SOC SCI 10A Chapter Notes - Chapter 5-6: Percentile, Asteroid Family, Standard Deviation
Document Summary
A standard distribution is one with a mean of 0 and a standard deviation of 1. These can have several different shapes when sketched. If you change the x score of an interval scale to another variable then the x score will have been linearly transformed or linearly rescaled: no cases are lost when the variables are rescaled. You can always go back and undo the rescaling process and recover the original data. Standardizing distributions linearly rescales the distribution so the mean is 0 and the standard deviation is 1. Standardization- replace each score of x by its deviations divided by the standard deviation. Standard scores are all referred to as z-scores. Standardization allows you to compare scores from one distribution with scores from the other directly. Standard scores are on an interval scale and that will tell you more information than scores on the ordinal level data scale will.
