Analysis and Presentation of Experimental Results With Examples Problems and Programs 1st Edition Costas Christodoulides

Analysis and Presentation of
Experimental Results With Examples

Problems and Programs 1st Edition
Costas Christodoulides
Visit to download the full and correct content document:
https://textbookfull.com/product/analysis-and-presentation-of-experimental-results-wit
h-examples-problems-and-programs-1st-edition-costas-christodoulides/
More products digital (pdf, epub, mobi) instant
download maybe you interests ...
Experimental Methods For Science And Engineering

Students An Introduction To The Analysis And
Presentation Of Data Les Kirkup
https://textbookfull.com/product/experimental-methods-for-
science-and-engineering-students-an-introduction-to-the-analysis-
and-presentation-of-data-les-kirkup/
Uncertainty analysis of experimental data with R 1st

Edition Benjamin David Shaw
https://textbookfull.com/product/uncertainty-analysis-of-
experimental-data-with-r-1st-edition-benjamin-david-shaw/
Statistical Analysis of Financial Data: With Examples

In R 1st Edition James Gentle
https://textbookfull.com/product/statistical-analysis-of-
financial-data-with-examples-in-r-1st-edition-james-gentle/
Computational and experimental analysis of functional

materials 1st Edition Reshetnyak
https://textbookfull.com/product/computational-and-experimental-
analysis-of-functional-materials-1st-edition-reshetnyak/
Recurrent Sequences Key Results Applications and
Problems Dorin Andrica
https://textbookfull.com/product/recurrent-sequences-key-results-
applications-and-problems-dorin-andrica/
Concise optics: concepts, examples, and problems 1st

Edition Ajawad I. Haija
https://textbookfull.com/product/concise-optics-concepts-
examples-and-problems-1st-edition-ajawad-i-haija/
Time Series Analysis and Its Applications With R

Examples 4th Edition Robert H. Shumway
https://textbookfull.com/product/time-series-analysis-and-its-
applications-with-r-examples-4th-edition-robert-h-shumway/
Eurocode Compliant Seismic Analysis and Design of R C

Buildings Concepts Commentary and Worked Examples with
Flowcharts 1st Edition Ioannis Avramidis
https://textbookfull.com/product/eurocode-compliant-seismic-
analysis-and-design-of-r-c-buildings-concepts-commentary-and-
worked-examples-with-flowcharts-1st-edition-ioannis-avramidis/
Survival Analysis with Interval Censored Data A

Practical Approach with Examples in R SAS and BUGS 1st
Edition Kris Bogaerts
https://textbookfull.com/product/survival-analysis-with-interval-
censored-data-a-practical-approach-with-examples-in-r-sas-and-
bugs-1st-edition-kris-bogaerts/
Undergraduate Lecture Notes in Physics
Costas Christodoulides
George Christodoulides
Analysis and
Presentation of
Experimental
Results
With Examples, Problems and
Programs
Undergraduate Lecture Notes in Physics (ULNP) publishes authoritative texts covering
topics throughout pure and applied physics. Each title in the series is suitable as a basis for
undergraduate instruction, typically containing practice problems, worked examples, chapter
summaries, and suggestions for further reading.
ULNP titles must provide at least one of the following:

• An exceptionally clear and concise treatment of a standard undergraduate subject.
• A solid undergraduate-level introduction to a graduate, advanced, or non-standard subject.
• A novel perspective or an unusual approach to teaching a subject.
ULNP especially encourages new, original, and idiosyncratic approaches to physics teaching
at the undergraduate level.
The purpose of ULNP is to provide intriguing, absorbing books that will continue to be the
reader's preferred reference throughout their academic career.
Series editors
Matthew Deady
Physics Program, Bard College, Annandale-on-Hudson, NY, USA
Morten Hjorth-Jensen
Department of Physics, University of Oslo, Oslo, Norway
Michael Inglis
SUNY Suffolk County Community College, Long Island, NY, USA
Heinz Klose
Humboldt University, Oldenburg, Niedersachsen, Germany
More information about this series at http://www.springer.com/series/8917

Costas Christodoulides George Christodoulides
•
Analysis and Presentation

of Experimental Results
With Examples, Problems and Programs
123
Costas Christodoulides George Christodoulides
Department of Physics, School of Applied Centre Valibel, Institute for Language
Mathematical and Physical Sciences and communication
National Technical University of Athens Université catholique de Louvain
Zografou, Athens GR15780 Louvain-la-Neuve
Greece Belgium
Code and examples for the book “Analysis and Presentation of Experimental Results”:
https://github.com/aperbook/APER.
ISSN 2192-4791 ISSN 2192-4805 (electronic)

ISBN 978-3-319-53344-5 ISBN 978-3-319-53345-2 (eBook)
DOI 10.1007/978-3-319-53345-2
Library of Congress Control Number: 2017932001
Translation from the Greek language edition: Amάktrη jai Paqotrίarη Peiqalasijώm
Aposekerlάsxm, © N.T.U.A. University Press, Athens 2009. All Rights Reserved.
© Springer International Publishing AG 2017
This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part
of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations,
recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission
or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar
methodology now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this
publication does not imply, even in the absence of a specific statement, that such names are exempt from
the relevant protective laws and regulations and therefore free for general use.
The publisher, the authors and the editors are safe to assume that the advice and information in this
book are believed to be true and accurate at the date of publication. Neither the publisher nor the
authors or the editors give a warranty, express or implied, with respect to the material contained herein or
for any errors or omissions that may have been made. The publisher remains neutral with regard to
jurisdictional claims in published maps and institutional affiliations.
Printed on acid-free paper
This Springer imprint is published by Springer Nature

The registered company is Springer International Publishing AG
The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland
Preface
The importance of the analysis and the presentation of experimental data cannot be
overstated. Good experimental results may be rendered useless by failing to keep to
certain rules in their presentation, either to an audience or in written form. It is our
intention to present in this book these methods at an introductory university level.
Those working in the experimental sciences, but also anyone involved in the
analysis of numerical data, may find the book useful.
The book is intended to be used as a textbook and this has determined its
characteristics: The theoretical proofs are given in considerable detail, many figures
are used, as well as a large number of examples and problems to be solved by the
reader. The vast majority of the examples are solved using four software packages:
Excel®, Origin®, Python and R. Most of the problems may also be solved using
these programs. Excel® is used due to its wide availability as a program for data
analysis, Origin® because it is an excellent program for creating graphical pre-
sentations of data. Python and R are used because they are free, open-source
programming languages, widely used in data science. Reference to these programs
is made using the symbols [E], [O], [P] and [R], respectively. The same symbols
also indicate that a certain problem may be solved using the corresponding
program.
The book may be used as a textbook for an introductory course on Data Analysis
and Presentation. It is hoped that it will provide a useful addition to the existing
literature.
Athens, Greece Costas Christodoulides

January 2017 George Christodoulides
v
Contents
1 Basic Statistical Concepts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

1.1 Tables, Distributions and Histograms . . . . . . . . . . . . . . . . . . . . . 1
1.2 Probability Density. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
1.3 Mean Value . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
1.4 Measures of Dispersion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
1.4.1 Range . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
1.4.2 Deviation from the Mean and Mean
Absolute Deviation from the Mean . . . . . ........... 19
1.4.3 Standard Deviation . . . . . . . . . . . . . . . . . ........... 20
2 Measurement Errors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
2.1 Errors of Measurements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
2.1.1 Accidental or Random Errors . . . . . . . . . . . . . . . . . . . . 39
2.1.2 Systematic Errors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
2.1.3 Personal Errors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
2.1.4 Occasional Errors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
2.1.5 The Errors in Reading the Indications
of Instruments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
2.2 Errors in Compound Quantities . . . . . . . . . . . . . . . . . . . . . . . . . 45
2.2.1 Error in a Sum or a Difference . . . . . . . . . . . . . . . . . . . 46
2.2.2 Error in a Product . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
2.2.3 Error in a Power . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
2.2.4 Error in a Quotient . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
2.2.5 The Use of Differentials . . . . . . . . . . . . . . . . . . . . . . . . 50
3 A Thought Experiment. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
3.1 The Thought Experiment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
Reference . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
4 The Statistical Analysis of Experimental Results . . . . . . . . . . . . . . . 77
4.1 The Mean and the Dispersion of the Results
of Measurements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
vii
viii Contents
4.2 The Standard Deviations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80

4.2.1 The Standard Deviation of the Measurements . . . . . . . . 80
4.2.2 The Standard Deviation of the Mean . . . . . . . . . . . . . . 90
4.2.3 The Relationship Between r and rx . . . . . . . . . . . . . . . 93
4.2.4 The Relationship Between sx and r and rx . . . . . . . . . . 94
4.3 The Standard Deviation of the Standard Deviation
of the Mean . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 98
4.4 Information Derived from the Measurement of x and sx . . . . .. 100
4.4.1 The Mean Value of the Results
of the Measurements and Its Standard Deviation . . . .. 100
4.4.2 The Statistical Distribution of the Results
of the Measurements . . . . . . . . . . . . . . . . . . . . . . . . . . . 103
4.4.3 Statistical Estimates for the Mean . . . . . . . . . . . . . . . . . 115
4.4.4 Summary of the Method of Analysis of the Results . . . 116
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121
5 The Presentation of Numerical Results . . . . . . . . . . . . . . . . . . . . . .. 123
5.1 Significant Figures and Rounding of Numbers . . . . . . . . . . . . .. 123
5.2 The Presentation of a Numerical Result of a Series of
Measurements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 124
5.3 The Number of Significant Figures Used in the Presentation
of Numerical Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 126
5.4 The International System of Units (S.I.) and the Rules
of Its Use . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 129
5.5 Recommendations on the Notation Used for Mathematical
Constants, Algebraic Parameters, Variables, Indices,
Mathematical Functions, Operators, Physical Units,
Elementary Particles and Isotopes . . . . . . . . . . . . . . . . . . . . . .. 134
Reference . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 136
6 The Propagation of Errors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 137
6.1 The Combination of Two Series of Measurements
of the Same Physical Magnitude . . . . . . . . . . . . . . . . . . . . . . .. 137
6.1.1 The Mean x of All the Measurements . . . . . . . . . . . . .. 138
6.1.2 The Standard Deviation sx of All the Measurements. .. 138
6.1.3 The Standard Deviation of the Mean rx
of All the Measurements . . . . . . . . . . . . . . . . . . . . . . .. 139
6.2 The Mean and the Standard Deviation of a Function
of Measured Quantities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 141
6.2.1 The Mean and the Standard Deviations
of a Function of One Variable . . . . . . . . . . . . . . . . . .. 141
6.2.2 The Mean and the Standard Deviation
of an Algebraic Sum . . . . . . . . . . . . . . . . . . . . . . . . . .. 146
Contents ix
6.2.3 The Mean and the Standard Deviations

of a General Function of Many Variables . . . . . . . . . .. 149
6.2.4 Another Approach to the Evaluation of the Mean
and the Standard Deviation of a Compound
Quantity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 155
6.3 The Error in Qðx; y; z; . . .Þ Due to the Errors in x; y; z; . . . . . .. 157
6.3.1 The Case of Asymmetrical Errors . . . . . . . . . . . . . . . .. 158
7 The Three Basic Probability Distributions . . . . . . . . . . . . . . . . . . . . 163
7.1 Histograms of Experimental Probabilities . . . . . . . . . . . . . . . . . . 163
7.2 The Binomial or Bernoulli Distribution . . . . . . . . . . . . . . . . . . . 166
7.2.1 The Normalization of the Binomial Distribution . . . . . . 178
7.2.2 The Mean Value of x for the Binomial Distribution . . . 178
7.2.3 The Standard Deviation of x from the Mean
for a Binomial Distribution . . . . . . . . . . . . . . . . . . . . . . 179
7.3 The Poisson Distribution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 180
7.3.1 The Normalization of the Poisson Distribution . . . . . . . 188
7.3.2 The Mean Value of x for the Poisson Distribution . . . . 188
7.3.3 The Standard Deviation from the Mean of x
for the Poisson Distribution . . . . . . . . . . . . . . . . . . . . . 188
7.4 The Normal or Gaussian Distribution. . . . . . . . . . . . . . . . . . . . . 192
7.4.1 The Normalization of the Gaussian Distribution . . . . . . 200
7.4.2 The Mean Value of x for the Gaussian Distribution . . . 200
7.4.3 The Standard Deviation of x from the Mean
for the Gaussian Distribution . . . . . . . . . . . . . . . . . . .. 200
7.4.4 Testing Whether a Set of Data Has a Gaussian
Distribution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 201
7.4.5 The Gaussian Distribution and the Phenomenon
of Diffusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 209
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 214
8 The Statistics of Radioactivity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 215
8.1 The Behavior of Large Samples. The Law of Radioactivity . . . . 215
8.2 Nuclear Disintegrations and the Binomial Distribution. . . . . . . . 218
8.3 Radioactivity and the Poisson Distribution . . . . . . . . . . . . . . . . . 231
8.4 The Counting Rate of Nuclear Disintegrations and Its Error . . . 234
Reference . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 238
9 Elements from the Theory of Errors . . . . . . . . . . . . . . . . . . . . . . . .. 239
9.1 The Normal or Gaussian Law of Errors . . . . . . . . . . . . . . . . . .. 239
9.2 The Lyapunov Central Limit Theorem . . . . . . . . . . . . . . . . . . .. 244
9.3 The Best Estimate that May Be Made for the Real Value
of a Magnitude, Based on the Results
of N Measurements of It . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 253
x Contents
9.4 The Weighting of Values . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 255

9.5 The Joint Probability Density for Two Random Variables . . . . . 272
9.6 The Probability Density of the Sum
of Two Random Variables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 277
9.6.1 The Probability Density of the Sum of Two Normally
Distributed Random Variables . . . . . . . . . . . . . . . . . . . 280
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 289
10 Comparison and Rejection of Measurements . . . . . . . . . . . . . . . . . . 291
10.1 The Problem of the Rejection of Measurements . . . . . . . . . . . . . 291
10.2 Chauvenet’s Criterion. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 293
10.3 Comments Concerning the Rejection of Measurements . . . . . . . 296
10.4 Comparison of the Means of Two Series of Measurements
of the Same Quantity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 298
11 The Method of Least Squares . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 301
11.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 301
11.2 The Theoretical Foundation of the Method of Least Squares . . . 302
11.3 The Fitting of Curves to Experimental Points . . . . . . . . . . . . . . 305
11.3.1 Straight Line . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 305
11.3.2 Polynomial . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 327
11.3.3 Other Curves. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 340
11.3.4 The Reduction of Non-linear Relations to Linear . . . . . 342
11.4 The Choice of the Optimum Function Fitted
to a Set of Experimental Results . . . . . . . . . . . . . . . . . . . . . . .. 346
11.5 The Fractional Absolute Deviation of the Experimental
Values from the Values of the Curve . . . . . . . . . . . . . . . . . . . . . 346
11.6 Smoothing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 348
11.7 The Error in a Value Read off a Smoothed Curve . . . . . . . . . . . 354
11.8 The Regression Line and the Coefficient of Correlation . . . . . . . 357
11.9 The Use of the Method of Least Squares in the Solution
of a System of Overdetermined Linear Equations . . . . . . . . . . . 362
11.9.1 Equations in Two Variables . . . . . . . . . . . . . . . . . . . . . 363
11.9.2 Equations in Three Variables . . . . . . . . . . . . . . . . . . . . 368
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 375
12 Graphs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 377
12.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 377
12.2 The Axes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 378
12.2.1 Linear Scales . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 378
12.2.2 Logarithmic Scales . . . . . . . . . . . . . . . . . . . . . . . . . . . . 380
12.2.3 Polar Diagrams . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 384
12.2.4 Other Matters Relating to the Axes and the Scales
of Graphs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 387
12.2.5 Legends of the Figure, Labels of the Axes
and the Units of Physical Magnitudes . . . . . . . . . . . . .. 391
Contents xi
12.3 The Points . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 392

12.4 The Curve . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 394
12.5 The Slope of the Curve . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 420
12.5.1 A Graphical Method of Evaluating the Errors da
and dk in the Parameters a and k of the Straight Line
y ¼ a þ kx . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 422
12.5.2 The Evaluation of Slopes of Straight Lines
in Graphs with Logarithmic Scales . . . . . . . . . . . . . . . . 431
Reference . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 447
13 The Written Report of the Results of an Experiment . . . . . . . . . . . 449
13.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 449
13.2 The Preparation of the Writing of the Report While Performing
the Experiment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 450
13.3 The Written Report of an Experiment . . . . . . . . . . . . . . . . . . . . 452
13.4 An Example of a Written Report of an Experiment . . . . . . . . . . 454
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 471
14 Appendices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 473
14.1 Appendix 1: Least Squares Straight Line y ¼ a þ kx.
The Errors in a and k . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 473
14.2 Appendix 2: Dimensional Analysis . . . . . . . . . . . . . . . . . . . . . . 483
14.2.1 The Dimensions of Physical Quantities . . . . . . . . . . . . . 483
14.2.2 The Dimensional Homogeneity of Equations . . . . . . . . 485
14.2.3 The Derivation of Relations Between Physical
Quantities Using Dimensional Analysis . . . . . . . . . . .. 486
14.3 Appendix 3: The Use of Random Numbers in Finding
Values x of a Variable x Which Are Distributed According
to a Given Probability Density Function Function f(x) . . . . . . .. 493
14.3.1 The Use of Random Numbers in Finding Values x
of a Variable x Which Are Distributed According
to a Given Probability Density Function f(x) . . . . . . .. 494
14.4 Appendix 4: The Values of the Fundamental Physical
Constants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 502
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 505
Answers to the Problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 507
List of Programs and Code Samples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 515
Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 521
Instructions on the Use of the Programs
and Code Samples
This book contains more than a hundred programs and code samples that
demonstrate how to analyse and present experimental data using Excel®, Origin®,
R and Python. You can download all these programs (or the ones you’re interested
in) from https://github.com/aperbook/APER. The companion website is organized
by software package, and then by book chapter. Additional programs and scripts,
including extensions to existing programs, which could not be included in the book
for space reasons, are available on the website.
The programs have been tested on the Windows®, Mac® and Linux platforms.
The Microsoft Excel® spreadsheets have been created using version 2016, and will
work with subsequent versions of the software. The OriginLab® Origin® files have
been created using version 2016 Pro of the software. Both Microsoft Excel (prod-
ucts.office.com/en/excel) and OriginLab Origin (www.originlab.com/Origin) are
proprietary software, and you will need a licensed version in order to open the files
provided with this book. The licensing costs will depend on the number of copies
of the software you wish to install; there are also academic and institutional licenses
available. The Excel spreadsheets can also be opened using an open-source, free
package such as LibreOffice Calc (www.libreoffice.org).
In recent years, open-source software projects have changed the landscape of
scientific and data analysis software. We have chosen to focus on two of the most
popular solutions, the R Project for Statistical Computing and the Python
general-purpose programming language.
R is a special-purpose programming language and an environment for statistical
computing. It is freely available (www.r-project.org) under the GNU General Public
License. A very large number of contributed packages extend R’s functionality and
can be automatically downloaded through the CRAN archive. To make the most
out of the R code provided with this book, we suggest that you install R along with
the RStudio graphical user interface, which is also free and open-source. You have
to install R before installing RStudio. To install R, visit https://cran.rstudio.com/
and chose your platform (Windows, Mac or Linux). You only need to install the
“base” distribution to run the code in this book. Binary distributions (executable
xiii
xiv Instructions on the Use of the Programs and Code Samples
files) are available for Windows and Mac. Linux users will usually install R using
their operating system’s package manager (e.g. the “apt-get” command in Ubuntu).
After installing R, visit https://www.rstudio.com/products/rstudio/download3/ and
select the appropriate installer for your platform. At the time of writing, the current
version of R is 3.3.1 and the current version of RStudio is 0.99.903.
Python is a general-purpose, interpreted programming language (www.python.org).
Python interpreters are available for several platforms, allowing the same plat-
form code to run on different operating systems. Three major Python projects
provide support for scientific programming, data analysis and graphics: NumPy
(www.numpy.org), SciPy (www.scipy.org) and Matplotlib (www.matplotlib.org).
In order to use the code provided in this book we suggest that you install a complete
Python distribution, containing all the necessary packages. The Anaconda distri-
bution has integrated over 720 packages to cover the needs of data scientists using
Python. Download the distribution from https://www.continuum.io/downloads by
choosing your platform (Windows, Mac or Linux, and 32- or 64-bit). The code
in this book has been tested to work with both version 2.x and version 3.x of
Python; however, we suggest that you download the installers for the latest version
(Python 3.5). After installing the Anaconda distribution of Python, you can launch
the IPython Qt Interactive Console, and run the code provided with this book
interactively.
Reference to these programs is made using the symbols [E], [O], [P] and [R], for
Excel, Origin, R and Python, respectively. The same symbols also indicate that a
certain problem may be solved using the corresponding program.
General Bibliography
J. Topping, Errors of Observation and their Treatment (Springer, 4th ed., 1972).
S.L. Meyer, Data Analysis for Scientists and Engineers (J. Wiley & Sons, New York, 1975).
L.G. Parratt, Probability and Experimental Errors in Science (Dover Publications, New York,
1971).
P.R. Bevington, Data Reduction and Error Analysis for the Physical Sciences (McGraw-Hill Book
Company, New York, 1969).
H.D. Young, Statistical Treatment of Experimental Data (McGraw-Hill Book Company, New
York, 1962).
J.R. Taylor, An Introduction to Error Analysis (The Study of Uncertainties in Physical
Measurements) (University Science Books, Sausalito, California, 2nd ed. 1997).
L. Kirkup, Data Analysis with Excel® (An Introduction for Physical Scientists) (Cambridge
University Press, 2002).
Chapter 1
Basic Statistical Concepts
In order to understand the theory of errors, as well as the methods of analysis and
presentation of experimental results, knowledge of the basic concepts of Probability
Theory and Statistics is required. The purpose of this chapter is the concise presen-
tation of the basic concepts from these fields that will be used in the book. Obviously,
the treatment will not be in great detail and the reader should consult the relevant
bibliography for this purpose. Those readers who have a satisfactory knowledge of
these subjects may omit this chapter and consult it whenever necessary.
1.1 Tables, Distributions and Histograms
Let us assume that we have performed N measurements xi ði ¼ 1; 2; . . .; NÞ of a

magnitude x, keeping the experimental conditions constant as far as possible. These
measurements may be considered to be a sample which was taken from a parent
population of the infinite number of measurements of this magnitude that may be
performed under the particular conditions. In Table 1.1, the results of 100 such
measurements have been recorded. This is the form in which the results appear
immediately after the measurements, although the number of measurements is very
rarely as high as this, except perhaps when they are acquired by an automated method.
If the number of the measurements is large enough, there may be values which
appear more than once. If we count the frequency of appearance of each value, we
may construct a table of this number, nr , as a function of the numerical value xr of
the measurement, such as Table 1.2. In this table, the values xr have been recorded
in order of increasing magnitude and the number r is given in the first column of the
table. The value of nr is the frequency of appearance of the r-th value. Obviously,
© Springer International Publishing AG 2017 1

C. Christodoulides and G. Christodoulides, Analysis and Presentation
of Experimental Results, Undergraduate Lecture Notes in Physics,
DOI 10.1007/978-3-319-53345-2_1
2 1 Basic Statistical Concepts
Table 1.1 100 measurements xi of the magnitude x (in mm)

i xi i xi i xi i xi i xi
1 9.5 21 9.9 41 10.2 61 9.9 81 10.1
2 10.3 22 10.2 42 9.9 62 10.2 82 9.5
3 10.1 23 9.9 43 10.1 63 9.5 83 9.6
4 9.9 24 10.2 44 9.6 64 9.6 84 10.4
5 10.0 25 10.0 45 10.0 65 10.2 85 10.2
6 10.3 26 9.9 46 9.9 66 9.5 86 9.9
7 9.7 27 10.1 47 10.0 67 10.1 87 9.6
8 10.0 28 9.7 48 9.7 68 10.0 88 9.7
9 10.3 29 10.1 49 10.3 69 9.7 89 9.8
10 9.8 30 10.0 50 9.8 70 9.8 90 10.2
11 10.2 31 9.9 51 10.0 71 9.8 91 10.4
12 9.8 32 10.4 52 9.8 72 10.3 92 9.9
13 10.2 33 10.3 53 10.0 73 9.9 93 10.3
14 10.2 34 10.4 54 10.3 74 10.4 94 9.9
15 9.7 35 9.6 55 10.1 75 10.0 95 9.5
16 10.1 36 9.7 56 9.7 76 10.1 96 10.0
17 10.1 37 10.0 57 10.4 77 9.6 97 10.0
18 10.2 38 9.8 58 10.0 78 10.4 98 9.8
19 9.8 39 10.1 59 10.0 79 9.6 99 9.9
20 10.4 40 9.9 60 10.1 80 10.2 100 10.3
Table 1.2 The values of Table 1.1 grouped in classes

r xr (mm) nr fr
1 9.5 5 0.05
2 9.6 7 0.07
3 9.7 8 0.08
4 9.8 10 0.10
5 9.9 14 0.14
6 10.0 15 0.15
7 10.1 12 0.12
8 10.2 12 0.12
9 10.3 9 0.09
10 10.4 8 0.08
R= 100 1
it is Rnr ¼ N. Table 1.2 is said to give the frequency distribution of the mea-
surements. If the frequencies nr are divided by the total number of measurements N,
the relative frequencies fr of the values xr are obtained and these are also given in
the table. By definition, it is Rfr ¼ 1. Due to this fact, fr is said to be normalized.
If the variable x can only take discrete values (e.g. number of objects, exam
grades between 1 and 10 etc.), the distribution is said to be discrete. If x varies in a
1.1 Tables, Distributions and Histograms 3
Table 1.3 The values of Class (mm) nr fr

Table 1.1 grouped in classes
of width 0.2 mm 9.40–9.59 5 0.05
9.60–9.79 15 0.15
9.80–9.99 24 0.24
10.00–10.19 27 0.27
10.20–10.39 21 0.21
10.40–10.59 8 0.08
R= 100 1
continuous manner (e.g. length, time, temperature etc.), the distribution is said to be
continuous.
In many cases, the N values xi are grouped together in a certain number of classes,
which have boundaries and width Dx, also known as bin width. For example, the
measurements of Table 1.1 may be distributed in classes of width Dx ¼ 0:2 mm,
having the following boundaries (in mm): 9.40–9.59, 9.60–9.79, 9.80–9.99, 10.00–
10.19, 10.20–10.39, 10.40–10.59. The result is shown in Table 1.3.
The boundaries of the classes must be stated in a clear manner, in order to avoid
mistakes during the distribution of the values in the classes. In Table 1.3 the
boundaries of the classes are given with two decimal points, while the values are given
with one. Another way of determining the boundaries of the classes is the following:
½9:49:6Þ ½9:69:8Þ ½9:810:0Þ ½10:010:2Þ ½10:210:4Þ ½10:410:6Þ
where the symbol [9.4 or 9.4], for example, means that in both cases the value 9.4 is
included in the class, whereas the symbol (9.6 or 9.6) means that the value 9.6 is not
included. A third way to write the class boundaries is:
9:49:6 9:69:8 9:810:0 10:010:2 10:210:4 10:410:6
where by writing 9.4 as a boundary we mean that the value 9.4 is included in the
class, whereas by writing 9:6 we mean that the value 9.6 is not included.
If the points (xr , nr ) are plotted in a diagram such as the one in Fig. 1.1. we have
a frequency polygon. If we plot the points (xr , fr ), we have a relative frequency
polygon (Fig. 1.1, right axis).
The frequency (or the relative frequency) of the classes may be used in the
drawing of a histogram of the measurements, as follows: The x-axis is divided into
regions which correspond to the classes of the distribution of the values and in each
region a column is raised, having a height proportional to the frequency xr (or the
relative frequency fr ). The result is a frequency histogram (or a histogram of rel-
ative frequencies, respectively). Shown in Fig. 1.2 is the frequency histogram and
the relative frequency histogram of the values of Table 1.3, in which the width of
the classes is Dx ¼ 0:2 mm.
Obviously, we need as many columns as possible in a histogram, so that we
detect any fine structure it may possess. At the same time, we need a large enough
Fig. 1.1 A polygon of the frequencies nr and of the relative frequencies fr of the 100
measurements of Table 1.1
number of measurements in the majority of the columns, so as to have a ‘smooth

statistical behavior’. To achieve this, the width of the classes may be decided using
the following practical rule:
If N is the number of the values to be recorded in the histogram and R is the
range of these values, then the width of the classes is chosen approximately
pffiffiffiffi
equal to R= N . Obviously this figure is suitably rounded.
As an example, if we have N = 100 measurements, the values of which have a
range of R ¼ 32 mm, the suggested value for the width of the classes is
pffiffiffiffiffiffiffiffi
Dx ¼ 32= 100 ¼ 3:2 mm, which should be rounded to 3 mm.
The values given by this rule may not always be used since the value of Dx is
decided after the particular facts of a situation are taken into account. For example,
it is usually desirable to have classes with widths equal to 1, 2 or 5 units, their
multiples or their sub-multiples.
Example 1.1 [E]

Given 100 random numbers, use Excel® in order to plot a histogram of them.
For the purposes of this example and using the random number generation func-
tionality of Excel® we enter, in column A, 100 random numbers, normally dis-
tributed with a mean of 5 and a standard deviation of 1. The details need not
concern us here. We will explain how this is done in later stages of the book. All we
need to know now is that we have 100 numbers in column A, in cells A1–A100. As
in Fig. 1.2, we will assume that the numbers given are the results of measurements
of a length in mm. It should be noted that the histogram is to be drawn using the raw
data of the measurements, and not after they have been counted and grouped in
classes as in Table 1.3.
Fig. 1.2 The frequency histogram and the histogram of the relative frequencies of the
measurements of Table 1.3. The width of the classes is Dx ¼ 0:2 mm
We note that the numbers lie within the range of 2.5 and 7.5 mm. In preparation
for plotting the histogram we enter the numbers 2.5, 3, 3.5, …, 6, 6.5, 7, 7.5 in cells
E1–E11. These will be the limits of the bins of the histogram. The bin width is thus
taken to be equal to Dx ¼ 0:5 mm.
In the window Data we open Data Analysis. This is an Add-in which has to be
installed if it is not already installed (To install, follow: File, Options, Add-ins,
Analysis ToolPak, OK). We open the Histogram dialog box.
With the cursor in Input Range: we move the cursor to A1 and, with the left
hand mouse button held down, we drag the cursor to cell A100. When the mouse
button is released, the range $A$1:$A$100 appears in the Input Range: box.
With the cursor in Bin Range: we move the cursor to E1 and, with the left hand
mouse button held down, we drag the cursor to cell E11. When the mouse button is
released, the range $E$1:$E$11 appears in the Bin Range: box.
We tick the Chart Output and press OK. The table giving the Frequency for
each of the bins and the Histogram shown below appear.
Bin Frequency
2 0
2.5 0
3 2
3.5 4
4 2
4.5 22
5 27
5.5 14
6 17
6.5 5
7 3
7.5 4
More 0
We will now modify the histogram.

We delete the words Frequency and Histogram. We may give a new title to the
histogram.
We change the X-label by double-clicking on the existing word Bin and
changing it to Result of the measurement, Dx (mm). Similarly, we change the
Y-label to Frequency, DN.
Double-click on the x-axis and open the Format Axis dialog box. Then open
Labels > Specify interval unit and set value to 2. The DN axis is satisfactory.
Double-click on one of the histogram’s bars and open the Format Data Series
dialog box. At Series Options, set the Gap Width to 10%. This sets the empty gap
between the columns to 10% of the bin width.
Clicking on the Fill and Line icon, , we open the Fill, Border dialog box. At
Fill > Color we set the color to gray.
The final histogram is shown in the following figure.
Example 1.2 [O]

Given 100 random numbers, use Origin® in order to plot a histogram of them.
For the purpose of this example and using the random number generation func-
tionality of Origin® we enter, in column A, 100 random numbers, normally dis-
tributed with a mean of 5 and a standard deviation of 1. The details need not concern
us here. We will explain how this is done in later stages of the book. All we need to
know now is that we have 100 numbers in column A. As in Fig. 1.2, we will assume
that the numbers given are the results of measurements of a length in mm. It should
be noted that the histogram is to be drawn using the raw data of the measurements,
not after they have been counted and grouped in classes as in Table 1.3.
We highlight column A by left-clicking on label A. Then press
Plot > Statistics > Histogram
The histogram shown in the adjacent figure appears. We will modify this
histogram.
1. Delete the text box containing A by left-clicking on the rectangle and then
pressing Delete.
2. We change the color of the bars by double-clicking on one of them and in the
window that opens pressing
Pattern > Fill > Color > LT Gray
3. In the same window, we change the gap between the bars
Spacing > Gaps Between Bars (in %) > 5
4. We change the labels of the axes: Double-click on the X-label and write Result
of the measurement, Dx (mm). Then Double-click on the Y-label and write
Frequency, DN.
5. We change the tick labels of the axes by double-clicking on one of them and
then
Scale > Horizontal > From 2 to 8.5 > Major Ticks, Value 1 > Minor Ticks,
Count 1
Scale > Vertical > From 0 to 25 > Major Ticks, Value 5 > Minor Ticks,
Count 4
Press Apply.
We want to show the tick labels on the right axis as well, so
Tick Labels > Right > Tick Show
Press OK.
6. We now draw the grid lines:
Grids > Horizontal > Major Grid Lines, Tick Show, Color Black, Style
Solid, Thickness 0.5 >
> Minor Grid Lines, Tick Show, Color Black, Style Solid, Thickness 0.5
Press OK. We will not draw vertical grid lines.
7. We modify Lines and Ticks:
Lines and Ticks > Top > Tick Show Line and Ticks > Major Ticks, Tick
Style In > Minor Ticks, Tick Style In
Then
Lines and Ticks > Right > Tick Show Line and Ticks > Major Ticks, Tick
Style Out > Minor Ticks, Tick Style Out
Press OK. We will not draw vertical grid lines. The final histogram is shown in
the figure below.
We export the graph for use as a jpg figure:
File > Export Graphs > Open Dialog… > Image Type, jpg > File Name,
give name > Path, give directory in which figure is to be stored > OK
Example 1.3 [P]

Using Python, plot the histogram of the 100 random numbers of Example 1.1 [E].
# We import the Python modules necessary for data analysis and
# graphics (numpy and matplotlib).
import numpy as np
import matplotlib.pyplot as plt
# We form a vector x with components the 100 random numbers of

#Example 1.1 [E].
x = np.array([4.78887, 4.17608, 4.08729, 6.26209, 4.52593, 4.62893, 5.90426,

5.75323, 4.07328, 5.63953, 4.91056, 4.73250, 5.72151, 4.24848, 4.90887, 4.67623,
5.90012, 4.65317, 4.79153, 5.45268, 5.30632, 4.26952, 5.65727, 5.18226, 4.38460,
4.54494, 4.00306, 5.97430, 4.22142, 4.96759, 7.08994, 4.81065, 4.53098, 4.86568,
5.47072, 4.96407, 7.11077, 6.69075, 5.16618, 6.30068, 5.29543, 3.61960, 2.94918,
4.22886, 2.71753, 5.54133, 4.13987, 4.10216, 5.91200, 5.47286, 5.15642, 5.56657,
5.77732, 5.83059, 4.28017, 5.15317, 6.12074, 6.88621, 4.91502, 4.86923, 7.11721,
5.74695, 4.20977, 6.94411, 4.69592, 5.76876, 4.21269, 3.37941, 4.93466, 5.82413,
4.82099, 4.07058, 4.02249, 4.94776, 4.45813, 4.43090, 4.65317, 4.74413, 5.48435,
5.03837, 6.11916, 3.35769, 7.19028, 5.31314, 4.23431, 6.12275, 4.03023, 4.59844,
5.67686, 5.74261, 4.48146, 4.97754, 4.75857, 4.30989, 4.77602, 3.41128, 5.17985,
5.09158, 3.14031, 3.74785])
# The histogram is produced with the following set of commands.
plt.hist(x, bins=10, align='mid', facecolor='grey')

plt.xlim(2, 8)
plt.xlabel('x (mm)')
plt.ylabel('Frequency')
plt.title('Histogram of x')
plt.show()
The histogram produced is shown here.
Example 1.4 [R]

Using R, plot the histogram of the 100 random numbers of Example 1.1 [E].
We form a vector x with components the numbers of Example 1.1 [E]:
> x <- c(4.78887, 4.17608, 4.08729, 6.26209, 4.52593, 4.62893, 5.90426, 5.75323,
4.07328, 5.63953, 4.91056, 4.73250, 5.72151, 4.24848, 4.90887, 4.67623, 5.90012,
4.65317, 4.79153, 5.45268, 5.30632, 4.26952, 5.65727, 5.18226, 4.38460, 4.54494,
4.00306, 5.97430, 4.22142, 4.96759, 7.08994, 4.81065, 4.53098, 4.86568, 5.47072,
4.96407, 7.11077, 6.69075, 5.16618, 6.30068, 5.29543, 3.61960, 2.94918, 4.22886,
2.71753, 5.54133, 4.13987, 4.10216, 5.91200, 5.47286, 5.15642, 5.56657, 5.77732,
5.83059, 4.28017, 5.15317, 6.12074, 6.88621, 4.91502, 4.86923, 7.11721, 5.74695,
4.20977, 6.94411, 4.69592, 5.76876, 4.21269, 3.37941, 4.93466, 5.82413, 4.82099,
4.07058, 4.02249, 4.94776, 4.45813, 4.43090, 4.65317, 4.74413, 5.48435, 5.03837,
6.11916, 3.35769, 7.19028, 5.31314, 4.23431, 6.12275, 4.03023, 4.59844, 5.67686,
5.74261, 4.48146, 4.97754, 4.75857, 4.30989, 4.77602, 3.41128, 5.17985, 5.09158,
3.14031, 3.74785)
> hist(x, main=''Histogram of x'', xlab=''x (mm)'', col=''grey'',

xlim=c(2, 8), las=1, breaks=7)
On pressing ENTER the histogram shown is produced.
Example 1.5 [E]

Using Excel®, plot a bar chart of the data in Table 1.2.
We enter the data ðxr ; nr Þ in columns A and B respectively. We highlight the cells
between A1 and B10.
We open the Insert window. From Recommended Charts we select the column
chart. The result is shown in the figure on the left.
We modify the bar chart in the same way as we did in Example 1.X [E]. The
final result is shown in the figure on the right.
Example 1.6 [O]

Using Origin®, plot a bar chart of the data in Table 1.2.
We enter the data ðxr ; nr Þ in columns A and B respectively. We highlight columns
A and B by left-clicking on label A and, holding the Shift or Control key down,
left-clicking on label B. Then select
Plot > Column/Bar/Pie > Column

The bar chart of ðxr ; nr Þ appears. We modify this bar chart by following the pro-
cedure of Example 1.2 [O]. The final result is shown in the figure.
Example 1.7 [P]

Using Python, plot a bar chart of the data in Table 1.2.
The categories are stored in the vector x and the frequency data in the vector y. The
width of the bar is defined in the variable width, and used in subsequent com-
mands to center the x-axis labels to the bars.
import numpy as np
import matplotlib.pyplot as plt
x = (9.5, 9.6, 9.7, 9.8, 9.9, 10, 10.1, 10.2, 10.3, 10.4)
y = [5, 7, 8, 10, 14, 15, 12, 12, 9, 8]
ind = np.arange(len(y))
width = 1/1.5
fig, ax = plt.subplots()
plt.bar(ind, y, width, color='grey')
ax.set_xticks(ind + width/2)
ax.set_xticklabels(x, ha='center')
plt.show()
The histogram shown is produced.
Example 1.8 [R]

Using R, plot a bar chart of the data in Table 1.2.
The categories of data are given in the second column, “9.5”, “9.6”, “9.7” etc. We
form a vector, x, with these as components. We also form a vector, Frequency,
with the frequencies of the x-values.
> xbarchart = data.frame(x=factor(c(''9.5'', ''9.6'', ''9.7'', ''9.8'', ''9.9'', ''10'',

''10.1'', ''10.2'', ''10.3'', ''10.4'')), levels = c(''9.5'', ''9.6'', ''9.7'', ''9.8'', ''9.9'',
''10'', ''10.1'', ''10.2'', ''10.3'', ''10.4''), Frequency = c(5, 7, 8, 10, 14, 15, 12, 12,
9, 8))
> barplot(xbarchart$Frequency, names = xbarchart$x, xlab=''x'', ylab=''Frequency'')
The chart shown on the right is produced.

1.2 Probability Density
If the height DN ¼ nr of each column of a histogram is divided by the total number

of measurements, N, and by the width Dx of the column, the probability density
1 DN
f ðxÞ ð1:1Þ
N Dx
is found. The same is achieved if the height of each column of a histogram of

relative frequencies is divided by its width, Dx. The term density is used, because
f ðxÞ gives the probability DN=N per unit x for a value to be found in the region of
x. The units of f ðxÞ are obviously the inverse of the units of x. Due to its definition,
if f ðxÞ is the probability density at a column of the histogram, then
DN
f ðxÞDx ¼ ¼ ðfraction of the measurements in the
N ð1:2Þ
particular column of the histogramÞ:
The sum Rf ðxÞDx over all the columns of the histogram will be equal to unity:
X 1X 1
f ðxÞDx ¼ DN ¼ N ¼ 1: ð1:3Þ
r
N r N
The histogram of f ðxÞ, therefore, has a total area equal to unity and is said to be
normalized.
The sum Rf ðxÞDx for the columns of the histogram between the limits of x1 and
x2 gives the fraction of the values of x which lie between x1 and x2 , or the prob-
ability for a value x of the magnitude x to lie between x1 and x2 .
The shape of the histogram obviously depends on the width Dx of its classes. If
we have a large number of values x of the magnitude x, the width may be chosen to
be very small and then the columns of the histogram tend to define a continuous
function f ðxÞ for the probability density, which is known as the probability density
function (also known as the density function or the frequency function, or, simply,
the density) of the values x of the random variable x. The graph of f ðxÞ is called
(normalized) frequency curve (Fig. 1.3). The function f ðxÞ, as probability per unit
range, takes only positive values.
The normalization condition is now,
Z 1
f ðxÞ dx ¼ 1: ð1:4Þ
1
1.2 Probability Density 15
Fig. 1.3 The transition of a histogram a to a frequency curve f ðxÞ b, as the width of the classes
tends to zero
Also, the partial sum of columns, mentioned above, takes the form
Z x2
f ðxÞdx Prfx1 x\x2 g ð1:5Þ
x1
and gives the probability for a value of x to lie between x1 and x2 . If the function
f ðxÞ is proportional to a function gðxÞ, which is not normalized, then the normalized
f ðxÞ is given by
gðxÞ
f ðxÞ ¼ R 1 ð1:6Þ
1 gðxÞdx
and
R x2
gðxÞdx
Prfx1 x\x2 g ¼ R 1
x1
ð1:7Þ
1 gðxÞdx
1.3 Mean Value
The mean of N measurements xi (i ¼ 1; 2; . . .; N) is defined as
1X N
x xi : ð1:8Þ
N i¼1
If the measurements can be grouped in a number of K classes in which the value

P
N P
K
xr (r ¼ 1; 2; . . .; K) has a frequency nr , then it will be xi ¼ nr xr and the mean
i¼1 r¼1
is given by
1X K
x¼ nr x r : ð1:9Þ
N r¼1
Since nr =N ¼ fr is the relative frequency of the value xr , this expression may

also be written as
X
K
x¼ f r xr : ð1:10Þ
r¼1
This quantity is also known as expected value or mathematical expectation EðxÞ

of the variable x, especially in the form of Eq. (1.10).
The weighted mean is examined in Sect. 9.4.
Example 1.9
Evaluate the mean of the measurements of Table 1.1, as these were grouped in
Table 1.2.
We will find the result using both Eqs. (1.9) and (1.10), so we construct two tables
with columns: r, xr , nr and fr , for r = 1, 2, … 10. We evaluate the products nr xr and
fr xr for each value of r.
r xr nr nr xr r xr fr f r xr
(mm) (mm) (mm) (mm)
1 9.5 5 47.5 1 9.5 0.05 0.475
2 9.6 7 67.2 2 9.6 0.07 0.672
3 9.7 8 77.6 3 9.7 0.08 0.776
4 9.8 10 98.0 4 9.8 0.10 0.980
5 9.9 14 138.6 5 9.9 0.14 1.386
6 10.0 15 150.0 6 10.0 0.15 1.500
7 10.1 12 121.2 7 10.1 0.12 1.212
8 10.2 12 122.4 8 10.2 0.12 1.224
9 10.3 9 92.7 9 10.3 0.09 0.927
10 10.4 8 83.2 10 10.4 0.08 0.832
Σ= 100 998.4 Σ= 1 9.984
P
K 1X K
The sum nr xr ¼ 998:4 mm gives x ¼ nr xr ¼ 9:984 mm.
r¼1 N r¼1
Another random document with
no related content on Scribd:
CHAPTER IV
SPACE AND TIME

§ 1. Are time and space ultimately real or only phenomenal? § 2. The space and
time of perception are limited, sensibly continuous, and consist of a
quantitative element together with a qualitative character dependent on
relation to the here and now of immediate individual feeling. § 3. Conceptual
space and time are created from the perceptual data by a combined process
of synthesis, analysis, and abstraction. § 4. They are unlimited, infinitely
divisible, and there is valid positive ground for regarding them as
mathematically continuous. Thus they form infinite continuous series of
positions. They involve abstraction from all reference to the here and now of
immediate feeling, and are thus homogeneous, i.e. the positions in them are
indistinguishable. They are also commonly taken to be unities. § 5. Perceptual
space and time cannot be ultimately real, because they involve reference to
the here and now of a finite experience; conceptual space and time cannot be
ultimately real, because they contain no principle of internal distinction, and
are thus not individual. § 6. The attempt to take space and time as real leads
to the difficulty about qualities and relations, and so to the indefinite regress. §
7. Space and time contain no principle of unity; there may be many space and
time orders in the Absolute which have no spatial or temporal connection with
each other. § 8. The antinomies of the infinite divisibility and extent of space
and time arise from the indefinite regress involved in the scheme of qualities
and relations, and are insoluble so long as the space and time construction is
taken for Reality. § 9. The space and time order is an imperfect phenomenal
manifestation of the logical relation between the inner purposive lives of finite
individuals. Time is an inevitable aspect of finite experience. How space and
time are transcended in the Absolute experience we cannot say.
§ 1. The problems which arise for the metaphysician from the fact
that the physical order, as it is presented to our senses, consists of
elements having position in space and time, are among the oldest
and most perplexing of all the riddles suggested by the course of our
experience. Adequate discussion of them would demand not only far
more space than we are at liberty to bestow on the topic, but such a
familiarity with the mathematical theory of order and series as is
scarcely possible to any one but an original mathematician. All that
we can do in the present chapter is to deal very superficially with one
or two of the leading problems, more with a view to indicating the
nature of the questions which Metaphysics has to face, than of
providing definite answers to them.
The fundamental problem for Metaphysics is, of course, whether
space and time are ultimate Realities or only appearances; that is,
would the whole system of Reality, as directly apprehended by an
absolute all-containing experience, wear the forms of extension and
succession in time, or is it merely a consequence of the limitations of
our own finite experience that things come to us in this guise? It may
indeed be urged that the contents of the universe must form an order
of some sort for the absolute experience, in virtue of their systematic
unity, but even so it is not clear that order as such is necessarily
spatial or temporal. Indeed, most of the forms of order with which we
are acquainted, both in everyday life and in our mathematical
studies, appear to be, properly speaking, both non-spatial and non-
temporal. Thus, e.g., it is seemingly by a mere metaphor that we
speak of the “successive” integers of the natural number-series, the
“successive” powers of an algebraical symbol, the “successive”
approximations to the value of a continued fraction, in language
borrowed from the temporal flow of events, the true relation involved
being in the first two cases the non-temporal one of logical
derivation, and in the third the equally non-temporal one of
resemblance to an ideal standard. The full solution of the
metaphysical problem of space and time would thus involve (1) the
discrimination of spatial and temporal order from other allied forms of
order, and (2) a decision as to the claim of this special form of order
to be ultimately coherent and intelligible.
The problem thus presented for solution is often, and usually with
special reference to the Kantian treatment of space and time in the
Transcendental Æsthetic, put in the form of the question whether
space and time are subjective or objective. This is, however, at best
a misleading and unfortunate mode of expression which we shall do
well to avoid. The whole distinction between a subjective and an
objective factor in experience loses most of its significance with the
abolition, now effected by Psychology, of the vicious Kantian
distinction between the “given” in perception and the “work of the
mind.” When once we have recognised that the “given” itself is
constituted by the movement of selective attention, it becomes
impossible any longer to distinguish it as an objective factor in
knowledge from the subjective structure subsequently raised upon it.
Kant’s adherence to this false psychological antithesis so completely
distorts his whole treatment of the “forms of intuition,” that it will be
absolutely necessary in a brief discussion like our own to deal with
the subject in entire independence of the doctrines of the Æsthetic,
which unfortunately continue to exercise a disproportionate influence
on the current metaphysical presentment of the problem.[141] It should
scarcely be necessary to point out that the metaphysical questions
have still less to do with the psychological problems, so prominent in
recent science, of the precise way in which we come by our
perception of extension and succession. For Metaphysics the sole
question is one not of the origin but of the logical value of these
ideas.
It is of fundamental importance for the whole metaphysical
treatment of the subject, to begin by distinguishing clearly between
space and time as forms of perception, and space and time as
conceptual forms in which we construct our scientific notion of the
physical order. One chief source of the confusions which beset the
Kantian view is the neglect of Kant and most of his followers to make
this distinction with sufficient clearness. We cannot insist too strongly
upon the point that the space and the time of which we think in our
science as containing the entire physical order, are not space and
time as directly known to us in sense-perception, but are concepts
elaborated out of the space and time of direct perception by a
complicated process of synthesis and analysis, and involving
abstraction from some of the most essential features of the space
and time of actual experience. The following brief discussion may
serve to illustrate the general nature of the relation between the two
forms of space and time, and to exhibit the leading differences
between them.
§ 2. Perceptual Space and Time. Both space and time, as we are
aware of them in immediate perception, are (1) limited. The space
we actually behold as we look out before us with a resting eye is
always terminated by a horizon which has a more or less well-
defined outline; the “specious present,” or portion of duration of
which we can be at any time aware at once as an immediately
presented content, has been shown by elaborate psychological
experimentation to have a fairly well-defined span. Whatever lies
outside this “span of attention” belongs either to the no longer
presented past or to the not yet presented future, and stands to the
sensible present much as the space behind my back to the actually
beheld space before my eyes. Of course, in either case the limits of
the actually presented space or time are not absolutely defined. To
right and left of the line of vision the visible horizon gradually fades
off into the indistinctly presented “margin of consciousness”; the
“sensible present” shades away gradually at either end into the past
and the future. Yet, though thus not absolutely defined, sensible
space and time are never boundless.
(2) Perceptual space and time are both internally sensibly
continuous or unbroken. Concentrate your attention on any lesser
part of the actually seen expanse, and you at once find that it is itself
an expanse with all the characteristics of the wider expanse in which
it forms a part. Space as actually seen is not an aggregate of minima
visibilia or perceptual points in which no lesser parts can be
discriminated; so long as space is visually or tactually perceived at
all, it is perceived as containing lesser parts which, on attending to
them, are found to repeat the characteristics of the larger space. So
any part of the “specious present” to which special attention can be
directed, turns out itself to be a sensible duration. Perceived space is
made of lesser spaces, perceived time of lesser times; the “parts”
not being, of course, actually distinguished from each other in the
original percept, but being capable of being so distinguished in
consequence of varying movements of attention.
(3) On investigating the character of our actual perception of
space and time, it appears to contain two aspects, which we may call
the quantitative and the qualitative. On the one hand, whenever we
perceive space we perceive a certain magnitude of extension,
whenever we perceive time we perceive a longer or shorter lapse of
duration. Different spaces and different times can be quantitatively
compared in respect of the bigness of the extension or the duration
comprised in them. On the other hand, the percept of space or time
is not one of mere extension or duration. It has a very different
qualitative aspect. We perceive along with the magnitude of the
extension the form of its outline. This perception of spatial form
depends in the last resort upon perception of the direction assumed
by the bounding line or lines. Similarly, in dealing with only one
dimension of perceived space, we never perceive length (a spatial
magnitude) apart from the perception of direction (a spatial quality).
The same is true of the perception of time. The lapses of duration we
immediately perceive have all their special direction-quality; the
“specious present” is essentially a simultaneously presented
succession, i.e. a transition from before to after. It must be added
that, in perceptual space and time, the directions thus perceived
have a unique relation to the perceiving subject, and are thus all
qualitatively distinct and irreversible. Direction in space is estimated
as right, left, up, down, etc., by reference to axes through the centre
of the percipient’s body at right angles to each other, and is thus for
any given moment of experience uniquely and unambiguously
determined. Direction in time is similarly estimated with reference to
the actual content of the “focus of consciousness.” What is actually
focal is “now,” what is ceasing to be focal is “past,” what is just
coming to be focal is “future” in its direction.[142]
This is perhaps the most fundamental and important peculiarity of
the space and time of actual perception. All directions in them are
unambiguously determined by reference to the here and now of the
immediate experience of an individual subject. As a consequence,
every individual subject has his own special perceptual space and
time; Geometry and Mechanics depend, to be sure, on the possibility
of the establishment of correspondences between these spatial and
temporal systems, but it is essential to remember that, properly
speaking, the space and time system of each individual’s perception
is composed of directions radiating out from his unique here and
now, and is therefore individual to himself.[143]
§ 3. The Construction of the Conceptual Space and Time Order of
Science. For the purposes of practical life, no less than for the
subsequent object of scientific description of the physical order, it is
indispensably necessary to establish equations or correspondences
between the individual space and time systems of different
percipients. Apart from such correspondences, it would be
impossible for one subject to translate the spatial and temporal
system of any other into terms of his own experience, and thus all
practical intercourse for the purpose of communicating directions for
action would come to an end. For the communication of such
practical directions it is imperative that we should be able mentally to
reconstruct the spatial and temporal aspects of our experience in a
form independent of reference to the special here and now of this or
that individual moment of experience. Thus, like the rest of our
scientific constructions, the establishment of a single conceptual
space and time system for the whole of the physical order is
ultimately a postulate required by our practical needs, and we must
therefore be prepared to face the possibility that, like other
postulates of the same kind, it involves assumptions which are not
logically defensible. The construction is valuable, so far as it does its
work of rendering intercommunication between individuals possible;
that it should correspond to the ultimate structure of Reality any
further than the requirements of practical life demand is superfluous.
The main processes involved in the construction of the conceptual
space and time of descriptive science are three,—synthesis,
analysis, abstraction. (a) Synthesis. Psychologically speaking, it is
ultimately by the active movements of individual percipients that the
synthesis of the individual’s various perceptual spaces into one is
effected. As attention is successively directed, even while the body
as a whole remains stationary, to different parts of the whole
expanse before the eye, the visual space which was originally “focal”
in presentation becomes “marginal,” and the “marginal” focal by a
sensibly gradual transition. When to the movements of head and
eyes which accompany such changes in attention there are added
movements of locomotion of the whole body, this process is carried
further, and we have the gradual disappearance of originally
presented spaces from presentation, accompanied by the gradual
emergence of spaces previously not presented at all. This leads to
the mental construction of a wider space containing all the
individual’s different presentation-spaces, the order in which it
contains them being determined by the felt direction of the
movements required for the transition from one to another.
As we learn, through intercommunication with our fellows, of the
existence for their perception of perceptual extension never directly
presented to our own senses, the process of synthesis is extended
further, so as to comprise in a single spatial system all the
presentation-spaces of all the individual percipients in an order once
again determined by the direction of the movements of transition
from each to the others. Finally, as there is nothing in the principle of
such a synthesis to impose limits upon its repetition, we think of the
process as capable of indefinite continuance, and thus arrive at the
concept of a space stretching out in all directions without definite
bounds. This unending repetition of the synthesis of perceived
spaces seems to be the foundation of what appears in theory as the
Infinity of Space.
Precisely similar is the synthesis by which we mentally construct a
single time system for the events of the physical order. Now means
for me the content which occupies the centre of attentive interest. As
attention is concentrated on the different stages in the realisation of
an interest, this centre shifts; what was central becomes first
marginal and then evanescent, what was marginal becomes central.
Hence arises the conception of the events of my own inner life as
forming a succession of moments, with a determinate order, each of
which has been a now, or point of departure for directions in
perceptual time, in its turn. As with space so with time, the
intrasubjective intercourse of man with man makes it possible for me
mentally to extend this conceptual synthesis of moments of time so
as to include nows belonging to the experience of others which were
already past before the first now of their experiences which I can
synchronise with a now of my own, and again nows of their
experiences relatively to which the last now which synchronises with
one of my own is past. The indefinite repetition of such a synthesis
leads, as before with space, to the thought of a duration reaching out
endlessly into past and future, and thus gives us the familiar concept
of the Infinity of Time.[144]
(b) Analysis. Equally important is the part played by mental
analysis in the formation of the conceptual space and time system.
As we have already seen, successive attention to lesser parts of a
presented extension, or a presented lapse, reveals within each
lesser part the same structure which belongs to the whole, and thus
establishes the sensible continuity of space and time. In actual fact,
the process of attending successively to smaller and yet smaller
portions of space and time cannot, of course, be carried on
indefinitely, but we can conceptually frame to ourselves the thought
of the indefinite repetition of the process beyond the limits arbitrarily
imposed on it by the span of our own attention. Thus, by an act of
mental analysis, we arrive at the concept of space and time as
indefinitely divisible, or possessed of no ultimately unanalysable last
parts, which is an indispensable pre-requisite of Geometry and
Dynamics.
This indefinite divisibility of conceptual space and time is not of
itself enough, as is often supposed, to establish their continuity in the
strict mathematical sense of the word; their continuity depends upon
the further assumption that whatever divides a series of positions in
space or events in time unambiguously into two mutually exclusive
classes, is itself a position in the space or event in the time series.
This assumption does not seem to be absolutely requisite for all
scientific treatment of the problems of space and time,[145] but is
demanded for the systematic establishment of the correspondence
between the spatial and temporal series and the continuous series of
the real numbers. Moreover, it seems impossible to assign any
positive content to the notion of a something which should bisect the
spatial or temporal order without occupying a position in that order.
Hence we seem inevitably led by the same analytical process which
conducts us to the conception of the spatial and temporal orders as
infinite series to think of them also as continuous series in the strict
sense of the term. The alternative conception of them as
discontinuous, if not absolutely excluded, does not seem to be called
for by any positive motive, and is incompatible with the complete
execution of the purposes which demand application of the number-
series to a spatial or temporal content.
(c) Abstraction. The part played by abstraction in the formation of
the conceptual space and time order out of the data of perception is
often overlooked by theorists, but is of fundamental importance, as
we shall see immediately. We have already learned that the most
significant fact about the time and space order of individual
experience is that its directions are unique, because they radiate out
from the unique here and now of immediate feeling. In the
construction of the conceptual space and time order we make entire
abstraction from this dependence on the immediate feeling of a
subject. Conceptual space contains an infinity of positions, but none
of them is a here; conceptual time an infinity of moments, but none
of them is a now. As the time and space of the conceptual order are
taken in abstraction from the differences between individual points of
view, no one point in either can be regarded as having more claim
than any other to be the natural “origin of co-ordinates” with
reference to which directions are estimated. We shall have repeated
opportunity in the remainder of this chapter to observe how important
are the consequences of this abstraction.
Abstraction also enters in another way into the construction by
which conceptual space and time are created. Actual perceived
space and time are indeed never empty, but always filled with a
content of “secondary” qualities. In other words, they are always one
aspect of a larger whole of fact. Extension is never perceived apart
from some further visual or tactual quality of the extended, temporal
lapse never perceived without some change in presented content,
however slight. But in constructing the conceptual space and time
system, we abstract altogether from this qualitative aspect; we think
solely of the variety of positions and directions in time and space
without taking any account of the further qualitative differences with
which they are accompanied in concrete experience. Thus we come
by the notion of an empty space and an empty time as mere systems
of positions into which various contents may subsequently be put.
Strictly speaking, the notion of an empty space or an empty time is
unmeaning, as the simple experiment of thinking of their existence is
sufficient to show. We cannot in thought successfully separate the
spatial and temporal aspects of experience from the rest of the
whole to which they belong and take them as subsisting by
themselves, any more than we can take timbre as subsisting apart
from musical pitch or colour-tone from saturation. We can, however,
confine our attention to the spatial-temporal system of positions
without taking into account the special secondary properties of the
extended and successive. It is from this logical abstraction that the
illusion arises when we imagine an empty set of spatial and temporal
positions as having first to exist in order that they may be
subsequently “filled” with a variety of contents.[146]
§ 4. Characteristics of the Conceptual Time and Space Order. The
following characteristics of the conceptual space and time created by
the construction we have just examined, call for special notice.
Conceptual space and time are necessarily taken, for reasons
already explained, to be unlimited, and indefinitely divisible. Though
it does not seem inevitable that they should be continuous, we
appear to be unable to attach any positive meaning to the notion of
their discontinuity, and, in the practical need for the application to
them of the complete number-series, we have a valid positive ground
for taking them as continuous. But space and time are thus resolved,
in the process of their conceptual construction, into continuous
infinite series of which the terms are spatial and temporal positions
or points. Unlike the parts of perceptual space and time, these
conceptual terms are not themselves spaces or times, as they
contain no internal multiplicity of structure. Conceptual space and
time are thus not wholes or aggregates of parts, but systems of
relations between terms which possess no quantitative character.
Between any two terms of the spatial, or again of the temporal,
series there is one unique relation, which is completely determined
by the assignment of the terms, their distance. In the temporal
series, which has only one dimension, you can only pass from any
one given term to any other through a series of intermediate terms
which is once and for all determined when the initial and final terms
are given, hence nothing is required beyond the terms themselves to
fix their distance. The spatial series is multi-dimensional, i.e. you can
pass from any one term in it to any second by an indefinite variety of
routes through intermediate terms, but it is still true that there is one
and only one such route which is completely determined when the
terms in question are known, namely, the straight line passing
through both. This straight line constitutes the unique distance of the
two points from each other.[147] Thus the genuine concept of which
those of space and time are species is not that of magnitude or
quantity, but of serial order.
Further, and this is a point of fundamental difference between
conceptual space and time, and the spaces and times of immediate
perception, any one position in either order, taken by itself, is
qualitatively indistinguishable from any other. All points of space, all
moments of time, are alike, or, as it is also phrased, conceptual
space and time are homogeneous throughout. It is not until you take
at least two terms of the spatial or temporal series and consider the
relation they determine, that distinction becomes possible. This
homogeneity of conceptual space and time is an inevitable
consequence of the abstraction from the immediate feelings of the
individual subject of experience involved, as we saw, in the process
of their construction. In our actual perception of spatial and temporal
extension, that part of perceived space and time which stands in
direct unity with immediate feeling is qualitatively distinguished as
the here and now from all the rest, and thus does not depend upon
the specification of a second spatial or temporal position for its
recognisability. Here is where I am, now is this felt present. And
similarly, every other part of the actually presented space and time
gets a unique qualitative character from its special relation to this
here and now; it is right or left, behind or in front, before or after.
When we abstract altogether from the unique relation with individual
experience which thus makes the here and now of perception, as we
do in constructing our conceptual space and time order, every
position alike becomes the mere possibility of a here or a now, and
as such mere possibilities the various positions are indistinguishable.
Practically, this homogeneity is important as the indispensable
condition for the quantitative comparison of different portions of
extension or duration.
An apparently inevitable consequence of the homogeneity of
conceptual space and time is the relativity of spatial and temporal
position. As we have seen, positions in conceptual space and time
are not distinguishable until you take them in pairs. In other words, to
fix one position in space or one date you have to give its relation to
another position or date, and similarly to fix this you must specify a
third, and so on indefinitely. To say where A is means to say how you
get to it from B, and B again is only known by the way it is reached
from C, and so on without end. Logically, this is a simple
consequence of the nature of space and time as conceptually
analysed into endless series. To specify any term in the series you
must give the unique relation it bears to some other term, its logical
distance. And, in a series which has neither first nor last term, this
second term cannot be defined except by its logical distance from a
third. In actual perception this difficulty is avoided, owing to the fact
that immediate feeling gives us the here and now from which all our
directions are measured. But in conceptual space or time there is
nothing to distinguish any one here which we may take as our “origin
of co-ordinates,” or any one now which we take as our present from
any other, and hence the endless regress seems inevitable.
It follows, of course, that in conceptual space and time there is no
principle by which to distinguish different directions. In perception
they can be distinguished as right and left, up and down, and so
forth. But since what is right to one percipient is left to another, in
conceptual space, where complete abstraction is made from the
presence of an individual percipient, there is neither right nor left, up
nor down, nor any other qualitative difference between one direction
and another, all such differences being relative to the individual
percipient. When we wish to introduce into conceptual space
distinctions between directions, we always have to begin by
arbitrarily assigning some standard direction as our point of
departure. Thus we take, e.g., an arbitrarily selected line ——— as
A B
such a standard for a given plane, and proceed to distinguish all
other directions by the angle they make with A B and the sense in
which they are estimated (whether as from B to A or from A to B).
But both the line A B and the difference of sense between A B and B
A can only be defined by similar reference to some other standard
direction, and so on through the endless regress.
Similarly with conceptual time. Here, as there is only one
dimension, the difficulty is less obvious, but it is no less real. In
conceptual time there is absolutely no means of distinguishing
before from after, past from future. For the past means the direction
of our memories, the direction qualified by the feeling of “no longer”;
the future is the direction of anticipation and purposive adaptation,
the direction of “not yet.” And, apart from the reference given by
immediate feeling to the purposive life of an individual subject, these
directions cannot be discriminated. In short, conceptual time and
space are essentially relative, because they are systems of relations
which have no meaning apart from qualitative differences in the
terms which they relate; while yet again, for the purpose of the
conceptual construction which yields them, the terms have to be
taken as having no character but that which they possess in right of
the relations.[148]
One other feature of the space and time construction is sufficiently
important to call for special mention. Space and time are commonly
thought of as unities of some kind. All spatial positions, it is usually
assumed, fall within one system of space-relations; all dates have
their place in one all-inclusive time. This character of unity completes
the current conception of the spatial and temporal order. Each of
those orders is a unity, including all possible spatial or temporal
positions; each is an endless, infinite, continuous series of positions,
which all are purely relative. There are other peculiarities, especially
of the current concepts of space, with which it is not necessary to
deal here, as they are of an accidental kind, not arising out of the
essential nature of the process by which the conception is
constructed. Thus it is probably a current assumption that the
number of dimensions in space is three and no more, and again that
the Euclidean postulate about parallels is verified by its constitution.
As far as perceptual space is concerned, those assumptions
depend, I presume, upon empirical verification; there seems to be no
reason why they should be made for the conceptual space-order,
since it is quite certain that a coherent science of spatial relations
can be constructed without recourse to them.[149]
§ 5. The question now is, whether the whole of this spatial and
temporal construction is more than imperfect, and therefore
contradictory, appearance. I will first state in a general form the
arguments for regarding it as appearance, and then proceed to
reinforce this conclusion by dealing with some special difficulties.
Finally, I propose to ask whether we can form some positive
conception of the higher order of Reality of which the spatial and
temporal series are phenomenal.
That the space and time order is phenomenal and not ultimate,
can, I think, be conclusively shown by a general argument which I
will first enunciate in principle and then develop somewhat more in
detail. An all-comprehensive experience cannot apprehend the detail
of existence under the forms of space and time for the following
reason. Such an experience could be neither of space and time as
we perceive them, nor of space and time as we conceptually
reconstruct them. It would not be of perceptual space and time,
because the whole character of our perceptual space and time
depends upon the very imperfections and limitations which make our
experience fragmentary and imperfect. Perceptual space and time
are for me what they are, because I see them, so to say, in
perspective from the special standpoint of my own particular here
and now. If that standpoint were altered, so that what are actually for
me there and then became my here and now, my whole outlook on
the space and time order would suffer change. But the Absolute
cannot look at the space and time order from the standpoint of my
here and now. For it is the finitude of my interests and purposes
which confine me in my outlook to this here and now. If my interests
were not bound up in the special way in which they are with just this
special part or aspect of the life of a wider whole, if they were co-
extensive with the life of that whole, every place and every time
would be my here and now. As it is, here is where my body is, now is
this particular stage in the development of European social life,
because these are the things in which I am primarily interested. And
so with all the other finite experiences in which the detail of the
absolute experience finds expression. Hence the absolute
experience, being free from the limitations of interest which condition
the finite experiences, cannot see the order of existence from the
special standpoint of any of them, and therefore cannot apprehend it
under the guise of the perceptual space and time system.
Again, it cannot apprehend existence under the forms of space
and time as we conceptually reconstruct them. For Reality, for the
absolute experience, must be a complete individual whole, with the
ground of all its differentiations within itself. But conceptual space
and time are constructed by deliberate abstraction from the relation
to immediate experience implied in all individuality, and
consequently, as we have just seen, they contain no real principle of
internal distinction, their constituent terms being all exactly alike and
indistinguishable. In short, if the perceptual time and space systems
of our concrete experience represent individual but imperfect and
finite points of view, the conceptual space and time of our scientific
construction represents the mere abstract possibility of a finite point
of view; neither gives a point of view both individual and infinite, and
neither, therefore, can be the point of view of an absolute
experience. An absolute experience must be out of time and out of
space, in the sense that its contents are not apprehended in the form
of the spatial and temporal series, but in some other way. Space and
time, then, must be the phenomenal appearance of a higher reality
which is spaceless and timeless.
§ 6. In principle, the foregoing argument appears to me to be
complete, but, for the sake of readers who care to have its leading
thought more fully developed, it may be re-stated thus. Perceptual
space and time cannot be ultimately real as they stand. They are
condemned already by the old difficulty which we found in the notion
of reality as made up of qualities in relation. Perceptual space and
time are aggregates of lesser parts, which are themselves spaces
and times; thus they are relations between terms, each of which
contains the same relation once more in itself, and so imply the now
familiar indefinite regress.[150] Again, when we try in our conceptual
space and time construction to remedy this defect by reducing space
and time altogether to mere systems of relations, the difficulty turns
out to have been merely evaded by such a process of abstraction.
For, so long as we keep rigidly to our conceptual construction, the
terms of our relations are indistinguishable. In purely conceptual
space and time, as we have seen, there is no possibility of
distinguishing any one direction from any other, since all are
qualitatively identical.
Indeed, it is obvious from first principles that when the sets of
terms between which a number of relations of the same type holds
are indistinguishable, the relations cannot be discriminated. To
distinguish directions at all, we must, in the end, take at least our
starting-point and one or more standard directions reckoned from it
—according to the number of dimensions with which we are dealing
—as independently given, that is, as having recognisable qualitative
differences from other possible starting-points and standard
directions. (Thus, to distinguish before and after in conceptual time,
you must at least assume some moment of time, qualitatively
recognisable from others, as the epoch from which you reckon, and
must also have some recognisable qualitative distinction between
the direction “past” and the direction “future.”) And with this reference
to qualitative differences we are at once thrown back, as in the case
of perceptual time and space, on the insoluble old problem of Quality
and Relation. The assumed starting-point and standard directions
must have qualitative individuality, or they could not be
independently recognised and made the basis for discrimination
between the remaining directions and positions: yet, because of the
necessary homogeneity of the space and time of conceptual
construction, they cannot have any such qualitative individuality, but
must be arbitrarily assumed. They will therefore themselves be
capable of determination only by reference to some other equally
arbitrary standard, and thus we are once more committed to the
indefinite regress. The practical usefulness of these constructions
thus depends on the very fact that we are not consistent in our use
of them. In all practical applications we use them to map out the
spatial and temporal order of events as seen in perspective from a
standpoint which is, as regards the conceptual time and space order
itself, arbitrary and indistinguishable from others.
§ 7. Instead of further elaborating this general argument, a task
which would be superfluous if its principle is grasped, and
unconvincing if it is missed, I will proceed to point out one or two
special ways in which the essential arbitrariness of the spatial and
temporal construction is strikingly exemplified. To begin with, a word
may be said about the alleged unity of space and time. It is
constantly taken for granted, by philosophers as well as by practical
men, that there can be only one spatial and one temporal order, so
that all spatial relations, and again all temporal relations, belong to
the same system. Thus, if A has a spatial relation to B and C to D, it
is assumed that there must be spatial relations between A and C, A
and D, and B and C, B and D. Similarly if A is temporally related with
B, and C with D. This view is manifestly presupposed in the current
conception of Nature, the “physical universe,” the “physical order,” as
the aggregate of all processes in space and time. But there seems to
be no real logical warrant for it. In principle the alleged unity of all
spatial and temporal relations might be dismissed, on the strength of
the one consideration that space and time are not individual wholes,
and therefore can contain no principle of internal structural unity.
This is manifest from the method by which the space and time of our
conceptual scheme have been constructed. They arose, as we saw,
from the indefinite repetition of a single type of relation between
terms in which we were unable to find any ultimately intelligible
principle of internal structure. But unity of structure cannot be
brought into that which does not already possess it by such mere
endless repetition. The result of such a process will be as internally
incoherent and devoid of structure as the original data. Hence space
and time, being mere repetitions of the scheme of qualities in
relation, cannot be true unities.
This becomes clearer if we reflect on the grounds which actually
warrant us in assigning position in the same space and the same
time to a number of events. For me A and B are ultimately in the
same space when there is a way of travelling from A to B; they are in
the same time when they belong to different stages in the
accomplishment of the same systematic purposes. Thus in both
cases it is ultimately from relation to an identical system of purposes
and interests that different sets of positions or events belong to one
space or one time. The unity of such a space or time is a pale
reflection in abstract form of the unity of a life of systematic purpose,
which is one because it has unique individual structure. It is in this
way, from the individual unity of the purpose and interests of my
ordinary waking life, that I derive the right to refer its experiences to
a single space and time system. Similarly, it is in virtue of the
inclusion of my own and my fellow-men’s purposes in a wider whole
of social systematic purpose that I can bring the space and time
relations of their experience into one system with my own. And
again, the sensible occurrences of the physical order belong to one
space and time with the space and time relations of human
experience, because of the varying ways in which they condition the
development of our own inner purposive life. But there are cases,
even within our own conscious life, where this condition appears to
be absent, and in these cases we do not seem to be able to make
intelligible use of the conception of a single time or a single space.
Take the case of our dreams. The events of my dreams stand in
spatial and temporal relations within the dream itself, but there would
be no sense in asking what are the spatial relations between the
places seen in my dreams and the places marked on the map of
England; or what are again the temporal relations between the
events of last night’s dream and those of this morning, or those of
the dreams of last week. Precisely because there is usually no
systematic identity of purpose connecting the dream with the waking
life or with other dreams, the time and space of the dream have no
position with respect to the time and space system of waking life, nor
those of one dream with relation to those of another.[151] Of course, it
may be said that the dream-space and dream-time are “imaginary,”
but the problem cannot be got rid of by the use of an epithet. To call
them imaginary is merely to say that they are not systematically
connected with the time and space of waking life, not to disprove
their genuineness as actual space and time constructions.
Similarly, if there are intelligent purposes of which our human
purposive life is debarred from taking account as such, as we urged
that there must be behind the phenomenal physical order, the time
and space within which those purposes are conceived and executed
would have no place in our spatial and temporal system. The
phenomenal events of the physical order would fall within our
system, but not the life of inner purpose of which that order is the
manifestation to our senses. Ultimately, in fact, all spaces and all
times could only form one spatial and temporal system on condition
that the infinite absolute experience views all its contents in spatial
and temporal form; then the various space and time systems
corresponding to the purposes of the various groups of finite
individuals would finally, for the infinite individual, form one great
system of time and space relations. But we have already seen that
the infinite experience cannot comprehend its contents in spatial or
temporal forms.
We infer, then, that there may be—indeed, if our interpretation of
the physical order is valid, there must be—a plurality of spaces and
times within the Real. Within any one such space or time all its
members are spatially and temporally interrelated, but the various
spaces are not themselves related in space, nor the various times
before or after one another in time. Their relation is the purely logical
one of being varying modes of the expression in a finite detail of the
underlying nature of the ultimate Reality.[152] For the absolute
experience they must be all at once and together, not in the sense of
being in “one space and time,” but in the sense of forming together
the systematic embodiment of one coherent ground or principle.
§ 8. Similar consequences, as to the phenomenal character of
space and time, follow from the consideration of the familiar Kantian
antinomies founded upon the concept of spatial and temporal infinity.
Space and time must be externally boundless and internally
indefinitely divisible, and yet again cannot be either. Freed from
unessential accessories, the argument for either side of the antinomy
may be stated thus. Space and time must be boundless because all
spatial and temporal existence means spatial and temporal relation
to a second term, itself similarly related to a third term. For precisely
the same reason both must be indefinitely divisible. Yet again, they
can be neither, since only the individual exists, and within such an
interminable network of relations between terms which are nothing
but the supporters of these relations there is no principle of individual
structure.[153] Thus the Kantian antinomies are a simple consequence
of the old difficulty about quality and relation. Space and time must
be mere relations, and the terms of those relations therefore
qualitatively indistinguishable; again, since they are relations they
cannot be relations between nothings or, what is the same thing,
between terms with no individual character. As in all cases where the
problem of relation and quality arises, it then conducts us to the
indefinite regress.
So long as we continue to look upon space and time as real, we
have therefore to choose between two equally illogical alternatives.
We must either arbitrarily refuse to continue the indefinite regress
beyond the point at which its difficulties become apparent, as is done
by the assertion that space and time have finite bounds or indivisible
parts, or we must hold that the absolute experience actually
achieves the summation of an unending series. With the recognition
that space and time are phenomenal, the result of a process of
construction forced on us by our practical needs, but not adequately
corresponding to the real nature of individual existence, the difficulty
disappears. Both sides of the antinomy become relatively true, in the
sense that for our practical purposes we must be content to adopt
now the one and again the other; both become ultimately untrue in
the sense that space and time, being constructions of our own, are
really neither finite nor infinite series, but are the one or the other
according to the purposes for which we use our construction.
§ 9. If spatial and temporal position and direction must thus in the
end be appearance, phenomenal of some more individual reality, we
have finally to ask, Of what are they the appearance? It is not
enough to say “of ultimate Reality,” or “of the Absolute.” Ultimately
this is, no doubt, true of space and time, as it is of everything else,
but we desire further to know if they are not proximately the
appearance of some special features of the inner physical life of the
lesser individuals which compose the Absolute. We naturally look for
some third term, in the nature of finite individuality, to mediate
between the structureless abstract generality of space and time
relation, and the perfect individual structure of the spaceless and
timeless Absolute Individual. We want, in fact, to connect the spatial
and temporal form which our experience wears, with some
fundamental aspect of our nature, as beings at once individual and
finite.
Nor is it particularly difficult to make the connection. When we
remember that space and time, as they actually condition our
perception and movement, are the space and time which radiate out
from an unique here and now of immediate feeling, it is fairly evident
that the spatial and temporal aspect of our experience is, as already
suggested, a consequence of that limitation of our attentive interests
which constitutes our finitude. It is the narrowness of my interests, or
at least of those which are sufficiently explicit to rise into the “focus”
of consciousness, that is reflected in the distinction of my here from
all the theres which are around me. Here is where my body is,
because of the specially intimate connection of the realisation of my
interests and purposes with those events in the phenomenal physical
order which I call the state of my body. Were my interests widened

Analysis and Presentation of Experimental Results With Examples Problems and Programs 1st Edition Costas Christodoulides

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Analysis and Presentation of Experimental Results With Examples Problems and Programs 1st Edition Costas Christodoulides

Uploaded by

Copyright:

Available Formats

Analysis and Presentation of

Experimental Results With Examples

Experimental Methods For Science And Engineering

Uncertainty analysis of experimental data with R 1st

Statistical Analysis of Financial Data: With Examples

Computational and experimental analysis of functional

Concise optics: concepts, examples, and problems 1st

Time Series Analysis and Its Applications With R

Eurocode Compliant Seismic Analysis and Design of R C

Survival Analysis with Interval Censored Data A

ULNP titles must provide at least one of the following:

More information about this series at http://www.springer.com/series/8917

Analysis and Presentation

ISSN 2192-4791 ISSN 2192-4805 (electronic)

Printed on acid-free paper

This Springer imprint is published by Springer Nature

Athens, Greece Costas Christodoulides

1 Basic Statistical Concepts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

4.2 The Standard Deviations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80

6.2.3 The Mean and the Standard Deviations

9.4 The Weighting of Values . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 255

12.3 The Points . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 392

1.1 Tables, Distributions and Histograms

Let us assume that we have performed N measurements xi ði ¼ 1; 2; . . .; NÞ of a

© Springer International Publishing AG 2017 1

Table 1.1 100 measurements xi of the magnitude x (in mm)

Table 1.2 The values of Table 1.1 grouped in classes

Table 1.3 The values of Class (mm) nr fr

½9:49:6Þ ½9:69:8Þ ½9:810:0Þ ½10:010:2Þ ½10:210:4Þ ½10:410:6Þ

9:49:6 9:69:8 9:810:0 10:010:2 10:210:4 10:410:6

number of measurements in the majority of the columns, so as to have a ‘smooth

Example 1.1 [E]

We will now modify the histogram.

Example 1.2 [O]

Example 1.3 [P]

# We form a vector x with components the 100 random numbers of

x = np.array([4.78887, 4.17608, 4.08729, 6.26209, 4.52593, 4.62893, 5.90426,

# The histogram is produced with the following set of commands.

plt.hist(x, bins=10, align='mid', facecolor='grey')

The histogram produced is shown here.

Example 1.4 [R]

> hist(x, main=''Histogram of x'', xlab=''x (mm)'', col=''grey'',

On pressing ENTER the histogram shown is produced.

Example 1.5 [E]

Example 1.6 [O]

Plot > Column/Bar/Pie > Column

Example 1.7 [P]

The histogram shown is produced.

Example 1.8 [R]

> xbarchart = data.frame(x=factor(c(''9.5'', ''9.6'', ''9.7'', ''9.8'', ''9.9'', ''10'',

The chart shown on the right is produced.

1.2 Probability Density

If the height DN ¼ nr of each column of a histogram is divided by the total number

is found. The same is achieved if the height of each column of a histogram of

1.3 Mean Value

The mean of N measurements xi (i ¼ 1; 2; . . .; N) is deﬁned as

If the measurements can be grouped in a number of K classes in which the value

Since nr =N ¼ fr is the relative frequency of the value xr , this expression may

This quantity is also known as expected value or mathematical expectation EðxÞ

SPACE AND TIME

You might also like