Why Randomize?
	• Members of the treatment (“TX”) and control groups tend to be comparable on all variables, known and unknown • Provides basis for statistical analysis