Date of this Version
From: The Computer and the Decision-Making Process, edited by Terry B. Gutkin and Steven L. Wise (Hillsdale, New Jersey, Hove & London: Lawrence Erlbaum Associates, 1991) .
A decade ago a scholar writing in a legal journal asked the question, "Can/Should Computers Replace Judges?" (D'Amato, 1977). The article explored problems involved in developing computer systems capable of making the difficult assessments and judgments required in judicial decision making. In discussing these problems, the author quoted extensively from Joseph Weizenbaum, who in a well-known critique of computerized psychotherapy, sagely asserted, "Since we do not now have any ways of making computers wise, we ought not now to give computers tasks that demand wisdom" (Weizenbaum, 1976). Nevertheless, the legal scholar concluded that any humanistic misgivings about computerized decision making are, at least for many kinds of functions performed by judges, outweighed by the considerable savings in time and money the new expert systems can provide.
If this volume had been published a decade earlier, we might have raised a comparable question: Can/should computers replace psychologists in the administration and interpretation of psychological tests? But that question is now moot. Computers already have replaced psychologists in many routine aspects of assessment. Computerized psychological testing (CPT) is making significant inroads in educational evaluation, personnel selection, occupational counseling, and mental health diagnosis. There is little doubt that computers will generate new methods of assessment in the foreseeable future.
Yet the question of whether CPT should replace psychologists has only recently received the attention given the question of how CPT might do so. Coincidental with the rise of computer-testing technology is the countervailing trend toward greater scrutiny of test use, particularly in employment and educational settings (Bersoff, 1983). We must carefully examine CPT to ensure that it does not unnecessarily create any new legal problems for testing, and in fact contributes to a high level of scientific and ethical merit in psychological testing practice.
As we have indicated elsewhere (Hofer & Bersoff, 1983), computerized tests may be vulnerable to many of the same legal attacks as conventional tests. Claims of cultural bias and other forms of unfairness are the predominant source of litigation involving tests, and such claims are likely to continue with any test showing disproportionate adverse impact on minorities or women, regardless of method of administration or interpretation. Although some types of litigation may become less likely by the switch to computers, especially challenges to the standardization and procedural regularity of the administration of the test itself, CPT could conceivably lead to new legal problems for developers and practitioners. A leading editorial in Science predicted a "flood of litigation involving unqualified users" of computerized tests (Matarazzo, 1983, p. 323).
THE LEGAL PROFESSION'S RESPONSE TO CPT
To this point, it is not so much a flood as a trickle. There is, to date, only one reported case even tangentially involving unqualified use of CPT that we have discovered (United States v. Curtis, 1974) and that case, while having its own intrinsic interest, is irrelevant to our concerns. The defendant advertised a "Computer Matching Institute" dating service, where couples were to be paired through testing by qualified psychologists and prompt computer processing. In fact, the defendant did not have the intent or capacity to match applications by computer or expert psychological testing, and simply hired clerks to match applications by hand. The court found a clear basis for a criminal indictment for fraud.
There is now one reported case directly concerned with CPT which is germane to those mental health professionals who purchase software for scoring and interpreting psychological tests. We discuss that case at some length in the section on intellectual property, which appears later in this chapter. Aside from that, the most interesting treatment of some of the legal issues raised by CPT is found in two advisory opinions written by state attorneys general.
The attorney general of Georgia (Unofficial Opinion, 1983) was asked by a judge of a county juvenile court if the interpretation of psychological tests administered to juveniles might be computerized. Apparently the judge was sufficiently concerned and unsure of the implications of CPT that an outside legal opinion was sought. The attorney general found no legal barriers to computerizing the testing process , so long as adequate steps were taken to protect the confidentiality of juvenile records, in this instance, by disguising the names of examinees so that no identifying information appeared in the computerized records. The replacement of names with identification codes before entry into electronic memory is common practice among testing companies and, along with safeguards required for all clinical material, should protect the confidentiality rights of clients. The opinion, however, does raise the concern that CPT might infringe unduly on the fundamental right to be protected against governmental "disclosure of personal matters" (Whalen v. Roe, 1977, p. 599).
In Kansas, the state board charged with licensing and regulating psychologists requested an opinion on several issues raised by CPT. One question is of great interest to many clinicians-whether CPT may be used by professions other than psychology. The Kansas attorney general (Attorney General Opinion, 1983), interpreting that state's laws, found nothing to prevent use of CPT by others if such use was consistent with their training and with their profession's code of ethics, and if they did not hold themselves or their work out to the public as "psychology" or "psychological." This issue is likely to be a continuing source of concern, and resolution may vary from state to state. For the most part, test developers and marketers have refrained voluntarily from providing clinical tests to nonpsychologists, but some CPT services have been less circumspect. A thoughtful analysis by state legislatures and professional organizations, such as the American Psychological Association (which has been studying the general problem of test user qualifications), of the responsibilities of CPT developers and users is required to protect the interests of the public.