Sherman v. Westinghouse Savannah River Co.

PER CURIAM: The four appellants—James Sherman, III (No. 04-2414), Annie B. Lott-Abney (No. 04-2417), Elvira Johnson (No. 04-2418), and Dianne S. Scott (No. 04-2420)— appeal from the judgments entered against them in the District of South Carolina on their race discrimination claims under the Civil Rights Act of 1866, 42 U.S.C. § 1981, and Title VII of the Civil Rights Act of 1964, 42 U.S.C. §§ 2000e to 2000e-17. In their consolidated appeals, the appellants challenge the district court’s denial of class certification on their disparate impact claims; the court’s exclusion of the report and proposed testimony of an expert witness; and the court’s awards of summary judgment against them on their individual disparate impact and disparate treatment claims. We possess jurisdiction over these appeals pursuant 28 U.S.C. § 1291. For the reasons explained below, we affirm in Nos. 04-2414 (Sherman), 04-2417 (LottAbney), and 04-2418 (Johnson), and we affirm in part and vacate and remand in part in No. 04-2420 (Scott). I. On October 31, 1997, ten African-American employees at the U.S. Department of Energy’s Savannah River Site (the “Site”) filed a class action complaint under § 1981 and Title VII on behalf of all African-American Site employees. The proposed class action eventually accumulated ninety-nine named plaintiffs, including the four appellants herein. The district court denied class certification and directed the plaintiffs to proceed individually. Of the ninety-seven plaintiffs who refiled individual complaints, all but five—the four appellants herein and Virginia Anderson—subsequently settled or otherwise dismissed their claims. We previously heard the appeal of Anderson, who also challenged the denial of class certification, as well as the entry of summary judgment on her individual claims. Our decision in that case— authored by our late and distinguished colleague Judge Widener—outlines the histoiy of the proposed class action and provides authority for the disposition of several issues in the present appeals. See Anderson v. Westinghouse Savannah River Co., 406 F.3d 248 (4th Cir.2005). The plaintiffs’ claims were brought against Westinghouse Savannah River Company; Bechtel Savannah River, Incorporated; The Babcock & Wilcox Savannah River Company, Incorporated; and British Nuclear Fuels Limited Savannah River Corporation. The defendants jointly operate and manage the Site, a 310-square-mile facility that processes radioactive and other hazardous waste. They operate as a “seamless organization” under a 1996 contract with the Department of Energy and have common personnel policies and prac *360 tices, including performance evaluation systems, promotion policies, and compensation policies, controlled by a single director of human resources. Various policies and practices of the defendants have been challenged by the plaintiffs in the proposed class action and their subsequent individual suits, under theories of both disparate impact and disparate treatment based on race. 2 A. Class Certification By way of their disparate impact claims, the plaintiffs variously attacked the defendants’ practices affecting five discrete aspects of employment—(1) promotions and evaluations, (2) hazardous job assignments, (3) salaries, (4) training, and (5) demotions. The plaintiffs initially sought certification of a single class, encompassing more than 4,000 African-American Site employees, for these claims. In a reply memorandum on the issue of certification, however, the plaintiffs suggested for the first time that the court could break the class into three separate subclasses for exempt, nonexempt, and craft employees. The district court denied class certification on May 25, 2000. The court determined that the plaintiffs’ claims lacked sufficient “commonality” and “typicality” to satisfy Federal Rule of Civil Procedure 23(a)(2) and (3). Moreover, the court ruled that the plaintiffs failed to demonstrate that at least one of the conditions of Rule 23(b) was met. Notably, the court considered the feasibility of the proposed subclasses, and concluded that none of them could “overcome the fatal deficiencies afflicting the larger proposed class.” The plaintiffs filed a motion for reconsideration of the class certification ruling, in which they abandoned their request for a single omnibus class, and instead asked for certification of as many as eleven different subclasses. On July 19, 2000, the district court denied the motion for reconsideration, rejecting the plaintiffs’ modified class certification request as untimely and lacking in merit. Thereafter, the plaintiffs filed their individual lawsuits. In her appeal, Anderson contended that the district court should have certified two subclasses corresponding to her two individual disparate impact claims, pertaining to the defendants’ systems for promotions and salaries. See Anderson, 406 F.3d at 273. We rejected Anderson’s contention on the ground that—having affirmed the awards of summary judgment on her disparate impact claims—she had “no valid claims which give her the same interest and cause her to suffer the same injury as the proposed class members she seeks to represent.” Id. at 274 (internal quotation marks omitted). 3 *361 B. Disparate Impact Claims 1. Sherman’s and Scott’s promotion system claims Sherman and Scott claim (as Anderson did before them) that the defendants’ system for promoting employees—the Competency Based Posting System (the “CBPS”)—has a disparate impact on African-American employees. Under the CBPS, which was instituted in the 1990s, open positions are submitted to the Site’s Human Resources Group (“HR”) for posting. For each open position, a hiring manager establishes the minimum qualifications, identifies and weighs the relevant “core” and “functional” competencies, serves on the three-member interview panel, and assists HR in the selection of the other interview panel members. 4 During the first of three stages under the CBPS, “self-nominating individuals” submit their qualifications to HR, by way of a “Personal History” form, a resume, or both. HR then identifies those applicants who meet the minimum qualifications for the open position, and forwards their application materials to the hiring manager. Next, during the CBPS’s second stage, the hiring manager (acting alone or with assistance of a committee) rates the applicants according to the weighted core and functional competencies of the open position, and selects the most qualified applicants for interviews. The interview panel conducts the interviews, produces written evaluations of the interviewees (based again on the core and functional competencies), and preliminarily identifies the most qualified candidate. If that candidate receives approval from the next level of management, as well as HR, the CBPS process moves to its third stage, during which the candidate is offered the position. Sherman and Scott (like Anderson) specifically contend that the CBPS has a disparate impact on African-American employees at the second and third stages, relying on expert testimony that there is a statistically significant under-representation of African-American employees succeeding at those two stages. By separate orders, the district court granted summary judgment to the defendants on Anderson’s, Sherman’s, and Scott’s CBPS-related disparate impact claims. Significantly, our panel majority in Anderson affirmed the *362 summary judgment award on Anderson’s claim, see 406 F.3d at 265-68, with one panelist disagreeing on the disposition of this single issue, see id. at 276-84 (Gregory, J., dissenting in part). 5 2. Scott’s and Johnson’s radiation exposure claims Scott and Johnson allege that the defendants’ job assignment practices have resulted in African-American Site employees being exposed to more radiation than their white co-workers. In support of their radiation exposure disparate impact claims, Scott and Johnson rely on the expert testimony of an epidemiologist, Dr. A. James Ruttenber. Importantly, the district court granted the defendants’ motion in limine to exclude Dr. Ruttenber’s testimony, at the conclusion of a hearing conducted pursuant to Daubert v. Merrell Dow Pharmaceuticals, Inc., 509 U.S. 579, 589, 113 S.Ct. 2786, 125 L.Ed.2d 469 (1993) (recognizing that “the trial judge must ensure that any and all scientific testimony or evidence admitted is not only relevant, but reliable”). The court determined that Dr. Ruttenber’s findings did not support the hypothesis that, during the relevant time period, African-American employees received higher radiation doses than white employees due to “management-initiated” job assignments. This was so, the court observed, because Dr. Ruttenber failed to ascertain the extent to which employee choice— rather than decisions made by managers— resulted in higher exposure job assignments. The court explained that “the problem that we have here is that there is no fit whatsoever between his analysis and the limited issue in this case, which is ... whether or not discriminatory management-initiated job assignments produced a difference in radiation dose for blaek[ ] and white[]” employees. The court further concluded that, although Dr. Ruttenber (as an epidemiologist) may have been qualified to conclude that African-American Site employees had higher radiation exposures than white employees, he went “beyond his realm of expertise” when he attempted to attribute such exposure rates to the defendants’ job assignment practices. And, the court noted and discussed “other flaws” in Dr. Ruttenber’s analysis. Thereafter, having excluded evidence (Dr. Ruttenber’s testimony) crucial to sustaining Scott’s and Johnson’s radiation exposure disparate impact claims, the district court awarded summary judgment to the defendants with respect to those claims. C. Disparate Treatment Claims 1. Sherman’s failure to promote claim Sherman claims that he was denied a promotion to a Lead Technical Specialist position because of his race. At the time Sherman applied for the promotion in December 1997, he had twenty-one years of experience working at the Site in various positions, including positions that arguably prepared and qualified him for the Lead Technical Specialist position. That position was awarded, however, to a white applicant (Jeffrey Stewart) who Sherman asserts was less qualified than him but the beneficiary of favoritism and preselection for the job. We focus herein on the facts tending to show that Stewart was preselected for the Lead Technical Specialist position. The *363 hiring manager for that position, Mike Hubbard, initially sought to promote Stewart without going through the CBPS process, but was ultimately required by HR to post the opening. The first posting listed minimum qualifications including a bachelor’s degree and 5-8 years of relevant experience. Before applications were submitted, the posting was revised to no longer require a bachelor’s degree; the revised posting listed minimum qualifications including a bachelor of science degree in engineering and 5-8 years of relevant experience, or a bachelor of science degree in another discipline and 9-11 years of relevant experience, or an associate’s degree and 13-15 years of relevant experience plus certain training, or a high school diploma and 15-17 years of relevant experience plus certain training. In the light most favorable to Sherman, the evidence indicates that the posting was revised so that Stewart (who did not possess a college degree) would minimally qualify for the Lead Technical Specialist position, and that Stewart was provided with special training in an area in which his qualifications would have yet been lacking. Stewart and two African-American applicants (but not Sherman) were named as finalists for the Lead Technical Specialist position. Sherman scored only one point below the two African-American finalists. Thus, had Stewart been ineligible for the job—as he would have been had the minimum requirements not been altered— Sherman would have been interviewed as a finalist. The members of the interview panel, including Hubbard, were all white. Stewart was selected for the Lead Technical Specialist position over the two African-American finalists. To prove his failure to promote disparate treatment claim, Sherman relies (as do the other appellants with respect to their disparate treatment claims) on the burden-shifting framework established by the Supreme Court in McDonnell Douglas Corp. v. Green, 411 U.S. 792, 93 S.Ct. 1817, 36 L.Ed.2d 668 (1973). 6 In awarding summary judgment to the defendants on this claim, the district court concluded that Sherman failed to forecast evidence sufficient to demonstrate that the defendants’ reasons for passing over Sherman were a pretext for discrimination. Regarding the favoritism shown by Hubbard to Stewart, the court observed that Sherman failed to demonstrate that Stewart’s preselection for the Lead Technical Specialist position resulted from racial animus. The court explained that the defendants’ “decision to lower the requirements for the position and [Hubbard’s] preference for one particular employee affected all applicants equally. Moreover, favoritism alone does not constitute actionable discrimination.” 2. Scott’s disparate treatment claims a. Scott’s failure to promote claim Scott alleges that, in 1998, she was denied a promotion to a Lead Technical Specialist position because of her race. Scott, *364 who started working at the Site in 1980 and holds a master of business administration degree (“MBA”), succeeded at the first stage of the CBPS (being deemed minimally qualified for the open position) but failed at its second stage (being selected for an interview). According to Scott, she was more qualified than the white employee (Susan Hatcher) who received the promotion, in that she had more years of work experience at the Site and a more advanced educational degree. The posting for the Lead Technical Specialist position listed, among the minimum qualifications, the equivalent of a technical degree and 3-5 years of experience, or an associate’s degree in a technical discipline and 8-10 years of experience, or a high school diploma and 12-14 years of certain practical experience. The posting also listed several areas of required knowledge. HR determined that Scott met the minimum requirements for the open position, and forwarded her application (which included a “Personal History” form but not a resume) to the position’s hiring committee. The hiring committee—which consisted solely of white members—indicated in its CBPS evaluation that Scott’s Personal History form contained “[inadequate information” to fully assess her candidacy or “to support selection for [an] interview,” and assigned Scott scores of “0” (for “not evident”) with respect to several of the weighted core and functional competencies. Scott (along with Bill Kesler, a white Site employee) received the lowest rating of all applicants for the position and was not selected for an interview. Hatcher, the prevailing candidate, began working at the Site in 1984, was promoted to a Laboratory Technician position in 1985, held a technical degree (in Environmental and Hazardous Material Management) directly relevant to the open Lead Technical Specialist position, and provided a detailed application specifically describing her experience, qualifications, and pertinent college class work. The district court granted summary judgment to the defendants on Scott’s failure to promote disparate treatment claim, based on its conclusion that Scott had failed to forecast evidence sufficient to demonstrate that the defendants’ proffered legitimate, nondiscriminatory reason for promoting Hatcher—that she “best fit the job description”—was a pretext for discrimination. The court observed that Scott’s only evidence was her own affidavit, which contained a self-assessment of her qualifications that was “irrelevant” and otherwise insufficient to create a genuine issue on the pretext question. b. Scott’s hazardous job assignment claim Scott also asserts that, based on her race, she was forced to take a job working with significantly more radioactive (and, thus, more dangerous) samples than those handled by her white counterparts. Scott had been hired to work at the Site in 1980 as a mail clerk, and had been promoted in 1981 to a position since retitled “technical analyst.” From 1981 to April 1998, Scott’s job responsibilities involved analyzing radioactive samples. In April 1998, Scott was temporarily assigned to another work group at the Site to assist in the disposal of “legacy chemicals,” i.e., chemicals dating back to the 1950s. Scott subsequently returned to her regular work group and, in November 1998, was involuntarily assigned to the post at issue, which involved receiving (rather than analyzing) radioactive samples. It appears from the record that this “sample receiver” job (an assignment within the scope of the broader technical analyst position) was a new post at the Site limited to working with “BNFL” samples, *365 i.e., highly radioactive samples that were significantly more radioactive than the samples more routinely handled within Scott’s work group. In this post, Scott was required, inter alia, to “split” BNFL samples that would have otherwise been too high in radiation for her co-workers to analyze. Prior to the creation of the new sample receiver job, the Site had previously analyzed just one set of the BNFL samples, and those samples had been handled by the two then-existing sample receivers along with all other incoming samples. The new sample receiver job was created to deal exclusively with the BNFL samples, based on lessons learned from experiences with the first set of BNFL samples. The new job was posted internally, but no Site employees applied for it—including the two sample receivers (Gina Robbins and Robin Wainwright, both white) who had handled the first set of BNFL samples. Even though there had not been any particular problems with Robbins’s and Wainwright’s handling of the first set of BNFL samples, neither of them were required to take the new sample receiver job, and they continued working as sample receivers handling samples that were not highly radioactive. Additional candidates for the new sample receiver job included thirteen other white employees in Scott’s work group who, as technical analysts (like Scott), met the minimum qualifications for the assignment. 7 In the spring of 1999, David Healy, a white employee who joined Scott’s work group after her assignment to the new sample receiver job, also became a sample receiver. Scott was not moved from her sample receiver job (working exclusively with the BNFL samples) at that time. Healy’s sample receiver assignment apparently occurred soon after Robbins shifted her job duties from receiving to analyzing samples. It is unclear whether Healy simply replaced Robbins working with non-highly radioactive samples, or whether Healy worked alongside Scott with the BNFL samples. It is also unclear whether Healy was voluntarily or involuntarily placed in the sample receiver job. The record suggests that Scott’s duties as a sample receiver exposed her to appreciably more dangerous conditions than those she had previously faced. For example, in the summer of 1999, Scott had a skin dose radiation reading of 149 MREM—well within the annual limit set by the federal government, but abnormally high for Scott and other employees in her work group (who were expected to have skin readings between 0 and 50 MREM). This reading caused one of Scott’s supervisors, Michael Polochko, to warn Scott to watch future skin dose readings for additional high numbers. Indeed, although *366 Polochko testified in his deposition that he had encountered abnormal readings for employees on “several occasions,” he specifically recalled only one other incident, and in that case (unlike Scott’s) the high number was attributed to an equipment malfunction. Polochko also acknowledged that, with respect to the annual whole-body exposure limits set by the Site, the limit for employees working with the BNFL samples (280 MREM) was higher than that for “most people” (150 MREM). Polochko refused to explain the discrepancy in the exposure limits. A fair inference, however, is that the defendants expected employees working with the BNFL samples to suffer greater exposures than other employees. And, the defendants have acknowledged that the less exposure, the better—as evidenced by their ALARA (“as low as reasonably achievable”) policy at the Site. In awarding summary judgment to the defendants on Scott’s hazardous job assignment disparate treatment claim, however, the district court concluded that Scott failed to establish a prima facie case of discrimination, in that she could not demonstrate that her assignment to the new sample receiver job constituted an “adverse employment action” that “had some significant detrimental effect.” 8 According to the court, the fact that Scott “could have been exposed to higher levels of radiation in the new job ... does not evidence a significant detrimental effect,” because “[h]er radiation exposure levels exceeded neither [the defendants’] nor the federal government’s limits, and [the defendants have] numerous checks and monitors to prevent employees from being harmed by radiation.” 3. Johnson’s hazardous job assignment claim Johnson claims that, in the early to mid-1990s, she and other African-American painters at the Site were exposed to radiation more often than their white counterparts. The district court granted summary judgment to the defendants on Johnson’s hazardous job assignment disparate treatment claim, for failing to establish a prima facie case of discrimination. The court explained that [t]he only evidence [Johnson] presented in connection with this claim is her own affidavit and the affidavit of another African American employee of [the defendants]. She also implies that it is the burden of [the defendants] to disprove [she] was not exposed to radiation more often than white employees by stating that [the defendant] offered no records comparing the assignments of its painter employees. The court concluded that, “[o]f course, it is not [the defendants’] burden to disprove [Johnson’s] claim, and [Johnson’s] assertions that she was exposed to radiation more often than similarly situated white employees, without more, is insufficient to create a prima facie case of discrimination.” 4. Lott-Abney’s demotion claim Lott-Abney alleges that she was illegally demoted in 1995 because of her race. Lott-Abney began working at the Site in 1972, and was working in an exempt position as an M-Area Records Management Coordinator in May 1995 during a Site- *367 wide reduction in force (“RIF”). LottAbney’s responsibilities involved collecting records from various facilities, packaging and processing the records, and sending them to the Site’s Record Control Department, as well as some procedure-writing duties. According to Lott-Abney, she was initially told that her work group would not be eliminated, but would instead be combined with a group of procedure writers in K-Area. Lott-Abney was subsequently notified, however, that her position was being eliminated. According to the defendants, the position ceased to exist upon the adoption of a new system requiring individual facilities to handle their own records. At the time of the RIF, a “bust back” policy was in place at the Site, under which certain exempt employees whose positions were eliminated could accept demotions to their former non-exempt positions. Rather than being laid off, Lott-Abney accepted a “bust back” to a non-exempt position as a Reactor Materials Operator, effective May 15,1995. Lott-Abney contends that, after the RIF, her former duties as Records Management Coordinator were given to Susan Power, a less senior white employee who had been a Production Supervisor in M-Area. Power was assigned post-RIF to K-Area, where she did some procedure writing and her title eventually became Training Instructor/Developer. To demonstrate that Power’s new duties were essentially the same as Lott-Abney’s old ones, Lott>-Abney relies on the testimony of a former co-worker, John Cummings, who stated in his affidavit that Lott-Abney “was a procedure writer and manager of document control for the M-Area” prior to the RIF, and that Power “was writing procedures [in the K-Area], just as Mrs. Lott-Abney had done as a portion of her responsibilities in M-Area, before she had been demoted.” Lott-Abney also points out that Power was given the same job title (“FLS-Operations Specialist”) and job code (“6646”) that Lott-Abney was assigned in her former Records Management Coordinator position. In granting summary judgment to the defendants on Lott-Abney’s demotion-related disparate treatment claim, the district court concluded, inter alia, that she failed to establish a prima facie case of discrimination. 9 Specifically, the court observed that Lott-Abney had not demonstrated that Power was assigned substantially the same duties that Lott-Abney had performed before her demotion—i.e., that Power had “filled” Lott-Abney’s former position. The court emphasized that, while Cummings’s affidavit reflected that both Lott-Abney and Power engaged in some procedure writing, Cummings acknowledged therein that such activity was only “a portion of [Lott-Abney’s] responsibilities” as Records Management Coordinator. Additionally, the court determined that there was no significance to LottAbney (as Records Management Coordinator) and Power (as Training Instructor/Developer) having the same job title and job code, because the evidence showed that they were “generic titles used in [the defendants’] database to identify employees for compensation purposes and reflect little about the employees’ actual responsibilities and duties.” II. After the district court entered final judgments against them, Sherman, Lott *368 Abney, Johnson, and Scott timely noted their appeals. They make numerous appellate contentions, challenging the district court’s denial of class certification; the court’s summary judgment awards on their individual disparate impact claims (and, with respect to the radiation exposure disparate impact claims, the exclusion of Dr. Ruttenber’s expert testimony); and the court’s summary judgment awards on them individual disparate treatment claims. We review a district court’s class certification decision for abuse of discretion. See Gregory v. Finova Capital Corp., 442 F.3d 188, 190 (4th Cir.2006). We also utilize an abuse of discretion standard when reviewing a district court’s Daubert evidentiary rulings on the relevance and reliability of expert testimony. See Bryte v. Am. Household, Inc., 429 F.3d 469, 475 (4th Cir.2005) (recognizing that “the district court has broad latitude in ruling on the admissibility of evidence, including expert opinion”). And, we review a district court’s grant of summary judgment de novo, viewing the facts in the light most favorable to the nonmoving party. See Nat’l City Bank v. Tumbaugh, 463 F.3d 325, 329 (4th Cir.2006). III. A. Class Certification Because none of the plaintiffs have made meritorious disparate impact 'claims, see infra Part III.B, they cannot constitute proper class representatives, and the issue of whether the district court abused its discretion in denying class certification is moot. See Anderson, 406 F.3d at 273-75. B. Disparate Impact Claims 1. Sherman’s and Scott’s promotion system claims Sherman and Scott contend that the district court erred in granting summary judgment on their CBPS-related disparate impact claims. Unfortunately for them, this contention is foreclosed by our prior Anderson decision, affirming the summary judgment award to the defendants on the essentially identical claim brought by Anderson. See McMellon v. United States, 387 F.3d 329, 332 (4th Cir.2004) (en banc) (recognizing “the basic principle that one panel cannot overrule a decision issued by another panel”). We are therefore constrained to affirm the court’s awards of summary judgment on Sherman’s and Scott’s claims. 2. Scott’s and Johnson’s radiation exposure claims Scott and Johnson challenge the district court’s awards of summary judgment on their radiation exposure disparate impact claims. As discussed above, the court made its summary judgment rulings after granting the defendants’ motion in limine to exclude the expert testimony of Dr. Ruttenber. Because the radiation exposure disparate impact claims depend on the expert testimony of Dr. Ruttenber, we first assess the propriety of the court’s exclusion of that evidence. Simply put, the district court did not abuse its discretion in ruling, at the conclusion of the Daubert hearing, that Dr. Ruttenber’s testimony must be excluded. Indeed, we agree with the court “that there is no fit ... between [Dr. Ruttenber’s] analysis and the limited issue in this case, [i.e.,] whether or not discriminatory management-initiated job assignments produced a difference in radiation dose for black[ ] and white[ ]” employees. See Daubert, 509 U.S. at 591-92, 113 S.Ct. 2786 (recognizing that, to satisfy relevancy requirement, expert testimony must “fit” by providing “a valid scientific connection to the pertinent inquiry as a precondition to admissibility”). As such, we need not *369 assess the other issues addressed by the court in excluding Dr. Ruttenber’s testimony. Because we affirm the ruling of the district court excluding Dr. Ruttenber’s testimony, we also affirm the court’s summary judgment awards on Scott’s and Johnson’s radiation exposure disparate impact claims. Those claims relied on Dr. Ruttenber’s testimony and are unsustainable without it. C. Disparate Treatment Claims 1. Sherman’s failure to promote claim Sherman maintains that the district court erred in granting summary judgment to the defendants on his failure to promote disparate treatment claim. We agree, however, with the district court that the evidence fails to demonstrate that Sherman was the victim of race discrimination. As Sherman himself asserts, the white employee given the promotion (Stewart) was preselected by the position’s hiring manager (Hubbard). While such preselection might demonstrate that Sherman and the other applicants were “unfairly treated, it does not by itself prove racial discrimination.” Blue v. U.S. Dep’t of Army, 914 F.2d 525, 541 (4th Cir.1990) (explaining that, “[i]f one employee was unfairly preselected for the position, the preselection would work to the detriment of all applicants for the job, black and white alike”); see also Anderson, 406 F.3d at 271 (invoking Blue in rejecting Anderson’s contention that preselection of employee for promotion constituted sufficient evidence for finding of pretext). Sherman acknowledges our Blue precedent, but contends that he can nevertheless show that Stewart’s preselection was discriminatory, based on evidence that the next three top-scoring applicants (including Sherman) were all African-American. The record reflects, however, that Hubbard selected Stewart for the Lead Technical Specialist position before even posting the opening (much less receiving applications for it and scoring the candidates). In these circumstances, we affirm the district court’s summary judgment award on Sherman’s failure to promote disparate treatment claim. 2. Scott’s disparate treatment claims a. Scott’s failure to promote claim Scott contends that the district court erred .in awarding summary judgment to the defendants on her failure to promote disparate treatment claim. We reject this contention and conclude that the court properly ruled that Scott’s claim failed for lack of evidence of pretext. Indeed, we agree with the court that Scott’s reliance on a self-assessment of her qualifications was wholly insufficient to sustain her claim. With particular respect to the proposition that Scott’s experience and MBA rendered her more qualified for the promotion than Hatcher, we observe that experience and education were only minimum qualifications for the open Lead Technical Specialist position, and that the promotion decision was made based on the weighted core and functional competencies. Cf. Anderson, 406 F.3d at 269 (recognizing, with respect to claim that Anderson was more qualified than promoted white employee based on educational background and years of service, that “Anderson cannot establish her own criteria for judging her qualifications for the promotion,” but rather “must compete for the promotion based on the qualifications established by her employer”). Of course, Scott’s total score on the competencies was brought down by the “0” marks she received in several areas for lack of information in her Personal History form on which she could be assessed. *370 Scott does not dispute that she failed to submit detailed information with her application materials for the hiring committee’s review. Rather, she suggests that the committee should have looked beyond her application materials to ascertain her prior work performance, and in doing so would have realized that she possessed the requisite skills for the open Lead Technical Specialist position. Along the same lines, the Anderson plaintiff contended that her scoring on a CBPS evaluation was inconsistent with prior positive performance reviews, and thus constituted evidence of pretext for the defendants’ failure to promote her. See 406 F.3d at 271-72. We rejected Anderson’s contention, explaining that “[w]e do not sit as a super-personnel department weighing the prudence of employment decisions made by the defendants,” and that “[w]e cannot require that different supervisors within the same organization must reach the same conclusion on an employee’s qualifications and abilities.” Id. at 272 (internal quotation marks omitted). Here, there is no evidence that the hiring committee evaluating Scott had any information before it other than what she included in her application, and we cannot say that the committee was obliged to seek out that which Scott herself did not provide. Accordingly, we affirm the district court’s grant of summary judgment on Scott’s failure to promote disparate treatment claim. b. Scott’s hazardous job assignment claim Scott also challenges the district court’s summary judgment award to the defendants on her hazardous job assignment disparate treatment claim. We agree with Scott that the court erred in concluding that she failed to establish her prima facie case. Contrary to the conclusion of the district court, Scott’s assignment to the new sample receiver job could be deemed an “adverse employment action” based on evidence that it subjected her to appreciably more dangerous conditions—i.e., greater exposure to potentially harmful radiation—than those she faced in her previous post. See Gunten v. Maryland, 243 F.3d 858, 868 (4th Cir.2001) (observing that if change in plaintiffs job assignment “truly had been significant,” such as by exposing her to more dangerous conditions, her contention that reassignment constituted adverse action “would have merit,” and agreeing “in principle that increased exposure to dangerous pathogens could adversely effect the terms, condition, or benefits of employment”), overruled on other grounds by Burlington N. & Santa Fe Ry. Co. v. White, 548 U.S. 53, 126 S.Ct. 2405, 165 L.Ed.2d 345 (2006). The relevant inquiry is whether Scott was indeed subjected to appreciably more dangerous conditions— not whether (as suggested by the district court) she suffered a radiation dose beyond those limits established by the federal government and the defendants. The defendants offer us an alternative ground for affirming the summary judgment award in them favor: they contend that Scott cannot demonstrate, in support of her prima facie case, that she was treated worse than “similarly situated” white employees. See Sterling v. Tenet, 416 F.3d 338, 345 (4th Cir.2005) (recognizing that, to establish prima facie case, plaintiff must “raise an inference of discriminatory intent by showing that she was treated worse than similarly situated employees of other races”). The defendants point to the fact, acknowledged by Scott, that three white employees in her work group (Robbins, Wainwright, and Healy) also worked as sample receivers. The record does not reflect, however, that Scott was “similarly situated” to Robbins, Wainwright, and Healy. Rather, the evidence shows that *371 Scott was the first Site employee ever placed in the sample receiver job created to deal exclusively with the highly radioactive BNFL samples. Although Robbins and Wainwright had previously worked with the Site’s first set of such samples, they were also handling less radioactive samples at the same time. Moreover, when they did not apply for the new sample receiver job dedicated to the BNFL samples, they (unlike Scott) were not forced into the assignment, even though it seems they would have been prime candidates for the job having already handled the Site’s first set of BNFL samples. Instead, Robbins and Wainwright remained sample receivers working only with non-highly radioactive samples. As for Healy, while it is possible that his sample receiver job involved working with the BNFL samples (the record is simply unclear on this point), there is no evidence before us suggesting that he was involuntarily assigned to do so. 10 Although the facts underlying Scott’s hazardous job assignment disparate treatment claim could certainly be cleai'er, the forecast evidence is at least sufficient to withstand summary judgment. We therefore vacate the district court’s summary judgment award on this particular claim only, and remand for such other and further proceedings as may be appropriate. 3. Johnson’s hazardous job assignment claim Johnson maintains that the district court erred in awarding summary judgment to the defendants on her hazardous job assignment disparate treatment claim. Upon reviewing the record, however, we are constrained to agree with the district court that Johnson’s evidence was wholly insufficient to establish a prima facie case of discrimination. In her own affidavit (apparently submitted to clarify her vague and sometimes contradictory deposition testimony), Johnson averred only that “I was forced to ‘dress out’ [for protection from radiation exposure] as often as every day from some point in thé early 1990’s until I was laid off in May 1995. White Painters with my same White Foreman, Eddie Hill, were not required to dress out as often as I was.” Simply put, this evidence cannot defeat the defendants’ summary judgment motion. See Bryant v. Bell Atl. Md., Inc., 288 F.3d 124, 134-35 (4th Cir.2002) (“These affidavits ... amount to no more than subjective beliefs, and such evidence, without more, is insufficient to create a genuine issue of material fact as to any discriminatory conduct on Bell Atlantic’s part.”). Accordingly, we affirm the district court’s summary judgment award on Johnson’s hazardous job assignment disparate treatment claim. 4. Lottr-Abney’s demotion claim Finally, Lott-Abney challenges the district court’s grant of summary judgment on her demotion-related disparate treatment claim. For the reasons discussed by the district court, however, we conclude that Lott-Abney failed to establish her prima facie case. Cf. Causey v. Balog, 162 *372 F.3d 795, 802 (4th Cir.1998) (observing that evidence was insufficient to show that employer “filled” plaintiffs position with someone outside protected class, where employer split plaintiffs former duties between two divisions); Blistein v. St. John’s College, 74 F.3d 1459, 1470 (4th Cir.1996) (concluding that plaintiff failed to establish he was “replaced” where evidence reflected that alleged replacement assumed “only some of [plaintiffs] former duties,” other such duties were given to “various individuals [in plaintiffs same] protected class,” and remaining “duties were not assumed by anyone”). We therefore affirm the district court’s summary judgment award on Lott-Abney’s illegal demotion claim. IV. Pursuant to the foregoing, we affirm the district court with respect to the claims of Sherman (No. 04-2414), Lot1>-Abney (No. 2417), and Johnson (No. 2418). We affirm in part on the claims of Scott (No. 04-2420); we vacate, however, the award of summary judgment on her hazardous job assignment claim, and remand for further proceedings on such disparate treatment claim only. Nos. Ob-UU, 0b-2bl7, and 0b-2kl8 AFFIRMED. No. 0b-2b20 AFFIRMED IN PART AND VACATED AND REMANDED IN PART. Footnotes [2] The plaintiffs have relied on evidence of race-related harassment in the workplace as relevant background. In particular, the plaintiffs allege that hangman's nooses and racist graffiti appeared throughout the Site, that there was open use of racial epithets, and that management did not remove the graffiti in a timely fashion or adequately respond to other instances of harassment. [3] With respect to the class certification issue, we further observed in Anderson that there was “no indication” that any plaintiff could successfully prosecute a disparate impact claim challenging the salary system, without the rejected testimony of plaintiffs’ proffered expert on that claim. 406 F.3d at 274. We acknowledged the possibility, however, that a class action might yet be appropriate with respect to a disparate impact claim attacking the promotion system. Id. at 275. Accordingly, the panel majority agreed to remand with the following instructions: Upon remand, if a proper plaintiff or plaintiffs with grievances similar to those of Miss Anderson with respect to that discrete portion of the [promotion] procedure presents himself to prosecute, himself, as a class representative, the district court should then decide whether a class action is maintainable and whether the then named plain *361 tiff should represent the class.... If no representative plaintiff so comes forward within a reasonable time, then the district court should strike the class action from the calendar and enter a final dismissal thereof. Id.; see also id. at 275-76 (Niemeyer, J., concurring in part and dissenting in part) (explaining disagreement with remand remedy). On October 17, 2005, the district court entered a final dismissal order in the original proposed class action, observing that five months had passed since the issuance of our Anderson decision and that, during that time, no proper representative plaintiff had stepped forward to pursue a class action on the promotion system-related disparate impact claim. [4] We detailed the CBPS in our prior Anderson decision. We explained therein, inter alia, that [u]nder the [CBPS], hiring managers can evaluate applicants in six core competencies: teamwork, leadership, communications, business results, self-management, and employee development. The hiring managers can also evaluate applicants using selected functional competencies that are specific to the position for which the manager is hiring. For example, a particular functional competency could be "proficient in heating, ventilation and air-conditioning design,” and the CBPS manual explains that "functional competencies can be derived from the responsibility section on the job description.” Each competency is assigned a weight, using a number from one to five, in which five is "most important relative to other competencies” and one is "least important relative to other competencies.” 406 F.3d at 256. [5] Anderson unsuccessfully petitioned this Court for rehearing en banc, with five judges voting to grant the petition, and seven judges voting to deny it. See Anderson v. Westinghouse Savannah River Co., 418 F.3d 393, 394 (4th Cir.2005). Judge Gregory dissented from the denial of rehearing en banc, criticizing the panel majority’s decision to affirm the summary judgment award on Anderson's CBPS-related disparate impact claim. [6] Under this framework, as we recognized in Anderson, the plaintiff can establish a prima facie case of racial discrimination in promotions "by showing that (1) she is a member of a protected group, (2) she applied for the position in question, (3) she was qualified for that position, and (4) the defendants rejected her application under circumstances that give rise to an inference of unlawful discrimination.” 406 F.3d at 268. "If a prima facie case is established, the burden then shifts to the employer to articulate some legitimate, nondiscriminatory reason for the decision not to promote.” Id. (internal quotation marks omitted). "After the employer states a reason for its decision, [the plaintiff] has the opportunity to show that the stated reason is a pretext for discrimination, and the trier of fact must determine if the plaintiff has proved that the employer intentionally discriminated against her because of her race.” Id. (internal citations omitted). [7] Viewed in the light most favorable to Scott, the record indicates that the defendants have offered varying reasons for the choice of Scott for the new sample receiver job. According to Scott, one of her supervisors, Marty Finney, informed her at the time of the assignment that she was selected for the new job because her advanced degree (an MBA) and administrative skills rendered her perfect for it. Finney testified in her deposition, however, that nothing about Scott’s education especially qualified her for the new sample receiver job. Moreover, Finney denied telling Scott that she was selected, at least in part, because of her educational background. Rather, Finney asserted that she told Scott she was selected because she lacked a permanent assignment after returning to her work group from the temporary assignment disposing of legacy chemicals. Meanwhile, Finney’s superior, Michael Polochko, testified that Scott was chosen because of her MBA and because Scott’s group had not had enough other work for her to do upon her return from the temporary assignment. In this appeal, the defendants offer the work shortage as a nondiscriminatory justification for Scott's selection for the new sample receiver job. [8] To establish the relevant prima facie case, the plaintiff must “raise an inference of discriminatory intent by showing that she was treated worse than similarly situated employees of other races,” and she must establish that her involuntary placement in a new position constituted an “adverse employment action.” See Sterling v. Tenet, 416 F.3d 338, 345 (4th Cir.2005). [9] Among the elements of the relevant prima facie case is that the plaintiff's former position “was filled by [a] similarly qualified applicantf] outside the protected class.” Hill v. Lockheed Martin Logistics Mgmt., 354 F.3d 277, 285 (4th Cir.2004) (en banc). [10] Notably, while the defendants assert in this appeal that they had a legitimate, nondiscriminatory reason for assigning Scott to the new sample receiver job (i.e., a work shortage in her group that left her without a full-time assignment), see supra note 7, they do not dispute that Scott has forecast sufficient evidence of pretext to survive summary judgment. In that regard, Scott relies on the evidence reflecting that the defendants have given varying reasons for choosing her for the new position, thus creating a genuine issue of fact with respect to pretext. See EEOC v. Sears Roebuck & Co., 243 F.3d 846, 852-53 (4th Cir.2001) ("[T]he fact that Sears has offered different justifications at different times for its failure to hire Santana is, in and of itself, probative of pretext.").