How to fail in the implementation of a risk assessment scheme or any other new procedure in your organization.

F orty years ago in this journal, Jay Haley wrote an article entitled ‘‘How to Fail as a Psychotherapist.’’ In his article, he outlined the ‘‘daily dozen’’ of what could be construed as sub-optimal psychotherapy practices (e.g., ‘‘Insist that the problem which brought the patient into therapy is not important’’; ‘‘Insist that only years of therapy will really change a patient’’; ‘‘Avoid the poor because they will insist upon results and cannot be distracted with insightful conversations’’; ‘‘Avoid evaluating the results of therapy’’). The true purpose of Haley’s article was, of course, to show that much knowledge exists about strategies for effective, ethical psychotherapy. More than a decade ago, Christopher Webster emulated Haley in a book chapter describing how to fail as an assessor of risk of violence (‘‘The Art of Being a Failure as an Assessor: Twenty Suggestions’’). Analogously, the real purpose was to create a framework for the conduct of assessments of risk of violence that would meet or exceed general professional standards. For example, Webster satirically exhorted readers to avoid clarifying the purpose of the evaluation, using a systematic approach to assessment, and obtaining outcome data. At the time when Webster’s chapter appeared, researchers were beginning to explore the predictability of institutional and community violence. Scholars intended, in part, to test earlier views that clinicians have very limited capacity for fulfillment of this task. The publication of a variety of risk assessment schemes helped subsequent researchers substantially. Although not originally intended as predictive devices, some of these instruments (e.g., the Hare Psychopathy Checklist Revised [PCL-R]) appeared to have potential as tools for forecasting violence. Although none of these many schemes have yielded truly impressive predictive power, most have performed better than would have been expected 30 or 40 years ago. Psychometric differences among contemporary instruments tend to be small—unsurprisingly, given that item content tends to overlap considerably. As Randy Otto and Kevin Douglas have shown, the new challenge is not to find instruments with acceptable predictive power but instead to ensure fidelity of application. Design of a risk assessment device may be easier than ensuring its true-topurpose application in forensic, civil mental health, and correctional settings. Because such instruments have a proven, albeit imperfect ability to separate the patients who present risks from those whom politicians can safely ignore, there will always be a market for a good implementation

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.