How can I get P-value

[HG]

Hi All,

I was asked to get P-values. Below is my sample data. My real data have more than 1000 observations and other variables. What procedures should I use? my variables are continuous variables.   All suggestions are much appreciated!

data have;

input var1 age;

cards;

50 20

55 21

70 25

80 18

90 19

40 50

30 60

70 25

20 80

15 90

20 70

100 18

110 21

130 20

5 90

;

proc means data= have noprint;

var var1;

output out=q p25= p50= p75=/autoname;

run;

data temp;

set have;

if var1<20 then group=1;

   else if var1<55 then group=2;

     else if var1<90 then group=3;

         else group=4;

  run;

  proc print;run;

proc sql;

  select group,count(*) as n, mean(var1) as m_var1,mean(age) as m_age

    from temp

        group by group;

quit;

/* output  */

                           group         n m_var1     m_age

                         ƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒ

                             1         2        10        90

                             2         5        32        56

                             3         4     68.75     22.25

                             4         4     107.5      19.5

        var1 by group

	1	2	3	4	P- value
N	2	5	4	4
Var1(mean)	10	32	68	107	???
Age (mean)	90	56	22	19	???
Other					???

[Doc_Duke]

HG,

In order for us to help you with the appropriate procedure for your data, we really need to better understand the question.

For instance,

Do you want to know if

is var1 is correlated with age?  There are several statistical tests for that, some with different underlying assumptions.

is age different by group membership?  also several statistical tests.

Doc Muhlbaier

Duke