Tukey estimate of power to which observations must
be raised to achieve additivity A transformation of the items in the scale that will reduce the nonadditivity component of the model. It is expressed as a power to which the values of the variables or items should be taken. For instance, if Tukey's estimate equals -1, you should take the reciprocal of each of your item or variable scores. If it equals 1/2, take the square root. If Tukey's estimate is small (close to zero), use the logarithmic transformation.
In making your transformations, it is recommended that you keep them simple by rounding power transformations to integers.
