Measurement and Data Processing: I. Uncertainties and Errors in Measurement and Results

11.
MEASUREMENT AND DATA PROCESSING

I. Uncertainties and errors in measurement and results
Quantitative data are obtained from measurements and always associated with random
errors determined by the apparatus, and by human limitations such as reaction times.
Error is defined as the difference between an observation (directly measured or
calculated) and the true value.
Error = Observation (practical value) – True value (Literature value)
There are two kinds of errors: Random and Systematic
(i) Random errors

These are caused by the limitations of the measuring apparatus and other
uncontrollable variables that are inevitable in any expt.
Random errors arise from the imprecision of measurements and can lead to readings
being above or below the “true” value. Random errors are observed when an
experimenter approximates a reading with an equal probability of being too high or too
low.
Some causes of random errors:
 Readability of the measuring instrument

 Effects of changes in the surrounding e.g. T. variations, air currents etc
 Insufficient data
 Observer misinterpreting the reading
 Decision on when indicator changes colour during an acid-base titration
Random errors can be reduced with the use of more precise measuring equipment or
their effect can be minimized through repeating measurements so that the random
errors cancel out. The random errors can never be completely eliminated.
(ii) Systematic errors

Systematic errors arise from a problem in the experimental set-up (experimental design
and procedure) that results in the measured values always deviating from the “true”
value in the same direction—that is, always higher or always lower.
Examples of causes of systematic error:
 miscalibration of a measuring device or poor insulation in calorimetry

experiments.
 Way in which reading is taken e.g. meniscus in a burette
 Overshooting the volume during titration
 Using an acid-base indicator whose end-point does not correspond to the
equivalent point of the titration
 Heat losses in an exothermic reaction leading to smaller T. changes
Page 1|6
The effect of a systematic error cannot be reduced by repeating the readings; but can
be reduced by modification to the experiment.
Systematic errors can be identified by comparison with the accepted literature values
e.g. If calculated value = - 55.8 KJmol-1 and Literature value = - 57.3 KJmol-1
% error = [(Expt. value – Lit. value) / Lit. value] x 100
% error = [(55.8 – 57.3) / 57.3] x 100 = 2.6%
% uncertainty due to random error ( i.e 55.8 ± 0.1 oC)
% random error = (0.1 / 55.8) x 100 = 0.2%
If the % calculated systematic error is greater than any % random error; it suggests that
the expt involves some systematic errors.
Precision and accuracy

Accuracy refers to how close a measurement is to the correct value of a particular
quantity.
Precision refers to how close several measurements of the same quantity are to each
other.
The smaller the systematic error, the greater the accuracy and the smaller the random
error, the greater the precision.
Consider the following examples:
(i) Volume readings (cm3) – 21.2, 21.3, 21.1, 21.3. 21.1, 21.2
The mean = 21.2
Thus volume = 21.2 ± 0.1 cm3 (high precision)
(ii) Volume reading (cm3) - 21.0, 21.9, 22.1, 21.2, 20.7, 21.5
The mean = 21.4
Thus volume = 21.4 ± 0.7 cm3 (low precision)
Note: lower precision indicates greater spread of the values
Uncertainties in calculated results

Check on the apparatus for the degree of tolerance. If not shown, then estimate the
uncertainty.
The uncertainty of a measurement is half the smallest division to which you take a
reading.
Uncertainties are usually quoted to 1 s.f. e.g. 69 ± 0.5 cm 3; 2.46 ± 0.005 g
Uncertainties can be stated as absolute and percentage errors.
Absolute errors indicate the magnitude of the uncertainty in the result in the same units
as the result e.g.
consider measuring 25.0 cm3 with a 25 cm3 pipette that measures to ± 0.1 cm3. Absolute
uncertainty is 0.1 cm3.
Percentage error indicates the uncertainty as a fraction of the result multiplied by 100%.
e.g. consider the above measurement. % error = (0.1 / 25.0) x 100 = 0.4%
Page 2|6
If a burette can measure 50.0 cm3 with an uncertainty of ± 0.1 cm3 is used to measure
50 cm3 of solution; % error = (0.1 / 50) x 100 = 0.2%
If the burette is used to measure 20 cm3, then the absolute uncertainty is still ± 0.1 cm 3,
but the % error = (0.1 / 20) x 100 = 0.5%
Significant figures
a) Addition or subtraction
The general rule is that the number of decimal places should be preserved e.g.
78.3 + 52.6 = 130.9
146.3 – 21.8 = 124.5
The absolute uncertainties can be added e.g add two volumes of 25.0 ± 0.1 cm 3
i.e. (25.0 ± 0.1 cm3) + (25.0 ± 0.1 cm3) = 50.0 ± 0.2 cm3
For addition and subtraction, the answer should have the same number of d.p. as the
quantity with the fewest number of d.p.
e.g. 3.247 + 41.36 + 125.2 = 169. 8 (1 d.p.) When
adding or subtracting, it is the number of d.p. that is important: e.g.
7.56 + 3.10 = 10.66
(3 sf) (3 sf) (4 sf)
22.36 – 5.16 = 17.20

(4 sf) (3 sf) (4 sf)
Page 3|6
b) Multiplication or division
% uncertainties should be used during calculation and then converted back to absolute
uncertainty when final result is presented.
The absolute uncertainty can be worked out from the % uncertainty using the equation:
Absolute uncertainty = (% uncertainty / 100) x value
e.g. If the final value of a calculation is 0.518 ± 1%

Then absolute uncertainty = 1/100 x 0.518 = 0.005
Thus final answer: 0.518 ± 0.005
When multiplying or dividing quantities with uncertainties, the % uncertainties should be

added. e.g.
What is the absolute uncertainty when 2.57 ± 0.01 is multiplied by 3.456 ± 0.007 and
how many sf can the answer be quoted?
2.57 x 3.456 = 8.88192
% uncertainties: 0.01/2.57 x 100 = 0.39% and 0.007/3.456 x 100 = 0.20%
Total uncertainty = 0.39 + 0.20 = 0.59%
Absolute uncertainty = 0.59/100 x 8.88192 = 0.05
Final answer: 8.88 ± 0.05
When multiplying or dividing a quantity with an uncertainty by a pure number, the

absolute uncertainty is multiplied/divided by that number so that the percentage
uncertainty stays the same e.g
If 12.12 ± 0.01 (% uncertainty = 0.08%) is multiplied by 3, the answer is 36.36 ± 0.03 (%

uncertainty = 0.08%). If 2.00 ± 0.03 (% uncertainty = 1.5%) is divided by 3, the answer
is 0.67 ± 0.01 (% uncertainty = 1.5%).
When multiplying or dividing, it is the number of s.f. that is important.

The number with the fewest s.f. used in the calculation determines how many s.f. should
be used in quoting the answer. e.g
i) Heat = 0.125kg x 7.2oC x 4.18 KJ Kg-1 oC-1 = 3.762 KJ = 3.8 KJ
ii) (3.14 x 2.751) / 0.64 = 13 (2 sf)
Page 4|6
II. Graphical techniques
Graphs provide an instant visual representation of trends in data.
By plotting a suitable graph to give a straight line, or some other definite relationship
between the variables, we can use graphs to predict unknown values.
Methods used to achieve this include:
 (i) measuring the intercept
 (ii) measuring the gradient
 (iii) extrapolation
 (iv) interpolation
Extrapolation – extending the graph to determine an unknown value that lies outside the
range of the values measured.
Interpolation – determining an unknown value within the limits of the values already
measured.
When plotting a graph, the convention is to plot the independent variable (manipulated)
on the horizontal axis (x-axis) and the dependent variable on the vertical axis (y-axis).
Graphs should be plotted using SI units wherever possible. All graphs should have:
 (i) a suitable title
 (ii) label the axes with both quantities and units clearly shown
 (iii) the scale of the axes should be appropriate – no uneven jumps and available
space used as effectively as possible
 (iv) plot all the points correctly
Once the points are plotted, determine whether to join the points directly or whether to
choose the line of best fit. Line of best fit may be a straight line or a smooth curve.
Page 5|6
Normally plot at least 5 readings.
Sometimes it is sensible not to just plot variables actually measured but to manipulate
them to give a straight-line graph. e.g. ideal gas equation, PV = K
Plot of this (P vs V) gives a curve – not easy to see relationship from a curve
Page 6|6

Measurement and Data Processing: I. Uncertainties and Errors in Measurement and Results

Uploaded by

Copyright:

Available Formats

Measurement and Data Processing: I. Uncertainties and Errors in Measurement and Results

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Measurement and Data Processing: I. Uncertainties and Errors in Measurement and Results

Uploaded by

Copyright:

Available Formats

11.

MEASUREMENT AND DATA PROCESSING

(i) Random errors

 Readability of the measuring instrument

(ii) Systematic errors

 miscalibration of a measuring device or poor insulation in calorimetry

% error = [(Expt. value – Lit. value) / Lit. value] x 100

% error = [(55.8 – 57.3) / 57.3] x 100 = 2.6%

% uncertainty due to random error ( i.e 55.8 ± 0.1 oC)

% random error = (0.1 / 55.8) x 100 = 0.2%

Precision and accuracy

Uncertainties in calculated results

22.36 – 5.16 = 17.20

e.g. If the final value of a calculation is 0.518 ± 1%

When multiplying or dividing quantities with uncertainties, the % uncertainties should be

When multiplying or dividing a quantity with an uncertainty by a pure number, the

If 12.12 ± 0.01 (% uncertainty = 0.08%) is multiplied by 3, the answer is 36.36 ± 0.03 (%

When multiplying or dividing, it is the number of s.f. that is important.

i) Heat = 0.125kg x 7.2oC x 4.18 KJ Kg-1 oC-1 = 3.762 KJ = 3.8 KJ

ii) (3.14 x 2.751) / 0.64 = 13 (2 sf)

You might also like