The Task Demonstration Model: A Concurrent Model for Teaching Groups of Students with Severe Disabilities

Article excerpt

Significant progress has been made in the past 10 years in implementing functional, community-based educational programs for people with severe mental retardation. However, there are many questions about the best procedures for teaching these students in their classrooms. The challenge is to identify teaching strategies that are both effective for students and practical for staff.

One of the more frequent debates has been whether these students should be taught individually or in groups (Reid & Favell, 1984). Traditionally, students with severe disabilities have been taught in one-to-one situations. Because of the finite amount of teacher time available, however, this approach has meant that students received relatively small amounts of instruction per day. Indeed, in observations of 30 classrooms, we have found direct instruction to occur less than 10% of the school day.

Group instruction involving a small number of students may significantly increase instructional time for each student. Several investigators have compared one-to-one teaching with group instruction and found no deficit in skill acquisition when students were taught in groups (Bourland, Jablonski, & Lockhart, 1988; Favell, Favell, & McGimsey, 1978; Fink & Sandall, 1980; Storm & Willis, 1978). Moreover, group instruction has been found to be superior in efficient use of teacher time (Favell et al., 1978; Fink & Sandall, 1980), opportunities for social contact (Storm & Willis, 1978), incidental learning (Biberdorf & Pear, 1977), and decreased rates of inappropriate behavior (Ranieri, Ford, Vincent, & Brown, 1984).

Although there are many advantages to group instruction, some concerns remain regarding its use. These include the difficulty of training students with heterogeneous skills (Westling, Ferrell, & Swenson, 1982), the type and difficulty of tasks appropriate for group instruction (Alberto, Jobes, Sizemore, & Duran, 1980; Favell et al., 1978), the decreased opportunities for individual students to respond during group instruction (Bourland et al., 1988), the question of whether the students should be taught the same or different tasks (Oliver 1983), and the selection of an appropriate group instruction model (Storm & Willis, 1978).

With respect to the latter, Reid and Favell (1984) have identified three general models of group instruction. In the sequential model, each student is taught individually in a sequential order while the other students in the group wait for their instructions. This is the most common alternative for persons with severe or moderate retardation, but it is in effect a series of one-to-one teaching situations. In the combination con-current/sequential model, some instruction is provided to all the students concurrently while other instruction is provided to individual students in sequential order. In the third model, the tandem individual-to-group model, the instruction begins with a one-to-one format and is systematically extended to include more students.

For each of these models, there is considerable time during group instruction in which the nontarget student does not have an opportunity to respond. The teacher may be requesting a response from another student, prompting another student's response, or correcting errors. During these periods, the nontarget student may decrease contact with the learning environment and increase problem behavior (Repp & Karsh, 1991; Repp and Karsh, in press). Since the major problem of students with severe disabilities is, by definition, that they take longer to learn tasks, unnecessary periods without instruction would seem counterproductive.

One group instruction model that has not been investigated with these students is a concurrent model in which all the students in the group respond concurrently on the same tasks throughout the teaching session. An advantage of a concurrent model is that each student receives many more opportunities to respond than in the previous models. …