Federal Court May Still Allow "Junk Statistics"

APA Monitor, September, 1998, page 50.

John L. Fleer, Ph.D., J.D., Fleer, Daugherty and Loftis, and Martin H. Williams, Ph.D, Independent Practice.

Although the Federal Courts have attempted to restrict and eliminate the testimony of so-called "junk science" experts, expert testimony based on what might be called "junk statistics" may continue to be admissible. Such testimony has recently been endorsed by the Ninth Circuit Court of Appeals. In United States v. Bighead, 131 F.3d 149; 128 F.3d 1329 (1997), the Court upheld the admissibility of expert statistical opinion in the absence of any minimal scientific foundation establishing the validity of that opinion.

In Bighead, the defendant was accused of sexually molesting his daughter. His daughter, now an adult, charged that her father had molested her beginning at age seven. Sexual contact allegedly continued until 1993 when, at the age of 17, she rejected her father's overtures. The daughter's testimony was problematic for the prosecution because there had been a significant delay in reporting the alleged assaults and because she was unable accurately to recall details of the alleged sexual activities.

To bolster the daughter's inconsistent testimony, the prosecution called a sexual abuse counselor who testified that victims of sexual abuse commonly delay reporting abuse and blend events together in their memories. As Appellate Judge Noonan stated in a dissenting opinion, "this expert's testimony 'marvelously imparted strength' to the daughter's otherwise suspect testimony. Profound doubts about the truth about the alleged victims story were transformed into evidence of the truth of the accusations." In effect, the expert persuaded the jury that the daughter's inconsistent testimony was itself evidence that she was a sexual abuse survivor. Inherent lack of credibility was transformed into prima facie evidence that the daughter's allegations were truthful.

The expert witness testified about her doctoral research which was "regarding statements from children who had been sexually abused." She had been the director of a research project "regarding child abuse on Indian reservations throughout the United States," Between 1987 and 1993, she had "interviewed 1200 children who had alleged sexual abuse."

On the basis of her experience, she was asked whether she had "formed or made observations with respect to general aspects of reporting incidents of child sexual abuse." She testified that "the most common circumstance with respect to reporting of sexual abuse is what is referred to as delayed disclosure." She also testified that "child sexual abuse is often chronic, is very rarely an isolated one time event" and that "the teenager develops something called a script memory. What that means is that the events, over many, many years, blend together. And so in the recounting, or the telling of that event, you are going to get little bits and pieces of information, details that come to the person differently each time they recount the event."

Based upon her interview experience, the expert concluded that the probability of delay in reporting the assault, and the probability of confusing different events, were higher in victims of sexual abuse than in victims of other types of crimes. Neither jurors nor the prevailing judges were able to recognize that the expert's conclusions failed to meet minimal scientific standards for carrying out such a survey and drawing these conclusions. The dissertation research in question had failed to distinguish between individuals making true or false claims of sexual abuse, and no effort had been made to compare systematically the reporting styles of sexual abuse victims, non-victims, or victims of other kinds of crimes.

In designing this type of study, one would conservatively assume that any sample of self-identified sexual abuse survivors might contain both true and false claimants, each respectively comprising an unknown portion of the sample ranging from 0 to 100%. Because true or false allegations of sexual abuse might have been present in the sample, conclusions about the reporting styles of real rather than merely purported victims could not reasonably have been drawn. Indeed, in terms of scientific inference, this study of the reporting styles of those making unsubstantiated claims of sexual abuse should have had no bearing on the prosecution's case.

The defendant was convicted, and the court of appeals affirmed the conviction and endorsed the prosecution's use of expert testimony based upon "professional experience." This result is of great interest to those of us who may have believed that, after Daubert, the courts had uniformly adopted stricter standards which would now prevent unwarranted, pseudo-scientific conclusions from being introduced in court. Because jurors and judges sometimes lack sufficient scientific sophistication to distinguish between valid and grossly invalid research conclusions, expert witnesses continue to exert a prejudicial effect in the courtroom in the guise of the scientific method.