• Teachers
• School Psychologists
• School & District Administrators
• State Departments of Education
• Researchers
• Parents

• NRCLD Home
• SLD and RTI Tools
• SLD Determination
• What is RTI?
• "How to" topics
• Changing roles of educators
• About NRCLD
• NRCLD research
• NRCLD publications
• NRCLD presentations
• NRCLD Contacts
• SLD and RTI resources

And Miles to Go...:
State SLD Requirements
and Authoritative Recommendations

Previous Page | Next Page
(Results) | (Trends)

Report Text

• Introduction
• SLD Definitions and Classification Criteria
• Historical Overview
• Questioning the Ability-Achievement Discrepancy
• Recommendations for Change in SLD Definition and Classification Criteria
• State Education Agency SLD Definitions and Criteria
• Methodology
• Results
• Analysis of Discrepancy Determination Methods
• Summary of Trends
• SLD Beliefs
• Further Analyses
• Discussion
• References

Figures and Tables

• Figure 1: Distribution of State SLD Prevalence
• Table 1: Consensus Conclusions: LD Summit Follow-Up (Bradley et al., 2002)
• Table 2: Components of State Definitions and Criteria
• Table 3: Methods used to determine the discrepancy
• Table 4: LD Beliefs

Analysis of Discrepancy Determination Methods

Although the discrepancy information was especially difficult to interpret, through numerous follow-up conversations with SEA LD authorities, analyses of the achievement-intellectual ability criteria in the 50 states were completed (see Table 3). According to these analyses, 19 states either did not attempt to operationalize discrepancy criteria or left it entirely to the discretion of LEA multidisciplinary teams. An additional two states established criteria that were difficult or impossible to implement consistently across districts (NY and NJ). A total of 21 states, in our judgment, provide little or no guidance to LEAs on discrepancy determination (42%).

For states that did provide criteria, the most frequent method accounted for regression effects either through regression tables (N=14 states) or achievement discrepancies determined from IQs corrected for regression (N=3 states). An additional 10 states established discrepancy criteria using standard score magnitudes (N=3) or standard deviation magnitudes (N=7). Two additional states listed in Table 3 (LA and KY) established what appear to us to be relatively objective discrepancy determination methods. In our judgment, 29 states (58%) establish clear guidance to local practitioners regarding the methods to determine the discrepancy and the magnitude to meet a standard for severe.

Direct comparisons of discrepancy determination methods to Mercer et al. (1996) are difficult for a number or reasons. Mercer et al.'s table did not separate standard score from standard deviation criteria and the methods listed by states that summed to well over 50. Therefore, our comparisons involve some judgments rather than direct comparisons of identical statistical summaries. With these cautions, it appears to us that the number of states providing guidance to LEAs regarding discrepancy method and magnitude has increased slightly to a clear majority of states. This trend might be regarded as positive.

There was a decline in the use of clearly inadequate methods such as Wechsler Verbal-Performance discrepancies (1 state in 1994 vs 0 states currently), discrepancies expressed as grade equivalents (3 states in 1994 vs 0 states currently) and a decline from 2 states to 1 state using a percentage discrepancy criterion. At the same time, there appears to be a substantial proportion of states (over 40%) that provide little guidance about discrepancy method or magnitude, a proportion that has changed little since 1994. Several intriguing analyses remain to be conducted to see if these state differences influence prevalence. For example, does a high degree of LEA discretion lead to higher, lower, or no effect on prevalence?

Previous Page | Next Page
(Results) | (Trends)