This occurs only when I'm trying to carry out an ANCOVA. Please help me!

Thanks in advance.

My first thought is to wonder whether you are using a continuous variable as a factor with many values and have a large dataset. In that case, you should treat the variable as a covariate. Otherwise, you may be able to resolve this by temporarily increasing the workspace size. See SET WORKSPACE in the Command Syntax Reference for details.

Hi, Jon. First of all, thank you for taking the time to answer my request. Let me tell you the details:

My dependent variable is Math scores. I am using 3 continuous variables as factors (Language scores, Classroom climate, Attitude towards Math) and only one variable (socioeconomic status) as covariate.

On the other hand, could you show me the steps for setting workspace, please. I just reviewed Command Syntax but I don't know how to do it. Thanks in advance.

You can enlarge the workspace by running the

SET WORKSPACE command (run SHOW WORKSPACE first to see how to set it back), but this is the wrong thing to do.

Don't use continuous variable as factors. If you have, say, 100 distinct values per variable, that's a million points to evaluate. The results, if you get them, would be worthless.

Use the continuous variables as covariates. Socioeconomic status, if it is a few discrete values, could be a factor. So I think you have these backwards.

Hi again, Jon. Thanks for your comments. Let me explain you:

I was given the assignment in my job to carry out a linear regression in SPSS, taking socioeconomic status as a control variable. You already know the independent variables (mentioned above). ¿Is it right to use all the independente variables as covariates? Please note that I just want to control socioeconomic status.

Kind regards,

Victor Eduardo

What is the nature of the SES variable? Is it categorical? If so you would use it as a factor. The other variables would be covariates.

No, SES is not a categorical variable. It's a continuous one. ¿What do you suggest me to do?

If all the variables are continuous - scale measurement level - then do ordinary regression - all the independent variables are covariates. You can do that with the REGRESSION procedure or UNIANOVA that you are using now, I presume.

It is entirely clear to me now. Thank you very much, Jon. I wish you and yours a Happy Holidays! We'll be in touch...