You can use biserial correlation coefficient

Don't use regression to give you a correlation, and don't use a biserial correlation coefficient -- you don't have a dichotomous variable here. You could use Spearman's rho or Kendall's tau. Kendall's is better if you have a small sample or outliers, so make sure you've thought about these aspects and examined your variables to inform your choice.

Bear in mind that in either case you're losing information by treating your continuous variable as though it's ordinal, but that can't be helped unless you want to make a slightly controversial choice and treat the ordinal variable as though it's continuous and use Pearson's r (some people would say you have a sufficient number of scale points to do this, but you'd need to check assumptions e.g. that your variables are roughly normally distributed and roughly linearly related to each other).

Dear Julia Aggio ,

If you have several restorative justice measures you mIshtiaq consider first carrying out a reliability analysis to combine them into a single scale variable. This would then allow you to undertake the kind of analysis that you want to do. The best technique for scale reliability is principal component analysis. Please see my study guide. You should also check out relationships before undertaking statistical analysis, e.g. by creating scatter plots in order to check whether parametric assumptions are met.

I think that you should Chi Square test or Logistics Regression test.

They can indicate the association between categorical data and quantitative data.

They are also non-parametric tests.