ABCs of Public Education 20002001

1
ABCs of Public Education2000-2001

• Diane N. Villwock, Ph.D.
• August 2000

2
ABCs of Public Education

• Accountability
• Back to Basics
• (Local) Control
• D.A.R.E.
• Explanations
• Graphing
• History
• Instruction
• Journalism
• Kick-ball
• Language
• Mathematics
• Nouns

• Oral exams
• Physics
• Quality
• Reading
• Science
• Textbooks
• Uniforms
• Varsity
• Writing
• eXcellence
• Yearbooks
• Zippers (on backpacks)

3
ABCs of Public Education

• Accountability
• Back to Basics
• (Local) Control

4
Basic Concepts

• School level accountability
• Cash incentives for good performance
• Consequences for poor performance
• Mostly based upon improvement in test scores
• One model all schools
• The measures for each school depend upon the
  grade levels at the school and the courses taught

5
Two Basic Measures

• Performance composite
• proportion of all students tested who were
  proficient on the included tests or meeting
  certain standards
• proficiency is defined as a level III or IV
• method of setting proficiency differs by test
• Implicationsone can only grow so far on this
  measurenowhere up to go from 100

6
The Second Measure

• Growth
• calculation within cohorts for EOG, HSCT and EOC
  --direct measure of learning
• index for growth across cohorts for writing at
  4th, 7th and 10th grades--measure of improvement
  of program over time
• Growth on EOGs and HSCT is possible and highly
  desirable for EVERY STUDENT (even Level IV at
  99th percentile)
• we are nowhere near having ceiling problems

7
Goals for Growth

• Two levels of goals for growth are calculated
• expected 100 of State growth prediction
• exemplary 110 of State growth prediction for
  EOG and HSCT and 105 of State growth prediction
  for EOC

8
Awards

• School of Excellence
• at least 90 proficient on the performance
  composite and at least expected growth
• School of Distinction
• at least 80 proficient on the performance
  composite regardless of growth

9
Other Labels

• No Recognition
• schools with at least 50 performance composite,
  but who did not make expected growth
• Low-Performing
• schools with less than 50 performance composite
  and did not make expected growth

10
Cash Incentives

• At schools with exemplary growth, certified staff
  are paid 1500 and teaching assistants 500
• At schools with expected growth, certified staff
  are paid 750 and teaching assistants 375

11
Elementary School Example

• see FPG Accountability Report
• Growth is calculated by subtracting the
  pre-test scale score from the post-test scale
  score actual mean scores
• Pre-test scores
• a pre-test is given at 3rd grade in August
• previous springs score in 4th-5th grades

12
Calculating Actual Growth

• 150.8 post-test, 143.3 pre-test
• difference is 7.5 points--growth
• students in model is the number of students for
  whom there were both pre- and post-test scores
• schools are responsible for obtaining scores for
  in-state transfer students

13
Regression Model to Set Goals

• DPI uses regression to predict expected and
  exemplary growth
• Variables used include
• pre-test mean for grade level
• adjustment for regression toward the mean
• average rate of growth for this subject in NC

14
Expected Score

• Expected score is the goal set by the State for
  a given school
• Expected growth is the amount of scale score
  points the mean is supposed to move between pre-
  and post-test
• Difference in growth is the difference between
  the actual and the goal

15
Example

• Actual growth of 7.5
• subtract expected growth of 6.2
• equals difference in growth of 1.3

16
Standard Growth

• Standard Growth is the Difference in Growth
  divided by the statewide standard deviation of
  growth
• this standardizes the scores across tests so that
  we can
• add up the standard growth scores
• compare standard growth scores to each other

17
Exemplary Growth

• Exemplary growth is 10 higher than expected
  growth on EOG
• Difference in growth is the actual score minus
  the exemplary growth amount
• Standard growth is the difference in growth
  divided by the state standard deviation for growth

18
Example of Standard Growth

• The difference in growth is 1.3
• The chart does not show the standard deviation of
  growth that is used to compute standard growth
• standard deviations are set (dont change)
• The standard growth for that 3rd grade reading is
  listed as 0.8
• Its calculated by the software that generates
  the report

19
Growth in Writing Index

• Scores are broken down into achievement levels I
  through IV (I means insufficient mastery of
  material, IV is superior mastery)
• Multiply the number of Level IVs by 3, Level IIs
  by 2 and add in the number of Level IIs
• Divide by 3

20
Examples

• 40 IV, 30 lII, 20 II, 10 I
• 3 x .40 1.20
• 2 x .30 0.60
• 1 x .20 0.20
• 1.80
• 1.80/30.60

• 25 IV, 25 lII, 25 II, 25 I
• 3 x .25 0.75
• 2 x .25 0.50
• 1 x .25 0.25
• 1.50
• 1.50/30.50

21
Growth Goals for the Indices

• The newest years index is compared to the
  average of the previous two years indices
• The expected and exemplary goal is at least 0.1
  units of improvement
• Baseline is the average of two previous years

22
Writing Index in Example

• The year one, year two, current year and baseline
  indices are not listed on the report
• Report gives the difference between the baseline
  and the current, found under expected
  difference heading
• and standard gain

23
Growth Composite

• Growth Composite is the sum of the standard
  growth scores under each of the two categories
  expected and exemplary
• The composite must be at least 0.1 for a school
  to have met its goal and to be eligible for cash
  awards
• The larger the composite, the greater the amount
  of growth demonstrated by the students
• Comparing the composite of two schools or a
  school over time is legitimate

24
Percent tested

• Percent tested is the proportion of the
  enrollment who were not exempt from testing who
  actually took the test
• Performance composite is the number of
  proficient students divided by the number tested
  across all measures

25
Footnotes

• Inclusion in growth calculations requires
• students with both pre- and post-test scores
• EOG--enrollment for at least 91 days
• HSCT--enrollment for at least 160 days
• Full precision -- rounding of calculations done
  only at the end

26
Middle School Example

• Look at MCDM Accountability report
• Parallels elementary school example
• EOG at 6th, 7th and 8th
• Writing at 7th
• Last years shows index for EOC, but for 2001
  there is a new measure for EOC tests

27
Predicted Means for EOC tests

• The expected mean for each course is predicted by
  a regression equation
• Scores used to predict means include some
  combination of 8th grade reading or math, Algebra
  I, Biology or English I, and the average
  performance in NC on each test
• See Prediction Formulas for EOC Performance
  hand-out for details

28
Predicted Mean EOC Scores

• Students included only if they have all the
  required scores in SIMS
• Length of enrollment doesnt matter, only that
  the district has the scores
• Implications--7th grade Algebra I cant be
  included in the prediction (requires 8th grade
  EOGs)
• State says 90 of students stayed in the
  calculations statewide

29
Derivation of the Equations

• State used a statistics program to create all
  possible regressions
• The simplest equation that passed tests of equity
  was chosen
• equity means that predictions for schools with
  different pre-test measures were equally accurate
• Reliability in .60s to .70s

30
EOC for Middle School

• See Worksheet for Computing Expected Growth/Gain
  Composite
• For Algebra I the report will show the
• actual EOC mean for all students in the model
• predicted average from the regression
• the difference between them
• standard growth

31
Also New for 2001 at Middle School

• The performance composite will include the
  percent of students who have met the computer
  proficiency standard
• percent of students who have passed both the
  multiple choice and performance tests

32
High School for 2000-2001

• See Worksheet for Computing Expected
  Growth/Gain
• All 10 multiple choice EOCs are included and use
  the new prediction formulas
• English II continues to be measured by the index

33
HS 2001 (continued)

• CUP/CTP--College University Prep/College
  Technical Prep
• Current years proportion must exceed 0.1
  compared to an average of the two past years
• Comprehensive Test (HSCT) is like EOG model

34
HS for 2000-2001 (III)

• Competency rate must increase by at least 0.1
  between 8th and 10th grade
• Dropout rate is being added, but the details are
  not yet available
• Note There is no difference between expected and
  exemplary goals for English II, CUP/CTP and
  Competency

35
HS Performance Composite Details

• Tests in the high school performance composite
• Algebra I II
• Geometry
• Biology
• Chemistry
• Physics
• Physical Science

• English I II
• US History
• ELP
• HSCT in math/reading
• Algebra I for students who took course in middle
  school

36
District Results for 1999-2000

• All schools showed exemplary growth for second
  consecutive year
• 4 Schools of Excellence--McDougle Elementary and
  Middle, Culbreth and Phillips Middle
• 8 Schools of Distinction--Carrboro, Ephesus,
  Estes Hills, FPG, Glenwood, Seawell, CHHS, East

37
(No Transcript)