Researchers are tackling a major problem of group randomized trials: likely positive intraclass correlation among observations of individuals within the group. They have developed an index of extra variability to understand the influence of groups on evaluating intervention effect. Paper