Articles Submitted
Abstract:
Using meta-analysis, randomized experiments
in education that either clearly did or clearly
did not experience attrition were examined for
the baseline comparability of groups. Results
from 34 studies suggested that the observed
measures of baseline comparability did not
differ more than would be expected given
sampling error. There was no relation
between either the overall or the differential
attrition rates and baseline comparability, a
finding which held up under sensitivity
analyses. In addition, although post-test effect
sizes were statistically significantly related to
the baseline (pretest) effect sizes, we
demonstrate a simple adjustment to the post-
test effect size that may result in a better
estimator of the effect size in the population.
Our results suggest caution is warranted when
applying blanket rules pertaining to attrition
that are meant to either serve as
inclusion/exclusion criteria or in scoring study
quality. Greater attention is needed to both
the reporting of attrition in primary studies
and to the synthesis of contextual influences
on attrition rates across domains.