Functional Definition: Implementation of policies, procedures, and practices in the general education instructional, behavioral, and intervention process and/or the special education identification, referral, evaluation, or eligibility determination process that results in a systemic, pervasive, persistent pattern of inappropriate over-identification/under-identification of students with disabilities of a specific racial/ethnic group as eligible for special education and related services or in a specific eligibility category.

Statistical Definition - NJOSEP used multiple measures to statistically determine disproportionate representation. The measures included three descriptive statistics: unweighted risk ratio; risk rate comparison; and a measure of impact comparing expected vs. observed numbers of students identified as eligible for special education. The measures also included a statistical test of significance - chi square.

Districts were ranked on each of the three measures for a three year period (2005-2006, 2006-2007, 2007-2008. For the purpose of identifying districts with disproportionate representation (over-representation/under-representation) of racial/ethnic groups in specific disability categories, NJOSEP:

• applied the chi-square to this pool of districts (regardless of rank) for each racial-ethnic group and for the disability categories of specific learning disability, mental retardation, other health impaired, emotionally disturbed, language impaired, and autism; and
• applied a measure of impact comparing expected vs observed numbers of students identified as eligible for special education.

The procedure used one statistical test of significance and one statistical procedure.

Calculations:

The steps followed in the analysis:

1. Select the districts which were already identified as having discrepancies in all categories for a particular race - ethnicity
2. Calculate the chi square test on the above districts using 6 X 2 Matrix as below.

   	MR	SLD	ED	LI	OHI	AUT
   Black Classified
   All others Classified

3. Choose all those districts whose Chi-Value is more than cutoff value of 11.07.
4. Calculate “Observed – Expected” count of students in individual categories such as MR, SLD etc. for the above selected districts.
   

   District	Observed	Expected	(Observed-Expected)
   ABCD	XX	YY	XX-YY

5. Choose the districts for significant discrepancy in individual category if their “Observed – Expected” count is at least 10 for the above categories i.e. MR, SLD, Ed etc.
   

   District	(Observed – Expected)	>10
   ABCD	XX-YY	YES

   Local districts that were identified for disproportionate representation completed a self-assessment review of district policies, procedures and practices and identified areas that may be contributing to disproportionate representation. Based on that review they have developed an improvement plan to address areas of compliance and/or continuous improvement.