I second David. Although with a small correction: descriptive statistics is NOT a version of regression. Also note that "triangulation" has a particular meanings in several areas, but not in data analysis/statistics, see https://en.wikipedia.org/wiki/Triangulation_(disambiguation).
Ok, let's say that some statistics obtained from regression models are sometimes used to describe a sample... (the sample mean being the most frequently used). If you also include quantile regression, all sample quantiles and derived quantities may also be called statistics from regression models.
No Sir. The issue of mixed methods has to do with your research design. I mean whether it is purely quantitative or qualitative or both. It is when both quantitative and qualitative design is adopted in single research study, you can now talk about mixed methods.