Jackson v. NASSAU CTY. CIVIL SERV. COM'N, 424 F. Supp. 1162 (E.D.N.Y. 1976)

*1163 *1164 Harvey W. Spizz, Spizz & McEvily, Community Legal Assistance Corp., Hempstead, N. Y., for plaintiffs.

A. Seth Greenwald, Asst. Atty. Gen., State of New York, New York City, James N. Gallagher, Deputy County Atty., of Nassau County, Mineola, N. Y., for defendants.

MISHLER, Chief Judge.

In June of 1973, the New York State Civil Service Commission gave an examination for the newly created permanent position of community service assistant, a paraprofessional berth in human services agencies throughout the state. Among the 294 candidates taking the examination in Nassau County were ten employees of the Nassau County Office of Consumer Affairs and the Nassau County Department of Social Services who served these agencies as provisional community service assistants. Although eight of the provisionals received passing grades on the exam, none of their scores were high enough to qualify them for appointment to the permanent positions. Consequently, the ten employees, eight of whom are black, either lost their jobs or were demoted to the lower paying but noncompetitive position of community service aide. The present litigation ensued.

*1165 The black plaintiffs, former provisional community service assistants, claim that the defendants, who include the New York State Civil Service Commission and the Nassau County Civil Service Commission,^[1] violated the provisions of Title VII of the Civil Rights Act of 1964, 42 U.S.C. § 2000e, as amended,^[2] and the federal civil rights laws, 42 U.S.C. §§ 1981 and 1983. Jurisdiction is predicated on 28 U.S.C. §§ 1331 and 1343(3) and (4) and 42 U.S.C. § 2000e-5(f). The white plaintiffs limit their claim to violations of the Due Process Clause of the fourteenth amendment and 42 U.S.C. § 1983. In an earlier proceeding, Judge Weinstein preliminarily enjoined the defendants from altering the salary status of those plaintiffs employed in the Office of Consumer Affairs or discharging the plaintiff Gibrilla. The case was then tried before this court without a jury. Although the facts are largely undisputed, some background would be helpful.

The position of community service assistant was created in 1968 by the federal government's Public Service Careers Program to provide lower-level social service jobs for welfare recipients and low income persons. Although the position was originally non-competitive, the eventual goal of the federal program was to create a competitive position that would blend into the state civil service system, providing a route for low income persons to the more desirable civil service jobs. (Defendants' Exhibit E; Mar. 2, 1976, tr. 172-73). During this period of its development, community assistant positions were established in Nassau County human service agencies,^[3] including consumer affairs and social services (welfare).

The Department of Consumer Affairs of the County of Nassau services the entire population of Nassau County, but places special emphasis on assisting the black communities by, for example, establishing educational programs for black consumers. Plaintiff Gibrilla, a community service assistant for the Office of Consumer Affairs, to cite one example, handles individual consumer complaints and lectures to various civic groups on consumer problems (Mar. 1, tr. 50-51). The Nassau County Department of Social Services provides such assistance to welfare recipients as organizing self-help drug addiction groups (Mar. 1, tr. 150-51) and furnishing advice on housing matters. In her capacity as community service assistant, plaintiff Melendez, for example, assisted welfare families in finding housing and was responsible for checking the bills received by the Department of Social Services when clients were placed temporarily in motels (Mar. 1, tr. 108-09).

Despite the broad range of activities performed by the plaintiffs, the duties of the various community service assistants require essentially the same basic skills.^[4] According *1166 to the job description of this position, a community service assistant must have the ability "to speak effectively"; "present oral and written reports"; "work effectively with others"; "to follow instructions"; and "to relate to the clientele and staff."^[5] (Plaintiffs' Exhibit 7). The job pays from $7,237.00 to $9,093.00, as compared with $6,498.00 to $8,107.00 for the position of community service aide. A candidate for community service assistant must pass the civil service examination in order to be placed on the eligible list. Appointment from the list to a permanent position is limited by the "rule of three", N.Y.Civ. Serv.L. § 61(1), to "one of the three persons certified by the appropriate civil service commission as standing highest on [the] eligible list."

The examination was given in Nassau County on June 2, 1973. Consisting of 60 questions divided into three sections, the test was designed as a "general selection battery for paraprofessionals in the human services area." (Plaintiffs' Exhibit 2). Fifteen questions involved the problems and demands of interviewing clients; fifteen questions were devoted to record keeping; and thirty questions tested ability to understand and appropriately respond to people with various personal problems in a variety of situations.

It was the first time a competitive exam had been given for the position of community service assistant. At the time the examination was given, the economy of metropolitan New York, including Nassau County, had so deteriorated that highly skilled and educated individuals were forced to look for work in areas previously left to members of low income groups.^[6] As a result, a large number of young whites with substantial educational backgrounds took the examination for community service assistant (Plaintiffs' Exhibit 1).^[7] The plaintiffs, whose work as provisional employees was highly praised by their supervisors, were unable to achieve scores sufficient to place them at the top of the eligibility list for appointment. Much of the plaintiffs' attack on the examination concerns its failure to test for the different skills required in the consumer and social service departments and to measure areas of knowledge specific to the duties performed by a community service assistant.

Because of the somewhat disparate nature of the claims of the black plaintiffs and those of the whites, we will deal with the two groups of plaintiffs separately. We turn first to the question of whether the examination violates Title VII of the 1964 Civil Rights Act, the basis for the black plaintiffs' claims.

Title VII of the 1964 Civil Rights Act was enacted to "achieve equality of employment opportunities and remove barriers that have operated in the past to favor an identifiable group of white employees over other employees." Griggs v. Duke Power Co., 401 U.S. 424, 429-30, 91 S. Ct. 849, 853, 28 L. Ed. 2d 158 (1971). The legislation created a significantly more rigorous standard for judging claimed discriminatory acts of private *1167 or public employers than exists with respect to purely constitutional claims of racial discrimination. Washington v. Davis, 426 U.S. 229, 247-48, 96 S. Ct. 2040, 2051, 48 L. Ed. 2d 597 (1976). In a challenge under Title VII to an employment exam, the complaining party has the burden of establishing a prima facie case of discrimination, usually defined as a showing that the test at issue has a racially disproportionate impact on applicants it "select[s] applicants for hire or promotion in a racial pattern significantly different from that of the pool of applicants." Albemarle Paper Co. v. Moody, 422 U.S. 405, 425, 95 S. Ct. 2362, 2375, 45 L. Ed. 2d 280 (1975); see Washington v. Davis, supra, 426 U.S. at 238-39, 96 S. Ct. at 2047. Once this showing is made, however, the burden shifts to the employer to prove that its tests are job-related, i. e., the entrance or promotion examination bears a reasonable relationship to the needs of the job for which it is given. Griggs v. Duke Power Co., supra, 401 U.S. at 431, 91 S. Ct. at 853; McDonnell Douglas Corp. v. Green, 411 U.S. 792, 802, 93 S. Ct. 1817, 1824, 36 L. Ed. 2d 668 (1973); Boston Chapter N.A.A.C.P., Inc. v. Beecher, 504 F.2d 1017, 1019 (1st Cir. 1974). Finally, if the employer does establish that his test is job-related, the complaining party still may prevail if he can show that "other tests or selection devices, without a similarly undesirable racial effect, would also serve the employer's interest in `efficient and trustworthy workmanship.'" Albemarle Paper Co. v. Moody, supra, 422 U.S. at 425, 95 S. Ct. at 2375, quoting McDonnell Douglas Corp., supra, 411 U.S. at 801, 93 S. Ct. at 1823.

In Nassau County, 286 candidates for the position of community service assistant took examination 64-914.^[8] Of the 175 candidates who identified themselves ethnically, 113 were white; 55 were black; 3 were Puerto Rican; and 4 fit other ethnic categories. (Plaintiffs' Exhibit 3). The passing grade was set at a score of 36 correct answers out of the exam's 60 questions, a percentage of 60%. Ninety-nine of the 113 identified whites attained this score or better for a passing rate of 88%. Of the 55 identified blacks who competed, 40 passed the exam, a pass rate of 73%. In percentage terms, the passing rate for identified blacks was 83% of the passing rate for identified whites, or put in other statistical terms, white candidates passed at a rate 1.2 times that of black applicants.^[9]

Although the black applicants did not do as well as the whites, we do not regard as discriminatory an exam that passed blacks at a rate better than 4/5 that of the passing rate for whites. A comparison with prior findings of discrimination in employment tests is revealing. In Vulcan Society of N. Y. City Fire Dept., Inc. v. Civil Serv. Comm'n., 490 F.2d 387 (2d Cir. 1973), a qualifying exam for the New York City Fire Department was held to have a discriminatory impact since white candidates passed the exam at a rate more than twice that of the minority candidates. Id. at 392. In Chance v. Board of Examiners, 458 F.2d 1167 (2d Cir. 1972), public school supervisory examinations, which white candidates passed at a rate 1½ times that of minority candidates, were found to have a discriminatory impact. Id. at 1173. See Kirkland v. New York State Dept. of Correctional Serv., 374 F. Supp. 1361, 1366 (S.D. N.Y.1974), modified, 520 F.2d 420 (2d Cir. 1975) (whites passed promotional exam for position of Correction Sergeant at a rate four times that of blacks and 2½ times that of hispanics); Bridgeport Guardians, Inc. v. Members of Bridgeport Civil Serv. Comm., 482 F.2d 1333, 1335 (2d Cir. 1973) (passing *1168 rate for white applicants for police department 3½ times that of minority candidates); Castro v. Beecher, 459 F.2d 725, 729 (1st Cir. 1972) (25% of black applicants and 10% of hispanic applicants passed police exam, while 65% of non-minority candidates achieved a passing score). See also Carter v. Gallagher, 452 F.2d 315 (8th Cir. 1971), modified en banc, 452 F.2d 327 (8th Cir.), cert. denied, 406 U.S. 950, 92 S. Ct. 2045, 32 L. Ed. 2d 338 (1972). Furthermore, the present case involves a relatively small number of applicants. Compare Vulcan Society of N. Y. City Fire Dept. v. Civil Serv. Comm'n., supra. The passing percentages would be substantially altered by a shift of several minority candidates from the fail to the pass column. In such a situation, some margin of error must be allowed. Plaintiffs bear a heavier burden in demonstrating a disproportionate impact when the number of test-takers is small and, in our opinion, a showing of a 4/5 passing rate does not satisfy that burden.^[10]

The plaintiffs also assert that discriminatory impact is evidenced by the fact that only one black candidate, or 2% of the identified black applicants, achieved a score sufficient to place him or her high enough on the eligibility list to be offered an appointment, while 7 whites, or 6% of the identified white applicants, achieved a score sufficiently high for appointment. Apparently, 17 candidates, identified and unidentified ethnically, all of whom had at least a raw score of 54, were offered the permanent position of community services assistant. Two black candidates, however, achieved raw scores of 53. (Plaintiffs' Exhibit 3). If each of these candidates had answered one more question correctly, thereby achieving a score of 54, then the comparative rates at the top of the eligibility list, i. e., 54 or above, would be 5.5% of the black applicants and 6% of the whites. See Bridgeport Guardians, Inc. v. Members of the Bridgeport Civil Serv., supra. Compare Vulcan Society of N. Y. City Fire Dept., Inc. v. Civil Serv. Comm'n., supra at 392. The failure of two black candidates to answer two questions correctly does not establish that the exam had a disproportionate impact with respect to those candidates who achieved a raw score of at least 54.^[11]

If arguendo examination 64-914 did have a discriminatory impact, defendants nonetheless have sustained their burden of showing that the examination was substantially related to job performance. As is common in Title VII litigation, this case led the court "deep into the jargon of psychological testing." Vulcan Society, supra at 394. The parties' expert witnesses engaged in the usual jousting over "content validation"^[12] versus "predictive and concurrent validation."^[13] However, after five years of Title VII litigation since the decision in Griggs, we think it is clear that "there is no single method for appropriately validating employment tests for their relationship to job performance." Washington v. Davis, supra, 426 U.S. at 247 n. 13, 96 S. Ct. at 2051 n. 13; Vulcan Society, supra at 394; Guidelines on Employment Selection Procedures ("Guidelines"), Equal Employment Opportunity Commission, 29 C.F.R. § 1607.5 (1975).^[14] Indeed, as Judge Lasker observed in his opinion in Kirkland *1170 v. New York State Dept. of Correctional Serv., 374 F.Supp. at 1371:

[N]o case in this Circuit has gone so far as to hold that failure to test an exam by criterion validation or to demonstrate the nonfeasibility of that approach justifies setting the exam aside even if it has been content validated.

But see Guidelines, § 1607.4(b). For the sake of completeness, however, we turn to a consideration of the feasibility of criterion validation of examination 64-914.

Predictive validation, which requires a showing that a candidate's test performance correlates with his job performance, is perhaps the most desirable method for determining the job-relatedness of a selection procedure. See Kirkland, 374 F. Supp. at 1370. However, predictive validation is not feasible in this case since, as plaintiffs' own expert acknowledged, an examination given for the first time cannot be validated for the group that was initially tested. (Mar. 2, tr. 125).^[15] Moreover, there was expert testimony that predictive validation is not appropriate where there are few job performances to measure against the test scores. (Mar. 3, tr. 333-34). In the present case, only seven community service assistant positions exist in Nassau County, an insufficient number with which to validate the examination.^[16]

Concurrent validation, which involves a comparison between the job performance of current employees and their relative scores on the examination, see Vulcan Society, supra at 394, is not possible where the examination is given for newly created positions, i. e., the job of permanent community service assistant. To the extent that the duties of a provisional community assistant are comparable to those of a permanent *1171 one, there is evidence that the examination has some concurrent validity. Eighty percent of the provisionals passed the examination and some scored grades within a few points of the highest grades achieved by any candidate. Although the plaintiffs' performance is by no means conclusive on the issue of concurrent validity, the relatively good showing of the plaintiffs, together with the fact that none of the parties introduced evidence that concurrent validation would be feasible here, suggest that no basis exists for assuming that the exam lacks concurrent validity and, further, that other methods of validation would be just as suitable.^[17]

Thus, content validation provides both a feasible and appropriate method of determining whether examination 64-914 is substantially related to the job requirements of the community service assistant. To prove content validation, the defendants must demonstrate that

they have formulated examination questions and procedures based on an analysis of the job's requirements, usually determined through empirical studies conducted by experts. An examination has content validity, then, if it elicits from the candidate information that is relevant to the job for which it is given.

Chance v. Board of Examiners, supra at 1174 (internal quotation marks omitted). Inquiry must be made both as to the method of preparation and the actual content of the examination. The poorer the quality of preparation, "the greater must be the showing that the examination was properly jobrelated, and vice versa." Vulcan Society, supra at 396.

The court's task in this case was easier than usual since there was relative agreement by the expert witnesses as to the content validity of the examination. Even the plaintiffs' expert witness conceded that there was a "reasonable effort to assume content validity. Going that far, I would have gone about it much the same way." (Mar. 2, tr. 99). The defendants' expert testified that "the procedures that were followed in the content strategy were those that are generally recognized by my profession." (Mar. 3, tr. 332). Moreover, extensive testimony was adduced by the defendants as to the development of the examination. Without repeating this entire line of testimony, it is clear that the examination at issue is the culmination of an extensive analysis by professionals of the concept of a paraprofessional in the human services, such as a community service assistant. The position of community service assistant itself was defined in terms of duties, required knowledge, skills and abilities, and needed training and experience. The materials testing interviewing and ability to work with people were developed by professional examiners with knowledge of the duties of paraprofessionals. The section on record keeping was based on sample records and forms submitted by a local social services agency.^[18]

*1172 In terms of actual content, the questions asked on the test were designed to confront the applicant with the type of situations that arise in the normal course of a community service assistant's duties. In addition to the interviewing and record keeping materials, candidates were tested on the best method of, inter alia, dealing with a client who refuses to keep appointments; obtaining information from a client who insists that he is being asked "personal" questions; handling arguments between clients; dealing with difficult and disruptive co-workers; and approaching clients who have emotional problems. (Plaintiffs' Exhibit 6). In sum, this was a professionally constructed examination with content validity.

The plaintiffs, however, contend that the examination is not job-related because the questions are not based on the specific activities of a community service assistant, which vary from the consumer agency to the social services agency. This argument misses the mark in several respects.

First, aside from the problem of the increased administrative burden of agency-by-agency testing, within each agency alone the duties of a community service assistant vary significantly. A paraprofessional in the welfare agency, for example, may be asked to perform a variety of tasks. Do the plaintiffs employed by the Nassau County Department of Social Services wish to be tested on how to organize a co-op supermarket in Hempstead? Or, should the exam concern the relocation of welfare families from motels to more permanent housing? Some of the consumer employees might wish to be tested on how to conduct an investigation of consumer fraud, while others might prefer an exam that tests ability to handle individual complaints by phone and make proper referrals. The point is that human services paraprofessionals may be asked to perform such a great variety of functions that the only feasible civil service examination for that position is one that tests, as examination 64-914 does, the basic skills underlying human services work: communication skills; the ability to work with people; the ability to respond appropriately to other people's emotional difficulties; and interviewing and record keeping skills.

Second, in making this argument the plaintiffs are asking, in effect, to be tested on knowledge they acquired in the course of their jobs as provisional community service assistants. Aside from the fact that such knowledge is more appropriately tested on a promotional exam (see testimony of Michael Dollard, Mar. 2-3, tr. 215, 320) there is a fundamental objection to the plaintiffs' contention. The purpose underlying the creation of the position of community service assistant was to provide eventual access *1173 to the civil service system for low income individuals. By demanding an examination that tests knowledge that only provisional community service assistants have been in a position to acquire, plaintiffs seek to defeat the mandate of the New York Civil Service System that appointments and promotions in the Civil Service "shall be made according to merit and fitness to be ascertained, as far as practicable, by examination which, as far as practicable, shall be competitive." N.Y. Con. art. V, § 6. See Kirkland v. New York State Dept. of Correctional Serv., 520 F.2d 420, 428 (2d Cir. 1975). The fact that plaintiffs are familiar with the specifics of their jobs and have discharged their responsibilities well does not mean that the other candidates could not acquire the same knowledge and perform the duties of community service assistant as well or better.

The unfortunate aspect of this case is that, although plaintiffs performed well in their jobs and, on the whole, did reasonably well on the examination, their scores were not high enough to qualify them for appointment to a permanent position. We have a great deal of sympathy for plaintiffs; it may be that they were victims of economic conditions that force highly educated young people to compete for jobs with former welfare recipients. Nonetheless, in the absence of evidence that the examination had a disproportionate impact on minorities and that the exam was not job-related, we are not about to halt a laudatory attempt to create civil service positions for human service paraprofessionals that, at least in better economic times, will be available to low income individuals on a competitive basis.

Here the plaintiffs bear the burden of demonstrating that alternative testing devices, free of a racial impact, would also serve the employer's interest in finding suitable employees. Albemarle Paper Co. v. Moody, supra, 422 U.S. at 425, 95 S. Ct. at 2375. The best plaintiffs are able to do is suggest that an oral examination, in connection with on-the-job evaluation and the service of a probationary period, would satisfactorily meet the needs of human service agencies. No evidence was introduced to support this assertion; nor did plaintiffs indicate the difference, if any, between the substance of their suggested oral examination and the presently-used written one or what advantages accrue from an oral test. It is worth noting that part of the selection process for permanent appointment as community service assistant involves a job interview, which certainly tests oral ability, as well as an on-the-job probationary period (Mar. 2, tr. 211). Plaintiffs have failed to sustain their burden of demonstrating the feasibility and utility of alternative selection methods.

The white plaintiffs assert that the absence of "ascertainable standards in the development of the examination" and the lack of job-relatedness of the examination deprived them of substantive due process rights under the fourteenth amendment. Our earlier discussion of the job-related qualities of examination 64-914 disposes of this claim as well.

Moreover, in order to make out a violation of substantive due process rights, plaintiffs must show that examination 64-914 impaired their liberty or property interests. As stated by then Circuit Judge Stevens:

The Fourteenth Amendment prevents the state from depriving any person of liberty or property without due process of law. As [Board of Regents v. Roth, 408 U.S. 564, 92 S. Ct. 2701, 33 L. Ed. 2d 548] squarely holds, the right to procedural due process is applicable only to state action which impairs a person's interest in either liberty or property. Certainly the constitutional right to "substantive" due process is no greater than the right to procedural due process. Accordingly, the absence of any claim by the plaintiff that an interest in liberty or property has been impaired is a fatal defect in her substantive due process argument.

Jeffries v. Turkey Run Consolidated Sch. Dist., 492 F.2d 1, 4 (7th Cir. 1974) (footnote *1174 omitted). Accord, Buhr v. Buffalo Public School District No. 38, 509 F.2d 1196, 1202 (8th Cir. 1974).

Here the absence of a liberty or property interest is fatal to plaintiffs' claims. Plaintiffs have no protected property rights in the position of permanent community service assistant, a job that neither they or the other applicants had ever held. Nor does failure on a civil service examination carry a stigma that impairs plaintiffs' liberty interests. See Arnett v. Kennedy, 416 U.S. 134, 94 S. Ct. 1633, 40 L. Ed. 2d 15 (1974); Board of Regents v. Roth, 408 U.S. 564, 92 S. Ct. 2701, 33 L. Ed. 2d 548 (1972); Perry v. Sindermann, 408 U.S. 593, 92 S. Ct. 2694, 33 L. Ed. 2d 570 (1972); Slochower v. Board of Higher Education of New York City, 350 U.S. 551, 76 S. Ct. 637, 100 L. Ed. 692 (1956).

This memorandum of decision and order contains findings of fact and conclusions of law required under F.R.Civ.P. Rule 52(a).

The complaint is dismissed. It is hereby

ORDERED that the Clerk enter judgment in favor of the defendants and against the plaintiffs dismissing the complaint.

[1] The other defendants are Gabriel Kohn, Chairman of the Nassau County Civil Service Commission; Adele Leonard, executive director of the Nassau County Civil Service Commission; Edward Witanowski and Edward Simmons, Commissioners of the Nassau County Civil Service Commission; the Nassau County Office of Consumer Affairs; James E. Picken, the Commissioner of Consumer Affairs; The Nassau County Department of Social Services; Joseph D'Elia, the Commissioner of Social Services; the New York State Department of Civil Service; and Ersa H. Poston, president of the New York State Civil Service Commission and head of the New York State Civil Service Department.

[2] In 1972, Title VII was amended to extend its reach to state and municipal government employees. 86 Stat. 103 (1972).

[3] The other major agencies in Nassau County that use community service assistants are the Nassau County Department of Health and the Nassau County Department of Mental Health. (Mar. 2, tr. 185).

[4] The Nassau County Civil Service Commission developed the following description of the "typical duties" of a community service assistant:

1. Assists in contacting, recruiting, and organizing welfare clients and other low income families into self-help groups and organizations.

2. Assists client groups emplore [sic] and utilize services and resources available to them from the Department and other community agencies.

3. Attends and helps lead client group meetings on welfare legislation, job training and opportunities, neighborhood improvement projects, etc.

4. Attends training classes and staff meetings.

5. Presents oral and written reports. (Plaintiffs' Exhibit 7; see Mar. 3, tr. 259).

[5] These five requirements or qualifications are known in the terminology of test development as the KSA (knowledge, skills and abilities) statement. The KSA statement here was developed by Michael Dollard, an associate personnel examiner with the New York State Department of Civil Service. (Mar. 3, tr. 259).

[6] We take judicial notice that, especially in 1973, conditions in the metropolitan job market compelled young college-educated men and women to look for work in fields that they would not consider in better economic times.

[7] According to a study of the candidates in Nassau and Erie Counties conducted after the examination had been given, "[a] higher preportion [sic] of whites have more relevant training ... which combines with the greater numerical strength of whites ... and the higher proportion of young whites to yield a high number of young whites with substantial training (24 whites, 30 or younger, with two or more years of relevant college)." (Plaintiffs' Exhibit 1; see testimony of Michael Dollard, Mar. 3, tr. 305).

[8] The position of community service assistant exists in a number of agencies throughout the state. The examination developed for statewide use on June 2, 1973, was denominated "Test 483-A." (Plaintiffs' Exhibit 6). As administered in Nassau County on that day, the exam was entitled "Examination No. 64-914." Test 483-A and examination number 64-914 are identical.

[9] The mean score for whites was 44.50; the mean score for blacks was 39.01. (Plaintiffs' Exhibit 3).

[10] The minority passing rate on examination 64-914 is well within the 80% rule set forth in the Uniform Guidelines on Employee Selection Procedures proposed by the Equal Employment Opportunity Co-ordinating Council:

A selection rate for any racial, ethnic or sex group which is less than four-fifths ( 4/5 ) (or eighty percent) of the rate for the group with the highest rate will generally be regarded as evidence of adverse impact, while a greater than four-fifths rate will generally not be regarded as evidence of adverse impact.

Proposed EEOC Reg. § 4(b) 41 Fed.Reg. 29016, 29017 (July 14, 1976).

[11] The plaintiffs advance, or appear to advance, two other arguments to support their contention that examination 64-914 had a discriminatory impact. First, plaintiffs point out that an item-by-item analysis of the 60 questions on the examination reveals that in Nassau County blacks scored worse than whites on 55 of the questions. (Plaintiffs' Exhibit 3). In addition, according to a statewide analysis of the results of the examination, "[o]f the 60 items in the care [sic], 35 items show between group differences in item difficulty of .10 or greater; 13 items show differences of .20 or greater." (Plaintiffs' Exhibit 2).

However, cases in this circuit are in agreement that

a showing that the overall examination procedure produced disparate results cannot be rebutted by fragmenting the process and demonstrating that separately the parts did not differentiate along racial or cultural lines.

Kirkland v. New York State Dept. of Correctional Serv., 374 F. Supp. at 1370. Accord, Vulcan, 360 F. Supp. 1265 at 1272. The converse of this principle is equally true: if the final result of the examination procedure does not evidence a discriminatory impact, cultural or racial bias in individual test items is irrelevant. Eligibility for appointment is determined by the score on the entire examination, not by performance on individual questions. Whatever the relative difficulty of the component parts of the examination for blacks and whites, in the end the passing rate for blacks was not sufficiently below that of the passing rate for whites to support a finding that the examination had a discriminatory impact.

This brings us to the plaintiffs' other argument, which concerns the statewide impact of the examination. Here there was a failure of proof. In the plaintiffs' pretrial memorandum and in the opening remarks at the trial (Mar. 1, tr. 25-26), mention is made that statewide nearly twice as many whites passed as blacks. Yet the plaintiffs' post-trial memorandum does not refer to these statistics at all. More importantly, it does not appear that these statistics were ever introduced into evidence. The list of plaintiffs' exhibits (Plaintiffs' Post Trial Memorandum, at 11-13) does not include a reference to statewide statistics, although the ethnic analysis of the results in Nassau County are highlighted in great detail. (see Plaintiffs' Exhibit 3). The various witnesses called by the parties did not discuss the statewide results of the examination, except with respect to the item-by-item analysis. (see testimony of Michael Dollard, Mar. 3, tr. 284-96). We therefore conclude that plaintiffs have failed to sustain their burden of proof with respect to the statewide impact of the examination.

[12] Judge Weinfeld, in his opinion in Vulcan, defined content validation as follows:

Content validation is less preferable than the criterion-related methods of test validation but nevertheless a professionally accepted means of establishing job-relatedness where the more desirable empirical methods are impractical. An examination has content validity if the content of the examination matches the content of the job. For a test to be content valid, the aptitudes and skills required for successful examination performance must be those aptitudes and skills required for successful job performance. It is essential that the examination test these attributes both in proportion to their relative importance on the job and at the level of difficulty demanded by the job.

360 F. Supp. at 1274 (footnotes omitted).

[13] To quote Judge Weinfeld once again:

Predictive validation consists of a comparison between the examination scores and the subsequent job performance of those applicants who are hired. If there is a sufficient correlation between test scores and job performance, the examination is considered to be a valid or job-related one. Concurrent validation requires the administration of the examination to a group of current employees and a comparison between their relative scores and relative performance on the job. Experts called by both sides agreed that this method is less desirable than predictive validation because of the possibility that some distortion may result from either the experience or lack of motivation of the current employees who participate in the examination for experimental purposes. Nevertheless concurrent validation has its uses, because in the absence of an examination which has been used and predictively validated in the past, it is the only method of empirically determining whether an examination adequately assesses what it is intended to assess prior to its administration to actual job applicants.

360 F. Supp. at 1273.

[14] According to the Guidelines of the Equal Employment Opportunity Commission, the following are minimum standards for test validation:

§ 1607.5 Minimum standards for validation.

(a) For the purpose of satisfying the requirements of this part, empirical evidence in support of a test's validity must be based on studies employing generally accepted procedures for determining criterion-related validity, such as those described in "Standards for Educational and Psychological Tests and Manuals" published by American Psychological Association, 1200 17th Street NW., Washington, D.C. 20036. Evidence of content or construct validity, as defined in that publication, may also be appropriate where criterion-related validity is not feasible. However, evidence for content or construct validity should be accompanied by sufficient information from job analyses to demonstrate the relevance of the content (in the case of job knowledge or proficiency tests) or the construct (in the case of trait measures). Evidence of content validity alone may be acceptable for well-developed tests that consist of suitable samples of the essential knowledge, skills or behaviors composing the job in question. The types of knowledge, skills or behaviors contemplated here do not include those which can be acquired in a brief orientation to the job.

See Albemarle Paper Co. v. Moody, supra, 422 U.S. at 431, 95 S. Ct. at 2378.

[15] The following is a description of a proper criterion-related validation study. In a

criterion-related validity study, as in any validity study, the first step is the undertaking of a job analysis, a task which results in a thorough understanding of the job as well as of the major or critical incidents of job performance. Based upon this job analysis, various criteria which represent major or critical work behaviors are selected or developed to be used as measures of job performance. Also based upon the job analysis is the selection or construction of a test designed to measure job performance. Once the criteria and the test have been obtained or prepared, a sample of applicants or employees are then administered the test and are tested for their performance on the criteria. Their test performance is then compared with their performance on the criteria to determine whether high performance on the test correlates with high performance on the criteria and vice versa.

Larson, Remedies for Racial Discrimination in Government Employment, 5 Colum.Human Rights L.Rev. 335, 364 (1973) (footnotes omitted) (hereinafter cited as Larson). The point is that the "ideal procedure involves the administration of the tests to an applicant sample, all of whom are hired and whose subsequent job performance provides the criterion data." A. Anastasi, Psychological Testing 415 (3d ed. 1968), quoted in Larson, supra at 367 n. 191. Here a proper predictive validation would have entailed hiring a broad sample of the applicants for community service assistant; administering the test to them; observing and rating their performance over a reasonable amount of time; comparing the sample candidates' test scores with their job performances; and, assuming the test was found to be valid, then using the test as a standard competitive civil service examination.

It may be that the civil service system should be required to validate an examination several years after it is given by measuring the job performances of the selected candidates against their test scores. But with respect to an examination given for a position that did not exist previously, the administrative burden and financial cost of such a procedure are unreasonable, particularly when content validation is available.

[16] Although there are other community service assistants in the 16 state agencies where the examination was given, the size of the measuring group would still be a handicap in attempting a predictive validation. To quote from the expert testimony of Dr. Grace Wright on behalf of the defendants:

Predictive validity is not the appropriate strategy in this instance for the reason that there are several, but the major reasons are insufficiency of subjects. I believe there was seven and our standards suggest that thirty be a minimum. Some might some standards say fifty. Other standards say in the hundreds because many cases wash out. So that this would not be the appropriate strategy on which to make employment decisions ... that affect people's lives and employment is an important decision in a person's life.

(Mar. 3, tr. 334).

[17] Another method of validation is "construct validation," which requires

identification of general mental and psychological traits believed necessary to successful performance of the job in question. The qualifying examination must then be fashioned to test for the presence of these general traits.

Vulcan, 490 F.2d at 395. However, since neither party offered evidence as to the construct validity of the examination and since this method may be the least preferred validation method, see Douglas v. Hampton, 168 U.S.App.D.C. 62, 512 F.2d 976, 986 (1975); Larson, supra at 363 & n. 174, there is no reason to analyze the examination for construct validity.

[18] According to the testimony of Michael Dollard (Mar. 2-3, tr. 195-326), the civil service commission normally develops an examination in response to requests from local agencies. The agencies submit background information on the position to be tested and an official job specification. The civil service examiner reviews the job description, with emphasis on the KSA (knowledge, skills, and abilities) statement, to see whether it conforms to other available examinations, in which case the requests are set aside. If a new examination needs to be developed, a job analysis is performed (usually based on the official job description submitted by the local agencies) and a mode of examination is selected, i. e., a written test as opposed to an oral exam. If a written test format is selected, a test plan is designed and submitted to local agencies for review. In some cases, the local agencies are asked to submit draft questions for inclusion in the examination.

Essentially this process was followed in developing test 483-A (see note 8 supra). In 1972, the state civil service commission was already planning an examination series for local human services paraprofessionals and had contacted an outside examiner for assistance in developing materials. Because a number of agencies had requested an exam for paraprofessionals, the material submitted by the agencies was used in developing the statewide paraprofessional examination. The examiner in charge of the paraprofessional test, Michael Dollard, who had five years of experience developing examinations for paraprofessionals, met with representatives of the State Department of Social Services to compare concepts of the paraprofessional position. At this meeting, according to the testimony of Mr. Dollard (Mar. 3, tr. 240),

it was agreed that generally their idea of what a paraprofessional in the human services area was was the same idea that we had. And that the test that we were developing for local paraprofessionals would, in fact, fit the positions they had in mind.

In addition, local social service agencies throughout the state were contacted and asked to submit draft multiple choice questions, based on the job analysis performed by Mr. Dollard, for an examination for paraprofessionals; individuals within the State Department of Social Services, who had experience supervising paraprofessionals, were also asked to submit draft questions; and a local social service agency was requested to provide samples of the forms and records used in the course of a paraprofessional's workday.