Understanding and Implementing Power Analysis for Treatment-Control Studies

Session Type: Professional Development Course

Abstract

Understanding statistical power is important for researchers, funders, and the education science field at large. This course will introduce the major concepts in power analysis (e.g., types of errors) and how they relate to statistical analyses. The course lectures will cover the treatment-control design for simple random samples (SRSs), SRSs with covariates, multilevel models (MLMs), and MLMs with covariates. Activities will include calculating power, sample sizes, and effect sizes using interactive computer programs provided by the instructor, and group activities. Discussion will include how to research assumptions and how to write about power. Free software (R and R studio) to perform power analysis will be introduced with interactive exercises. This course is appropriate for researchers involved in the planning and execution of education studies: professors at all levels, postdocs, graduate students, and staff at research firms.

Sub Unit

• Professional Development and Training Committee

Instructor

• Eric Hedberg, NORC at the University of Chicago