Truth and traceability in physics and metrology
Measurement Physics Metrology Uncertainty
IOP Publishing
2018
EISBN 9781643270968
1. Basics of metrology.
1.1. Regular or constant errors.
1.2. Where traceability begins.
1.3. Judging measurement results.
1.4. True values and traceability.
1.5. Consistency.
1.6. Measuring errors
2. Some statistics.
2.1. Measurands and random variables.
2.2. Fisher's density.
2.3. Confidence intervals.
2.4. Non-uniqueness of the empirical covariance.
2.5. Breakdown of statistical inference.
2.6. Curing hypothesis testing
3. Measurement uncertainties.
3.1. One measurand.
3.2. Two and more measurands.
3.3. Random errors.
3.4. Bias.
3.5. Overall uncertainty.
3.6. Error propagation at a glance
4. Method of least squares.
4.1. Geometry of adjustment.
4.2. Linear systems.
4.3. Quintessence of the method of least squares
5. Fitting of straight lines.
5.1. True straight line.
5.2. Fitting conditions.
5.3. Straight line (I).
5.4. Straight line (II).
5.5. Straight line (III)
6. Features of least squares estimators.
6.1. Uncertainties.
6.2. Weighted least squares.
6.3. Transfer of true values.
6.4. Fundamental constants of physics
7. Prospects.
7.1. Revising the error calculus.
7.2. Redefining the SI base units
8. Epilogue.
8.1. Verification by experiment.
8.2. Deciding by reasoning.
8.3. What is right, what is wrong?
Metrological data is known to be blurred by the imperfections of the measuring process. In retrospect, for about two centuries regular or constant errors were no focal point of experimental activities, only irregular or random error were. Today's notation of unknown systematic errors is in line with this. Confusingly enough, the worldwide practised approach to belatedly admit those unknown systematic errors amount to considering them as being random, too. This book discusses a new error concept dispensing with the common practice to randomize unknown systematic errors. Instead, unknown systematic errors will be treated as what they physically are--namely as constants being unknown with respect to magnitude and sign. The ideas considered in this book issue a proceeding steadily localizing the true values of the measurands and consequently traceability.
1.1. Regular or constant errors.
1.2. Where traceability begins.
1.3. Judging measurement results.
1.4. True values and traceability.
1.5. Consistency.
1.6. Measuring errors
2. Some statistics.
2.1. Measurands and random variables.
2.2. Fisher's density.
2.3. Confidence intervals.
2.4. Non-uniqueness of the empirical covariance.
2.5. Breakdown of statistical inference.
2.6. Curing hypothesis testing
3. Measurement uncertainties.
3.1. One measurand.
3.2. Two and more measurands.
3.3. Random errors.
3.4. Bias.
3.5. Overall uncertainty.
3.6. Error propagation at a glance
4. Method of least squares.
4.1. Geometry of adjustment.
4.2. Linear systems.
4.3. Quintessence of the method of least squares
5. Fitting of straight lines.
5.1. True straight line.
5.2. Fitting conditions.
5.3. Straight line (I).
5.4. Straight line (II).
5.5. Straight line (III)
6. Features of least squares estimators.
6.1. Uncertainties.
6.2. Weighted least squares.
6.3. Transfer of true values.
6.4. Fundamental constants of physics
7. Prospects.
7.1. Revising the error calculus.
7.2. Redefining the SI base units
8. Epilogue.
8.1. Verification by experiment.
8.2. Deciding by reasoning.
8.3. What is right, what is wrong?
Metrological data is known to be blurred by the imperfections of the measuring process. In retrospect, for about two centuries regular or constant errors were no focal point of experimental activities, only irregular or random error were. Today's notation of unknown systematic errors is in line with this. Confusingly enough, the worldwide practised approach to belatedly admit those unknown systematic errors amount to considering them as being random, too. This book discusses a new error concept dispensing with the common practice to randomize unknown systematic errors. Instead, unknown systematic errors will be treated as what they physically are--namely as constants being unknown with respect to magnitude and sign. The ideas considered in this book issue a proceeding steadily localizing the true values of the measurands and consequently traceability.
