| (March Newsletter 2003 - 3) True though that may be, the first question that a finder of fact should ask is: �How good is an actuarial approach�? Just because it is better than clinical assessment (estimated to be wrong two out of three times) does not mean that it is accurate enough to be clear and convincing or beyond a reasonable doubt. (Id.) The highest estimate of accuracy is about 80%. (Id. at 4.) This is not bad for behavioral science, but is it high enough to use to deprive someone of their constitutional rights? (Liberty Interest) Improvements to psychological testing What is clear is that forensic psychology and psychiatry are, in response to pressure from many areas, including the forensic, making clear strides in research efforts that are improving the predictability of psycho-logical testing. The question for the advocate is whether the predictive value is currently sufficient. When the best psychological prediction is in the neighborhood of 80%, the jurist must make the argument as to whether this is sufficient to the standard of �beyond a reasonable doubt� or even the �clear and convincing� standard of proof. Standards must be met Psychological and psychiatric evaluations that do not meet current scientific standards are excludable. (See Daubert, supra; Kumho, supra. See also Frye v. U.S., 293 Fed. 1013 (D.C. Cir. 1923).) In general, there are two measures against which a psychological test will be measured: 1. Validity; and 2. Reliability. In addition, a test�s acceptance by the scientific community will be affected by its ability to be applied generally to different populations and by its likelihood of producing false positive negatives. All of this data must be produced or the test will not meet the Daubert standard and can be excluded. Editor�s Note: Daubert established that �general acceptance� is not a necessary precondition to the admissi-bility of scientific evidence under the Federal Rules of Evidence. Judges must make preliminary assessments about whether the testimony�s underlying reasoning or methodology is scientifically valid and properly can be applied to the facts at issue. Many considerations will bear on the inquiry, including whether the theory or technique in question can be -- and has been -- tested, whether it has been subject to peer review and publica-tion, its known or potential error rate, the existence and maintenance of standards controlling its operation, and whether it has attracted widespread acceptance within a relevant community. The inquiry is a flexible one, and its focus must be solely on principles and methodology, not on the conclusions they generate. Throughout, the judge should also be mindful of other applicable rules. However, under the Federal Rules of Evidence, trial judges have the task of ensuring that an expert�s testi-mony rests both on a reliable foundation and is relevant to the task at hand. Pertinent evidence based on scien-tifically valid principles will satisfy those demands. The inquiries of the District Court and the Court of Appeals focus almost exclusively on �general accept-ance� as gauged by publication and the decisions of other courts. Validity Standard Validity refers to the degree that a test measures what it purports to measure. The most basic is �false validity�. A test has face validity if it appears to measure what it is supposed to measure. A column of addition and multiplication problems would have face validity as a test for math skill. Face validity can be deceiving. Whereas the example of the arithmetic problems probably has a correlation between the test and skill tested, if the underlying assumptions are erroneous, even face validity will not rescue the test from worth-lessness. On a simple level, consider astrology. Predictions can (and have been) made based on scientifically unsupported assumptions about celestial events and behavior. On a more immediate level, one can question the true validity of psychological prediction from �meta-analyses� if improperly conceived. If one of the defini-tions in the meta-analyses is violence, but the definition of violence in the individual studies vary widely, that analysis is flawed, even though it has apparent or face validity. Validity measured against known test The generally accepted method for testing validity is to measure the instrument, test, or procedure against a known test that has highly accepted validity. The Wechsler intelligence test, therefore, were measured against Binet tests which were older and generally accepted as good measures of that construct. Validity studies should be available for any given instrument, test, or procedure in order to meet the Daubert standard. Test must distinguish between groups An instrument, test, or procedure will have questionable validity if it fails to distinguish between groups. If the number of false positives or false negatives is too high, the test has little or no validity since it cannot dis-tinguish between groups. (Author�s Note: False positives are individuals who are predicted to behave in the manner the procedure is designed to detect but in fact do not. False negatives are individuals who are not pre-dicted to behave in the manner the procedure is designed to detect but in fact do.) In Kansas v. Hendricks, 541 U.S. 346 (1997), the court found that the commitment scheme of Kansas was constitutional because it only confined the most egregious cases. If no psychological instrument, test, or procedure can distinguish between the most egregious cases and cases that would not meet the Hendricks standard, then there is no way to determine who should not be committed. (Editor�s Note: In reality, the Kansas standard is a �False Positive� given the fact that not all that are committed under Minnesota�s Civil Commitment Scheme act as predicted. In addition, this standard not only affects the most egregious cases, it civilly commits those that do not meet the Hendricks standard.) Reliability Standard Reliability means that the test consistently measures what it purports to measure if given multiple times. A test is reliable if given to the same individual on different occasions and produces similar results unless the con-struct being measured is changeable over time. A person�s weight will vary over time and a scale is not unre-liable because it produces different results if sufficient time passes. If a trait is not changeable over time, then a reliable test will produce the same result if given to the same individual multiple times. Thus, a tape measure should produce the same height on an adult when measuring the person multiple times. |
|