Calculate mean score for every respondent, and than use scores to analyze two groups (male vs female). Analyse method and/or test selection depends on your research question.

Do you want to combine all items into a total score as Kenan notes, do you want to compare men/women on each item separately, or do you have subscales? If you want to make comparisons on means, a simple independent group t-test could be used, whether it is used on a single total score, or on each item separately (20 t-tests).

You are comparing the means of two sets of ordinal data. The test criterion must be nonparametric. Then you can use Wlcoxon signed rank test, Mann Whitney U test, or Kruskal Wallis test.

Normally male/female comparing we calculate in demographic information and secondly, you calculate their means/SD. As a result, you are able to found those test which are required according to your variable and research.

Kenan and Paul have provided vital and useful lead as to what to do on this issue. Be specific, because there are various ways to arrive at what you are looking for, but being specific makes it easy and less clumsy.