Computers have become an integral part of modern life. No longer are they mysterious, giant electronic machines that are stuck away in some remote site at a university or government facility requiring a bunch of engineers with PhDs to operate. Computers are everywhere—doing tasks that were once considered to be sheer human drudgery (managing vast unthinkable inventories with lightening speed), happily managing chores that no one could accomplish (like monitoring intricate internal engine functions), or depositing a letter to a friend all the way around the world in microseconds, a task that used to take months.
Computers have served in several capacities in the field of psychological assessment since their introduction almost a half
I would like to express my appreciation to Reneau Kennedy for providing case material used in this chapter.
century ago, although initially only in the processing of psychological test information. Over the past several decades, their uses in mental health care settings have broadened, and computers have become important and necessary aids to assessment. The benefits of computers to the field of psychology continue to expand as technology becomes more advanced, allowing for more sophisticated operations, including integrative test interpretation, which once was the sole domain of humans. How can an electronic and nonintuitive gadget perform a complex cognitive process such as psychological test interpretation (which requires extensive knowledge, experience, and a modicum of intuition)?
The theoretical rationale underlying computer-based test interpretation was provided in 1954 when Meehl published a monograph in which he debated the merits of actuarial or statistical (objective) decision-making methods versus more subjective or clinical strategies. Meehl's analysis of the relative strengths of actuarial prediction over clinical judgment led to the conclusion that decisions based upon objectively applied interpretive rules were ultimately more valid than judgments based on subjective strategies. Subsequently, Dawes, Faust, and Meehl (1989) and Grove and Meehl (1996) have reaffirmed the finding that objective assessment procedures are equal or superior to subjective methods. More recently, in a meta-analysis of 136 studies, Grove, Zald, Lebow, Smith, and Nelson (2000) concluded that the advantage in accuracy for statistical prediction over clinical prediction was approximately 10%.
In spite of the common foundations and comparable rationales that actuarial assessment and computerized assessment share, they are not strictly the same. Computer-based test interpretation (CBTI) can be either clinical or actuarial in foundation. It is an actuarial task only if its interpretive output is determined strictly by statistical rules that have been demonstrated empirically to exist between the input and the output data. A computer-based system for describing or predicting events that are not actuarial in nature might base its interpretations on the work of a clinician (or even an astrologer) who hypothesizes relationships using theory, practical experience, or even lunar phases and astrology charts.
It is important in the field of psychological assessment that the validity of computerized assessment instruments be demonstrated if they are to be relied upon for making crucial dispositions or decisions that can affect people. In 1984 the Committee on Professional Standards of theAmerican Psychological Association (APA) cautioned psychologists who used interpretive reports in business and school settings against using computer-derived narrative test summaries in the absence of adequate data to validate their accuracy.
Was this article helpful?