Module 8 Producing Descriptive Statistics

Question 1

What is proc means?

Answer 1

Gets us basic statistics for numeric values

Question 2

Proc means options

Answer 2

maxdec, max, min, mean, mode, N, NMISS, range, stdev, sum, etc

Question 3

optional VARS statement

Answer 3

specified which numeric variables to use in the analysis

Question 4

optional BY statement

Answer 4

performs a separate analysis for each level of variables in the list

Question 5

what must we do before using BY?

Answer 5

SORT THE DATA!

Question 6

optional CLASS statement

Answer 6

separate analysis for each level but no sorting needed

Question 7

optional TYPES statement

Answer 7

specifies a combination of CLASS variables

Question 8

optional TABLES statement

Answer 8

calculates frequencies, crosstabulation with *, all vars used in tables must be in the class statement

Question 9

What does () do in type statement?

Answer 9

gives the descriptive statistic required across all observations in the data set

Question 10

OUTPUT OUT syntax

Answer 10

OUTPUT OUT = data-set statistic(variable-list) = name-list

Question 11

value of TYPE

Answer 11

depends on the level of interaction.
The observation where TYPE has a value of zero is the grand total.

Question 12

Practice: Frequencies for the variables rank, grade, race, and gender

Answer 12

proc freq data = one;
tables rank grade race gender/ list missing;
run;

Question 13

Frequencies on all CHARACTER variables in the data set;

Answer 13

proc freq data=one;
tables character/ list missing;
run;

Question 14

one way PROC FREQ options

Answer 14

Must include / in the TABLES statement
LIST: display counts in list format;
MISSING: includes missing values in frequencies and percentages;
NOCUM: suppresses cumulative frequencies;
NOPERCENT: suppresses printing of percentages;
OUT = dataset: writes out a data set containing frequencies

Question 15

two-way PROC FREQ options

Answer 15

CROSSLIST: displays crosstabulations in list forat with totals
NOCOL: suppresses column perentages;
NOROW: suppresses row percentages

Question 16

Cross tabulation

Answer 16

use an asterisk between variables;
proc freq data = one;
tables genderraceanothervar; /* can be more than two vars*/
run;

Question 17

PROC FORMAT syntax

Answer 17

proc format;
value format-name range1 = ‘label1’
range2 = ‘label2’;
run;

Question 18

Rules for format names

Answer 18

** must begin with a dollar sign ($) if the format applies to character data;
** must be a valid SAS name (up to 32 characters, including $ sign if needed);
** cannot be the name of an existing SAS format;
** cannot start or end in a number;
** The only characters allowed are underscores (_);

Question 19

When to use periods when doing proc format?

Answer 19

The format name does not end in a period(.) when specified in a VALUE statement.
However, it must be referenced with period(.) at the end when on the FORMAT statement.

Question 20

proc format example

Answer 20

proc format;
* format for a numeric variable;
value survresponse 1=’Yes’
2=’No’
3=’Did Not Answer’;
* format for a character variable;
value $racecode ‘W’=’White’
‘B’=’Black’
‘H’=’Hispanic’;
* format that uses value ranges of a numeric variable;
value agegroup 13 - <20 =’Teen’
20 - <65 =’Adult’
65 - HIGH=’Senior’; **Note: you can use the keywords HIGH and LOW to refer to extreme values for you range;
run;