Survey Analysis

[ballardw]

First question because it relates to appropriate procedures: is your survey from a complex sample meaning other than a simple random sample?

The first step would be similar but if you have complex sample then you need to look at Proc Surveyfreq to describe the sample used, otherwise Proc Freq is the general tool for Chi-square analysis.

You can do a lot of analysis by using the  grouping features in Proc freq.

You can use a tables statement like:

tables (sex education marital <other demographics variables> ) * (Q28- Q59 Q72-Q79) /chisq;

to get summaries and chisq tests for all of the variables in the first set of parentheses crossed with the listed variables in the second set.

I suggest directing the output to an ODS document for later reading as there will be A LOT of output (on the order of 400 comparisons).

Legibility will be increased by making sure the formats are applied to your variables.

I have no idea what your "create appropriate scale" would mean.

[djmangen]

Am I correct that the part of your question that pertains to "making scales" refers to combing the items in, e.g., Job Satisfaction, into one overall score that combines the several items listed in that section?

Also, is your statement regarding "reliability testing" related to those same subsets of items.

Assuming that I am correct, why are you even mentioning the other background measures such as age, etc.?  What role, if any, do they play in the questions you are posing.