “female employees who were overweight” did not constitute direct evidence of gender discrimination). Moreover, Langenbach’s comments cannot constitute direct evidence because Conti fails to argue Langen-bach’s involvement in the employment de*923cisions she challenges. See Cushman-Lagerstrom, 72 Fed.Appx. at 332 (affirming summary judgment on a gender-discrimination claim where alleged comments did not come from a “primary decision-maker” in the plaintiffs termination). Finally, Conti’s statistical evidence is inappo-site. Although she argues that AAM never hired female as a top-ranking executive, and that only a small percentage of low-ranking executives are female, her explanation of what constitutes a “top-ranking” executive is entirely subjective — AAM hired Conti herself directly into executive ranks — and the statistical evidence does not analyze any component of AAM’s hiring process. The statistical evidence is thus unreliable as direct proof of gender discrimination. See id. at 331 (rejecting statistical evidence as unpersuasive where the factors analyzed proved unhelpful). Conti cannot demonstrate that the record supplies direct evidence that AAM discriminates on the basis of gender.
Conti similarly fails to support her claim via circumstantial evidence. As the majority acknowledges, when considering circumstantial evidence of discrimination, this court employs the familiar burden-shifting framework articulated by the Supreme Court in McDonnell Douglas Corp. v. Green, 411 U.S. 792, 93 S.Ct. 1817, 36 L.Ed.2d 668 (1973). Conti bears the initial burden of establishing a prima facie case of discrimination. If she succeeds in doing so, the burden shifts to AAM to articulate a legitimate, nondiscriminatory reason for its actions. Finally, should AAM satisfy its burden, the burden returns to Conti to prove by a preponderance of the evidence that the legitimate reasons offered were only a pretext for discrimination. Id. at 802, 93 S.Ct. 1817; DiCarlo v. Potter, 358 F.3d 408, 414-15 (6th Cir.2004).
To establish a prima facie case of gender discrimination, Conti must demonstrate that she is: (1) a member of the protected class; (2) subject to an adverse employment action; (3) qualified for the job; and (4) treated differently than similarly situated male employees for the same or similar conduct. Humenny v. Genex Corp., 390 F.3d 901, 906 (6th Cir.2004). Although Conti contends that AAM subjected her to a variety of adverse employment actions— failing to promote her, demoting her in June 2003, paying her an incommensurate salary, and failing to train her — she waived all but her failure-to-promote claim by failing to argue them in the district court. In fact, the district court only addressed Con-ti’s ELCRA claims as failure-to-promote and retaliation claims. See St. Marys Foundry, Inc. v. Employers Ins. of Wausau, 332 F.3d 989, 996 (6th Cir.2003) (noting that this court exercises “discretion to rule on an issue not decided below only in exceptional cases”) (citation and internal quotation marks omitted).
Conti’s failure-to-promote claim fails on the first prong of the McDonnell Douglas analysis, largely for the reasons set forth in the district court’s opinion. Specifically, Conti’s allegations that AAM neglected to promote her to Vice President or Director positions in the Procurement and Materials Departments cannot support a prima facie case of gender discrimination because she cannot demonstrate that AAM treated her differently than similarly situated male employees. See Humenny, 390 F.3d at 906 (“To show that she was treated differently than similarly situated males for the same or similar conduct, Appellant must show that ‘all relevant aspects’ of her employment situation are ‘nearly identical’ to those of the alleged similarly situated male employees.”).
To be “similarly situated,” the male employees with whom Conti compares herself “must have dealt with the same supervisor, have been subject to the same standards, and engaged in the same conduct without *924such differentiating or mitigating circumstances that would distinguish their conduct or the employer’s treatment of them for it.” Id. at 906 (internal citation and quotation marks omitted). But, as the district court rightly noted, Conti does not have a similar level of education or professional experience when compared to those who held Vice President or Director positions in her respective departments. Unlike others named to higher executive positions, Conti did not have graduate-level degrees in engineering or other technical fields (she holds an undergraduate degree in communication and advertising and an MBA). Although the majority joins Conti in disputing the relevance of these considerations for the VP of Procurement position, the fact that the two men who held the position during the period in question, Abdallah Shanti and Alberto Satine, both had advanced education in specialized engineering fields, corroborates the relevance of those professional qualifications for the position. I would conclude that Conti failed to set forth a prima facie case of ELCRA gender discrimination because she does not prove that AAM treated similarly situated male employees differently.
B.
Conti next alleges that her June 2003 transfer to the position of SEIT-New Model Launch in the Materials Department constituted retaliation because it occurred shortly after she told outside counsel in an unrelated discrimination case that Langenbach told her not to hire women who were overweight or mothers. The meeting with outside counsel occurred on April 11, 2003. To support a prima facie retaliation claim, Conti must demonstrate that: (1) she engaged in protected activity; (2) AAM knew of the protected activity; (3) AAM took an adverse action against her; and (4) a causal connection exists between the protected activity and the adverse action. Niswander v. Cincinnati Ins. Co., 529 F.3d 714, 720 (6th Cir.2008).
Without discussing or arguing against the district court’s conclusion that Conti failed to demonstrate that Shanti, her direct supervisor at the time, knew of her protected conduct, the majority relies largely on the necessity of deposing Dauch to vacate the district court’s decision on Conti’s retaliation claim. As set forth above, I disagree with the majority’s position on the discovery order; and regardless, the district court properly disposed of Conti’s retaliation claim.
In support of her claim, Conti points to: (1) the fact that Earhart retained his Director position for nearly two years despite problems with his performance; and (2) the close temporal relationship between her April 11, 2003 meeting with outside counsel and her June 9, 2003 transfer. But, as the district court concluded, these arguments cannot sustain Conti’s retaliation claim.
Earhart’s role as a Director prior to becoming co-Directors with Conti is not relevant to his later role as co-Director. And once the two became co-Directors, Shanti transferred both on the same day. In fact, AAM transferred Earhart to a non-executive position (without eligibility for bonuses), while AAM placed Conti in a different executive-level position.
Furthermore, without more, temporal proximity “does not demonstrate a causal connection between the protected activity and any adverse employment action.” West v. Gen. Motors Corp., 469 Mich. 177, 186, 665 N.W.2d 468 (2003); see Arendale v. City of Memphis, 519 F.3d 587, 606 (6th Cir.2008) (“[Tjemporal proximity alone will not support an inference of retaliatory discrimination when there is no other compelling evidence.”) (citation and internal quotation marks omitted). This court previously held that two months is not *925necessarily a sufficiently close temporal connection. See Arendale, 519 F.3d at 606 (rejecting the plaintiffs claim that a two-month proximity between an EEOC charge and an alleged adverse employment action constituted — without more— sufficient evidence of causation); Hafford v. Seidner, 183 F.3d 506, 515 (6th Cir.1999) (deeming “two to five months” to be a “loose temporal proximity ... insufficient to create a triable issue [on causation]”). I would affirm the district court’s rejection of Conti’s retaliation claim on summary judgment.
C.
Next, Conti alleges that she is able to support a prima facie claim under the Equal Pay Act because she received a lower salary than her immediate male predecessor after AAM promoted her to the position of Manager of the New Model Launch department. Specifically, when she took on the New-Model-Launch position, she received a $117,000 annual salary, while Mike Pannucci received $171,000 before vacating his role as Director of the New Model Launch department.
The EPA prohibits employers from paying two employees who perform equal work at unequal rates of pay on the basis of gender. 29 U.S.C. § 206(d)(1). As the majority notes, establishing a prima facie EPA claim requires Conti to demonstrate that AAM pays male employees at a higher rate “for equal work on jobs the performance of which requires equal skill, effort, and responsibility, and which are performed under similar working conditions.” Corning Glass Works v. Brennan, 417 U.S. 188, 195, 94 S.Ct. 2223, 41 L.Ed.2d 1 (1974). Notably, this court assesses whether a job is “substantially equal” for EPA purposes on a case-by-case basis, and does not require that jobs be identical for a plaintiff to make out an EPA claim. Beck-Wilson v. Principi 441 F.3d 353, 359-60 (6th Cir.2006). And unlike Conti’s Title VII claims, the EPA does not require a plaintiff to show discriminatory intent. Id. at 360. If Conti is able to set forth a prima facie case, AAM may still justify the wage difference under any of four affirmative defenses: (1) a seniority system; (2) a merit system; (3) a system that measures earnings by quantity or quality of production; or (4) any other factor other than gender. Id. at 360 (citing Buntin v. Breathitt County Bd. of Educ., 134 F.3d 796, 799 (6th Cir.1998)). I conclude that the district court properly rejected Conti’s EPA claim.
Conti fails to make out a prima facie claim because she compares her job with other jobs that are not “substantially equal” for EPA purposes. Relying on Buntin, the majority holds that Conti met her prima facie burden “by demonstrating a wage differential between herself and her predecessor” after she became Manager of the New Model Launch department. 134 F.3d at 799. But the majority reads Buntin out of context. The Buntin court held that a wage differential is sufficient where a plaintiff also demonstrates that “substantial similarity” exists between the compared jobs. See id. (citing to a case requiring “substantial similarity” between the jobs held by the plaintiff and her male predecessor). In other words, Buntin merely affirms the requirement of substantial similarity; the majority incorrectly suggests that a wage differential — without more — suffices to support a prima facie EPA claim.
Conti argues that as the Manager of the New Model Launch department, she “performed the same work Pannucci performed or more and with fewer resources.” Appellant’s Br. at 47. In support, Conti highlights that she handled Sarbanes-Ox-ley audits and IT system administration— duties Pannucci did not have — without the benefit of Pannucci’s two direct reports. *926But Conti gives us no reason why audits and IT system administration should warrant a different pay scale, and viewing the facts in Conti’s favor does not require us to credit “bald assertions” or legal conclusions. Anspach ex rel. Anspach v. City of Philadelphia, 503 F.3d 256, 260 (3d Cir.2007); see also Burch v. City of Nacogdoches, 174 F.3d 615, 621 n. 11 (5th Cir.1999) (“Construing the facts in the light most favorable to Burch does not require us to credit otherwise unsupported assertions.”).
Moreover, as the district court pointed out in rejecting Conti’s EPA claim, Conti “was paid more in the New Model Launch position than Satish Raheja, who held the position [as a Manager] for more than eight years before she did, and whose tenure in the position ended less than a year before [Conti] took it over.” Furthermore, Pannucci had an engineering background and over forty years of experience in the industry, and when he transferred to the role as Director of New Model Launch, AAM merely preserved his previous salary. And although Conti concedes that Pannucci added responsibilities to Raheja’s job, which accounted for the adjusted pay and different role, she fails to note that Pannucci also deemed Conti incapable of fulfilling the job responsibilities he took on when he acted as the Director of New Model Launch, saying:
She didn’t have the knowledge base or skill sets to ask the right questions, to look at [mechanical] processes and be able to tell if they were functioning properly, if they were doing the right job, if — I mean, it was just based on product and process knowledge. She was — she was good at what the job was originally before I came into it, it was tracking more or less the trail, the paperwork trail, was this — was the billing material done? Was the purchase orders put in place? You know, those kinds of things. But to look beyond that, to look at the tooling, to look at the equipment, to look at the fixtures, to look at the design and question what this was going to mean to the plant, what it was going to mean to the tooling, what it was going to mean to the equipment, what it was going to mean to the financial bottom line. She did not comprehend and she couldn’t ask those questions.
I would conclude that Conti fails to demonstrate her prima facie case because she cannot prove that she shouldered the same responsibilities as Manager as did Pannuc-ci as Director.
III.
I respectfully dissent from the majority’s decision to reverse the district court on Conti’s discovery claim and summary judgment claim.