Academic outcomes of students with learning disabilities.

Abstract

This study reports results from a multi-district demonstration project: Systemic Restructuring to Include Children with Disabilities jointly conducted by the Louis de la Parte Florida Mental Health Institute and the Department of Special Education at the University of South Florida. Academic outcomes of 89 students with learning disabilities included in general education were measured. Inclusion was measured by assessing the number of hours per week a student was present in the general education classroom. Results showed that, on average, students with learning disabilities who spent more time in inclusive settings had higher achievement scores.

**********

The national drive towards integrating more students with disabilities into general education classrooms and merging the governance of special and general education stems from the Regular Education Initiative (REI) of the 1980s (Weintraub, 1991). The REI began a wave of reform that shifted the emphasis away from mainstreaming towards a more systematic approach of educating students with disabilities within the general education classroom. The term "inclusion" was popularized during the early 1990s and represents a philosophical departure from mainstreaming. Inclusion represents a commitment to educate each child with a disability to the maximum extent appropriate in the school and classroom he/she would otherwise attend, and requires only that the child will benefit from the class, not necessarily be able to do the work.

The individual nature of inclusion efforts and the diversity among practices and prospective audiences makes the collection and synthesis of information a challenging feat. Recent attempts to examine inclusion have failed to arrive at a consensus on the best placement for delivering instruction to students with disabilities (Halvorsen & Sailor, 1990; Lipsky & Gartner, 1995; Zigmond, Jenkins, Fuchs, Deno, Fuchs, Baker, Jenkins, & Couthino, 1995). Moreover, the results of meta-analyses conducted on the effectiveness of mainstreaming in meeting the needs of exceptional students have been inconclusive (Baker, Wang, & Walberg, 1994; Brady, Swank, Taylor, & Freiberg, 1988; Carlberg & Kavale, 1980). For this reason, some researchers conclude that serving children with learning disabilities in the general classroom is only marginally effective, while others suggest that the potential of general education to provide an atmosphere that promotes learning without stigmatizing children has yet to be fully explored.

Method

Participants

Students for this study were selected from three schools in the Systemic Restructuring to Include Children with Disabilities project using the following criteria: (a) identification by the school system as having a specific learning disability; and (b) participation in the project for at least two years. Using these criteria, 89 students from the 3 schools, 2 elementary and 1 middle, qualified for the study. The demographics of each school varied, as did the number of special education students qualifying for the study at each site. Demographic information on participants in the project is shown in Table 1, see issue's website <http://rapidintellect.com/AEQweb/sum02.htm>

Typically, in studies examining the effect of general education on students with disabilities, inclusion has been defined as a categorical variable in terms of placement. In this study, inclusion was a continuous variable reflecting actual hours spent in general education classrooms. In turn, this reflects exposure to general education curriculum and instruction. Using presence, rather than placement, as an indicator of inclusion provided a more sensitive and powerful analysis of the effects of inclusion on student outcomes. The models of inclusion at the three schools differed in the amount of support provided in the inclusive setting, ranging from mainstreaming and collaborative consultation models to team teaching.

Procedures

Previous research was used to identify student characteristics commonly associated with academic achievement (Easton & Engelhard, 1982). Several key components have been identified in the literature as influencing student outcomes in school. The number of days a student attends school, along with other factors such as intelligence, gender, age, and socioeconomic status (SES), have been found to be related to achievement gain (Easton & Engelhard, 1982). Multiple regression analyses were used to test the hypothesis that hours in general education predict student outcomes, above and beyond the prediction provided by the other independent variables known to be related to academic outcomes. Year 2 was omitted from the analyses due to missing academic achievement data. To determine if the predictor variables were more potent at one stage (year) versus another, outcome measures for each year were examined individually.

Measures

Academic outcomes, specifically reading and mathematics achievement, were measured using total scores in the mathematics and reading domains from the Woodcock-Johnson Tests of Achievement-Revised (WJ-R) and the Kaufman Test of Educational Achievement (KTEA), because these were the most frequently administered achievement tests at the three schools. Independent variables included full scale IQ, SES, grade, attendance (i.e., days absent), and average number of hours per week in general education. Full scale IQ was measured by a standardized test of cognitive ability administered as part of the psychoeducational assessment conducted to determine eligibility for special education services. SES was measured by student eligibility to receive either free or reduced lunch versus full-paid lunch in the government-sponsored school lunch program. Presence in general education classrooms was measured by the number of hours per week a student was scheduled to spend in general education academic settings. This information was gathered from Individual Educational Programs (IEP) found in the cumulative and special education folders of participants each year of the study.

Results

The results of this study are reported in two sections: (1) descriptive statistics for all study variables in years 1 and 3; and (2) inferential statistics to assess the effect inclusion in general education on reading and mathematics achievement scores. Table 2 shows that the average number of hours participants were included in general education each year varied greatly, ranging from 11.4 to 15.1 in the 1st year and 8.8 to 15.9 in the 3rd year. On the average, females, non-minorities, and students in the "full-pay lunch" category were included in general education more hours per week than were other categories of students. See website <http://rapidintellect.com/AEQweb/sum02.htm>

Analyses of students' reading achievement each year revealed that scores were higher for females, non-minorities, students on full-pay lunch, and students at school 2 in both years. Students' mathematics achievement scores each year revealed similar patterns to hours of inclusion and reading achievement. Non-minorities, students on full-pay lunch, and those at school 2 had higher mathematics achievement scores. In year 1, the mean mathematics score for females was higher than males; however, the mean score for males was higher in year 2. Next, a series of one-way analyses of variance and bivariate regressions were completed to examine unique contributions of each independent variable to the dependent variables. SES and school site were both statistically significant predictors of hours of inclusion (p<. 01). Students who reportedly were from homes with higher incomes were more likely to be included in general education. The mean hours per week of inclusion for students from school 2 and school 3 was statistically significantly higher than for students at school 1 (p<. 01).

Standardized regression coefficients for numbers of academic hours of inclusion and each continuous independent variable (i.e., full scale IQ, SES, grade, race, gender, and attendance) were computed. Full scale IQ was statistically significantly correlated with hours per week of inclusion in year 3. Significant positive correlations indicate that students with higher IQ scores tended to spend more time in general education during Year 3. Days absent from school and student grade level did not correlate significantly with hours of inclusion in either year.

Reading Achievement

Standard multiple regressions were used to test the relationship between reading achievement in years 1 and 3, and the independent variables of full scale IQ, absences, SES, grade, race, gender, and site. In year 1, two of the independent variables made statistically significant contributions to the prediction of reading achievement after adjusting for other variables, full scale IQ and academic hours in general education (t=2.93, p<.01) and (t=2.89, p<.01), respectively. In year 3, only academic hours in general education (t=3.154, p<.01) contributed significantly to prediction of reading achievement after adjusting for other variables. This indicates that students with higher IQs and those who spent more time in general education classrooms had higher reading scores. The remaining variables of race, gender, SES, grade, site, and

absences did not significantly contribute additional variance explaining reading achievement.

Mathematics Achievement

A second series of regression analyses was performed to test the relationships between mathematics achievement and the independent variables of full scale IQ, absences, SES, grade, race, gender, site, and hours of inclusion. In year 1, only full scale IQ (t=4.433, p<.0001) contributed significantly to prediction of mathematics achievement after adjusting for other variables. In year 3, full scale IQ (t=2.456, p<.05) and hours in general education (t=4.504, p<.0001) made statistically significant contributions to the prediction of mathematics achievement after adjusting for other variables. This demonstrates that students with higher IQ scores and those who spent more time in general education had higher mathematics scores.

Discussion

Results of this study support research findings that students with higher IQs tend to perform better academically in school (Boulon-Diaz, 1992; Hall, 1990) and that, on average, students with learning disabilities who spend more time in inclusive settings have higher achievement scores (Deno et al., 1990; Schulte, Osborne, & McKinney, 1990). However, the statistically significant correlation between IQ and academic hours in general education raises difficulties in interpreting the association between integrated settings and outcomes for students with learning disabilities.

Although marked individual differences occur in outcomes among students with learning disabilities, a variety of variables, including IQ, SES, and self-esteem, have been related to academic achievement (Kavale, 1988). Results of this study are consistent with other correlational studies using full scale IQ as a predictor variable. Both full scale IQ and hours in general education were consistent predictors of reading and mathematics achievement. Although interpretation of the regression analyses does not imply causality, it does provide further evidence of a relationship between inclusion in general education and academic outcomes. This study supports the conclusion that school placement decisions are an important variable in predicting student outcomes. The variable average hours per week inclusion contributed significantly to the prediction of reading achievement both years and to mathematics achievement in year 3. Other factors often associated with school achievement, including gender, race, age, SES and attendance did not contribute significantly to variation in reading and mathematics achievement scores. The composite set of eight predictor variables accounted for approximately 43% of the variance in reading and mathematics achievement scores, except for the prediction of mathematics scores in year 3. The regression equation for mathematics in year 3 accounted for 53% of the variance.

Demographic Variables

Race and gender were not originally designated as independent variables, but they were included in the final regression equations because of the significant correlations among both variables and the hours included in general education. Other studies using race in multiple regression analyses have reported a correlation between ethnicity and school achievement (Neisser, 1986). Some researchers challenge the use of ethnicity as a predictor in models of achievement (Penny & Bond, 1991), while others argue that statistically removing its effect from achievement will promote a better understanding of the unique effect of other factors, such as SES or hours in general education (Caldas, 1993). In this study, SES was not a statistically significant predictor of academic outcomes. Analyses showed that SES was a statistically significant predictor of hours of inclusion in year 3 (p<. 01). Since academic success is more prevalent in higher SES groups (Lee-Corbin & Evans, 1996), the true influence of hours in general education may be overstated. Gender has also been reported to significantly influence academic achievement, particularly at the secondary level (Hill, 1989). However, neither SES nor gender statistically significantly contributed in the prediction of academic achievement in this study. Similarly, the student variables of grade and attendance showed no statistically significant effect on student performance.

Inclusion

Although attitudes of key personnel and inclusive practices and programs implemented at schools were used as criteria for selection of schools in the study, teacher instruction and classroom environment factors for each student were not assessed. As the actual amount of time a student was engaged in academic tasks was not assessed, the criterion of a student's "presence" as a proxy for academic inclusion does not reflect the nature or quality of a student's participation in a general education classroom. In fact, an alternative interpretation of the data in this study is that the positive relationship between inclusion and achievement is a factor of the implicit criteria used by teachers when selecting students for inclusion in general education. Improved academic outcomes may be simply a validation of a team's initial decision to include a student for few or numerous hours, rather than a validation of the service delivery setting. Research also suggests that successful inclusion requires classroom environments that promote acceptance and social development (Stevens & Slavin, 1995). The level of involvement of students with disabilities is determined in part by teacher attitudes and behavior rather than administrative decisions (Schumm & Vaughn, 1992). Given the complexity and idiosyncrasies of schools, identifying and specifying actual changes taking place that are contributing to improved student outcomes is difficult. This is particularly problematic in special education research where "we are often attempting to measure behavior and performance at the extremes of the normal distribution, wherein lies the maximum error of measurement" (McKinney, 1994, p. 222).

Conclusion

Currently, the field lacks an adequate research base for determining the most effective system for delivering special education services. Due to the complexity of research on inclusive services for students with disabilities, most studies are still descriptive and fail to provide conclusive evidence concerning effects on academic achievement. Overall, inclusive environments in this study proved effective in predicting student achievement. However, differences among students with LD, including IQ, suggests that future research on inclusive environments should examine how student entry variables interact with other input variables in an effort to attribute improved outcomes to specific school or teacher variables, such as service delivery model or classroom instruction.

References

Baker, T., Wang, M., & Walberg, H. (1994). The effects of inclusion on learning. Educational Leadership, 52, 33-35.

Boulon-Diaz, F. (1992, August). The effects of intelligence, social class, early development and pre-school experience on school achievement of Puerto Rican children. Paper presented at the annual convention of the American Psychological Association, Washington, DC.

Brady, M. P., Swank, P. R., Taylor, R. D., & Freiberg, H. J. (1988). Teacher-student interaction in middle school mainstreamed classes: Differences with special and regular education students. Journal of Educational Research, 81 (6), 332-339.

Caldas, S. J. (1993). Reexamination of input and process factor effects in public school achievement. The Journal of Educational Research, 86, 206-214.

Carlberg, C. K., & Kavale, K. (1980). The efficacy of special versus regular class placement for exceptional children: Meta-analysis. Journal of Special Education, 14 (3), 295-309.

Deno, S., Maruyama, G., Espin, C., & Cohen, C. (1990). Educating students with mild disabilities in general education classrooms: Minnesota alternatives. Exceptional Children, 57 (2), 150-161.

Easton, J.Q., & Engelhard, G. (1982). A longitudinal record of elementary school absence and its relationship to reading achievement. Journal of Educational Research, 75, 269-274.

Hall, A. M. (1990). Techniques for evaluating the effectiveness of programs for the learning disabled: Determining program factors which influence student outcomes. Georgia State University, Dissertation Abstracts International, 51 (A), 0819.

Halvorsen, A.T., & Sailor, W. (1990). Integration of students with severe and profound disabilities: A review of research. In R. Gaylord-Ross (Ed.), Issues and Research in Special Education (Vol. 1, pp. 110-172). NY: Teachers College Press.

Hill, C. (1989, October). Predictors of seventh-grade mathematics achievement. Paper presented at the annual meeting of the Northern Rocky Mountain Education Research Association, Jackson WY.

Kavale, K. A. (1988). Using meta-analysis to answer the question: What are the important, manipulable influences on school learning? School Psychology Review, 17, 644-650.

Lee-Corbin, H., & Evans, R. (1996). Factors influencing success or underachievement of the able child. Early Child Development and Care, 117, 133-144.

Lipsky, D., & Gartner, A. (1995). The evaluation of inclusive education programs. NCERI Bulletin, 2 (2), 1-8. Washington, DC: Office of Educational Research and Improvement.

MacMillan, D. L., & Hendrick, I. G. (1993). Evolution and legacies. In J. I. Goodlad & T.C. Lovitt (Eds.), Integrating general and special education (pp. 23-48). New York: MacMillan.

McKinney, J. D. (1994). Methodological issues in longitudinal research on learning disabilities. In Vaughn, S. & Bos, C. (Eds.). Research issues in learning disabilities: Theory, methodology, assessment, and ethics (pp. 202-230). New York: Springer-Verlag.

Neisser, U. (1986). New answers to an old question. In U. Neisser (Ed.), The school achievement of minority children: New perspectives (pp. 1-17). Hillsdale, NJ: Lawrence Erlbaum Associates.

Penny, J., & Bond, L. (1991). Use of ethnicity as a predictor of achievement: A reply to Klingele and Warrick. Journal of Educational Research, 84, 133-134.

Schulte, A.C., Osborne, S.S., & McKinney, J.D. (1990). Academic outcomes for students with learning disabilities in consultation and resource programs. Exceptional Children, 57, 162-172.

Schumm, J. S., & Vaughn, S. (1991). Making adaptations for mainstreamed students: General classroom teachers' perspectives. Remedial and Special Education, 12(4), 18-27.

Stevens, R., & Slavin, R. (1995). Effects of a cooperative learning approach in reading and writing on academically handicapped and nonhandicapped students. The Elementary School Journal, 95 (3), 241-262.

Weintraub, F. (1991). The REI debate: What if everybody is right? In J. W. Lloyd, N. Singh, & A. C. Repp (Eds.) The regular education initiative: Alternative perspectives on concepts, issues, and models (pp. 17-27). Sycamore, IL: Sycamore.

Zigmond, N., Jenkins, J., Fuchs, L., Deno, S., Fuchs, D., Baker, J., Jenkins, L., & Couthino, M. (1995). Special education in restructured schools: Findings from three multi-year studies. Phi Delta Kappa, 76 (7), 531-540.

K. Sarah Hall, Ph.D., is an Assistant Professor in the Department of Special Education. Her current research interests include educational technology and the physical activity and nutrition of girls with physical disabilities.