Response to Intervention (RTI) vs the Discrepancy Model

What is RTI?

RTI (Response To Intervention) is a rather new and somewhat controversial approach to the identification of a learning disability. In some ways it is a simpler approach, in other ways it is much more complex. Before discussing the process of RTI it is important to fully understand both the definition of a learning disability as well as the more traditional method for identification - The Discrepancy Model.

The definition of a Learning Disability:

Simply put, a learning disability exists when a student experiences significant difficulty learning and progressing in school due to underlying difficulty 'processing' certain types of information. In other words, to be considered 'learning disabled' the student must be 'underachieving' because his/her brain has difficulty understanding or coping with certain types of information. It is very important to note that the underachievement must be related to how the brain processes information and NOT due to lack of instruction, behavioral/motivational issues, generally low intellectual ability, etc., etc. Although all States address the general aspects of a learning disability as defined by IDEA, each State is given the responsibility to establish its own criteria for eligibility for special education services. For many years (since the inception of IDEA) a major component of all State LD eligibility criteria has been the inclusion of some form of 'severe discrepancy' component in order to objectively establish the severity of a student's underachievement. With the reauthorization of IDEA in 2004 (referred to as the 'Individuals with Disabilities Education Improvement Act of 2004'), however, school districts are no longer required to formally evaluate 'severe discrepancy' and instead are given the option of documenting underachievement through 'a process that determines if the child responds to scientific, research-based intervention as a part of the evaluation procedures'. Response to Intervention (RTI) is such an alternative process.

The discrepancy model:

Through the traditional discrepancy model, a learning disability has been determined primarily through a combination of cognitive (intellectual) and academic (achievement) testing. When a 'severe discrepancy' between ability and achievement is found, along with indication of underlying information processing issues, a learning disability can be identified and special education services can be provided. Each State establishes its own formula for determining when a 'discrepancy' can be considered 'severe'.

Problems with the discrepancy model:

Although the traditional discrepancy model provides a rather objective means for identifying a learning disability (just plug in the numbers and see if there is a 'severe discrepancy') there are a couple of very significant limitations to this process. First, it is extremely difficult to accurately determine 'severe discrepancy' for young students (prior to 2nd or 3rd grade). Students arrive in kindergarten with varying levels of exposure to education and developmentally progress at different rates even without the impact of any possible learning disabilities. It generally takes a few years for these developmental differences to 'even out'. Similarly, academic and cognitive assessment instruments are not highly accurate or reliable at early ages. So even though early intervention is considered a key factor to helping LD students become successful in school, the discrepancy model has proven at least somewhat ineffective when it comes to early identification. Another rather significant problem with the discrepancy model is that the 'severe discrepancy' cutoff value is somewhat arbitrarily determined and fails to adequately take into consideration 'error of measurement' which occurs in all forms of standardized assessment. In one sense, the severe discrepancy value merely provides a 'probability' that the student's achievement is significantly discrepant from his/her ability. Due to error of measurement and other factors, for some students an achievement score below the cutoff value is really not as 'severe' as indicated whereas for other students a score above the cutoff value is quite severe. As a result, strict adherence to the severe discrepancy model will definitely result in the identification of some students who do not actually have a learning disability while failing to identify other students who do.

Enter RTI:

RTI or 'Response To Intervention' is a new approach to the identification of a learning disability which is intended to alleviate the problems associated with the discrepancy model. With RTI, there is no need to formally evaluate cognitive abilities or academic achievement. All that is required is documentation that the student has been unable to respond to appropriate interventions which have been provided within the classroom. In other words, as long as the interventions within the classroom have been provided consistently over a reasonable period of time and have been based upon sound and scientifically proven educational principles, if the student continues to demonstrate significant underachievement within the classroom he/she can be considered eligible for special education services related to a learning disability. It is very important to note that the interventions provided within the classroom are 'mainstream' interventions and are not intended to simulate special education services. It is also important to note that even with RTI the student's underachievement should be related to an underlying cognitive processing issue.

Problems with RTI:

While RTI clearly provides a method of identification which 'bypasses' the typical problems of the traditional discrepancy model, it has serious problems and limitations of its own. First, there are no clear guidelines provided or objective means to determine what are or are not considered appropriate forms or levels of intervention. In other words, it is an extremely subjective process. While some mainstream teachers naturally provide appropriate interventions which may even enable LD students to succeed, other teachers will have considerable difficulty providing any level of appropriate intervention. In addition, some students (most notably those with lower cognitive abilities) will naturally struggle to keep up with their classmates regardless of any intervention which may be provided. As such, while some truly LD students may not be identified through RTI, many more non-LD underachievers will be found eligible for LD services.

Below is a chart comparing both positive and negative aspects of the traditional discrepancy model and RTI:

Traditional Discrepancy Model

 Response to Intervention Model



Objective, easy to apply and understand

Not practical or accurate for young students

Uses statistical properties to establish a predictable LD population size

Requires a student to reach a certain level of 'failure' before being identified

Allows examiners to evaluate learning style and information processing skills during testing

Does not adequately consider error of testing - will identify some non-LD students while failing to identify some LD students

Time-consuming assessment process for both student and examiner




Can be applied to students of all ages

Does not clearly differentiate between LD students and other pervasive 'underachievers'

Does not require a student to 'fail' for a significant period of time before being identified

Will likely identify virtually all 'slow learners' as LD

Forces schools to be 'proactive' in providing mainstream interventions

Does not have any formal means of cognitive 'processing' evaluation built into the process

Requires little if any educational disruption for testing


Can be inappropriately influenced by parents, teachers, or others who simply want a student to be identified

Obviously, there is no easy or entirely accurate or reliable means of evaluating/identifying a learning disability. Regardless of the method chosen, assessment teams need to be aware of the possible pitfalls and use their very best professional judgment when determining if a student really does or does not have a learning disability. Best practice might suggest using some combination of discrepancy model and RTI to be most respectful of the individual needs of all students.

back to top



Return to the LDinfo Web Site to find out about any of the following topics (and more):

Learning disabilities - what is a learning disability (LD or SLD)?

Dyslexia: Dyslexia is a reading disability or reading disorder

Dysgraphia Dysgraphia is a writing disability or disorder

Dyscalculia Dyscalculia is a math disability or disorder

What is an attention deficit disorder (ADD, AD/HD, ADHD)?

Gifted LD: Can a student be gifted and LD?

Emotional/Behavioral issues and LD: Do LD students experience behavior problems or depression?

Section 504: What is a Section 504 plan?

What is special education?

What is processing?

What is a severe discrepancy?

What is a nonverbal learning disability (nonverbal LD or NLD)?

What is a central auditory processing disorder (CAPD)?

What is IDEA?


LDinfo Home


Copyright © 2006-2017 LDinfo Publishing
All rights reserved. Any reuse or republishing of the text or images on this web site without prior written consent of the copyright holder is prohibited.