Upload
mabyn
View
29
Download
1
Embed Size (px)
DESCRIPTION
How much are we learning? Advances in school evaluation in New York and São Paulo. May 19, 2010. Jesse Margolis. Instituto Braudel. How much are we learning?. General observations about São Paulo schools Shared characteristics with New York Notable differences - PowerPoint PPT Presentation
Citation preview
How much are we learning?
Advances in school evaluation in New York and São Paulo
May 19, 2010
Jesse Margolis
Instituto Braudel
2
• General observations about São Paulo schools
– Shared characteristics with New York
– Notable differences
• Brief presentation of New York City’s accountability system
• School evaluation in São Paulo
– Best practices
– Opportunities for improvement
– Recommendations
How much are we learning?
3
Schools in São Paulo share challenges with New York• Both states enroll a large number of public school students…
– New York State has 2.7 million students enrolled in 4.6K public schools– São Paulo State has 7.8 million students enrolled in 11.8K public schools (approx.
10 Federal, 5.7K State and 6.1K Municipal)
• …and contain big cities with a substantial share of public school enrollment– New York City has 1.0 million students enrolled in 1.5K public schools– The city of São Paulo has 2.0 million students enrolled in 1.6K public schools
(approx. 2 Federal, 1.1K State and 500 Municipal)
• In both cities, a significant share of school-age children are enrolled in private schools…– In New York City, 19% of all students are enrolled in private schools, compared with
10% in the rest of the state and 10% in the entire United States– In the city of São Paulo, 18% of all students are enrolled in private schools,
compared with 14% in the rest of the state and 12% in all of Brazil
• …leaving the public school system primarily for low-income families– In New York City, 66% of public school students are eligible for free lunch due to low
family income– In São Paulo, 68% of public school parents in the State system report an income of
R$1,275 (about US$700) or less per month
Source: NYSED, Censo Escolar 2008, SARESP Cuestionário 2008; Note: excludes “eucação infantil” for São Paulo and PK for NYC.
4
• General observations about São Paulo schools
– Shared characteristics with New York
– Notable differences
• Brief presentation of New York City’s accountability system
• School evaluation in São Paulo
– Best practices
– Opportunities for improvement
– Recommendations
How much are we learning?
5
• State: regulation and oversight• Establish standards• Certify teachers• Evaluate schools and districts• Define physical boundaries and
governance of school districts• Redistribute funding between
wealthier and poorer school districts
• City: run schools• Build buildings• Hire principals, teachers and other
staff & set salaries• Develop an admissions system for
students• Decide on details of curriculum &
professional development• Pay staff and other expenses
Responsibilities for school governance are divided differently in São Paulo and New York
0%
20%
40%
60%
80%
100%
ElementarySchols
MiddleSchools
High Schools
Per
cent
age
of S
tude
nts
São Paulo: State and municipal governments both regulate and run schools
StateSchools
MunicipalSchools
New York: State regulates school system and city runs schools
2.5MM 2.5MM 1.5MM
Source: Censo Escolar 2008
6
0
50,000
100,000
150,000
200,000
Accoun
tants
Registered
nurses
Sales(m
anufac
turin
g)
ES,M
S&HS
Teache
rs
Policeofficers
Officesu
perviso
rsCa
rpen
ters
Exec
secretary
Truc
kdrivers
Mainten
ance
workers
Accoun
tingclerks
Custom
erse
rvicereps
Secretaries
Nursingaide
s
Secu
ritygu
ards
Janitors
andcle
aners
Offic
ecle
rks,
gene
ral
Retailsalesp
ersons
Teac
hera
ssist
ants
Waiters
Homehe
alth
aide
sStoc
kclerks
Cash
iers
Chief executivesPhysiciansLawyers
FinancialmanagersOperationsmgr.
Teachers in New York State earn above the median salary, but below most other professions requiring a college degree
US$ 65.899 / year
Ave
rage
Sal
ary
($U
S /
year
)
Note: includes categories 25-2031, 25-2022 and 25-2021 for high, middle, and elementary school teachers. Does not include kindergarten, pre-school, or special education teachersSource: Occupational Employment Statistics (OES) Survey, May 2008. Bureau of Labor Statistics and Department of Labor
7
0
2,000
4,000
6,000
8,000Prof
(ens
inosu
perio
r)Ad
voga
dos
Gerentes
operaç
ões
Gerentes
deap
oio
Secretáriasex
ecutivas
Enferm
eiros
Professo
resEF
&EM
Cond
utores
(distrib)
Policiais
Aux.
enferm
agem
Escritu
rário
sem
geral
Vend
edores
emlojas
Alve
naria
Guarda
sevigias
Garçon
seba
rmen
Embe
leza
men
to
Recepc
ionistas
Telemarke
ting
Cozin
heiro
sTrab
alha
doresag
rícolas
Manu
tenç
ãode
edifício
s
Costurade
roup
as
Serviço
sdo
méstic
os
Aten
d.de
crec
he/id
ososMédicos
DiretoresgeraisEngenheiroscivis
Dirigentesde empresas
Reps
comerciais
Source: PNAD 2008
Ave
rage
Sal
ary
($R
/ m
onth
)In São Paulo State, the same is true, though salaries across the board are lower
0
2,000
4,000
6,000
8,000
10,000Mé
dico
s
Enge
nheiroscivis
Cirurgiões-den
tistas
Enge
nheirosmecân
icos
Prof
(ens
inosu
perio
r)
Pesq
.Econô
mica
Contad
ores
Journa
listas
Advo
gado
s
Diretoresde
prod
ução
Psicólog
os
Enferm
eiros
Farm
acêu
ticos
Fisioterap
eutaseafins
Professo
resEF
&EM
8
Among professions in São Paulo with a high share of university degrees, elementary and secondary teachers are the largest and lowest paid group
Note: includes only those professions where 80% or more of respondents said that their last level of education studied was “Superior – graduação” or “Mestrado ou doutorado”“Source: PNAD 2008
Ave
rage
Sal
ary
($R
/ m
onth
)
9
Full-time teachers in São Paulo spend a larger share of their contractual time teaching students
0
5
10
15
20
25
30
35
40
45
NYC (High Poverty) NYC (Low Poverty) São Paulo
Teac
hing
per
iods
(50
min
utes
)
Teaching(22 periods)
Preparation(8 periods)
Professional Activity (5 per.)
Lunch/Transition(6 periods)
41 periods(34hr 10min / week)
Teaching(25 periods)
Preparation(5 periods)
Professional Activity (5 per.)
Lunch/Transition(6 periods)
41 periods(34hr 10min / week)
Teaching(33 periods)
HTPC* (3 periods)
Prep. (4 periods)
40 periods(33hr 20min / week)
In or out of school (not teaching)In school (not teaching)In school (teaching)
* HTPC is shared planning and professional development timeSource: NYC DOE teachers contract (2007-2009); DECRETO Nº 55.078, DE 25 DE NOVEMBRO DE 2009
10
46% of public school teachers in the São Paulo State system report working in two or more schools
0%10%20%30%40%50%60%70%80%90%
100%
Elementary SchoolTeachers (4th
Grade)
Middle and HighSchool
(PortugueseLanguage)
Middle and HighSchool (Math and
Physics)
Middle and HighSchool (Chemistry
and Biology)
Respondents: 7,102 15,874 17,542 11,763
Per
cent
age
of T
each
ers
Source: Qúestionario SARESP 2008
One School
Two Schools
Three or more
11
Across all grade levels, average class sizes in São Paulo are larger than in New York
Note: Sao Paulo class size is for Regular Education in Urban, State Schools. New York City class size is for General Education and Gifted & TalentedSource: 2008 Censo Escolar for São Paulo; 2009/10 Class Size Report for New York City
31 32 31 3235 35 35 36
3736 35
22 2325 25
26 27 28 27 27 27 27
0
5
10
15
20
25
30
35
40
2 3 4 5 6 7 8 9 10 11 12
Ave
rage
Cla
ss S
ize
São Paulo State (Avg. = 34.7)(Urban schools)
Grade Level
New York City (Avg. = 26.1)
12
Teacher absenteeism is two and a half times greater in São Paulo State than in New York City
15%
6%
0%
2%
4%
6%
8%
10%
12%
14%
16%
São Paulo(2008)
New York City(2006)
Per
cent
of S
choo
l Day
s
Average Teacher Absence Rate
Source: São Paulo figure is based on the 2008 school year from Veja interview with Paulo Renato Souza on 10/24/09. New York City figure is from Chicago Tribune article “Teachers miss days; poor kids miss out; Educators at some struggling schools take most time off, analysis shows, September 25, 2006.Teacher quote from “Paralisação dos professores em SP tem baixa adesão”, Estado de São Paulo, March 9, 2010
Recent Policies in São Paulo Have Focused on Teacher Attendance
• Teacher Bonus: beginning in 2008, teacher performance bonuses of up to 22% of salary are proportional to their attendance (67% attendance minimum)
• Teacher Raises: beginning in 2010, teachers must meet attendance minimums to be eligible for raises of 25%
• “I don’t even think of going on strike, because if I am absent, I lose the one chance of earning a real raise”
• São Paulo Public School Teacher, March 9, 2010.
13
More high school students in São Paulo go to school at night
43.6%
5.7% 5.5%
0%
5%
10%
15%
20%
25%
30%
35%
40%
45%
50%
São Paulo State Schools São Paulo Private Schools New York City PublicSchools
Enrollment: 981,767 132,593 331,437
Per
cent
of H
igh
Sch
ool S
tude
nts
Source: Censo Escolar 2009 (Sinopse Table 2.16); DOE ATS System as of 4/02/10
14
• Students in São Paulo go to school for fewer hours
– The school year is longer (200 days in São Paulo vs. 185 days in New York)
– The school day for the student is shorter (4 hours to 5 hours and 20 minutes in São Paulo vs. 6 to 7 hours in New York)
Other observations about students
• Students in São Paulo take more courses at the same time
– São Paulo: students are normally enrolled in 10 to 13 courses
– New York: students are normally enrolled in 5 to 7 courses
15
• Students in São Paulo
V go to school for fewer hours each week
V and study more subjects at the same time
• Teachers in São Paulo
V teach more hours each week
V have larger class sizes
V and are more likely to teach in multiple schools
Middle and High School Teachers in São Paulo have a substantially higher student load than in New York
Teachers have a much higher student load
Teachers need to know the names and track the progress of many more students
16
• General observations about São Paulo schools
– Shared characteristics with New York
– Notable differences
• Brief presentation of New York City’s accountability system
• School evaluation in São Paulo
– Best practices
– Opportunities for improvement
– Recommendations
How much are we learning?
17
New York: Accountability Framework
WH
AT
EVALUATE ENFORCE CONSEQUENCES ENABLE
HO
W
Progress ReportsGrades based on student outcomes
Learning Environment SurveyParent, teacher, and student surveys
Quality ReviewsScores based on performance management criteria
DemandEnrollment demand
Rewards Monetary bonuses for principals and a pilot program for teachers
ConsequencesImmediate restructuring of chronically failing schools. Target setting for other D/F schools; if no improvement, leadership change and/or closure
EmpowermentPrincipals have more autonomy over budget, hiring and professional development
Periodic Assessments Diagnose and track progress
Achievement Reporting and Innovation System (ARIS)Fully integrated knowledge and data management
Children First IntensiveHands-on data training through Inquiry Teams
18
The Progress Report was designed with several guiding principles in mind:
• Outcomes, not inputs: Focus the conversation on student outcomes rather than inputs
• Performance and Progress: Measure performance and progress of individual students
• City and Peer Comparison: Compare schools to similar schools as well as a citywide standard
• Simple, Valid, Verifiable: Ensure that schools can verify and re-create metrics, so schools understand how they are measured and how they can improve their performance
Progress Report guiding principles
19
Progress Report measures
School Environment
15 points
Student Performance
25 points
Student Progress
60 points
Additional Credit
Up to 15 points
• Learning Environment Survey results
• Attendance
• Student test scores in ELA and Math (median proficiency and % Level 3/4)
• Individual student progress on ELA and Math test scores
• Graduation rates (4-year and 6-year)
• More points for diplomas indicating college-readiness
• Exemplary progress on test scores with high need students
• Exemplary progress in credit gains with high need students
• Learning Environment Survey results
• Attendance
• Credit accumulation
• Regents completion and pass rates
Elementary, Middle, and K-8 Schools
High Schools
20
• General observations about São Paulo schools
– Shared characteristics with New York
– Notable differences
• Brief presentation of New York City’s accountability system
• School evaluation in São Paulo
– Best practices
– Opportunities for improvement
– Recommendations
How much are we learning?
21
SARESP is a good instrument to measure student learning
• Performance Scale: Uses Item Response Theory (IRT) to assign each student a score for each subject, based on a scale from 0 to 500, with the majority of students between 125 and 400
• Proficiency Levels: In each subject, the scale is divided into four levels:
• Advanced
• Adequate
• Basic
• Below Basic
22
IDESP is a simple way to combine two different objectives: improve student performance and improve promotion rate
• IDESP is based on two indicators:
• Performance Indicator: Based on the proficiency levels of students on standardized Portuguese Language and Math tests
• Promotion Indicator: Grade promotion rate of the students
IDESP = Performance Indicator X Promotion
Indicator
23
IDESP has a well thought out system of long and short-term targets
Source: IDESP Technical Note, Secretaria Estuadual de Educação do Estado de São Paulo
• Long term targets (2030): The São Paulo State Education Department (SEE-SP) established long-term goals based on simulating IDESP results for developed countries in the OECD
• School-specific targets: SEE-SP established intermediate targets for each school, based on the distance between its starting point and the long-term goal
24
IDESP is relatively simple to calculate, though the State’s current explanation could be improved
The Performance Indicator (ID) isbased on the proficiency levels• Av = percentage of students at the advanced level • Ad = percentage of students at the adequate level• B = percentage of students at the basic level• AB = percentage of students below basic (abaixo do básico)
A. Two-step formula described in the Nota Tecnica de IDESP
Step 1: Calculate “Desfasagem” Step 2: Calculate Performance Indicator
B. Equivalent one step formula
IDjs = (10 x Av) + (6,66 x Ad) + (3,33 x B) + (0 x AB)
Note: an alternative formula for “B” would be IDjs = (10 x Av) + (20/3 x Ad) + (10/3 x B) + (0 x AB)
25
Positive: IDESP is given strong teeth by the bonus program
22.3%
4.6%
0%
5%
10%
15%
20%
25%
São Paulo State New York City
Per
cent
of A
nnua
l Sal
ary
Maximum Bonus as a Percent of Average Annual Teacher Salary
Fonte: New York City bonus is $3,000 per teacher provided to schools in a 200-school pilot project that meet Progress Report targets. São Paulo bonusis up to 2.9 additional monthly salaries for teachers whose schools improve their IDESP scoreSource: teacher salaries based on PNAD 2008 for São Paulo and Bureau of Labor Statistics May 2008 Occupational Employment Statistics survey
26
• General observations about São Paulo schools
– Shared characteristics with New York
– Notable differences
• Brief presentation of New York City’s accountability system
• School evaluation in São Paulo
– Best practices
– Opportunities for improvement
– Recommendations
How much are we learning?
27
Change in IDESP from 2007 to 2008
0
1
2
3
4
5
6
7
8
9
10
0 1 2 3 4 5 6 7 8 9 10
IDESP 2007
IDES
P 20
08
Note: each school is represented by a separate point.Source: Boletim IDESP
0
1
2
3
4
5
6
7
8
9
10
0 1 2 3 4 5 6 7 8 9 10
IDESP 2008
IDES
P 20
09
28
Change in IDESP from 2008 to 2009
Recommendation: Interview school personnel to identify best practices and review testing micro-data to identify potential weaknesses in the reliability of test score and passing rate indicators.
Note: each school is represented by a separate point.Source: Boletim IDESP
29
Recommendation: incorporate percentage of students taking the test into IDESP, potentially by multiplying the test taking percentage by the IDESP score
There is no incentive for weaker students to take the test
Source: Boletim SARESP 2008; Censo Escolar 2008; Author interviews.
93%85%
78%
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
5th Grade 9th Grade 12th Grade
Average Test Taking Percentage (SARESP 2008)
Test Taking Percentage by School (SARESP 2008, Grade 9)
• “There were many schools that only sent the good students, and if you only send the good students, your IDESP is better. The great failure of IDESP is that it does not consider the number of students who take the test”
-- High School Principal, São Paulo State
0
50
100
150
200
250
40% 50% 60% 70% 80% 90% 100%
30
“SARESP is not a tool for improving education...we don’t have access to the student test results”
- High School Teacher, São Paulo State
Individual student or item results are not provided to the schools, limiting SARESP’s use in improving education
Recommendation: Develop a plan to provide individual test results to schools and parents.
“Who are the students who are below basic; we don’t know What did they get wrong; we don’t know...How am I going to work with the students this year if I don’t know what they got wrong”
- High School Curriculum Director, São Paulo State
Source: author interviews
31Note: parent survey is given once per year to parents of all students who take the SARESP (5 th, 7th, 9th, 12th grades)Source: SARESP 2008 Questionário dos pais
The parent survey asks many questions that can paint a picture of a school’s learning environment
Parents: To what degree to you agree or disagree with the following statements:
Communication with parents• 2.1. I receive information from the school about my child’s progress.• 2.11. The school always holds parent meetings to inform me about my child.
Parent Involvement• 2.17. When there is a problem I am quickly called to the school.• 2.19. The school recognizes that parents’ opinions are important.
Student Safety• 2.3. My child is safe in school.• 2.16. My child feels safe in school.• 2.23. This school has many students with behavioral problems.
Opinion about teachers• 2.4. The school’s teachers respect the students.• 2.12. I believe the teachers are very capable.
Opinion about the school• 2.15. I would like my child to study in another school.• 2.24. If I could pay, my child would study in private school.
32Source: SARESP 2008 Questionário dos pais
48%40%
57%
30%
36%41%
38%
23%
17% 19%5%
48%
58%
32%
10%
0%10%20%30%40%50%60%70%80%90%
100%
2.1. I receiveinformation fromthe school about
my child’sprogress
2.19. Theschool
recognizes thatparents’
opinions areimportant
2.3. My child issafe in school
2.12. I believethe teachers are
very capable
2.15. I wouldlike my child to
study in anotherschool
Completely Agree
Partly Agree
Disagree
1,174,660 1,119,737 1,130,132 1,195,661 1,136,501
The survey could give valuable system-wide information for the State of São Paulo
Per
cent
age
of R
espo
nden
ts
33Source: SARESP 2008 Questionario dos pais
58%
48%
40%
57%
30%
77%
61%
69% 71%
9%
27% 29%
22%
37%
43%
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
2.1. I receiveinformation fromthe school about
my child’sprogress
2.19. The schoolrecognizes that
parents’ opinionsare important
2.3. My child issafe in school
2.12. I believethe teachers are
very capable
2.15. I would likemy child to studyin another school
Per
cent
age
that
Com
plet
ely
Agr
ee
São
Pau
lo S
tate
Sch
ool A
Sch
ool B
Recommendation: publish results of parent, teacher, and student surveys, aggregating results to maintain confidentiality
The survey could provide parents, teachers, and students with an insider’s perspective into school quality
34
4.5
5.25.7
6.1
0
1
2
3
4
5
6
7
D C B A
N* = 2 11 138 886
NYC Simulation: On average, schools with better Progress Report letter grades had higher simulated IDESP scores in 2009
Sim
ulat
ed ID
ES
P (2
009)
Note: Includes Ensino Fundamental I (4th grade) and Ensino Fundamental II (8th grade) separately, so schools with both grades may appear twice in the chart.
IDESP Simulation: However, there was substantial variation around these trends
R = 0.20
0
20
40
60
80
100
120
140
2 3 4 5 6 7 8 9 10
IDESP Simulado (2009)
Res
ulta
do n
o B
olet
im d
e Pr
ogre
sso
(200
9)
Note: Includes Ensino Fundamental I (4th grade) and Ensino Fundamental II (8th grade) separately, so schools with both grades may appear twice in the chart.
Simulated IDESP Score (2009)
Pro
gres
s R
epor
t Ove
rall
Sco
re (2
009)
36
1. The two instruments measure different things
• NYC Progress Reports include parent survey as 10% of the grade
• IDESP includes student grade promotion rates
2. NYC Progress Reports include more students
• IDESP only includes the last year of elementary school and last year of middle school
• Progress Report includes all students that take the test: 3rd – 5th in elementary school and 6th – 8th in middle school
3. 75% of the NYC Progress Report is based on a comparison between similar schools; IDESP is calculated equally for all schools in the state.
4. The NYC Progress Report measures students’ individual progress; the change in IDESP compares the results of one group of students in a particular grade one year with a different group of students in the same grade the following year
Hypotheses to explain difference between IDESP and NYC Progress Reports
Recommendations: Consider “Additional Credit” for: • Individual student progress• Performance relative to similar schools
37
Recommendation: Continue developing tools to help principals, teachers, and other school professionals do their job better
• Excel tool that allows the school to simulate student test performance and grade promotion results and see the impact on target achievement and teacher bonus
IDESP Modelo
Escola:
Lingua Portuguesa 2009 Atual* 2010 Proj. 2009 Atual* 2010 Proj. 2009 Atual* 2010 Proj. EF1 EF2 EMAvançado (AV) 4% 4% 1% 1% 0% 0% IDESP 2009 2.93 2.08 0.79Adequado (AD) 21% 21% 11% 11% 17% 17% Meta 2010 3.09 2.21 0.88
Básico (B) 48% 48% 56% 56% 31% 31% Proj 2010 2.88 2.06 0.79Abaixo do Básico (AB) 26% 27% 32% 32% 52% 52%
Total 99% 100% 100% 100% 100% 100%
Indicador de Desempenho (LP) 3.4615 3.400 2.7451 2.700 2.1839 2.167
MatimáticasAvançado (AV) 2% 2% 1% 1% 0% 0%Adequado (AD) 14% 14% 4% 4% 2% 2%
Básico (B) 48% 48% 58% 58% 19% 19%Abaixo do Básico (AB) 36% 36% 37% 37% 79% 79%
Total 100% 100% 100% 100% 100% 100%
Indicador de Desempenho (Mat) 2.7677 2.733 2.3039 2.300 0.7471 0.767
Taixa de Promoção 94.0% 94.0% 82.4% 82.4% 54.0% 54.0%
IDESP 2.93 2.88 2.08 2.06 0.79 0.79
Meta para 2010 3.09 2.21 0.88
% de meta alcançada 0% 0% 2%
* No cálculo atual de 2009, os porcentagems em cada nível estão com uma grau de precisão maior de o que esta disponivél aqui Isso resulta que as sumas das porcentagems pode não igualar 100% por causa de "rounding"
Ensino Fundamental I Ensino Fundamental II Ensino Médio
2.93
2.08
0.79
3.09
2.21
0.88
2.88
2.06
0.79
0.00
0.50
1.00
1.50
2.00
2.50
3.00
3.50
EF1 EF2 EM
IDES
P
IDESP 2009Meta 2010Proj 2010
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
I understand how the IDESP isrelated to the teacher bonus
Per
cent
of R
espo
nden
ts
n = 37
Completely agree
Agree
Disagree
Completely disagree
Survey with Teacher Coordinators (April, 2010) IDESP Modeler
38
• General observations about São Paulo schools
– Shared characteristics with New York
– Notable differences
• Brief presentation of New York City’s accountability system
• School evaluation in São Paulo
– Best practices
– Opportunities for improvement
– Recommendations
How much are we learning?
39
• Short-term
• Simplify the explanation of IDESP and implement active training effort so that school staff understand the calculation
• Further study why certain schools had large jumps in IDESP to identify best practices or test-security weaknesses
• Incorporate the percentage of students taking the test into the calculation of IDESP
• Provide school-wide parent, teacher and student survey results to the public, aggregating results at the school level to maintain confidentiality
• Long-term
• Develop a plan to provide individual student test results to schools and parents
• Consider incorporating individual student progress and performance relative to similar schools into the measure
• Continue developing tools to help school principals, teachers and other staff do their job better
Recommendations