\(x =\) number of measurements
\(S = n_1+n_2+n_3+.......+n_x=\) sum of original mesaurements
We know the average of the original measurements is \(\frac{S}{x} = 7.3\)
The corrected measurement is given by \(m = -0.5n+10\).
Here, the original measurement is being multiplied by \(-0.5\) and then \(10\) is being added to it. Let incorporate that into the average.
New Average \(= \frac{(-0.5)n_1+10+(-0.5)n_2+10+........+(-0.5)n_x+10}{x}\)
\(=\frac{(-0.5)(n_1+n_2+n_3+........+n_x)+10x}{x}\)
\(=\frac{(-0.5)S}{x}+\frac{10x}{x}\)
\(=(-0.5)(7.3)+10\)
\(=6.35\)
Clearly, the average is going through the same operations as the original mesaurements. The original average is being multiplied by -0.5 and this product is being added to 10 to get the corrected average.
To find the corrected standard deviation, there are two ways ->
1. Adding a constant to all numbers in the set does not change the standard deviation, but multiplying all the numbers in the set by a constant scales the standard deviation by |constant|.
If the original SD was 2.9, then the corrected standard deviation will be \(|-0.5|*2.9 = 1.45\) (because the original values in the set were multiplied by constant \(-0.5\). Adding \(10\) to all the values in the set did not affect the \(SD\))
2. The formula for Standard Deviation is \(SD = \sqrt{\frac{Σ(x_i - mean)^2}{n}}\), where \(x_i\) is each value in the set and n is the number of values in the set. To find the corrected standard deviation, we can just plug in the new corrected values in place of \(x_i \) and mean.
New \(SD = \sqrt{\frac{Σ[((-0.5)n_i+10) - ((-0.5)mean_{old}+10)]^2}{x}}\)
\(=(-0.5) \sqrt{\frac{Σ(n_i+10) - (mean_{old}+10)^2}{x}}\)
\(=(-0.5) \sqrt{\frac{Σ(n_i+10 - mean_{old}-10)^2}{x}}\)
\(=(-0.5) \sqrt{\frac{Σ(n_i - mean_{old})^2}{x}}\)
We know \(\sqrt{\frac{Σ(n_i - mean_{old})^2}{x}} = 2.9\)
\(=(-0.5)(2.9)\)
\(=-1.45\)
But \(SD\) cannot be negative, because it's essentially a measure of distance, and only the magnitue matters in distance.
New \(SD = |-1.45| = 1.45\)
Takeaway ->
1. When adding/subtracting/multiplying/dividing a constant to/from all numbers in the set, do the same operations on the old mean to get the new mean.
2. Adding/subtracting a constant to/from all numbers in the set has no affect on the standard deviation. Multiplying/dividing a constant with all the numbers in the set scales the standard deviation by |constant|.