First of all, the computers in the labs in LeConte should have PROC INSIGHT loaded onto 
them.  If you've purchased a version of SAS for your home computer, and it doesn't have 
PROC INSIGHT loaded, or PROC INSIGHT doesn't seem to work on it, you can ask the computer 
services people to load PROC INSIGHT onto your version.  If they won't or you can't, 
here are some alternate instructions that will produce similar graphs/analyses as 
PROC INSIGHT.  You can use these instructions for the homework; however, the graphs won't 
look as nice as those PROC INSIGHT produces, so if you can use PROC INSIGHT, do so.

Don't forget that if you print the output via a MS Word document, you should use a font 
as "Courier New" or "SAS Monospace" (which may be the default when copying SAS output into 
Word).

Basic summary statistics
------------------------------

PROC UNIVARIATE gives these.  See the examples on the course web page.

----------------------------------------------------------------
Bar graphs for categorical variables
----------------------------------------------------------------

For example, consider the data in problem 2.9.  Instead of PROC INSIGHT, you can use 
PROC CHART.  If the data set is called ceoeduc and the variable is called educlevel, use 
the following commands instead of the PROC INSIGHT commands:

proc chart data=ceoeduc;
vbar educlevel;
run;


--------------------------------------------------------------------
For stem-and-leaf, Q-Q (normal probability) plots, and box plots
--------------------------------------------------------------------

PROC UNIVARIATE will give these if you include the PLOT FREQ option.  For example, if our 
data set is called gasdata and the variable of interest is called mileage, use the following
 commands:


PROC UNIVARIATE DATA=gasdata PLOT FREQ;

VAR mileage;

TITLE 'Summary of the Gas Mileages for the 100 Cars';

RUN;


------------------------
For Histograms
------------------------

To get histograms, we can use PROC CHART again.  For example, for the data in problem 2.30, 
where satscore is the data set and the variables are sat1990 and sat2000 and the score 
differences are called differ, use the following commands to get histograms for sat1990, 
sat2000, and differ:

proc chart data=satscore;
vbar sat1990 sat2000;
vbar differ;
run;

Note that here the midpoints of the class intervals are what is labeled on the axis of the 
histogram.


------------------------------------------------------------------------------------------
Alternate code for problems 2.44, 2.103, 2.104 (instead of the corresponding PROC INSIGHT 
commands):
------------------------------------------------------------------------------------------

/* For problem 2.44:  */

proc univariate data=womenpow;
var age;
run;

proc chart data=womenpow;
vbar age / type=percent;
run;

/* For problem 2.103:  */

proc univariate data=down PLOT FREQ;
var downtime;
run;

/* For problem 2.104:  */

proc univariate data=sanit PLOT FREQ;
var score;
run;