Z-score = (x-mean) / SD
= the number of standard deviations a datum is from the mean of the data set.
Convert a datum x to its distance in SDs from the mean of the data set.
Data set mapped/transformed/normalized to z-scores:
each datum x: its number of SDs from the mean: x → z
z = (x-mean) / SD
x z
--- ---
x̄ → 0
x̄+s → 1
x̄+1.5s → 1.5
x̄+2s → 2
x̄+3s → 3
x̄-s → -1
x̄-2s → -2
Within ±2 is "usual". [2,2]
≤-2 (-∞,-2] or ≥2 [2,∞) is "statistically significant",
i.e. maybe important.
mean=100 SD=10 Do this group in your head.
x=100 z=________
x=90 z=________ x=110 z=________
x=80 z=________ x=120 z=________
x=70 z=________ x=130 z=________
x=95 z=________ x=105 z=________
x=85 z=________ x=115 z=________
x=75 z=________ x=125 z=________
mean=127 SD=12 Do this group by hand/head.
x=127 z=________
x=115 z=________ x=139 z=________
x=103 z=________ x=151 z=________
x=91 z=________ x=163 z=________
Or use the SW:
x→Z-score→CDF ***
x=121 z=________ x=133 z=________
x=109 z=________ x=145 z=________
x=97 z=________ x=157 z=________