Computational Details

In the following descriptions "X_R" refers to any particular item in the raw data set and "X_T" refers to the transformed value of "X_R"

For data with positive skew:_T

Square root transformation:
If all values of XR are equal to or greater than 1.0	XT = square_root(XR)
Otherwise	XT = square_root(XR+C1) whereTC1 is a constant selected to ensure that the lowest value of XR+C1 is equal to 1.0.
Natural logarithm transformation:
If all values of XR are equal to or greater than 1.0	XT = loge(XR)
Otherwise	XT = loge(XR+C1) whereTC1 is a constant selected to ensure that the lowest value of XR+C1 is equal to 1.0.
Inverse transformation:
If all values of XR are equal to or greater than 1.0	XT = 1−(1/XR)+C2 whereTC2 is a constant selected to ensure that the lowest value of XT is equal to 1.0.
Otherwise	XT = 1−(1/[XR+C1])+C2 whereTC1 is a constant selected to ensure that the lowest value of XR+C1 is equal to 1.0; and C2 is a constant selected to ensure that the lowest value of XT is equal to 1.0.

For data with negative skew:_T

Square root transformation (reflected):
In all cases	XT = C2−square_root(−XR+C1) whereTC1 is a constant selected to ensure that the lowest value of −XR+C1 is equal to 1.0; and C2 is a constant selected to ensure that the lowest value of XT is equal to 1.0.
Natural logarithm transformation (reflected):
In all cases	XT = −loge(−XR+C1)+C2 whereTC1 is a constant selected to ensure that the lowest value of −XR+C1 is equal to 1.0; and C2 is a constant selected to ensure that the lowest value of XT is equal to 1.0.
Inverse transformation (reflected):
In all cases	XT = (1/[−XR+C1])+C2 whereTC1 is a constant selected to ensure that the lowest value of −XR+C1 is equal to 1.0; and C2 is a constant selected to ensure that the lowest value of XT is equal to 1.0.