The Office of Special Education Programs (OSEP) of the U.S. Department of Education has commissioned a design for the Special Education Elementary Longitudinal Study (SEELS), to be developed by SRI International, with support from the Research Triangle Institute.

As part of a comprehensive OSEP program of longitudinal research, SEELS will involve a large, nationally representative sample of students in special education who are ages 8 through 12 when the sample is selected (Fall 1999). Students will be selected randomly from rosters of students in special education; rosters will be provided by local education agencies (and special schools) that are selected for and agree to participate in the study. Statistical summaries generated from SEELS will generalize to special education students nationally as a group, to each of the 13 federal special education disability categories, and to each single-year age cohort. Information about students will be collected repeatedly as they transition from elementary to middle school and from middle to high school. Thus, SEELS will provide the first national picture of the experiences and outcomes of students in special education as they move through these crucial years of their educational careers.

To enable SEELS to be maximally useful to the wide range of audiences that will be interested in it, OSEP is relying on the collective expertise of a task force whose members represent many of those audiences: parents; teachers, principals, and related service practitioners; researchers; local, state, and federal policy makers; and advocacy organizations. The task force is being asked to identify the range of information needs that could potentially be met by SEELS, translate those information needs into research questions, and help to envision a conceptual framework to guide the study in addressing those questions. Because the range of information needs the task force may articulate is likely to exceed the capacity of any single study to meet them, the task force also is being asked to suggest priorities among research questions.

The purpose of this paper is to provide background information to assist the task force in identifying the critical research questions that SEELS must address in order to be successful and in developing a conceptual framework that encompasses those research questions. The paper begins with an overview of the SEELS design process and current plans for the SEELS sample, timeline, and instrumentation. The remainder of the document is devoted to the SEELS conceptual framework and is intended to stimulate discussion about the relative importance of different conceptual issues, key study questions, and likely interrelationships among factors.

The design process will include developing the SEELS timeline, sample specifications, instrumentation, and data collection and analysis plans, as well as recruiting a sample of local education agencies (LEAs) and special schools from which students will be selected for the sample. The table below documents key study design activities and their associated completion dates.

Table 1
SEELS DESIGN ACTIVITIES AND SCHEDULE
Solicit input from research advisory panel	10/23 and following
Convene task force	11/23/98
Task force meeting summary	11/27/98
Draft project time line and data collection plan	12/24/98
Draft project sampling plan	12/24/98
Draft project data analysis plan	12/24/98
Final project time line and data collection plan	1/25/99
Final project sampling plan	1/25/99
Final project data analysis plan	1/25/99
Instrument analysis	2/8/99
First interim sampling report	2/24/99
Second interim sampling report	3/24/99
Conceptual outlines and item specifications for new instruments	3/24/99
Final sampling report	6/24/99
Drafts of instruments	5/7/99
Revise and pilot test instruments	6/24/99
Report of pilot test	7/8/99
Final sampling fractions	7/26/99
Final instruments and OMB clearance forms	8/24/99

The SEELS student sample, to be selected in the Fall of 1999, will be a nationally representative sample that will generalize to: students in special education ages 8 to 12 nationally as a group as they age, each of the five single-year age cohorts within the full sample, and students in each of the 13 federally defined special education disability categories in the full sample. Additional characteristics of the sample include the following:

• In order to generate statistical summaries with adequate precision in the study’s out years SEELS will require a sample of approximately 1,300 students in each of the 13 federal special education disability categories (with the exception of deaf/blind) at the time of the last data collection. Assuming that is the 2003-2004 school year (although the duration of the study has not yet been fully determined), the total sample is estimated to be approximately 15,800 students at the outset (the 1999-2000 school year).
• To generate the needed student sample, SEELS will need to select students from between 400 and 450 LEAs. These LEAs are likely to be selected via a stratified sampling strategy to ensure representation of region, district size, and student poverty. The primary LEA sample could be augmented by a sample of students from special schools. The LEA sample will be drawn in January 1999.
• Students will be randomly selected from district rosters according to specified sampling fractions for each age group and disability category. SEELS students will attend a large and diverse group of schools, estimated to number more than 2,000.

The sheer size and wide geographic distribution of the student sample has important implications for the kinds and frequency of data collection that can be afforded. The wide range in age and disability characteristics of students also pose significant challenges in ensuring that measures and instrumentation are appropriate to that student diversity.

The specific data collection schedule for SEELS is still under discussion. However, it is certain that SEELS will be longitudinal in nature so that it can track changes in student statuses and outcomes. There are three general options available in sequencing the main data collection activities:

• Fixed waves of data collection — Fixed-time data collection involves collecting information on all sample members at specified calendar periods, regardless of the age or educational stage of the youth. For example, data could be collected during the spring of years 1, 3, and 5; or in the fall and spring of selected years, etc.
• Data collection tied to specified milestones in a student’s experience — This approach calls for information to be collected about sample members when they experience specified events (e.g., at a transition from elementary to middle school, at a particular birthday). This approach has the advantage of illuminating key ages or milestones by collecting data on each child at that key point, regardless of when in the course of the study that point occurs.
• A combination of the two approaches — It is also possible to combine these approaches so that, for example, information is collected from all students at two or more specified time periods (e.g., in the first school year of the study and again in the last school year of the study), with intermediate data collection focusing on specified milestone events (e.g., the transition from elementary to middle and middle to high school). This would permit a wide-ranging description of the full sample at two or more time points (yielding trend data) as well as permitting an in-depth look at the milestone events of interest for all students who experience them in the course of the study.

Decisions regarding the study timeline depend in part on the specification of study’s conceptual framework and it’s associated research questions. So, recommendations of the task force regarding research questions will influence those design decisions. It is currently anticipated that the first SEELS data collection efforts will take place in the spring of 2000.

SEELS is intended to be far-reaching in terms of conceptual domains that it will address. Thus, SEELS will collect data from multiple data sources (e.g., parents, teachers, principals) using several data collection methods (e.g., telephone interviews, collection or abstraction of school records, mail surveys, teacher assessments or reports of student performance, direct assessments). As the task force considers the SEELS research agenda, it is important to consider the validity and logistics of obtaining data from different sources and in different ways. Table 2 illustrates potential information sources for different conceptual domains. Table 3 indicates design-related questions to be considered by the task force.

X Possible primary respondent for items relating to domain.

BACK

NEXT