Plaintiffs, four women formerly employed as Assistant Professors at defendant Cornell University (“Cornell”), commenced this action in June, 1980, alleging inter alia they were denied professorial tenure because of their gender in violation of Title VII of the Civil Rights Act of 1964 (“Title VII” or the “Act”).1 42 U.S.C. §§ 2000e et seq. They now appeal from Chief Judge Munson’s granting of Cornell’s motion for summary judgment and entry of a final judgment under Fed.R.Civ.P. 54(b) dismissing those claims, 579 F.Supp. 349. We affirm.
BACKGROUND
The plaintiffs allege that each was the victim of discriminatory treatment on the basis of sex in Cornell’s decisions to deny them tenure and that Cornell’s tenure criteria and procedures have an illegal disparate impact within the meaning of Griggs v. Duke Power Co., 401 U.S. 424, 91 S.Ct. 849, 28 L.Ed.2d 158 (1971).
Before proceeding to more specific issues, we briefly detail the procedural sequence and structural context of tenure decisions prevailing at Cornell at the relevant times. Cornell has eleven separate academic colleges. Each college is composed of a number of departments which pursue studies in more specific areas, e.g., the Sociology Department in the College of Arts and Sciences.
The initial formal step in a tenure decision is the vote of the tenured faculty *89members of a department. A positive vote at the departmental level is automatically reviewed by the dean of the college in question. If the dean determines that the departmental decision is supported by the available documentation, an ad hoc committee of tenured faculty from the college and university will be appointed to make an independent inquiry and recommendation to the dean. The dean is also free to seek advice from any source. In the College of Arts and Sciences, the Standing Committee on Appointments and Tenure is available for such a purpose. The dean’s final recommendation is based on his or her independent judgment of the merits of the appointment and the long term interests of the university.
Where the departmental recommendation is negative, a somewhat different review process is utilized. Because the departmental faculty is generally the group most familiar both with the relevant area of study and the work of a particular candidate and because tenure necessarily results in a collegial relationship within that department which may last for decades, negative decisions at the departmental level are given substantial deference by reviewing authorities at Cornell. Until recently, no review whatsoever of a negative decision was available. Now, however, such decisions are reviewed by the college deans and a grievance procedure is available which allows higher authorities to review negative decisions to ensure that proper procedures were followed and that the negative decision was neither arbitrary nor capricious, terms which encompass sex discrimination.
Viewing the record as established through discovery in the light most favorable to plaintiffs, we turn now to more particularized matters.
(a) Dr. Zahorik
Dr. Donna Zahorik was evaluated for tenure in the Psychology Department during the academic year 1977-78. She was denied tenure while three of four male candidates considered at the same time were ultimately tenured. Although the Chairperson of the Psychology Department, Dr. Halpern, had chaired the committee of a successful male candidate and was familiar with Dr. Zahorik’s area of specialization, he declined to serve on the departmental committee charged with marshaling evidence with respect to Dr. Zahorik. Moreover, Dr. Zahorik’s committee as constituted did not act as a vigorous advocate of her candidacy, in contrast to the actions of a similar committee constituted for a second successful male colleague.
After review of the materials marshaled by the departmental committee, the Psychology Department voted four in favor, three against and five abstentions on Dr. Zahorik's candidacy. The department chairperson concluded that this unusual vote was not an endorsement of her candidacy.
Dr. Zahorik appealed the departmental decision to the Acting Dean of the College of Arts and Sciences, who, in light of the unusual vote, referred the issue back to the department for reconsideration. Subsequently, after the submission of additional materials, the department again voted to deny her tenure. This vote was six against, five in favor and one abstention. Dr. Zahorik appealed this second vote, and Harry Levin, then Dean of the College of Arts and Sciences, denied her appeal as did the Provost.
The departmental file concerning Dr. Zahorik contains highly favorable evaluations which praise her as an outstanding and productive scholar, more tempered assessments which rate her work highly but question her productivity, and highly critical statements that she lacks creativity and that graduate students were unwilling to conduct research under her supervision. There is no evidence that the judgment of any of the negative commentators was influenced by the fact that she is a woman other than a statement by an abstaining member of the department that Dr. Zahorik’s problem in attracting graduate students was that she was too “feminine” in that she was too “unassuming, unaggres*90sive, unassertive and not highly motivated for vigorous interpersonal competition.”
(b) Dr. Laws
Dr. Judith Laws was the first woman to be hired in a tenure-track position by the Sociology Department at Cornell. She was reviewed for tenure during 1976. The department voted against her promotion by a six to four vote. Dr. Laws appealed to Dean Levin of the College of Arts and Sciences. Confining his review to the procedure followed, Dean Levin recommended departmental reconsideration after the solicitation of additional material on Dr. Laws, including opinions from reviewers outside Cornell. The additional materials were obtained, and Dr. Laws also elected to present a colloquium on her empirical research. This colloquium, according to one of her supporters, went very badly. The second tenure vote, held in December 1976, was also unsuccessful, and Dr. Laws again took an appeal. As in the case of Dr. Zahorik, this second appeal was denied by the Dean and ultimately by the Provost.
Dr. Laws’ file also contains strongly supportive evaluations as well as highly critical ones. The negative views on her work were more widely and strongly held than in Dr. Zahorik’s case. The Chairperson of the Department' thus wrote Dean Levin: “[Doctor Laws’] teaching record is weaker now than when she began at Cornell; and her research has declined to zero while her writing is shifting toward the polemic and away from the major forms of theoretical work.”
(c) Dr. Glasse
Dr. Antonia Glasse was evaluated for tenure in the Department of Russian Literature in the 1975-76 academic year. In December, 1975, the Department voted two to one to recommend tenure. The two members supporting Dr. Glasse cited her perfect command of Russian, her access to Soviet literary circles and her scholarly approach. One of the two, however, expressed concern over her teaching and described her published work as “relatively little.” The negative voter, the chairperson and most prestigious member of the department, questioned the significance of Dr. Glasse’s scholarly work and particularly questioned whether it would be wise to grant a fourth tenure position in a department of only five members, thus blocking the upward path of more promising scholars.
Dean Levin appointed an ad hoc committee from the College of Arts and Sciences to review the departmental recommendation. This committee endorsed Dr. Glasse for tenure, although it concluded that “our recommendation is ... on the whole measured rather than enthusiastic.” An outside reviewer consulted by the ad hoc committee also supported Dr. Glasse for tenure. Three Cornell Slavic language professors, not members of the Department of Russian Literature, provided Dean Levin with letters regarding Dr. Glasse’s candidacy; two strongly opposed it and one strongly supported it. Dean Levin also sought counsel from the College’s Standing Advisory Committee on Promotions. It unanimously recommended that she should not be granted tenure.
Dean Levin concluded that Dr. Glasse’s candidacy should be denied. He explained in a letter to her that he could not grant tenure on recommendations based principally on the argument that no one better was presently available. Dr. Glasse asserts that a male scholar of markedly inferior credentials was granted tenure in the Department of Russian Literature one year before Dean Levin’s denial of her candidacy.
(d) Dr. Farris
Dr. Charlotte Farris was appointed to the Department of Community Service Education (“CSE”) in Cornell’s College of Human Ecology in January, 1974 and was reviewed for tenure during the academic year 1978-79. The CSE Personnel Committee formed a Department Review Committee of tenured faculty both to marshal the evidence and to make a recommendation to the tenured faculty. This committee solicited *91opinions from a variety of sources and reviewed Dr. Farris’ career at Cornell. It concluded that her greatest strengths and greatest interests lay in public service and that her teaching time was below that originally expected of her. The committee recommended that Dr. Farris not be promoted unless the department intended to increase its role in the area of public service. Absent such an intention the committee believed her “limited publications,” “lack of investment in curriculum and program,” and “mixed evaluations” in research and teaching weighed against tenure. The CSE faculty then voted unanimously to deny her tenure. As with Dr. Laws and Dr. Zahorik, a second tenure vote was taken after Dr. Farris had been given an opportunity to supplement her file. Nine faculty members voted against her candidacy and one voted for it.
(e) Statistical Evidence of Discrimination
Plaintiffs offered certain statistical data indicating that during the four-year period in which their tenure decisions were made, 65.3% of the male candidates in three of Cornell’s colleges (Human Ecology, Arts and Sciences, Architecture) were granted tenure (47 of 72 candidates), while only 42.1% of the female candidates were similarly favored (8 of 19 candidates). There is no breakdown among the various departments involved.
The data from which these particular statistics were culled included tenure decisions in the three colleges from July 1, 1975 to June 30, 1981. Plaintiffs chose not to use that data, however, and requested Dr. Zahorik to separate out those decisions made between July 1, 1975 and February 1, 1979. Although this three and one-half year period corresponds to the time during which the plaintiffs were considered for tenure, the elimination of the period after February 1, 1979 required “estimates” by Dr. Zahorik as to when particular decisions were made and excluded almost 50 tenure decisions, thus reducing the sample considerably.
(f) The Proceedings in the District Court
Plaintiffs filed their complaint in June, 1980. Class certification was denied on January 21, 1982. On January 27, 1983, Cornell moved for summary judgment and filed a 44 page statement of material facts which it contended were not in dispute. Plaintiffs did not file a responding statement of facts allegedly in dispute, as required by N.D.N.Y.R. 10(c). Nor did they set forth their statistical evidence in their papers answering the motion for summary judgment or present their legal theory of disparate impact. As a consequence, Chief Judge Munson might properly have assumed that plaintiffs admitted the defendants’ entire statement of material facts and ignored the statistical evidence. However, the statistical data was resurrected at oral argument before him from the class certification papers and he considered it in his opinion.
Plaintiffs also complain of inadequate opportunity for discovery. The lack of discovery weighs heavily in favor of the party opposing summary judgment, Landmark Land Co. v. Sprague, 701 F.2d 1065, 1070 (2d Cir.1983); Quinn v. Syracuse Model Neighborhood Corp., 613 F.2d 438, 445 (2d Cir.1980). However, the plaintiffs here had some two and one-half years for discovery before the motion for summary judgment was filed and they explicitly assured the district court that sufficient discovery had been conducted. In those circumstances, the claim of lack of discovery cannot be entertained on appeal.
DISCUSSION
(a) Discriminatory Treatment
Discriminatory treatment is established under Title VII by proof that plaintiffs were treated less favorably than others solely because of their race, color, religion, sex or national origin. “Proof of discriminatory motive is critical,” International Brotherhood of Teamsters v. United States, 431 U.S. 324, 335 n. 15, 97 S.Ct. 1843, 1854 n. 15, 52 L.Ed.2d 396 (1977), *92although that motive can be established by circumstantial as well as direct evidence, United States Postal Service Board of Governors v. Aikens, 460 U.S. 711, 103 S.Ct. 1478, 1481 n. 3, 75 L.Ed.2d 403 (1983). Normally, discriminatory treatment is proven pursuant to the three-part test established in McDonnell Douglas Corp. v. Green, 411 U.S. 792, 802-04, 93 S.Ct. 1817, 1824-25, 36 L.Ed.2d 668 (1973). First, the plaintiff has the burden of proving by a preponderance of the evidence a prima facie case of discrimination. This may be accomplished by evidence that the plaintiff was a member of the protected group, was qualified for tenure, and was not granted tenure in circumstances permitting an inference of discrimination. Second, if the plaintiff has established a prima facie ease, the burden of producing evidence shifts to the defendant “to articulate some legitimate, nondiscriminatory reason for the employee’s rejection.” Id. at 802, 93 S.Ct. at 1824. Third, should the defendant carry this burden, the plaintiff may offer evidence that the defendant’s ostensibly legitimate reasons were not genuinely held but were merely a pretext for discrimination. While the McDonnell Douglas test does shift the burden of production once a plaintiff has offered a prima facie case, the burden of persuasion remains upon the plaintiff at all times. Texas Department of Community Affairs v. Burdine, 450 U.S. 248, 101 S.Ct. 1089, 67 L.Ed.2d 207 (1981).
We have previously noted that tenure decisions in an academic setting involve a combination of factors which tend to set them apart from employment decisions generally. Lieberman v. Gant, 630 F.2d 60, 64 (2d Cir.1980). First, tenure contracts entail commitments both as to length of time and collegial relationships which are unusual. Lifetime personal service contracts are uncommon outside the protected civil service but even there difficulties in collegial or professional relationships can be eased by transfers among departments. Professors of English, however, remain in that department for life and cannot be transferred to the History Department.
Second, academic tenure decisions are often non-competitive. Whereas in other employment settings a decision not to hire one person is usually the flip side of a decision to hire another, a decision to grant or not grant tenure to a particular person does not necessarily affect the future of other tenure candidates. In some cases, of course, the number of tenure slots is fixed and an affirmative tenure decision necessarily excludes other candidates. Even in such cases, however, the effect on those excluded is uncertain since the immediate alternate candidates have no assurance that they would have received tenure had a slot been available. For the same reason, a denial of tenure to one person does not necessarily lead to tenure for another. The number of tenure slots available may be flexible and, even where fixed, there may be no pressing need to fill vacancies since teaching chores can be discharged by nontenured faculty.
Third, university tenure decisions are usually highly decentralized. The decision at the departmental level is of enormous importance both because of the department’s stake in the matter and its superior familiarity with the field and with the candidate. Authority to overrule departmental decisions may exist, particularly in the case of affirmative decisions since the downside risk of affirmative decisions is greater than that of negative ones, but the deference given to departmental decisions grows as a case travels up the chain of authority.
Fourth, the number of factors considered in tenure decisions is quite extensive. The particular needs of the department for specialties, the number of tenure positions available, and the desired mix of well known scholars and up-and-coming faculty all must be taken into account. The individual’s capacities are obviously critical. His or her teaching skills, intelligence, imagination, willingness to work, goals as a scholar and scholarly writing must be evaluated by departmental peers and outsiders *93asked to render advice. The evaluation does not take place in a vacuum, however, but often in the context of generations of scholarly work in the same area and always against a background of current scholarship and current reputation of others. Moreover, universities and departments within them occupy different positions in the academic pecking order and the standard of “excellence” may vary widely according to the ability to attract faculty.
Fifth, tenure decisions are a source of unusually great disagreement. Because the stakes are high, the number of relevant variables is great and there is no common unit of measure by which to judge scholarship, the dispersion of strongly held views is greater in the case of tenure decisions than with employment decisions generally. As the present record amply demonstrates, arguments pro and con are framed in largely conclusory terms which lend themselves to exaggeration, particularly since the stauncher advocates on each side may anticipate and match an expected escalation of rhetoric by their opponents. Moreover, disagreements as to individuals may reflect long standing and heated disputes as to the merits of contending schools of thought or as to the needs of a particular department. The dispersion of views occurs within departments themselves but is accentuated by the solicitation of opinion from students, faculty from other departments and faculty from other universities. Where a broad spectrum of views is sought and the candidate suggests certain persons as refer-rants, a file composed of irreconcilable evaluations is not unusual.
*/^The context and nature of tenure decisions rarely benefit Title VII plaintiffs seeking to prove that a particular tenure decision was influenced by sex or race. No tenure candidate is without blemishes and a resort to illegitimate considerations can be hidden in the weighing of the numerous factors which are relevant to a tenure decision. Because of the decentralized nature of the decision-making process, comparisons which might tend to show unlawful discrimination are hard to come by. A denial of tenure by an English department simply cannot be compared with a grant of tenure in the physics or history departments. Even within a single department comparisons are difficult because the number of decisions within a particular period may be quite few, the decisions sometimes may be non-competitive and tenure files typically contain positive as well as negative evaluations, often in extravagant terms, sufficient to support either a grant or denial of tenure.
Courts, moreover, are understandably reluctant to review the merits of a tenure decision. Lieberman v. Gant, 630 F.2d at 66. Where the tenure file contains the conflicting views of specialized scholars, triers of fact cannot hope to master the academic field sufficiently to review the merits of such views and resolve the differences of scholarly opinion. Moreover, the level of achievement required for tenure will vary between universities and between departments within universities. Determination of the required level in a particular case is not a task for which judicial tribunals seem aptly suited. Finally, statements of peer judgments, as to departmental needs, collegial relationships and individual merit may not be disregarded absent evidence that they are a facade for discrimination.
Tenure decisions are not exempt under Title VII, however, and plaintiffs seeking to show that forbidden purposes lurk in a tenure decision have available methods of challenging such decisions. Departures from procedural regularity, such as a failure to collect all available evidence, can raise a question as to the good faith of the process where the departure may reasonably affect the decision. Conventional evidence of bias on the part of individuals involved may also be available. Finally, the McDonnell Douglas three-part test described above requires, a response by the university once a prima facie case has been established. Given the elusive nature of tenure decisions, we believe that a prima facie case that a member of a protected class is qualified for tenure *94is made out by a showing that some significant portion of the departmental faculty, referrants or other scholars in the particular field hold a favorable view on the question. If the other element of the first step of the McDonnell Douglas test is present, namely circumstances permitting an inference of discrimination, the burden of production is then shifted to the university, and its response may be attacked both on its logic and the substantiality of the underlying file.
However, for a plaintiff to succeed in carrying the burden of persuasion, the evidence as a whole must show more than a denial of tenure in the context of disagreement about the scholarly merits of the candidate’s academic work, the candidate’s teaching abilities or the academic needs of the department or university. Absent evidence sufficient to support a finding that such disagreements or doubts are influenced by forbidden considerations such as sex or race, universities are free to establish departmental priorities, to set their own required levels of academic potential and achievement and to act upon the good faith judgments of their departmental faculties or reviewing authorities.
We conclude that the evidence in the record does not raise a disputed issue of material fact as to differential treatment because of sex as to any of the plaintiffs. Summary judgment was thus properly granted on this claim.
Dr. Zahorik alleges procedural irregularities, citing Dr. Halpern’s refusal to serve on the committee marshaling the file for the departmental vote and the failure of her committee to support her candidacy vigorously. We fail to see how Dr. Halpern’s abstention from such service was prejudicial to Dr. Zahorik in view of his tepid attitude toward her candidacy. Nor are we prepared to hold that committees marshaling the file on a candidate either must advocate promotion to tenure whatever the true views of those who serve upon them or that their failure to do so'is evidence of forbidden discrimination. Dr. Zahorik’s conclusory assertion that two men granted tenure are less well qualified adds nothing to her claim since the record at best indicates a difference of opinion in evaluation of scholarly merit. Dr. Zahorik also emphasizes the letter in her file describing her as too “feminine.” The author, however, was discussing not her gender but rather the effect of her personality on graduate students seeking strong minded committee chairpersons.
Dr. Laws concedes that the second vote in her case remedied any procedural irregularities in the initial departmental consideration of her candidacy. On the merits, her file demonstrates that some of her peers had considerable doubt about her future as a scholar. Like Dr. Zahorik, she claims that a less qualified male received tenure in the department some years before she did but also fails to demonstrate more than a disagreement as to scholarly merits.
In the cases of Drs. Zahorik and Laws, the demonstration of significant support for their candidacies by scholars in their fields and evidence of the relatively simultaneous promotion of males was sufficient to shift the burden of production. However, Cornell has produced evidence of reasoned doubt as to the merits of their candidacies and the record is barren of any evidence that gender influenced the decisions. Such evidence of pretext is essential since Drs. Zahorik and Laws have the burden of persuasion on the whole case. Since neither Dr. Zahorik nor Dr. Laws could prevail at trial, summary judgment was properly granted.
As to Dr. Farris, there is also no evidence of discrimination. Reconsideration of the original negative departmental judgment was denied by a nine to one vote on the grounds that she had not fulfilled academic expectations in the way of teaching, had not produced material of high quality, had not organized her material well and had limited publications. It is not clear that Dr. Farris established a prima facie case as to her qualifications for tenure. However, there is also no evidence that her gender influenced the decision. In such *95circumstances, summary judgment was properly granted as to her tenure claim,
Dr. Glasse’s candidacy was somewhat stronger than those of the other plaintiffs since both her department and an ad hoc committee recommended tenure. Nevertheless, there was again evidence of reasoned doubt. Her recommendations were lukewarm, and she was not without substantial opposition. The weakness of her candidacy might reasonably alarm a dean responsible for long-range planning. Since there is no evidence that the negative opinions concerning Dr. Glasse were held in bad faith, a trier of fact could not find a violation of Title VII unless it inferred discrimination solely from Dean Levin’s rejection of a departmental recommendation. However, there is also no evidence that his decision was based on gender. Indeed, he had ordered departmental reconsideration in the case of two of the other plaintiffs. Universities are free to vest authority in designated officials to override departmental decisions, and the exercise of that authority is not itself evidence of discrimination based on sex.
Plaintiffs’ proffer of statistical data adds nothing to their claim of discriminatory treatment. The weakness of the data is self-evident. It has been deliberately culled from a broader sample and includes “estimates” as to when particular decisions were made. Moreover, tenure decisions at Cornell are not made by a single authority but are highly decentralized. Gross statistics are thus meaningless absent a departmental breakdown. Finally, even if statistics reliably demonstrating that 65% of male candidates during an appropriate period were granted tenure as against only 42% of female candidates was sufficient by itself to permit a finding of discrimination against some women at Cornell — a proposition which the caselaw does not support — such data would not be sufficient to prove that the particular plaintiffs were among the women who were not treated neutrally. More particularized evidence relating to the individual plaintiffs is necessary to show discriminatory treatment.
(b) Disparate Impact
Violations of Title VII may occur when certain facially neutral selection criteria are found to cause a disparate impact on a class protected under Title VII. International Brotherhood of Teamsters v. United States, 431 U.S. at 335 n. 15, 97 S.Ct. at 1854 n. 15. The disparate impact theory has been used mainly in the context of quantifiable or objectively verifiable selection criteria which are mechanically applied and have consequences roughly equivalent to results obtaining under systematic discrimination. Griggs v. Duke Power Co., 401 U.S. 424, 91 S.Ct. 849, 28 L.Ed.2d 158 (1971) (aptitude and intelligence tests and educational requirements); Dothard v. Rawlinson, 433 U.S. 321, 97 S.Ct. 2720, 53 L.Ed.2d 786 (1977) (height and weight requirements); Grant v. Bethlehem Steel, 635 F.2d 1007, 1018 (2d Cir.1980), cert. denied, 452 U.S. 940, 101 S.Ct. 3083, 69 L.Ed.2d 954 (1981) (requirement of previous experience as a foreman with the employer). Plaintiffs who allege a forbidden disparate impact are required to prove a causal connection between the challenged selection criterion and the disparate impact itself, Pouncy v. Prudential Insurance Co. of America, 668 F.2d 795, 800-01 (5th Cir.1982), while employers can defend use of the challenged selection criterion on grounds of “a legitimate business reason,” id. at 800. Job-relatedness is thus a defense to Title VII liability, New York City Transit Authority v. Beazer, 440 U.S. 568, 587, 99 S.Ct. 1355, 1366, 59 L.Ed.2d 587 (1979).
Plaintiffs claim that the process by which tenure decisions are made at Cornell has an illegal disparate impact because it is highly subjective and leads to bias against women. We reject the claim because it is not supported by evidence that Cornell’s selection criteria have resulted in the requisite discriminatory impact and because the criteria and procedures utilized by Cornell *96are legitimately related to the position of tenured professor.
Where the challenged selection criteria are facially neutral and there is no evidence of a discriminatory intent, courts have insisted upon clear and convincing evidence of a substantially discriminatory effect before applying the disparate impact theory. See, e.g., International Brotherhood of Teamsters v. United States, 431 U.S. at 342 n. 23, 97 S.Ct. at 1858 n. 23; Griggs v. Duke Power Co., 401 U.S. at 430 n. 6, 91 S.Ct. at 853 n. 6; Grant v. Bethlehem Steel Corp., 635 F.2d at 1010. Plaintiffs’ proof in the instant case falls far short of the showing necessary to undermine facially neutral criteria. Even after the sample was diminished, the data offered shows that a relatively substantial proportion of women candidates (42%), achieve tenure at Cornell. It also shows that of the decisions taken into account men had a higher promotion rate (65%), but Title VII does not forbid all employment decisions which result in departures from an ideal statistical norm. Evidence of systematic exclusion by the mechanical application of facially neutral criteria is necessary.
Additionally, Cornell’s selection criteria, however difficult to apply and however much disagreement they generate in particular cases, are job related. Accomplishments and skills in scholarship and teaching are obviously relevant to employment in tenured professorships. A decentralized decision-making structure founded largely on peer judgment is based on generations of almost universal tradition stemming from considerations as to the stake of an academic department in such decisions and its superior knowledge of the academic field and the work of the individual candidate. It would be a most radical interpretation of Title VII for a court to enjoin use of an historically settled process and plainly relevant criteria largely because they lead to decisions which are difficult for a court to review.
Affirmed.