CHM 256: EVALUATION OF

CHAPTER 2 EXPERIMENTAL DATA

 TYPES OF
SIGNIFICANCE ERROR :
FIGURE
 SYSTEMATIC ERROR
(DETERMINATE)
 RANDOM ERROR
(INDETERMINATE)
CHAPTER
Q TEST 2
PRECISION
AND
ACCURACY
CONFIDENCE
LIMIT  MEAN
 MEDIAN
 ABSOLUTE ERROR, RELATIVE
ERROR, % RELATIVE ERROR
 DEVIATION
 STANDARD DEVIATION
 RELATIVE STANDARD DEVIATION
 SIGNIFICANCE FIGURES
- The meaningful digits in a measured or calculated quantity.
- The number that reflects the precision of the measurement.
RULES:
• Any digit that is not zero is significant
1.234 kg 4 significant figures
12.3 mL 3 significant figures
• Zeros between non zero digits (numbers) are significant
606 m 3 significant figures
1.2034 mol 5 significant figures

•Zeros to the left of the first non zero digit ( 0 used to locate
decimal point) are NOT significant
0.08 L 1 significant figures
0.0012 L 2 significant figures
 SIGNIFICANCE FIGURES

•Zero existing to the right of the decimal point & zero at the
end & in the middle of the numbers are significant
2.0 mg 2 significant figures
0.2900 g 4 significant figures
0.00402 mL 3 significant figures

• Numbers that do not contain decimal points, zeros after the

last nonzero digit may or may not be significant.
400 cm 1or 2 or 3 significant figures
4 x 102 1 significant figures
4.0 x 102 2 significant figures
 SIGNIFICANCE FIGURES

How many significant figures are in each of

the following measurements?
24 mL 2 significant figures

3001 g 4 significant figures

0.0320 m3 3 significant figures

6.4 x 104 molecules 2 significant figures

560 kg 2 or 3 significant figures

 SIGNIFICANCE FIGURES involve CALCULATION

Addition or Subtraction
The number of significant figures in the result is set by the
original number that has the smallest number of decimal
points.
89.332 3 s.f after decimal point
+1.1 1 significant figure after decimal point
90.432 round off to 90.4

3.70 2 significant figures after decimal point

-2.9133 4 s.f after decimal point
0.7867 round off to 0.79
+-
LEAST
DECIMAL
POINT
 SIGNIFICANCE FIGURES involve CALCULATION
Multiplication or Division
The number of significant figures in the result is set by the
original number that has the smallest number of significant
figures
4.51 x 3.6666 = 16.536366 = 16.5
5 s.f
3 sig figs round to
3 sig figs

6.8 ÷ 112.04 = 0.0606926 = 0.061

5 s.f
2 sig figs round to X‚
LEAST
2 sig figs SIGNIFICANCE
FIGURES
 EXERCISE

Calculate and state the answers to

the correct number of significant
figures.

i) 17.020 cm + 0.145 m – 12.3 cm

ii) 397.32 ÷ 6.351

8
EXERCISE
 TYPES & SOURCES OF ERROR

DETERMINATE INDETERMINATE
ERRORS ERRORS
(SYSTEMATIC) (RANDOM)

10
 TYPES OF ERROR
1. DETERMINATE ERRORS (SYSTEMATIC)
 Definition: A constant error that originates from a fixed
cause, such as error in the design of an equipment of exp
 Determinate errors are nonrandom and occur when
something is wrong with the measurement.
 The mean of a set data to differ from the accepted value
 Arise from known sources that can be identified
 It can be avoided or corrected.
 A l l measuring devices are sources of determinate errors
 TYPES OF ERROR
1. DETERMINATE ERRORS (SYSTEMATIC)
Known as
SYSTEMATIC
ERROR
Results are
either high/low
- A systematic error can be recognized from multiple
measurements (replicate measurements) where the results
either high or low when compared to the true value.
- EXAMPLE: The measurements of copper’s density give the
values;
9.54, 9.55 and 9.56 g/cm3, but the true value of copper’s
density is 8.96 g/cm3.
- It can be suspected that systematic error occur because all
of the measured values are consistently too high.
DETERMINATE ERRORS (SYSTEMATIC)
INSTRUMENT ERROR OPERATIVE ERROR / METHOD ERROR
PERSONAL ERROR
 Faulty equipment  Personal errors:  Most serious error
 Uncalibrated  Poor sample  Incomplete
glassware preparation reaction/slow
(pipettes,  Incorrect reading  Impurities in
burettes, vol flask) of meniscus reagent
 Glassware not at  Color of solution  Reagent not
calibration Temp. at endpoint of stable
 Avoided by proper titration  Slight solubility
calibration and  Mathematical of precipitate.
maintaining the errors  Side reaction
equipment  Prejudice in
estimating  Avoided by proper
measurement method selection &
(bias) development
 Reduced by training,
experience,discipline
13
 TYPES OF ERROR INSTRUMENT
1. DETERMINATE ERRORS (SYSTEMATIC) ERROR

Any glassware used for quantitative measurements (pipettes,

burets) are potential source of systematic error :
(a)if they are not calibrated
(b)the solution temperature is different from the temperature of the
glassware
(c)Inaccuracy in determining the level of meniscus

For electronic measuring apparatus, systematic errors can come

about as the result of low batteries, poor contacts within the device,
sensitivity to temperature and humidity instrument malfunction.

HOW TO AVOID/ CORRECT

Always remember to do calibration of glassware or instruments
 TYPES OF ERROR
1. DETERMINATE ERRORS (SYSTEMATIC) METHOD
ERROR

HOW TO AVOID/ CORRECT

i)Using standard sample
Analysis using standard reference materials
Materials that contain one or more analyses at exactly known
concern.

ii)Performing ‘blank determination’.

In a blank determination, all steps of the analysis are
performed in the absence of the sample.
Blank solution: Solution that contains the solvent and all
reagents in an analysis except the analyte)

iii)Variation in sample size.

iv)Perform second independents and reliable analytical method
 TYPES & SOURCES OF ERROR
2. INDETERMINATE ERRORS (RANDOM)
 Occur when a system of measurement is extended to its
maximum sensitivity (successive measurements)
 Errors that originate from unknown sources (cannot be
identified). They are caused by many uncontrollable variables
and therefore cannot be totally eliminated.
 Exist in every measurement
 The detection of is difficult because they are VERY SMALL and
NOT positively identified or measured.
 The indeterminate errors are responsible for DEVIATIONS that
occur in a series of experimental data

16
DETERMINATE ERRORS INDETERMINATE ERRORS
(SYSTEMATIC) (RANDOM)
STATISTICAL PROCEDURES TO HANDLE ANALYTICAL
DATA
MEAN
ERROR

ACCURACY MEDIAN

PRECISION RANGE
 MEAN (AVERAGE), x
 Mean is the AVERAGE reading of all data or
measurements that are obtained from an experiments.
Mean can be calculated by dividing the sum of replicate
measurements by the number of measurements in the
set.

Mean for sample: x =

xi
i1
n
where n is the number of samples.
 MEAN for SAMPLE &
POPULATION
• SAMPLE • POPULATION
• A SMALL set of data • LARGE SET of data
<20 (>20)
n N
 xi  xi
x = i1 µ = i1
n N
where n is the number where N is very large
of samples. for real population
number
(usually small numbers
<20)
(µ is the true mean )
 MEAN (AVERAGE), x

EXAMPLE

• Calculate the mean of a population. Our

example will use the ages of four young
children: { 5, 6, 8, 9 }.
5+6+8+9
x= =7
4

21
 MEDIAN
 Value that is in the middle result of a set of data.
 Median of a set of replicate data is the middle result
when the data are arranged by increasing in size.

EXAMPLE Find the median for the following data:

5, 6, 8, 9,
Median = 7

RANGE
Difference in magnitude between highest value &
lowest value
xhighest - xlowest
PRECISION & ACCURACY
PRECISION

1. Precision is how close the measured values are to

each other measurements that been made in exactly
the same way.
2. Precision describes the reproducibility of results, the
agreement between two or more.
3. Expressed in terms of DEVIATION
4. If the DEVIATION between the measurements is small,
your data are PRECISE.
ACCURACY

1. Accuracy is how close a measured value to the

actual (true) value or expert value which we believe to
be correct.
2. Expressed in terms of ERROR/ABSOLUTE ERROR

Error, E = ( Value obtained - True Value )

3. If the error is small, then our measurements are

ACCURATE and it shows the accuracy of the results
obtained.
4. The errors are reported as % error
or relative error.
 CANNOT have
accuracy without
PRECISION vs ACCURACY GOOD precision.
But a precise result
can have determinate/
systematic error.

Low accuracy, low precision Low accuracy, high precision

High accuracy, low precision High accuracy, high precision

 WAYS TO DESCRIBE
PRECISION & ACCURACY

Precision Accuracy
o Absolute
o Deviation from
error
mean
o Relative
o Deviation from error
median o Percentage
o Range error
o Standard
Deviation
 37

ABSOLUTE ERROR, E

E = ( xi - xt )

Where:
xi value obtained experimentally,
xt is the true value

The sign of absolute error indicate whether the value

in question is either high or low as compared to true
value.
 RELATIVE ERROR, Er

xi - x t Relative error from mean is Er

=
Er =
x - mean
xt i

mean

PERCENT RELATIVE ERROR, %E

xi - x t
% E= x x 100
t
 DEVIATION, D
Deviation(d) = How much each measurement
differs from the mean
d = xi –

RELATIVE DEVIATION
% deviation (d) = deviation x 100
mean
d = xi –
 STANDARD DEVIATION for
sample, (s)
Standard Deviation,(s) is
n more significant quantity
S=  (xi  x )2 in that measures the
i 1 PRECISION or scatter
n1 of sample data set

The n – 1 term: represents the DEGREE OF FREEDOM.

In the calculation of standard deviation, the degree of
freedom is reduced by ONE.
 STANDARD DEVIATION for
population, ()
When N  , x  and s  
N
True Standard Deviation ,  =  (x i μ) 2
i1
N

% RELATIVE STANDARD
DEVIATION (% RSD)
s
%RSD  x 100%
x
EXAMPLE

• Calculate the mean and the standard

deviation of the following set of
analytical results:

15.67 g, 15.69 g, 16.03 g

32
EXAMPLE
xi xi - x (xi – x)2
15.67
15.69
16.03
∑ ∑ ∑

47.39
x  15.80
3

s
1 N

N 1 i1
 
xi  x
2

0.0819
s  0.20g
31 33
EXERCISE

A student carried out an experiment to measure the boiling

of an organic compound at atmospheric pressure. The
results obtained are as follows.
Exp.1. 54.9 oC
Exp.2 54.4 oC
Exp.3 1. oC
Exp.4 2. oC
 The true boiling for the compound is 54.0 oC. Calculate:
(i) Error
(ii) Percentage error (Relativeerror)
(iii) Mean
(iv) Deviation
(v) Percentage/relative deviation
(vi) Median
SOLUTION

i) Error = Measured value – truevalue

ii ) %/Relative error = Error x 100

Truevalue
iii) Mean:
Mean 

Mean  #

iv)Deviation = Measured value – mean

v) % deviation = deviation x 100

mean

vi) Median:
 Q TEST
• Deciding whether an outlying value (OUTLIERS)
in a set of replicate results should be
RETAINED or REJECTED in calculating the
mean for the set of data.

OUTLIER
Outlying value that appear to be
different from all other data in a set
of replicate measurements
Can be the results of error
 Q TEST
STEPS TO APPLY Q TEST
OUTLIER
 Arrange the data in an increasing order.
 Determine Q by using the formula;
 Calculate the difference between the suspect value
and its nearest neighbour, (a).
 Calculate the range (difference between highest and
lowest values), (w).

Qexp =
a (Gap btw outlier and closest no to it)

w (Range)

Qexp is then compared to the Qtable (from table).

If the value of Qexp  Qtable, the questionable result/outlier
can be REJECTED due to a systematic error.
 If Qexp < Qtable , the data is retained, NOT an outlier
Q TEST

TIPS?
EXAMPLE
Trial IV appears incorrect, check using Q-test at 90% confidence
whethertrial IV shouldbe rejected or accepted.
Trial I II III IV V VI VII
Volume 25.75 25.62 25.52 25.21 25.65 25.60 25.71
of HCl used
(mL) STEPS?

Solution:
25.21 25.52 = 0.31
= = 0.57
25.75  25.21 0.54

QExp > QTable ie. 0.57 >0.51 , therefore the reading

25.21 should be rejected
EXERCISE

• An analysis of iron ore gives the following

results;
33.78%, 33.84%, 33.80% and 33.15%.
Which of the values should be removed at
95% confidence?

41
 Arrange the data

a =?
w=?
Q calculated = a = =
w
Q table =
Q calculated ? Q table
Qcalc Reject or accept?
 CONFIDENCE LIMIT

The true value of the mean, µ for a population of

data can never be determined exactly because such
a determination requires that infinite number of
measurements be made.

Therefore limits are set within which experimentally

determined mean, (sample mean) can be found.
 CONFIDENCE LIMIT
• Range within which true value,µ might fall with a given
probability.
• The range of estimated value of population mean is called
the confidence interval and the limit of this range is called
the confidence limit.
• Confidence Level: The probability of population mean lies
within the confidence interval (90%, 95% or 99%)

• Confidence interval for µ:

• s = standard deviation
• N = number of sample
• t = statical factor (from the table)
– Degree of freedom (N –1)
– 90%, 95% or 99% depend on
the question
 CONFIDENCE LIMIT
STEPS TO APPLY

1. Calculate experimental mean

2. Calculate standard deviation
3. Check t-factor table for value at 90/95/99%
confidence and degree of freedom
4.Use confidence interval formula to calculate the limit
Confidence limit = x ± ts/√N.

5. Write a conclusion
• The true value lies within ? to ?.
Select a confidence level (95% is good) for the number of samples analyzed
(= degrees of freedom -1).
N
Confidence limit = x ± ts/√N.
It depends on the precision, s, and the confidence level you select.

©Gary Christian, Analytical Chemistry, 6th Ed. (Wiley)

EXAMPLE

• Calculate the confidence limit at 95%

confidence level. Our example will use
the ages of four young children: { 5, 6,
8, 9 }.

ts 7
(3.18)(1.8)
x 4
N 7  2.862
47
EXAMPLE

• Conclusion: It means that there is 95%

chance that the true mean, µ lies within
the range 7 ± 2.862 (4.138 to 9.862)

ts 7
(3.18)(1.8)
x 4
N 7  2.862
48
EXERCISE

• A set of volumetric calibration measurement give

observed values of 4.995, 5.004, 4.998, 4.999 and
4.997 mL. Calculate the 90 % and 95 % confidence
intervals for the volume
ts
µ= x 
• For 90 % confidence intervals for N
the volume:

1. Find x and s
x: ? and s: ?
2. Look up the Table under ? % and degree of
freedom (N- 1)
t = ?
Answer ?
EXERCISE

• For 95 % confidence intervals for the volume:

1.Find x and s
x: and s:
2. Look up the Table under ? % and degree of
freedom (N- 1)
t =
Answer: ?