 The function of the analyst is to

obtain a result as near to the

true value as possible by the
correct application and use of
the analytical method and
procedure.
 The level of confidence in the
results will be very small unless
there is a knowledge of the
accuracy and precision of the
method used as well as being
aware of the sources of error in
the measurement.
 Accuracy and Precision

 Statistics

 Errors

 Calibration Curves
 Statistics

 The true or absolute value of a quantity

cannot be established experimentally, so that
the observed value must be compared with
the most probable value.

 Statistics provide a means of quantifying the

precision of a set of measurements.
 Mean
 It is found that the results of a series of
determinations will vary slightly.
 The average value (mean) is accepted as
the most probable.

x
x=
n
 Estimates of Precision
 Standard Deviation
 Variance
 Relative Standard Deviation
 Coefficient of Variation
 Standard Deviation
 Defined as the square root of the sum of
the squares of the deviation from the
mean divided by the total number of data
minus one.

 ( x - x)2
s=
n-1
 Standard Deviation for a large set

 ( x - x)2
s=
n
 Variance
 Is the square of the standard deviation.

 ( x - x) 2
s2 =
n-1
 Relative Standard Deviation

 A further measure of precision is known as the

Relative Standard Deviation (R.S.D.).

R.S.D. = s / x
 Coefficient of Variation

 This measure is often expressed as a

percentage as the coefficient of variation
(C.V.)

R.S.D. = 100s / x
 Classification of Errors

 Systematic

 Random
 Random Error  Systematic Error
 Affects precision  Affects accuracy
 Within run precision -  Proximity to the
repeatability truth
 Between run precision -  Determinate error
reproducibilty or bias.
 Indeterminate error
 Systematic Errors
 These are errors which can be avoided, or
whose magnitude can be determined.

There are three important types of

systematic error.
 Operational and personal errors.
Factors for which the analyst is responsible.
 Instrumental and reagent errors.
Uncalibrated equipment, unexpected
reactions.
 Errors of the method used.
Incorrect sampling, incompleteness of a
reaction.
 Random Errors
 The slight variations which occur in successive
measurements of exactly the same method.

 Due to causes over which the analyst has no

control.

 If a sufficiently large number of measurements are

taken it can be shown that these errors lie on a
Gaussian curve.
 Propagation of errors
 It is important to note that the procedure for
combining random and systematic errors are
completely different.

 Random errors to some extent cancel one

another out whereas every systematic error
occurs in a definite and known sense.
 If a and b have a systematic error of +1 then
the systematic error in x given by
 x = a + b is +2.

 If however, a and b have a random error of

 , the random error in x is not  
 Propagation of errors
 Two types

 Linear combinations
 sums and differences

 Multiplicative expressions
 Linear Combinations

 In the case of the final value y calculated

from the linear combination of measured
quantities a, b, c, etc

sy =  sa2 + sb2 + sc2 + ...

 Multiplicative Expressions

 In the case of the final value y calculated

from an expression of the type y = ab / cd

2 2 2 2
sy  s a   sb   s c   s d 
=   +  +  + 
y a b c d
  Multiplicative Expressions 2

 In the case of the final value y calculated

from an expression of the type y = ab / cd

2 2 2
    
 100 =   100  +   100  +   100 
sy sa sb sc
y a  b  c 
 Calibration Curves

 When carrying out an analysis it is often

necessary to carry out a calibration
procedure by using a series of standards
(samples) each having a known
concentration.

 A calibration curve is constructed by plotting

the response of the standards against the
concentration.
 Calibration Curves

 There are two statistical tests which should be

applied to a calibration curve.

 To ascertain the linearity of the curve.

 To evaluate the best straight line through the

data points.
 Correlation Coefficient
 In order to establish whether there is a linear
relationship between two variables x1 and y1
the Pearson’s correlation coefficient is used.

nx1y1 - x1y1
r=
[nx12 - (x1)2] [ny12 - (y1)2]
 Linear Regression

 Once a linear relationship has been shown to

have a high probability by the value of the
correlation coefficient, then the best straight
line through the data points has to be
estimated.

 This can often be done by visual inspection

but it is better to evaluate it by linear
regression – using the method of least squares.
 Linear Regression

 The equation of a straight line is y = ax + b

where y the dependent variable is plotted as
a result of changing x, the independent
variable.

 To obtain the regression line y on x the slope

of the line “a” and the intercept on the y-
axis “b” are given by the following
equations.
 Linear Regression

nx1y1 - x1y1
a=
nx12 - (x1)2

b = y - ax
 Calibration Curves

R2 = 0.9992
y = 1.2469x -
0.0219
 Significant Figures

 There are a number of rules for

computations which you need to be aware
of and be familiar with.
 Significant Figures

 Retain as many significant figures in a result

or in any data as needed but the last digit
only will give one uncertain figure.

 A volume which is known to be between 30.5

ml and 30.7 mL should be written as 30.6 mL,
but not 30.60 mL as the latter would imply
that the value lies between 30.59 ml and
30.61 mL
 Significant Figures

 In rounding off quantities to the correct

number of significant figures add one to the
last figure retained if the following figure
(which has been rejected ) is 5 or over.
 In addition or subtraction, there should be
in each number only as many significant
figures as there are in the least accurately
known number.
 e.g. 168.11 + 7.045 + 0.6832
 should be written as:
 168.11 + 7.05 + 0.68
 Significant Figures
 In multiplication or division, retain in each
factor one more significant figure than is
contained in the factor having the largest
uncertainty.
 i.e. 1.26 x 1.236 x 0.6834
should be written as:
1.26 x 1.236 x 0.683
 Data can be presented visually as:
 Data tables
 Graphs

 Be easy to understand.
 Project a message.
 Does not require extensive reference to the
text.
A solution of compound X is reacted with
reagent Y which causes the production of a
light absorbing product. The amount of light
absorbing product indicates how much
compound X is present. The sample
containing X is divided into three aliquots
and the experiment is repeated three times
at three wavelengths (360, 420 and 540 nm).
 The results are presented in a table.
Results of the three experiments
Absorbance Absorbance Absorbance
values for reaction values for reaction values for reaction
#1 at 360, 420 and #2 at 360, 420 and #3 at 360, 420 and
540 nm 540 nm 540 nm

0.875 (360) 0.885 (360) 0.823 (360)

0.253 (420) 0.250 (420) 0.244 (420)

0.164 (540) 0.165 (540) 0.157 (540)

Absorbance values for the “experiments”
at 360,420 and 540 nm
Trial # A360 A420 A540

1 0.876 0.253 0.164

2 0.885 0.250 0.163

3 0.825 0.244 0.157

 0.9
0.8
0.7
0.6
0.5
Absorbance Trial 1
0.4
0.3 Trial 2
0.2 Trial 3
0.1
0
A360 A420 A540
Wavelength
 0.9
0.8
0.7
0.6
0.5 A360
Absorbance
0.4 A420
0.3 A540
0.2
0.1
0
Trial 1 Trial 2 Trial 3
a) Calibration of apparatus
b) Running a blank determination
c) Running a control determination
d) Use of independent methods of analysis
e) Running parallel determinations
f) Standard addition
g) Internal standards
h) Amplification methods
i) Isotopic dilution
j) Double checking items
Analyst has no control on random errors, but systematic
errors can be reduced by the following methods.
 Quality equipment and instruments use. Make sure
you’re using up to date and precise measuring devices
that do not have any defects while doing the
experiments.
 Calibration of apparatus: By calibrating all the
instruments and equipment used, errors can be
identified, and appropriate corrections applied to the
original measurements to minimize the errors.
 Control determination: A standard (reference)
material is used in the experiment in identical
experimental conditions to identify, quantify and
minimize the errors.
Blank determination: By omitting the analyte in the
sample, a determination is carried out in identical condition
to identify and minimize the errors occurs due to impurities
present in reagent.
 Independent method of analysis: It is carried out to
maintain accuracy of the result e. g. Iron (III) is first
determined gravimetrically by precipitation method as iron
(III) hydroxide and then determined titrimetrically by
reduction to the iron (II) state.
 Parallel determination: Instead of a single determination,
duplicate or triplicate determinations are carried out to
identify and minimize the possibilities of accidental errors.
 Standard addition: This method is generally applied to
physico-chemical procedures such as polarography,
potentiometry and spectrophotometry. A known amount
of analyte is added to the sample and analyzed.
 Internal standards: It is used in spectroscopic and
chromatographic determination. A similar compound as
the analyte is added to all samples and blanks and
analyzed.
 Amplification methods: It is used when a very small
amount of material is to be measured which is beyond
the limit of the apparatus.
 Isotopic dilution: It is used for the compounds
containing radio-active isotopes.
 Human error: To reduce the impact of human
error, personnel (analysts) need to double-check all
observations, recordings, and measurements.
You can easily complete this process by double-
entering all findings on two separate worksheets or
files and then later comparing them to check for
correct recordings.