Is anyone aware of a way to exclude cases for a large number of variables? I've got about 200 variables with interval level data ranging from 0-5. I want to exclude cases with a zero in these variables. The 1-5 represents an attitude on a Likert scale; the zero represents the respondents inability to adequately answer the question. It is also important that that cases that are excluded, only by exclude for analyses of that variable. For example, if respondent A answers 5 for variable X and 0 for variable Y, respondent A's response to variable X should still be used in OLS.

I'm open to suggestions.

Thanks.

Keith Leek ---------------------------------------------- Keith M. Leek Marketing Manager Notre Dame Food Services 210 South Dining Hall Notre Dame, IN 46556-1043

VOICE: 219-631-5589 FAX: 219-631-7994 ---------------------------------------------- "What is perceived as real, is real in its consequences." -Thomas Hobbes