Module 8: Producing Descriptive Statistics

Question 1

What is PROC MEANS?

Answer 1

A procedure that helps get basic statistics for the numeric variables in a data set.

Question 2

List some options for proc means.

Answer 2

Maxdec = n, missing, max min, median, mean, mode, n, nmiss, range, stdev, sum

Question 3

What do the options Maxdec = n, Missing, N, and Nmiss do?

Answer 3

Maxdec: n rounds the values to n decimals
Missing: treats missing values as a seperate group of values
N: number of non-missing values
Nmiss: number of missing values

Question 4

What are optional statements that you can use in proc means?

Answer 4

var: specifiies varaibles to use
by: seperates into levels by by-variable(s)*
class: also performs separate analysis for each levels**
types: used to specify a combination of CLASS variables to produce
output out=: data-set output-statistic list

*needs to be sorted
**note that all type var need to be in class

Question 5

To save out the output to a data set, you need to:

Answer 5

1) Use a noprint option in the proc means statement
2) Specify the descriptive statistics required on the OUTPUT statement

Question 6

What is the type automatic variable?

Answer 6

Value displaying the type of interation. And zero value is the grand total.

Question 7

What is the difference between one-way and two-way frequencies?

Answer 7

One-way freq counts are for 1 variable
Two-way freq counts are for 2 variables

Question 8

Why do we use PROC FREQ?

Answer 8

To create tables showing the distribution of categorical variables in a data set.

Question 9

How do you add options to PROC FREQ? Also list those options.

Answer 9

Option must appear after a slash (/) in the TABLES statement
Options: list, missing, nocum nopercent, and out =

Question 10

What do these options do: list, missing, nocum, nopercent.

Answer 10

List: displays couns in list form
Missing: includes missing values in freq and percentages
Nocum: supresses cumulative freq
Nopercent: suppresses printing of percentages

Question 11

Write the general syntax for multi-way freq/cross tabulations with PROC FREQ.

Answer 11

proc freq data=dataset;
tables var1*var2;
run;

Question 12

What are the specific options for multi-way freqs?

Answer 12

Crosslist: displays crosstabulations in list format with totals
Nocol: supresses column percentages
Norow: supresses row percentages

Question 13

What is the general syntax for user defined formats?

Answer 13

Proc format;
value format-name range = ‘label’

Note: semicolon goes at the end of the LAST range and label…

when formating multiple ranges in one value code.

Question 14

What are the rules for format names?

Answer 14

1) Must begin with a $ if the format applies to character data
2) Must be a valid SAS name (up to 32 characters)
3) Cannot be the name of an existing SAS format
4) Cannot stort or end in a number
5) Only characters allowed are underscores

Question 15

Write a proc format for numeric variables, character variables, and one with value ranges of a numeric variable.

Answer 15

proc format;
1) value survresponse 1 = ‘Yes’;
2) value $racecode ‘W’ = ‘White’;
3) value agegroup 13 - <20 = ‘Teen’;

Note: can use 65 - HIGH in range to indicate extreme values (or LOW)

Question 16

T/F: The keyword OTHER can be used to assign a format to any values not listed in the VALUE statement.

Answer 16

True

Question 17

T/F: You cannot assign non-continous ranges in a format.

Answer 17

False: You can. Separte them with a comma.
Ex. value odd 1,3,5,7 = ‘Odd’;

Question 18

Assign the format survresponse to the variable q3, assuming q3 has numeric values.

Answer 18

data dataset;
set dataset
format q3 survresponse.;
run;

Question 19

Write the syntax to assign new formats to a new varaible.

Answer 19

newvar = put(oldvar, format);

Question 20

How do you permanently store a format?

Answer 20

Specify the libname statement with the libreference.
Ex. proc format lib = libname;

Question 21

How do you display user-defined formats?

Answer 21

Add the keyword FMTLIB in the proc format. Displays all formats in catelog along with descriptions of values.
Ex. proc format lib = formatlib fmtlib;