Instructions

 

Read the case below.  Provide the information that is requested at the end of the case.  Be careful in reading the case.  Scrutinize the information in the case as well as the information in the “Statistical Data” section.

 

The length of the assignment is not of importance.  The length should be dictated by the explanations you provide surrounding your recommendations.  What is of importance is whether or not you make the correct recommendation, and whether that recommendation is based on an analysis of the correct information.

 

Good luck.

 

The Delta Intelligence Test

 

For the past ten years, Eastern Amalgamated Paper Company (EAPCO) has used the Delta Intelligence Test (DIT) to select employees for its unskilled production jobs.  There are approximately 1,000 employees in these jobs, and the average turnover rate in EAPCO plants is high – about 35 percent each year.  Pay for employees in these jobs is around $20,000 per year.  Job applicants have been considered for employment only if they score above the national average score of all employees taking the DIT.  This national average score is published yearly by Selection Systems, Inc. (SSI), the developer of the DIT, and is based on the test scores of more that 100,000 job applicants.  This year’s national average score was 22.  Test-retest reliability data are also available from SSI.  Some regular users of the DIT are asked to give the DIT to a sample of their employees twice, with the second administration following the first by one week.  Typical test-retest data reported by SSI indicate coefficients of stability of around .85.

 

In analyzing the firing patterns of the company, Paul Hawke, the new manager of human resources for EAPCO, noticed that a large number of black applicants failed the test.  Out of every hundred white applicants, eighty-five passed the DIT, and approximately fifty were hired for positions in the company.  In contrast, out of every hundred black applicants, thirty passed the DIT, and only ten were hired.  Hawke was concerned about the potential legal implications of this hiring system.  He found no evidence that systematic job analyses had been conducted on the unskilled production jobs or that EAPCO had validated the DIT against any measure of job performance.  Validity data on the DIT were available from SSI.  These data typically consisted of the results of concurrent validation studies conducted by other organizations using the DIT.  Hawke decided that it was time for an internal EAPCO validation study to be done.

 

Hawke contacted Validation Consulting, Inc. (VCI), a local firm, and hired it to conduct the study.  One week later, Steven Peabody-Reynolds, a partner at VCI, arrived at Hawke’s office.  Peabody-Reynolds spent the morning discussing the current selection system with various mangers and first-line supervisors.  He returned one week later and presented Hawke with a plan for validating the DIT.  The study was designed as a concurrent validation procedure.  Peabody-Reynolds recommended using just the Kentucky plant in the study to save money.  There were a hundred unskilled production workers at the Kentucky plant – ninety whites and ten blacks.  Most of the employees had been with the company for quite some time and had not taken the DIT since their initial hiring.  Fifteen of the employees in the proposed study (all white) had been at the Kentucky plant for eighteen years or more.  Hawke agreed to the plan, and one week later the study was conducted at the Kentucky plant.  All employees in the study were given the DIT (time 1).  One week later (to allow VCI to check the test-retest reliability of the DIT), the same employees were given the DIT again (time 2), and their supervisors were asked to rate their overall work performance using a 7-point scale where 1 = very unsatisfactory performer, 2 = unsatisfactory performer, 3 = satisfactory performer, 4 = good performer, 5 = very good performer, 6 = excellent performer, and 7 = truly exceptional performer.  Performance ratings and the DIT scores were returned to Peabody-Reynolds.  He conducted statistical analyses on the data and reported to Hawke that he had found a significant correlation (.85) between DIT (time 2) scores and performance ratings.  Hawke concluded that the DIT was related to job performance and was a valid test for use in screening employees to fill unskilled production jobs.

 

Six months after the study was completed, James Wilson, a black job applicant, took and failed the DIT.  He then filed a charge of racial discrimination against EAPCO.  The EEOC attempted to conciliate between Wilson and EAPCO, but the company refused to offer any sort of settlement that was acceptable to Wilson.  The EEOC decided to file suit against EAPCO in federal court.  Hiring data from EAPCO’s personnel files clearly indicate a prima facie case of discrimination.  However, the company is basing its case on the arguments that the use of the DIT is justifiable because DIT scores are related to job performance and that the use of the DIT has substantial utility.  With the around 1,000 applications for unskilled production jobs each year, EAPCO feels that the DIT is important, inexpensive ($20 per applicant) screening device in the hiring process.

 

Through various legal maneuvers, the case against EAPCO has been delayed several times.  During this period, Paul Hawke has resigned as human resource director and Sherri Downer, a local HR consultant with significant HR experience in a variety of industries, has been hired as Hawke’s replacement.  The CEO of EAPCO has indicated some concern as to whether Paul Hawke was handling the Wilson case in the best interests of the company.  The CEO has asked Ms. Downer to review the situation and make a recommendation as to whether EAPCO should seek to settle the discrimination case with Wilson.  EAPCO’s lawyers have indicated that they believe Wilson will accept a settlement of around $200,000.  The CEO asked Ms. Downer, in making her analysis of the situation, to consider all the important issues.  These include the following:

 

·         The reliability of the DIT

·         The validity of the DIT

·         The adequacy of the validation study conducted by VCI

·         The overall effectiveness of the current DIT-based selection system

 

The CEO wants Ms. Downer to make a recommendation, based on her analysis of these critical issues, as to how EAPCO should proceed in the case brought against them by Mr. Wilson.

 

Put yourself in the position of Ms. Downer.  Conduct your analysis and make a recommendation, supported by sound arguments, to the CEO.

 

 

Statistical Data

 

·         The correlation between DIT scores collected at Time 1 and those collected at Time 2 = .59

·         The correlation between DIT (Time 2) and performance ratings are as follows:

§         total sample = .31

§         black employees = -.14

§         white employees = .57

§         employees with 18 years experience or more = .85

·         The average DIT (time 2) score for each level of performance rating

§         Performance rating of 1: 21.5 (2 people)

§         Performance rating of 2: 21.4 (4 people)

§         Performance rating of 3: 22.9 (20 people)

§         Performance rating of 4: 24.1 (14 people)

§         Performance rating of 5: 24.4 (27 people)

§         Performance rating of 6: 25.0 (29 people)

§         Performance rating of 7: 30.3 (3 people)

·         EAPCO’s typical selection ratio:

§         To estimate the selection ration, the following data are pertinent:

-          Typical turnover rate = 35% of 1000 employees = 350 job openings/year

-          Approximately 1000 job applicants/year

-          Selection ratio = openings/applicants = 350/1000 = .35

·         An estimate of the base rate of success for job applicants:

§         To estimate the base rate of success, the only data you have available are present performance ratings on 100 people.  Given the rating scale used, ratings of 1 and 2 represent some form of unsatisfactory performance, while ratings of 3 or above indicate some level of satisfactory performance.  A frequency distribution of performance scores indicates that only 7 people in the sample were rated as unsatisfactory.  Thus, a crude estimate of the BRS would be 93 percent, since 93 percent of the sample (who had been selected on the basis of a testing procedure – the DIT – which correlates .31 with performance) seem to be able to produce a satisfactory level of performance on their jobs.