Full opinion text
MEMORANDUM AND ORDER ATLAS, District Judge. This is a Title VII employment discrimination case. Plaintiffs Ed Coleman, Carl McClore, Charlotte Allen, Reginald Carter, Tony Williams, Gaines Owens and Joseph Sloan contend that they have suffered disparate treatment on account of their race while employed at Defendant Exxon Chemical Corporation (“Defendant” or “Exxon”). Specifically, Plaintiffs allege that the competitive ranking system Exxon uses to rate its employees and determine their salary allows the racial and/or gender biases of supervisor-rankers to run unchecked. Plaintiffs contend that their rankings have been artificially depressed due to unlawful animus on the part of some of the rankers, and that they have suffered a corresponding loss of pay. Defendant has moved for summary judgment as to all Plaintiffs. After extensive briefing by the parties and oral argument, the motions are ripe for adjudication. Having considered the parties’ briefs, all matters of record and the applicable authorities, the Court concludes that Defendant’s motions for summary judgment as to all Plaintiffs should be granted. I. BACKGROUND All of the Plaintiffs work at Exxon’s Baytown Olefins Plant (“BOP”) in Bay-town, Texas. By and large, Plaintiffs began working at BOP in the late 1970s. All hold or have held the position of first-line supervisor (“FLS”). FLSs, as well as various “specialists,” are grouped together for ranking purposes (the “FLS/Specialist rankings”). McClore, Allen, Williams and Sloan currently are FLSs. Owens moved from FLS to training coordinator in the late 1980s. However, he refers to himself as a FLS and continues to be included in the FLS/Specialist rankings. Carter moved from a FLS position into contract administration in 1998, as did Coleman in 1999. At BOP, a FLS’s job performance is evaluated in two ways. First, each FLS receives a personal evaluation from the second-line supervisor (“SLS”) who directly supervises him or her. The evaluation, or “performance and coaching” (“PAC”) worksheet, is a standardized form listing various performance criteria in which the employee is rated “better” than, “similar to,” or “below” most. The PAC also provides room for the SLS’s comments. Second, all FLSs and comparable specialists are competitively ranked against each other. The rankings are performed by SLSs, with review and supervision from human resources and senior management at BOP. A new rank list is generated each year. The rank list is divided into five quintiles, the first quintile being the highest. A FLS’s quintile is a significant factor in determining his salary. Although only the quintile ranking impacts salary, the forty-odd FLSs at the plant are given individual rankings from a high of 99 to a low of 1. Although Plaintiffs were aware that a competitive ranking system was in place prior to 1998, FLSs were not informed of their exact place on the rank list. In May 1998, an employee found a copy of the 1998 draft rank list on the company computer system. The list was quickly circulated among the FLSs, many of whom were dismayed to learn their rankings after receiving predominantly positive evaluations in their individual PAC forms. The following chart shows each Plaintiffs rankings for the years 1996 to 2000: _1996_1997_1998_1999_2000 Coleman_79_75_65_60_69 McClore_35_38_43_42_35 Allen_11_19_14_12_20 Carter_33_22_28_20_N/A Williams_55_54_49_47_62 Owens_30_22_14_12_13 Sloan_N/A_N/A_40_30_24 See Redacted Rank Lists (Ex. D to McClore Motion). After the disclosure of the rank list, several African-American FLSs met to discuss their perception that minorities and women were clustered in the bottom quintiles of the rank list. While the average rank on the 1998 rank list for all FLSs was 52, the average rank for minorities was 38 and the average rank for women was 14. See Skewing Data (Ex. D to Coleman/McClore Response). Separate data (in which women appear to be grouped with minorities) show the quintile distribution of the sixteen minority FLSs: First Quintile 1 employee Second Quintile 2 employees Third Quintile 4 employees Fourth Quintile 4 employees Fifth Quintile 5 employees See Profile of 1998 Ranking of First Line Supervisors, attachment to Affidavit of Carl McClore (Ex. C to Coleman/McClore Response) (“McClore Affidavit”). On June 30, 1998, several FLSs sent a letter to the plant manager at BOP, expressing their dissatisfaction with the rankings of minority FLSs and requesting an investigation into the ranking system. See Letter to D.G. Blake, June 30, 1998, attachment to McClore Aff. In response, the plant manager Del Blake and Human Resources Manager Gail Culberson met with a group of minority employees. See Notes of Meeting, July 28, 1998 (Ex. 25 to McClore Motion). Culberson later conveyed the employees’ concerns with the ranking system to the SLSs. See Affidavit of Gail Culberson (Ex. A to McClore Motion) (“Culberson Affidavit”), ¶ 26. In addition, Blake commissioned a task force to examine the current ranking system and make suggestions for improvement. Cul-berson Aff., ¶ 27. Both Coleman and McClore were members of the task force. Id. The task force made a number of recommendations, many of which were adopted. See Ranking Implementation Team Summary of Key Recommendations (Ex. 137 to McClore Motion) (showing which recommendations of the task force were implemented). However, Defendant and Plaintiffs agree that no significant changes were made as a result of the task force: in Defendant’s view, the concept of competitively ranking FLSs remains intact; and in Plaintiffs’ view, problems of gender and racial bias persist. See Culberson Aff., ¶ 32; Coleman Aff., ¶ 8; McClore Aff., ¶ 6. Plaintiffs filed a charge with the Equal Employment Opportunity Commission (“EEOC”) on February 24, 1999. After receiving a right to sue notice, Plaintiffs filed this case on December 30, 1999. See Complaint [Doc. # 1]. Plaintiffs allege that the ranking system disparately impacted minorities and women. Id., ¶¶ 7, 12. Plaintiffs also allege in their Complaint that because the ranking system allowed the racial biases of SLSs to run unchecked, they have suffered disparate treatment and pattern and practice race discrimination. Id., ¶¶ 7, 8. Plaintiffs contend that their depressed rankings resulted in lost pay and promotions. Id., ¶¶ 8, 10. Plaintiffs also originally alleged hostile environment and retaliation claims. Id., ¶ 15. At the Court’s request, Plaintiffs subsequently clarified their claims. See Supplemental Response, at 1-5. Plaintiffs now raise only disparate treatment claims on the basis of race (and, for Charlotte Allen, gender). Id. at 1. Gaines Owens also contends he has suffered retaliation since complaining to BOP management about the ranking system’s impact on minorities. Id. Plaintiffs explicitly have relinquished any disparate impact, pattern and practice, and hostile environment claims. Id. at 2 n. 1, 3 n. 3. In addition, Plaintiffs no longer contend that they were denied promotions as a result of artificially low rankings. Id. at 2. 11. SUMMARY JUDGMENT STANDARD In deciding a motion for summary judgment, the Court must determine whether “the pleadings, depositions, answers to interrogatories, and admissions on file, together with the affidavits, if any, show that there is no genuine issue as to any material fact and that the moving party is entitled to judgment as a matter of law.” Fed. R. Civ. P. 56(c); Heritage Bank v. Redcom Laboratories, Inc., 250 F.3d 319, 325 (5th Cir.2001). Material facts are those facts that might affect the outcome of the suit under the governing law. Martin’s Her-end Imports, Inc. v. Diamond & Gem Trading U.S.A. Co., 195 F.3d 765, 773 (5th Cir.1999). “Factual controversies are construed in the light most favorable to the nonmovant, but only if both parties have introduced evidence showing that an actual controversy exists.” Edwards v. Your Credit, Inc., 148 F.3d 427, 432 (5th Cir.1998). Summary judgment is inappropriate if there is a “dispute in the substantial evidence, that is, evidence which is of such quality and weight that reasonable and fair-minded men in the exercise of impartial judgment might reach different conclusions. Consequently, a mere scintilla of evidence is insufficient to present a question for the jury.” Chaney v. New Orleans Public Facility Mgmt., 179 F.3d 164, 167 (5th Cir.1999). A nonmovant may not defeat summary judgment with allegations or denials in the pleadings or with “unsubstantiated or eonclusory assertions that a fact issue exists.” See, e.g., Morris v. Covan World Wide Moving, Inc., 144 F.3d 377, 380 (5th Cir.1998). In the absence of any proof, the court will not assume that the nonmovant could or would prove the necessary facts. Little v. Liquid Air Corp., 37 F.3d 1069, 1075 (5th Cir.1994) (en banc). III. DISCUSSION Plaintiffs contend that, as a result of race and gender discrimination at BOP, their salaries have been artificially depressed. In response, Defendant contends that Plaintiffs’ salaries are determined by their rankings, and that these rankings are the result of a facially neutral, and indeed explicitly anti-discriminatory, process. Plaintiffs have replied with the following types of evidence: (1) evidence that the ranking system allows for SLS racial bias; (2) evidence that Plaintiffs were ranked significantly lower than “similarly situated” White FLSs; (3) statistical evidence purporting to show a “gross disparity” between the rankings of White male FLSs and minority and female FLSs; and (4) evidence of racially charged remarks and incidents involving various SLSs. A. Analytical Framework Title YII prohibits employers from discriminating “against any individual with respect to his compensation, terms, conditions, or privileges of employment, because of such individual’s race, color, religion, sex, or national origin.” 42 U.S.C. § 2000e-2(a)(1). A plaintiff may prove a claim through direct evidence or circumstantial evidence. See Russell v. McKinney Hospital Venture, 235 F.3d 219, 222 (5th Cir.2000); Urbano v. Continental Airlines, 138 F.3d 204, 206 (5th Cir.1998). “Direct evidence is evidence which, if believed, proves the fact [of intentional discrimination] without inference or presumption.” Brown v. East Mississippi Electric Power Ass’n, 989 F.2d 858, 861 (5th Cir.1993); accord Brady v. Ft. Bend County, 145 F.3d 691, 712 (5th Cir.1998). When a plaintiff presents credible direct evidence that discriminatory animus in part motivated or was a substantial factor in the contested employment action, the burden of proof shifts to the employer to establish by a preponderance of the evidence that the same decision would have been made regardless of the forbidden factor. Brown, 989 F.2d at 861. A plaintiff can also prove a case of race discrimination through circumstantial evidence, using the familiar burden-shifting framework established by McDonnell Douglas Corp. v. Green, 411 U.S. 792, 93 S.Ct. 1817, 36 L.Ed.2d 668 (1973). See Russell, 235 F.3d at 222. The plaintiff first must establish a prima facie case of race discrimination. The employer must then respond by producing a legitimate, nondiscriminatory rationale for its decision. Once the employer offers a legitimate, non-discriminatory reason for the plaintiffs termination, the burden shifts back to the plaintiff to “raise a genuine issue of material fact as to whether the employer’s proffered reason was merely a pretext for [discrimination].” Medina v. Ramsey Steel Co., Inc., 238 F.3d 674, 680 (5th Cir.2001). “[A] plaintiffs prima facie case, combined with sufficient evidence to find that the employer’s asserted justification is false, may permit the trier of fact to conclude that the employer unlawfully discriminated.” Reeves v. Sanderson Plumbing Products, Inc., 530 U.S. 133, 148, 120 S.Ct. 2097, 147 L.Ed.2d 105 (2000). “The ultimate determination, in every case, is whether, viewing all of the evidence in a light most favorable to the plaintiff, a reasonable factfinder could infer discrimination.” Crawford v. Formosa Plastics Corp., 234 F.3d 899, 902 (5th Cir.2000) (citing Reeves, 530 U.S. at 142, 120 S.Ct. 2097). “In making this determination, a court should consider ‘the strength of the plaintiffs prima facie case, the probative value of the proof that the employer’s explanation is false, and any other evidence that supports the employer’s case... I” Id. (citing Reeves, 530 U.S. at 148-49, 120 S.Ct. 2097). B. Plaintiffs’ Prima Facie Case A plaintiff may state a prima facie case of race (or sex) discrimination by showing that (1) he was a member of a protected class; (2) he was qualified for the benefit or promotion he sought; (3) that he was denied the benefit or promotion, and (4) that similarly situated employees outside the protected class were treated more favorably (ie., that the employment decision of which he complains was differentially applied to him). Rubinstein v. Administrators of Tulane Educ. Fund, 218 F.3d 392, 399 (5th Cir.2000), cert. denied, — U.S. —, 121 S.Ct. 1393, 149 L.Ed.2d 316 (2001). Defendant initially contended that Plaintiffs could not meet the fourth prong of the prima facie test because they had no evidence that the ranking process was “differentially applied” to them. See Coleman Motion, at 12. Plaintiffs responded, without further explication, that they “did not receive the same salary increases as did similarly or less qualified White employees.” Allen, et al., Response, at 27. Plaintiffs later produced a list of White male FLSs that they contend are similarly situated because the PAC evaluations of the White male FLSs contain ratings and comments similar to those contained in Plaintiffs’ PAC evaluations. Defendant contends that the specifically identified White male FLSs are not proper comparators, first because most had different supervisors than Plaintiffs and thus the PAC evaluations were filled out by different persons; and, second, because the PAC evaluations constitute individualized feedback with little or no bearing on the comparative rankings at issue in this case. It is well established that a plaintiff may support a claim of disparate treatment by showing that his employer gave “preferential treatment to [another] employee under ‘nearly identical’ circumstances.” Okoye v. Univ. of Texas Houston Health Sci. Ctr., 245 F.3d 507, 514 (5th Cir.2001) (quoting Little v. Republic Refining Co., Ltd., 924 F.2d 93, 97 (5th Cir.1991)). The “nearly identical” standard, when applied at the McDonnell Douglas “third step,” the pretext analysis, is a stringent standard. Employees with different responsibilities, different supervisors, different capabilities, different work rule violations or different disciplinary records are not considered to be “nearly identical.” See Okoye, 245 F.3d at 514-15 (and cases cited therein). The Fifth Circuit has held that the “nearly identical” standard applies equally to the fourth prong of the prima facie case. See Williams v. Trader Publishing Co., 218 F.3d 481, 484 (5th Cir.2000); Bennett v. Total Minatome Corp., 138 F.3d 1053, 1062 (5th Cir.1998); Mayberry v. Vought Aircraft Co., 55 F.3d 1086, 1090 (5th Cir.1995). However, the Fifth Circuit has appeared reluctant to reject discrimination claims by applying the “nearly identical” standard at the prima facie stage of the analysis. See, e.g., Okoye, 245 F.3d at 513 (assuming, without deciding, that pri-ma facie burden was met and deciding case on issue of pretext); Nieto v. L & H Packing Co., 108 F.3d 621, 623 n. 5 (5th Cir.1997) (noting “confusion” in distinguishing between prima facie and pretext analyses, and “proceeding] directly to the ultimate question [of discrimination]” where “strict application of the burden-shifting framework [was] not particularly helpful”). To the extent Plaintiffs rely primarily on the individually identified comparators, the Court is unpersuaded. The Court’s analysis is set forth below in Section III.D.l (Comparator Evidence). However, given the lack of Fifth Circuit authority addressing in detail employment discrimination claims at the prima facie stage for failure to show “nearly identical” circumstances, given the lack of briefing by the parties on this issue, and given the parties’ views expressed at oral argument that the entire rank list represents Plaintiffs’ proof of their prima facie cases, the Court will make no decision as to whether Plaintiffs in fact have met their prima facie burden on this theory. The Court’s analysis on the second and third steps of the McDonald Douglas test resolves this case. Thus, the Court assumes for the summary judgment motions that Plaintiffs have met their prima facie burden. C. Defendant’s Legitimate, Nondiscriminatory Rationale Defendant contends that Plaintiffs’ rankings, and therefore salary, have not been artificially depressed as a result of racial or gender bias. Instead, Defendants contend that Plaintiffs’ rankings are simply the result of Exxon’s race and gender neutral ranking process. A detailed description of the ranking process follows. Defendant has established that during the pertinent years, FLSs have been ranked by SLSs, with review and supervision from human resources and senior management at the plant. See Culberson Aff., ¶¶ 23; Lattimer Aff., ¶¶5, 11, 14-16; Deposition of Mike Williams (Ex. I to Plaintiffs’ Response) (“M. Williams Deposition”), at 35-41, 47-52, 73-74, 77, 87, 101-107. It appears that the SLS group comprises approximately twenty individuals. See Ranker-Rankee Matrix (Ex. 47 to McClore Motion); Notes of Meeting, July 28,1998. While the exact number of SLSs presently at the plant is unclear, in the current group of SLSs, five are African-American. M. Williams Dep., at 16. At least two are female. At the time the 1998 rank list was prepared, two SLSs were African-American. Coleman Aff., ¶ 33. The ranking process is conducted in several stages. First, the human resources manager holds a “kickoff meeting” for all SLSs. See Culberson Aff., ¶ 4; Lattimer Aff., ¶ 5; M. Williams Dep., at 47-48. The kickoff meeting includes a discussion of the procedures and objectives of the ranking process, as well as a question and answer session. The kickoff meeting is the primary means of educating SLSs about the ranking process. M. Williams Dep., at 49-51. The human resources manager discusses the seriousness of the activity, as well as the need to be fair and open. The human resources manager also mentions the company’s commitment to diversity at the kickoff. Id. at 51-52. Second, each SLS assembles performance evaluation information on the FLSs who report to him. For each of these FLSs, the SLS solicits information from the FLS, as well as from the FLS’s “customers,” co-workers, subordinates, and/or contractors who have worked with the FLS. The FLS and SLS usually agree on which “customers” should be approached for feedback. In addition, the FLS is given the opportunity to provide a self-evaluation, including accomplishments, goals, training desired or areas for improvement. Once a SLS has finished gathering information, he prepares a summary (a form called a “PS-III”) which he discusses with the FLS. Culberson Aff., ¶¶ 5-6; Lattimer Aff., ¶¶ 6-7. As part of the information-gathering process, an SLS may fill out a draft PAC form. Lattimer Aff., ¶ 7. While the information-gathering process is intended to be highly interactive, it is unclear in practice how much input a FLS has. For instance, McClore contends that he was only rarely asked by his supervisors to provide a list of his accomplishments or other information for the ranking process. McClore Aff., ¶ 13. Indeed, while SLSs were encouraged to fill out PS-III forms, they were not required to do so. See Confidential Interoffice Memo regarding 1999 FLS/Specialist Rank Input, April 23, 1998 (Ex. D to Coleman/McClore Response), at unnumbered first page. Third, each SLS selects those FLSs he will rank. The SLS ranks those FLSs who report directly to him, as well as any other FLSs with whose, work he has familiarity. Lattimer Aff., ¶ 8; M. Williams Dep., at 28-29. The SLS must weigh each ranking he gives a FLS according to his familiarity with the FLS’s work using a three-point scale, “1” for a direct supervisory relationship, “2” for a functional or indirect supervisory relationship, and “3” for direct and recurring business contact. See Confidential Interoffice Memo regarding 1999 FLS/Specialist Rank Input, at unnumbered fourth page. A SLS’s rankings are not counted if he ranks fewer than three FLSs. Id. In ranking a FLS who is not a direct report, the SLS may consider the PS-III form or PAC (individualized evaluation) completed by the FLS’s direct supervisor. M. Williams Dep., at 54-56. Once he has decided which FLSs to rank and has assigned a familiarity code to each, the SLS ranks those FLSs from top to bottom according to their performance. Although SLSs are not required to use any particularized criteria in this ranking, they are encouraged to consider the performance dimensions of leadership, communication, safety/health, job knowledge and people development. See Confidential Interoffice Memo regarding 1999 FLS/Specialist Rank Input, at unnumbered third page. SLSs Tracy Lattimer and Joe Wolf both state they are generally confident with the relative ranking list they are able to produce through this process. Lat-timer Aff., ¶ 9; Affidavit of Joseph Wolf (Ex. E to McClore Motion) (“Wolf Affidavit”), ¶ 13. Fourth, the individual rank lists prepared by the SLSs are combined according to a computer algorithm. The resulting list, known as the “first pass” rank list, is distributed to all SLSs. Fifth, all SLSs meet to determine FLS rankings. The session usually lasts a full day and is held outside the plant premises. The session begins with the human resources manager making opening remarks, stressing that the rankers should only consider FLS performance for the preceding year and emphasizing Exxon’s policy on equal employment opportunity and nondiscrimination. Culberson Aff., ¶ 10. The human resources manager is present throughout the ranking session. Each SLS may bring the draft PAC or PS-III forms for the FLSs for whom he intends to give input. Lattimer Aff., ¶7; M. Williams Dep., at 64. Each FLS is considered in turn. The FLS’s direct supervisor generally gives a presentation of the FLS’s accomplishments, strengths and weaknesses. The FLS’s performance is compared with the performance of those FLSs in the same range on the first pass rank list. If the SLSs cannot differentiate the performance of two or more FLSs, they may elect to “cluster” (ie., give the same numerical ranking to) those FLSs. See e.g., M. Williams Dep., at 89-90. Because SLSs are aware that quintile rankings affect pay, the discussion is particularly intense when a quintile ranking is at stake. M. Williams Dep., at 115. The process continues until consensus among the SLSs has been reached on the entire rank list. Lattimer Aff., ¶¶ 12-13; M. Williams Dep., at 80, 104. The finished product is called the draft rank list. Sixth, the human resources staff runs “skewing data” on the draft rank list. The skewing data reveals the average ranking for each subgroup of FLSs. Subgroups are created for FLSs working in different departments of the plant, for specialists versus FLSs, and for minorities and females. Skewing data has been generated since at least 1993. M. Williams Dep., at 102. Seventh, the SLSs meet again to consider the skewing data and to give their opinions of that year’s ranking process. At that time, the SLSs may decide to adjust the rankings given the skewing data. However, such adjustments are rare. Lattimer Aff., ¶ 13; M. Williams Dep., at 103. With senior management (ie., department heads, plant manager and human resources manager) present, SLSs are given the opportunity to express their concerns about aspects of the ranking process or to explain certain ranking decisions. M. Williams Dep., at 106-07. During this process, the human resources manager offers suggestions for improving the following year’s ranking process, based on her own perceptions and the feedback gleaned from the SLSs. Id. at 35-36. Finally, members of senior management meet to finalize the rank list. Management may decide to make adjustments to the rank list, but these adjustments are generally small and infrequent. Culberson Aff., ¶ 15. Defendant contends that the process is inherently inimical to discrimination and contains safeguards to prevent unlawful biases from affecting the outcome. Defendant notes that the process requires inputs from multiple sources as well as vocal debate and explicit justification of ranking decisions among the SLSs. SLSs are reminded throughout the process to consider performance only, to focus on the single year in issue, and to keep in mind Exxon’s commitment to nondiscrimination, which is stressed both at the kickoff and ranking meetings. SLSs are aware that skewing data will be generated showing the relative positions of minority and female FLSs. Defendant contends that the practice of circulating and considering the skewing data, and SLSs’ awareness of this practice, increase SLSs’ accountability to make nondiscriminatory decisions. The record is uncontradicted that the ranking process is facially neutral. The ranking process protects against overt discrimination. Indeed, there is no evidence of a race- or gender-based comment being made during the ranking process. Defendant has thus met its burden to articulate a legitimate reason for the rankings of the Plaintiffs during the years in issue. The Court next will examine each Plaintiffs individualized evidence of anecdotal disparate treatment or discrimination, as well as Plaintiffs’ statistical evidence purporting to show disparities in the salaries of minority FLSs and White male FLSs. D. Plaintiffs’ Evidence of Discrimination, Pretext or Falsity 1. Comparator Evidence As mentioned above, Plaintiffs have submitted a list of so-called “comparators,” White male FLSs who were ranked higher than Plaintiffs, and thus allegedly received more favorable treatment. Plaintiffs claim these individuals are “similarly situated” to Plaintiffs. Plaintiffs contend the comparators are similarly situated because the comparators’ PAC evaluations contain comments and ratings similar to those contained in Plaintiffs’ PAC evaluations. See generally Comparator List. Defendant first contends that the comparisons are meaningless if a Plaintiff and his or her purported comparator did not have the same supervisor. For instance, Plaintiff Allen contends that she was “similarly situated” to six White male FLSs, all of whom were ranked above her for the years 1995 through 1998. However, as Defendant points out, none of these comparators had the same supervisor as Allen in any given year. See Comparator’s Supervisors (Ex. B to Final Submission). The same is true for the comparators produced by Plaintiffs Coleman, Owens, Sloan, and Williams. Id. The fact that a similarly situated individual outside a Plaintiffs protected class was treated more favorably than that Plaintiff may constitute probative evidence of disparate treatment. However, the more favored treatment must have occurred under “nearly identical” circumstances. Where the comparator and the Plaintiff have different supervisors, their situations are not “nearly identical.” Okoye, 245 F.3d at 514; Little, 924 F.2d at 97. Accordingly, Allen, Coleman, Owens, Sloan and Williams have failed to produce specific probative comparator evidence. The other two Plaintiffs, respectively, have identified at least some comparators who had the same supervisor as the Plaintiff. For instance, Carl McClore compares himself to Doug Olin, who had the same supervisor as McClore in 1998. Reginald Carter compares himself to Harvey Moore. and Alan Tingle. Carter, Moore and Tingle had the same supervisor in 1995 and 1996. Defendant objects to this comparative evidence because it is manifestly incomplete and because the PACs are used primarily for a counseling function, which is materially different from the comparative ranking process under challenge. The Court agrees that this comparator evidence based on PACs is materially incomplete, and has minimal, if any, probative value. First, this comparator evidence, insofar as it is designed to show substantial similarity of circumstances with Plaintiffs in the ranking process and thus to serve as circumstantial evidence of falsity or pretext in Defendant’s explanation of its ranking process, the evidence misses the mark. While a FLS’s direct supervisor may play a large part in the rank that the FLS achieves, each FLS is ranked by various SLSs in preparation for the ranking session. See supra at 14 — 17. The evidence does not reveal which SLSs’ rankings had a material effect on a Plaintiffs ultimate rank in any year. It is impossible to tell which SLSs ranked particular FLSs, how many SLSs ranked each of the Plaintiffs or their respective purported comparators, whether the same SLSs ranked a Plaintiff and his purported comparators, or how a Plaintiff fared against his comparators in those SLSs’ initial rankings. A single PAC evaluation by one supervisor, standing alone, is merely an isolated piece of evidence singularly unhelpful in showing that SLSs as a group — in the ranking process — discriminated against a Plaintiff. Plaintiffs’ efforts to establish “substantial similarity” between themselves and particular comparators, based on comments on a single PAC form, even forms about two FLSs written by one SLS, are also impeded by the nature and purpose of the PACs. The PAC form consists of a list of performance categories in which the FLS must be rated within three simple categories, “better” than, “similar to,” or “below” most other FLSs. The SLS is provided limited space to make substantive comments. The PAC form thus inherently limits the SLS’s description of the deficiencies or positive aspects of a FLS’s performance. SLS Tracy Lattimer states, [Although I and other SLSs receive a lot of training in filling out the individual performance evaluation (PAC) during various management courses we take at Exxon, I feel more confident in my ability to rank individuals competitively than to put into words a description of the quality of their performance on a PAC form.... [To] me it is more straightforward and objective to competitively rank the overall contribution of two individuals than it is to describe them separately and qualitatively. Lattimer Aff., ¶ 9. SLS Joe Wolf also notes his confidence in constructing a relative ranking for the four FLSs reporting to him. Wolf Aff., ¶ 13. Plaintiffs have not produced the testimony of any SLSs who contradict these assessments. The PAC forms constitute a performance counseling. Although Plaintiffs characterize the comments they have isolated from individual PAC forms as “articulated reasons for low ranking,” see Comparator List, passim, the PAC comments are merely suggestions for improvement. The PAC forms are prepared by FLSs’ direct supervisors. In the years on which Plaintiffs focus (1995-98), a PAC form did not necessarily reflect the views of SLSs or their third party sources of information, other than the FLS’s direct supervisor. In any event, not surprisingly, many FLSs have the same or similar suggested areas of improvement {e.g., better communication skills, increased initiative). As Defendant points out, even the two highest ranked FLSs’ PAC forms contain these bromides. See Redacted PAC forms (Ex. D to Final Submission). The fact that two differently ranked FLSs had the same recommended area for improvement says nothing about the degree of the perceived problem or the relative importance of the identified area to each FLS’s particular job duties. Two FLSs’ receipt of one, or even several, similar characterizations of areas for improvement- in a given year, therefore, does not mean that they actually are in “nearly identical” circumstances. Similarly, purported comparators identified in Plaintiffs’ affidavits based on anecdotal evidence cannot be considered “similarly situated” or “nearly identical” to the Plaintiffs. Plaintiffs’ attempts to compare their rankings to those of White FLSs who appear to have had the same performance issues as Plaintiffs, or who worked on the .same projects or committees as Plaintiffs, ultimately are unavailing in light of the holistic evaluation of each FLS’s performance by multiple sources which takes place at the ranking sessions. Plaintiffs have submitted no evidence that the issues they point to were dispositive in the rankings. Finally, Plaintiffs’ personal, subjective views of their own performance are not probative evidence of falsity or pretext in this case. Several Plaintiffs in their affidavits attempt to rebut the suggestions for improvement listed in their PACs. See McClore Aff., ¶ 12; Affidavit of Charlotte Allen (Ex. A to Allen, et al., Response) (“Second Allen Affidavit”), ¶¶ 6, 7; Carter Aff., ¶¶ 7, 8, 11-14; Owens Aff., ¶ 16; Sloan Aff., ¶¶ 14-15. The failure of this evidence lies not in its “subjective” nature per se, but in the fact that the suggestions for improvement listed in the PACs cannot be equated with “articulated reasons for low ranking.” . In summary, Plaintiffs have attempted to show that performance criticisms they received on PACs were pretextual, by identifying White male FLSs who received the same criticisms but were higher-ranked; or that the performance criticisms were false, by affidavit testimony regarding their own perceptions of the fairness of the criticisms. As discussed above, there are several flaws in Plaintiffs’ evidence. The suggestions for improvement contained in the PACs are not Defendant’s “articulated reasons” for a Plaintiffs ranking. A PAC form represents only the opinions of the SLS filling out the form, whereas a FLS is ranked by several SLSs at the ranking session. For these reasons, Plaintiffs’ comparator evidence fails to raise a fact issue as to pretext or falsity for any Plaintiff. 2. Statistical Evidence a. Applicable Legal Standard Plaintiffs rely heavily on statistical evidence they have submitted purporting to show a gross disparity between the pay of minority FLSs and White FLSs. A plaintiff may raise an inference of discrimination—ie., prove a prima facie case—if gross statistical disparities are shown in the analysis of pertinent data. EEOC v. American Airlines, Inc., 48 F.3d 164, 173 n. 9 (5th Cir.1995); Anderson v. Douglas & Lomason Co., Inc., 26 F.3d 1277, 1290 (5th Cir.1994). A showing of gross statistical disparity requires a statistical difference of at least two standard deviations. Anderson, 26 F.3d at 1292; Lopez v. Laborers Int’l Union Local # 18, 987 F.2d 1210, 1214 (5th Cir.1993) (difference greater than three standard deviations is prima facie proof that the selection system is not random); Frazier v. Garrison Indep. School Dist., 980 F.2d 1514, 1524 n. 29 (5th Cir.1993). A difference of two standard deviations is roughly equivalent to a .05 level of statistical significance. Kadas v. MCI Systemhouse Corp., 255 F.3d 359, 361-62 (7th Cir.2001). Plaintiffs have offered their statistical analysis as evidence of pretext, not to meet their prima facie burden. See Coleman/McClore Response, at 21. However, “[w]hile statistical evidence ‘may be probative of pretext in limited circumstances,’ it ‘usually cannot rebut the employer’s articulated nondiscriminatory reasons.’ ” Scott v. Univ. of Mississippi, 148 F.3d 493, 510 (5th Cir.1998), abrogated on other grounds, Kimel v. Florida Bd. of Regents, 528 U.S. 62, 120 S.Ct. 631, 145 L.Ed.2d 522 (2000) (quoting EEOC v. Texas Instruments, Inc., 100 F.3d 1173 (5th Cir.1996)). Statistics “standing alone [are] not likely to establish a case of individual disparate treatment.... To establish disparate treatment, the statistics must be accompanied by other evidence.” Rummery v. Illinois Bell Telephone Co., 250 F.3d 553, 559 (7th Cir.2001); Deloach v. Delchamps, Inc., 897 F.2d 815, 820 (5th Cir.1990). b. Analysis of Plaintiffs Statistical Evidence For the reasons discussed below, the Court concludes that Plaintiffs’ statistical evidence suffers from serious methodological flaws, and is therefore inadmissible under Rule 702 of the Federal Rules of Evidence and the standards for expert testimony under Daubert v. Merrell Dow Pharmaceuticals, Inc., 509 U.S. 579, 589, 113 S.Ct. 2786, 125 L.Ed.2d 469 (1993), Kumho Tire Co. v. Carmichael, 526 U.S. 137, 156-57, 119 S.Ct. 1167, 143 L.Ed.2d 238 (1999), and their Fifth Circuit progeny. In order to test the hypothesis that Plaintiffs suffered discrimination resulting in pay lower than that received by White male FLSs, Plaintiffs’ statistician Ronald Luke performed a regression analysis. See Report of Ronald T. Luke, Ph.D., J.D. (Ex. F to Coleman/McClore Response) (“Luke Report”). Luke attempts to determine the extent to which the independent variables of minority status and tenure (i.e., years worked at Exxon) influenced the dependent variable of salary for the years 1995 to 2000. Luke concludes that there was a statistically significant difference between the salaries of minority FLSs and White male FLSs. Luke Report, at 2-3. Plaintiffs have also submitted the Affidavit of John F. Scoggins (Ex. A to Coleman/McClore Sur-Reply) (“Scoggins Affidavit”). Scoggins’ analysis purports to show there is a statistically significant difference between the salaries of minority and White male FLSs in each year from 1995-1999, even when the data are not pooled. See Scoggins Aff., at 3^4. Defendant has introduced evidence from its own expert, Dwight Steward, who criticizes several fundamental assumptions on which Plaintiffs’ statistical evidence is based. See Affidavit of Dwight Steward (Ex. E to Coleman/McClore Reply) (“Steward Affidavit”); Second Affidavit of Dwight Steward (Ex. C to Final Submission) (“Second Steward Affidavit”). First, Steward contends that the sample size is too small to be meaningful. Second, Steward contends that Luke and Scoggins erred in comparing White male FLSs to “minority” (African-American, Hispanic and White female) FLSs. Third, Steward contends that Luke and Scoggins’ variable for “tenure” is flawed and/or irrelevant. Sample Size Between 1995 and 2000, the FLS population at BOP ranged between forty-one and fifty-one individuals. See Redacted Rank Lists. In any given year, there were between eight and ten African-American FLSs. Id. “Whether a sample is too small to yield meaningful results is a determination made by the district court on a case-by-case basis.” Anderson, 26 F.3d at 1289 n. 20. The Fifth Circuit has cautioned against the use of statistics involving small sample sizes. “The reason for [the] hesitance is obvious: ‘the smaller the sample size, the greater the likelihood that the underrepresentation reflects chance rather than discriminatory practices.’ ” Stendebach v. CPC Int’l, Inc., 691 F.2d 735, 738 (5th Cir.1982) (citing Williams v. Tallahassee Motors, Inc., 607 F.2d 689, 693 (5th Cir.1979)). In the instant case, the sample sizes, especially those for African-American FLSs, are quite small. The Court does not hold that these sample sizes are insufficient as a matter of law; however, any statistical analysis derived from such a small universe is far from conclusive and must be subjected to close scrutiny for reliability. Inclusion of Hispanic Males and White Females in Protected Group.Plaintiffs rely on a comparison of the salaries of White male FLSs to those of so-called “minority” FLSs, which Plaintiffs define as African-American, Hispanic and White female FLSs, collectively. See Luke Report, at 1. Defendant contends this methodology is improper because “historical experiences concerning discrimination have differed by race and gender.” Steward Aff., ¶ 8. More pertinently, at oral argument Defendant contended that in a race discrimination claim, it is invalid and illogical to include Hispanies and White females along with African-Americans because both Hispanies and White females are outside the protected class. The Court agrees with this latter analysis. All Plaintiffs contend they have suffered discrimination because they are African-American. Therefore, both Hispanies and White females, as individuals outside the protected class in this case, could be comparators and their more favorable treatment could constitute evidence of disparate treatment of African-Americans. Indeed, Plaintiffs implicitly concede this point when they contend that FLS Robert Arau-jo, a Hispanic male, is a comparator for Charlotte Allen. See Comparator List, at 2-3. To support their race discrimination claim, Plaintiffs must rely on data concerning only African-American FLSs. The proper comparison is between African-American FLSs and non-Hispanic White males. Plaintiffs have offered no evidence of a gross statistical disparity between African-American FLS salaries and non-Hispanic White male FLS salaries. Indeed, Scoggins himself notes that the disparity in salary for African-Americans is “never quite statistically significant” for any year between 1995 and 2000. Scoggins Aff., at 5. Defendant’s expert Steward reaches the same conclusion. Steward has produced the following table, which compares the average salaries of African-American FLSs and non-Hispanic White male FLSs: _African-American_White Male_Standard Year_FLSs_FLSs_Ratio_Deviations' 1995_$67,300_$70,900_94.9%_1.32 1996_$66,580_$71,841_92.7%_1.93 1997_$68,722_$72,043_95.4%_1.31 1998_$70,133_$72,947_96.1%_1,06 1999_$73,088_$75,254_97.1% 0.89 2000$74,522 $74,484_100.1%0.02 See Table C to Second Steward Aff. According to Steward, there was no “gross disparity” (i.e., a difference of at least two standard deviations) between the salaries of African-American and White male FLSs in any year under study. It appears that Plaintiffs’ experts were only able to obtain a statistically significant finding by including with African-American data salaries of Hispanic FLSs and White female FLSs. See Scoggins Aff., at 5. Average female FLS salaries at BOP are significantly lower than average White male FLS salaries and average African-American FLS salaries. See Scoggins Aff., at 2. As the Court has held above, however, the comparatively low salaries of Hispanic and White female FLSs do not aid Plaintiffs in their claim for discrimination on account of their race, African-American. Plaintiffs’ statistics accordingly fail to show a “gross disparity” in the treatment of African-American FLSs and White male FLSs. “Tenure” Variable. In their regression analyses, Plaintiffs’ experts Luke and Scoggins include a variable for each employee’s total number of years at Exxon. Defendant contends this tenure variable is factually unfounded and thus improper. The Court finds that Plaintiffs’ experts have based their calculations on an assumption not supported by the record. According to BOP Human Resources Manager Gail Culberson, salary “factors in years of experience as a first line or specialist then the quintile.” Culberson Dep., at 118 (emphasis added). Plaintiffs have produced no evidence to the contrary. Thus, Luke and Scoggins should have differentiated between years worked before the employee was promoted to FLS and the years after promotion to FLS. Plaintiffs’ experts’ failure to use each FLS’s years of seniority as a FLS creates further unreliability in Plaintiffs’ experts’ methodology. Statistical Evidence Conclusion. Plaintiffs’ statistical analysis suffers from several serious methodological flaws. Fundamentally, Plaintiffs’ analysis is premised on assumptions that are not supported by the summary judgment record, are illogical or are contrary to applicable legal doctrine. It is improper for Plaintiffs’ experts to compare non-Hispanic White males to an aggregate group of “minority” (African-American, Hispanic, and White female) FLSs in a case in which all Plaintiffs claim discrimination on the basis of their race. It is also unwarranted for Plaintiffs’ experts to assume that a FLS’s seniority for salary purposes included all years worked at Exxon, as opposed to only those years as a FLS. Plaintiffs’ experts’ putative statistically significant finding is derived only by using unreliable methodologies. Thus, Plaintiffs’ statistical evidence is inadmissible under Fed. R. Evid. 702 to show a gross disparity between the salaries of African-American and non-Hispanic White male FLSs. Moreover, if the appropriate comparison is made, i.e., between African-American FLSs and non-Hispanic White male FLSs, using Plaintiffs’ own raw data, the data fail to reveal a disparity of at least two standard deviations, and thus Plaintiffs fail to satisfy through their statistical evidence their summary judgment burden to show a “gross disparity” between Defendant’s treatment of members of the protected class and the favored group. 3. Racial Comments and Incidents Plaintiffs have submitted evidence (and a summary) of racial comments made by various employees at BOP and principally rely on these matters as direct evidence of Defendant’s discriminatory intent. See Racial Comments Brief. In addition, each Plaintiffs affidavit contains a list of incidents involving SLSs which Plaintiffs contend are racially charged. A supervisor’s repeated use of racial epithets may constitute direct evidence that a contested employment decision was motivated by racial animus. Reeves, 530 U.S. at 151, 120 S.Ct. 2097; Brown v. East Mississippi Electric Power Ass’n, 989 F.2d at 861-62. Remarks may serve as evidence of race (or sex) discrimination if they are (1) race or sex related, (2) proximate in time to the employment decision, (3) made by an individual with authority over the employment decision at issue, and (4) related to the employment decision. Medina v. Ramsey Steel Co., Inc., 238 F.3d 674, 683 (5th Cir.2001); Brown v. CSC Logic, Inc., 82 F.3d 651, 655-56 (5th Cir.1996). While the remarks need not be made in the “direct context” of the employment decision, see Russell, 235 F.3d at 226 (citing Reeves, 530 U.S. at 152, 120 S.Ct. 2097), “[c]omments that are ‘vague and remote in time’ are insufficient to establish discrimination.” Brown, 82 F.3d at 655-56 (internal citation omitted); Russell, 235 F.3d at 229 n. 19 (“remarks that the plaintiff wholly failed to tie to any potentially relevant time frame” could not constitute evidence of pretext sufficient to defeat summary judgment); see also EEOC v. Texas Instruments, Inc., 100 F.3d 1173, 1181 (5th Cir.1996) (remarks must be “direct and unambiguous”). a. Ed Coleman Coleman was supervised by the following SLSs during the period for which evidence has been submitted: Mark Cardinal, 1995; Mark Stephens, 1995-97; Mark Cardinal, 1997-98; and Alex Mott, 1998 to the present. Coleman does not contend that Alex Mott is racially biased against him. However, Coleman believes that both Mark Cardinal and Mark Stephens discriminated against him in various ways between at least 1995 and 1998. Coleman states that in 1997, Bob Metzger, a SLS and “respected member of Exxon’s management team,” told him that Cardinal and Stephens were “messing with him” and that they had attempted to move him as far down the rankings as they could. Coleman Aff., ¶¶ 14, 16. However, Coleman has no evidence to show that even if Cardinal and Stephens were “messing with him,” it was on account of his race. Coleman contends that on one occasion Cardinal went out of his way to obtain negative feedback on Coleman’s performance. Coleman Aff., ¶ 17; Coleman Dep., at 75-85. However, the incidents Coleman recounts do not evince a racially hostile attitude. In any event, in his deposition, Coleman admitted that he has no basis to contend that Cardinal solicited negative information because of his race and not in the exercise of appropriate supervisory inquiry as mandated by the ranking process and ordinary management practices. Nor is there any evidence that these were steps taken as to Coleman alone. Id. at 84-85. Coleman also states that Cardinal discounted his opinions because he was Black. Coleman Aff., ¶¶ 17-19. However, Coleman has produced no evidence from Cardinal or from any other person, or objective facts to permit the inference of racial bias in Cardinal’s decisions. Rather, Coleman’s descriptions reveal Cardinal’s and Coleman’s legitimate differences of opinion on several management decisions. Plaintiffs conclusory, subjective belief that he has suffered discrimination by Cardinal is not probative of unlawful racial animus. Auguster v. Vermilion Parish School Board., 249 F.3d 400, 403 (5th Cir.2001). Coleman also contends that, in 1993, Cardinal made a racially charged remark. Specifically, Coleman describes a discussion regarding racial diversity and an article entitled, “White Privileges,” that apparently discussed “the privileges that Whites had in this society because of racism.” Coleman Aff., ¶20. According to Coleman, Cardinal did not agree with the article, “blamed the efforts to end racism as something that would hurt his children’s opportunities,” and allegedly stated that he almost hated that his children were born White. Id. Cardinal’s remarks appear to be opposition to affirmative action, which position per se does not equate with racial bias. Further, Cardinal’s remarks, as recounted by Coleman, are too ambiguous to support a reasonable inference of racial bias against Coleman in job-related decisions. In addition, Coleman states that two individuals at the plant said the word “nigger” in his presence. Coleman Aff., ¶ 31. However, these epithets were uttered in 1981 and 1985, see Racial Comments Brief, one by an individual who retired in 1984 (and therefore had no influence on the ranking process during the time period relevant to this case), and one by an individual not a SLS. See Second Affidavit of Jo Kay Morreale (Ex. A to Final Submission) (“Second Morreale Affidavit”), ¶ 8. Thus, this evidence is not material to Coleman’s claims. Finally, Coleman states that the two African-American SLSs who participated in the 1998 rankings, Mike Williams and Melvin Parker, told him they felt that race played a part in the rankings. Coleman Aff., ¶ 33. These comments, while potentially significant, sound only in speculation as recounted. There is no evidence in the record that establishes the reason or the circumstances of Parker’s comment, or that fleshes out Parker’s view. As described, Parker’s alleged comment was merely a thought, without reference to any particular FLS or any indication as to the impact on the ranking results. As to Williams, in his deposition, Williams testified that race plays a part in the rankings only because people “can’t leave [their] biases at the door always.” M. Williams Dep., at 125. However, Williams clarified, “I think there’s an— there’s an effort to try and control [individual biases] with the checks and balances that we have in place with the management involvement ... but I’m not in an individual’s head when they’re making their choices, so I can’t say one way or not.” Id. Williams’s personal belief that race influences the rankings is grounded on his assumptions about society, but he has no specific examples of inappropriate comments or silences by SLSs at the ranking session. An African-American supervisor’s philosophical generalizations about race relations is too speculative to serve as probative evidence of discrimination against a particular employee. The incidents or remarks on which Coleman relies do not constitute probative evidence sufficient to support a verdict in Plaintiffs favor for race discrimination by the SLSs who participated in his rankings. Nor has Coleman adduced probative evidence of falsity of Defendant’s explanation of the rankings, which is a legitimate explanation for Coleman’s rank and salary for the years Plaintiffs’ claim are in issue, 1995-1998. b. Carl McClore From 1995 until the present time, McClore has been supervised by SLSs John Mabry, 1995-97; Rob Peplinski, 1997-98; and Joe Wolf, 1999 to the present. McClore characterizes his working relationships with his supervisors as “amicable” and “very good.” McClore Dep., at 30-31. Although McClore does not believe that his last three supervisors bore him any racial animus, he does contend that earlier supervisors treated him less well than his White co-workers. McClore contends that SLS Bob Metzger, his direct supervisor at some unspecified time before 1995, “was much less cordial to the black employees than he was to the white employees.” McClore Aff., ¶ 14. McClore provides no detail for his general observation and thus deprives it of probative value. In addition, at some unspecified time in the 1990s, McClore and another employee were teasing Metzger. In response, Metzger stated, “Well, at least I’m not black.” Id This exchange with Metzger is not probative of racial animus. The absence of specification of the time frame of Metzger’s alleged hostile comment also deprives it of probative value, rendering it a mere stray remark. McClore testifies further that part of the reason he believed Metzger was “against” him is because “a couple of years ago, [Metzger] came to [McClore] and [they] spoke and [Metzger] gave [McClore] a semi-apology, in so many words, and stated that he [Metzger] would be proud to support [McClore] henceforth.” Deposition of Carl McClore (Ex. C to McClore Motion) (“McClore Deposition”), at 200. McClore states that Metzger never mentioned or hinted that his prior opposition was race-based, and admits that the opposition might have been based purely on work performance. Id at 200-01. McClore states, without further elaboration, that Mark Stephens “started the wheels to get [him] demoted, and he never attempted any such thing with a white employee.” McClore Aff., ¶ 15. It is unclear if Stephens was McClore’s direct supervisor at this time. This incident, to which McClore specifies no date, appears to consist of Stephens asking another employee, Allen, if she thought McClore would be happier working in a different part of the plant when Stephens observed McClore having issues with a technician who reported to him. First Allen Aff., ¶ 12. This incident is ambiguous at best; it is mere speculation by Allen that Stephens’s reason for suggesting a less desirable job was because McClore was Black. Id McClore contends that Ken Breedlove, his supervisor in the late 1980s, discriminated against him. McClore Dep., at 144. McClore states that Breedlove appeared to resent McClore’s being able to solve a problem at the plant, and that Breedlove remarked “it wasn’t as though [others who hadn’t solved the problem were] stupid.” McClore Aff., ¶ 18. McClore states, “I had no doubt but that he made that comment because I was black and had accomplished something that a white male had not been able to do.” Id Again, McClore’s interpretation of this ambiguous incident is entirely speculative. The incident also is remote in time. McClore also recounts an occasion when Breedlove was unwilling to obtain equipment for McClore, see McClore Aff., ¶ 17, when he did so for his White subordinates. This incident apparently also occurred in the late 1980s. Accordingly, it is too remote in time to be probative. None of these incidents involving Breedlove constitute proof of racial bias, and all are, at best, vague or ambiguous actions that lack value to support McClore’s claim of discrimination. Finally, McClore contends that he and other minorities have been affected by a subtle racism: The way racism works in the ranking process is really a function of who is doing the ranking, and who their favorite people happen to be.... [I]t is much easier for a white male to be comfortable with, and like and respect, another white male than it is for that same white male to like and respect someone who is different from himself: a female, or a black or Hispanic male. It is simply easy for SLSs to choose someone like themselves to stand up for — particularly if they are not challenged on what they are doing. With this type of socialization going on at BOP, the white male SLSs have more exposure to other white males in the ranking group — and thus will be more likely to either recommend or go along with a higher ranking for someone they know and like. McClore Aff., ¶ 8. MeClore’s beliefs about White SLSs at BOP are stereotypical generalizations, based on speculation; they are not proof of racial animus against McClore, in specific rankings or salary decisions. See Auguster, 249 F.3d at 403. The remarks, incidents and theories McClore offers do not constitute probative evidence of racial animus by any SLS. c. Charlotte Allen Allen lists a number of remarks she contends exhibit racial or gender bias. Allen’s anecdotal evidence, while disturbing, is consistent in that all the incidents are extremely remote in time. Allen first states that in 1978, when she first began work at BOP, she told Bob Metzger that her goal was to become BOP’s first female supervisor and he replied, “What makes you think you will ever be qualified to be a supervisor?” Sometime before 1983, Stephens criticized Allen for being too motherly. Allen Dep., at 27. Sometime in the 1980s, an individual working at the plant, Randy Gillman, told Allen that he wouldn’t let his wife work, and that she was taking a job away from a man. Sometime in the 1980s, a BOP employee asked Allen who was taking care of her children. The above-cited remarks are too remote in time to constitute probative evidence of the claims at issue in this case. In addition, some of the comments were not made by SLSs. To constitute probative evidence of discrimination, remarks must be proximate in time to the adverse employment decision at issue and made by an individual with authority over the employment decision. Brown, 82 F.3d at 655-56; Medina, 238 F.3d at 683. Allen also recounts an incident, which apparently occurred some time in the 1980s, in which SLS Mark Stephens expressed concern about a Black technician being promoted to FLS because the last two technicians promoted were Black. First Allen Aff., ¶ 12. However, Allen states that Stephens admitted that the Black technician up for promotion was the most qualified candidate. Id. In addition, Allen indicates that although Stephens expressed concern that some White males in the department might be upset, the Black technician did get the job. Id. Accordingly, the remark is both too remote in time and too ambiguous to constitute evidence of racial bias. Allen testifies that, in 1990 or 1991, she reported to SLS Kevin Smith. Smith asked her to complete a project by a certain deadline, and when she told him it could not be done, he assigned a White male to follow her the next day. Second Allen Aff., ¶ 4. In addition to being remote in time, there is nothing about the incident that creates an inference that Smith treated Allen differently because of her race or sex. None of the above-mentioned incidents or comments constitutes probative evidence that raises a fact question as to whether any of the SLSs harbored racial (or gender) animus. d. Tony Williams Since 1995, Williams has been supervised by SLS Dave Veillon, 1995; Robert Hardesty, 1996; Mark Stephens, 1997-98; and Richard Ng, 1999. See Comparator’s Supervisors; Deposition of Tony Williams (Ex. A to Williams Motion) (“T. Williams Deposition”), at 97; T. Williams Aff., ¶ 15. It is unclear who Williams’s current supervisor is. See T. Williams Aff., ¶ 15. Williams’s complaints are directed towards Mark Stephens, who was also his supervisor in years prior to the relevant time period for this case. First, Williams contends that before being promoted to FLS, he was required by Stephens to undergo a trial period as a “step-up” supervisor. T. Williams Aff., at ¶ 4. Williams successful