ERIC ED351839: Results of a Panel Longitudinal Study with Systematic Replication: Graduates of Preschool Special Education Programs in Washington and New Mexico.

26 Pages·1991·0.3 MB·English

by  ERIC

DOCUMENT RESUME
ED 351 839
EC 301 666
AUTHOR
Stile, Stephen W.; And Others
TITLE
Results of a Panel Longitudinal Study with Systematic
Replication: Graduates of Preschool Special Education
Programs in Washington and New Mexico.
SPONS AGENCY
Department of Health and Human Services, Washington,
D.C.
PUB DATE
91
CONTRACT
90DD016803
NOTE
30p.; The document was supported by a grant from the
New Mexico Special Education Unit and by the New
Mexico University Affiliated Program.
PUB TYPE
Reports
Research/Technical (143)
EDRS PRICE
MF01/PCO2 Plus Postage.
DESCRIPTORS
*Disabilities; Educational Needs; Elementary
Education; Longitudinal Studies; *Mild Disabilities;
*Outcomes of Education; Preschool Children;
*Preschool Education; Prevention; Research
Methodology; *Special Education; *Student
Placement
IDENTIFIERS
New Mexico; Washington
ABSTRACT
This panel longitudinal study, originally conducted
in Washington and systematically replicated in New Mexico,
investigated the long-term educational placements of students served
in preschool special education programs. Subjects were the 4338
children who graduated from preschool programs for children with
disabilities during two 5-year periods. Findings indicate that early
childhood programs for children with disabilities rarely reduce the
need for future special education services and then only for students
with mild disabilities. About 15 percent of students were able to
remain in regular education fox up to 5 years without special
education services. The paper reviews past longitudinal studies,
discusses methodological issues involved in longitudinal studies, and
discusses implications of the study's findings for policymakers and
practitioners. (Contains 12 references.) (JDD)
***********************************************************************
Reproductions supplied by EDRS are the best that can be made
from the original document.
**********************************************************************
U.S. DEPARTMENT OF EDUCATION
Office of Ecucational Research and Improvement
EDUCATIONAL RESOURCES INFORMATION
CENTER (ERIC)
This document has been reproduced as
received from the person or organization
originating it
improve
C Minor changes have been made to
reproduction Quality
Panel Longitudinal Study
docu-
Points of view or opinions stated in this
official
ment do not necessarily represent
OEM position or policy
1
Results of a Panel Longitudinal Study with
Systematic Replication : Graduates of
Preschool Special Education Programs in
Washington and New Mexico
Stephen W. Stile
New Mexico University Affiliated Program
Brenda Hudson
Emporia State University
Michael Ames
Pat Kelly
June LeCrone
New Mexico State University
Eugene Edgar
University of Washington
"PERMISSION TO REPRODUCE THIS
MATERjAL HAS BEEN GRANTED BY
TO THE EDUCATIONAL RESOURCES
INFORMATION CENTER (ERIC)"
Running head: RESULTS OF A PANEL LONGITUDINAL STUDY
This study was supported by a discretionary grant from the New Mexico Special Education Unit
and by the New Mexico University Affiliated Program (UAP) under DDHS Grant No.
90DD016803.
2
BEST COPY AVAILABLE
Panel Longitudinal Study
2
Abstract
in Washington (N=1659) was
A panel longitudinal study conducted
Subjects were the
(N=2679).
systematically replicated in New Mexico
children with disabilities
4338 children who graduated from programs for
Data were collected by trained local
during two five-year periods.
instruments/procedures validated in the
education agency personnel using
Initial placements and stability of
earlier Washington studies.
Recommendations based upon
placements were analyzed for both states.
findings are made for policy-makers and practitioners.
Pane] Longitudinal Study
3
with Systematic Replication:
Results of a Panel Longitudinal Study
Programs in Washington and New
Graduates of Preschool Special Education
Mexico
disabilities have become
Preschool programs for children with
These programs have developed
accepted practice in the United States.
professional opinion that early intervention
in response to value-based
the impact of disabilities on future
can substantially reduce
future intervention services;
acquisition of skills; reduce need for
handicapping conditions and provide
alleviate development of secondary
Legislators have been lobbied to support
needed support to families.
the premise they will
the additional costs of preschool programs on
education by preschool graduates and,
result in reduced need for special
special education for school-aged
therefore, reduced overall costs of
children.
considerable research effort,
Despite twenty years of programs and
outcomes of early childhood programs
there are few definitive data on
how
their families and much debate on
for children with disabilities and
&
data (Shonkoff, Hauser-Cram, Krauss,
best to interpret the available
As Guralnick has explained it
1991).
Upshur, 1988, Guralnick, 1989;
(1989):
and decision-makers to assess prior
Studies available for professionals
generation of
99-457 were part of a first
to the passage of P.L.
form an
Understandably, these studies did not
efficacy research.
of this period were struggling to
exemplary data base, as the programs
in a context of limited resources
balance intervention and evaluation
(p.2)
and experiences.
4
Panel Longitudinal Study
4
policy decisions
empirical base is limited,
Although the current
foreseeable future, the
at least for the
have been made that ensure,
conducted in an
The current studies were
continuation of such programs.
issues and to discuss
additional light on the
attempt to shed
of the data.
policy-maker and practitioner use
Past Longitudinal Studies
(Bricker, 1989) and
difficult to conduct
Longitudinal studies are
However, many questions
difficult analysis issues.
often pose even more
answered through such
advocates can only be
raised by preschool program
measurement of
(1988) have put it, "the
As Shonkoff, et al.
studies.
the traditional focus
effectiveness must move beyond
early intervention
at long-term impacts
effects and look increasingly
on short-term
(p.89)."
study which explained
of Colorado conducted a
In 1983, the State
education preschool
graduates of special
educational placement of
This study found that
Maddox, 1984).
(Edgar, McNulty, Gaetz, &
programs
education with no special
placed in regular
31.4% of graduates were
support
regular education with
37.1% were placed in
education services,
placed in self-contained
education and 31.4% were
services from special
replication was
Based on this study, a
special education programs.
results
determine if similar
State in order to
conducted in Washington
major
Fischer, 1988). Several
(Edgar, Heggelund, &
would be obtained
Colorado.
data as compared to
in the Washington
discrepancies were found
without
in regular education
students were placed
For example, fewer
31.4% in Colorado).
(16% in Washington,
special education support
Panel Longitudinal Study
5
data on type of
Since the Colorado study did not collect
could not be considered
disability, the Colorado and Washington data
data on
In addition, the Washington study collected
equivalent.
found a high rate of stability
subsequent placements after preschool and
districts, approximately 90% of
in initial placement. In the Washington
of placement as their initial
the graduates remained in the same type
from preschool.
placement up to four years after graduation
reported by Edgar,
A continuation of the Washington study was
initial Washington
Heggelund, and Fischer (1988) which confirmed the
While
of placement.
data both as to initial placement and stability
the post-presWool
these data provided some additional insight to
with the original
educational placements, the overall discrepancy
For instance, what accounts
Colorado data raised some serious issues.
placement rates in regular
for the rather large discrepancy in initial
Could this be solely due to
education between Colorado and Washington?
there substantial
the types of students in the two studies, or were
Also, stability data
services?
differences in the quality of preschool
entire population rather than for
were only calculated for the
While these data are useful for practitioners
individual students.
to remain in a given
(e.g., "how many preschool graduates can we expect
status of individual
placement"), the data do not address the placement
remain in the same place-
students (e.g., "does the same individual
teachers and parents.
This latter question is important for
ment")?
Methodological Issues
(Borg, & Gall, 3989).
Longitudinal studies can be of several types
1
Panel Longitudinal Study
6
individuals (e.g., all preschool
Panel studies select one group of
specific year) and follow
graduates from a specific district and of a
data at several points in
those same individuals over time, collecting
of individual change as
This method allows for the intepretation
time.
In contrast, cohort studies sample from
well as that of the population.
but may not collect data on the same
a set population over time
This method allows for
individuals at each point in time.
within the
generalization to the population but not to individuals
population.
larger
With any study, a question as to generalization to a
One method of addressing broader external
population must be addressed.
of a study using
generalization is to conduct systematic replication
noted in replications,
different populations. If similar findings are
population (i.e., all preshool
believability is increased for the larger
The current study is a series of panel
students with disabilities).
by using multiple
longitudinal studies replicated both within state
cohorts) and by conducting the study in
panels (i.e., several graduating
Mexico).
two states (Washington and New
Uses of the Results
"what are the
The basic question addressed by this study was:
served in preschool special
long-term educational placements of students
These data may be used to address the policy
education programs?"
reduce need for future
"do preschool special education programs
issue:
In addition, the data may be used to
special education services?"
about individual students.
assist practitioners who must make decisions
Panel Longitudinal Studies
7
Procedures
earlier studies in
Procedures used in this study were developed in
1984).
Colorado and Washington (Edgar, et al.,
Subjects
Twenty (20) school districts were selected using a
Washington.
long-term preschool
convenience sample of districts (districts with
All
the study).
special education programs willing to participate in
Gender and
preschool graduates from 1984 to 1988 were included.
ethnicity data on these students are found in Table
1.
Insert Table 1 about here
Seventy Four of 88 (84%) school districts in the state
New Mexico.
The nonparticipating districts were invited
participated in the study.
Students from graduating panels from 1987
but chose not to be included.
Basic demographic data on these students are
to 1991 were included.
also found in Table 1.
Instrumentation.
used in both states.
The same demographics and placement form was
its use and paid to
Local school district personnel were trained in
Students are found eligible for
collect data on a yearly basis.
developmentally delayed in
preschool services based on classification as
designate the
Therefore, professional judgment was used to
both states
This procedure undoubtedly
specific type of disability of each student.
to specific type of
resulted in some misclassification of students as
Panel Longitudinal Study
8
Overall, however, we believe disability classifications are
disability.
generally equivalent across states.
In New Mexico, special education placement options are listed as
regular education with itinerant special education services; B:
A:
regular education with special education resource room support; C:
special education with some integration in regular education and D:
In Washington, the three original
self-contained special education.
placement categories were, regular education with no special education
support; regular education with special education support and
The New Mexico researchers merged the
self-contained special education.
A and B and C and D placements and added the regular education with no
Thus, for purposes of analysis, both
special education support option.
states had the same three placement categories.
Data Collection.
Data were collected annually by the local data collecters.
Students who had moved out of the district or who died were placed in
"other" category.
the
.
Data Analyses
Mexico) who
All children from one state (Washington or New
graduated from preschool in the same year were placed in a panel
For Washington there were five panels
(defined by year of graduation).
For New Mexico there were also five
1987 and 1988).
1986,
1985,
(1984,
Within each panel, the data
1991).
1989, 1990 and
panels
1988,
(1987,
(i.e., mild retardation,
were analyzed by specific disability
etc.).
communication disorders, multihandicapped, vision impairments,
where similar
Collapsed groups were formed for these disabilities
9
Panel Longitudinal Study
9
mildly
BD, LD formed a group identified as
patterns were noted (i.e., CD,
collapsed groups as
All analyses were then performed on
handicapped).
panel.
well as the total group for each
Results
Initial Placements
of
Figure 1 describes initial placements
Washington.
As shown, initial placement in
Washington graduates during 1984-1988.
(Level I) ranged from 87. to 247..
regular education with no support
special education support placements
Initial regular education with
Initial placements in special
(Level II) ranged from 16% to 307..
Initial
III) ranged from 45% to 63%.
education classrooms (Level
regular
in Washington were:
placements for the five-year period
(21%)
(17%), regular education with support
education with no support
(52%).
and special education
Insert Figure 1 about here
considered, the percentages of
When types of disability were
These data are
varied dramatically.
initial placements by level
for 14% of the
Level I was the placement option
presented in Figure 2.
with mild
retardation, 24% of the students
students with mild mental
Of
handicapping conditions.
disabilities and 8% of those with severe
29%
Level II, 12% had mild retardation,
students initially placed in
handicaps.
conditions and 16% had severe
exhibited mild handicapping
mild
64% of the students with
Level III was the choice for
10