Testing regime change in SAS I have two samples, A and B. I run the same model for A and B separately to obtain the estimates. I now want to test whether the parameters for models in A are the same with the parameters in B using SAS. Please advise. 
 A: There is not much detail given in your question (what type of model?) and there is probably a way to combine your data into one full model to test for differences between A and B.
Having said that you could construct confidence intervals for the difference between the two parameters and see if 0 is in the confidence interval.  This shouldn't be too hard since SAS will give you the standard error associated with each estimate.
Edit:
If it is a simple linear model you could bring in a dummy variable main effect (1 for group B, 0 for group A).  This would test for same intercepts.  Interacting the dummy variable with the other parameter main effect will test for same slope.
A: Pool your data and create a dummy variable that allows you to identify the two samples. 
First, estimate the model:
$$Y = b_0 + b_1*X$$
Second, estimate the model:
$$Y = b_0 + b_1*X + b_2*D + b_3 * (D*X)$$
Where $D$ is the dummy variable that identifies the two samples
Compute the test statistic
$$ \frac{SSR_1 - SSR_2}{SSR_1} * \frac{n-2*k}{k}  $$
Where $SSR_1$ and $SSR_2$ are the sum of squared residuals of models 1 and 2, $n$ is the number of observations and $k$ is the number of parameters estimated in model 1 (here: $k = 2$). This test statistic follows an F distribution with degrees of freedom $(k, n-2*k)$.
This procedure can be generalized to the more than one explanatory variable. 
In SAS you can easily implement this approach using proc reg. All the information you need to compute the test statistic can be found in the output of proc reg. Note that you could also use the test statement:
proc reg data = pooled;
model Y = X D XD;
test D = 0, XD = 0;
run;

In econometrics this procedure is called Chow Test. You can e.g. refer to Wooldridge's Introductory Econometrics, or any other textbook. 
