The UNIVARIATE Procedure

Example 1: Univariate Analysis for Multiple Variables

Procedure features:
 VAR statement

This example computes the univariate statistics for two variables.
`options nodate pageno=1 linesize=80 pagesize=72;`
 ```data statepop; input State \$ citypop_80 citypop_90 Noncitypop_80 Noncitypop_90 Region @@; label citypop_80='1980 metropolitan pop in millions' noncitypop_80='1980 nonmetropolitan pop in millions' citypop_90='1990 metropolitan pop in millions' noncitypop_90='1990 nonmetropolitan pop in million' region='Geographic region'; datalines; ME .405 .443 .721 .785 1 NH .535 .659 .386 .450 1 NY 16.144 16.515 1.414 1.475 1 NJ 7.365 7.730 .A .A 1 PA 10.067 10.083 1.798 1.799 1 DE .496 .553 .098 .113 2 ...more lines of data... IA 1.198 1.200 1.716 1.577 3 MO 3.314 3.491 1.603 1.626 3 MT .189 .191 .598 .608 4 ID .257 .296 .687 .711 4 HI .763 .836 2.02 .272 4 ;```
 ```proc univariate data=statepop; var citypop_90 citypop_80;```
 ``` title 'United States Census of Population and Housing'; run;```

 Univariate statistics for both analysis variables appear on separate pages. Because each population value is unique, the mode is missing. By comparing the two sums in the Moments table, you find that the metropolitan population increased by 20 million (197.7 - 176.9) in ten years. By comparing the two medians in Basic Statistical Measures table or the Quantiles table, you find that the 1990 median metropolitan population increased to 2.423 million.