Group Size and Measure of Group-Level Properties : A Different Perspective

Group Size and Measure of Group-Level Properties : A Different Perspective

  Bliese and Halverson (1998) argued that a corrected eta-squared rather than eta-squared should be used for analyzing group level effects. However, the proposed corrected eta-squared approach has the following limitations: (1) it fails to specify a criterion for justifying the aggregation of individual-level data to the group level and (2) it fails to specify a criterion for analyzing and allowing for within-group effects. In this paper, we suggest several alternatives that go beyond these limitations. We also suggest an additional procedure for justifying the aggregation of individual-level data to the group level.