Q1.
Will including a new value in the data increase or decrease the mean?
Answer
It depends entirely on where the new value sits compared with the present mean.
- New value greater than the mean → the mean increases.
- New value less than the mean → the mean decreases.
- New value equal to the mean → the mean does not change.
Old data 4, 6, 8 → mean 6
Include 12 → (4 + 6 + 8 + 12) ÷ 4 = 30 ÷ 4 = 7.5 (up)
Include 2 → (4 + 6 + 8 + 2) ÷ 4 = 20 ÷ 4 = 5 (down)
Include 12 → (4 + 6 + 8 + 12) ÷ 4 = 30 ÷ 4 = 7.5 (up)
Include 2 → (4 + 6 + 8 + 2) ÷ 4 = 20 ÷ 4 = 5 (down)
Why it happens: with n values and mean a, adding x gives a new mean of (na + x) ÷ (n + 1). Subtracting the old mean,
(na + x)/(n + 1) − a = (na + x − na − a)/(n + 1) = (x − a) ÷ (n + 1).
So the shift has the same sign as x − a. In balance language: a value dropped on the right makes that side heavier, so the balance point has to slide right until the distances even out again.
(na + x)/(n + 1) − a = (na + x − na − a)/(n + 1) = (x − a) ÷ (n + 1).
So the shift has the same sign as x − a. In balance language: a value dropped on the right makes that side heavier, so the balance point has to slide right until the distances even out again.
Tip: the formula also tells you how much the mean moves — the gap x − a shared out over n + 1 values. A big class barely notices one new value; a small group feels it a lot.