Professional Documents
Culture Documents
Contents :
1. Definitions.......................................................................................................................2
2. Accuracy and Precision...................................................................................................2
3. Significant Figures...........................................................................................................3
4. Types of Error...................................................................................................................3
5. Statistical Analysis............................................................................................................4
5.1 Arithmetic Mean...............................................................................................................4
5.2 Deviation from the Mean..................................................................................................4
5.3 Average Deviation.............................................................................................................4
5.4 Standard Deviation............................................................................................................4
6. Probability of Errors.........................................................................................................5
6.1 Probable Error...................................................................................................................7
7. Limiting Errors..................................................................................................................8
8. Systems of Units of Measurement....................................................................................8
9. Standards of Measurement..............................................................................................12
1
1.1 Definitions
To illustrate the distinction between accuracy and precision, two voltmeters of the same
make and model may be compared. Both meters have knife-edge pointers and mirror-backed
scales to avoid parallax, and they have carefully calibrated scales. They may therefore be read
to the same precision. If the value of the series resistance in one meter changes considerably,
its reading may be in error by a fairly large amount. Therefore the accuracy of the two meters
may quite different.
Precision is composed of two characteristics : conformity and the number of significant
figures to which a measurement may be made. Consider, for example, that a resistor, whose
true resistance is 1,384,572 Ω , is measured by an ohmmeter which consistently and
repeatedly indicates 1.4 M Ω . But can the observer "read" the true value from the scale? His
estimates from the scale reading consistently yield a value 1.4 M Ω . This is as close to the
true value as he can read the scale by estimation. Although there are no deviations from the
observed value, the error created by the limitation of the scale reading is a precision error. The
2
example illustrates that conformity is a necesary, but not sufficient, condition for precision
because of the lack of significant figures obtained. Similarly, precision is a necesary, but not
sufficient, condition for accuracy.
Significant figures convey actual information regarding the magnitude and the
measurement precision of a quantity. The more significant figures, the greater the precision of
the measurement.
For example, if the resistor is specified as having resistance of 68 Ω , its resistence
should be closer to 68 Ω than to 67 Ω or 69 Ω . If the value of the resistor is described as
68.0 Ω , it means that its resistance is closer to 68.0 Ω than it is to 67.9 Ω or 68.1 Ω . In 68
Ω there are two significant figures; in 68.0 Ω there are three.
Often, however, the total number of digits may not represent measurement precision.
For example, the population of a city is reported closer to 380,000 than 370,000 or 390,000.
Since in this case the population can be reported only to two significant figures, how can large
numbers be expressed? A more technically correct notation uses power of ten, 38x104 or
3.8x105.
Another way of expressing the result of measurement indicates the range of possible
error. For example, the voltage may be expressed as 117.1 + 0.05 V, indicating that the value
of the voltage lies between 117.05 V and 117.15 V.
When a number of independent measurements are taken in an effort to obtain the best
possible answer (closest the true value), the result is usually expressed as the arithmetic mean
of all readings, with the range of possible error as the largest deviation from that mean. When
two or more measurements with different degrees of accuracy are added, the result is only as
accurate as the least accurate measurement.
Errors may come from different sources and are usualy classified under three main
headings :
Gross errors : largely human errors, among them misreading of instruments, incorrect
adjustments, and computational mistakes.
Systematic errors : shortcomings of the instruments, such as defective or worn parts,
and effects of the environment on equipment or the user.
3
Random errors : those due to causes that cannot be directly established because of
random variations in the parameter or the system of measurement.
Deviation is the departure of a given reading from the arithmetic mean of the group of
readings. If the deviation of the first reading, x1 , is called d 1 , and that of the second reading,
x 2 , is called d 2 , and so on, then the deviations from the mean can be expressed as
d 1 = x1 − x d 2 = x2 − x d n = x n − x .................................................(1-2)
D=
d 1 + d 2 + d 3 + ... + d n
=
∑d ......................................................................(1-3)
n n
4
of data is the square root of the sum of all the individual deviation squared, divided by the
number of readings. Expressed mathematically :
σ=
d 12 + d 22 + d 32 + ... + d n2
=
∑d i
2
................................................................(1-4)
n n
In practice, of course, the possible number of observations is finite. The standard
deviation of a finite number of data is given by
d 2 + d 22 + d 32 + ... + d n2
σ= 1 =
∑d i
2
..................................................................(1-5)
n −1 n −1
Table 1-1 shows a tabulation of 50 voltage readings that were taken at small time
intervals and recorded to the nearest 0.1 V. The nominal value of the measured voltage was
100.0 V. The result of this series of measurements can be presented graphically in the form of
a block diagram or histogram in which the number of observation is plotted. against each
observed voltage reading. The histogram of figure 1-1 represents the data of table 1-1.
Table 1-1. Tabulation of voltage readings [1]
Number of
Voltage reading readings
(volts)
99.7 1
99.8 4
99.9 12
100.0 19
100.1 10
100.2 3
100.3 1
5
Figure 1-1. Histogram showing the frequemcy of occurance of the 50 voltage readings of
table 1-1. The broken curve represents the limiting case of the histogram when a
large number of readings at small increments are taken. [1]
The possibilities as to the form of error distribution curve can be stated as follows :
(a) Small errors are more probable than large errors.
(b) Large errors are very improbable
(c) There is an equal probability of plus and minus errors so that the probability of a given
error will be symmetrical about the zero value.
The error distribution curve of figure 1-2 is based on the Normal law and shows a
symmetrical distribution of errors. This normal curve may be regarded as the limiting form of
the histogram of figure 1-1 in which the most probable value of the true voltage is the mean
value of 100.0 V.
6
Figure 1-2. Curve for the Normal law. The shaded portion indicates the region of probable
error, where r=+0.6745 σ . [1]
The are under the Gaussian probability curve of figure 1-2, between the limits + ∞ and -
∞ , represents the entire number of observations. The area under the curve between the + σ
and - σ limits represents the cases that differ from the mean by no more than the standard
deviation. Integration of are under the curve within the + σ limits gives the total number of
cases within these limits. For normally dispered data, following the Gaussian distribution,
approximately 68 % of all the cases lie between the limits of + σ and - σ from the mean.
Corresponding value of other deviations, expressed in terms of σ , are given in table 1-2.
7
If, for example, a large number of nominally 100- Ω resistors is measured and the mean
value is found to be 100.0 Ω , with a standard deviation (S.D.) of 0.20 Ω , we know that on
the average 68 % (or roughly two-thirds) of all resistors have values which lie between limits
of + 0.2 Ω of the mean.
To specify and perform calculations with physical quantities, the physical quantities
must be defined both in kind and magnitude. The standard measure of each kind of physical
quantity is the unit; the number of times the unit occurs in any given amount of the same
quantity is the number of measure.
In science and engineering, two kinds of units are used: fundamental units and derived
units. The fundamental units in mechanics are measures of length, mass, and time, and they
are often called the primary fundamental units. Measures of certain physical quantities in the
thermal, electrical, and illumination disciplines are also represented by fundamental units.
These units are used onl when these particular classes are involved, and they may therefore be
defined as auxiliary fundamental units.
All other units which can be expressed in terms of the fundamental units are called
derived units. Ecery derived unit originates from some physical law defining that unit.
A derived unit is recognized by its dimensions, which can be defined as the complete
algebraic formula for the derived unit. The dimensional symbols for the fundamental units of
length, mass, and time are L, M, and T, respectively. The dimensional symbol for the derived
unit of area is L2.
8
Table 1-3. Decimal Multiples and Submultiples [1]
9
Table 1-4. Basic SI Quantities, Units, and Symbols [1]
10
Table 1-6. Fundamental, Supplementary, and Derived Units [1]
11
Table 1-7. English Into SI Conversion [1]
12
measurement technology allow. International standards are periodically evaluated and
checked by absolute measurements in terms of the fundamental units. These standards are
maintained at the International Bureau of Weights and Measures and are not available to the
ordinary user of measuring instruments for purposes of comparison or calibration.
The primary (basic) standards are maintained by national standards laboratories in
different parts of the world. The National Bureau of Standards (NBS) in Washington is
responsible for maintenance of the primary standards in North America. Other national
laboratories include the National Physical Laboratory (NPL) in Great Britain and, the oldest
in the world, the Physikalisch-Technische Reichsanstalt in Germany. The primary standards,
again representing the fundamental units and some of the derived mechanical and electrical
units, are independently calibrated by absolute measurements at each of the national
laboratories. Primary standards are not available for use outside the national laboratories. One
of the main functions of primary standards is the verification and calibration of secondary
standards.
secondary standards are the basic reference standards used in industrial measurement
laboratories. These standards are maintained by the particular involved industry and are
checked locally againts other reference standards in the area. The responsibility for
maintenance and calibration of secondary standards rests entirely with the industrial
laboratory itself. Secondary standards are generally sent to the national standards laboratories
on a periodic basis for calibration and comparison against the primary standards. They are
then returned to the industrial user with a certification of their measured value in terms of the
primary standard.
Working standards are the principal tools of a measurement laboratory. They are used to
check and calibrate general laboratory instruments for accuracy and performance or to
perform comparison measurements in industrial applications.
13