PAC 302 Module 1 Foundations of Psychological Assessment Example

One Interview Is Not an Assessment: Testing, Clinical Judgment and Actuarial Prediction at an Outpatient Intake

Student Name

Psychology and Addiction Studies Program, Aspen University

PAC 302: Assessment Procedures in Addiction Studies

Instructor Name

Month Day, Year

One Interview Is Not an Assessment: Testing, Clinical Judgment and Actuarial Prediction at an Outpatient Intake

Lakeview Counseling and Recovery, an outpatient agency in Grand Rapids, Michigan, invented for this course, admits about forty new clients a month. Each new client meets whichever counselor is free for an hour-long unstructured interview. The counselor then writes a narrative, assigns a diagnosis and recommends a level of care, such as standard outpatient or intensive outpatient. Counselors are experienced and caring, but supervisors have noticed that two counselors interviewing similar clients often reach different conclusions. Last month, one client was placed in standard outpatient care and was hospitalized for alcohol withdrawal a week later; another, with a similar history, was sent to intensive care she did not need and stopped coming. This paper examines the foundations of assessment and what they suggest for Lakeview's intake. The agency and its clients are composites, created for teaching; the research is real.

Testing and Assessment

Testing and assessment are often used interchangeably, but they are different activities. Testing is the administration of a standardized instrument that produces a score, such as a screening questionnaire for alcohol problems. Assessment is a broader process in which a clinician gathers information from multiple sources, including tests, interviews, history, observation and reports from others, and integrates it to answer a question, such as what diagnosis fits or what level of care a client needs. A test provides data; an assessment interprets data in context. Lakeview's intake is an assessment that uses only one source of information, an unstructured interview.

How Valid Are Psychological Tests?

Meyer et al. (2001) reviewed meta-analyses of the validity of psychological tests and compared them with meta-analyses of medical tests. They found that the validity of many psychological tests was comparable to that of many widely accepted medical tests, such as those used to detect disease. Psychological testing, in other words, was not less scientific than much of medicine. A second finding was equally important. When different methods were used to assess the same characteristic, such as a client's self-report, a clinician's rating and an observer's report, the methods often agreed only modestly. The authors argued that this was not a flaw but a reason to use multiple methods, because each captures different aspects of a person. An assessment based on a single method, such as an interview alone, systematically misses information that other methods would provide.

Clinical and Actuarial Judgment

Dawes et al. (1989) contrasted two ways of combining information to make predictions and decisions. In clinical judgment, a professional weighs the information in their head and reaches a conclusion. In actuarial, or mechanical, judgment, information is combined by an explicit rule, such as a statistical formula or a scoring system, derived from data about outcomes. The authors reviewed comparisons and found that actuarial methods equaled or exceeded clinical judgment in virtually every study, even when clinicians had access to more information. They attributed the advantage to clinicians' susceptibility to inconsistency, to giving too much weight to vivid information and to difficulty learning from feedback.

Grove et al. (2000) quantified this in a meta-analysis of studies comparing clinical and mechanical prediction across fields, including psychology and medicine. Mechanical prediction was, on average, about ten percent more accurate. It was substantially better in a large share of studies, the two approaches were about equal in roughly half and clinical prediction was better in a small minority. Giving clinicians interview data did not close the gap. The authors concluded that mechanical methods should be used where they are available and validated.

What This Means for Lakeview's Intake

The two cases from last month illustrate the risk of the current approach. The client hospitalized for withdrawal had reported drinking a fifth a day and two previous withdrawal seizures, information that a structured withdrawal history would have flagged for a higher level of care. The client sent to intensive treatment she did not need had described a crisis vividly, the kind of vivid information Dawes and colleagues found can mislead clinical judgment.

Where Clinical Judgment Still Belongs

The research does not suggest removing counselors from assessment. It suggests using structured tools where they are validated and reserving judgment for what tools cannot do. Counselors are needed to build rapport, so that clients answer screening questions honestly; to recognize unusual situations that a rule does not cover, such as a client whose cultural background makes an item's meaning different; to integrate results that conflict; and to explain findings to clients in ways they can use. The proposed intake places the counselor at the center of interpretation while giving that interpretation a more reliable foundation.

Fairness and Consistency

Consistency is also a fairness issue. Under the current intake, the same history can lead to different levels of care depending on which counselor is free, which means clients receive different decisions for reasons unrelated to their needs. Structured instruments and explicit criteria apply the same questions and standards to every client. They are not free of bias, and the agency would need to check that its screens perform well for the populations it serves, but they make the basis of each decision visible and open to review in a way that an unstructured interview does not.

Implementation and Costs

The change requires training counselors in the structured interview, adopting free validated screens and building the level-of-care criteria into the electronic record. It adds about fifteen minutes per intake. Lakeview would review the change after six months by comparing agreement between counselors, early dropout and unplanned hospitalizations before and after.

Conclusion

Lakeview's single unstructured interview is an assessment built on one method, and the research suggests it misses information and varies between counselors. Meyer and colleagues show that psychological tests can be as valid as medical tests and that multiple methods capture more, while Dawes, Faust and Meehl and Grove and colleagues show that structured, actuarial approaches modestly but consistently outperform unaided judgment. A multimethod intake, with counselors interpreting structured findings, would give Lakeview's clients more consistent and safer decisions.

References

Dawes, R. M., Faust, D., & Meehl, P. E. (1989). Clinical versus actuarial judgment. Science, 243(4899), 1668-1674. https://doi.org/10.1126/science.2648573

Grove, W. M., Zald, D. H., Lebow, B. S., Snitz, B. E., & Nelson, C. (2000). Clinical versus mechanical prediction: A meta-analysis. Psychological Assessment, 12(1), 19-30. https://doi.org/10.1037/1040-3590.12.1.19

Meyer, G. J., Finn, S. E., Eyde, L. D., Kay, G. G., Moreland, K. L., Dies, R. R., Eisman, E. J., Kubiszyn, T. W., & Reed, G. M. (2001). Psychological testing and psychological assessment: A review of evidence and issues. American Psychologist, 56(2), 128-165. https://doi.org/10.1037/0003-066X.56.2.128