Attachment 2

NYS Sampling Methodology for Some Federal Indicators
in the 2005-2010 State Performance Plan

NYS will collect data from a statewide representative sample of school districts on six federal indicators. No district will report on all indicators every year except New York City. All school districts will provide data on all six indicators distributed over a six-year period beginning with the initial year in which data are collected for each indicator. The six indicators are as follows:

NYS has distributed all school districts among six statewide representative samples. These six groups of school districts were tested with Analysis of Variance (ANOVA) and there was no statistical difference among the six groups of school districts on the population variables listed in the table below. These population variables were from the 2000 decennial census.

 

Census 2000 Population Variables Used to Ensure Each Sample of School Districts is Similar

population

female poverty head of household

n households in POV

n children in families

n unempl over 16

n house classif in POV

n children w/single parent

n in workforce

n  households w/ no plumbing

n children 5~17 in poverty

n unempl 1999

n total Households

n 5~17

persons not in POV

n one room Households

n 5~17 relevant for school

n classif in POV

n occupied Households

n less than 5

n children in 1 parent family

n over 25 not graduate of HS

female head of household

n children in families

n total over 25

 

New York City is the only LEA in the state with a total enrollment of 50,000 or more students, so it will be represented in each of the six samples.

All school districts will have a choice of reporting data on all eligible students for each federal indicator or submitting data on a randomly selected sample of students. The minimum number of students required for these indicators can be obtained by using the sampling calculator provided by the State and the guidelines provided below. The vast majority of school districts will need to submit data on all eligible students on most indicators. For some large school districts if it will be less burdensome to report on a sample of students, the methodology described below (totally random sampling) is likely to produce a sample that is representative of the school district in terms of all variables, since every eligible student has the same chance as another student to be selected for the sample. 

SED will require that LEAs maintain documentation as described below if they choose to report data on a sample of students. The totally random sampling methodology and required documentation should eliminate selection bias.  School districts will be required to over-sample as described below for indicator 8 where poor response rate is a known issue. Also, school districts will be encouraged to provide surveys for indicator 8 in a variety of ways to improve the response rate.  The Department will attempt to prevent missing data by first describing precisely what the State needs to collect, providing technical assistance and then following up with school districts to request missing data. The completeness of data collection will improve after the first year and will continue to improve as long as requirements remain unchanged. All issues of confidentiality will be handled in accordance with the rules and procedures in FERPA. The Department will also guard against divulging personally identifiable information by not reporting results when there are less than five students for whom data are available or when those results can be easily calculated based on other data provided 
 

Federal Indicator Number

Eligible Population of Students From Which a Random Sample Must be Selected

Minimum Number of Students in the Sample

Method for Selecting Students

Required Documentation

7

Entry - all children who are referred for preschool special education programs and/or services.

Exit - all children who received preschool special education programs/or services for at least six months and are declassified or are within their last six months of eligibility for preschool special education services and  the annual review meeting for whom entry evaluation data are available.

Use a sampling calculator. Require 95% confidence interval and plus or minus 5% margin of error.

Random selection using a random number table.

Documentation period is seven years.

Maintain list of all eligible students, copy of Random Number Table used, beginning random number for selecting students and list of all students who were selected and their number.

8

Every preschool and school-age student with a disability who is provided special education programs and/or services in a district-operated program or under contract with other service providers.

Use a sampling calculator. Require 95% confidence interval and plus or minus 10% margin of error. Expect 10% response rate, so require over-sampling by 90% of minimum number identified by the calculator.

Same as above.

 Same as above.

11

For preschool and school-age students: All preschool and school-age students for whom parental consent for an initial evaluation was received during the school year (July 1-June 30).

Use a sampling calculator. Require 95% confidence interval and plus or minus 2% margin of error.

Same as above

Same as above

12

All children who are referred for special education programs and/or services from Part C to Part B prior to age 3 during the school year (July 1-June 30).

Use a sampling calculator. Require 95% confidence interval and plus or minus 2% margin of error.

Same as above

Same as above

13

All students with disabilities ages 15-21 who are provided special education services in district-operated programs or under contract with other service providers.

All students up to 30.

 

New York City sample 100 students

Same as above

Same as above

14

All students with disabilities who are no longer in secondary school but received some special education program and/or service during the school year (July 1-June 30) in district-operated programs or under contract with another service provider. (Include all students who left with a credential, reached maximum age for educational services or dropped out.)

School districts with less than 100 students with disabilities exiting, survey all students.

School districts with 100 or more students use the sampling calculator. Require 95% confidence interval and plus or minus 5% margin of error.

Same as above

Same as above

 

The table below demonstrates a schedule for data collection from the six samples of school districts on the six federal indicators listed above.  Please note.

 

School Year

Schedule for Reporting Data on Some Federal Indicators

Sample 1

Sample 2

Sample 3

Sample 4

Sample 5

Sample 6

2005-06

8

11

12

13

14 (Year 1)

7 - entry

2006-07

7 - exit

8

 

11

 

12

14 (Year 2)

13

14 (Year 1)

 

2007-08

14 (Year 1)

 

7-exit

8

11

12

14 (Year 2)

13

2008-09

14 (Year 2)

13

14 (Year 1)

 

7-exit

8

 

11

 

12

 

2009-10

12

14 (Year 2)

13

14 (Year 1)

 

7-exit

 

8

 

11

 

2010-11

11

 

12

 

14 (Year 2)

13

14 (Year 1)

 

7-exit

 

8

7-exit

2011-12

 

 

 

14 (Year 2)