An Efficient Scores Test for Comparing Several Measuring Devices

Abstract

We develop an omnibus test for the equality of biases and variances in several measuring devices, assuming Grubbs' model. The omnibus test statistic is based on the method of efficient scores, and is the sum of two components which are the scores statistics for testing equality of biases and variances separately. An advantage of the efficient scores test over competing tests is that the scores test is easily modified to allow symmetric non-normal responses. An application of the proposed test is discussed.

• Analysis of Variance
• Grubbs' Model
• Hypothesis Testing
• Variance Components

Additional information

Notes on contributors

Edward J. Bedrick

Dr. Bedrick is a Professor in the Department of Mathematics and Statistics. His email address is [email protected].