| Interaction styles of special education teachers were investigated using semi-structured observation of 13 student-teacher pairs during one-on-one language arts instruction of elementary students with cognitive disabilities, to study teacher use of verbal and nonverbal directions and responses. Teacher directions included command, question, suggestion, full physical prompt, partial physical prompt and gestural prompt. Likewise, teacher response variables included elaboration, repeat, acknowledgment, compliance, imitation, and nonverbal acknowledgment. Student task-behaviors of engage, intrusive/disruptive, on-task, and off-task provided a context for understanding differences in teacher styles. The results showed that, similar to studies of mother-child interaction in developmental disabilities, special educators' interaction styles were more directive than responsive; they used directions over twice that of responses. Gestural prompts and questions were the most frequently used directions. Upper elementary students with cognitive disabilities exhibited task-conducive behaviors (engage and on-task) over 90% of time during one-on-one instruction. Teacher directions---individual as well as aggregate measures---were correlated with the rate dimension, but not the duration dimension, of student active task engagement, implying short-lived effects of teacher directions. Thus, while teachers may observe a desired change in student performance, the change may not be durable and, hence, may not be as desirable as they had believed. From results of sequential analyses (Yule's Q index and Allison Liker z scores), aggregates of teacher directions were followed by student active engagement, and aggregates of teacher responses by student passive task-orientation (i.e., on-task). This higher quality feedback from student responses, together with the outerdirected characteristic of students with cognitive disabilities, is postulated as a potential mechanism that maintains a high level of teacher directiveness. Furthermore, the pattern of sequential relationships changed as the level of student engagement varied (vs. level of teacher directiveness), suggestive of a child-driven model of teacher-child instructional interactions. Similarities and differences between correlational and sequential analyses as well as implications of the current findings are also discussed. |
