Search the site...

SASCRUNCH TRAINING
  • Home
  • Member's Area
  • How to Start
  • SAS Interface
  • Creating a Data Set
  • Practical SAS Training Course
  • SAS Certified Specialist Training Program
  • Proc SQL Course
  • Introduction to Time Series Analysis
  • SAS Project Training Course
  • Full Training / Membership
  • Sign up
  • About us
  • Contact us
  • Home
  • Member's Area
  • How to Start
  • SAS Interface
  • Creating a Data Set
  • Practical SAS Training Course
  • SAS Certified Specialist Training Program
  • Proc SQL Course
  • Introduction to Time Series Analysis
  • SAS Project Training Course
  • Full Training / Membership
  • Sign up
  • About us
  • Contact us
Sentry Page Protection
Please Wait...
Data Analysis [1-15]


Proc Univariate
Proc Univariate is commonly used in Exploratory Data Analysis.

It provides a full set of statistics for you to understand the distribution of the data.

Example

Copy and run the DEMO data set from the yellow box below.
Picture

Open the DEMO data set in SAS Studio.

The data set contains a list of event participants along with their age and income.

Now, enter and run the code below on SAS Studio:

Example

Proc Univariate Data=Demo;
Var Age;
Run;

By default, Proc Univariate displays 5 sets of statistics:



1. Moment

In statistics, a moment is a measure of the shape of the data.
Picture

The first four moments are:
  • 1st: Mean
  • 2nd: Variance
  • 3rd: Skewness
  • 4th: Kurtosis

In our example, all four moments are listed in the first table.

Picture

2. Basic Statistical Measures
Picture

The second table shows the basic statistical measures such as:
  • Mean
  • Median
  • Mode
  • Standard Deviation
  • Variance
  • Range
  • Interquartile Range


​3. Tests for Location
Picture

The three tests for locations are:
  • Student’s  test
  • sign test
  • Wilcoxon signed rank test

4. Quantiles

Picture

5. Extreme Observations
Picture

The statistics give us a good idea how the data is distributed.

The average age of the events participant is 36.6 with the standard deviation of 12.4.

The skewness is 0.3133, which shows the distribution being slightly skewed to the right.

The kurtosis is 0.1568. This indicates a "flag" distribution as opposed to a "peak" one.

Understanding the distribution of the data is an essential step prior to performing the more complex statistical analysis.

Proc Univariate is a quick way to do just that!

Exercise

Take a look at the DEMO data set again.

Run Proc Univariate on the INCOME variable.

Briefly describe the distribution of the INCOME.
Next

Need some help? 

Get Hint
Get Solution

Fill out my online form.
Already a member? Go to member's area.