ABSTRACT
With the conceptual introduction in Chapter 3, we now come to more detailed
discussion on meta-analysis for a specific data type. The first and probably
most commonly seen meta-analysis is of binary or binomial data where the
number of successes are considered in a sequence of independent studies each
of which has success with probability p. We begin this chapter with two real
datasets in Section 4.1 and then describe the meta-analysis models associated
with this type of data in Section 4.2 with step-by-step R implementation using
the first dataset. The second dataset will be used in Section 4.3 to illustrate
the application of R package meta corresponding to the meta-analysis methods
detailed in Section 4.2 with discussion in Section 4.4.