Samuel CARROLL, Et Al., Plaintiffs-Appellants, v. SEARS, ROEBUCK & COMPANY, Defendant-Appellee

WISDOM, Circuit Judge: This case raises issues related to hiring, job assignment, training, promotion, compensation, and termination at the Shreveport facilities of the defendant, Sears, Roebuck & Company (“Sears”). The plaintiffs, two black employees of Sears in Shreveport, filed this class action under Title VII, Civil Rights Act of 1964, 42 U.S.C. § 2000e to 2000e-17 (1976), alleging that employment practices at Sears discriminated against blacks. The district court rejected these allegations. We affirm in part and reverse in part. I. Background Sears is a national chain of department stores and related facilities which employs 375,000 people. It has divided the nation into five administrative areas, or territories, which report directly to Sears’s corporate headquarters in Chicago. During the time period relevant to this lawsuit, the Shreveport facilities were located in the Southwest Territory, with headquarters in Dallas. The Shreveport facilities consist of the Southern Avenue Store, the Retail Distribution Center, the Central Service Unit, and the Southside Store. Sears employs ap *187 proximately 600 people in Shreveport in seven EEO job categories, although most of its employees work in sales. 1 In November 1971, Sears hired Samuei Carroll as a part-time truck helper after Carroll had requested full-time employment. Sears never promoted Carroll to a full-time position despite numerous requests on Carroll’s part for promotion. In April 1973, Sears terminated Carroll for insubordination after Carroll was involved in a heated argument over failing to report for work on what was usually his day off. Carroll contends that he was terminated because he is black. Sears hired Charles W. Grant in August 1971 as a part-time truck helper. Grant, after repeated requests for promotion to full-time status, was promoted to a full-time position. Sears laid-off Grant in 1975 during a large-scale reduction in force, and Sears did not rehire him when he reapplied for work. Grant contends that he was refused promotion and laid off by Sears because he is black. Sears contends that Grant’s layoff was based solely on his lack of seniority and that he was not rehired because he failed to comply with Sears’s policy that all applicants for employment must pass a certified physical examination by a company physician. Carroll and Grant filed this class action under Title VII on July 5, 1977. Their complaint alleged that Sears discriminated across the board in its employment practices. The district court certified the following class on May 14, 1980: All present black employees of defendant Sears, Roebuck and Company’s Shreveport facilities on or after December 22, 1972; all future black employees of defendant’s Shreveport store; and all unsuccessful black applicants who applied for employment at defendant’s Shreveport facilities and who were rejected on or after December 22, 1972. At trial, the plaintiffs largely relied on statistics and expert testimony to establish their prima facie case. Sears responded with extensive statistical evidence and expert testimony. The district court rejected all the charges of discrimination against Sears. The district court held that the plaintiffs failed to show that Sears’s employment practices had a disparate impact on blacks in hiring, job placement, promotion, training, testing, or termination. The district court found the plaintiffs’ statistical evidence to be flawed or inconclusive. The district court concluded that the presence of a successful affirmative action program at Sears was proof that Sears did not discriminate and that the plaintiffs’ allegation of disparate treatment, including the individual Title VII claim of Carroll, was meritless. On appeal, Carroll contests a number of the district court’s factual findings and legal conclusions. Carroll contends that the district court erred in accepting testimony that directly contradicted pre-trial stipulations and in refusing to give any credence to the testimony of a former personnel manager at Sears in Shreveport. Carroll’s main contention is that the district court erred in failing to find that the plaintiffs established a prima facie case of racial discrimination in defendant’s hiring, promotion, job classification, training, termination, and testing practices. Carroll argues' that the district court misinterpreted the statistics presented to it and erred in assessing the effect of Sears’s use of subjective employment criteria on hiring, job classifi-. 1 cation, and promotion decisions. Sears contends that the district court properly found that there was no discrimination based on the statistical evidence and testimony before it. II. Preliminary Issues Two preliminary issues must be resolved before we turn our attention to the evidence on discrimination. First, Carroll *188 contends that the district court erred in not giving any credibility to the testimony of Ms. Esther Colbert Wilkins, the personnel manager of Sears from June 1972 to October 1977, because she had resigned from Sears and unsuccessfully conducted a sex discrimination suit against Sears in 1980. Second, Carroll contends that the district court made evidentiary rulings that were erroneous. According to Carroll, the district court allowed Sears to introduce a wage book as evidence which was wrongfully withheld from the plaintiffs by Sears. The district court also allowed testimony that Sears usually promoted and transferred its employees within EEO job categories. Carroll argues that allowing this testimony was contrary to facts stipulated to in the pre-trial order and to the testimony in the case. On issues of credibility, the district court’s findings cannot be reversed unless they are clearly erroneous. Robbins v. White-Wilson Medical Clinic, Inc., 5 Cir. 1981, 660 F.2d 1064, 1066. The district court’s determination that Ms. Wilkins’s testimony that was unfavorable to Sears would not be given credence is not clearly erroneous. In April 1980 the district judge ruled that Ms. Wilkins’s allegations of sex discrimination against Sears were unfounded. The district court concluded, based on her earlier trial and on her testimony in this litigation, that she “harbors strong feelings of animosity and resentment toward Sears.” Carroll v. Sears, Roebuck and Company, 514 F.Supp. 788, 795 n. 4 (W.D.La.1981). The district court had a reasonable basis for its conclusion, and its credibility determination with respect to Ms. Wilkins is not clearly erroneous. We also reject Carroll’s contentions that other evidentiary rulings were in error. The district court did not err in allowing the wage book into evidence. Sears stated at trial that the wage book was never requested during discovery, and it gave the information in the wage book to the plaintiffs in answers to interrogatories and in a computer tape containing hiring statistics. Record on Appeal at 2088-97, 2133. There is no error in allowing the wage book into evidence when the plaintiffs had access to all the information contained in it. We also find that the district court did not err in accepting testimony that promotion at Sears is usually within EEO categories. The parties did stipulate in the pre-trial order that Sears “routinely transfers employees to different departments” within the Shreveport facilities, but this is not a concession by Sears that transfers between departments are the usual practice at Sears. There is a difference between a usual practice and a permitted departure from that practice which is routine. The pre-trial stipulation did not preclude Sears from proving this point or prevent the district court from considering it. III. The Class Claim A. Disparate Impact The plaintiffs contend that the evidence presented to the district court established a prima facie case of racial discrimination under the disparate impact theory in the hiring, promotion, job classification, training, termination, and testing practices of Sears. We conclude, however, that the proof adduced at trial does not fit the disparate impact model. The use of subjective criteria to evaluate employees in hiring and job placement decisions is not within the category of facially neutral procedures to which the disparate impact model is applied. Carpenter v. Stephen F. Austin State University, 706 F.2d 608 (5th Cir. 1983); Pegues v. Mississippi State Employment Service, 669 F.2d 760, 765 (5th Cir. 1983); Pouncy v. Prudential Insurance Co. of America, 5 Cir.1982, 668 F.2d 795, 800— 01; Dickens & Beaver, Disparate Treatment & Disparate Impact Under Title VII — The Difference, Evidentia, Aug. 1982, at 1; see generally D. Baldus & J. Cole, Statistical Proof of Discrimination §§ 1.23 & 1.25 (1980 & Supp.1982) [“Baldus & Cole”]. With respect to training, promotion, termination, and job classification, the plaintiffs have failed to focus on a facially neutral, objective employment practice that *189 the disparate impact model was designed to; test. Under Pouncy, the disparate impact model can no longer serve as a “vehicle from which to launch a wide ranging attack on the cumulative effect of a company’s employment practices.” Pouncy, 668 F.2d at 800. We must therefore analyze the. plaintiffs’ challenge to Sears’s employment; practices under the disparate treatment! model. The only employment practice subject to analysis under the disparate impact model is Sears’s testing of employees. See Griggs v. Duke Power Co., 1971, 401 U.S. 424, 91 S.Ct. 849, 28 L.Ed.2d 158; Pouncy, 668 F.2d at 800. The use of a test is a specific, facially neutral selection criterion that can be shown to have a causal connection to a racial imbalance in the work force. This selection criterion is the kind of employment practice to which the disparate impact model traditionally has applied. At trial, the plaintiffs presented evidence that Sears’s employment tests 2 were not job related and that blacks on the average scored much lower than whites. Sears responded with testimony that its tests were job-related, properly validated, and non-discriminatory. The district court, however, did not shift the burden to Sears to prove job relatedness because of its finding that Sears did not discriminate against blacks in its hiring and promotion practices. The district court also found that the tests did not select applicants for hire, but were only one factor in hiring or promotion decisions. 3 We also find that the plaintiffs failed to make a prima facie showing of discrimination against blacks in hiring or promotion as a result of the tests used at Sears. The flaw in the plaintiffs’ proof was its failure to establish the required causal connection between the challenged employment practice (testing) and discrimination in the work force. See Pouncy, 668 F.2d at 801. Because the test does not have a cut-off and is only one of many factors in decisions to hire or promote, the fact that blacks score lower does not automatically result in disqualification of disproportionate numbers of blacks as in cases involving cut-offs. In Griggs, for example, applicants were required to achieve a minimum satisfactory score on two professionally prepared aptitude tests to qualify for placement. Griggs, 401 U.S. at 427-28, 91 S.Ct. at 851-52, 28 L.Ed.2d at 162. In this case, unlike Griggs, the plaintiffs have not shown through their statistics that the lower scores of blacks is a factor in employment decisions that causes fewer blacks to be selected. The plaintiffs’ proof on testing was fundamentally flawed in another respect. The only proof of discrimination as a result of testing was that 45.3 percent of hirees were black whereas 56.5 percent of the applicants were black. The plaintiffs contend that this disparity results both from testing and the use of subjective hiring criteria, yet they offer no method from which this Court can ascertain whether a significant part of this disparity results from testing. The plaintiffs simply have not shown that testing, independent of other factors that may affect the racial balance of the workforce, is causally related to discrimination in the number of blacks hired or promoted. The causal requirement recognizes that under-representation of blacks might result from *190 any number of factors, and it places an initial burden on the plaintiff to show that the specific factor challenged under the disparate impact model results in the discriminatory impact. The plaintiffs, by failing to isolate the discriminatory effect of the practice they challenged, did not meet this burden. 4 B. Disparate Treatment Although the plaintiffs structured their evidence to fit a disparate impact case, the record also shows that the plaintiffs contend that Sears discriminated against black applicants by treating them differently from whites. The plaintiffs’ model of proof therefore presents a class-wide disparate treatment case, and we analyze it under this theory of discrimination. To prove disparate treatment, plaintiffs must prove discriminatory intent and demonstrate more than “the mere occurrence of isolated or ‘accidental’ or sporadic discriminatory acts.” International Brotherhood of Teamsters v. United States, 1977, 431 U.S. 324, 336, 97 S.Ct. 1843, 1854, 52 L.Ed.2d 396, 416. They must show by a preponderance of the evidence that discrimination was the [defendant’s] standard operating procedure — the regular rather than the unusual practice.” Id. at 336, 97 S.Ct. at 1854, 52 L.Ed.2d at 416. A prima facie case of disparate treatment may be established by the use of statistics if a “gross” disparity in the treatment of workers based on race is shown. 5 See Hazelwood School District v. United States, 1977, 433 U.S. 299, 97 S.Ct. 2736, 53 L.Ed.2d 768. Gross statistical disparities, standing alone, may justify an inference of discriminatory motive. See Payne v. Tra-venol Laboratories, Inc., 5 Cir.1982, 673 F.2d 798, 817; Williams v. New Orleans Steamship Association, 5 Cir.1982, 673 F.2d 742, 749. Other evidence is probative to prove disparate treatment: The statistical showing of disparate result may also be buttressed with evidence of a history of discrimination practiced by the employer, individual instances of discrimination, and opportunities to discriminate in the employer’s decision-making processes. If the statistical disparity is insufficient alone to establish a prima facie case, the plaintiff may get over his or her initial hurdle by combining statistics with historical, individual, or circumstantial evidence. See Payne, 673 F.2d at 817. If the plaintiffs establish a prima facie case, the defendant must discredit their evidence or provide a nondiscriminatory explanation for the apparently discriminatory result in order to prevail. See Payne, 673 F.2d at 817; Williams, 673 F.2d at 741. 6 1. Hiring Practices The plaintiffs contend that Sears discriminated against blacks in its hiring practices. The evidence at trial showed that Sears accepted applications for jobs at its personnel office six days a week. The hiring criteria used by personnel interviewers at Sears are highly subjective for many job positions. For sales positions, applicants are evaluated on their appearance and *191 on attributes such as articulateness, maturity, aggressiveness, and friendliness. For other jobs, such as automobile mechanics, the qualifications required are more objective, and qualified, experienced applicants are required. The plaintiffs presented statistical evidence based on applicant flow data to show that Sears’s hiring procedures resulted in discrimination. Between 1973 and 1979, 56.5 percent of the job applicants were black, whereas only 45.3 percent of the applicants hired were black. Based on the assumption that blacks, and therefore black applicants at Sears, possessed the physical and personality characteristics desired by Sears in the same proportion as whites, the plaintiffs’ expert concluded that the expected number of black hirees at Sears absent discrimination would equal their percentage in the applicant pool. According to the plaintiffs, that the percentage of hirees is less shows the existence of discrimination. The district court found the plaintiffs’ statistics to be unreliable because the raw, unadjusted applicant flow data fails to take into account that certain jobs at Sears required special qualifications 7 and that certain factors artifically inflated the number of black applicants at Sears. 8 We do not reach this issue because we conclude that these statistics do not establish a gross disparity in treatment even at their unadjusted level. Sears hired 45.3 percent of all black applicants during the relevant period and increased its number of black employees by 22.3 percent. This increase occurred despite a decrease of 13 percent in Sears’s workforce during this period and despite low turnover at the Shreveport stores. The 45.3 percent figure is significantly higher than black representation in the population (30 percent) or in the workforce (28 percent). When Sears’s expert took the qualifications of the applicants into account and assigned appropriate weight to the occurrence of each job category in the SMSA in proportion to its occurrence in the Sears’s workforce, he concluded that the actual number of blacks hired exceeded the expected number of blacks for all positions except laborer. The plaintiffs’ statistics do not establish a gross disparity in the number of blacks hired or provide evidence of purposeful discrimination. The plaintiffs have also failed to produce convincing evidence of a history of discrimination or individual instances of discrimination. Absent this evidence and in the light of the evidence that Sears Affirmative Action Program has resulted in a substantial increase in the number of blacks hired, we hold that the plaintiffs failed to establish their prima facie case of discrimination in hiring under the disparate treatment theory. 2. Job Assignment The plaintiffs next contend that Sears assigns a disproportionate number of blacks to part-time positions and to lower paying, less prestigious jobs. With respect to assigning too many blacks to part-time positions, the district court found virtually *192 no evidence of discrimination. During the period for which statistics are available, Sears assigned 79.3 percent of white applicants and 85.3 percent of black applicants to part-time positions. Sears’s expert, Dr. Levine, testified that this disparity is not statistically significant. We conclude that there was no discrimination in the assignment of workers into part-time or full-time jobs. The district court also found that the plaintiffs’ statistical evidence did not show that Sears assigned blacks into lower status job classifications in disproportionate numbers. The district court found that their data base was inaccurate, that they failed to account for job qualifications in categories where experience or training was needed, and that they made no definitive indication concerning which jobs are in fact “lower status”. Dr. Levine also testified that the actual number of black hires into each EEO job category except “laborers” exceeds the expected number of black hires into these categories. Based on this evidence, the district court properly concluded that the plaintiffs did not show that blacks were placed in lower status jobs in disproportionate numbers. 3. Promotions The promotional practices at Sears warrant strict scrutiny by this Court. Sears’s promotion criteria are predominantly subjective. In the pretrial order, Sears stipulated that its criteria for promotion include fitness and demonstrated ability, employees’ past work histories and abilities to produce, annual evaluations, and the opinions of supervisors. Sears has no written criteria or guidelines for promotion, and does not post notices concerning specific job openings or promotion opportunities. It is undisputed that a majority of the supervisors are white. Given the importance of supervisory ratings and opinions and the presence of unwritten, subjective criteria for promotion, black employees face a greater risk of discrimination at Sears. See James v. Stockham Valves & Fittings, 5 Cir.1977, 559 F.2d 310, 346; Rowe v. General Motors Corp., 5 Cir.1972, 457 F.2d 348, 358-59; Stacy, Subjective Criteria in Employment Decisions Under Title VII, 10 Ga. L.Rev. 737 (1976). The plaintiffs’ exhibits showed the number and percentage of blacks in various categories in its workforce. These figures show the relevant pool for promotion because Sears promotes workers from within its own labor force. See Johnson v. Uncle Ben’s Inc., 5 Cir.1980, 628 F.2d 419, 425, vacated, 1981, 450 U.S. 902, 101 S.Ct. 1967, 68 L.Ed.2d 290; James v. Stockham Valves, 559 F.2d at 331, 341. Over time, the percentage of black employees in the upper-level positions should correspond to the percentage of black employees in the workforce as a whole if no discrimination is taking place. See Rowe, 457 F.2d at 357. The plaintiffs contend that their statistics show an underrepresentation of blacks at the higher level positions at Sears when compared to the number of blacks in the entire workforce. First, the plaintiffs argue that discrimination against blacks takes place in promotion from part-time to full-time positions. We conclude that the statistics do not bear out this allegation. From 1974 to 1980, the number of black full-time employees increased from 21.7 percent to 32.7 percent of the workforce during a time when the total full time workforce decreased by 15.6 percent. Over this period, the number of full-time positions decreased from 424 to 358 while the number of black full-time employees increased from 92 to 117. The percentage of black full-time employees has increased along with the increases in the black workforce, although there is a small lag in the percentages. The district court explained this lag with evidence that many part-time employees did not want to be promoted and that there is a low turnover *193 at the Shreveport stores. Based on this evidence, we hold that the plaintiffs have failed to show a gross disparity in the number of blacks and whites promoted from part-time to full-time positions that would support a finding of disparate treatment. 9 The question whether blacks have been discriminated against in promotion to upperlevel jobs is more troublesome. The number and percentage of blacks have increased dramatically in recent years in many of what appear to be the more desirable job categories, such as sales and craftsmen, and have decreased in jobs of apparently lower status, such as tire installer, warehouseman, and stock clerk. Most notable, however, is that Sears never had a salaried black employee during the period covered by this lawsuit even though there were 22 to 26 such positions in its workforce. Of the division managers, the total number of blacks varied from 14.5 percent in 1974 to 18.7 percent in 1977 even though the workforce was 26.8 percent black in 1974 and 34.3 percent black in 1977. For the category “other managers,” only 2 of 49 were black in 1974, and there were no blacks of 28 in 1977. These comparisons, and especially the “inexorable zero” found in the salaried employee category, show a gross disparity in the treatment of blacks in promotion to top managerial positions. 10 See Teamsters, 431 U.S. at 342 n. 23, 97 S.Ct. at 1858 n. 23, 52 L.Ed.2d 419 n. 23. The plaintiffs also supported their statistics with convincing anecdotal evidence. A black employee, Monroe Smith, was promoted to installation manager, a position held by whites both before and after Smith. When the white employees before and after Smith held the job, it was a salaried position; when Smith held the job, it was not. This evidence, when combined with the great underrepresentation of blacks in high managerial and the presence of a promotional system which is inherently prone to discriminate against blacks, compels our conclusion that the plaintiffs made a prima facie case of disparate treatment in the promotion of blacks to manager and salaried positions. The only evidence to rebut this conclusion is testimony by Sears that its Shreveport stores had low turnover in these positions. This testimony was vague and conclusory, and it was not supported by any statistics showing how long employees remain in salaried or managerial positions or how many of these positions were available during the relevant time period. This evidence does not rebut the plaintiffs’ showing of discrimination. Nor do we accept Sears’s argument, relied on by the district court, that the promotion of employees to salaried positions occurred at the Southwest Territorial Office and is therefore outside the scope of this lawsuit. The defendant in this case is Sears. The Shreveport store is not a separate entity. Sears presented substantial evidence on its national affirmative action program and personnel policies, and Sears did not object to evidence concerning the promotion of blacks to salaried positions. We reject the contention that Sears is absolved from discriminatory policies in its promotional practices simply because the discrimination originated from the territori *194 al office rather than in Shreveport. We hold that Sears did not rebut the plaintiffs’ prima facie case of disparate treatment in the promotion of blacks. 4. Training The plaintiffs contend that their evidence shows an underrepresentation of black employees in Sears’s training programs. 11 The plaintiffs presented evidence that the employees black representation was 16 percent in the SEI courses, 11 percent in technical training courses, and 18 percent in the management training courses. These figures are below the percentage of blacks in Sears’s workforce, and the plaintiffs argue that this apparent underrepresentation is proof of discrimination. We reject this contention. The SEI programs were available to all employees on request and approval of these requests was given as a matter of course. There was ample evidence to support the district court’s conclusion that the low number of blacks taking the course reflect the percentage of blacks interested in these courses and that no discrimination is present. We reach the same conclusion with respect to the management training courses. Any interested individual could attend the course, and there was no evidence that black individuals who were interested in attending were denied the opportunity to do so. In fact, the most recent management training course trained five blacks and seven whites. For the technical training courses, the plaintiffs failed to make the relevant comparison. Not all black or white employees were eligible for this training because they were given to employees in technical spe-cialities. When the eligibility of employees is taken into account, 59 percent of eligible whites and 57.1 percent of eligible blacks attended. There was also no class member who testified that he was denied technical training. Based on this evidence, we conclude that the plaintiffs failed to prove a prima facie case of discrimination against blacks in the area of job training. 5. Terminations The plaintiffs’ final contention is that Sears discriminated against blacks in firing. They presented evidence which they argued was proof that a disproportionate number of blacks were fired. The district court focused on firings for subjective reasons, concluding that if a black was fired for an objective reason, such as tardiness, issuing bad checks, or misappropriating company property, there can be no reasonable argument that the termination was based upon racial considerations. The district court accepted the testimony of Dr. Levine that the proportion of black laborers and service employees fired during this period approximated the representation of blacks in those job categories from 1974 to 1979. 12 The district court found that the number of blacks fired in other categories *195 was higher than might be expected, but the small numbers involved and the lack of a significant disparity did not support an inference of discrimination in termination. We hold that the district court’s finding of no discrimination in terminations is not clearly erroneous. The district court appropriately focused on subjective terminations because proper grounds were shown to exist for the terminations based on objective criteria. On the evidence before it, the district court could reasonably conclude that the number of terminations in each job category reflected the racial composition of the job category and did not show discrimination. We hold that the plaintiffs failed to prove a prima facie case of discrimination in terminations. 6. Conclusion This Court has recently stated that a district court, in assessing statistical presentations in Title VII class actions, should “accept what figures are available; allow for imperfections, skewing factors, and margins of error; and then take the figures for what they are worth.” Phillips v. Joint Legislative Committee, 5 Cir.1981, 637 F.2d 1014, 1025. With the exception of promotion, the district court properly evaluated the statistical and nonstatistical evidence to conclude that the plaintiffs failed to establish a prima facie case of discrimination. We note that the plaintiffs’ statistical evidence was often not as focused or helpful as it might have been. — We again warn plain-' tiffs that their statistical evidence should show more than a racial imbalance in the workforce, but should prove that' the chal'lenged practices result in discrimination. \If LtM plaintiffs’ statistics fail to establish a causal link between the challenged practice and discrimination, courts can infer discrimination only when the “inexorable zero” or gross statistical disparities in the “bottom line” compel this inference, as with the statistics on promotion in this case. We therefore affirm the district court’s holding that the plaintiffs did not establish a prima facie case in the hiring, job assignment, training, and firing practices at Sears, but we reverse the district court’s holding that Sears did not discriminate in the promotion of blacks to division manager, other manager, and other salaried positions. IV. Carroll’s Individual Claim The remaining issues concern Carroll’s contention that he received disparate treatment at Sears in the areas of hiring, promotion, training, and termination. 13 In a disparate treatment case involving an individual, the plaintiff initially has the burden of showing a prima facie case of discrimination. If the plaintiff succeeds in this showing, the burden shifts to the employer to rebut the presumption of discrimination by producing evidence that the plaintiff was rejected, or someone else was preferred, for a legitimate, non-discriminatory reason. If this rebuttal is successful, the plaintiff must demonstrate that the proffered reason was not the true reason for the employment decision. The plaintiff bears the ultimate burden of persuading the court that he has been a victim of intentional discrimination throughout the trial. See Texas Department of Community Affairs v. Burdine, 1981, 450 U.S. 248, 101 S.Ct. 1089, 67 L.Ed.2d 207. Carroll contends that Sears discriminated against him when it hired him for a part-time rather than a full-time position. jThe district court found this allegation to ¡be meritless. Carroll did not establish that Sears was seeking applicants for a full-time position when he applied or continued to seek full-time applicants after Carroll was hired part-time. Carroll admitted on direct examination that Sears hired him part-time because no full-time positions were available at the time of hiring. We hold that Carroll’s mere suspicion that his hiring for a *196 part-time position was discriminatory did not establish disparate treatment in his hiring. See Long v. Sapp, 5 Cir.1974, 502 F.2d 34, 37. Carroll next contends that he was not promoted to a full-time position because he is black. Two supervisors at Sears testified that Carroll was unpleasant and could not be counted on to perform his work. Based on this evidence, the district court found that Carroll was not promoted because of his temperment and lack of dependability rather than his race. This finding is not clearly erroneous, and we hold that there was no discrimination against Carroll in the area of promotions. Carroll also contended that there was discrimination against him in the area of training. Carroll testified that he was unaware of training opportunities because no notices were posted and no one told him about training programs. The district court rejected this contention, relying on testimony from several witnesses stating that notices about training programs were always posted in conspicuous places at Sears. Carroll has no one to blame but himself for his failure to look at the notices. We hold that Carroll failed to establish that he was denied training based on his race. Carroll’s final contention concerns his termination. 14 The district court found that Carroll was asked to work on his usual day off. Carroll complained, and Ms. Wilkins gave him permission to take the day off conditioned on his clearing the matter with Lowrey. 15 Carroll, however, never secured Lowrey’s approval and took the day off anyway. Ms. Wilkins called Carroll to determine why he was absent, and Carroll became belligerent. Carroll continued his belligerent speech with Vic Parrish, Sears’s Operating Superintendent. Parish, Wilkins, and Lowrey agreed that Carroll should be terminated for insubordination, although Wilkins later testified at trial that Carroll’s reaction was understandable given the circumstances. The district court found that Carroll was terminated not because he was black, but for insubordination, and we hold that this finding is not clearly erroneous. The district court’s finding is also supported by the testimony that Carroll was “a mean guy who said anything that came to his mind” and undependable. Given these facts, we hold that Carroll’s termination did not result from discrimination, but from his inexcusable conduct at work. V. Conclusion On the issue of testing, we affirm the district court’s holding that the use of tests did not have a disparate impact on black employees. We affirm the district court’s holdings with respect to Carroll’s individual claims of disparate treatment. For the class claims of disparate treatment, we affirm the district court’s holdings with respect to hiring, job assignment, training, and termination, but reverse on the holding of no discrimination in promotion. On remand, we instruct the district court to fashion a remedy for discrimination in promotion that is consistent with the holding of this opinion. AFFIRMED in part, REVERSED in part, and REMANDED. Footnotes [1] “EEO”, or “EEO-1”, job categories are seven general job categories which cover the specific job classifications at Sears. The name derives from the “EEO-1” report which Sears and other government contractors fill out so that their employment practices may be reviewed for compliance with the civil rights laws. The seven categories include: (1) officials and managers; (2) sales; (3) clerical; (4) craftsmen; (5) operatives; (6) laborers; and (7) service. [2] Sears administers several employment tests to its employees, including the “Sears Test of Mental Alertness” (STMA), the “Sears Clerical Test” (SCT), and the “Sears Arithmetic Skills Test” (SAST). These tests attempt to measure an applicant’s or employee’s “aptitude and ability.” Sears also used the "Thurstone Temperament Schedule” (TTS) to measure personality, and administers certain knowledge and skill tests to applicants for skilled positions, such as the “Mechanical Aptitude Schedule” and the “Mechanical Knowledge Inventory”. Sears gives different combinations of these tests to different employees, depending on the position into which the employee is hired. For an overview of employee testing under Title VII, see C. Sullivan, M. Zimmer & R. Richards, Federal Statutory Law of Employment Discrimination § 2.2 (1980 & 1980 Supp.). [3] The district court found that there was no pass/fail cutoff score associated with the test. The test alone would neither qualify nor disqualify an individual from being hired or promoted. [4] For example, in Griggs, the plaintiff proved not just a racial imbalance in the employer’s work force, but also that the challenged educational and testing practices resulted in the hiring of a lower percentage of blacks than whites. Griggs, 401 U.S. at 430 n. 6, 91 S.Ct. at 853 n. 6, 28 L.Ed.2d at 163 n. 6. See James v. Stockham Valves & Fittings Co., 5 Cir.1977, 559 F.2d 310, in which the plaintiff satisfied the causal connection by showing that blacks were greatly underrepresented in three job positions for which high tests scores were required. We also note that the plaintiffs did not produce a single witness who testified that his test scores kept him from being promoted or hired. [5] This Court has recognized repeatedly the limits of statistical evidence. Statistics must be used meaningfully and properly. The Supreme Court has cautioned “that statistics are not irrefutable; they come in infinite variety and, like any other kind of evidence, their usefulness depends on all the surrounding facts and circumstances.” Teamsters, 431 U.S. at 340, 97 S.Ct. at 1857, 52 L.Ed.2d at 418. [6] Findings of discrimination and discriminatory intent are subject to the clearly erroneous standard of review. See Pullman-Standard v. Swint, 1982, 456 U.S. 273, 102 S.Ct. 1781, 72 L.Ed.2d 66; Hill v. K-Mart Corp., 5 Cir.1983, 699 F.2d 776, 781, [7] There was substantial evidence that many of the positions at Sears required objective qualifications that would not necessarily be randomly distributed throughout the population as the plaintiffs assumed. For example, auto mechanics were required to have certain levels of experience and service mechanics needed a state license for employment in electronics repair. The plaintiffs did not adjust their applicant flow data to account for qualifications required by Sears. See D. Baldus & J. Cole, § 4.1 at 109; Smith & Abram, Quantitative Analysis and Proof of Employment Discrimination, 1981 U.Ill.L.Rev. 33, 58-59. [8] The district court cited several factors which would have led to a disproportionate number of black applicants at Sears. Sears is in close proximity to the State Employment Office, and a significant number of unemployed blacks applied for work at Sears in order to remain eligible for Unemployment compensation. Sears has recruited blacks actively under its affirmative action program and has had significant contacts with area high schools, trade schools, and colleges to contact potential black hirees. The applicant flow also includes a large influx of black applicants from the Job Corps and various drug and alcohol rehabilitation programs. The district court concluded that these factors resulted in an artificially high black applicant flow and made the use of applicant flow data inherently suspect. See Baldus & Cole, § 4.1 at 109 & Supp. 35, § 9 at 292. [9] We note that no statistics were presented by either party on the number of promotions from full-time to part-time during the relevant period and the percentage of these promotions that went to whites or blacks. Without these statistics, which obviously would be the most helpful in determining if discrimination took place, we can analyze only the increases in the numbers and percentages of blacks in the full-time positions, part-time positions, and the overall workforce to determine if discrimination is present. [10] We also note that this discrimination was prolonged in nature. Sears made little or no gain in the promotion of blacks to managerial or salaried positions between 1972 and 1979. This supports the conclusion that discrimination in promotion was the “regular rather than the unusual practice.” Teamsters, 431 U.S. at 336, 97 S.Ct. at 1855, 52 L.Ed.2d at 416. [11] Sears offers many different kinds of training programs at the local and territorial level. The Sears Extension Institute (SEI) courses are locally-conducted, self-improvement courses which are available on request to all employees. The SEI courses cover a variety of subjects and are not required for hiring or promotion. Technical courses are open to eligible employees (service technicians, mechanics) and are taught in Dallas. The number of openings in these classes is limited. The division manager training course is a 13-14 week course taught in the Southern Avenue Store by the staff. This course is open to anyone who is interested in taking it. [12] Dr. Levine broke down the number of blacks and whites fired for subjective reasons within the seven EEO categories as follows: Subjective Terminations 1974-79 EEO Categories No. Blacks No. Whites Officials and Managers 0 3 Sales 0 7 Clerical 6 1 Craftsmen 5 7 Operatives 0 4 Laborers 10 5 Service 6 1 Dr. Levine examined the files of every company-initiated termination from 1974 to 1979 and determined which occurred for subjective reasons, such as unsatisfactory group adjustment, *195 unsatisfactory quantitative performance, and unsatisfactory qualitative performance. [13] Charles Grant did not appeal the district court’s disposition of his individual claims. The relief available to Grant on appeal is that provided to him by virtue of his membership in the class action. [14] In Thompson v, LeLand Police Department, 5 Cir.1980, 633 F.2d 1111, 1114, this Court set forth the elements of a plaintiff’s burden of proof to establish a prima facie case of discriminatory discharge: (1) he is a member of the protected class; (2) he was discharged; (3) the discharge was not for a valid reason; and (4) the defendant sought applicants for the vacant position. [15] The issue whether Ms. Wilkins personally gave Carroll the day off was hotly contested. The district court found that she did not, based on her admission that only Lowrey, as Carroll’s ultimate superior in the warehouse, had the final authority to give Carroll the day off. This finding is reasonably supported by the evidence and is not clearly erroneous.