The discrepancy model is the traditional method used to determine whether a student has a learning disability. It works by comparing two scores: a student’s result on an IQ test and their performance on academic achievement tests. If a significant gap exists between the two, the student may qualify for special education services. For decades, this was the dominant approach in American schools, though it has faced sharp criticism and is now one of several options available to educators.
How the Discrepancy Model Works
The core logic is straightforward. A student takes a standardized intelligence test, such as the Wechsler Intelligence Scale for Children, and separately takes academic achievement tests in areas like reading, writing, or math (the Woodcock-Johnson Achievement Test is a common example). Evaluators then compare the two sets of scores. If a student’s IQ suggests they should be performing at a higher academic level than their achievement scores reflect, that gap is the “discrepancy.”
The size of the gap required to qualify varies by state and school district, but the underlying assumption is consistent: a learning disability exists when a student’s intellectual ability and their actual academic performance don’t match. A child who scores in the average range on an IQ test but reads far below grade level, for instance, would show the kind of discrepancy this model is designed to catch. The idea is that the student has the cognitive capacity to perform better, and something specific is interfering with learning in that academic area.
The “Wait to Fail” Problem
The most persistent criticism of the discrepancy model is that it forces students to fall significantly behind before they can get help. Researchers and educators have long called it a “wait to fail” approach, and the data backs that label up. Under this model, the average age at which students were identified with a reading disability was 10 years old, roughly third or fourth grade. By that point, a child has already spent years struggling without targeted support.
This is particularly damaging because research consistently shows that early intervention for reading difficulties is far more effective than trying to remediate problems later. A student who gets intensive reading support in kindergarten or first grade has a much better trajectory than one who doesn’t receive services until age 10. The discrepancy model, by its design, requires a measurable gap to develop between ability and achievement. For young children, that gap simply hasn’t had time to widen enough to meet the threshold. So the students who would benefit most from early help are the ones least likely to qualify for it.
The concern became serious enough that leading researchers in special education and school psychology held a formal summit in 2001, producing a series of papers arguing that schools needed to provide powerful early reading interventions rather than waiting for children to fall far enough behind to qualify. This momentum directly influenced changes in federal education law.
Other Limitations
Beyond the timing problem, the discrepancy model has technical weaknesses that affect its accuracy. IQ and achievement test scores both carry measurement error. When you compare two imperfect scores to find a gap between them, that error compounds. A student might qualify one day and not the next simply because of normal score fluctuation, not because their learning profile actually changed.
The model also struggles with students whose IQ scores are on the lower end of average. These students may have genuine learning disabilities but never show a large enough gap between their IQ and achievement to qualify. Their academic struggles get attributed to limited ability rather than a specific learning difference, and they miss out on services entirely.
There are equity concerns as well. IQ tests can reflect differences in language exposure, cultural background, and educational opportunity. Students from under-resourced schools or non-English-speaking households may score lower on IQ measures for reasons unrelated to their actual cognitive ability, which can mask a true discrepancy or create a misleading one.
Response to Intervention as an Alternative
The main alternative that emerged is Response to Intervention, commonly called RTI. Rather than testing for a gap between IQ and achievement, RTI focuses on how a student responds to increasingly intensive instruction. Every student receives high-quality classroom teaching (Tier 1). Those who fall behind get additional targeted support in small groups (Tier 2). Students who still struggle despite that intervention move to even more intensive, individualized help (Tier 3). If a child continues to underperform even with strong, evidence-based instruction at every level, that persistent difficulty may indicate a learning disability.
The philosophical shift is significant. RTI doesn’t wait for failure. It identifies struggling students early and provides support immediately, using the student’s response to that support as diagnostic information. A child who catches up with targeted help probably didn’t have a learning disability. A child who doesn’t respond, even to well-designed intervention, likely does. The identification process and the intervention happen simultaneously rather than sequentially.
Where Things Stand Now
When the Individuals with Disabilities Education Act was reauthorized in 2004, the updated law made a key change: states could no longer be required to use the IQ-achievement discrepancy model as the sole method for identifying learning disabilities. States gained the option to use RTI or other alternative approaches. Some states moved away from the discrepancy model entirely. Others kept it as one available method alongside RTI. A handful still rely on it heavily.
In practice, many school districts now use a combination. They may begin with an RTI framework to catch struggling students early and provide tiered support, then use more traditional testing (which can include discrepancy analysis) as part of a comprehensive evaluation when a student is referred for special education. The discrepancy model hasn’t disappeared, but it is no longer the default gatekeeping mechanism it once was. The broader trend in the field has moved toward identifying learning difficulties through a student’s actual response to quality instruction rather than a single snapshot comparison of two test scores.