Power Analysis

Power analysis helps estimate the required sample size for a study.

Most of the time I conduct a power analysis for RCT designs or QED designs, using the PowerUP! software.  The program is an Excel file downloadable from:

http://teams.mspnet.org/index.cfm/webinars/webinar_info?id=500

Posing with PowerUP! author, Dr. Dong.  Nov. 9th 2017 at American Evaluation Association's annual meeting (Washington, DC)

Occasionally I conduct a power analysis for a survey design and address a question  of how many people we need to reach to achieve a sufficient statistical power.   I usually use Excel for this when the item of interest is a binary variable.

What Works Clearinghouse considers an effect size of .25 (or greater) "substantively important."  See p. 26, table footnote at https://ies.ed.gov/ncee/wwc/Docs/referenceresources/wwc_procedures_v3_0_standards_handbook.pdf

J. Cohen (1988) has an idea of .2 small effect, .5 medium effect, .8 large effect.

Cohen, J. Statistical power for the behavioral sciences (2nd ed.). Hillsdale, NJ: Erlbaum (1988).  See page 5 of  http://www.wmich.edu/evalphd/wp-content/uploads/2010/05/Effect_Size_Substantive_Interpretation_Guidelines.pdf .

 

 

Power analysis for survey analysis (and a lot of other study settings).