Consequences of LD Identification Changes
In this final section, consequences of LD identification changes are discussed including the questions established by the symposium organizers. Data are discussed where possible.
1. Are psychologists, diagnosticians, and teachers ready for changes in LD identification?
Generally, a great deal of change will be necessary in the roles of psychologists and diagnosticians. Less change in teachers' roles has occurred in SEAs and LEAs implementing RTI in LD identification. Changes in instructions will be prompted by the availability of graphs, direct measures that can be used frequently, comparisons of progress to goals, and expectations for instructional changes when progress does not meet goals. Changes in LD identification will enhance progress toward improving general and special education interventions.
Continuing education for diagnosticians, psychologists, special education directors, and others will be necessary for effective implementation of RTI in LD identification. The general parameters of that continuing education have been discussed and validated (Reschly & Grimes, 1991). Continuing education in critical areas and implementation of RTI will be markedly enhanced by "manualized" supports for practitioners that prompt, guide, and support implementation of functional assessment, standard protocol treatments, progress monitoring, effective instructional principles, and effective behavior change principles. Manualized or protocol-based assessment and treatments exist today in medicine, psychology, and education. One example is the development of standardized, norm-referenced tests. The manualized supports are increasingly available in curriculum based measurement and other areas (Good & Kaminski, 1996, 2003; Kratochwill & Bergan, 1990). RTI implementation will be enhanced by manualized assessments and treatment protocols.
Continuing education is a must for all of us. I like this perspective. "I have saved my course notes from graduate school. With the exception of statistics, almost everything I was taught no longer is true." Bardon, (1981, p. 202). My corollary is that I did not do that well in statistics. The continuing education needs are considerable and should not be underestimated. We also should not underestimate, however, the attractiveness of adopting practices that make a documented difference in the achievement, behavior, and emotional regulation of children and youth.
2. Do professional roles change as a result of RTI-based LD identification?
I have a data-based answer that applies to school psychologists and diagnosticians. My experience is that relatively little change in teachers' roles is necessary, other than increased emphasis on scientifically-based instruction.
Changes in school psychology practices are summarized in Figures 2, 3, 4, and 5. The comparisons are between Iowa school psychologists from the Heartland Area Education Agency (N=65) and a nationally representative sample of school psychologists (N=1054). The Heartland response rate was 100%. The national response rate was 74%. Changes in roles and assessment are quite dramatic. The Iowa sample was comparable or perhaps a bit better on dimensions of job satisfaction.
Figure 2: Current Roles of School Psychologists in the U.S. and Iowa: Hours per Week
Figure 3: School Psychology Assessment: Times per Month
Figure 4: Assessment of Educational Skills: Times per Month: U.S. and Iowa
Figure 5: School Psycholgoists' job Satisfaction in the U.S. and Iowa
Roles changed in terms of the amount of time devoted to assessment for special education eligibility determination (less in RTI) (see Figure 2), amount of time devoted to direct intervention and problem solving consultation (more in RTI), kinds of assessment (see Figure 3), and kinds of educational assessment (see Figure 4). School psychologists in RTI focus on assessment that is relevant to interventions with markedly less assessment using ability tests, visual-motor measures, and projectives. Behavior observation by school psychologists using systematic observation methods occurs much more frequently in RTI. Assessment of academic skills is dominated by curriculum-based measurement (CBM) and curriculum-based evaluation (CBE) (Howell & Nolet, 2000). Both CBM and CBE are directly related to academic interventions. Systematic behavior observations are devoted to assessing academic learning opportunities and students' enabling/disabling behaviors related to learning and positive classroom behaviors. Direct measures are used to define problems, assess degree of need, determine progress toward goals, determine changes in programs, and evaluate outcomes. The dramatic assessment differences in traditional and RTI illustrate effectively the relative focus on determining LD eligibility through cognitive processes and severe discrepancy on one hand and functional, intervention-related measures useful for identification and interventions on the other.
The history and current data on RTI implementation in IA, MN, IL, and SC clearly indicate that identification can be accomplished with assessment procedures that are related to interventions. The outcomes of those interventions, in turn, are the basis for subsequent decisions about eligibility, programming, and special education continuation/exit.
3. Will LD prevalence vary markedly if RTI is implemented?
LD prevalence already varies by a factor of 3 times in the U.S. (KY =2.76% vs. RI at 9.46%) (See Figure 6). Moreover, LD prevalence within states varies markedly now. It is unlikely that these differences will become greater. LD prevalence differences have been difficult to explain. No satisfactory explanations have appeared in the literature. For example, in a recent analysis, we determined the number of points required by the states with explicit requirements (N=30) in order for a student with an IQ of 100 to meet the severe discrepancy requirement. The range was 15 to 30. The correlation between the number of points required and LD prevalence should be negative (requiring larger discrepancies should produce low prevalence). In fact, the correlation between the stringency of states' severe discrepancy requirements and prevalence was not statistically significant.
LEAs and SEAs implementing RTI do not have large increases or decreases in the proportion of students in special education or in LD. There is no reason to believe that RTI will markedly increase or decrease LD prevalence.
4. Will differences be exacerbated between children and youth with LD in SEAs and LEAs?
First, different groups of students with LD vary differ significantly today. It is unlikely that these differences will become larger with RTI than they are now. Enormous variations in children and youth with LD exist across LEAs within the same state (Gottlieb Alter, & Gottlieb, 1999; Gottlieb, Alter, Gottlieb, and Wishner, 1994; Peterson & Shinn, 2002). Some studies suggest that these differences between LD groups are closely related to LEA achievement levels. Gottlieb et al. reported that the LD students in a suburban New York City district would be slightly above average achievers in New York City and, therefore, unlikely to be referred to special education. Peterson and Shinn reached a similar conclusion. Findings from studies in southern California also suggest that levels of classroom and local school attendance center achievement significantly affect the characteristics of children and youth identified as LD (Bocian, Beebe, MacMillan, & Gresham, 1999; Gresham, MacMillan, & Bocian, 1998; MacMillan, Gresham, & Bocian, 1998). Traditional LD is in part a function of where you are and the performance levels of peers.
5. Who is responsible to ensure that the procedures are implemented fully and with fidelity in special education or general education?
In places with RTI LD identification procedures in place, the personnel responsible for implementing procedures in the traditional system are responsible for implementing procedures in the RTI system. No additional or new personnel are required. RTI has been implemented without additional resources or new kinds of personnel. Professional roles and activities do change as noted previously.
6. Is it impossible to implement RTI as an LD identification method with good fidelity?
This is a challenge. Others have suggested significant barriers to implementation (Fuchs, Mock, Morgan, & Young, 2003). Implementation fidelity, however, is not a problem unique to RTI.
Significant continuing education is necessary to do RTI well (Upah & Tilly, 2002). Implementation fidelity is a significant problem; however, before dismissing RTI because it is difficult to implement, it is essential to consider the considerable benefits of closer links between LD identification and treatment. Moreover, implementation problems are rife in current LD identification policies and practices.
Significant proportions of students diagnosed as LD and placed in special education currently do not meet the existing LD classification criteria established by the states. In several studies as many as 50% of all students in LD programs do NOT meet the existing classification criteria (Bocian et al., 1999; Gottlieb et al. 1994; Kavale & Reece, 1992; MacMillan et al., 1998; Norman & Zigmond, 1980; Shepard, 1980; Shepard, 1983, Shepard & Smith, 1983; Shepard, Smith, & Vojir, 1983; Smith, Coleman, Dokecki, & Davis, 1977). Consistency of decision making from school to school is an enormous problem in the current system that has existed at least since the mid-1970s. The current system has failed on this criterion.
The question is, Will children be served more effectively if we devote resources to ensuring that practitioners apply more rigorously current LD criteria by focusing on better compliance with discrepancy criteria and processing disorders? Or does it make more sense in terms of outcomes of children and youth to focus on overcoming RTI implementation problems so that closer ties are established between LD identification and LD treatment. Finding the "right" kids is not the primary special education challenge. Doing something with them that changes academic and behavioral trajectories, increases competencies, and expands opportunities is the real challenge. RTI is more closely related to improving special education.
7. How will LD identification be accomplished at the middle and high school levels where a significant proportion of students with LD are identified now?
First, some of the identification of students with LD at the high school level is questionable in that it appears to be motivated by gaining more time on the SAT and ACT college aptitude tests rather than a need or desire for specially designed instruction. Many of the parents and students seeking accommodations on high stakes tests do NOT want special education IEPs and special education programs, per se. These students do not need special education because they are earning good grades in school subjects and do not need specially designed instruction. They want accommodations on tests, usually more time, not special education.
These students are by definition not eligible for LD under the legal auspices of IDEA or state special education statutes. Special education eligibility requires documented adverse impact of the disability on educational performance and need for special education (IDEA, 1997, 1999). These students may be eligible for Section 504 accommodations and should be evaluated in relation to those requirements.
The standard RTI criteria are relevant to middle and high school students. The RTI criteria, low performance related to peers in (a) relevant domain(s), insufficient response to high quality interventions, documented adverse impact on classroom performance, and documented need for special education, are equally applicable at the middle and high school levels. It is important to note, however, that documentation of both adverse impact and need for special education are crucial to LD decision making regardless of the system used.
8. How will LD be differentiated from other education disabilities if a cognitive measure is not used as part of the assessment?
For some, this distinction is not important (President's Commission on Excellence in Special Education, 2002, see p. 20; Tilly, Reschly, & Grimes, 1999). For many others the distinction is very important. I suspect that most states will retain LD as a diagnostic category and will not adopt a non-categorical approach to identification of high incidence disabilities.
The principal issue for people strongly committed to the LD diagnostic construct is making sure that a clear distinction exists between LD and mental retardation (MR). The LD researcher group suggested screening for adaptive behavior and intellectual ability to rule out MR (Bradley et al., 2002). Simple and time efficient methods exist to accomplish that distinction. The overwhelming majority of states are likely to retain the traditional high incidence categories such as LD, MR, and ED. Retention of traditional high incidence categories is not a barrier to implementation of RTI.
9. Will LD still mean average ability?
The answer depends at least in part on what is regarded as average ability. States typically do not operationalize what is meant by average other than to require that ability must be above the level of MR (Reschly et al, 2003). Studies over the last 25 years of students in LD programs leads to at least some skepticism about the average ability assumption. Early studies suggested that the average IQ of students with LD was in the mid to high 80s, suggesting that over half were below average if average is defined as by the IQ range of 90-110 (Norman & Zigmond, 1980; Smith et al, 1977). More recent studies yield similar results. The average IQ of students in LD programs is about 2/3 SD below the population average. The National Longitudinal Transition Study (U.S. Department of Education, 1992) reported a mean of 87.1. Smith et al. reported a mean of 87.1. The highest reported IQ mean was 96.5 (Kavale & Reece, 1992); however the mean IQ of the student population in that state was 112. The assumption of average ability has not been implemented effectively in the traditional system where half or more of all students in LD have IQs below the average IQ range of 90-110. There is no reason to expect the average IQs of children and youth identified with LD will be even more below average with implementation of RTI.
10. Will LD still mean unexpected low achievement?
This core concept of LD can be retained by considering non-IQ indices to establish expectations. Other approaches to establish expectations for achievement include (a) Unexpected low learning rate even with high quality interventions, (b) Unexpected low performance in a specific academic skill or subject area compared to performance in other areas, (c) Unexpected low performance in relation to age/grade level norms, and (d) Unexpected low performance in relation to other possible explanations such as MR, ED, sensory impairment, and inadequate opportunity to learn.
11. Is it premature to move away from the severe discrepancy approach to identification of LD?
This question can be best answered with a question, Is it ever premature to move away from a method that causes harm? At the very least, we must move away from the use of severe discrepancy in the first 3 grades. Severe discrepancy causes harm at those ages.
I think it is premature to mandate immediate elimination of severe discrepancy in every SEA and LEA. It is time to encourage and support the development and implementation of alternatives to severe discrepancy and processing disorder conceptions of LD. Models exist (Reschly, Tilly, & Grimes, 1999). Some places are using alternative methods without any identifiable harm and tangible benefits to children and youth with achievement problems. There is no reason to believe that RTI methods cause harm or deprive students of legal protections and interventions that they need.
12. Will RTI as a method of LD identification strengthen or diminish services to students with LD, strengthen or diminish LD as a viable diagnostic construct, strengthen or diminish the LD infrastructure of parent and professionals organizations, and strengthen or diminish special education?
There will be different answers to these questions across the participants at this symposium. My answer is that changing to RTI as the LD identification method is part of addressing the fundamental problem in special education and LD policies and practices, the outcomes problem. Current LD identification methods are not supported by scientific evidence and they are not linked to the interventions needed by children and youth. LD as a diagnostic entity, LD parent and professional associations, and special education in general will be strengthened if there are links between identification and treatment and if treatment programs are more effective. The greatest threat to special education and LD is not RTI; rather, it is undocumented outcomes and persistence of practices unrelated to effective interventions. Failure to improve outcomes IS an enormous threat to the entire LD and special education infrastructure.
LD identification through RTI has the enormous advantage of establishing links between identification and treatment. RTI establishes several of the basic elements known to be closely related to effective treatment including direct measures, graphing, formative evaluation, etc. Traditional LD identification sends children to special education with psychometric profiles that are irrelevant to effective treatment and likely cause harm by distracting teachers and others from focusing on variables that do make a difference in effective interventions. It is time for change.
Previous Page | Next Page
(Implications) | (Summary)
