An intervention designed to correct affective and cognitive biases was tested in the context of a civil commitment hearing of a sexually violent predator. Potential differences between a college student mock jury sample and a more representative, juror venire sample in reaction to these bias correction interventions were explored. In the first of two experiments, undergraduate mock jurors (n = 130) demonstrated a leniency effect when the sex offender's attorney acknowledged jurors' emotional reactions and motivated them to thoughtfully weigh the evidence. The second experiment failed to replicate these findings with a more ecologically valid sample (n = 300). Several differences between samples were found: representative jurors, as opposed to undergraduates, were sensitive to differences between pure clinical and actuarial expert testimony; and measures of intrinsic cognitive effort predicted verdicts for undergraduates, but not for representative jurors. Copyright © 2011 John Wiley & Sons, Ltd.