Assessing the Sensitivity of Early Head Start Intervention Findings to Manipulated Randomization Threats

• In Event: Roundtable Session 15
  In Roundtable Session: Exploring the Methodological Implications of Conducting Research on Evaluation

Abstract

This study systematically examined how the magnitude and type of typical randomization problems affected applied research and evaluation results. Secondary data from a randomized, national early childhood study were used to examine the manipulated effects of threats to randomization on selected child developmental outcomes. The goal was to increase understanding of the differential effects of randomization disruption and the potential utility of subsequent post-randomization statistical adjustment techniques in ameliorating such effects.

Authors

• Sheridan Green, Clayton Early Learning

• Susan Hutchinson, University of Northern Colorado