Is that a theoretical question or do you have actual data?

The statistical analysis plan always depends on different factors, e.g. on how the variables were measured and how large your sample is.

Assuming that age, gender and civil status were measured as independent categorial variables and knowledge, skills and attitute as dependent continous variables, a multivariate analysis of variance (MANOVA) could be run - but statistical prerequisites must be checked.

If you want to split the sample into binary groups (men vs women, old vs young, rich vs. poor) you can use MANCOVA. If you want to keep it as continuous variables (precise age, precise income) you should use regression.

Depends mainly on the nature of the community and ten research and statistical implied treatment and division of variables (continued, independent). Multiple analysis of variance can be used( MANOVA) after verification