Statistical Treatment of Experimental Data

This book is primarily intended for researchers wishing to analyse experimental data using statistical methods. Statistical concepts and methods which may be employed to treat experimental data are explained, and the ideas and reasoning behind statistical methodology are clarified. Formal results are illustrated by many numerical worked examples mainly taken from the laboratory. Concepts, practical methodology, and worked examples are integrated in the text. Consideration is given in this work to a large number of practical topics which are often omitted from standard texts. These include; obtaining an approximate confidence interval for a function of some unknown parameters; testing for outliers, stabilization of heterogeneous variances, and significant differences between means; estimation of parameters after performing tests; deciding what numbers of significant figures to quote for sample means and variances; straight-line and polynomial regression, through the origin or not, using weighted points, and testing the homogeneity of a set of such lines or curves.
The numerous examples which are provided throughout the text will serve as models for the various problems encountered by the readers when employing statistical methods to treat experimental data. Neither a strong mathematical background nor a prior knowledge of probability or statistics is required in order to make use of this work. In addition to research workers in universities and industry, the book will be of use for first-year students of statistics, and would be expecially suitable as the basis of a graduate course in experimental sciences.
 
This book is primarily intended for researchers wishing to analyse experimental data using statistical methods. Statistical concepts and methods which may be employed to treat experimental data are explained, and the ideas and reasoning behind statistical methodology are clarified. Formal results are illustrated by many numerical worked examples mainly taken from the laboratory. Concepts, practical methodology, and worked examples are integrated in the text. Consideration is given in this work to a large number of practical topics which are often omitted from standard texts. These include; obtaining an approximate confidence interval for a function of some unknown parameters; testing for outliers, stabilization of heterogeneous variances, and significant differences between means; estimation of parameters after performing tests; deciding what numbers of significant figures to quote for sample means and variances; straight-line and polynomial regression, through the origin or not, using weighted points, and testing the homogeneity of a set of such lines or curves.
The numerous examples which are provided throughout the text will serve as models for the various problems encountered by the readers when employing statistical methods to treat experimental data. Neither a strong mathematical background nor a prior knowledge of probability or statistics is required in order to make use of this work. In addition to research workers in universities and industry, the book will be of use for first-year students of statistics, and would be expecially suitable as the basis of a graduate course in experimental sciences.