U is not a parameter in your model. S2U is. U is the realization of the random effect from the distribution of U you have defined with your RANDOM statement. No starting values are needed for U. You do need a starting value for its variance, though ... S2U.  ... View more

Sounds like a classic repeated measures. Seeing the data would help a bit more, but from what you describe a REPEATED statement would be the way to go. Use the TYPE= to specify the covariance structure you wish for the 4 repeated measures on each subject. If each subject received one and only one treatment, then you would not be able to estimate a SUBJECT*TREATMENT interaction. If you model TIME as a CLASS effect and have 1 observation per time point per subject, then you will not be able to model a SUBJECT*TIME interaction. It is possible to model TIME as a continuous effect and set up a random coefficients model for this data with random int time / subject=subject However, you get much more flexibility in the covariances with the REPEATED approach. ... View more

You have one observation for each rep and location combination, so the rep(location) effect is confounded with the residual variance. You would need multiple observations for each rep and location combination to estimate a rep(location) effect. Remove the RANDOM statement and you get the expected results for this data and model - including the LINES output ... View more

If you have a repeated statement like repeated time / type=cs; then you are not estimating anything in the r-matrix except a diagonal element. MIXED might partition the variance into two components (with TYPE=CS), but the components will sum to the residual variance you obtain when you drop the repeated statement completely. So, this repeated statement is the same as having no repeated statement at all. Specifically, this repeated statement is not correlating any observations in your data.  ... View more

GROUP=TREATMENT will estimate the TYPE=CS structure separately for each level of treatment. If you suspect you have heterogeneity in the variances across levels of treatment, then that is the way to handle it. You do need sufficient data to estimate the variances for each treatment. If you are having trouble using GROUP= then that can be a sign that you do not have sufficient data to estimate the variance separately.  ... View more

With a repeated statement like repeated time / type=cs subject=subject; you are modeling the R matrix as a block diagonal structure with a common covariance on all the observations from the same level of SUBJECT.  With your data description, if SUBJECT takes on a unique value for for each unique subject in your data (ie, subjects 1,2,3 in block a and subjects 4,5,6 in block b), then this syntax will be fine If you have subjects nested in blocks (ie, subjects 1,2,3 in block a and subjects 1,2,3 in block b) and subject 1 in block a is a different subject than subject 1 in block b, then you need subject=subject(block) on your repeated statement, A statement like repeated time / type=cs subject=subject*time is likely incorrect. The subject= effect on the repeated statement is the experimental unit on which repeated measures were taken. If time is your repeated measure, then it is unlikely that you took repeated measures over time on the experimental unit subject*time.  ... View more

PROC ADAPTIVEREG will get you started. This simple example should help ODS Graphics and PROC ADAPTIVEREG  ... View more

That RANDOM statement is almost right. If you have block 1-4 in site A and block 1-4 in site B then you do need to nest block within site. Your RANDOM statement could be    random block(site);  or random int / subject=block(site); ... View more

As @PaigeMiller posted (and citing Rick's excellent blog post), the normality condition is on the residuals, not the dependent variable. You can run MIXED and check the plot of your residuals to see if they appear normal "enough" to you.  ... View more

PROC ADAPTIVEREG can be helpful in situations where you do not fully understand the nature of the relationship between the predictor(s) and the response. ... View more

On top of @KevinScott 's and  @SteveDenham 's excellent suggestions, try simplifying the model a bit. You have 132 observations (13 with missing values) and have year, site(year) and block(site year) as random effects. Try removing block(site year) as a random effect and see if that helps. You may not have enough data to get a variance component down to the block level. ... View more

Adding the R option to the REPEATED statement will show you the block of R for the first subject. You can specify multiple subject numbers using R=1,2,4,8 to print the block of R for the 1st, 2nd, 4th, and 8th subject. I have debugged lots of issues with mixed models over the years by looking at R (and G and V on the RANDOM statement) to see what kind of crazy covariance structure I have (mis)applied to my data.  ... View more