Since your variables under consideration are all categorical, in terms of regression methods the logistic regression model will suit your data. One thing missing in your question is the nature of your dependent variable. If the dependent variable is dichotomous (two (2) outcomes), then the Binary Logistic Regression model will best fit your model but if the outcome of the dependent variable is more than two (2), then the Multinomial Logistic Regression model will fit your data.

For resources on Binary Logistic Regression model read Sandro Sperandei's material entitled "Understanding Logistic Regression Analysis" and for the second model see attached file.

I think what you want is to look up works on the analysis of Likert scales in psychology. It may also help to look at the difference between Likert and Likert-type data.

Using Likert as a key word is specific enough that many web resources can be found. There are also a wide range of text books available on the analysis of survey data or Likert-type data. The books in psychology often discuss converting survey data into composite variables. Things like questions 2,7,15, and 32 are combined in "this" way to get a quantitative measure of how much a person loves eating mango versus mangosteen or durian.

I am not sure regression is the right tool here. You might want to consult a local statistician about possibly using factor analysis, or similar methods.

@ Joseph Otoo thank you the dependent variable has 4 categories i got your answer thank you i am going to read the article u sent.