Software Reviews : Topics in Research Methods: Power,Version 1.0 Publisher: Conduit,University of Iowa,Oakdale Campus,Iowa City,IA 52242 (telephone: [phone]) Year of Publication: 1987 Materials: Two diskettes (original and backup copies),instructor notes (two pages),student notes (five pages) Price: $50 Machine Specificity: IBM,Apple System Requirements: DOS 2.1-3.2,one 360K disk drive,color graphics (CGA,EGA) Effectiveness: Fair User-Friendliness: Good Documentation: Fair

Abstract

POWER uses a series of simulated experiments to introduce students to the statistical concepts of quality of experimental design and error variance. The experiments are based on a real-life comparison of electricity use in homes with and without a signaling device for home cooling. POWER allows the student to modify the experiment in six variable areas and manipulate combinations of those variables to influence changes in the power of the test and in the amount of statistical error in the experimental setting. POWER is useful for demonstrating the concepts of power and error variance, but it does not allow for the introduction of independent data for student-generated investigations.