January 2017
Beginner
882 pages
203h 41m
English
Variables that measure income are typically right skewed. This skewness makes the mean and SD less useful as summaries. A transformation changes the scale of measurement in a fundamental way that alters the distribution of data. The log transformation is the most useful transformation for business data because it shows variation on a relative scale. Economics sometimes use logs to measure utility. Variation on a log scale is equivalent to variation among the percentage differences. The log transformation also allows us to measure association between income and other factors using the correlation.