Connell focuses on the role that deliberation has on the juror's perception of group functioning, measured here by the construct of group climate. Her results suggest individual juror characteristics do not have a direct effect on sentencing outcomes; rather, the level of group climate acts as a mediating variable between individual characteristics and outcomes. Trial level characteristics directly predict sentencing and indirectly operate through the level of group climate. Group climate is the strongest predictor of outcomes, with juries who have more positive perceptions of group climate more likely to return the death penalty.