122
rather than student-teacher interaction. Their classes are large and
they do not have an opportunity to interact with students on a one-to-
one basis. Although Spanish is the primary language of instruction in
the bilingual classes, students respond in written English exercises.
English is the language of evaluation. It is possible that had the
bilingual teachers rated the students in English, the correlation may
have had a stronger relationship to total achievement. The reason for
the lack of correlation between the bilingual teachers' ratings and total
language score is not known. This lack of significant correlation is
not a reflection of the teachers' ability to teach or dedication to
their students, to be sure.
Composition of Language Scores
A written Spanish language score ivas achieved for each student by
collapsing the total Spanish spring LAB percentile score with the Inter-
American percentile reading score, the LAS sentence comprehension, and
written story retelling subsections percentile scores. The same pro
cedure was performed to obtain the written English language score,
with the exception of the Inter-American reading score which was not
available in English. A total language score was obtained in each
language by combining the total written language score with a weighted
oral language score. Analysis of the subtests making up the total score
reveals that some components are more significantly correlated with total
language ability than are others.