25/02/1446

#$: !" ‫ا‬ ‫ر ا‬ ‫ا‬

' ( ‫ ا‬: ‫ا !& ى‬ ‫ع رب‬
‫ول‬+‫ ا‬: ‫ا& م‬ ‫ما‬ ‫ا‬
2024/2025 : ‫ا م‬ ‫ر‬ : ‫ا‬

Analytical chemistry
Chapter 2
Data Handling and Spreadsheets in Analytical
Chemistry
Instructor:
Dr. Abdulbaset Alzazaee

Outline
2.1 Accuracy and precision

2.2 Determinate Errors

2.3 Indeterminate Errors

2.4 Expressing Accuracy

2.5 Standard deviation

2.6 Confidence limit.

2.7 Rejection of Results

2

1
25/02/1446

Objectives of the chapter

1. Understand the meaning of accuracy and precision.
2. Understand the difference between systematic and
random errors.
3. Express the results in proper significant figures after
algebraic operations.
4. Describe absolute errors and their relationship to
accuracy.
5. Use spreadsheets for analytical calculations.
6. Understand the propagation of errors and be able to
calculate a confidence interval.
7. How to retain or reject analytical data. 3

Introduction
What are analytical results?

Analytical results usually have:

Numerical data and a Unit:

1. What you need to know about this numerical data?

Quantity of error .
How far is it from the true value?

2. What you need to know about the unit?:

Relationship to other units .
4

2
25/02/1446

Introduction
What are analytical results?
An example of an Analytical result:
The blood glucose level is:
10.0 mmole/L

Numerical data Unit

1. Numerical data
Quantity of error in 10.0 (e.g: ±0.1 , ± 0.5 or ± 1)
Is 10.0 close to the true value
2. The unit
Relationship to mmole/L to mg/dL . 5

5 5

Introduction
What are analytical results?

• In this chapter we will investigate

how to deal with the Numerical data
and in later chapters will deal with
the Units.

3
25/02/1446

Introduction
What are analytical results?
10.0 mmole/L

What we need to know about this

Numerical data :
a. How much the doubt in it (How
significant).
b. How much the error /range of
confidence.
c. How far is it form the true value.
7

Introduction
The need for Statistics
To reduce the error and be able to
judge numerical data:
data should be repeated more than
2 times .
Accuracy and Precision of the data
then can be analyzed .
We need to use Statistics to treat the
collected data.
8

4
25/02/1446

Introduction
Some Useful Statistics Terms
▫ Mean – Average of a set of values
▫ Median – Mid-point of a set of values.
▫ Population – A collection of an infinite
number of measurements. N infinity
▫ Sample – A finite set of measurements which
represent the population.
▫ True value (true mean)- (m), mean value for
the population.
▫ Observed Mean mean value of the sample
set

10

2.1 Accuracy and Precision:

There is a Difference

What is Accuracy ?
The degree of agreement between measured
value and the true value.

What is Precision?
Precision: degree of agreement between
replicate measurements of same quantity.

5
25/02/1446

11

2.1 Accuracy and Precision:

There is a Difference
To understand the difference between
Accuracy and Precision let us propose an
analysis of a real value µ showed this three
replicate measurements : X1, X2 and X3 ?
Accuracy Precision
µ X1

X1 X2 X3 X2 X3

12

2.1 Accuracy and Precision:

There is a Difference
Accuracy Precision

Accurate precise

Inaccurate Not precise

6
25/02/1446

13

2.1 Accuracy and Precision:

There is a Difference
To understand the difference between Accuracy and
Precision lets see this example:
In previous Lab. Final Exam the real value was 5.0 mL
and I selected these results (all in mL):
Student (1): 5.0 5.6 4.5
Which student has:
Student (2): 8.2 5.8 5.5
1) the best Accuracy
Student (3): 5.0 5.0 5.1 2) the best Precision
Student (4): 9.3 9.2 9.3 3) the worst Accuracy
Student (5): 4.7 5.0 5.3 4) the worst Precision
Student (6): 6.7 6.5. 6.6

14

2.1 Accuracy and Precision:

There is a Difference
To understand the difference between Accuracy and
Precision lets see this example:
In previous Lab. Final Exam the real value was 5.0 mL
and I selected these results (all in mL):
Student (1): 5.0 5.6 4.5 high precision high accuracy

Student (2): 8.2 5.8 5.5 low precision low accuracy

Student (3): 5.0 5.0 5.1 high precision high accuracy

Student (4): 9.3 9.2 9.3 high precision low accuracy

Student (5): 4.7 5.0 5.3 high precision high accuracy

Student (6): 6.7 6.5. 6.6 high precision low accuracy???

7
25/02/1446

15

2.1 Accuracy and Precision:

There is a Difference

How to control and manage the precision of

the results?
To answer this question we should study the
classes of Errors:
1) Determinate errors (systematic errors).
2) Indeterminate errors (Random errors).

16

2.2 Determinate Errors - They Are Systematic

• Determinate Errors or Systematic Errors :
occurs when something wrong with the
measurement.

• Properties:
1) Could take one direction (increasing
(+ve ) or reduction (–ve)).
2) Either additives or multiplicative.
3) Can be avoided

This error correlates with the accuracy of

the results

8
25/02/1446

17

2.2 Determinate Errors - They Are Systematic

From the previous example can you find which

students had determinate or Systematic Errors if you
know the real value should be 5.0 :
Student (1): 5.0 5.6 4.5

Student (2): 8.2 5.8 5.5

Student (3): 5.0 5.0 5.1

Student (4): 9.3 9.2 9.3

Student (5): 4.7 5.0 5.3

Student (6): 6.7 6.5. 6.6

18

2.2 Determinate Errors - They Are Systematic

Detection of determinate errors :

1) Addition of a known amount of a
standard and measure its recovery
(recovery analysis).
2) Using reference materials.

9
25/02/1446

19

2.2 Determinate Errors - They Are Systematic

Some Common Determinate Errors
1. Instrumental errors: examples such as:
a) Faulty equipment
b) Uncalibrated weights
c) Uncalibrated glassware

2. Operative errors:
Personal errors (You have to avoid by gaining
experience).

3. Method errors:
Needs more attention.

20

2.2 Determinate Errors - They Are Systematic

Some Common Determinate Errors

3. Method errors: not easy to minimize
Sources include:
a) Coprecipitaion of impurities
b) Slight solubility of precipitate
c) Side reactions
d) Incomplete reactions
e) Impurities in reagents

How to solve it?

By running reagent blank (?).

10
25/02/1446

21

2.3 Indeterminate Errors - They Are Random

errors
• Cannot be avoided
• But it can be calculated (using Mathematical
Laws of Probability).
• Follow Normal distribution or Gaussian
Curve
We are 95% certain
that the true value falls
within 2σ (infinite
population), if there is
no systematic error.

22

2.3 Indeterminate Errors - They Are Random

errors
From the previous example can you find which
students have (Random errors) :
Student (1): 5.0 5.6 4.5

Student (2): 8.2 5.8 5.5

Student (3): 5.0 5.0 5.1

Student (4): 9.3 9.2 9.3

Student (5): 4.7 5.0 5.3

Student (6): 6.7 6.5. 6.6

This error correlates with the precision of the
results.

11
25/02/1446

23

2.3 Indeterminate Errors - They Are Random

errors
This error correlates with the precision of
the results

Statistical data treatment is used to measure the

random errors in the results or precision of
these results such as:
1) Standard Deviation
2) Coefficient of Variation
3) Range of Data
4) Confidence Interval about Mean Value

24

Random Errors vs. Systematic Error

You can’t have accuracy without good precision.
But a precise result can have a determinate or systematic error.

(a)large random errors (b) small random errors, (c) small random errors, no
(LOW ACCURACY large systematic error. systematic error).
AND PRECISION) (LOW ACCURACY BUT (HIGH ACCURACY AND
HIGH PRECISION) PRECISION)

12
25/02/1446

25

2.4. Ways of Expressing Accuracy

• Absolute Errors: difference between true
value(µ)
µ) and measured value (x).
• Mean Errors: difference between true value and
mean value.
• Relative Error: Absolute or Mean Errors
expressed as a percentage of the true value
Relative Error % =((µ
=((µ-x)/µ)∗
µ)∗100
µ)∗100 %

• Relative Accuracy: measured or mean value

expressed as a percentage of true value
Relative Accuracy %=((x/µ)∗
µ)∗100
µ)∗100 %

26

2.4. Ways of Expressing Accuracy

Example 2.4 :
The mean result of an analysis is 36.97 g
compared with accepted value of 37.06 g.
What is the %relative error.

Solution:
Absolute Error = µ-x
True value - measured value
37.06 -36.97 = 0.09 g
Relative Error = µ-x/µ
µ x 100 %
=0.09 g/ 37.06g x 100 %
= 0.2428 %=0.2 %

13
25/02/1446

27

2.5. Standard Deviation

The Most Important Statistic
Standard Deviation (σ) of Estimated Standard
an infinite set of Deviation (s) of a finite set
experimental data is of experimental data
theoretically given by (usually N < 30)

xi = individual
measurement xi = individual
µ = mean of infinite measurement
number of measurements x = mean of N number of
(true value) measurements
N = number of N = number of
measurements measurements

28

2.5. Standard Deviation

The Most Important Statistic
Example :
Find the mean and the standard deviation for
the following Analysis values 5.0, 5.6, 4.5 mL
Solution:
Data :
Individual measurements (xi) are:
5.0, 5.6, 4.5 mL
The number of measurements (N )=3

14
25/02/1446

29

i xi xi-xɸ (xi-xɸ)2
(xi-

1 5.0 0.0 (0.0)2=0.0

2 5.6 0.6 (0.6)2=0.4
3 4.5 -0.5 (-0.5)2=0.3

 i= N 2
 ∑ (x i − x ) 
S =  i =1
±  = ± (0 + 0.4 + 0.3) = ± (0.7 ) = ±0.6
(N − 1) (3 − 1) (2)

30

2.5. Standard Deviation

The Most Important Statistic
Relative standard deviation (RSD), or Coefficient
of Variation (CV)
Refer to the previous example and find RSD or CV for
the result.
Referring to the previous solution:
The solution =x̅ ± S = 5.0 ±0.6
S
RSD or CV = ( ) × 100%
x
0.6
= ( ) × 100% = 12 %
5.0

15
25/02/1446

31

2.6. Confidence Limit

How sure are you?

• Confidence interval: range within which the

true value might fall.

• The limits of this range is called confidence

limit.

• Probability (confidence level): is the likelihood

that the true value fall within the range.

32

2.6. Confidence Limit

How sure are you?

t: statistical factor that depends on the number of

degrees of freedom (ν = N-1).

Values of t at different confidence levels and degrees

of freedom are shown in table .

16
25/02/1446

33

Select a confidence level (95% is good) for the number of samples

analyzed (= degrees of freedom +1).
Confidence limit = x ± ts/√N.
It depends on the precision, s, and the confidence level you select.

34

2.6. Confidence Limit

How sure are you?

17
25/02/1446

35

2.6. Confidence Limit

How sure are you?
• The mean = 93.50%
• s=0.075 %
• N=3 N-1=2 if we checked the previous table
when v = 2 and at 95% t=4.303

= (93.50 ± 0.19)%
• The value is sitting between 93.31% and 93.69%

36

2.7. Rejection of a Result:

The Q Test
• If you did volumetric titration for a sample 4
replicates and it showed these volumes:
11.0 mL
9.5 mL
9.8 mL
9.3 mL
• Which volume is an outlier?

• Do you reject this value?

You need to do Q test to be able to decide

18
25/02/1446

37

2.7. Rejection of a Result:

The Q Test
• The Q test is used to determine if an “outlier” is due to
a determinate error.
If it is not should be retained (falls within the
expected random error).
• The ratio Q is calculated by:
arranging the data in decreasing order of numbers:

a
Q =
w
a: difference between “outlier” and nearest sorted
result.
w: range of results.

38

2.7. Rejection of a Result:

The Q Test
Arrange the data in decreasing order of
numbers:
11.0 mL 9.3 mL
9.5 mL 9.5 mL
9.8 mL a 9.8 mL
Q=
9.3 mL w 11.0 mL
• a =11.0 - 9.8 mL = 1.2 mL
• w = 11.0 - 9.3 mL = 1.7 mL
Q = 1.2/1.7 = 0.7059
Refer to table and the Qcalc < Qtable
retained (not rejected).

19
25/02/1446

39

QCalc = outlier difference/range.

If QCalc > QTable, then reject the outlier as due to a systematic error.

40

2.7. Rejection of a Result:

The Q Test

• Data:
• Number of analyses = 4
• 103, 106, 107, 114 meq/L
• 114 is the suspect result

20
25/02/1446

41

Solution:
a (the difference with the nearest) =114 -107 =7 meq/L

w (range of results)= 114 -103 = 11 meq/L

Q = (a/ w) = 7/11 = 0.636

Refereeing to table 3.3 Q95 of 4 observations = 0.829

Since , it is higher than 0.636 114 is retained (not

rejected)

21