ERIC Number: ED518894
Record Type: Non-Journal
Publication Date: 2011
Reference Count: 26
Power Analysis for Anticipated Non-Response in Randomized Block Designs
Pustejovsky, James E.
Society for Research on Educational Effectiveness
Recent guidance on the treatment of missing data in experiments advocates the use of sensitivity analysis and worst-case bounds analysis for addressing non-ignorable missing data mechanisms; moreover, plans for the analysis of missing data should be specified prior to data collection (Puma et al., 2009). While these authors recommend only that missing data should be anticipated in the analysis plan of an experiment, it is logically consistent that anticipation of missing data should be incorporated into another important step in the planning process for a randomized control trial (RCT): power analysis. The paper develops power analysis formulas for the worst-case lower bound and sensitivity lower bound and compares power for these estimands to the power to detect treatment effects assuming missingness at random (MAR). The methods described provide a valid and principled approach for researchers to study the consequences of anticipated levels of missing data for planned experimental designs. The paper also illustrates how these methods can be used for cost-benefit analysis, weighing measurement approaches that are less expensive but yield lower response rates against approaches that are more expensive but yield higher response rates.
Descriptors: Statistical Analysis, Data, Sample Size, Response Rates (Questionnaires), Research Design
Society for Research on Educational Effectiveness. 2040 Sheridan Road, Evanston, IL 60208. Tel: 202-495-0920; Fax: 202-640-4401; e-mail: email@example.com; Web site: http://www.sree.org
Publication Type: Reports - Research
Education Level: N/A
Authoring Institution: Society for Research on Educational Effectiveness (SREE)