Even though you haven't described the context of your study, there are many statistical methods that can be used. The particular method chosen depends on the data. No method is chosen to wrongfully obtain statistical significant results.
Yes, Roman I can see the change. Statistical validity of a method depends whether the data have satisfied the assumptions underlying the method. For example if the data are of the interval or ratio scale and are normally distributed and independent then parametric methods are validly used. Otherwise nonparametric ones are validly used. Another determinant is the sample size. Using a large sample size tends to make the method robust; that is, it makes it less vulnerable to violations of the assumptions and therefore more valid.