I do not have hypotheses for my research and I have got only categorical data. what kind of data analysis can i conduct? also I do not have hypotheses for my research and I have got only categorical data. what kind of data analysis can i conduct for correlation between variables? also is chi-square used  only for hypotheses testing? can it be used for exploratory analysis? and how to test if there is moderation relationship between variables?
 A: You can use the chi-square as a basis for exploratory data analysis, if you don't focus on the p-value. Assuming we're talking about a two-way table (rather than say a goodness of fit test)
You can, for example:


*

*transform it to a measure of association (such as Cramer's V)

*decompose it into single df comparisons.

*calculate contributions to chi-square, or their signed square root (/Pearson residuals) as a way of detecting where deviations from independence occur
Any of these could be used as exploratory tools.
The question about moderation depends on precisely what you mean in this context. Can you give an example of the kind of thing you're trying to find?
A: First, it is always appropriate to have objective even if they cant be coined in to testable hypothesis, but reading through your question you do seem to have one!. You can use a Chi-square (or Fisher's exact test) test to test the null hypothesis the two categorical variables of interest are independent (not associated).
Exploratory analysis can be viewed broadly, but if the particular case includes inferential statistics then using a chi-square test can be part of it.
You can use a Mantel-Haenszel Chi-square to test for a moderation relationship if you have a third variable where you will be assessing your primary association of interest within levels of this third variable.
A: Research that is devoid of hypotheses or goals probably should not be undertaken.  The likelihood of a meaningful result that anyone cares about is very low.
