COLORADO DEPARTMENT OF EDUCATION Colorado State Board of Education RULES FOR THE LONGITUDINAL ANALYSIS OF STUDENT ASSESSMENTS 1 CCR 301-72 [Editor’s Notes follow the text of the rules at the end of this CCR Document.] _________________________________________________________________________

1.00 Statement of Basis and Purpose.

The statutory basis for these rules is found in § 22-2-107(1)(c) and § 22-7-604.3, C.R.S. Colorado Revised Statutes § 22-7-604.3 requires the state board of education to promulgate rules which include, but are not limited to: adoption of a longitudinal growth model and related procedures for identifying (i) students who make at least a year’s academic growth in a year’s time, (ii) students demonstrating adequate academic growth, and (iii) schools that are associated with the highest rates of student academic growth and are thereby eligible for a Governor’s Distinguished Improvement Award.

2.00 Definitions.

2.00 (1) Department.

“Department” means the Colorado Department of Education created and existing pursuant to § 24-1- 115, C.R.S.

2.00 (2) State Board.

“State Board” means the State Board of Education created pursuant to Section 1 of Article IX of the State Constitution.

2.00 (3) CSAP.

“CSAP” means the Colorado Student Assessment Program.

2.00 (4) Academic Peers.

“Academic Peers” are all students in the same grade being tested in the same subject and having a similar CSAP score history in that subject prior to the current year.

2.00 (5) Quantile Regression.

“Quantile Regression” is a statistical regression procedure where various quantiles are predicted from an existing set of data rather than predicting the mean response as in typical regression procedures.

2.00 (6) Student Growth Percentile.

A “Student Growth Percentile” is the position of a student’s current CSAP scale score relative to academic peers. The student growth percentile provides a measure of academic growth (i.e. relative position change) where students who have similar academic score histories provide a baseline for understanding each student’s progress.

3.00 Implementation Procedures.

Code of Colorado Regulations 1

3.00 (1) Longitudinal Growth Model

1. 3.00 (1) (a) The longitudinal model to be used in the State of Colorado for the analysis of student academic growth will utilize a Quantile Regression technique to estimate Student Growth Percentiles for each student in the state who completes at least two successive CSAP tests in at least one academic subject. The individual Student Growth Percentiles estimated through use of a Quantile Regression technique will be used to identify which students have made at least a year’s growth in a year’s time and to identify which students are making adequate academic growth. Medians of the individual student growth percentiles will be computed for schools based on all students for whom qualified scale scores are available as defined by the Department. The median Student Growth Percentiles computed for each school will serve as the indicators of student growth associated with specific schools. Median Student Growth Percentiles will also be computed for each district and will serve as the indicators of student growth associated with specific districts.
2. 3.00 (1) (b) By August 15th (fifteenth) of each school year beginning in 2008, the Department will report to each school district in the state and each charter school in the state a median growth percentile for each school by subject, grade, performance level, and student group as defined by the Elementary and Secondary Education Act as amended. CDE shall also report median growth percentiles in a manner that relates them to school achievement levels.

3.00 (2) Identification of Students Making a Year’s Growth in a Year’s Time

1. 3.00 (2) (a) The 50th (fiftieth) Student Growth Percentile will be defined as typical student growth in any given year. Due to measurement imprecision inherent in the CSAP and uncertainty introduced when utilizing statistical models, Student Growth Percentiles will not provide perfect measures of student growth. A margin of error, to be determined by the Department and subject to change over time, will be added and subtracted from the 50th (fiftieth) growth percentile to arrive at a range of Student Growth Percentiles that are statistically indistinguishable from the 50th (fiftieth) Student Growth Percentile. The margin of error is to be initially set at 15 (fifteen) percentile points such that students for whom Student Growth Percentiles are at or between the 35th (thirty-fifth) and 65th (sixty- fifth) percentile will be considered to have made typical growth or be indistinguishable from those who did make typical growth. Typical growth shall be considered a year’s growth in a year’s time.
2. 3.00 (2) (b) Growth below the margin of error below the 50th Percentile shall be considered Low growth and growth above the margin of error above the 50th Percentile shall be considered High growth.
3. 3.00 (2) (c) By August 15th (fifteenth) of each school year beginning in 2008, the Department will report to each school district in the state and each charter school in the state the individual student growth percentiles for each student and which students attending the respective district or school made low growth, typical growth, and high growth.

3.00 (3) Identification of Students Making Adequate Academic Growth

1. 3.00 (3) (a) The definition of adequate academic growth is dependent upon the current scale score of individual students. Adequate academic growth is defined as a growth rate sufficient for a student to reach a performance level of proficient or advanced in a subject area within one, two, or three years.

Code of Colorado Regulations 2

1. 3.00 (3) (b) By August 15th (fifteenth) of each school year beginning in 2008, the Department will report to each school district in the state and each charter school in the state the individual student growth percentiles for each student, which students attending the respective district or school did make adequate academic growth and which students did not, and the amount of growth needed for a student to reach partially proficient, proficient, and advanced performance within one, two, and three years. Descriptors of the amount of growth needed shall include the categories of Low, Typical, High, and the actual growth percentile needed.

3.00 (4) Governor's Distinguished Improvement Awards.

In 2008 and each year thereafter, recipients of the Governor's Distinguished Improvement Awards will be identified through use of the longitudinal growth model adopted pursuant to these rules and awarded in accordance to § 22-11-301 et seq., C.R.S. Awards shall be made to schools that in the previous two school years met or exceeded a median growth percentile target set by the Department and based on reading, writing, and mathematics scores considered collectively. The single median computed for each school each year will be compared to an appropriate target median that may vary based on school type and size. School types will include elementary, middle, and high schools. School size categories will be determined by the Department and may be revised by the Department as necessary. _________________________________________________________________________ Editor’s Notes History Entire rule emer. rule eff. 03/06/2008.

Entire rule eff. 04/30/2008.

Code of Colorado Regulations 3