PAWS: Power analysis and Width of Confidence Inverval for Sample Size Estimation

The Power Analysis and Width of a confidence interval for Sample size estimation (PAWS) application is to help researchers in the planning stage of their cluster randomized design studies. The program provides information about optimal sample size combinations (number of clusters and cluster size) based on several criteria: statistical power, accuracy in effect size estimation, and cost. PAWS estimates the cluster and individual-level sample size combination, resulting in the desired power with the lowest cost. Also, the program will calculate the cluster and individual-level sample size combination that provides the smallest width of confidence intervals of effect sizes with the lowest cost.

Developers

Sunthud Pornprasertmanit

W. Joel Schneider

Program

Get program here. While you are installed, you might see the warning that the publisher cannot be trusted. If you load from this site, you can trust this program. Unfortunately, this program is written by Visual Basic and will only support the Window platform.

The source code is available at my github page.

Please report any bugs or give me any suggestions by email

Materials

Software Manual

Thesis Cover

Thesis Table of Contents

Thesis Content

Presentation

Pornprasertmanit, S., & Schneider, W. J. (October, 2010). Efficient sample size for power and desired accuracy in Cohens'd estimation in two-condition cluster randomized design. Paper presented at the weekly colloquium in the Center of Research Methods and Data Analysis, University of Kansas. Presentation.

Pornprasertmanit, S., & Schneider, W. J. (April, 2010). Sample size estimation in cluster randomized design: PAWS-CRD. Presentation given for Sunthud Pornprasertmanit's thesis defense at the Department of Psychology, Illinois State University. Presentation.

Pornprasertmanit, S., & Schneider, W. J. (February, 2010). Sample size estimation in two-group cluster randomized designs. Presentation given for Sunthud Pornprasertmanit's proposal presentation at the Department of Psychology, Illinois State University. Presentation.

Pornprasertmanit, S., & Schneider, W. J. (October, 2009). Sample size estimation for two-group cluster randomized designs. Presentation given at the Quantitative Professional Seminar at the Department of Psychology, Illinois State University. Presentation.