| What
is Scientifically Based Research?
Replication
and Meta-analysis
The
range of contexts to which the findings of any one experimental
or quasi-experimental study can be applied is necessarily
limited by the contexts in which it is conducted. Claims
of effectiveness gain in credibility when they are based
on a large number of studies, each of which repeats the
essential hypothesis tests but varies such factors as
implementation contexts and the particular instruments
used to measure dependent variables. The techniques
of meta-analysis (Cooper, 1998), pooling together and
statistically analyzing the results of many experiments
and quasi-experiments, can often reveal trends not evident
in single studies.
|
