Monday, October 24, 2005

In defense of psychological testing

Another "blast from the past" article that I'd like to share, as one who has spent much of his career developing, administering, interpreting, or investigating psychological tests (esp. intelligence tests). All emphasis added by this blog's (benevolent) dictator :)

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.

In 1996 the American Psychological Association's (APA) Board of Professional Affairs (BPA) established a Psychological Assessment Work Group (PAWG) to (a) evaluate contemporary threats to psychological and neuropsychological assessment services and (b) assemble evidence on the efficacy of assessment in clinical practice. Based on the review of this group, it was concluded that “there substantial evidence to support psychological testing and assessment.”
The group put the strength of psychological testing correlations into perspective via comparisions with other correlations (e.g., in medicine) that have resulted in signficant policy and health decisions. A few examples from the paper:
  • Taking aspirin on a regular basis helps to reduce the risk of dying from a heart attack (r = 0.02)
  • Impact of chemotherapy on breast cancer survival (r = .03)
  • The association between a major league baseball player's batting average and his success in obtaining a hit in a particular instance at bat (r = .06)
  • The value of antihistamines for reducing sneezes and a runny nose (r = .11)
  • The impact of Viagra on improved sexual functioning (r = .38)
The APA PAWG group further concluded:
  • In many respects, these findings highlight how challenging it is to consistently achieve uncorrected univariate correlations that are much above .30.
  • For those who may be inclined to square the values and feel discouraged, we recommend an alternative, which is to re-conceptualize effect size magnitudes.
  • Instead of relying on unrealistic benchmarks to evaluate the findings…it seems that psychologists studying highly complex human behavior should be rather satisfied when they can identify replicated univariate correlations among independently measured constructs that of the magnitude observed for antihistamine effectiveness (r = .11: Table 1, Entry 16), college grades and job performance (r = 16; Table 1, Entry 24), or criminal History and recidivism (r= .18; Table 1, Entry 29). Furthermore, it appears that psychologists generally should be pleased when they can attain replicated univariate correlations among independently measured constructs that approximate the magnitude seen for gender and weight (r = .26; Table 1, Entry 39), reliability and validity (r = .33; Table 1, Entry 47), or elevation above sea level and daily temperature (r = .34; Table 1, Entry 48). Finally, psychologists probably should rejoice when they find replicated evidence that uncorrected univariate correlations are of the same magnitude as those observed for gender and arm strength (r = .55. Table 1, Entry 58) or for latitude and and daily temperature (r = .60; Table 1, Entry 59).
  • The validity coefficients found for psychological tests frequently exceed the coefficients found for many of the medical and psychological interventions. Taken together, the extensive array of findings offers compelling support for the value of psychological testing and assessment
More importantly, the APA PAWG group stated (emphasis added by blogmaster)
  • Although psychological tests can assist clinicians with case formulation and treatment recommendations, they are only tools. Tests do not think for themselves, nor do they directly communicate with patients. Like a stethoscope, a blood pressure gauge, or an MRI scan, a psychological test is a dumb tool, and the worth of the tool cannot be separated from the sophistication of the clinician who draws inferences from it and then communicates with patients and other professionals
  • Formal assessment is a vital element in psychology's professional heritage and a central part of professional practice today. This review has documented the very strong and positive evidence that already exists on the value of psychological testing and assessment for clinical practice. We have demonstrated that the validity of psychological tests is comparable to the validity of medical tests and indicated that differential limits on reimbursement for psychological medical tests cannot be justified on the basis of the evidence.

No comments: