Methods of Data Processing in Measurements and Metrological Models

Abstract

The development of methods of data processing over the past 50 years is outlined. Distinctive features of data processing as a division of the theory of measurements are identified. The essential role and diversity of metrological models are emphasized. The significance of estimation of systematic measurement errors, including their randomization, is noted. The relationship between regression and confluent models is discussed within the framework of a discussion of joint measurements. Validation of data processing algorithms is considered as a practical approach to the comparison of the quality of different algorithms. The role of models in the determination and comparison of precision characteristics is demonstrated using the Allan variance as an example.