This study sought to broaden the scope of teacher training in English by incorporating an experimental design in practice teaching. The control group, 20 ninth graders whose IQ's averaged 95, studied narrative-descriptive writing and transformational grammar, writing the same number of compositions as the two experimental groups. The low experimentals, 23 ninth graders whose IQ's averaged 95, studied narrative-descriptive writing and sentence-combining models presented without the abstractions of grammatical terminology. The high experimentals, 26 ninth graders whose IQ's averaged 120, studied the same curriculum as the low experimentals. All three groups wrote the same number of compositions on the same topics. Precompositions and postcompositions were analyzed for the number of T-units, words per T-unit, and sentence-combining transformations (SCT's) per T-unit, the latter restricted to those SCT's taught. Results showed that the sentence-combining approach is more effective as a means of increasing the syntactic fluency of low IQ students than is the formal study of transformational grammar, and that high IQ students become more mature syntactically regardless of the treatment. (Author/JM)