(i)Arithmetic Mean

The most probable value of the measured variable is the arithmetic mean of the readings.
Theoretically, an infinite number of readings would give the best results, but in practice only
finite readings are available.
The arithmetic mean is given by:
∑
̅
Where x1, x2, x3… xn are the readings and n is the number of readings.

(ii)Dispersion
 Dispersion denotes the extent to which the values are dispersed or scattered about the
central value.
 Dispersion is also called as scatter or spread.
 It is very important to know the dispersion since it is an indication of the degree of
consistency or precision and regularity of the data.
 A large dispersion indicates that some factors involved in the measurement process are
not under control.
 The data with less dispersion is considered to be a reliable data.
 Some measures of dispersion are range, deviation and average deviation.
(iii)Range
The simplest measure of dispersion is the range. Range is defined as the difference
between the greatest and least values of the data.

(iv)Deviation
Deviation is the departure of the observed reading from the arithmetic mean of the group
of readings.
Let d1, d2, d3…..dn be the deviation of readings x1, x2, x3…..xn from the mean value ̅ .
Then,
̅
̅
…………
̅
The algebraic sum of deviations gives,
̅ ̅ ̅
̅
But since ̅ ̅

̅ ̅
Hence the algebraic sum of deviations is zero.
(v)Average Deviation
The average deviation is an indication of precision of the instrument used for making the
measurements. Highly precise instruments provide readings with low average deviation.
Average deviation is defined as the sum of absolute values of deviations divided by the
number of readings.

The average deviation may be expressed as:

| | | | | | ∑| |
̅

(vi)Standard deviation (Root Mean Square Deviation)

An important term in the analysis of the random errors is the standard deviation or root
mean square deviation.
The standard deviation is defined as the square root of sum of individual deviations
squared, divided by the number of readings.
If the number of readings is greater than 20, then the standard deviation is denoted by σ
and is defined as:

∑
√ √

If the number of readings is less than 20, then the standard deviation is denoted by s and
is defined as:

∑
√ √

(vii)Variance
The variance is the just the mean square deviation, without square root.If the number of
readings is greater than 20, then the variance is defined as:
∑

If the number of readings is less than 20, then the standard variance is defined as:
∑

Normal or Gaussian curve of errors

  The normal or Gaussian curve of errors is the basis for the major part of study of random
effects.
 According to the law of probability, the normal occurrence of deviations from average
value of an infinite number of measurements or observations can be expressed by

√
Where x is the magnitude of deviation from mean
y is the number of readings at any deviation x (probability of occurrence of
deviation x)
h is a constant called precision index
 The Gaussian curve can also be expressed by the equation
( ⁄ )
√
 The Normal or Gaussian Probability Curve is the plot of y against x.
 The Normal Probability Curve is as shown in fig.1.5.
y
𝑦
√𝜋

-3σ -2σ -σ -r 0 r σ 2σ 3σ
x

Fig.1.5. Normal Probability Curve

 The Normal Probability Curve is symmetrical about the arithmetic mean value.
 The area under the normal probability curve is unity.
 A higher value of precision index h indicates higher precision of data.
 Higher the value of h, sharper is the normal probability curve.
Probable Error
 Probable error (r) is a convenient measure of precision.
 Half of the observed values lie between , this is the reason why it is called as
probable error.
 If r is found using n measurements and if an additional measurement is made, then the
chances are 50 – 50 percent that the new value will lie between – r and + r.
 The probable error is expressed as:
Average deviation for the normal curve
The average deviation for the normal curve is given by:
̅
Standard deviation for the normal curve
The standard deviation for normal curve is given by:

The probable error in terms of standard deviation of normal curve is:

The probable error in terms of average deviation of normal curve is:

̅
The probable error of one reading is:

The probable error of mean is:

√
Standard deviation of mean
The standard deviation of mean is given by:

√
Standard deviation of standard deviation
The standard deviation of standard deviation is given by:

√ √

1.9 Odds
 The probability of occurrence can be stated in terms of odds.
 Odds are the number of chances that a particular reading will occur when the error limit
is specified.
 For example, if the error limits are specified as , the chances are that 50% of
the observations will lie between the above limits or in other words, we can say that the
odds are 1 to 1.
 The odds can be calculated using the relation:
  The Table below shows the probability and odds for various deviations:
Deviation Probability Odds

0.5 1 to 1
±0.6745σ
0.6828 2.15 to 1
±σ
0.9546 21 to 1
±2σ
0.9974 256 to 1
±3σ
 For other values of deviation, similar calculations can be done by referring to the
probability tables.
1.10 Uncertainty Analysis
 The statistical analysis discussed in section 1.8 is valid only for data obtained from multi
sample test.
 For data obtained from single sample test, it is not possible to observe their scatter by
plotting a frequency distribution curve.
 Kline and McClintock proposed a method based upon probability and statistics which
analyzes the data employing the uncertainty distribution rather than frequency
distribution.
 Uncertainty distribution is the error distribution the experimenter believes would exist if
the situation permits multi sampling.
 According to Kline and McClintock, the single sample result may be expressed in terms
of mean value and an uncertainty interval based upon the stated odds.
 The result of the single sample test may be expressed as:
̅
Where, ̅ is the arithmetic mean of several readings or the value of one reading if only
one value is available, is the uncertainty interval, is the odds or chances that the true
value lies within the stated range based on the opinion of the experimenter
 For example, consider the result of temperature measurement expressed as:

This means that there is an uncertainty of in the result and the experimenter is
ready to bet that the readings would fall in the range mentioned with odds of 20 to 1 or
the chances of the readings falling in the above mentioned range is 95.23%.
Propagation of Uncertainties
Suppose X is a function of several variables x1, x2, x3,…….xn

Let wx be the resultant uncertainty and wx1, wx2, wx3,…….wxn be the uncertainties in independent
variables x1, x2, x3,…….xn respectively.
Then the uncertainty in the result is given by:

√( ) ( ) ( )

Problems:
1. The expected value of the voltage across a resistor is 80 V. However, the measurement
gives a value of 79 V. Calculate (i) absolute error, (ii) relative error, (iii) % relative error
(iv) relative accuracy, and (v) % of accuracy.
Solution:
True value of voltage across resistor, At = 80V
Measured value of voltage across resistor, Am= 79V
(i) Absolute error,

(ii) Relative error,

(iii) % relative error,

(iv)Relative accuracy, | |
(v) % relative accuracy,

2. A voltage has true value of 1.5V. An analog indicating instrument with a scale range of
(0 – 2.5) V shows a voltage of 1.46V. What are the values of absolute error and correction?
Express the error as a fraction of true value and the full scale deflection.
Solution:
True value of voltage, At = 1.5V
Measured value of voltage, Am= 1.46V
The full scale deflection can be found from the range of voltmeter as, FSD = 2.5V.
(i) Absolute error,

(ii) Correction =
(iii) Error as fraction of true value (or) Relative error,

(iv) Error as fraction of full scale deflection

3. The resistance of a circuit is found by measuring the current flowing and the power fed
into the circuit. Find the limiting error in the measurement of resistance when the limiting
errors in the measurement of power and current are respectively ±2.5% and ±2.
Solution:
Relative limiting error in Power measurement,
Relative limiting error in Current measurement,
We know that,
Power,

The relative limiting error in resistance calculation can be found using the relation:

[ ]
[ ]
[ ]

Hence the limiting error in the measurement of resistance is ±6.5%.

4. Three resistors of following values:

are combined in series to obtain a resistance R. Calculate the limiting error in R.
Solution:
Relative limiting error in measurement of R1,

Relative limiting error in measurement of R2,

Relative limiting error in measurement of R3,

Since R is found by combining three resistors in series,

The relative limiting error in R is given by:

[ ]

[ ]

[ ]

Hence the limiting error in the measurement of resistance is ±5%.

5. The Power drop across a resistor is calculated from the values of voltage drop across it
and current passing through it. The relative limiting error in voltage drop is given by ±2%
and the relative limiting error in current is given by ±5%. Calculate the limiting error in
the power measured.
Solution:
Relative limiting error in Voltage measurement,
Relative limiting error in Current measurement,
We know that,
Power,

The relative limiting error in Power can be found using the relation:

[ ]
[ ]

Hence the limiting error in the measurement of power is ±7%.

6. In a Wheatstone bridge, the unknown resistance P is found using known resistors R, S

and Q using the relation: . The values of R, S and Q are given by:
. Compute the value of resistance P and
the limiting error.
Solution:
The values of resistors are R = 100, S = 1000, Q = 50
Relative limiting error in measurement of R,
Relative limiting error in measurement of S,
Relative limiting error in measurement of Q,
The value of P is given by:
The relative limiting error in P is given by:

[ ]
[ ]

Hence the limiting error in the measurement of the resistance P is ±1.75%.

The value of P can be given as:
7. A set of independent ten measurements were made to determine the weight of a lead
shot. The weight in grams is 1.57, 1.597, 1.591, 1.562, 1.577, 1.58, 1.564, 1.586, 1.55, and
1.575. Determine the arithmetic mean, deviations, average deviation, standard deviation,
variance, probable error of one reading, probable error of mean and range.
Solution:
a) The arithmetic mean is given by
̅

̅
b) The deviations are:
̅
̅
̅
̅
̅
̅
̅
̅
̅
̅
c) The average deviation is:
̅ | | | | | | | | | | | | | | | | | | | |

| | | | | | | | | | | | | | | | | | | |

d) The standard deviation is:

Since the number of readings is less than 20,

√
 √

e) The Variance is given by:

e) Probable error of one reading is given by:

f) The probable error of mean is:

√ √

g) The range is:

8. A temperature is measured 100 times with variation in apparatus and procedures. After
applying the corrections, the results are:
Temperature ( °C ) 397 398 399 400 401 402 403 404 405
Frequency of Occurrence 1 3 12 23 37 16 4 2 2
Calculate the arithmetic mean, mean deviation, standard deviation, the probable error of
one reading, the standard deviation of mean, the probable error of mean and the standard
deviation of standard deviation.
Solution:
The computations can be done in a table as under:
Frequency
Temperature of Deviation
T*f f*d d2 f * d2
(T) occurrence (d)
(f)
397 1 397 – 3.78 – 3.78 14.288
398 3 1194 – 2.78 – 8.34 14.288 23.185
399 12 4788 – 1.78 – 21.36 7.728 38.020
400 23 9200 – 0.78 – 17.94 3.168 13.993
401 37 14837 0.22 8.14 0.608 1.708
402 16 6432 1.22 19.52 0.048 23.814
 403 4 1612 2.22 8.88 1.488 19.714
404 2 808 3.22 6.44 4.928 20.737
405 2 810 4.22 8.44 10.368 35.618
17.808

Total 100 40078 102.8 191.08

a) Arithmetic mean is given by:
̅ ∑
∑

b) Mean deviation is given by:

̅ ∑
∑

̅
c) Since the number of readings is greater than 20, the Standard Deviation is given by:
∑
√ ∑

√ √

d) The probable error of one reading is

e) The probable error of mean is

√∑

f) The standard deviation of the mean is

√∑

g) The standard deviation of standard deviation is

√ √
9. The table given below lists a sample of experimental data:
Value 3 4 5 6 7 8 9 10 11
Frequency of
1 2 3 6 7 6 4 2 1
Occurrence
Calculate the arithmetic mean, mean deviation, standard deviation, the probable error of
one reading, the standard deviation of mean, the probable error of mean and the standard
deviation of standard deviation.

Solution:
The computations can be done in a table as under:
Frequency
Value of Deviation
V*f f*d d2 f * d2
(V) occurrence (d)
(f)
3 1 3 – 4.0625 4.0625 16.5039 16.5039
4 2 8 – 3.0625 6.125 9.3789 18.7578
5 3 15 – 2.0625 6.1875 4.2539 12.7617
6 6 36 – 1.0625 6.375 1.1289 6.7734
7 7 49 – 0.0625 0.4375 0.0039 0.0273
8 6 48 0.9375 5.625 0.8789 5.2734
9 4 36 1.9375 7.75 3.7539 15.0156
10 2 20 2.9375 5.875 8.6289 17.2578
11 1 11 3.9375 3.9375 15.5039 15.5039
Total 32 226 46.375 107.8748
a) Arithmetic mean is given by:
̅ ∑
∑

̅
b) Mean deviation is given by:
̅ ∑
∑

̅
c) Since the number of readings is greater than 20, the Standard Deviation is given by:
∑
√ ∑

√ √

d) The probable error of one reading is

e) The probable error of mean is

√∑

f) The standard deviation of the mean is

√∑

g) The standard deviation of standard deviation is

√ √

10. Two resistors R1 and R2 are connected in series and then in parallel. The values of
resistors are: . Calculate the uncertainty in the
combined resistance for both series and parallel combinations.
Solution:
Uncertainty in measurement of R1,
Uncertainty in measurement of R2,
Series Combination
When resistors are combined in series, their resultant value is:

The uncertainty in R is given by:

 √( ) ( )

√
√
Hence the uncertainty in series combination of resistances is ±0.56%
Parallel Combination
When resistors are combined in parallel, their resultant value is:

( )

( )

The uncertainty in R is given by:

√( ) ( )

√
√
Hence the uncertainty in parallel combination of resistances is ±0.12%
11. A certain resistor has a voltage drop of 100.3 V and a current of 4.2 A. The
uncertainties in measurement of voltage and current are ±0.1 V and ±0.02 A respectively.
Calculate the power dissipated in the resistor and the uncertainty in the power
measurement.
Solution:
The value of voltage drop across resistor, V = 100.3V
The value of current through resistor, I = 4.2 A
Uncertainty in measurement of voltage,
Uncertainty in measurement of current,
The value of power dissipated is:
The uncertainty in measurement of power is:

√( ) ( )

√
√

Hence the uncertainty in measurement of power is ±2.05 W.

The value of power can be expressed as:

12. The power drop in a resistor 100 Ω ± 0.2% is calculated using the voltage drop across it
which is given by 100V± 1.5%. Calculate the uncertainty in measurement of power.
Solution:
The value of resistor, R = 100 Ω
The value of voltage drop across resistor, V = 100 V
Uncertainty in measurement of resistance,
Uncertainty in measurement of voltage,
The power drop in a resistor is given by:

The uncertainty in measurement of power is:

√( ) ( )

( )
 ( )

√( ) ( )

Hence the uncertainty in measurement of power is ±3.01%