Reading in the CompuStat financial statement database, I have the following variable: However, when I execute the following code, zero observations are identified (and I've verified in the dataset that observations should have been identified): DATA TEST02; Set TEST01; If DataDate > 201231; Run;   However, when I execute the following code based on integers using the 1/1/1960 date convention, I receive the correct number of observations: DATA TEST02; Set TEST01; If DataDate > 22281; Run;   Why is SAS considering this the 1/1/1960 convention when the format is YYMMDD8, and how do I treat this as YYMMDD8 as it is indicated in the attributes? The column attributes are identical for the original dataset and the TEST02 dataset. Thank you in advance for your assistance. ... View more

I have a dataset of approximately 450 observations, 250 from Country #1 and 200 from Country #2.  Here are my variables in my dataset:   Country (either #1 or #2) Variable of Interest #1 Variable of Interest #2 Variable of Interest #3 Variable of Interest #4 Control #1 Control #2 Control #3 Dependent Variable #1 Dependent Variable #2 Dependent Variable #3   For example, in Proc Reg:  Model DV1 = VariableInterest#1 VariableInterest#2 VariableInterest#3 VariableInterest#4 Control#1 Control#2 Control#3;   I am working on coding MANOVA using Country as "Class".  I'm not very familiar with MANOVA and have tried a few codes, but nothing has worked.  Any help on coding MANOVA for this data to determine any variance between DVs #1-3?  Thanks! ... View more