Frequency & Sampling Distributions

Dr. Philip N. Howard

Department of Communication

University of Washington

	Frequency Distribution	Sampling Distribution
Units in curve	Observations (e.g. individuals, firms cities, etc.)	Estimates of parameters calculated from repeated, identical samples of the same population
Normal Distribution	Rarely	Usually
Basis	Empirical: cases systematically sampled from a defined population	Theoretical (e.g. central limits theorem) although they can be simulated
Average value	Sample mean	‘True’ or population mean
Distribution unit	Standard deviation	Standard error
Primary use	Description of a specific sample	Inference; hypothesis testing about a population
Value to social science	Limited beyond description-very few measures are normally distributed	Great-the foundation for almost all quantitative work
Increased sample	Doesn’t change shape	Distribution becomes more peaked