Analyzing the Setup
We begin with a dataset of n=50 observations. The initial mean is given as xˉold=15.
The total sum of these observations is calculated using the definition
xˉ=n∑x. Therefore, the old sum is:
∑xold=50×15=750
Unmasking the Culprit
We are informed that one observation x was recorded incorrectly and should have been x′. We are given the constraint x′+x=70.
After replacing the incorrect value, the new mean becomes
xˉnew=16. The new sum is:
∑xnew=50×16=800
The difference between the new sum and the old sum is 800−750=50. This difference represents the net change: x′−x=50.
We now solve the system of linear equations:
1) x′+x=70
2) x′−x=50
Adding these equations yields 2x′=120, which gives x′=60. Substituting this back, we find the incorrect value x=10.
The Energy of the Data
To find the variance, we utilize the formula:
σ2=n∑x2−(xˉ)2
Given the old variance
σold2=22=4, we substitute the known values:
4=50∑xold2−152
This simplifies to
4=50∑xold2−225. Solving for the sum of squares, we obtain:
∑xold2=50×229=11450
The Final Restoration
We now perform the data correction by removing the contribution of the incorrect value and adding the correct one:
∑xnew2=∑xold2−x2+(x′)2
Substituting our values:
∑xnew2=11450−102+602=11450−100+3600=14950
Finally, we calculate the new variance using the corrected sum of squares and the new mean:
σnew2=5014950−162
The first term simplifies to
299, and the second term is
256. Subtracting these gives:
σnew2=299−256=43
The correct variance is 43.