In 1999, Texas enacted legislation allowing for the civil commitment of violent sexual offenders after their release from prison. Currently, 20 states have enacted such laws, and the United States government also has passed legislation allowing for the civil commitment of sexually violent predators at the federal level. In these hearings, the state is responsible for showing that the offender possesses a "behavior abnormality" making them likely to engage in sexually violent predatory behavior following their release. An evaluation is conducted by a mental health professional (a psychiatrist or a doctoral level psychologist) to determine the presence of this abnormality, and, in most cases, a hearing is conducted in front of a jury in which the expert witnesses presents his or her findings. On some occasions, the defense also hires an expert witness to offering opposing testimony.

In Texas, all of the Sexually Violent Predator (SVP) civil commitment hearings occur in the Montgomery County Courthouse. Researchers in this study were granted access to the jurors immediately following each hearing over a two-year period. Two questionnaires were administered to all jurors who volunteered to participate in the anonymous study. One questionnaire addressed specific factors of each hearing (details of the offenses, offender characteristics, expert witness testimony, victim characteristics, influence of any character witnesses, etc.). Jurors rated each of these 15 factors on the degree of importance each carried and the level of influence each variable had on their decisions as to whether the respondent had a behavior abnormality. The second questionnaire addressed each expert testifying in the hearing and required the jurors to rate the effectiveness of the expert in terms of his or her testimony. Also, a credibility scale was created in order to allow for an additional rating of "credibility" for each expert.

The study shows that jurors are most influenced by factors that have been empirically shown to be poor predictors of likelihood of sexual reoffending and sexually violent reoffending. This is the case even when the research supporting or refuting each factor is explained to the jurors. Jurors are conversely least influenced by factors which have been empirically shown to be the best predictors of recidivism (e.g. actuarial scores such as the Static-99, MnSost-R, and tests such as the Psychopathy Checklist Revised – PCL-R).

In addition, results show that jurors rate state experts as more influential and credible than expert witnesses for the respondent, but that the presence of an opposing witness - either expert or lay person - significantly reduces the ratings of credibility and influence of the respondent expert witnesses.

All of the hearings in this study ended in civil commitment save for one which ended in a hung jury. Future studies with mock jurors may control for the manner in which testimony is presented to jurors. In addition, future studies with actual jurors may be able to provide for more variability in trial outcome.

KEY WORDS: Sexually violent predator, juror decisions, risk assessment, risk communication to jurors