Uncertainties
Absolute
Some common instruments found in the laboratory
Ruler ± 1mm
Vernier callipers ± 0.1mm
Micrometer screw gauge ± 0.01mm
Thermometer ± 1° (or ½°)
Stopwatch ± 0.1s or ± 0.2s (reaction time)
Protractor ± 1° (or ½°)
Digital meters ± last dp, (eg 2.46V ± 0.01V)
Analogue meters – probably ±1 μA for a 0 – 100 μA range,
±0.1mA for a 0 – 10 mA range
Percentage uncertainty
For a single value:
uncertainty
100
value
For repeated readings:
½ range
100
mean value
Example: consider the following current readings: 0.50 A, 0.54 A, 0.46 A
½ 0.08
Percentage uncertainty = 100 8%
0.50
Combining Uncertainties
When multiplying/dividing quantities – add percentage errors
Examples:
if percentage uncertainty in I is 8%, percentage error in I2 = 8% + 8% = 16%
if percentage uncertainty in I is 8%, percentage error in I3 = 3 x 8% = 24%
if percentage uncertainty in I is 8%, percentage error in √I = ½ x 8% = 4%
Uncertainties in tables on paper 5
There are often large numbers of calculations to do. It is possible to work out percentage
uncertainties and then work out the absolute uncertainty
The absolute uncertainty is calculated by
1. Uncertainty = maximum value – value
2. Uncertainty = value – minimum value
3. Uncertainty = ½ (maximum value – minimum value)
Determining absolute uncertainties in quotients
Care needs to be taken with the use of maximum and minimum quantities.
In general, if r = p/q, then the largest and/or smallest value of R needs to be determined first.
A. largest r = largest p / smallest q
B. smallest r = smallest p / largest q
The absolute uncertainty is calculated by
1. Uncertainty = maximum value – value
2. Uncertainty = value – minimum value
3. Uncertainty = ½ (maximum value – minimum value)
Example
Determine resistance R when V = 3.6 0.2 V and I = 0.23 0.01 A
R = 3.6 / 0.23 = 15.7 Ω
(since both V and I are to 2sf, R should be to 2 or 3 sf at this stage)
A largest R = 3.8 / 0.22 = 17.3 Ω
B smallest R = 3.4 / 0.24 = 14.2 Ω
1. Uncertainty = 17.3 – 15.7 = 1.6
2. Uncertainty = 15.7 – 14.2 = 1.5
3. Uncertainty = ½ (17.3 – 14.2) = 1.6
Therefore R = 16 2 or 15.7 1.5 or 15.7 1.6, 15.7 2
All of these would be credited in a table of results on paper 5.
The calculations could also be determined using percentage uncertainties:
Percentage uncertainty in V = 0.2/3.6 x 100 = 5.5%
Percentage uncertainty in I = 0.01/0.23 x 100 = 4.3%
Percentage uncertainty in R = 5.5% + 4.3% = 9.8% = 10%
Note, 10% of 15.7 is 1.57 which is close to 1.5 or 1.6.
Uncertainties in logarithms in tables on paper 5
For logarithms to base 10, the absolute uncertainty is calculated by
1. Uncertainty = lg (maximum value) – lg (value)
2. Uncertainty = lg (value) – lg (minimum value)
3. Uncertainty = ½ (lg (maximum value) – lg (minimum value)
For natural logarithms, the absolute uncertainty is calculated by
1. Uncertainty = ln (maximum value) – ln (value)
2. Uncertainty = ln (value) – ln (minimum value)
3. Uncertainty = ½ (ln (maximum value) – ln (minimum value)
Example:
Consider V having the value of 3.6 0.2 V.
ln V = ln 3.6 = 1.28 or 1.281
Since V is to two significant figures, ln V should be to two (or three) decimal places.
Absolute uncertainty is either:
1. ln 3.8 – ln 3.6 = 0.05(4)
2. ln 3.6 – ln 3.4 = 0.05(7)
3. ½ x (ln 3.8 – ln 3.4) = 0.05(6)
Therefore ln 3.6 0.2 = 1.28 0.05
Significant figures in the uncertainties would not penalised in table of results.
Uncertainties in gradients and y-intercepts
Gradient
1. Uncertainty = gradient of line of best fit – gradient of worst acceptable line
2. Uncertainty = ½ (steepest worst line gradient – shallowest worst line gradient)
y-intercept
1. Uncertainty = y-intercept of line of best fit – y-intercept of worst acceptable line
2. Uncertainty = ½ (y-intercept of steepest worst line – y-intercept of shallowest worst
line)
Note:
If there is a false origin, the y-intercept will need to be calculated by substituting a plot on the
line into y = mx + c. To determine the y-intercept of the worst acceptable line by this method,
a plot is taken from the worst acceptable line and substituted into y = mx + c and the gradient
is from the worst acceptable line.
Uncertainties in quantities of the form lg k = y-intercept
To determine the value for k
k = 10 y-intercept
Uncertainty in k
1. Uncertainty = 10 best y-intercept - 10 worst y-intercept
2. Uncertainty = ½(10 max y-intercept of shallowest line - 10 y-intercept of steepest line)
Uncertainties in quantities of the form ln k = y-intercept
To determine the value for k
k = e y-intercept
Uncertainty in k
1. Uncertainty = e best y-intercept - e worst y-intercept
2. Uncertainty = ½(e max y-intercept of shallowest line - e y-intercept of steepest line)
