Chapter 4: Random Errors

in Chemical Analysis

杜采潓 助理教授
醫藥暨應用化學
E-mail: tshudu@[Link]

Skoog, West, Holler, and Crouch, Fundamentals of Analytical Chemistry, 10e. © 2022 Cengage. All Rights Reserved. May not be
scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part.

[Author Name], [Book Title], [#] Edition. © [Insert Year] Cengage. All Rights Reserved. May not be scanned, copied or duplicated, or posted to a publicly
accessible website, in whole or in part.
歐亞書局 1
學習目標
※Random Error Sources

※ Samples and Populations

※ Properties of Gaussian Curves

※ Standard Deviation

※ Relative Standard Deviation (RSD)

※ Coefficient of Variation (CV)

2
資料來源

Skoog. West. Holler. Crouch.:

FUNDAMENTALS OF ANALYTICAL
CHEMISTRY, 10e.

3
4A The Nature of Random Errors

Random, or indeterminate, errors can never be

totally eliminated and are often the major source
of uncertainty in a determination. Random errors
are caused by the many uncontrollable variables
that accompany every measurement.
The accumulated effect of the individual
uncertainties, however, causes replicate results to
fluctuate randomly around the mean of the set.
 4A-1 Random Error Sources

Skoog. : FUDAMENTALS OF ANALYTICAL CHEMISTRY, 10e, p. 52

5
Gaussian curve

When the same procedure is applied to a very large

number of individual errors, a bell-shaped curve like
that shown in Figure 4-2c results. Such a plot is called
a Gaussian curve, or normal error curve.

6
p. 95
Gaussian curve

Figure 4-2
Frequency distribution for
measurements containing
(a) 4 random uncertainties;
(b) 10 random uncertainties;
(c) a very large number of
random uncertainties.

Skoog. : FUDAMENTALS OF
ANALYTICAL CHEMISTRY, 10E, p. 53
7
4A-2 Distribution of Experimental Results
From experience with many determinations, we find
that the distribution of replicate data from most
quantitative analytical experiments approaches that of
the Gaussian curve shown in Figure 4-2c.
In Table 4-2, the mean is calculated with Excel’s
function AVERAGE( )
The median is calculated using Excel’s function
MEDIAN( ).

8
4A-2 Distribution of Experimental Results

The maximum value is found with MAX() and the

minimum value with MIN( ).
Spread of data results directly from an accumulation
of all random uncertainties in the experiment.
The spread in a set of replicate measurements is the
difference between the highest and lowest result.

9
p. 53
4A-2 Distribution of Experimental Results
Table 4-2

10
Skoog. : FUDAMENTALS OF ANALYTICAL CHEMISTRY, 10e, p. 54
4A-2 Distribution of Experimental Results

Table 4-3

Skoog. : FUDAMENTALS OF ANALYTICAL CHEMISTRY, 10e, p. 54

11
4A-2 Distribution of Experimental Results

Skoog.: FUDAMENTALS OF ANALYTICAL CHEMISTRY, 10e, p. 55

12
4A-2 Distribution of Experimental Results

Sources of random uncertainties in the calibration of

a pipet include(1)Visual judgments, such as the level
of the water with respect to the marking on the pipet
(2) variations in the drainage time and in the angle of
the pipet as it drains; (3) temperature fluctuations,
and (4) vibrations and drafts that cause small
variations in the balance readings.

13
FEATURE 4-1
If you flip a coin 10 times, how many heads will you
get? Try it, and record your results. Repeat the
experiment. Are your results the same? Ask friends or
members of your class to perform the same
experiment and tabulate the results. The table below
contains the results obtained by several classes of
analytical chemistry students over an 18-year period.

Number of 0 1 2 3 4 5 6 7 8 9 10
Heads
Frequency 1 1 22 42 102 104 92 48 22 7 1

14
p.55
FEATURE 4-1
Add your results to those in the table, and plot a
histogram similar to the one shown in Figure 4F-1.
Find the mean and the standard deviation (see Section
4B-3) for your results and compare them to the values
shown in the plot. The smooth curve in the figure is a
normal error curve for an infinite number of trials with
the same mean and standard deviation as the data set.
Note that the mean of 5.06 is very close to the value of
5 that you would predict based on the laws of
probability. As the number of trials increases, the
histogram approaches the shape of the smooth curve
and the mean approaches 5.

15
FEATURE 4-1

120

100
μ = 5.04
80
σ = 1.62
frequency

60

40

20

0
0 1 2 3 4 5 6 7 8 9 10
Number of heads
Figure 4F-1 Results of a coin-flipping experiment by 395 students
over an 18-year period.
16
Skoog.: FUDAMENTALS OF ANALYTICAL CHEMISTRY, 10E, p. 55
4B STATISTICAL TREATMENT OF RANDOM
ERRORS
Analytical data can follow distributions other than the
Gaussian distribution.
For example, experiments in which there is either a
successful outcome or a failure produce data that
follow the binomial distribution.
However, we often use a Gaussian distribution to
approximate these distributions.

17
4B-1 Samples and Populations
Typically in a scientific study, we infer information
about a population or universe from observations
made on a subset or sample.
A population is the collection of all measurements of
interest to the experimenter, while a sample is a
subset of measurements selected from the population.
Statistical laws have been derived for populations;
often they must be modified substantially when
applied to a small sample because a few data points
may not represent the entire population.

18
4B-2 Properties of Gaussian Curves
Figure 4-4a shows two Gaussian curves in which we
plot the relative frequency y of various deviations
from the mean versus the deviation from the mean.
As shown in the margin, curves such as these can be
described by an equation that contains just two
parameters, the population mean m and the
population standard deviations. The term
parameter refers to quantities such as m and s that
define a population or distribution.

19
4B-2 Properties of Gaussian Curves

The equation for a Gaussian curve has the form

Statisticians find it useful to differentiate between the

sample mean and the population mean. The sample
mean is the arithmetic average of a limited sample
drawn from a population of data.

20
4B-2 Properties of Gaussian Curves

Skoog.: FUDAMENTALS OF ANALYTICAL CHEMISTRY, 10E

21
4B-2 Properties of Gaussian Curves

The sample mean is found from

where N is the number of measurements in the sample

set. The same equation is used to calculate the
population mean μ

22
Properties of Gaussian Curves

In the absence of systematic error, the population

mean is also the true value for the measured quantity.
The probable difference between and μ decreases
rapidly as the number of measurements making up the
sample increases; ordinarily, by the time N reaches 20
to 30, this difference is negligible.
The population standard deviation σ, which is a
measure of the precision of a population of data, is
given by the equation

23
Properties of Gaussian Curves

where N is the number of data points making up the

population.
Figure 4-4b shows another type of normal error curve
in which the x-axis is now a new variable z , defined
as

24
Properties of Gaussian Curves

Because it appears in the Gaussian error curve

expression, the square of the standard deviation σ2 is
also important. This quantity is called the variance.
A normal error curve has several general properties: (a)
The mean occurs at the central point of maximum
frequency, (b) there is a symmetrical distribution of
positive and negative deviations about the maximum,
and (c) there is an exponential decrease in frequency
as the magnitude of the deviations increases. Thus,
small uncertainties are observed much more often than
very large ones.

25
Properties of Gaussian Curves

68.3% of the area beneath a Gaussian curve for a

population lies within one standard deviation (±1σ) of
the mean μ. Roughly 68.3% of the values making up
the population will lie within these bounds.
Furthermore, approximately 95.4% of all data values
are within ±2σ of the mean and 99.7% are within ±3σ.

26
p.101
Properties of Gaussian Curves

FEATURE 6-2
Calculating the Areas Under the Gaussian Curve

It is easier to use the form of Equation 6-3 with the

variable z, so our equation becomes

27
Properties of Gaussian Curves

Since there is no closed form solution, the integral

must be evaluated numerically. The result is

Likewise, if we want to know the area under the

Gaussian curve 2σ on either side of the mean, we
evaluate the following integral.

28
Properties of Gaussian Curves

Skoog.: FUDAMENTALS OF ANALYTICAL CHEMISTRY, 10E

29
Properties of Gaussian Curves

Skoog.: FUDAMENTALS OF ANALYTICAL CHEMISTRY, 10E

30
Properties of Gaussian Curves

Skoog.: FUDAMENTALS OF ANALYTICAL CHEMISTRY, 10E

31
4B-3 The Sample Standard Deviation: A Measure of
Precision
the sample standard deviation s is given by the
equation

The quantity N – 1 is called the number of degrees

of freedom.

32
The Sample Standard Deviation

where the quantity represents the deviation di

of value xi from the mean .
Equation 6-1 is replaced by the number of degrees of
freedom (N – 1 ). When N – 1 is used instead of N, s
is said to be an unbiased estimator of the population
standard deviation σ.
The sample variance σ2 is also important in statistical
calculations. It is an estimate of the population
variance σ2.

33
An Alternative Expression For Sample Standard
Deviation
※

34
The Sample Standard Deviation

EXAMPLE 4-1
The following results were obtained in the replicate
determination of the lead content of a blood sample:
0.752, 0.756, 0.752, 0.751, and 0.760 ppm Pb.
Find the mean and the standard deviation of this set
of data.
To apply Equation 6-5, we calculate

35
The Sample Standard Deviation

36
The Sample Standard Deviation

Substituting into Equation 4-5 leads to

37
Standard Error of the Mean
The standard deviation of each mean is known as the
standard error of the mean and is given the symbol sm.
The standard error of the mean, sm, is the standard
deviation of a set of data divided by the square root of the
number of data points in the set.

38
4B-4 Reliability of s as a Measure of Precision
The probability that these statistical tests provide
correct results increases as the reliability of s becomes
greater.
When N is greater than about 20, s is usually a good
estimator of s, and these quantities can be assumed to
be identical for most purposes.
The rapid improvement in the reliability of s with
increases in N makes it feasible to obtain a good
approximation of s when the method of measurement
is not excessively time consuming and when an
adequate supply of sample is available.
39
Pooling Data to Improve the Reliability of s
If we have several subsets of data, we can get a better
estimate of the population standard deviation by
pooling (combining) the data than by using only one
data set.
The pooled estimate of s, which we call spooled, is a
weighted average of the individual estimates.

40
Pooling Data to Improve the Reliability of s

FEATURE 4-4
Equation for Calculating the Pooled Standard
Deviation
The equation for computing a pooled standard
deviation from several sets of data takes the form

41
Pooling Data to Improve the Reliability of s

where N1 is the number of results in set 1, N2 is the

number in set 2, and so forth. The term Nt is the total
number of data sets pooled.

42
Pooling Data to Improve the Reliability of s

EXAMPLE 4-2
Glucose levels are routinely monitored in patients
suffering from diabetes. The glucose concentrations in
a patient with mildly elevated glucose levels were
determined in different months by a
spectrophotometric analytical method. The patient was
placed on a low-sugar diet to reduce the glucose levels.
The following results were obtained during a study to
determine the effectiveness of the diet. Calculate a
pooled estimate of the standard deviation for the
method.
43
Pooling Data to Improve the Reliability of s

44
Pooling Data to Improve the Reliability of s

For the first month, the sum of the squares in the next
to last column was calculated as follows:

The other sums of squares were obtained similarly.

The pooled standard deviation is then

45
4B-5 Variance and Other Measures of Precision
Chemists ordinarily use the sample standard deviation
in reporting the precision of their data.
Variance (s 2)
The variance is just the square of the standard
deviation. The sample variance s2 is an estimate of
the population variance σ2 and is given by

46
Relative Standard Deviation (RSD) and Coefficient of
Variation (CV)
The relative standard deviation, RSD, is sometimes
given the symbol sr.

The result is often expressed in parts per thousand (ppt)

or in percent

The relative standard deviation multiplied by 100% is

called the coefficient of variation (CV).

47
Spread or Range (w)
The spread, or range (w), is another term that is
sometimes used to describe the precision of a set of
replicate results.

48
Spread

EXAMPLE 4-3
For the set of data in Example 4-1, calculate (a) the
variance, (b) the relative standard deviation in parts
per thousand, (c) the coefficient of variation, and
(d)the spread.
In Example 4-1, we found
x = 0.754 ppm Pb and s = 0.0038 ppm Pb

49
Spread

50
4C STANDARD DEVIATION OF CALCULATED
RESULTS
Ofen we must estimate the standard deviation of a
result that has been calculated from two or more
experimental data points,
4C-1 Standard Deviation of a Sum or Difference

The variance of y, is given by

Hence, the standard deviation of the result sy is

51
4C STANDARD DEVIATION OF CALCULATED RESULTS

Skoog.: FUDAMENTALS OF ANALYTICAL CHEMISTRY, 10e 52

4C-2 Standard Deviation of a Product or Quotient
For example, in the case of

we obtain the relative standard deviation sy/y of the

result by summing the squares of the relative standard
deviations of a, b, and c and then extracting the square
root of the sum:

53
Calculate the standard deviation of the result

EXAMPLE 4-4
Calculate the standard deviation of the result of

First, we must calculate the standard deviation of the sum and

the difference. For the difference in the numerator,

and for the sum in the denominator,

54
Calculate the standard deviation of the result

We may then rewrite the equation as

The equation now contains only products and quotients,

and Equation 4-12 applies. Thus,

0.107

To obtain the absolute standard deviation, we write

Sy= y ． 0.107 = 1.725 x 10-6 ． (± 0.107) = ±0.185x10-6
and round the answer to 1.7(±0.2) x 10–6 .

55
4C-3 Standard Deviations in Exponential
Calculations
Consider the relationship

56
Calculate the standard deviation of the result

EXAMPLE 4-5
The solubility product Ksp for the silver salt AgX is
4.0 (±0.4) × 10-8. and the molar solubility is
Solubility ＝ (Ksp)1/2 ＝(4.0 × 10-8 ) 1/2＝ 2.0 × 10-4M
What is the uncertainty in the calculated solubility of AgX ?
Substituting y ＝ solubility, a = Ksp, and
x = 1/2 into Equation 4-13 gives

57
Calculate the standard deviation of the result

58
4C-4 Standard Deviations of Logarithms and
Antilogarithms
The last two entries in Table 6-4 show that for
y = log a

and for y antilog a

59
Calculate the standard deviation of the result

EXAMPLE 4-6
Calculate the absolute standard deviations of the results of the
following calculations. The absolute standard deviation for each
quantity is given in parentheses.

Referring to Equation 6–14, we see that we must multiply the

relative standard deviation by 0.434:

Thus
60
Calculate the standard deviation of the result

61
Calculate the standard deviation of the result

62
4D REPORTING COMPUTED DATA

A numerical result is worthless to users of the data

unless they know something about its quality.
Therefore, it is always essential to indicate your best
estimate of the reliability of your data. One of the best
ways of indicating reliability is to give a confidence
interval at the 90% or 95% confidence level.
Another method is to report the absolute standard
deviation or the coefficient of variation of the data.

63
EXAMPLE 4-8
A 3.4842-g sample of a solid mixture containing
benzoic acid, C6H5COOH (122.123 g/mol), was
dissolved and titrated with base to a phenolphthalein
end point. The acid consumed 41.36 mL of 0.2328 M
NaOH. Calculate the percent benzoic acid (HBz) in
the sample.

64
p.118
%HBz =

Since all operations are either multiplication or

division, the relative uncertainty of the answer is
determined by the relative uncertainties of the
experimental data. Let us estimate what these
uncertainties are.
65
Calculate the standard deviation of the result
1. The position of the liquid level in a buret can be
estimated to ±0.02 mL(Figure 4-5). In reading the
buret, two readings(Initial and final) must be made,
so that the standard deviation of the volume sV will
be

The relative uncertainty in volume sV/V is then

2. Generally, the absolute uncertainty of a mass

obtained with an analytical balance will be on the
order of ± 0.0001 g. Thus the relative uncertainty of
the denominator sD/D is 66
Calculate the standard deviation of the result

3. Usually we can assume that the absolute uncertainty in

the concentration of a reagent solution is ± 0.0001,
and so the relative uncertainty in the concentration of
NaOH, sc/c is

Sc

67
Calculate the standard deviation of the result
4. The relative uncertainty in the molar mass of HBz is
several orders of magnitude smaller than any of the
three experimental values and will not be significant.
Note, however, that we should retain enough digits in
the calculation so that the molar mass is given to at
least one more digit (the guard digit) than any of the
experimental data. Thus, in the calculation, we use
122.123 for the molar mass (we are carrying two extra
digits in the instance).
5. No uncertainty is associated with 100% and the 1000
mmol HBz, since these are exact numbers.

68
Calculate the standard deviation of the result
Substituting the three relative uncertainties into
Equation 4-12, we obtain

Therefore the uncertainty in the calculated result is

0.03% HBz, and we should report the result as 33.75%
HBz or, better, 33.75 ( ±0.03)% HBz.

69
SUMMARY
Important Equations

Population standard deviation, σ Sample standard deviation, s

N N N

 ( xi − μ )
2
(x − μ) d
2 2
i i
σ= i =1
s= i =1
= i =1

N N −1 N −1

Standard error of the mean, sm Pooled standard deviation, spooled

s N1 N2 N3

(x − x1 ) +  ( x j − x2 ) +  ( xk − x3 ) +
2 2 2
sm = i
N i =1 j =1 k =1
spooled =
N1 + N2 + N3 + − Nt
Relative standard deviation, RSD
s Coefficient of variation, CV
RSD = st =
x

歐亞書局