POLI 30 Lecture Notes - Lecture 4: Univariate, Standard Deviation, Frequency Distribution
Document Summary
Dimensionality of display should not exceed dimensionality of data. Univariate = one variable at a time, not looking at relationships. Distribution means the values a variable takes and how often it takes them. Simple frequency distributions and measures of central tendency and of spread. Calculate the percentage of cases with each value. % = number of cases x 100/ n. This only works for variables with a limited number of categories-often used with public opinion data. Add all the values and divide by the number of cases. Mean or average is often a weird value that doesn"t actually exist. Mean is highly influenced by outliers- nonrobust. The middle value- half are smaller and half are bigger. The median doesn"t care about the extreme values. Good when data isn"t highly skewed or doesn"t have big outliers. Gives every data point some influence- uses all the information. Protects against big outliers, and delivers meaningful values.