Model Programs
As Frede (1998) makes clear, determining the effects of curricula
or teaching methods on young children is a complex and
difficult task. A number of problems result from the difficulties of
measuring learning and development in young children. Standardized
tests of cognitive ability in early childhood are of questionable
validity (Kamii, 1990). Measures of social development
are problematic, since they often fail to discriminate adequately
among children (Datta, 1983). Different curricular approaches
have different goals; thus different outcomes should be expected,
and comparing the programs on the same outcome measures can
bias findings in favor of one approach or another. The same type
of bias can occur in trying to measure treatment implementation:
the appropriate observation techniques for one approach may fail
to discern important practices or failures in implementation of
another approach.