Under the right conditions, pre-employment tests can vastly improve corporate productivity. This paper discusses the conditions, expected productivity gains, and issues associated with pre-employment testing. Special attention is paid to issues of bias, legal rulings, validity generalization, exaggerated expectations, test quality, misuse of tests, publishers' claims, alternative assessment techniques, and the use of honesty tests.
The effectiveness of pre-employment testing is a function of three factors' (1) the correlation between test scores and job productivity; (2) the percentage of applicants being hired; and (3) the proportion of applicants which are classified as successful by a test. These factors are reviewed and major testing issues -- test bias, legal issues, validity generalization, exaggerated expectations, quality of the test, the responsibilities of test users, publishers' claims about testing, alternative assessment techniques, and the ethical issues of honesty tests -- are discussed. Recommendations are made to industry regarding testing and documentation practices.
Although frequently attacked as invalid, demeaning, biased, illegal, and irrelevant; pre-employment testing procedures appear to be increasing in popularity. The American Society for Personnel Administration found 39% of 360 companies surveyed were testing more in 1985 than in 1980, and 44% were considering even more testing. A 1988survey of 245 human resource executives by the Bureau of National Affairs (a publisher) found that 63% of surveyed companies ask applicants to supply work samples or take performance tests, while 30% require ability tests, and 25% test for job knowledge. A new referral system being considered by the U.S. Employment Service, a part of the Department of Labor, could result in the testing of several million applicants annually.
Many prominent companies and organizations are making extensive use of tests. The Illinois Department of Employment Security used a written multiple choice test to screen 50,000 blue collar applicants at Diamond Star. At American Telephone and Telegraph, testing is a routine part of hiring and promotion through the second layer of management. International Business Machines uses skill and aptitude tests to evaluate applicants for about 75% of entry level jobs. Manpower expects to test over 700,000 applicants this year. Corporate executives, state officials, and federal policymakers are discovering that the judicious use of formal assessment procedures may lead to increased efficiency and productivity. The benefits of testing appear to outweigh its costs and concerns.
Part of the resurgence of testing is attributable to clearer definitions of acceptable practice. The landmark case of Griggs v. Duke Power (1971) results in a legal precedent requiring defendants to demonstrate adequate validity. In 1978, the Equal Employment Opportunity Commission established "Uniform Guidelines on Employee Selection Procedures." In 1974 and again in 1986, the American Psychological Association, the National Council on Measurement in Education, and the American Educational Research Association adopted professional standards for educational and psychological tests. And in 1987, the Society for Industrial and Organizational Psychology issued the third edition of its own principles for the validation and use of personnel selection procedures.
Legal precedents and federal and professional guidelines help both the test developer and its users. The developer can conduct appropriate studies and prepare necessary documents. When assured that tests meet legal and professional standards, potential customers can use them with greater confidence.
Under the right conditions, pre-employment testing can vastly improve corporate productivity. But, testing is marked with real issues that employers are often ill-equipped to handle. What does an employer do about Black applicants who, on average, score lower than whites on standardized tests? …