Full opinion text
ORDER ON PENDING MOTIONS McKINNEY, District Judge. I. MOTIONS TO STRIKE Rule 56(e) requires that affidavits supporting or opposing a motion for summary judgment set forth facts that would be admissible in evidence, and show “affirmatively that the affiant is competent to testify to the matters” presented. Fed.R.Civ.P. 56(e). Courts may permit an affidavit to be supplemented or opposed by depositions, answers to interrogatories, or further affidavits. Id. Although the evidence presented in opposition to or support of summary judgment does not have to be in admissible form, it must be admissible in content. See Winskunas v. Birnbaum, 23 F.3d 1264, 1267-68 (7th Cir.1994) (evidence must be of “evidentiary quality,” meaning that a change in form but not in content would make the evidence admissible at trial). Plaintiffs offering scientific or technical evidence have the additional burden of establishing the qualifications of the experts, the reliability of their testimony, and its usefulness to the jury. Daubert v. Merrell Dow Pharm., Inc., 509 U.S. 579, 587 n. 10, 113 S.Ct. 2786, 125 L.Ed.2d 469 (citing Bourjaily v. United States, 483 U.S. 171, 175-76, 107 S.Ct. 2775, 97 L.Ed.2d 144 (1987)). Defendant, Indiana Bell Telephone Company, Inc., (“IBT” or “Indiana Bell”), has moved to strike the testimony of plaintiffs’ potential expert witnesses Michael Murphy (“Murphy”), Richard Wertheimer (“Wertheimer”), Marc Bendick (“Bendick”), Lance Seberha-gen (“Seberhagen”), and Jeffrey Rasmussen (“Rasmussen”). The Court need not address the motion as it relates to Murphy’s testimony, because plaintiffs have indicated they do not intend to call him as an expert, and thus have effectively withdrawn his testimony from contention. See Pls.’ Brf. In Oppos. To Def.’s Mot. To Str. Exp. Test, at 8, n. 1. In addition, plaintiffs’ relatively weak arguments in favor of the relevance and reliability of Rasmussen’s report, coupled with Indiana Bell’s proffered evidence that he did not actually write this report, renders it untrustworthy and inadmissible as an expert opinion. See Rasmussen Dep. at 33, 36-37, 42, 43, 46-47, 52. Even if it did not suffer these defects, Rasmussen’s report would not support the contention that Indiana Bell purposely targeted older workers during its 1992 workforce resizing program (“WRP”). The report is based on the assumption that employees who had worked five years or less were completely excluded from the resizing process. The evidence shows otherwise. At most, it shows that these workers were not automatically included in the mechanical phase of the evaluation process. Yet, they were evaluated individually and separately by their supervisors, using the same criteria as the others in phase one. The report’s conclusion that younger workers were “protected more from the downsizing process at a probability level of less than 5 times out of a million,” is not trustworthy because its premise is false. Plaintiffs have not offered sufficient evidence to raise a factual issue about whether these employees were excluded, and defendant’s' evidence to the contrary adequately supports a finding that they were included. Thus, Rasmussen’s opinion bears no relationship to a material fact in dispute, for there was ample evidence to show that the alleged “pz’otected” group did not exist. Not only are Rasmussen’s methods and assumptions unreliable, but his opinion suffers from irrelevance. For all of these reasons, it must be excluded. The report offered by Seberhagen is the same report as was filed in the related case, Adams v. Ameritech Services, Inc., IP 93-420-C, which the Court found to be inadmissible. It is equally inadmissible here. Seberhagen’s testimony about the disparate impact on older workers of use of the criterion “growth potential” during the downsizing is not only irrelevant, in that it relates to a theory of recovery not available under the ADEA, but is also not helpful to the jury. The criterion Seberhagen analyzed was “growth potential, which he construed as being defined by Indiana Bell as ‘promotional potential.’ ” Def.’s App. 3, Seberhagen Rep. at 2. Based on what plaintiffs’ counsel had told him, Seberhagen reported that there were no “detailed rating scales, behavioral guidelines, or other written explanations to assist in defining these terms.” Id He characterized the term as a “vaguely-defined rating factor,” and concluded that its use would produce unreliable measurements. Each rater would have to supply his or her own definition, he opined, meaning that the results would vary depending on the rater. Id. By his own admission, however, Seberha-gen was not concerned about how the terms were actually used or applied by either Indiana Bell or Ameritech Services, Inc. (“ASI”). Seberhagen Dep. at 38. Rather, as he understood it, the point of his report was to give his “general professional opinion about the use of growth potential and promotional potential” as described only in the documents he examined. Id. Seberhagen examined only the proposed selection process (Pis.’ Ex. 1) for Indiana Bell, and did not know which company used, or had access to, which documents. Id. Again, he said that did not matter to him. Id. In other words, he gave his opinion about the reliability and validity of these terms in the abstract, without reference to the actual evaluation process used by either company. He testified that he only looked at growth potential as the selection procedure, and that it was “a very limited question that I was asked to look at.” Id. at 109. The evidence shows that many different criteria were used to select the employees for termination, making Seberha-gen’s view of the facts distorted. According to Seberhagen, the unreliable scores that would result from use of such vague terms would be invalid in a competitive selection situation, and were likely to have an “adverse impact on older employees.” Id. at 3. Seberhagen’s explanation for the latter conclusion, however, is that the “common stereotype that ‘You can’t teach an old dog new tricks’ could have worked against older employees in the ratings.” Id. Such obvious speculation by an expert is precisely the type of testimony that should be kept from a jury. Moreover, the likelihood of confusing the jury with testimony by an expert on the meaning of a phrase found in a document, when jurors are capable of discerning that meaning themselves from the context, would be immense. It is also highly prejudicial to allow someone bearing the mantle of “expert” to testify about the meaning or effect of a term, based on so little information about how the term was actually used. For these reasons also, Seberhagen’s testimony is irrelevant, unreliable, and not helpful to the factfinder. Turning to Wertheimer’s and Bendick’s opinions, the Court finds that, as in the Adams case, the defendant here has challenged the validity of the methods used by Wertheimer, the relevance of his opinion to the demonstrable facts of this case, and its relative probative value. Because Bendick’s report is primarily a vouching statement for Wertheimer’s work, its admissibility will depend on whether its antecedent is admissible. Bendick submitted the same report for both cases, in which he described his credentials as a labor economist, with a Ph.D. from the University of Wisconsin. Def.’s App. 5, Ben-dick Report at 1. He reports spending most of his professional time researching and consulting on employment issues, with a smaller portion of that time devoted to litigation. According to Bendick, his testimony has been accepted in thirteen federal cases and he has assisted in more than seventy-five cases. Id. Bendick was asked by plaintiffs’ counsel to “consult with another economist, Dr. Richard Wertheimer, while he was producing his analyses in this case; to review and evaluate Dr. Wertheimer’s analyses; and to supplement Dr. Wertheimer’s work with such additional analyses as [he] judged appropriate.” Id. Bendick reported that he agrees with Wertheimer’s analytical approaches, believes they were “implemented with appropriate professional care,” and that they can be relied on as sound. Id. at 2. He also agrees with the conclusions that Wertheimer drew from his statistical analyses, especially certain enumerated conclusions. To supplement the work that Wertheimer performed, Bendick added his opinion about the re-employment prospects for older, former Indiana Bell employees. Id. at 5. He began by explaining that the majority of terminated employees who were in their forties and fifties would likely seek re-employment, rather than retire. Id. at 6. Bendick illustrated this point using data collected from one terminated worker, Lawrence Shelton, from which he projected that the “vast majority” of Indiana Bell employees age forty and above, who were selected under the resizing plan, would be concerned about whether they could reasonably expect to find a new job. Id. at 7. That concern, Bendick opined, would have influenced whether they would “volunteer” to take the early retirement package. Id. The point of his analysis was to discredit the defendant’s accounting for the employees who volunteered for early retirement. Based on Bendick’s own independent research, and that of others, he predicted that “most such workers [would] face unfavorable labor market circumstances and limited employment opportunities.” Id. He concluded further that “only a small proportion” of the employees selected for termination at Indiana Bell were likely to have “truly volunteered]” for termination. Id. at 11. Bendick further states that this “line of reasoning” raises “serious questions” about the characterization of Indiana Bell employees who accepted the early retirement package as volunteers. Id. He opined that it would be more likely that they accepted the offer “because they believed (perhaps because they have been informed, directly or indirectly, by their employer) that they were about to be involuntarily terminated.” Id. He found this hypothesis supported by an analysis of the percentage of employees, aged forty or more, who had been selected and who volunteered for termination, with the number of the same population who had not been selected and who had volunteered. The former group was 40%, while the latter was 7%. Id. at 12. This analysis, according to Bendick, strengthened Wertheimer’s conclusion that it was inappropriate to classify employees who had been selected for termination, yet accepted the early retirement package, as volunteers. Id. Indiana Bell challenges the basis, validity and relevance of Bendick’s opinions. Initially, the defendant points out that Bendick stated in his report that he and Wertheimer remained in contact throughout the period in which Wertheimer was producing his analy-ses, and that they discussed this ease. Yet, he has admitted that was not true. Id. at 2. During his deposition, Bendick said that he and Wertheimer had only discussed Wer-theimer’s difficulty reading certain databases for this case, and the remainder of their discussions were about an unrelated case. Bendick Dep. at 12-13,15. He also admitted that he did not read or review Wertheimer’s report before it was submitted, so he could not have offered any substantive assistance therewith. Thus, he did not actually “consult” with Wertheimer. Defendant argues that Bendick did not objectively evaluate the Wertheimer report, or provide any reasons for his opinion as to its validity. Instead, he merely agreed with it in its entirety, in effect rubber-stamping it. The defendant also points to Bendick’s testimony during his deposition that he did “nothing to confirm, evaluate, or otherwise analyze” the report. Ben-dick Dep. at 85-87. However, upon close review of his testimony, the Court notes that Bendick actually said that he “read [the report] thoroughly,” he thought he had discussed certain points with Wertheimer, he believed that he made certain calculations, he checked numbers as to their plausibility, and he “thought about it a good deal.” Bendick Dep. at 87. He also believed that he checked the report against some external sources before rendering his opinion, but he did not say which sources. Id. at 87-88. Although defendant has overstated the paucity of evidence supplied by Bendick’s analysis, it is evident that Ben-dick’s was not a thorough review of Wer-theimer’s work. Bendick admitted that he did not check the “arithmetic processes,” or attempt to verify “those sorts of things.” Id. at 78. He also stated that he did not “independently look at the data and his different data sources,” or make an independent assessment about which data source Wertheimer chose to use. Id at 79. Finally, Bendick said that he relied on his personal knowledge of Wertheimer’s work and reputation when he stated that he agreed with Wertheimer’s analytic approach, and that it was implemented with appropriate professional care. Id at 80. In light of this testimony, Ben-dick’s vouching for Wertheimer’s report adds little to its value; neither does it detract from it. As a separate source of evidence, however, Bendick’s report provides little assistance and has scarce probative value. Its value will rise and fall with the value of Wertheimer’s report. Turning to that report, the Court notes that the primary objection voiced by the defendant is that Wertheimer’s analyses are unreliable in that he received little or no accurate information about the Indiana Bell workforce resizing program (“WRP”). The information he had, defendant asserts, is not the type of information on which an expert would normally rely to form an opinion. In addition, the opinions based on this information are not helpful to the jury in understanding the facts of the case. They are not relevant. At best, Wertheimer’s statistical evidence can prove that the outcome of the WRP was not the result of chance, which does not help a jury decide, in an age discrimination case, what actually caused the outcome. At worst, this evidence may confuse the jury and lead them to infer causation merely from impact because of the “expert’s clout.” For both of these reasons, Indiana Bell has moved to strike Wertheimer’s reports under Daubert. In the alternative, even if the Court finds the evidence admissible under Daubert, Indiana Bell argues it should be excluded under Rule 403 of the Federal Rules of Evidence. Having just reviewed Wertheimer’s work in an order for the related case, Adams v. Ameritech Corp., IP 93-420-C, the Court will assume familiarity with the framework by which such expert testimony is evaluated. For purposes of economizing judicial resources, not to mention the natural resources in our forests, the Court will endeavor here to distinguish between the testimony he has offered in the two cases. There has been no suggestion from the defendant that Wer-theimer lacked the proper credentials to conduct a statistical analysis of the WRP and its results, so this discussion will focus primarily on the scientific validity of his studies, the reliability of his conclusions, and their relevance to this action. In the related case, the Court found that Wertheimer’s analyses were flawed in that they merely eliminated random chance as an explanation for the statistical disparity in selection rates between older and younger workers, without accounting for any other independent variable that might explain the association between age and termination rates. In essence, he assumed that any anomalies in the data were caused by age discrimination. Other variables might include favored job skills, education, experience, performance or self-selection. Also, his analytical framework was built on inaccurate assumptions about the make-up of the workforce at ASI and the distribution of necessary skills among members of that workforce. Crucial to the decision to exclude Wertheimer’s report in the related ease is the fact that his assumptions and analyses did not accurately reflect what occurred during the downsizing. In fact, Wertheimer did not know what had occurred, and he did not account for it in his analysis. In essence, the Court refused to allow Wertheimer’s testimony in Adams because he failed to demonstrate, either in his opinion or in his subsequent testimony, that he employed an appropriate scientific method for determining the significance of the difference in termination rates for workers under the age of forty and those age forty or more. That finding is even more appropriate here. As indicated in the introduction to Wertheimer’s report, the task he was given by plaintiffs’ counsel was to “analyze the age structure of Indiana Bell’s management workforce [and] the process of selecting employees for terminations associated with the 1992-1993 Workforce Resizing Program.” Def s Tab 1. This stated purpose provides the only evidence of the “specific question that is under investigation.” See Moore’s Federal Practice: Reference Manual on Scientific Evidence 423 (Fed.Jud.Ctr.1995) (“Moore’s Reference Manual”). “Research begins with a clear formulation of a research question. The data to be collected and analyzed must relate directly to the immediate issue; otherwise, appropriate inferences cannot be drawn from the statistical analysis.” Id. (emphasis added). These statements, found in the chapter on “Research Design” in Moore’s Reference Manual, provide the Court with the first step of the scientific method for statistical research. Once a clear formulation of the question is posed, the next step is to decide which model to use to evaluate the question. Id. Finally, the expert should interpret the results, either in terms of statistical or practical significance. The first problem noted with Wertheimer’s report is that the “research questions” are unrelated to the real issues in the ease. There is no dispute about the age structure of Indiana Bell’s management workforce pri- or to the WRP. Nor is there any dispute about the fact that a greater percentage of workers aged forty or more were selected for termination (56 of the 59 selectees). What is crucial for ADEA purposes is a determination of what caused the disparity. The second problem is that Wertheimer said he was going to “analyze the process of selecting employees for termination,” when he in fact only analyzed the results of the termination process. Making matters worse, he then worked backwards from the results and “analyzed” (or rather speculated) that the process treated older workers differently. He did this without researching any other possible explanations for the observed disparity, indeed, without even knowing anything specific about the WRP process used by Indiana Bell. Wertheimer 5/22/95 Dep. at 161-62. In fact, Wertheimer characterized the information he had received from plaintiffs’ counsel as an “impressionistic view” of the process, without any detailed or specific information. Id. Nevertheless, Wertheimer concluded that the “evidence in [his] report is consistent with the Indiana Bell 1992-1993 Workforce Resizing Program treating employees at least age 40 differently from employees under age 40.” Wertheimer Report at 21 (emphasis added). This conclusion is not based on an application of the scientific method for statistics, because it infers from statistics that merely demonstrate an association between two variables, that one variable caused the other. For that conclusion to be valid, the scientific method would require more than the cursory treatment applied by Wer-theimer. The most appropriate scientific method for determining causation is a controlled experiment, but observational studies may be useful under certain circumstances. Moore’s Reference Manual at 347, 350. For example, when the association between the two variables is seen in several different types of studies among different groups, the chance that the observed association is due to a defect in the study or a unique characteristic in the group is reduced. Or, when the results of a study hold while the effects of plausible confounding variables are taken into account by appropriate statistical techniques, the causation conclusion would be supported. Likewise, when there is a plausible explanation for the effect of the independent variables, so that the causal link does not depend only on the observed association, a conclusion of causation would be more reliable. Moore’s Reference Manual at 351-52. These circumstances increase the likelihood that a valid causal relationship has been identified. None of those circumstances are present here. Wertheimer did not conduct any other studies, using any different populations at Indiana Bell, nor did he account for the effects of plausible confounding variables by any method. Finally, Wertheimer explored no other plausible explanations for the effect between the independent variables. In other words, Wertheimer simply inferred from the fact that a greater number of older workers were selected for termination, that it was their age that caused them to be terminated. His methodology is not scientifically valid for answering the research question he actually posed (analyze the process of termination), nor is it valid for understanding the issue for which the plaintiffs seek to employ his opinions (showing disparate treatment). To be admissible, testimony must be relevant to the facts the party intends to prove under the legal theory that has been chosen to support the party’s claim. See Charles Alan Wright & Kenneth W. Graham, Jr., Federal Practice And Procedure: Evidence § 5162. “Relevance is a relationship between the evidence offered and the fact it is supposed to prove.” Id.. A piece of evidence may be excluded as irrelevant in one of two ways. First, it may provide satisfactory proof of a fact in dispute, but the fact for which it is offered is not a material fact in the case.^ Id. Second, it may be directed at a material fact, but not have any value as proof of that fact. Id. A material fact is a “fact that is of consequence to the determination of the action.” Fed.R.Evid. 401. Wertheimer’s analyses and opinions miss the relevance mark in both ways. First, his efforts at measuring the statistical significance of the disparity between selection rates for older versus younger workers relates to the “fact” of whether there was a difference between the number of older and younger workers who were selected. That fact, however, was not seriously in dispute, nor is it material to the issue of whether Indiana Bell chose the older workers for termination because of their age. Second, his assessment of the significance of the disparity, although directed at the material fact of causation, has no value in proving that fact. This is because, as noted above, he did not use an appropriate method for determining causation, and the circumstances that would allow the use of an observational study were not present. Thus, his testimony cannot be considered relevant, and would not be helpful to the finder of fact. Even if Wertheimer’s testimony were to be considered reliable and helpful under Dau-bert, it would not pass the Rule 403 balancing test. The probative value of his questionable methodology, incomplete research, and shaky conclusions, would be substantially outweighed by the distinct possibility of confusing the jury about the relevant issues. It would also be outweighed by the danger of misleading them into inferring causation from the fact that an expert found the disparity in selection rates “statistically significant.” Statistical significance is an expression of the probability of achieving a certain result by chance. It is often measured by p-values. When a statistician computes the p-value for any set of data, he or she is determining the probability of getting, just by chance, test data as extreme as the actual data obtained, given that the null hypothesis is true. Moore’s Federal Practice: Reference Manual on Scientific Evidence 402 (Fed.Jud. Ctr.1995) (“Moore’s Reference Manual”). A “null hypothesis” is the hypothesis that there is no difference between two groups from which samples are drawn. Id, at 401. For example, the null hypothesis in this case would be that there is no difference between Indiana Bell employees who are under forty and those forty or more in terms of the criteria used to select them for termination. Thus, if the selection rates found in samples of the two age groups at Indiana Bell are not the same, then the p-value would give the probability that this data resulted from “the luck of the draw.” See Moore’s Reference Manual 378. Large p-values are consistent with the null hypothesis, and small p-values undermine the hypothesis. Id. at 402. However, p does not express anything about the accuracy of the null hypothesis, or the probability that it is true. Rather, it is computed by assuming the hypothesis is true. Id. In this case, a small p-value would be consistent with a disparate impact by the selection process. Id. Wertheimer’s report indicates the p-value for the difference in the termination rates between the two age groups was quite small (0.0000004). Wertheimer Rep. D-2, l. 15. If the null hypothesis is true, then the following conclusions drawn from Wertheimer’s analysis might be useful. Wertheimer found the disparity in termination rates between the two age groups “highly statistically significant.” Use of the term “significant” in this context describes the degree of difference between the expected results, in light of the null hypothesis, and those actually obtained. Essentially it is just a label for certain kinds of p-values. Id. at 380. Its meaning is no different from, and it is subject to the same limitations as, p-values. Id. “Significant differences are evidence that something besides random error is at work, but they are not evidence that this ‘something’ is legally or practically important.” Id. That is why statisticians distinguish between statistical and practical significance. Id. “Significance comes no closer to expressing the probability that the null hypothesis is true than does the underlying 23-value.” Id. at 381. In other words, by first assuming that the two age groups will have equal rates of being selected for termination, Wertheimer tested for the actual selection rates. Next, he compared the actual results to the expected results, given his underlying assumption of no difference between the two age groups, and found the observed difference in selection rates “statistically significant.” Although Wertheimer’s conclusion is internally consistent with the model he structured, it simply means that, given the hypothesis of equality, the fact that the outcome was different suggested that something besides random error was at work. For him to take the next step and conclude that the results he obtained are consistent with the claim that Indiana Bell’s WRP treated older workers differently goes beyond the realm of scientific methodology and into speculation. The usefulness of statistical evidence cannot be underrated, particularly for cases involving possible discrimination. When the appropriate methods have been employed for collecting all of the data and information needed to reach the conclusion offered by the expert, and the conclusion fits the facts at issue, the court will not hesitate to entrust to the jury the task of weighing that evidence. However, when, as here, serious doubts exist about the accuracy and scope of the information considered by the expert, and about the appropriateness of the methods employed to arrive at his conclusions, this Court will not leave it up to the jury to sort it all out. For all these reasons, the Court finds that Wertheimer’s reports are inadmissible, and cannot form the foundation on which the plaintiffs build them theory of recovery under the ADEA. Because Bendick’s report is a mirror of Wertheimer’s results, it is also inadmissible. Defendant’s motion to strike all of the experts’ reports, opinions and testimony is GRANTED. II. MOTIONS FOR SUMMARY JUDGMENT A. Pattern or Practice Claims The ADEA does not prohibit an employer from taking actions that have a disparate impact on older workers. See 29 U.S.C. § 623(f)(1). Disparate impact as a theory of liability under Title VII provides a means of dealing with the residue of past discrimination. Troupe v. May Dept. Stores Co., 20 F.3d 734, 738 (7th Cir.1994). Past discrimination is not an issue in an age-discrimination claim, because the protected trait, age, is not immutable, it changes over time. Rather, the ADEA prohibits treating older workers differently because of their age, or taking any action based on stereotypical ideas about older workers’ capabilities without considering their actual abilities. Consequently, the plaintiffs’ proof must show either the existence of the prohibited intent to discriminate on the basis of age, or the universal treatment of older workers without accounting for them individual abilities. If there is no direct evidence of bias, as is the case here, the plaintiff will have to prove intent using circumstantial evidence. Employees may enforce the ADEA against their employer either in their individual capacities or as nominal plaintiffs in an opt-in class action. See 29 U.S.C. § 626(b) (incorporating 29 U.S.C. § 216(b) by reference); Grayson v. K Mart Corp., 79 F.3d 1086, 1096 (11th Cir.), cert. den., — U.S. —, 117 S.Ct. 435, 136 L.Ed.2d 332 (1996). The employees here have done both. To state an individual claim of age discrimination under the ADEA, plaintiffs employ either the direct method of proof, or use the indirect, burden-shifting, method described in McDonnell Douglas v. Green, 411 U.S. 792, 93 S.Ct. 1817, 36 L.Ed.2d 668 (1973). The type of evidence that may be offered under either method includes both direct and circumstantial evidence. See Troupe, 20 F.3d at 736. Using the direct method, a plaintiff may prove discrimination by providing evidence that the fact-finder may interpret as the employer’s acknowledgment of discriminatory intent. Chiaramonte v. Fashion Bed Group, Inc., 129 F.3d 391, 396 (7th Cir.1997). For it to succeed, such evidence must clearly demonstrate the motivation of the person who made the contested employment decision. Id. (citing Cheek v. Peabody Coal Co., 97 F.3d 200, 203 (7th Cir.1996)). Essentially, the plaintiff must prove that age was the determining factor in that decision. Id. When plaintiffs sue as class representatives, based on an alleged pattern or practice of age discrimination, their burden is slightly different from that necessary to prove an individual claim. A pattern or practice claim’s prima facie case is established by showing that unlawful discrimination has been the company’s standard operating procedure, the regular rather than the unusual practice. EEOC v. Chicago Miniature Lamp Wks., 947 F.2d 292, 297 (7th Cir.1991); Coates v. Johnson & Johnson, 756 F.2d 524, 532 (7th Cir.1985). Usually, the proof offered for a pattern or practice claim consists of “statistical evidence demonstrating substantial disparities in the application of employment actions as to ... the protected group, buttressed by evidence of general policies or specific instances of discrimination.” Coates, 756 F.2d at 532 (citing International Bhd. of Teamsters v. United States, 431 U.S. 324, 336, 360, 97 S.Ct. 1843, 52 L.Ed.2d 396 (1977)). Evidence, either direct or circumstantial, of discrimination against particular individuals may form part of the proof. Essentially there are three types of circumstantial evidence that will satisfy a plaintiffs evidentiary burden. The first, and the most commonly used, type includes evidence such as suspicious timing, ambiguous oral or written statements, behavior toward or comments directed at other employees in the protected group, and other “bits and pieces” from which a finder of fact could draw an inference of discriminatory intent. Troupe, 20 F.3d at 736. Second is evidence that employees who are similarly-situated to the plaintiff, other than in the characteristic on which the employer is forbidden from basing an employment decision, received systematically better treatment from the employer. Id. This is the category in which statistical evidence usually falls, commonly referred to as comparative evidence. The final category includes evidence that the plaintiff was qualified for the job in question, but was passed over in favor of, or replaced by, a person not having the forbidden characteristic, and the employer’s stated reason (or explanation) is unworthy of credence (pretextual). Id. Each of these categories of evidence is sufficient of itself to support a judgment for the plaintiff, depending on its strength in relation to the other evidence. Id. The first category, however, often contains various pieces of evidence, none of which are conclusive by themselves, that together may “compose a convincing mosaic” of discrimination. Id. at 737. Nevertheless, the relative strength of each part of the mosaic must be measured against the remaining evidence. If the plaintiffs evidence of discrimination is offset by the defendant’s evidence to the contrary, it may not form a part of the mosaic. Two characteristics of this case further complicate the plaintiffs’ evidentiary burden: 1) it is a class-based pattern or practice claim; and 2) the terminations occurred during a large, corporate-wide restructuring and resizing. The plaintiffs have presented no direct evidence of discrimination, attempting instead to use circumstantial evidence from all three categories to prove their case under both methods, or at least demonstrate a genuine issue of material fact. The foundation of their circumstantial proof is the statistical analyses that were conducted by the plaintiffs’ experts. The Court, however, has deemed that evidence to be unreliable, irrelevant, and confusing, and has excluded it from consideration on this motion. What evidence remains must now be examined. The most useful type of evidence for proving a pattern or practice claim of discrimination is statistical evidence of disparate promotion, hiring, or discipline practices when the employee is a member of the protected class. See Mozee v. American Commercial Marine Serv. Co., 940 F.2d 1036, 1042 (7th Cir.1991) cert. denied 506 U.S. 872, 113 S.Ct. 207, 121 L.Ed.2d 148 (1992). Another common type of evidence is that of a general policy of discrimination against the protected group around the company. Id. Here, the plaintiffs have attempted to present both types of evidence in support of their pattern or practice claim. Even though the Court has excluded the experts’ reports and opinions about the results of the WRP, it cannot ignore the actual results. Fifty-six of fifty-nine employees who were selected for termination under the WRP were forty years of age or more. This result justifies paying close attention to the proffered evidence of a general policy of age discrimination around the company, and any anecdotal evidence. In Mozee, a class action pattern or practice case involving claims of race discrimination, the plaintiffs accused their employer of engaging in a subjective promotion and discipline system that adversely affected African-Americans. Id. at 1042-43. They offered statistical evidence to support a disparate impact claim regarding promotions to certain jobs, and the district court found that their statistics were not defeated by the employer’s “more refined” statistical analysis. Id. at 1043. This type of evidence is probative of both disparate impact and of a pattern or practice of disparate treatment. Id. at 1042. The plaintiffs also succeeded in presenting enough statistical evidence to show a disparate impact on African-Americans by the employer’s practice of administering discipline. Id. at 1043. For their class pattern or practice claim based on alleged disparate treatment by the employer, the Mozee plaintiffs added anecdotal evidence from their individual claims to the statistical evidence. Id, The court also considered non-statistical evidence of class-wide discrimination. Id. “Often plaintiff classes make out pattern or practice showings through a combination of strong statistical evidence of disparate impact coupled with anecdotal evidence of the employer’s intent to treat the protected class unequally.” Id. at 1051. In Mozee, the district court found, after trial, that the employer had violated Title VII by its employment practices that had a disparate impact on, and constituted a pattern or practice of intentional discrimination against, African-American employees. Id. at 1050. On appeal, the Seventh Circuit found the anecdotal evidence of disparate treatment of African-Americans strong enough to affirm the district court’s finding of classwide disparate treatment, but vacated and remanded the trial court’s findings on the issue of disparate impact of the employer’s disciplinary practices. Id. Consequently, the court noted that, in light of the flaws in the plaintiffs’ statistics for the disparate impact claim, the trial court “must reevaluate as well its finding of class-wide disparate treatment.” Id. That finding rested in part on “what appears to be a strong statistical disparity in terminations,” which may be significantly affected by the re-evaluation of the discipline claims. Id. The Mozee court also had non-statistical evidence of a general policy of discriminating, such as evidence that the employer stamped the initials of a local African-American civil rights activist on the files of each African-Anerican job applicant. Id. Other evidence included a sign found in the rigging yard that said “No Nigger Riggers,” evidence that only whites were allowed to request lateral bids, and that the employer had a “dismal” record of affirmative action compliance and implementation. Id. at 1044. This evidence was considered sufficient for a finding of a pattern or practice of disparate treatment. Id. at 1051. When measured against the type of evidence presented in Mozee, the plaintiffs’ evidence here does not come close to being strong enough to offset the unreliability of their statistical evidence. Even if it did, one must be careful when applying a case such as Mozee, which dealt with race discrimination in a non-RIF setting, to an ADEA-based pattern or practice case emerging from a corporate downsizing. First, a plaintiff in a race discrimination case may recover under either a disparate treatment or a disparate impact theory of liability, unlike in an age discrimination claim, where a disparate impact theory is not available. Second, when the alleged discrimination occurs in the ordinary employment context the court is focusing on an individual firing or failure to promote and it is easier to assess the validity, subjectivity, or objectivity of the employer’s decision. Whereas, in a RIF context with multiple employment decisions made at one time, all of the employees could be meeting their employer’s legitimate expectations, but some of them will be terminated because they compare less well with those who remain. The legitimate reason for the terminations in question is presumed to be the force reduction, and not any alleged individual misconduct or inadequate performance. Third, when an employer can show that it used a detailed, criteria-based, decentralized process for selecting those to be discharged, the RIF becomes virtually impregnable to claims of discrimination. Nevertheless, the plaintiffs here have designated “six major pieces of evidence,” in addition to their statistical evidence, that they contend demonstrate a pattern or practice of discrimination. Pis.’ Brf. in Oppos. To the Mot. Of Def. for Summ. J. On Plfs’ Clms. of a Pattern or Practice of Age Discrim, at 40. The Court will consider each piece in turn. They have also pointed to evidence that individual employees were allegedly selected for termination or demotion because of their age. However, to elevate individual claims to the level of a pattern or practice requires evidence that the discriminatory treatment is typical of the employer’s employment practices, or that such treatment was motivated by a policy of discrimination against the protected class that pervades the entire business. King v. General Elec. Co., 960 F.2d 617, 623 (7th Cir.1992). For example, one way to satisfy this evidentiary requirement would be to identify a policy of discrimination in the employer’s other employment practices that is reflected in the practice at issue. Id. This is not easy, for the employee must prove by “significant evidence” that the employer’s policies include the regular, purposeful, less-favorable treatment of members of a protected group. Id. In an attempt to make that showing, the plaintiffs point to evidence that in late May or early June, 1992, Chairman Bill Weiss (“Weiss”) discharged four senior managers in the Ameritech Corporation hierarchy, three of whom were in their fifties. Plfs.’ Ex. 32, Weiss Dep. at 52. According to the plaintiffs, these managers were “replaced” by substantially younger managers who Weiss dubbed “young team captains.” Plfs.’ Ex. 5, Rao, The Painful Remaking of Ameritech. The flaw in this argument is that no evidence connects the departure of four senior executives at Ameritech Corporation with the plaintiffs’ selections for termination at Indiana Bell. Weiss was not a decision-maker at Indiana Bell. Moreover, Weiss’ designation of the four men as leaders in the process of examining how to refocus Ameritech Corporation’s business occurred three months before the discharge of the senior executives. Plfs.’ Ex. 32, Weiss Dep. at 52. The younger individuals included Gary Drook, president of Ameritech Publishing, Dick Brown, president of Illinois Bell, Barry Allen, president of Wisconsin Bell, and Richard Notebaert, president of Indiana Bell, and their ages at the time were between forty-three and forty-seven. Id. at 52, 57. The senior executives who left included Robert Barnett, age fifty-one and vice-chairman and president of Ameritech Bell Group, Bruce DeMaeyer, fifty-three and head of Ameritech Mobile Communications, Kenneth Millard, forty-five and president of Michigan Bell, and Harold D’Orazio, fifty-two and a senior executive at Ameritech Services. Id. at 53-61. None of the younger managers succeeded to any of the positions held by the senior executives who left, with the possible exception of Notebaert, who became president of Ameritech Services sometime in late May or early June of 1992, after D’Orazio took early retirement. Id. at 175-76. Rather, when these managers were appointed to the “lead team,” as Weiss put it, each was to retain all of his former responsibilities. Id. Although a contemporaneous newspaper account characterized their appointments as a decree “that these individuals would run the company,” Weiss denied having said or meant that. Id. at 53. Even if their appointments could be considered “better treatment,” none of the younger executives could be considered sub stantially younger. See Hartley v. Wisconsin Bell, Inc., 124 F.3d 887, 893 (7th Cir.1997) (ten-year difference in age is presumed “substantial”). In addition, given the fact that the younger managers were appointed months before the senior executives left, the Court cannot characterize their appointments as “replacing” the senior managers, and refuses to endorse that inference as reasonable. Another incident offered by the plaintiffs as evidence of Indiana Bell’s alleged general policy of treating older workers worse than younger ones is the selection of approximately twelve managers in salary grades seven and above for termination in September 1992. According to the plaintiffs, all of those selected were more than forty years old, and eight of them were fifty or older. The documents cited by the plaintiffs, however, only included salary grades six, seven and eight, making it difficult to assess the evidentiary support for this contention. See Plfs.’ Ex. 16. Moreover, the plaintiffs have not shown that age was a determining factor for the termination of these managers, only that each manager was within the ADEA-protected age category. In fact, no evidence was presented about the selection process used. This represents another attempt by plaintiffs to prove then-case using the disparate impact method. Without evidence of how these managers were selected, or evidence that they were selected because of their age, or because of age-based stereotypes, the Court cannot find that these discharges provide any evidence of a general policy of discriminating against older workers. The remaining four pieces of key evidence proffered by the plaintiffs fare no better in proving the pattern or practice case. Plaintiffs spend most of their time discussing IBT president Thomas Reiman’s (“Reiman”) comment about having a continual influx of “new young crazy people” to keep the business from “drying up.” Pis.’ Ex. 33, Reiman Dep. at 299. Although this remark could be interpreted as indicating an intent to replace terminated employees with young people, that inference is not reasonable in light of its context. Reiman was responding to questions during a video conference session shortly after he became president of IBT. Id. at 297-98. One question was whether IBT would begin to promote people again, after years of a promotion freeze. Pis.’ Ex. 2, Trans. Of Videotape. Reiman first acknowledged that the question reflected one of the challenges to the business, and before giving a direct response he discussed how to “develop folks” and keep them challenged, without necessarily giving them promotions. Id. He then admitted that he did not foresee the company being in a position, in the next five years, to give the number of promotions it had in the past. Id. He also contrasted promoting with developing, and suggested that his goal was to change “the whole status of what training and development is.” Id. Specifically, he said that, as opposed to being something that is just “fit in” whenever there was time, training and development would become a “major focus” of the company. Id. Reiman remarked that when people have done a specific job for a while and are becoming a little bored, the company needs to have a system in place that will allow them to transfer and be retrained. Id. In addition to all this, he said, the company needed to be able to promote people, and to find a way to hire new people. The latter comment seemed an afterthought, as did the subsequent remark, which forms the basis for plaintiffs’ claim of age-related bias. In the context of his expressed concern for keeping employees challenged, developing them, hiring new people, and by implication developing the business, Reiman said, “The business will absolutely dry up and die if we don’t have a continued influx of new young crazy people-"Id. At his deposition, Reiman explained that by “new young crazies,” he was referring to a subset of the types of workers IBT needed to hire to maintain diversity, which would consist of people who did not think the same way that he might think, or that an engineer might think. Reiman Dep. at 299. He specifically noted that his remarks at the video conference were not related to the subsequent downsizing, but were instead tied to ideas that had been generated during his Breakthrough Leadership sessions. Id. Likewise, when explaining how the company would make room for new hires and promotions, Reiman said, “we’re gonna [sic] have to find a way for people that aren’t performing up to the standards that we, as employees of the company, set exit the business .... we’re goin’ to have to find a way to get the ... poor performers out of the business and replace those people with new hires and open up opportunities for promotion.” Pis.’ Ex. 2. Thus, even if he were drawing a connection between hiring “young crazies” and terminating incumbents, Reiman clearly expressed the criteria that would be used for selecting those to be discharged: performance. According to Reiman, who was a team captain for the “Breakthrough! Leadership” project, the subsequent resizing program was a response to the company’s change from a regulated monopoly to a competitive environment. Id. at 227. “One of the things that stood out was that there was a big distance between the front line person who dealt with the customer and where the authority and the information in the company resided. And so the effort was on the one hand to reduce that bureaucracy ... and on the second to get our costs more in line.” Id. at 227-28. That distance was populated by what is commonly called middle management. The size of the workforce was also a problem, Reiman noted, both from the cost standpoint and from an efficiency standpoint. Id. at 230. For example, a business office service representative would have to go through five layers of management to get to someone who had the authority to say “yes” to a customer. Id. The goal of the resizing was to reduce that to only one or two layers, “or even better still, to empower that person on their own to do it.” Id. Instead of indicating the intent to replace an aging workforce with younger employees, Reiman’s comment, taken in context, falls into a general category of “stray remarks” that by themselves do not prove actionable discrimination, even if repeated. See Shager v. Upjohn Co., 913 F.2d 398, 402 (7th Cir.1990). The remark is but one sentence buried in two single-spaced, typed pages of the transcript of this portion of the video conference. It also constitutes a random reference to hiring, that was only tangentially related to the bulk of Reiman’s remarks. Not only does the remark constitute a departure from the main thrust of Reiman’s other remarks, but it also does not reflect age bias. An expressed willingness to hire people who see things differently than the rest of the company’s employees, does not constitute a preference for younger workers, and an inclination to weed out older ones. The plaintiffs argue that Reiman’s remark is ambiguous, and that its meaning should be discerned by a jury. The Court cannot agree. In order for this “fact” to require deliberation by a jury, it would need to be material to the outcome of the case. That is its most significant failing. Reiman was not a decision-maker in the sense that he could control who among the employees at IBT would be selected for termination. He had no involvement in establishing the actual selection process, or in the mechanical creation of “at-risk” lists, or in discussing and ranking those employees who were at risk. See McFeeley Aff. ¶ 15. His only involvement was to draw lines on lists of ranked employees in the various salary grades, below which employees would be asked to leave. No evidence has been presented to show that his involvement in the process could have affected its outcome. Moreover, the remark at issue was made several months before the 1992-1993 Workforce Resizing Program began, and nearly six months before any of the plaintiffs’ terminations. For these reasons, the Court finds that Reiman’s remark is too remote in time and subject-matter to be a material fact in dispute. Frequently during this litigation, the plaintiffs have taken off-hand, vague and generalized remarks and offered them as an indication that the defendant’s motive was to discriminate against them because of their age. It is not clear whether plaintiffs mean that the overall decision to resize was intended to weed out older workers, or if the actual selection decisions were made with that intent. If their theory is the former, it is defeated by voluminous evidence of the self-analysis conducted by Ameritech and its subsidiaries to determine how to improve their business and become more competitive. It also pales in the light of frequent contemporaneous, non-discriminatory, explanations for the decision to resize, and the specific instructions in the resizing program’s selection guidelines that age should not be a factor in any decisions. On the other hand, if plaintiffs see these remarks as supporting a finding that each decision during the selection process was made with the intent to get rid of older workers, they have failed to account for the fact that' the quoted general remarks, and the sentiment plaintiffs wish the Court to read into them, were not made by the relevant decision-makers. The evidence is that the process used by IBT for selecting those to be terminated was developed by Robert McFeeley (“McFeeley”), IBT’s director of human resources at the time. See McFeeley Aff. ¶¶ 3,13,14. McFeeley received no input from IBT president, Reiman, regarding the criteria to use to determine which managers should be discharged, or about any .particular results that were to be achieved. Id. ¶ 15. Likewise, McFeeley’s boss, Fred Peters, made it clear that the process was to select people based on “performance and skills, not age, years of service, pension status, or any other prohibited characteristic.” Id. The plaintiffs have failed to offer any evidence to contradict McFeeley’s testimony regarding these matters, which further indicates the absence of evidence connecting Reiman’s remark to the overall selection process. Neither can his remark be associated with any individual termination decisions, as a review of the selection process will demonstrate. The first step of the selection process was for human resources managers to mechanically rate employees within each salary grade according to their most recent evaluations of managerial skills, assessments of promotional potential, and their past two-years’ performance. Id. ¶ 20. The greatest weight was given to the information about the employees’ performance over the past two years. Id. Each of these three components of the “mechanical” ranking process reflected information that had been gathered and documented in the employees’ files long before any decision had been made about resizing, using criteria that had been in place at IBT for many years. The next step involved hands-on rankings of those who were placed on the “at-risk list” by their scores in the mechanical process. Id. ¶ 26. The department heads were allowed to review these lists and add to or subtract from them as long as they provided written explanations for those decisions and obtained approval from the vice president of human resources, Fred Peters. Id. The departmental reviews resulted in only a few changes, the most common reasons for which were a decline in performance that justified adding an employee, or deleting an individual who had unique skills or abilities that could not be easily replaced. Id. ¶ 27. After the “at-risk” lists were finalized, eight ranking committees were convened, comprised of managers at salary grades six or above, to discuss and rank all of the employees on the lists. Id. at 33. The eight committees met in mid-October 1992 for ranking sessions that were facilitated by a member of the human resources department. The process of ranking involved giving each committee member a summary of the data that caused the employee to be on the list, and making sure that at least one person on the committee was familiar with the work performance of each employee to be ranked. Id. ¶ 34. The committee numerically ranked each employee based on discussions during the ranking sessions. Id. ¶ 37. At the time of the rankings, the committee members had been provided with no information about any employee’s age, pension status, race, gender, or other protected characteristic, nor did they know how many employees in any group would ultimately be selected for termination. Id. ¶¶ 34, 39. Once the rankings were completed, the committees’ task was finished. The lists of ranked managers were then given to IBT’s officers, who had to determine where the lines would be drawn between those to be retained and those to be terminated. Id. ¶40. McFeeley attended these officer sessions, held in late October 1992, and noted that the officers did not change any of the rankings, or base their decisions about the lines on any employee’s age. Id. Instead, then decisions were governed by the number of managers who had volunteered for early retirement in light of the overall goal of reducing the workforce by ten to fifteen percent, the business needs of the company, and the number of employees they predicted would accept the early retirement package. Id. ¶ 41. Thus, assuming that Reiman’s comment reflected a preference for younger workers, his participation in the “line-drawing” could not have affected the rankings of any employees or influenced the overall number of older workers who were selected. When viewed in their historical context and in the context of the other evidence, Reiman’s remarks do not justify that assumption, or reflect any intent to select workers for termination because of their age. On the contrary, it is apparent that the resizing program grew out of a corporate-wide restructuring of the business to become more competitive. Although many people in management positions at IBT would be terminated by the program, their terminations occurred after each employee was carefully evaluated by a group of higher level managers, using specific criteria relating to the projected business needs identified by IBT’s strategic planners. That a disproportionate number of those people were over the age of forty does not mean that the entire resizing program was an intentional plan to discriminate against older workers. So far, none of the plaintiffs’ evidence creates an issue of fact for trial. The fourth category of evidence proffered by the plaintiffs to show that IBT used age to determine who would be selected for termination is the alleged “excluded younger workers.” According to the plaintiffs, IBT excluded seventy workers who had five years or less of service from the selection process, and of these seventy the median age was twenty-seven. In the plaintiffs’ own words, “Reiman and Indiana Bell decided to exclude those managers hired within the last five years from the resizing Program by presumptively excluding them from the ‘át risk’ lists.” Pis.’ Brf. at 41. In support of this contention, they cite a document entitled “Proposed Selection Process,” Pis.’ Ex. 1, and another one that appears to be a list of Indiana Bell employees with less than five years of service, sorted by age. Pis.’ Ex. 19. The first exhibit is McFeeley’s initial proposal to the IBT officers for how to select employees for termination. See McFeeley Aff., ¶ 20. It contains some hand-written notes on it, including the note “Add to Basic List” next to the section covering what to do with managers who have less than five years of service. Pis.’ Ex. 1. Plaintiffs make much of the fact that the first statement in this section is “These managers will be excluded from the basic process.” Id. Standing alone, this statement would seem to support the plaintiffs’ theory that those employees with five years or less were excluded from the process, which would raise factual questions about this alleged “different treatment” of younger employees. However, the statement does not stand alone, in spite of plaintiffs’ assertions to the contrary. In addition to the hand-written note mentioned above, the exhibit álso contains the following instructions for handling these employees: -HR will research performance indicators and provide to Department. -Departments will review performance skills and potential. -Departments will indicate any that should be considered “at risk”. -HR will maintain a company list. Id. Also included in this exhibit is a page with sections titled “Treatment of Missing Data,” and “Less Than 5 Years Service.” Id. McFeeley suggested that the latter category be handled by “selecting all,” researching merit increases, and researching any other available da