"Whole language" is typically endorsed by faulty research
termed "qualitative research" by its proponents
The results are "descriptive" – for example: "teachers
use phonics in whole language" is a "research finding"
There is no separation of groups that receive the experimental
treatment from "control" groups not receiving treatment that can be used
to compare outcomes
There are no definitive conclusions because there is nothing
to compare with
There does not appear to be any random assignment of subjects
(teachers/ schools/ students/ or programs) in such studies
Recent teacher training programs employ these "results"
to document effectiveness of whole language programs and to disparage phonics
School administrators seem unaware of the unscientific
nature of the studies
Many reading curricula are chosen by educators using outcomes
of this kind of "research"
Current teachers are frequently not trained to teach reading
Typically more balanced view of programs
The studies have long duration and are carefully designed
so that only one component of the reading program is changed at a single
There are always "control" groups that do not receive
the expermental treatment for the duration of the study, so that results
with and without treatment can be compared
Groups are carefully matched and the subjects of the study
(teachers/ schools/ students/ or programs) are randomly assigned
Study outcomes can be replicated if conditions are duplicated.