In this work, using regular integral transformations on time scales, we
generalize the concept of statistical convergence. This enables us not only
to unify discrete and continuous cases known in the literature but also to
derive new convergence methods with choices of appropriate transformations
and time scales. This is a continuation of our earlier work and includes
many new methods. We obtain sufficient conditions for regularity of kernel
functions on time scales and also we prove a characterization theorem for
the generalized statistical convergence. At the end of the paper we display
some applications and special cases of our results.