Units and Dimensions - Absolute & Relative Errors
Description:
.fraction {
display: inline-block;
vertical-align: middle;
margin: 0 0.2em 0.4ex;
text-align: center;
}
.fraction > span {
display: block;
padding-top: 0.15em;
}
.fraction span.fdn {border-top: thin solid black;}
.fraction span.bar {display: none;}
.intsuma {
position: relative;
display: inline-block;
vertical-align: middle;
text-align: center;
}
.intsuma > span {
display: block;
font-size: 70%;
}
.intsuma .lim-up {
margin-bottom: -1.0ex;
}
.intsuma .lim {
margin-top: -0.5ex;
}
.intsuma .sum {
font-size: 1.5em;
font-weight: lighter;
}
.intsuma .sum-frac {
font-size: 1.5em;
font-weight: 100;
}
To decrease the errors in measurement, arithmetic mean (A.M.) is calculated to estimate the most correct value.
If ‘n’ measurements were taken as,
a1, a2, a3, .....an
Then arithmetic mean is defined as,
amean = a1 + a2 + a3 + ..... + ann = n∑i=1ain
The A.M. is much closer to the true value.
The magnitude (absolute value without the sign) of difference between A.M. and the individual measurements is called the Absolute Error. Therefore,
Δa1 = |a1 - amean|
Δa2 = |a2 - amean|
⋮
Δan = |an - amean|
Finally, the mean absolute error is defined as −
Δamean =
Δa1 + Δa2 + Δa3 + ....+ Δann = n∑i=1Δain
The absolute error is expressed with the measurement in the following way −
a = amean ± Δamean
Example − Following measurements are made of an iron rod in the laboratory
10 cm, 12 cm, 9 cm, 8 cm, 13 cm
Find the mean absolute error.
Method − First we calculate the A.M. as,
amean = 10 + 12 + 9 + 8 + 135 = 525 = 10.4 cm
Now calculating absolute error for each measurement, Δan = |an − amean|
|10.4 - 10| = 0.4cm
|10.4 - 12| = 1.6 cm
|10.4 - 9| = 1.4 cm
|10.4 - 8| = 2.4 cm
|10.4 - 13| = 2.6 cm
Now, mean absolute error = 0.4 + 1.6 + 1.4 + 2.4 + 2.65 = 1.68 cm
Hence, the measurement can be expressed as −
length = 10.4 ± 1.68 cm
Relative and Percentage Error
Relative Error is defined as −
Relative Error = Δameanamean
Percentage Error is defined as −
Percentage Error = Relative Error × 100
Relative and percentage errors are unit-less.
Example −
Person ‘A’ designs 1 m ruler with mean absolute error of 1 cm.
Relative errorA = 0.01 m1 m = 0.01
Person ‘B’ designs 10 km road with mean absolute error of 1 cm.
Relative ErrorB = 0.01 m1000 m = 0.00001
Clearly, Person ‘B’ is much better at his job compared to Person ‘A’. This aspect can only be revealed by Relative error. (and not by Absolute error)