Terry A. ROCHA, Plaintiff-Appellant, v. GREAT AMERICAN INSURANCE CO., Defendant-Appellee

850 F.2d 1095 47 Fair Empl.Prac.Cas. 208, 46 Empl. Prac. Dec. P 38,073, 26 Fed. R. Evid. Serv. 214 Terry A. ROCHA, Plaintiff-Appellant, v. GREAT AMERICAN INSURANCE CO., Defendant-Appellee. No. 87-3340. United States Court of Appeals, Sixth Circuit. Argued March 15, 1988. Decided June 24, 1988. James B. Helmer, Jr. (argued), Ann Lugbill, Cincinnati, Ohio, for plaintiff-appellant. Louis F. Gilligan, Cincinnati, Ohio, Kalvin M. Grove (argued), Cary Schwimmer, Fox and Grove, Chartered, Chicago, Ill., for defendant-appellee. Before JONES and GUY, Circuit Judges, and CONTIE, Senior Circuit Judge. NATHANIEL R. JONES, Circuit Judge. — 1 — Plaintiff-appellant Terry A. Rocha appeals from a jury verdict entered in favor of Great American Insurance Company ("Great American" or "the Company"), the defendant-appellee, in this action for age discrimination. For the following reasons, we find that the court erred in its continuous commentary and interrogation of plaintiff's witnesses, thus depriving Rocha of the fair and impartial trial to which he was entitled. Accordingly, we reverse and remand this action, thereby granting Rocha's request that he be given a new trial.I. — 2 — Plaintiff-appellant, Terry Rocha, was hired by the defendant-appellee, Great American, in 1970 as the Director of its Statistical Accounting Department. From 1970 to 1983, Rocha reported directly to a Mr. Clovy Reed, Great American's Chief Financial Officer. In 1983, Reed, at the age of 52, voluntarily retired. Upon his retirement, he recommended that Robert Adams, a company employee in his late thirties, become his successor. The President of Great American followed that suggestion and thereafter Rocha reported directly to Adams. — 3 — While there was and is no dispute that Rocha was knowledgeable and adequately performed his statistical accounting duties, his superiors did have several discussions with him concerning his abrasive management style. According to Reed, several employees complained that Rocha was unduly critical and aloof and that he gave them the impression "that he was on one level and they were on another one." J.App. at 509. Reed also testified that he told Rocha that he had to change his managerial attitude because he was causing people to quit the Company. Rocha's supervisors, however, never made any entry in Rocha's personnel file regarding his alleged management problems, and Rocha testified that Reed's comments to him about his management style were routine suggestions made in the course of "casual conversation" and were not official reprimands. — 4 — Nevertheless, after Reed left the Company, Rocha's new supervisor, Robert Adams, also concluded that Rocha's treatment of his subordinates had a detrimental effect on the morale of Great American's Statistical Accounting Department. On that basis, Adams decided to relieve Rocha of his supervisory role over the department. Adams testified that he was particularly concerned about two key people in the department, John Reyman, a 49 year old Assistant Director, and 55 year old Gerry Recker, both of whom had expressed unhappiness with Rocha's management style. — 5 — After Rocha was removed from his supervisory role, Adams created a position for Rocha in which he could succeed by utilizing his technical expertise without having to manage people. Although company policy did not require Adams to create such a position, Adams testified that he did so because, in view of Rocha's years of service to the Company, he wanted to give him an opportunity to make a fresh start in a position he could perform. — 6 — The job created for Rocha was a special projects position where he could use his accounting expertise and experience to accomplish projects assigned to him by Assistant Vice Presidents Gary Gruber and Robert Maly. In his new position, Rocha reported directly to 29 year old Gruber. When Rocha was placed in this newly created position in October 1984, he retained his same job title (Director), his same salary ($57,000), and his same benefits. — 7 — According to the Company, Rocha did not perform his new job satisfactorily. At trial, the Company's witnesses pointed to several instances in which projects assigned to Rocha were not completed, and the person who had originally assigned the project ended up having to complete it on his own. Again, there was no mention in Rocha's personnel file of any problems regarding his performance in his new position, and Rocha testified that none of his supervisors ever told him that there were problems with his work. At trial, Rocha did admit that certain long-term projects were not completed, but he explained that this was because certain people with whom he was to consult on the projects left the Company sometime after they were assigned to him. — 8 — Rocha also testified that, on April 18, 1985, Adams called him into his office and told him that his position was to be eliminated and that his employment with Great American was terminated, effective April 19, 1985. Rocha testified that Adams explained that the cost of Rocha's position could not be justified by the benefit the Company was receiving. At the time of his discharge, Rocha was 53 years old and had been with the Company for 14 1/2 years. Adams testified, however, that Rocha's age had nothing to do with the decision to eliminate his position. Instead, as seen in Rocha's termination report, the Company and Adams indicated that Rocha had been terminated for "lack of work." After Rocha's position was abolished, the work he had been doing was assigned to the same people who had originally been responsible for it before his position was created. — 9 — On October 30, 1985, Rocha filed this action in the United States District Court for the Southern District of Ohio against Great American alleging that his discharge was in violation of the Age Discrimination in Employment Act ("ADEA"), 29 U.S.C. Sec. 621 et seq. (1982). Together with the complaint, Rocha served interrogatories and a document request upon the Company. The Company objected to much of Rocha's requested discovery and a motion to compel was filed. After a hearing on August 8, 1986, Chief Judge Carl Rubin granted the plaintiff's motion in large part and assessed sanctions against Great American for unreasonably withholding discovery. Specifically, Judge Rubin ordered the Company to provide the personnel files of all Great American employees at the same level in the Company as Rocha, at the level above, and at the level below; and to provide the names of other discharged employees, information on other lawsuits on age discrimination, and information on job openings that occurred during the relevant period. The court also informed the parties that a federal magistrate would monitor the discovery process throughout the remainder of the case. — 10 — Thereafter, Rocha was provided with over 350 personnel files of employees similarly situated to Rocha, which included the files of 200 or so employees in the field (i.e., not in the Cincinnati office). Rocha, however, was not satisfied, and on November 4, 1986, filed with the magistrate a request to enforce Judge Rubin's discovery orders. Specifically, Rocha claimed that the defendant had not produced the files of employees at the "manager" level--which plaintiff claimed was the level immediately below him--and had not provided hiring and job application materials relevant to an "arrangement" plaintiff believed defendant had with the University of Cincinnati to hire its recent business school graduates. Plaintiff asserted that these items of information, along with other information not provided, were essential to his statistical expert's determinations. — 11 — Magistrate Robert Steinberg enforced Judge Rubin's order requiring the defendant to produce the files at one level above and one level below plaintiff's job level. However, because Great American represented that the level below plaintiff was that of "assistant director," not "manager" as plaintiff claimed, the magistrate ruled that if plaintiff's expert needed the files of the manager-level employees in order to obtain an adequate sample group, "plaintiff must submit an affidavit from his expert statistician to that effect." J.App. at 21. The magistrate also ordered that certain "informal" personnel files, kept not in the personnel office but by an employee's supervisor, were to be produced for employees one level below and one level above that of plaintiff. — 12 — Subsequently, after plaintiff's expert filed an affidavit claiming a need for the files of the manager-level employees, the magistrate, in a January 21, 1987 letter to plaintiff's counsel, denied further production of documents. As the magistrate noted, plaintiff's expert had stated in his affidavit that "it is desirable to have a pool of at least 100 of the most similarly situated employees for certain types of statistical analyses which should be conducted in this case." J.App. at 46. However, the magistrate concluded that because plaintiff already had over 300 personnel files "of employees similarly situated," there was no reason to order the defendant to produce further files. — 13 — Plaintiff appealed some of the magistrate's discovery rulings to Judge Rubin. A hearing was held on February 20, 1987 to resolve those matters. As to the request for the manager-level files, Judge Rubin agreed with the magistrate that those files did not have to be provided. Judge Rubin also accepted the defendant's representation that no documents existed regarding its purported "arrangement" to hire University of Cincinnati's business school graduates. Finally, Judge Rubin noted that the defendant had complied with the magistrate's order to produce to Rocha the so-called "informal personnel files." — 14 — With the discovery stage completed, the case was set for trial. On February 26, 1987, however, Judge Rubin entered an order that the case would be tried by the Honorable Hubert L. Will, Senior Judge for the Northern District of Illinois, as part of the University of Cincinnati Law School's Smith Tyler Judge in Residence Program. Pursuant to this order, the parties and their counsel reported to Judge Will at the Law School on March 2, 1987. — 15 — A jury was selected on March 2, 1987 and the three-day trial began. The plaintiff's case was based primarily upon the testimony of Rocha and the expert testimony of Dr. William Feinberg. At the close of plaintiff's case, Judge Will denied the defendant's motion for a directed verdict. Thereafter, defendant proceeded with the presentation of its defense which consisted primarily of the testimony of Adams and Reed, Rocha's former supervisors, as well as the testimony of Dr. Robert Wessel, a statistical expert who attempted to refute the testimony of plaintiff's expert. After the defense rested, plaintiff indicated that he had no rebuttal testimony to offer. The court then denied defendant's second motion for a directed verdict and sent the case to the jury. After deliberating for ninety minutes, the jury returned a verdict in defendant's favor. Rocha did not move for a new trial pursuant to Fed.R.Civ.P. 59, but instead appealed directly to this court. — 16 — Rocha's primary contention on appeal is that he is entitled to a new trial because the trial judge unfairly and prejudicially injected himself and his opinions into the case, repeatedly disparaged the probative value of plaintiff's evidence in the presence of the jury, continually interrupted plaintiff's counsel's examination of witnesses, and made several adverse and erroneous evidentiary rulings during the course of the trial. Additionally, Rocha claims that the court erroneously limited the scope of plaintiff's cross-examination of defendant's expert witness, thus denying counsel the opportunity to impeach his credibility. Rocha also claims that certain limits upon discovery, as well as discovery abuses by the defendant, prejudiced him and deprived him of a fair trial. Finally, Rocha contends that the court erred in refusing to give a jury instruction on whether the defendant wilfully violated the ADEA. — 17 — Because we find Rocha's first contention to be dispositive, we do not address the remaining issues raised by the parties in this appeal. In reaching our decision, however, we do note that this court cannot disturb the jury verdict and judgment below unless Rocha has established error so prejudicial that he was denied substantial justice. Fed.R.Civ.P. 61; Prater v. Sears, Roebuck & Co., 372 F.2d 447, 448 (6th Cir.1967) (per curiam). And, even if error is shown, it must be disregarded if it did not affect Rocha's substantial rights. Moreover, if despite the alleged error Rocha still does not have sufficient evidence to support a jury verdict in his favor, any error, committed by the trial court, could not have affected Rocha's substantial rights, and therefore, is harmless error. See Hallmark Industry v. Reynolds Metals Co., 489 F.2d 8, 14 (9th Cir.1973), cert. denied 417 U.S. 932, 94 S.Ct. 2643, 41 L.Ed.2d 235 (1974). — 18 — At the outset, we point out that our role in this appeal is not to retry the case. Rather, both sides have had the opportunity to present their case to the jury. Thus, the only issue here is whether Rocha can establish error so prejudicial that he was denied substantial justice. Because we think he has met that burden, we reverse the judgment for Great American. II. — 19 — Courts have long recognized that while trial judges have wide discretion to comment upon the evidence presented during a trial and to bring certain facts deemed important to the jury's attention, such activities must be performed with "strict neutrality and utmost impartiality." Travelers Insurance Co. v. Ryan, 416 F.2d 362, 364 (5th Cir.1969). The Travelers Insurance court cited an opinion by this court, Nunley v. Pettway Oil Co., 346 F.2d 95 (6th Cir.1965), in holding that the trial judge's comments exceeded the permissible scope of fair judicial play. Travelers Insurance, 416 F.2d at 364. In so doing, that court set forth its reasoning, which provides, in pertinent part, as follows: — 20 — It is precisely because of the important functions assigned to a trial judge, especially in a federal trial where these functions include the right to comment upon evidentiary matters, that the judge's duty must be performed with strict neutrality and utmost impartiality. By reason of his role, quickly observed by jurors, the judge is a figure of over-powering influence, whose every change in facial expression is noted, and whose every word is received attentively and acted upon with alacrity and without question. — 21 — The instant case affords a clear illustration of the force of these observations.... There is a difference between comment by the trial judge upon evidentiary matters, which if fair and unbiased is permissible, on the one hand, and comment by the judge upon the ultimate factual issues to be decided, on the other hand. The latter is not permitted. — 22 — As [stated] below, the trial judge in our case gave the jury his views upon ultimate issues of fact.... This was an invasion of the ultimate fact-finding function of the jury, prejudicial to the rights of the appellant. — 23 — Id. (emphasis in original). — 24 — This court, in Nunley, 346 F.2d at 99, recognized the same limitations upon a trial judge's right to comment on the evidence, citing a decision from the Supreme Court, Quercia v. United States, 289 U.S. 466, 53 S.Ct. 698, 77 L.Ed. 1321 (1933), as dispositive of the issue. See Quercia, 289 U.S. at 470, 53 S.Ct. at 699 ("[The] privilege of [a] judge to comment on the facts has its inherent limitations. His discretion is not arbitrary and uncontrolled, but judicial, to be exercised in conformity with the standards governing the judicial office.... This Court has accordingly emphasized the duty of the trial judge to use great care that an expression of opinion upon the evidence 'should be so given as not to mislead, and especially that it should not be one-sided'...."). — 25 — In light of this law, we now review the trial judge's statements and questions.1 III. — 26 — Because Rocha's case was so heavily reliant on statistical evidence of discrimination, his strongest argument on appeal pertains to Judge Will's handling of the expert testimony. Specifically, Rocha argues that Judge Will erred to his substantial prejudice by repeatedly interrupting the testimony of his expert, Dr. William Feinberg; by taking over the direct examination of his expert without reason; and by unfairly and prejudicially commenting on the substance of his expert's testimony and methodology. Rocha claims that the court's continual interference is easily seen by examining the record of his expert's testimony. Rocha contends that the court's comments and interjections number well into the hundreds. See J.App. at 307-89. Indeed, when examining the record, we find such prejudicial comments and biased interrogation. For instance, Judge Will made the following comment regarding one of Dr. Feinberg's statistical studies:It isn't a complete picture because it doesn't indicate those who came in from Great America's other offices into Cincinnati who were not here in [approximately 1982]. — 27 — J.App. at 423 (emphasis added). Also, the judge observed with regard to another of Dr. Feinberg's studies: — 28 — I really don't know what the reasons were [for the Company's increased use of "lack of work" terminations], and you don't either because you didn't look behind the terminations. I think it seems to me it is not only significant that there's a difference between what happened in those under forty and over forty, but it's also significant to see what happened to those in the same age group to see whether or not there was a change in policy. If the policy with respect to the forties and over didn't change, but the policy with respect to the under forties did change, that doesn't mean there's been discrimination with respect to those over forty, does it? — 29 — J.App. at 464 (emphasis added). And, at the conclusion of Feinberg's testimony, the court stated: — 30 — Thank you, Dr. Feinberg. You know what you did and why you did it. — 31 — J.App. at 471-72 (emphasis added). These comments, and others like them, certainly make it clear that the trial judge had some uncertainty with regard to the plaintiff's statistical evidence. While plaintiff's statistical evidence could be categorized as weak by this circuit's standards, see, e.g., Simpson v. Midland-Ross Corp., 823 F.2d 937, 944 (6th Cir.1987), where the court rejected the "assertion that a decreasing average employee age coupled with other questionable statistics proves the existence of impermissible age discrimination," the court's comments and actions seem more appropriate for a cross-examiner than a fair and impartial judge. The jury was surely affected by such behavior and certainly that was to the prejudice of plaintiff's substantial rights. — 32 — Similarly, even though the court has broad discretion to interrogate all witnesses, including expert witnesses, see e.g., Fed.R.Evid. 614(b), that "authority is, of course, abused when the judge abandons his proper role and assumes [the role] of [an] advocate...." Fed.R.Evid 614(b), advisory committee note. In this case, that is precisely what happened. Judge Will's interrogation of Dr. Feinberg and plaintiff's other witnesses exceeded his right to have matters more fully elaborated and to clarify testimony inarticulately or reluctantly given. See United States v. Vida, 370 F.2d 759, 768 (6th Cir.1966), cert. denied, 387 U.S. 910, 87 S.Ct. 1695, 18 L.Ed.2d 630 (1967). This circuit has ordered new trials in civil cases when such interrogations or interjections become impermissible and exceed the scope of fair judicial play. See, e.g., Watson v. Chesapeake & Ohio Ry. Co., 287 F.2d 662 (6th Cir.1961); Knapp v. Kinsey, 232 F.2d 458 (6th Cir.), cert. denied, 352 U.S. 892, 77 S.Ct. 131, 1 L.Ed.2d 86 (1956). Indeed, in Knapp, this court stated: — 33 — [T]he right of the trial judge to participate in the proceedings and to interrogate witnesses is not an unlimited one. Such participation should be exercised in conformity with the standards governing the judicial office. The judge should exercise self-restraint and preserve an atmosphere of impartiality. When the remarks of the judge during the course of a trial, or his manner of handling the trial, clearly indicate a hostility to one of the parties, or an unwarranted prejudgment of the merits of the case, or an alignment on the part of the Court with one of the parties for the purpose of furthering or supporting the contentions of such party, the judge indicates, whether consciously or not, a personal bias and prejudice which renders invalid any resulting judgment in favor of the party so favored. — 34 — Id. at 466 (citations omitted and emphasis added). — 35 — Because we find that Judge Will's comments fall within the Knapp reasoning, we hold that the court's actions constitute plain, i.e., reversible, error. — 36 — However, even if plain error is shown, a reversal and remand is in order only if all of plaintiff's evidence, including his statistical evidence, taken together as a whole, was sufficiently probative as a matter of law to support a verdict that Rocha's age was a determining factor in defendant's decision to discharge him. — 37 — For statistics to be valid and helpful in a discrimination case, "both the methodology and the explanatory power of the statistical analysis must be sufficient to permit an inference of discrimination." Segar v. Smith, 738 F.2d 1249, 1274 (D.C.Cir.1984), cert. denied sub nom. Meese v. Segar, 471 U.S. 1115, 105 S.Ct. 2357, 86 L.Ed.2d 258 (1985). As this court stated recently: — 38 — Unless the statistics, standing alone or in comparison, are sufficient to lead the mind naturally to the conclusion sought, they have no probative value; they do not move the proof one way or another. — 39 — Simpson v. Midland-Ross Corp., 823 F.2d 937, 944 (6th Cir.1987) (emphasis added). — 40 — Plaintiff's statistical evidence consists of four studies which, Dr. Feinberg suggests, tend to show: 1) that the average age of the Company's employees had decreased from 1982 to 1985; 2) that a "suspiciously" large drop in the number of employees at the higher management levels of the Company who were between the ages of 50 and 59 (plaintiff's age group) had occurred between 1982 and 1986, even though the total number of management employees had increased; 3) that those employees in the upper-salary groups who were over 50 years old were more likely to be terminated or to leave the company than were younger employees at the same salary level; and 4) that in 1985, employees over 40 years old were more likely to be terminated for "lack of work" (the reason given for plaintiff's termination) than younger employees, although that was not true in 1982. — 41 — Although these statistics may not be as strong as statistical evidence found in other cases, there was certainly enough evidence presented for the court to hold that Rocha had presented a prima facie case of age discrimination. Indeed, plaintiff's evidence demonstrated 1) that his job duties had been assumed by younger employees upon his termination; 2) that he had adequately and knowledgeably performed his statistical accounting duties; 3) that no documentation of his alleged performance problems existed despite company policy to the contrary; and 4) that he was a member of the protected age group. See Chappell v. GTE Products Corp., 803 F.2d 261, 265-66 (6th Cir.1986), cert. denied, --- U.S. ----, 107 S.Ct. 1375, 94 L.Ed.2d 690 (1987) ("To make out a prima facie case, the plaintiff must show that he was in the protected age group (40-70), that he was discharged, that he was qualified for the position and that he was replaced by a younger person."). As such, Rocha made out a prima facie case of age discrimination. Indeed, even Judge Will was sufficiently satisfied with the evidence presented so as to deny the defendant's motions for directed verdict and send the case to the jury. Therefore, we reject defendant's argument that Rocha's evidence failed to establish a prima facie case. See United States Postal Service Bd. of Governors v. Aikens, 460 U.S. 711, 713-14, 103 S.Ct. 1478, 1480-81, 75 L.Ed.2d 403 (1983). — 42 — We also believe that Rocha presented enough evidence which could have supported a jury verdict in his favor if that evidence had been considered by the jury without tainted commentary from the court. However, because the jury's view of Rocha's proof was tainted, the verdict rendered was unfavorable to him. Therefore, inasmuch as Judge Will felt that Rocha had the right to present his case to the jury, we believe that right should have been exercised unhindered by biased commentary and interjection. Thus, we hold that Rocha's substantial rights were prejudiced and therefore he is entitled to a new trial. However, we do not hold that a reasonable jury could find only for Rocha. Rather, we hold only that he is entitled to present his case to a fair and impartial tribunal. — 43 — Therefore, for all the foregoing reasons, the judgment of the lower court is REVERSED and this action is REMANDED for further proceedings consistent with this court's opinion. — 44 — RALPH B. GUY, Jr., Circuit Judge, dissenting. — 45 — I respectfully dissent for three reasons: (1) the plaintiff did not make out a prima facie case and the defendant should have had its motion for directed verdict granted; (2) the trial judge played an active role in the trial but did not intrude to the point where plaintiff was prejudiced; and (3) in this age of the "forensic expert," a trial judge acts within the bounds of discretion when an "expert" is extensively questioned. I. — 46 — The plaintiff did not present any direct evidence at trial showing that he was discharged because of his age. Instead, plaintiff sought to prove age discrimination through the use of circumstantial and statistical evidence. Most of the circumstantial evidence consisted of plaintiff's personal observations and his general impression that older employees were leaving the company and being replaced by younger employees. In support of his claim of a pattern or practice of age discrimination, plaintiff gave five examples of older employees being replaced by younger employees. The record shows that three of the five individuals cited by the plaintiff were not "replaced" after they left the Company. Moreover, with respect to the two remaining individuals, one of them was transferred to California where he continues to work for the Company, and with respect to the other, there is nothing in the record to indicate whether his retirement was voluntary or involuntary. This anecdotal evidence regarding the retirement, transfer, or discharge of five out of defendant's 3,200 employees was insufficient to establish a "pattern or practice" of age discrimination. — 47 — Plaintiff also sought to introduce a list compiled by a co-employee which, according to plaintiff, supported his claim of age discrimination. The trial judge refused to admit the list into evidence because it was not limited to those individuals who had been discharged by the Company but, rather, supposedly included every individual who had left the home office of the Company for any reason. Therefore, the trial judge found that the list lacked probative value. I agree. — 48 — Plaintiff also attempted to establish a prima facie case by citing examples of what he refers to as "age-related discriminatory statements." For example, plaintiff testified that his superior, Mr. Adams, had told him that there were already "enough problems with age" at the Company. Ironically, the record shows that Mr. Adams' remark was made in the context of a meeting with the plaintiff in which he admonished the plaintiff that he should "not refer to age in dealing with employees because it had a potential for creating a problem." The impetus for Adams' remarks to plaintiff was a complaint from one of the plaintiff's subordinates who was upset about plaintiff's comments on her age. Apparently, plaintiff had told her that she had to adapt to the use of personal computers, and that her resistance to change was a sign of her age. The other "discriminatory" statements cited by plaintiff are either ambiguous or unsupported by the record, or both. I find that these types of isolated generalized statements are insufficient to establish a prima facie showing of age discrimination. See Chappell v. GTE Products Corp., 803 F.2d 261, 268 n. 2 (6th Cir.1986), cert. denied, --- U.S. ----, 107 S.Ct. 1375, 94 L.Ed.2d 690 (1987). — 49 — Plaintiff also attempted to establish the existence of age discrimination through the use of statistical evidence. Plaintiff's expert witness, Dr. Feinberg, testified that the average age of defendant's workforce was lower in 1985 than in 1982. Dr. Feinberg testified that, all things being equal, the average age of the workforce should increase over time and that the decline in the average age was "consistent with age discrimination." According to Dr. Feinberg's statistics, the decline in the average age of the entire Company between 1982 and 1985 was only .2 years, and the decline in age of the workers at the Cincinnati home office was only one year. This court has previously found that such slight changes in the average age of the workforce have no probative value in an age discrimination claim. See Simpson v. Midland-Ross Corp., 823 F.2d 937 (6th Cir.1987). Moreover, the methodology of the plaintiff's statistical expert is faulty in that it fails to differentiate between those workers who voluntarily retired, transferred, or resigned from those who were discharged against their will. See Wilkins v. Eaton Corp., 790 F.2d 515, 523 (6th Cir.1986). — 50 — Dr. Feinberg also presented evidence showing that the number of management level employees between the ages of fifty and fifty-nine dropped significantly from the year 1982 to the year 1986, even though the total number of management employees increased. Once again, however, Dr. Feinberg failed to account for voluntary retirements, resignations, or transfers among the group of employees over fifty years of age. This omission rendered the analysis essentially meaningless with respect to the issue of whether management employees suffered age discrimination through involuntary discharges or forced resignations. In Simpson, this court stated: "[W]e reject any assertion that a decreasing average employee age coupled with other questionable statistics proves the existence of impermissible age discrimination." 823 F.2d at 944 (citing Chappell, 803 F.2d at 268; Wilkins, 790 F.2d at 523). As in Simpson, the statistical evidence presented in this case was insufficient to create even an inference of age discrimination on the part of the defendant in its treatment of managerial level employees. A "mere scintilla" of evidence will not support a verdict in favor of a plaintiff in an age discrimination case. Simpson, 823 F.2d at 944. II. — 51 — It cannot be gainsaid that the trial judge played a very active role in this trial. It may well be that the demonstration setting for this trial contributed to the amount of judicial involvement. This cannot be used as an excuse, however, if the trial judge was, in fact, so intrusive as to prejudice the plaintiff with the jury. The majority has set out plaintiff's "best shot" examples of the alleged prejudicial comments made by the trial judge. Although some of the comments were critical of the expert's methodology, they appear to be accurate criticisms. There is no doubt that if a trial judge has serious doubts as to whether a proposed witness even qualifies as an expert, further inquiry should be conducted outside the presence of the jury. The problem that arises, however, in this age where the "forensic expert" populates the judicial landscape in ever increasing numbers, is that there is a plethora of experts who look good on paper and do not reveal their shortcomings until they start testifying. Although one would hope that the adversary system would be an adequate safeguard against misinformation, such is not always the case. What frequently happens is one expert is called to offset another, and a jury is left with little guidance on how to choose between the two. We instruct juries that they are free to disregard the testimony of an expert, but we give them no criteria for doing so. It is no accident that Federal Rule of Evidence 706 allows a trial court to not only appoint its own expert but to actually call that expert to testify. A trial should still be a search for the truth and, if a few "experts" become tarnished in the process, it is a cheap price to pay. In this context, one must be careful of how the word "prejudice" is interpreted. All evidence that is offered by either side is intended to be prejudicial to the other side. We exclude prejudicial evidence only when its probative value is too slight. It may be that a trial judge's questioning of a witness results in "prejudice" to one side or the other, but if the prejudice stems from the exposing of fallacies or improper methodology, so be it. — 52 — I want to make it crystal clear that I am not suggesting that plaintiff's expert here was other than an honorable witness and a person of integrity. However, one can possess these characteristics and still be wrong. Also, I am not writing a paean to trial judge activism. It is permissible to fill a void, but taking over a trial is to be avoided. On balance here, however, I conclude that plaintiff received a fair, if not perfect, trial and that is all to which he was entitled. — 53 — I would affirm the jury verdict. Footnotes [1] Although the defendant would have us believe that plaintiff loses on all of his contentions because he did not properly object at the trial court level to each of the judge's comments, such an argument is erroneous. In this instance, the plaintiff was not required to object to each improper comment or action in order to obtain appellate review. See Newman v. A.E. Staley Mfg. Co., 648 F.2d 330, 335 (5th Cir. Unit B 1981): When a timely objection [is] not made at trial, ... appellate review is not precluded. "Where a trial judge's comments were so prejudicial as to deny a party an opportunity for a fair and impartial trial, the absence of objections will not preclude our review since counsel will be loathe to challenge the propriety of a trial judge's utterances for fear of antagonizing him and thereby prejudicing a client's case." Although not precluded, appellate review is limited to review for plain error. (citations omitted and emphasis added.)