Self-Determination and Access to the General Education Curriculum

By Lee, Suk-Hyang Wehmeyer, Michael L; Palmer, Susan B; Soukup, Jane H; Little, Todd D

Addressing federal mandates regarding both the delivery of transition services and access to the general education curriculum has been a challenge for secondary special educators. A practice common to both initiatives, however, has been efforts to promote self-determination. This study examined the impact of promoting self- determination as an instructional strategy to examine (a) the relationship between self-determination and access to the general education curriculum and (b) the impact of promoting self- determination as a curriculum augmentation on access to the general education curriculum for high school students with disabilities. The study implemented a randomized trial control group design in which classroom observations were conducted to determine student access to the general education curriculum before and after the implementation of an instructional model to enable students to self-regulate learning. Analyses were conducted using multilevel model methods. The findings indicated the potential positive impact of promoting self-determination on access to the general education curriculum for students with disabilities as well as the degree to which curriculum modifications and other supports are needed to further access and student progress. Keywords: access to the general education curriculum; self-determination; curriculum modification; students with disabilities; student-directed learning

The Individuals with Disabilities Education Act (IDEA) requires that the Individualized Education Programs (IEPs) of all students with disabilities address student needs pertaining to transition services and ensure involvement with and progress in the general education curriculum [Sec.602(a)(19), 1414(d)]. The 2004 reauthorization of IDEA, the Individuals with Disabilities Education Improvement Act (IDEIA), continued to emphasize both transition services and access to the general education curriculum by requiring documentation of both academic and functional achievement [Sec.602 (34)(A)]. These dual IDEA requirements, however, may present a challenge for secondary special education teachers (Johnson, 2002; Johnson, Stodden, Emanuel, Luecking, & Mack, 2002; Kochhar-Bryant & Basset, 2002), primarily because teachers may perceive them as being mutually exclusive (Blalock & Patton, 1996; Katsiyannis, Zhang, Woodruf, & Dixon, 2005; Wehmeyer, Sands, Knowlton, & Kozleski, 2002). That is, the middle and high school curriculum places an increased emphasis on academic achievement (Deshler et al., 2001; Mastropieri & Scruggs, 2001), which may be perceived by teachers as competing with the transition requirements.

The result, too often, may be that teachers and others view efforts to provide transition services and to promote access to the general education curriculum as incompatible and, in essence, choose to do one or the other (Kochhar-Bryant & Bassett, 2002). It is important, as such, to develop effective instructional strategies to align these two educational initiatives for youth with disabilities (Johnson, 2002; Kochhar-Bryant & Bassett, 2002; Wehmeyer, 2003). Examining the best practices shared between these two areas would be a starting point to ensure that the educational programs of students with disabilities address both. One such area of common practice has been the area of self-determination. Promoting the self- determination of youth with disabilities has become best practice in transition services (Momingstar, Kleinhammer-Tramill, & Lattin, 1999; Trainor, 2005; Wehmeyer, Agran, & Hughes, 1998). Research has shown that students with disabilities who are more self-determined when they leave school achieve more positive adult outcomes (Wehmeyer & Palmer, 2003; Wehmeyer & Schwarte, 1997).

In addition, promoting self-determination has been identified as a means to enhance access to the general education curriculum for students with disabilities (Wehmeyer, Field, Doren, Jones, & Mason, 2004). The 1997 amendments to the IDEA introduced statutory and regulatory language pertaining to ensuring access to the general education curriculum for students receiving special education services and required that the IEP of all students receiving special education services include a statement of (a) how the student’s disability affects involvement and progress in the general curriculum, (b) the program modifications or supports for school personnel that will be provided for the child to be involved and progress in the general curriculum, and (c) the special education and supplementary aids and services to be provided to ensure a student’s involvement in and progress in the general curriculum (34 C.F.R. [section] 300.347(a)(3)). The IDEA regulations defined the “general education curriculum” as “the same curriculum as for nondisabled children” (Rules and Regulations, 64 C.F.R. 12592, 1999). The intent of these mandates was to align practice in special education with school reform efforts in general education, efforts that have been codified in the No Child Left Behind (NCLB) Act.

The 2004 IDEIA continued these IDEA ’97 requirements and extended them, mandating that schools ensure that the IEP team includes someone knowledgeable about the general education curriculum and that it meet at least annually to address any lack of expected progress in the general education curriculum. Finally, the proposed regulations to IDEIA 2004 (issued in June 2005) prohibit a student with a disability from being removed from the general education setting based solely on needed modifications to the general education curriculum. The general education curriculum is defined as the same curriculum as provided to students without disabilities and, in practice, is defined by the academic content and student performance standards in each state mandated by NCLB. It is evident, thus, that promoting student access to, and progress in, the general education curriculum remains a focal point of federal policy and national school reform efforts affecting students with disabilities.

Promoting self-determination can, theoretically, promote access to the general education curriculum in a number of ways. First, component elements of selfdetermined behavior, such as goal setting, problem solving, and decision making, are frequently found as elements of local and state content and student achievement standards (Wehmeyer et al., 2004). Furthermore, as a means for access, teaching students skills such as goal setting, problem solving, and self-regulation can serve as curriculum augmentations that, in turn, can promote student involvement with and progress in the general education curriculum (Knowlton, 1998; Wehmeyer et al., 2004).

Curriculum augmentation refers to efforts to augment or expand the curriculum to include instruction on skills or strategies that help students succeed within the general education curriculum. Such curriculum augmentations include (a) strategies for learning (e.g., learning-to-learn strategies, mnemonic strategies), (b) strategies for test taking, (c) strategies for organization, and (d) self- regulation strategies (e.g., selfmonitoring, goal setting, problem solving; Bashinski & Wehmeyer, 2002). Research has supported the efficacy of these types of curriculum augmentations as a mean to enhance access for students with disabilities by promoting student self-directed learning and academic achievement across multiple content areas (i.e., science, math, social studies, etc.), such as mnemonic strategies (Mastropieri & Scruggs, 1991; Scruggs & Mastropieri, 2000), test-taking strategies (Hughes & Schumaker, 1991; Hughes, Schumaker, Deshler, & Mercer, 1988), and organization strategies (e.g., managing time and handling learning materials; Bryan, Burstein, & Bryan, 2001; Hughes, Ruhl, Schumaker, & Deshler, 2002). Self-regulation strategies, such as goal setting, selfmanagement, self-monitoring, and problem solving, have been widely used to increase students’ learning and on-task behaviors across various academic domains (i.e., reading, writing, spelling, math problem solving, etc.) and appropriate behavior in the general education classroom (Deshler, Ellis, & Lenz, 1996; Johnson, Graham, & Harris, 1997; King-Sears & Bonfils, 1999; Maccini & Hughes, 2000; Sexton, Harris, & Graham, 1998; Shimabukuro, Prater, Jenkins, & Edelen-Smith, 1999). In sum, these curriculum augmentation strategies have the potential to play a critical role in enhancing student access to the general education curriculum by enabling students to learn and apply skills that support enhanced self- directed learning (Palmer, Wehmeyer, Gipson, & Agran, 2004). In addition, given that augmenting the curriculum to teach students to self-direct learning, set goals, and solve problems also promotes self-determination, the implementation of such curriculum augmentations can serve to both improve transition outcomes and promote access to the general education curriculum (Wehmeyer et al., 2004).

Despite the potential benefit, recent research on the degree to which curriculum modifications were provided to promote access to the general education curriculum for students with disabilities showed limited use of curriculum augmentations (Soukup, Wehmeyer, Bashinski, & Bovaird, 2007; Wehmeyer, Lattin, LappRincker, & Agran, 2003). Wehmeyer, Lattin, et al. (2003) observed 33 middle school students with intellectual disabilities in general education classrooms for a total of 6,585 minutes. They found that the percentage of intervals in which students engaged in activities linked to the general education curriculum using some form of curriculum adaptation was just 2.78% and there were no curriculum augmentations evident. Recently, Soukup et al. (2007) observed 19 elementary students with intellectual and developmental disabilities for a total of 1,140 minutes and recorded the occurrence of curriculum adaptations and augmentations. The results of this study mirrored the Wehmeyer, Lattin, et al. (2003) findings. Students most often were provided accommodations, followed by adaptations. There were no occurrences of curriculum augmentations observed for any group. Recent research, however, supports the hypothesis that promoting self-determination as a curriculum augmentation (i.e., teaching goal setting, problem solving, and self-monitoring) has the potential to promote access to the general education curriculum for students with and without disabilities. This research has documented (a) the positive relationship between the level of a student’s self- determination and academic achievement of students without disabilities (Goldberg & Cornell, 1998; Hardre & Reeve, 2003); (b) the positive effect of teaching choice making, self-regulation, or problem-solving skills as a means to replace and reduce problem behaviors, which have been identified as a significant barrier to learning in the general education classroom for students with disabilities (Martin, Mithaug, Peterson, VanDycke, & Cash, 2003; Mayer, Lochman, & Acker, 2005; Shogren, Faggella-Luby, Bae, & Wehmeyer, 2004); (c) the positive effect of learning-to-leam strategies or cognitive strategies to promote student interaction with and progress in the general education curriculum for students with learning disabilities (Guastello, Beasley, & Sinatra, 2000; Scruggs & Mastropieri, 2000); (d) the impact of teaching students directed learning strategies (e.g., selfmonitoring, self- evaluation, or self-instruction) to enable students with disabilities to achieve educationally relevant goals in the context of general education classroom (Agran et al., 2005; Copeland, Hughes, Agran, Wehmeyer, & Fowler, 2002; Hughes, Copeland, et al., 2002; Wehmeyer, Yeager, Bolding, Agran, & Hughes, 2003); and (e) evidence of the direct benefit in student attainment of goals linking the general education curriculum from instruction to promote selfdetermination and student-directed learning (Agran, Cavin, Wehmeyer, & Palmer, 2006; McGlashingJohnson, Agran, Sitlington, Cavin, & Wehmeyer, 2004; Palmer et al., 2004).

In spite of the emerging evidence for the potential benefit of promoting self-determination as a curriculum augmentation, to date there has been little research to directly examine the relationship between self-determination and access to the general education curriculum for high school students with disabilities and to investigate the impact of promoting self-determination as a curriculum augmentation on access to the general education curriculum. This study addressed two aspects of this research gap by examining (a) the relationship between self-determination and overall access to the general education curriculum and engaged and nonengaged academic behavior and (b) the impact of promoting self- determination as a curriculum augmentation on access to the general education curriculum.

Method

Participants and Settings

The initial pool of participants included 45 high school students with disabilities recruited from 11 campuses at seven suburban school districts in the Midwest. To be included in the sample, students must have received instruction in core content areas (e.g., English, math, social studies, or science) in the general education classroom. Although current (e.g., within 2 years) scores from standardized intelligence tests were not available for most students, an indicator of support for academic participation was collected. To provide an indicator of student level of functioning, teachers responded to items asking them to rate, on a Likert-type scale ranging from 1 (no support needed) to 5 (total support needed), the degree to which students needed support to function overall (e.g., independent living, daily care, community integration, etc.), referred to subsequently as Overall Support Needs, and to acquire new knowledge and skills, referred to subsequently as Learning Support Needs.

To examine the impact of promoting self-determination as a means to enhance student access to the general education curriculum, analyses were conducted on results from 42 students because two of the 45 students initially recruited were suspended from school during the study period and one student was not available for the second observation. Students were randomly assigned, at the level of the high school campus, to a control or treatment condition. Twenty of the students who completed the study were in the treatment group, with 22 in the control group. The characteristics of each group are described in Table 1. Caucasians were the largest racial group in both the control (75%, n = 15) and experimental group (68%, n = 15). The average of Overall Support Needs in the control and experimental group was 2.54 (SD = 1.01) and 3.25 (SD = .79), respectively, and mean Learning Support Needs was 2.64 (SD = 1.00) and 3.10 (SD = 1.07), respectively. The Arc’s Self-Determination Scale (SDS; Wehmeyer & Kelchner, 1995) total score in both groups averaged 101.9 (SD =16.10) for the control group and 96.00 (SD = 17.34) for the treatment group. Students in the control group were observed in English (27%, n = 6), math (27%, n = 6), social studies (27%, n = 6), and science class (18%, n = 4), respectively. In the experimental group, interventions with and observations of students took place in English (25%, n = 5), math (20%, n = 4) social studies (40%, n = 8), and science (15%, n = 3) classes. Overall, 29 general education teachers were involved in this study (17 teachers for the control group and 12 teachers for the experimental group). The average number of years teaching for general education teachers in the control and experimental group was 6.9 years (SD = 6.74) and 9.9 years (SD = 7.10), respectively. Three students in the control group and two students in the experimental group were observed in a class taught by both a general education and a special education teacher in a coteaching arrangement.

To examine the relationship between selfdetermination and access to the general education curriculum, analyses were conducted on a subset of the total sample whose scores on the SDS (Wehmeyer & Kelchner, 1995) fell at or above the median total score for the sample. We opted to identify this subgroup for analysis because we were interested in the degree to which being self-determined contributed to access to the general curriculum and, as such, opted to select those students who were in the top 50% of selfdetermination scores as representative of students who are more self-determined. From the total sample of 45 students, 21 students from 9th through 12th grade were identified for this high self- determination group. The mean age of the 21 students was 16.41 (SD = 1.32), ranging from 14.4 to 19.2. Eleven students were male (M age = 16.53, SD = 1.41) and 10 students were female (M age = 16.29, SD = 1.28). Nineteen participants (91%) were students with learning disabilities and the rest of the students were identified as having attention-deficit disorder (ADD), attention-deficit/hyperactivity disorder (ADHD), or another health impairment. Their mean score on the SDS was 112.94 (SD = 9.59). The average Overall Support Needs and Learning Support Needs for students in the high self- determination group were 2.8 (SD = .77) and 2.8 (SD = .93), respectively. Students were observed in English (24%, n = 5), math (38%, n = 8), science (28%, n = 6), and social studies (9%, n = 2) classes.

For all students, parental consent was obtained prior to any data collection. We also obtained consent from all teachers in the study, both special and regular education.

Procedures

The study used a pretest-posttest randomized trial control group design. For all students, baseline data pertaining to self- determination using multiple studentand teacher-report measures and access to the general curriculum using a Windows PC-based data collection system were obtained immediately on receipt of informed consent. In addition, goal attainment data were collected at the posttest for students in the treatment group. Special education teachers with students in the experimental group received training regarding the intervention to promote self-determination: the SeIfDetermined Learning Model of Instruction (SDLMI; Wehmeyer, Palmer, Agran, Mithaug, & Martin, 2000). The SDLMI is a model of teaching designed to enable teachers to teach students to engage in self-regulated and self-directed learning. The SDLMI is based on the component elements of self-determined behavior, the process of self- regulated problem solving, and research on student-directed learning. It is appropriate for use with students with and without disabilities across a wide range of content areas and enables teachers to engage students in the totality of their educational program by increasing opportunities to selfdirect learning and, in the process, to enhance student self-determination.

Implementation of the model consists of a threephase instructional process: set a goal (Phase 1), take action (Phase 2), and adjust goal or plan (Phase 3). Each instructional phase presents a problem to be solved by the student. The student solves each problem by posing and answering a series of four Student Questions per phase that students learn, modify to make their own, and apply to reach self-selected goals. The four questions differ from phase to phase but represent identical steps in the problem-solving sequence: (a) identify the problem, (b) identify potential solutions to the problem, (c) identify barriers to solving the problem, and (d) identify consequences of each solution. The solutions to the problems in each phase lead to the problem-solving sequence in the next phase. Each question is linked to a set of Teacher Objectives that, in essence, describe the outcomes desired by having the student answer the question. Each instructional phase includes a list of Educational Supports that teachers can implement to enable students to selfdirect learning. In each instructional phase, the student is the primary agent for choices, decisions, and actions, even when eventual actions are teacher-directed. Previous research has verified the efficacy of the SDLMI as a teaching model to enable students with cognitive disabilities, including students with intellectual disabilities and students with learning disabilities, to attain educationally valued goals (Agran, Blanchard, & Wehmeyer, 2000; Wehmeyer et al., 2000). One of the developers of SDLMI and the lead researcher trained all special education teachers in the experimental group to implement the SDLMI with their students. In addition, training also included instruction with regard to data collection activities. Researchers worked with special education teachers to identify content areas within the general education curriculum for each student in which goals could be set that would be tied directly to state and local standards and to activities in the general education classroom. Special educators communicated with general education teachers as necessary to determine appropriate content areas and tasks. The lead researcher provided technical assistance to teachers throughout the study. Once special education teachers were trained to implement the SDLMI, they were responsible for working through Phases 1 and 2 with the student. These phases, respectively, result in students setting an educational goal and developing an action plan that includes a self-monitoring plan to achieve the goal set. Once students were supported by the special education teacher using the SDLMI, students implemented their action plan and self-monitored their progress toward their goal in the context of their relevant general education context class. General education teachers were wholly responsible for instruction in the classroom once students began implementing their action plan, although special education teachers provided supports to students with regard to using the model’s questions when necessary.

The duration each student used the SDLMI in the general education classroom varied depending on each school’s academic calendar or schedule, when the student’s parent provided informed consent, and when baseline data collection could be completed. The mean duration (e.g., number of weeks) in which the intervention was implemented, beginning with Phase 1, was 10.1 weeks (SD = 3.18), ranging from 4 to 16 weeks. The mean duration (e.g., number of weeks) in which the intervention was implemented in the general education classroom (e.g., action plan was implemented and Phase 3 completed) was 8.1 weeks (5D = 2.95), ranging from 2 to 11 weeks. The variability in the length of time in which students were engaged with the SDLMI was a function of several factors, including those mentioned previously (school calendar, return of consent, completion of baseline data collection) as well as the speed with which the student used the model to reach the goal set.

Fidelity to treatment for implementation of the SDLMI was monitored by three types of fidelity measurement (Fixen, Naoom, Blase, Friedman, & Wallace, 2005): (a) a context fidelity measure that describes the necessary precursors to high-level performance (e.g., completion of training), (b) a compliance fidelity measure that provides an outline of the core intervention components and their use by the practitioner, and (c) a competence fidelity measure that illustrates how well the practitioner is performing the core intervention components of an evidence-based program or practice. For the context fidelity indicator, all special education teachers received training from the same trainer on the SDLMI for about an hour. Compliance fidelity was monitored through ongoing support and communication to facilitate teachers to implement the SDLMI. For this, regular notices to announce important agendas and schedules of implementation were sent via e-mail before moving on to the next phase of the SDLMI. All teachers and students followed the same procedures regarding implementation of the SDLMI. Competence fidelity was evaluated by reviewing worksheets completed by the participating students to achieve their goals related to core content areas, which included student responses to the Student Questions in each of the three phases and checklists completed by teachers that show Teacher Objectives and Educational Supports to which teachers referred while teaching students the model. In total, 97% of student worksheets of each phase and teacher objectives and educational supports were completed except for Phase 3 worksheets for two students who were dismissed from special education service. Competence fidelity also was calculated for each student by dividing the number of weeks in which students actually used their self- monitoring sheet or study guide to work on and monitor their goals and plans by the total number of weeks from the beginning of Phase 2 (Take Action) to Phase 3 (Adjust Goal or Plan). The overall mean of the student competence fidelity was 73% (SD = 24.47, ranging from 27%100%). Analysis of variance indicated no significant differences in student competence fidelity between 13 students with teachers who had previously been introduced to the model and seven students with teachers who had not received any training prior to the start of the study.

After students in the experimental group implemented the SDLMI in conjunction with their special education teachers and finished all phases of the SDLMI (described below), second observations were conducted for all students in both the experimental and control groups in the same sequence as their baseline data collection.

Instrumentation

Measuring self-determination. Two measures were used to determine student self-determination. The SDS (Wehmeyer & Kelchner, 1995) is a 72-item selfreport measure that provides data on self-determination through the measurement of four essential characteristics of self- determined behavior: autonomy, self-regulation, psychological empowerment, and self-realization (Wehmeyer, 1996a). A total of 148 points are available on the scale, with higher scores indicating higher levels of self-determination. The SDS was developed and normed with 500 adolescents with cognitive disabilities (Wehmeyer, 1996b). It was demonstrated to have adequate reliability and validity when used with adolescents with cognitive disabilities. Construct validity was determined by multiple means, the first of which was a factor structure analysis. The scale had adequate internal consistency (Cronbach’s alpha = .90). The second measure, The AIR Self-Determination Scale (AIR; Wolman, Campeau, Dubois, Mithaug, & Stolarski, 1994) includes Student, Educator, and Parent versions. For the purposes of this study, the Student and Educator versions of the scale were used. The AIR-Educator (AIR-E) and AIR- Student (AIR-S) scales consist of 30 questions and 18 questions, respectively, that provide data on students’ capacity and opportunity for self-determination rated on a scale from 1 (never) to 5 (always). The AIR-S and AIR-E were developed and normed with 450 students with and without disabilities and their teachers in California and New York (Wolman et al., 1994). Both versions of the scale were demonstrated to have adequate reliability and validity in the measurement of self-determination for students with and without disabilities.

Measuring access to the general education curriculum. A momentary time-sampling method using a Windows PC-based data collection system called the Access Version of the Code for Instructional Structure and Student Academic Response (Access CISSAR) was used to collect data concerning student access to the general education curriculum. The Access CISSAR is an expanded version of a direct observational system-the MainStream Version of the Code for Instructional Structure and Student Academic Response (MS-CISSAR; Carta, Greenwood, Schulte, Arreaga-Mayer, & Terry, 1988)-a component of the EcoBehavioral Assessment System Software (EBASS) that has been designed to collect classroom observational data using notebook computers (Greenwood, Carta, Kamps, Terry, & Delquadri, 1994). The MS-CISSAR is an observation system that focuses on the individual student as the observer’s target and collects data on 105 individual codes in 13 categories across three variables: student behavior (3 categories), teacher behavior (5 categories), and classroom ecology (5 categories). Data are collected in the 13 categories of variables during a 60-second interval composed of three 20-second observation intervals for each student behavior, teacher behavior, and ecology variable. One event may be recorded for each of the 13 categories during each interval in full-minute increments and data entry is limited to four active keys to reduce the probability of erroneous entries. The MS-CISSAR was designed with rigorous technical scrutiny during its development and field testing. Test-retest reliability averaged .85 overall (Greenwood et al., 1997). The MS-CISSAR’s divergent validity was demonstrated through correlation of students’ higher levels of academic responding in the classroom, with posttest gains on the Metropolitan Achievement Test-Basic Scale (Greenwood, Arreaga-Mayer, & Carta, 1994). The Access CISSAR was designed to collect additional data on when and how opportunities were made available for students to access the general education curriculum (Bashinski & Wehmeyer, 2002). The Access CISSAR includes all original student behavior and ecology categories included in the MS- CISSAR and all but three of the teacher behavior variables. To capture data indicating the degree to which students had access to the general education curriculum, the Access CISSAR also included new observational categories based on research by Wehmeyer, Lattin, et al. (2003). These include (a) whether peers of target student (e.g., the student with a disability being observed) were engaged in a task linked to any general education standard or a grade-level standard; (b) whether a target student was engaged in a task linked to any general education standard or a grade-level standard; (c) whether a target student was engaged in a task linked to an IEP goal; and (d) whether accommodations, curriculum augmentations, or curriculum adaptations were provided. Accommodations refer to supplementary aids or services that enable students to be engaged with and respond to the same curriculum with peers without disabilities but that do not change or modify the general education curriculum in any way, such as paraprofessional or peer support, having a note taker, adjusting the classroom environment or the time to complete tasks or assessments, or the implementation of assistive technology. Curriculum adaptations refer to modifications to the ways in which curriculum content is represented or presented to the student or the way in which students respond to the curriculum (e.g., adjusting reading or cognitive demand, nonprint or enhanced content, content delivered through technology, nontraditional responsefs] to instruction or nontraditional instructional materials). As previously reviewed, curriculum augmentations involve efforts to augment or expand the curriculum to teach skills or strategies to facilitate students’ success within the general education curriculum-including leaming-to-learn strategies, test- taking strategies, organization strategies, and self-regulation strategies-that enable students to self-direct learning and more effectively engage with and have access to the curriculum. All Access CISSAR variables can be reentered at any time during data collection. The coder can simultaneously code as many access- related variables as observed.

The lead researcher, who was the primary data collector, received one-on-one training from a trainer who had been skilled to mastery on the MS-CISSAR and received training on the Access CISSAR data collection system. Training began with an instrument calibration process: A test of an observer’s capacity to collect data from a videotaped classroom simulation in agreement with a standard set by the original MS-CISSAR’s software developers. Reliability training for the MS-CISSAR also included live, in-school practice and in- school reliability sessions. After receiving an overall reliability rating of 97.49% agreement with a trainer for three in-school training sessions (a total of 61 minutes) on the MS-CISSAR, the primary observer received training on the Access CISSAR, obtaining a reliability score of 97.85% for that version, including Access CISSAR and MS-CISSAR. The trainer also served as a second observer for this study. Additional observers were skilled to mastery using the same training process and trainer as discussed previously. Overall reliability ratings of agreement for the two additional observers training subsequently on the three in-school sessions across primary observer and two secondary observers were 98.18% (a total of 64 minutes between primary and one second observer), 95.75% (a total of 86 minutes between primary and the other second observer), and 98.61% (a total of 50 minutes between second observers), respectively.

Access scores. For purposes of this study, student access to the general education classroom was operationally defined as a function of whether students were engaged in the behaviors and activities recorded by the Access CISSAR toggles, for example, (a) whether peers of the target student were engaged in a task linked to any general education standard or a grade-level standard; (b) whether a target student was engaged in a task linked to any general education standard or a grade-level standard; (c) whether a target student was engaged in a task linked to an IEP goal; and (d) whether accommodations, curriculum augmentations, or curriculum adaptations were provided. Specifically, we calculated an overall access score, as indicated in Equation 1, by combining assigned point values for the F4 toggle (participant engaged in task linked to any general education standard) or F5 toggle (participant engaged in task linked to grade-level standard) plus F7 toggle (curriculum accommodations), F8 toggle (curriculum augmentations), and F9 toggle (curriculum adaptations).

This access score was developed by Soukup et al. (2007) to calculate the degree to which children with disabilities in elementary school were accessing the general curriculum during observed sequences. During each observation, each minute the F4 toggle (any general education standard) was activated counted as 1 point. If the F5 toggle (grade-level general education) was activated, then 3 points were tallied for the total access score. When any F7 toggle (accommodation) was coded during an observation interval concurrently with either F4 or F5 toggles, then 1 point was counted toward the total access score. In any interval in which an F8 toggle (augmentation) or F9 toggle (adaptation) were coded concurrently with F4 or F5 toggles, 3 points were added to the total access score. Thus, the resulting access score for any given observation interval per student could range between zero and 10. The weighted value for each variable was based on the hypothesized importance of that variable, as determined by Soukup et al., for students with disabilities to achieve access to the general curriculum. In essence, the gold standard for access should be that students are working on grade-level general education standards (F5) and receive needed accommodations and curriculum modifications (F7- F9). The weighting system was put in place to better differentiate between circumstances in which students were receiving instruction and curriculum modifications that met or approached that gold standard and circumstances when only one component was in place (e.g., only paraprofessional support).

Goal Attainment Scaling (GAS). GAS was used to assess student attainment of educational goals linked to core content areas in the general education curriculum for students in the experimental group. The GAS is a rubric to compare individual student progress on goals across multiple goal areas and provides a choice of five levels of predicted achievement (Kiresuk, Smith, & Cardillo, 1994). After students set a goal linked to the general education curriculum, goal outcomes for each goal using a 5-point continuum of possible outcomes were developed with teacher support. After completion of the study, teachers selected the outcome that best described the students’ progress toward achieving the goal. Raw scores were converted to standardized T scores (Kiresuk, Smith, & Cardillo) with a mean of 50 (an acceptable outcome) and a standard deviation of 10. Scores of 40 or less indicate outcomes that teachers found less acceptable.

Data Collection

All students were observed in their general education class for a total of 60 minutes, once at baseline and once postintervention, for 30 minutes each time. The mean number of weeks between the first observation and second observation for control group students was 5.5 weeks (SD = .17, range = 5-8 weeks). The mean number of week for students in the treatment group was 7.2 weeks (5D = 2.68, range = 2- 11 weeks), adjusted to account for interruptions in student attendance and teacher scheduling of treatment. Observations were conducted in typical classroom sessions. Immediately before each data collection session, the researcher (and reliability coder, if scheduled) met with teachers to clarify all information regarding content-related standards (on/off grade-level standard) to be addressed by the day’s lesson; IEP goals addressed by the activities; and types of accommodations, curriculum adaptations, and augmentations that would be implemented. During each observation, the coder was seated to record data where she was able to hear and see a study participant but not intrude on instructional activities. First, the observer entered student information required by the Access CISSAR software and coded the initial access toggle settings before beginning the coding interval. The observer then began the formal coding session by entering information that corresponded to the situation observed using the MS-CISSAR protocol’s “look, record, and rest” pattern.

Data Analyses

Relationship between self-determination and access. For the entire first observation, data analysis to address the first question was conducted on 21 out of the 45 students who were in the initial pool of participants and interrater reliability was obtained through an MS-CISSAR reliability check report (Greenwood & Hou, 1995). Cohen’s Kappa statistic (Sax, 1997) was used to calculate an index of interobserver agreement. Kappa values higher than .60 are generally considered to be adequate levels of agreement (Hartmann & Woods, 1982). Power analysis (Sample Power 2.0) was conducted to identify the power for the sample of 21 students with six predictor variables (i.e., the self-determination scores of four subscales of SDS, ATR-S, and AIR-E) and four independent variables (i.e., access score, academic response, competing response, curriculum accommodations). With the six predictor variables and an average R^sup 2^ of .41 (n = 21) across four dependent variables, observed power was .79. Within the data set, there was a small amount of missing data on a number of variables (approximately 1.0%). Because of the potential deleterious effects of not including all available data in the analysis process, the expectationmaximization (EM) imputation algorithm using the Missing Value Analysis procedure within the SPSS program was used to impute the missing data (Croy & Novins, 2005) before data were analyzed. Because data from each of the 30 observation intervals (Level 1) were nested within each of the 21 students (Level 2), multilevel (two-level) regression analyses were employed to analyze the data to examine the relationship between self-determination and access to the general education curriculum. The multilevel model can describe how students change over time and how these changes vary across students by including fixed effects that do not vary across individuals and random effects that have values that vary randomly within and/or between individuals (Singer, 1998; Singer & Willett, 2003). To provide a baseline against which we can compare more complex models, the unconditional means model (null model or empty model) that examined variation in interval (Level 1) outcomes across students was provided. Based on the unconditional mean model, which does not have any predictor, a Level 2 predictor model was provided to examine the effect of studentlevel predictors, including the self- determination scores of four subscales of SDS and the AIR-S and AIR- E. Given that all dependent variables, except the overall access score, were dichotomous categorical variables, multilevel (two- level) regression analyses for categorical outcomes using Mplus were employed to find the best model fit to explain the relationship between predictors and dependent variables.

Impact of intervention. Interrater reliability was conducted for 84 observations of 42 students on which data analysis regarding the second question targeted using the Cohen’s Kappa statistic. In terms of power analysis, the Optimal Design software for longitudinal and multilevel research was employed (Raudenbush, Spybrook, Liu, & Congdon, 2005). Assuming 30 intervals were set up for each student, it was calculated how many students were needed to achieve 80% power at the level of .05 with the effect size of .35. To achieve 80% power, it was necessary to have at least 40 students, and thus the sample was configured accordingly.

The data regarding the impact of the intervention have a hierarchical structure as follows: Data from each of the 30 observation intervals (Level 1) are nested within each observation time (first and second observations; Level 2), the observation times are nested within each of the 42 students (Level 3), and the students are nested within groups (experimental/control; Level 4). Therefore, a multilevel model having four levels was employed to analyze the data. To address the impact of implementation of curriculum augmentation on access to the general education curriculum (access score), SAS PROC MIXED was employed, which allows for both fixed and random effects of continuous variables in the multilevel model (Singer, 1998; Wang, 1997). To address the impact of the intervention on student variables associated with access in the general education curriculum (academic responses and competing responses), SAS GLIMMIX macro was used, which addresses categorical variables because the academic and competing responses are dichotomous categorical variables (Guo & Zhao, 2000).

Results

Analysis of variance to determine any differences between control and treatment groups on level of disability (learning support needs) or baseline level of self-determination indicated that the two groups did not differ significantly. As an index of the interrater reliability on all observations for both groups, the average Cohen’s Kappa for 27% of the 84 observation sessions was .891, ranging from .792 to .968. Overall reliability rating of agreement for the 27% of reliability sessions was 97.42%.

Relationship Between Self-Determination and Access to the General Education Curriculum

Table 2 provides outcomes for the comparison of fit between the unconditional means model and the Level 2 predictor model. On the access score, there were no significant differences between the unconditional models and the Level 2 predictor model, Deltachi^sup 2^(1, n = 630) = 2.54, p = .11, although SDS Section 4 (Self- Realization) had a negative relationship with the access score. However, there were significant differences between the unconditional means model and the Level 2 predictor model on the academic response, Deltachi^sup 2^(3, n = 630) = 13.12, p < .01. The SDS Section 2 (Self-Regulation) had a positive relationship with academic response. Even though it was slightly smaller than p < .10, the SDS Section 2 was included in the model because the model was still significantly different from the unconditional model. SDS Section 3 (Psychological Empowerment) had a strongly positive relationship with academic response. On the other hand, SDS Section 4 (Self-Realization) had a negative relationship with academic response. Along with academic response, the Level 2 predictor model was significantly different from the unconditional model on the competing response, Deltachi^sup 2^(3, n = 630) = 16.36, p < .01, variable. Unlike the academic response variable, though, SDS Section 2 had a negative relationship and SDS Section 4 had a positive relationship with the competing response. Similar to SDS Section 2, SDS Section 3 had a negative relationship with the competing response variable. In terms of the occurrence of curriculum modifications, a significant difference between the null model and the Level 2 predictor model was found only in the F7 (accommodations) access toggle, Deltachi^sup 2^(3, n = 630) = 16.30, p < .01. SDS Section 2 and SDS Section 3 showed positive relationships with the F7 toggle. On the other hand, the scores of AIR Educator had a negative relationship with the FV toggle. Most curriculum modifications observed were accommodations (315 intervals, 50% of the overall 630 intervals). Among the subtoggles for accommodations, only paraprofessional support was observed and it dominated, with 49.5% (312 intervals) of overall intervals in which the use of a paraprofessional was observed. (There were no subtoggles identified for the remaining three intervals in which the F7 toggle was on.) Curriculum augmentations (F8) toggles were not observed in any single interval throughout all 630 intervals. Curriculum adaptations (F9) were observed in 23.8% (150 intervals) of the overall intervals. In terms of interrater reliability of the entire baseline observations, the average Cohen's Kappa for 42% of the 45 observation data in which two coders recorded data simultaneously was .896, ranging from .792 to .968.

Impact of Self-Determination on Access to the General Education Curriculum

Access to the general education curriculum. Table 3 presents both fixed and random effects of the multilevel model for the access score, student academic response, and competing response. For the access score, the random effects for both intercept and slope (observation) were significant, which means that the initial access scores at the first observation and the changes of such scores from the first observation (Time 1) to the second observation (Time 2) varied substantially across students. However, there were no significant fixed effects of observation time, group, and interaction between time and group. Similar results were found in the academic and competing response. That is, there were no significant fixed effects of observation time, group, and interaction between time and group on both academic response and competing behavior even though the random effects for slope (observation) showed that the changes of such responses from the first observation (Time 1) to the second observation (Time 2) varied substantially across students. The random effects for intercept on both responses demonstrated that there were significant variances across students on the initial academic responses at the first observation, whereas initial competing responses did not vary across students.

Curriculum modifications: Accommodation, augmentation, adaptation. Similar patterns were repeated in both fixed and random effects of multilevel model for the accommodations (F7) and curriculum adaptations (F9; see Table 4). That is, there were no significant fixed effects of observation time, group, and interaction between time and group on the accommodations and curriculum adaptations even though the random effects for slope (observation) revealed that the changes of accommodations and adaptations from the first observation (Time 1) to the second observation (Time 2) varied significantly across students. Although the random effects for intercept, which address the variance of initial observation data across students, were significant on accommodations, those of adaptations were not significant. For curriculum augmentations (F8), the model did not converge. This is most likely due to the small number of curriculum augmentations (F8). Curriculum augmentations were observed in the second observation period in the experimental group for only 16 of 2,520 intervals observed across groups and times (0.6%). In terms of F7 and F9, each toggle was coded in 1,647 intervals (65%) and 432 intervals (17%), respectively, of 2,520 intervals observed across groups and times on 42 students. Paraprofessional support dominated the categories of F7 (accommodations). That is, 97% of intervals (1,601 intervals) of the 1,647 intervals in which an accommodation was coded involved paraprofessional support. However, intervals in which paraprofessionals directly interacted with target students were significantly less frequent, with only 25 of 2,520 intervals in which interaction occurred (0.9%). General education teachers mostly led the class activities (94%; 2,367 of 2,520 intervals) and special education teachers directly interacted with target students for only 3% (83 intervals) of total intervals. All 16 intervals of F8 (augmentations) were related to learning-to-learn strategies. Enhanced content that was the most frequently observed category in F9 (adaptations) was observed in 302 intervals (70%) of the 432 intervals in which adaptation was coded. Given that there were no significant differences regarding the types of curriculum modifications across groups and time, the frequencies of student academic and competing responses observed were investigated across all intervals (N = 2,520) in which no accommodations or curriculum modifications were present and across intervals in which accommodations and curriculum modifications did occur (accommodations = 1,384 intervals, 55%; curriculum augmentations = 16 intervals, 0.6%; curriculum adaptations =185 intervals, 7%) along with F4 or F5. Figure 1 provides descriptive information about differences in academic and competing responses as a function of the presence or absence of and types of accommodations and curriculum modifications. As seen in Figure 1, even though there were few intervals when curriculum augmentations and curriculum adaptations were provided, the frequencies of student academic responses were higher, and those of competing behavior during the interval were lower, than when none were observed.

GAS. The GAS scores for all of the students in the experimental group yielded a mean of a 52.80 (SD = 11.28), which indicated that students attained their goals at a better-than-expected level. Sixty- five percent of the GAS scores (n =13) were at the expected level or higher than the expected level. Two students who achieved the highest level of goal attainment showed 100% of student competence fidelity. Thirty-five percent of the students (n = 7) achieved less than an expected level of goal attainment.

Social validity. Social validity was assessed by participating students and teachers in the experimental group through the student Phase 3 worksheet and teacher feedback forms. Eighteen students completed the worksheet. Thirteen students mentioned that they achieved their learning goal, which included note taking, asking for help, completing work/assignments, improving reading and vocabulary skills, becoming more organized, increasing study time, increasing grades, and so forth. To the question about whether the process removed barriers to learning, students reported that they became more organized, were less stressed, were more confident, increased their participation, tracked work, understood their assignments better, and used better study habits at home. Along with student feedback, teacher feedback was collected regarding the effect of the SDLMI, the self-monitoring sheet and study guides, and any other effects and consideration to be improved. Among the nine teachers in the experimental group, eight teachers completed the feedback forms. Overall, teachers reported that the SDLMI worked well for their students’ learning goals and could be used for other students. Teachers also reported that the SDLMI was effective in generalizing skills from special education into the general education class, establishing habits regarding organization to achieve better control and responsibility, improving class grades, being more aware of their (student) needs, and communicating and collaborating with general education teachers.

Discussion

The premise of this study was that promoting self-determination can be used as an instructional strategy to ensure access to the general education curriculum as well as to address needs regarding transition services. The study provides important, though certainly preliminary, information about this premise and provides more information about student access to the general education curriculum.

Access to the General Education Curriculum

First, this study provided more information about the degree to which students with disabilities are provided the supports they need to succeed in the general education curriculum, and the outcomes of this study mirrored findings from Soukup et al. (2007) and Wehmeyer, Lattin, et al. (2003). Accommodations were most frequently observed and followed, albeit distantly, by curriculum adaptations. Paraprofessional support was the dominate subcategory of accommodations. Curriculum augmentations were rarely observed, and in fact, there were no intervals in the first observation in which a curriculum augmentation was observed. In general, it seems evident that there is a need to ensure that students receive a wider array of accommodations and are provided curriculum adaptations and augmentations.

Relationship Between Self-Determination and Access

First, we would note a caveat with regard to our capacity to draw conclusions about the impact of self-determination on access to the general education curriculum from these data. We operationalized access by calculating a weighted score reflecting whether students were working on tasks linked to on or off grade-level standards and whether they were receiving any accommodations or curriculum modifications to promote their engagement with the learning tasks. Similar to findings from Soukup et al. (2007) and Wehmeyer, Lattin, et al. (2003), however, the use of curriculum modifications was very limited in this study. There were, in essence, no intervals in which curriculum augmentations were observed, very few in which curriculum adaptations were seen, and the almost sole accommodation was the presence of a paraprofessional.

This introduces a serious limitation in examining the degree to which self-determination contributes to or promotes self- determination. The primary way that self-determination is hypothesized to contribute to access to the general education is as a curriculum augmentation that enables students to self-direct learning more effectively and to engage with the curriculum in more meaningful ways (Agran et al., 2006; Knowlton, 1998; Palmer et al., 2004; Wehmeyer et al., 2004; Wehmeyer, Yeager, et al., 2003). Furthermore, once students are more self-directed learners, one could hypothesize that their enhanced self-determination might lead them to advocate more for supports and modifications that would enable them to more effectively progress in the general education curriculum, although it should be stated that curriculum modifications will still be a primary responsibility of the teacher and not the student. As a result, however, of the lack of curriculum modifications and accommodations, the access score generated by observations consists almost exclusively of the degree to which students were working on tasks linked to on- or off-grade standards. Decisions pertaining to what standards to address are the domain of teachers and a function of lesson and unit planning and might only be influenced by student self-determination very indirectly through teacher’s perceptions and understandings of student capacity. As such, findings from these studies should be interpreted within the context of the current reality of limited curriculum modifications and accommodations.

Given this caveat, this study found that self-determination did not have a statistically significant relationship with overall student access to the general education curriculum. That is, as noted, self-determination status was not a good predictor of student access to the general education curriculum because overall access scores were highly dependent on externally determined factors pertaining to on/ off-grade standard. This issue was mirrored in the findings regarding the second question. That is, there were no significant differences in the access scores between experimental and control groups across times. There were, however, promising findings when one moved beyond simply the overall access scores. The data analysis indicated that self-determination was a strong predictor of increased student academic response (e.g., a proxy measure of student engagement activities) and decreased competing response (e.g., problems or other behaviors that disrupt student engagement). In particular, the self-regulation score from the SDS had a strong positive relationship to the academic response variable and a negative relationship with the competing response variable. Unexpectedly though, the self-realization section of the same measure had relationships in the opposite directions. This section contributes only a small portion of the variance to total self- determination scores and did not affect any other outcomes. Given the small sample size, it seems most likely that this is accounted for by problems with measurement and not any truly negative relationship between self-awareness or self-knowledge and student engagement.

Impact of Interventions to Promote Self-Determination on Access

With regard to the impact of the intervention, students who were involved with the SDLMI were able to achieve at, in general, a higher than expected rate their self-set goals linked to the general education curriculum. This is a finding that has been replicated several times (Agran et al., 2000, 2006; Palmer et al., 2004; Wehmeyer et al., 2000) and confirms the presumption that promoting self-determination can, in fact, promote access to the general education. Second, although there were no between-group, time, or group differences on access, academic responses, or competing responses indicators, descriptive information about the percentage of intervals on academic and competing behavior as a function of the observation of curriculum modifications and accommodations suggested that more academic responses and fewer competing responses occurred during intervals in which curriculum augmentations and adaptations were provided than during the intervals that did not have any curriculum modifications. Limitations

The findings mentioned should be interpreted with consideration for the following limitations. First, the intervention was implemented in a relatively short period of time and outcomes may reflect the fact that students did not have the opportunity to become more familiar with and integrate the strategies taught through the model into their learning activities. Only longerterm studies will be able to fairly assess the impact of self- determination on access. Second, despite all efforts to observe participating students in similar class activities and contexts across the first and second observation period, it was not possible to control for all differences between these observations. Third, participants in this study were primarily, although not exclusively, students with learning disabilities. Considering that students with more severe disabilities may need more intensive support or curriculum modifications, one must be cautious in generalizing the findings to other populations of students with disabilities.

Conclusions

All in all, given the lack of curriculum modifications and the relatively limited time in which the intervention to promote self- determination was provided, the findings can be interpreted as providing admittedly limited, but generally promising, evidence of the relationship between and impact of self-determination on access to the general education curriculum. Student self-determination and self-regulation positively predicted student engagement and negatively correlated with competing behavior. As important, the study confirmed that students with disabilities can achieve educational goals linked to the general education curriculum through instruction to promote self-determination and student-directed learning. Finally, the fact that students are provided very few curriculum modifications and only stock accommodations, primarily in the form of the presence of a paraprofessional, confirmed previous research and adds urgency to the need to rectify this situation. It is evident, as such, that there are both multiple benefits from and reasons to expand the degree to which students with disabilities receive instruction to promote self-determination and student- directed learning using methods and strategies such as the SDLMI across disability categories along with students without disabilities.

It is also obvious, though, that training and support for teachers is needed to increase the possibility that strategies to promote self-determination and studentdirected learning will, in fact, be implemented. As revealed in this study, student access was affected more by instructional decisions or actions of the teacher than anything the student did, and the curriculum modifications and accommodations provided were both few and limited in scope and potential impact. Of importance, given that in these studies general education teachers were primarily responsible for academic instruction, it is critical to provide general and special education teachers with extensive training on curriculum modifications and classroom accommodations. This is equally true for paraprofessionals, who provide support for students as an accommodation. In addition, teachers need to consider and implement a wider array of accommodations that are linked to student access.

References

Agran, M., Blanchard, C., & Wehmeyer, M. L. (2000). Promoting transition goals and self-determination through student-directed learning: The Self-Determined Learning Model of Instruction. Education and Training in Mental Retardation and Developmental Disabilities, 35, 351-364.

Agran, M., Cavin, M., Wehmeyer, M. L., & Palmer, S. (2006). Participation of students with severe disabilities in the general curriculum: The effects of the Self-Determined Learning Model of Instruction. Research and Practice for Persons with Severe Disabilities, 31, 230-241.

Agran, M., Sinclair, T., Alper, S., Cavin, M., Wehmeyer, M., & Hughes, C. (2005). Using self-monitoring to increase following- direction skills of students with moderate to severe disabilities in general education. Education and Training in Developmental Disabilities, 40, 3-13.

Bashinski, S. M., & Wehmeyer, M. L. (2002). Access CISSAR (Version 1.0) [Computer software]. Lawrence: University of Kansas.

Blalock, G., & Patton, J. R. (1996). Transition and students with learning disabilities: Creating sound futures. Journal of Learning Disabilities, 29,7-16.

Bryan, T., Burstein, K., & Bryan, J. (2001). Students with learning disabilities: Homework problems and promising practices. Educational Psychologist, 36, 167-180.

Carta, J. J., Greenwood, R., Schulte, D., Arreaga-Mayer, C., & Terry, B. (1988). Code Instructional Structure and Student Academic Response: Mainstream Version (MS-CISSAR). Kansas City: Juniper Gardens Children’s Project, Bureau of Child Research, University of Kansas.

Copeland, S. R., Hughes, C., Agran, M., Wehmeyer, M. L., & Fowler, S. E. (2002). An intervention package to support high school students with mental retardation in general education classrooms. American Journal of Mental Retardation, 107, 32-45.

Croy, C. D., & Novins, D. K. (2005). Methods for addressing missing data in psychiatric and developmental research. Journal of the American Academy of Child and Adolescent Psychiatry, 44, 1230- 1240.

Deshler, D., Ellis, E. S., & Lenz, R. K. (1996). Teaching adolescents with learning disabilities (2nd ed.). Denver, CO: Love Publishing.

Deshler, D. D., Schumaker, J., Bulgren, J., Lenz, K., Jantzen, J., Adams, G., et al. (2001). Making learning easier: Connecting new knowledge to things students already know. Teaching Exceptional Children, 33(4), 82-85.

Fixen, D. L., Naoom, S. F., Blase, K. A., Friedman, R. M., & Wallace, F. (2005). Implementation research: A synthesis of the literature. Tampa: University of South Florida.

Goldberg, M., & Cornell, D. (1998). The influence of intrinsic motivation and self-concept on academic achievement in 2nd/3rd grade students. Journal for the Ed