Abstract
Empirical data showed the Class Within A Class (CWC) model for collaborative co-teaching of language interventions to be more effective than the traditional pull-out model for elementary school-age children. Interventions in the collaborative setting were co-planned and co-presented by the general education teacher and speech language pathologist. Parallel to these lessons the same strategic interventions were delivered to students who were served in traditional speech-langague therapy. Students with language impairments (N=11) were matched to students with language impairments served in the pull-out model (N=11) and each group was matched to students with normal language skills (total N=44). Effectiveness was measured by pre-post tests using language samples told from stories with social integration and self-concepts assessed for the same six month interventions period. Clause density (sentence complexity) was the single feature that distinguished students with language impairments and students with normal language on pre-test data. Clause density showed significant gains for students with language impairments served in the collaborative setting and not for students served in the pull-out model. The study provides a model to evaluate effectiveness with curriculum-based assessment, social integration and self-concepts of students (age 7-12) in the elementary school-age population.
Introduction
In 1975, the passage of PL 94-142 brought educational services to all students with disabilities. One of the six principles addressed within this law was the least restrictive environment (Turnbull, 1997), which continues to challenge both general and special educators due to lack of substantial research on which service delivery models are most effective (Marston, 1996). Speech-language pathologists are among those special educators who must determine the most effective service delivery for students with language impairments. While a number of articles about language in inclusive settings have advocated for changes in service delivery, many are supported only by opinion and theory without research data to support either pull-out or inclusionary models (Ferguson, 1992; Montgomery, 1992; Prelock, Miller, & Reed, 1995; Roller, Rodriquez, Warner, & Lindahl, 1992). Empirical data are needed to show the effectiveness of individual service delivery models.
Problem Statement
Early programs to implement PL 94-142 interpreted "least restrictive
environment" as individual or small group instruction in a pull-out model. Will
(1986) argued that although these programs were planned to support instruction,
they were inferior because this model denied access to grade-level curriculum.
Furthermore, within the pull-out setting students are socially isolated and
often teased by peers, which results in lowered self-concepts of the very
students whom these programs were intended to help.
The reauthorization of
Individuals with Disabilities Education Act (IDEA) (1997) mandates movement
toward a unified system that blends general and special education services for
students. Speech and language services are included in this shift in service
delivery (Montgomery, 1992; Westby, Weston, & Murphy, 1994). The purpose of
this study was to evaluate the effectiveness of language services provided in
the Class Within A Class (CWC) model (Hudson, 1989) compared to the pull-out
model (Simon, 1991). The CWC model advocates that special education and general
education teachers work together as a team. Emphasis is on equalizing learning
opportunities through effective teaching strategies and not curriculum
adaptations (Hudson, 1989). The focus is on making the curriculum accessible to
all students, not on providing separate lessons for students identified as
needing additional help as is found in pull-out models. Other purposes included
determining the effect of the two service delivery models on students' social
integration and self-concept when compared by groups.
Design
A quasi-experimental research design was used because students with
language impairments cannot be randomly chosen. Subjects were students from
second through sixth grade. Eight grade level matches were divided into four
groups for a total of 32 subjects. All subjects were considered to be within a
normal range of ability. Matches were made by age (within six months) and
grade, enrollment in the same school district so the grade-level curriculum was
consistent and coming from homes where English was the only language. Students
with language impairments were matched to students whose language was
considered to be normal then matched by location where services were delivered.
Each areas of analysis was measured with pre- and post-tests to allow
comparisons among groups when parallel interventions were delivered to subjects
within the matched groups. Three hypotheses were developed as a means of
studying the components of the educational programs criticized by Will (1986).
Thus, language skills were assessed along with social status rankings and
self-concept. Separate methods of analysis were conducted for each.
Hypothesis One: There will be a significant difference in outcome data of language skills of: (a) students with language impairments served in collaborative/inclusive settings, (b) students with normal language skills who were exposed to language interventions in the collaborative/inclusive settings, (c) students with language impairments served in the traditional pull-out model of service delivery, and (d) students with normal language skills with no interventions.
Hypothesis Two: There will be a significant difference in outcome data for social status rankings of: (a) students with language impairments served in collaborative/inclusive settings, (b) students with normal language skills who were exposed to language interventions in the collaborative/inclusive settings, (c) students with language impairments served in the traditional pull-out model of service delivery, and (d) students with normal language skills with no intervention.
Hypothesis Three: There will be a significant difference in outcome data for the self-concepts of: (a) students with language impairments served in collaborative/inclusive settings, (b) students with normal language skills who were exposed to language interventions in the collaborative/inclusive settings, (c) students with language impairments served in the traditional pull-out model of service delivery, and (d) students with normal language skills with no interventions.
Methodology
Language skills, social integration of students, and self-concept were each analyzed. Language samples were taken from stories retold from wordless books that were transcribed and analyzed with Systematic Analysis of Language Transcripts (Miller & Chapman, 1993). These samples were analyzed as measures of language growth for three levels: discourse, semantic, and syntactic. Social integration was measured by the Revised Class Play (Masten, Morison, & Pelligrine, 1985) which was administered to classes of subjects to provide social status rankings by peers. Positive and negative roles assigned to subjects by classmates were analyzed for the level of social integration revealed. Finally, to derive a score for students' self-concept, the Piers-Harris Children's Self-Concept Scale (Piers, 1984) was given to each subject.
Results
Hypothesis One was supported by data obtained from language samples. The two most significant findings involved discourse and syntactic level. The discourse level variable ranked overall growth of stories told. Results were significant for students with language impairments served in general education settings, as were both groups of students with normal language skills. However, students served in the pull-out setting did not make significant gains. Syntactic analysis of language samples was based on clause density, which shows complexity at the sentence or utterance level. Only students with language impairments served in collaborative/inclusive settings made significant gains on this measure. These two findings support the statement by Will (1986) that interventions in isolated settings are not as effective as those delivered in inclusive classrooms.
Hypothesis Two was not supported by the data. Only one significant difference was found between pre- and post-test data for social status rankings. That difference was for gains in negative roles assigned to students served in collaborative settings. This finding contradicts Will's (1986) criticism that students are stigmatized when taken out of the classroom for special services. In this study it was the group of students served in the collaborative setting who were assigned more negative roles by their peers.
Hypothesis Three was not supported by the data from this study either. That is, no significant differences were found or self-concept scores for any group studied, suggesting that student's self-concept did not appear to be affected by type of service delivery model used.
Conclusion
Equalizing learning opportunities is the focus of the CWC (Hudson, 1989)
service delivery model. The results of this study showed that teaching
complexity at the utterance or syntactic level was most effectively taught by
using the CWC (Hudson, 1989) service delivery model. Clause density was found
to increase for students with language impairments who were served in
collaborative/inclusive settings and not for any of the other three groups. As
a measure of complexity of syntax, clause density scores showed that
information was integrated and relationships were shown through more complex
utterances. The significance of gains in clause density supports Scott's (1988)
claim that curriculum-based interventions are one way to teach complex
sentences.
Another significant finding related to narrative rankings for
the discourse level. Students with language impairments served in collaborative
settings, and students with normal language skills exposed to language
interventions in the collaborative setting as well as those who received no
interventions made statistically significant gains. This was not the case for
students with language impairments served in pull-out service delivery models.
Both these findings, therefore, support Will's claim and advocacy for offering
interventions in collaborative settings.
Will's (1986) criticism that
services delivered in isolated settings result in less social integration and
lowered self-concept of students was not supported by this research. Thus,
social integration as measured by social status rankings and self-concept of
students served in isolated settings were not found to be significantly
different for either of these areas of analysis.
Research Implications and Summary
Further research to document the benefits of collaborative co-teaching
models is needed. Specifically, the effectiveness of the CWC (Hudson, 1989)
model on equalizing learning opportunities should be documented with additional
empirical data. Also, the syntactic measure of clause density should be studied
in terms of the level of integration of content from curriculum.
In the
present study, the CWC (Hudson, 1989) model of collaborative co-teaching
practice was more effective for language interventions delivered to students
with language impairments at the elementary school level. The inclusive
practice using the CWC (Hudson, 1989) model was found to be more effective for
students when served for language interventions underscoring the general
education classroom as the least restrictive environment.
References
Ferguson, M. (1992). Implementing collaborative consultation: An introduction. Language, Speech, and Hearing Services in Schools, 23 (4), 361-361.
Hudson, F. (1989). Class within a class staff development workshop. Module III and module IV. Jefferson City: Missouri Department of Elementary and Secondary Education.
IDEA (1997). Individuals with disabilities education act amendments of 1997. (IDEA). Horsham, PA: LRP Publications.
Marston, D. (1996). A comparison of inclusion only, pull-out only and combined service models for students with mild disabilities. The Journal of Special Education, 30 (2), 121-132.
Masten, A. S., Morison, P., & Pelligrine, D. S. (1985). A revised class play method of peer assessment. Developmental Psychology, 21 (3), 523-533.
Miller, J. F. & Chapman, R. S. (1993). SALT: Systematic analysis of language transcripts. Madison, WI: Language Analysis Laboratory, Waisman Center.
Montgomery, J. (1992), Clinical Forum: Implementing collaborative consultation, Perspectives from the field: Language, speech and hearing services in schools. Language, Speech, and Hearing Services in Schools, 23 (4), 363-364.
Piers, E. V. (1984). Piers-Harris children's self-concept scale, revised manual. Los Angeles, CA: Western Psychological Services
Prelock, P. A., Miller, B. L., & Reed, N. L. (1995). Clinical Exchange: collaborative partnerships in a language in the classroom program. Language, Speech, and Hearing Services in Schools, 26 (3), 286-292.
Roller, E., Rodriguez, T., Warner, J., & Lindahl, P. (1992). Integrations of self-contained children with severe speech-language needs into the regular education classroom. Language, Speech, and Hearing Services in Schools, 23, 365-366.
Scott, C. M. (1988b). Producing complex sentences. Topics in Language Disorders, 8 (2), 44-62.
Simon, C. S. (1991). Communication skills and classroom success assessment and therapy methodologies for language and learning disabled students. San Diego, CA: College Hill Press.
Turnbull, R. (1997). Significant recent cases interpreting IDEA before 1997 amendments and likely impact of amendments on case law. Working paper. Manuscript in process. Kansas City, KS: University of Kansas, Department of Special Education.
Westby, C. E., Weston, S., & Murphy, M. (1994). The vision of full inclusion: don't exclude kids by including them. Journal of Childhood Communication Disorders, 16 (1), 13-22.
Will, M. (1986). A shared responsibility. Exceptional Children, 52 (5), 411-415.
 |
 |