AERA March 25, 2008 Delaware’s Growth Model and Results from Year One.

AERA

March 25, 2008

Delaware’s Growth Model and Results from Year One

Background Information

• Invitation for states to submit proposals to use a growth model

• Pilot project – up to ten states

• Model must demonstrate that it can raise student achievement and enhance school accountability

• “Bright Line” principles of NCLB upheld

• DE first submitted proposal in March 2006 – was denied

• DE revised/resubmitted proposal September 2006

• USED approved for use in 2006-07 with one condition– Cannot use Confidence Interval

• Calculate AYP by original and growth models

• Report both original and growth models

Why did we submit?

• To ensure more valid and reliable accountability determinations

• To monitor various subgroups progress

• To support our value of continuous improvement and longitudinal student growth

What model did we propose?

• Value Table Model

– Maintains emphasis on performance levels (standards based achievement)

– Values longitudinal student growth

– Gives schools credit for moving students towards proficiency

– Values growth especially below the standard

Who chose Delaware’s model?

• Committee of Stakeholders

– District administrators, school administrators, teachers, parents, community

How do value tables work?

• Values are placed in a table to indicate points earned from one year to the next

• Calculate the average growth value for the school and each subgroup in reading and math

• Compare average growth to the target

Value Table for Grade 3

Grade 3 Level

Grade 2 Level

Level

1A

Level

1B

Level

2A

Level

2B

Proficient

Below 0 0 0 200 300

Meets 0 0 0 0 300

Value Table for Grades 4-10

Year 2 Level

Year 1 Level

Level

1A

Level

1B

Level

2A

Level

2B

Proficient

Level 1A 0 150 225 250 300

Level 1B 0 0 175 225 300

Level 2A 0 0 0 200 300

Level 2B 0 0 0 0 300

Proficient 0 0 0 0 300

Growth Value Targets Table

Reading Math

2006 186 123

2007 204 150

2008 204 150

2009 219 174

2010 237 201

2011 252 225

2012 267 249

2013 285 276

2014 300 300

Delaware’s Accountability System:

Adequate Yearly Progress (AYP)

2006• Participation (ELA, Math)• Performance (ELA, Math)

– Total School• Original Status /

Safe Harbor

– Subgroup• Original Status /

Safe Harbor

• Other Academic Indicators

2007• Participation (ELA, Math)• Performance (ELA, Math)

– Total School and Subgroup• Growth• Original Status /

Safe Harbor

• Other Academic Indicators

How to meet AYP

• Meet Growth Value Targets or Original Status Targets in ELA and math

• Meet Participation Targets in ELA and math

• Meet the Other Academic Indicator

Ratings

Ratings are determined by the combination of:

• AYP Determination

• State Progress Determination

• Accountability History

Ratings Table

AYP STATE PROGRESS

STATE ACCOUNTABILITY DETERMINATION

AFTER 2 CONSECUTIVE

YEARS

A A Superior

A M Superior

A B Commendable

M A Superior

M M Commendable

M B Commendable Academic Review

B A Academic Review Academic Progress

B M Academic Review Academic Progress

B B Academic Review Academic Watch

Schools facing appropriate consequences per NCLB

How to be classified as

Under Improvement

• Two consecutive years not meeting AYP in same content area

-- ELA

-- Math

-- Other Academic Indicator

What Happened in 2007?

• 193 Schools with ratings– 146 schools (76%) used growth model for

rating• 89 schools (46%) used growth model and made

AYP• 57 schools (30%) used growth model and did not

make AYP

– 47 schools (24%) used original model for rating (all made AYP)

Comparison of Growth to Original

• 82 schools (42%) made both growth and original

• 57 schools (30%) did not make growth or original

• 7 schools (4%) made growth but not original• 41 schools (21%) made original but not

growth• 6 schools used original model only – K, K-1,

K-2 schools – no growth available

Correlation of Growth and Original Models – A look at those schools that did not make

AYP (57 schools)

• Missed reading in growth model also missed reading in original model for same subgroups

• Missed math in growth model may or may not have missed math in original model for same subgroups

• Missed special education in reading or math in growth model also missed in original model

A look at Reading for Those Schools That

Only Made Original Model (47 schools)

Subgroup Met OM Target

CI Safe Harbor Safe Harbor CI

All Students 87% 11% 2% 0%

Black 57% 40% 3% 0%

Hispanic 84% 8% 0% 8%

White 98% 0% 0% 2%

Special Education

4% 21% 42% 33%*

Low Income 65% 28% 5% 2%

A Look at Math for Those Schools That Only Made Original Model (47 schools)

Subgroup Met OM Target

CI Safe Harbor Safe Harbor CI

All Students 99% 0% 2% 0%

Black 89% 8% 1% 0%

Hispanic 92% 8% 0% 0%

White 100% 0% 0% 0%

Special Education

17% 33% 29% 29%

Low Income 86% 12% 0% 0%

Some Observations

• Growth model results will be similar to Original– Principles of NCLB, 100% proficient by 2013-14– Different models produced similar determinations

• Growth models only help when real growth is occurring– Showing growth in low achieving students but rate of

improvement is differs by subgroups

• If AYP was not met:– ½ of the students maintained their level from previous year– 1 in 4 improved– 1 in 4 regressed– Similar pattern across subgroups

Contact Information

• Robin Taylor– rtaylor@doe.k12.de.us– 302-735-4080