求高手指点

【求助】本人是sas菜鸟，利用以下程序算出了循环回归的斜率的平均值meanslope，现想检验一下meanslope是否显著>0，请问高人：如何给出检验meanslope>0的t统计量值（5%显著性水平）。
data tes;
input t1 t2  t3;
datalines;
-0.0367        -0.1254        -0.0078
-0.0292        -0.1817        0.0431
0.0141        0.0108        0.0143
0.0139        0.0302        0.0399
-0.0106        -0.0946        0.1088;
run;
%macro fun;
data est;  *create an empty dataset to load values;
length est 8.;
delete;
run;
%do i=1 %to 2;   *2=3-1;
proc reg data=tes outest=out(keep=t&i rename=(t&i=est));  *output each estimate dataset,leave one interested;      
   model t%eval(&i+1)=t&i;
quit;
proc append base=est data=out force;*put all output datasets together to a bigger one 'est';
run;
%end;
proc sql;
create table result as
   select avg(est) as MeanSlope from est;*get the avg slope;
quit;
proc print; run;*output to listing window;
quit;
%mend fun;
%fun

哥们T检验有个前提哦
你这个属于单侧的T检验
前提是：正态 方差齐
要进行正态性检验和方差齐性检验
但是你只有一个数据
呵呵

I guess, it is to test the mean of slopes; and very few t’s are just for illustration here. So I would suggest

PROC TTEST DATA = EST H0 = 0 SIDES = U;
   VAR EST;
RUN;

thanks a lot

if you use ttest from SAS9.2
you need to specify alpha=0.1

i.e.

PROC TTEST DATA = EST H0 = 0 SIDES = U alpha=0.1;
   VAR EST;
RUN;

if ttest assumptions are not satisfied, you're suggested to use
PROC NPAR1WAY

我用的是sas9.1, 程序执行后的结果是：
" PROC TTEST DATA=est H0 = 0 SIDES= U ALPHA=0.05;
                                ------
                                22   200
WARNING: 忽略第二数据集引用。
ERROR 22-322: 语法错误，期望下列之一: ;, ALPHA, CI, COCHRAN, DATA, H0.
ERROR 200-322: 该符号不可识别，将被忽略。"

不知如何修改？