Title: The use of formative assessments: Perspectives from district and school leaders
1The use of formative assessmentsPerspectives
from district and school leaders
Leslie Nabors Oláh Consortium for Policy
Research in Education, University of Pennsylvania
CCSSO Summer Leadership Training
Conference, Boulder, CO
August 2, 2006
leslieno_at_gse.upenn.edu
2Acknowledgements
- This presentation is based upon work supported
by the National Science Foundation under contract
number (REC-0529485). Opinions expressed in this
presentation are those of the author and do not
necessarily reflect the views of the National
Science Foundation. - Margaret Goertz, Nancy Lawrence, and John
Weathers (CPRE) have contributed greatly to this
research. We would like to thank the two
participating school districts for their time and
for their thoughtful responses to our questions.
3Research Overview
- Framework based on the new Center on Continuous
Instructional Improvement (CCII), funded by the
Hewlett Foundation. - The purpose of assessment is to inform decisions
about how to respond pedagogically to students
current state in ways that increase learning.
4 Cycle of Instructional Improvement
DISTRICT CONTEXT
Teaching and learning Curriculum Goals Assessm
ent use Incentives for improvement Data systems
Accountability State District School
Professionalization Professional
development Norms about assessment
Interpret evidence
Use evidence to improve instruction
(differentiate, etc.)
Cycle of Improvement
Gather evidence about standard achievement and
student learning
Carry out improved instruction
School-level culture and routines
Data
School leadership
Schedules
Professional development
SCHOOL CONTEXT
Marshall S. Smith, Hewlett Foundation, 2004
5Common Language
- Formative assessment
- Benchmark assessment
- Coach
6Research Questions
- How do elementary teachers learn from formative
assessment data in mathematics and apply that
knowledge to instructional decisions about
content and pedagogy? - What policies and conditions affect the ability
of teachers to learn from and make use of
assessment data?
7Sites
- Nine schools chosen based on AYP status, math
performance, and socio-demographic
characteristics. - Abington, PA
- Small, suburban district (7 elementary schools).
- Has used Everyday Math (EDM) since 1991.
- District-created formative assessments every 4-6
weeks. - Philadelphia, PA
- Large, urban district (177 elementary schools).
- Has used Everyday Math (EDM) since 2003.
- One school uses Investigations on waiver.
- Co-created (district and Princeton Review)
benchmarks given every six weeks. - Scores posted on SchoolNet Information Management
System. - Monthly SchoolStat principal meetings with
district leaders.
8Analysis
- Findings based on initial round of interviews
with district leaders, school principals, and
math coaches in Spring 2006. - Qualitative analysis of participant interviews.
- We will observe instruction and interview
teachers in SY 2006-2007.
9Cycle of Assessment and Instruction Abington, PA
Day 28-29 Teachers receive practice end-of-unit
test (benchmark) from math science
coordinator (MSC) (graded, but not counted) a
few days before end-of-unit test. Teachers
administer and score practice test.
Day 1-28 Teachers follow EDM curriculum.
Day 30 Low scores may be reported to ECS or Math
Coach to help with remediation.
Day 30 Teachers use flexible grouping based on
test results.
Day 39 ECSers and teachers (with or without MSC)
may meet to discuss instruction, individual
remediation, or further professional development
as a result of test scores.
Day 39 MSC reviews scores with ECSers once a
month, with an eye toward concepts, classes,
individual remediation, enrichment, and
individual items.
Day 30-32 Teachers review material for 2-3 days.
Day 38 Teachers examine data individually for
flexible grouping.
Day 33 Teachers receive end-of-unit test
(benchmark) from math science coordinator and
administer to students.
Day 38 ECSers print out spreadsheets (by class)
and put in binder for MSC.
Day 35-37 Teachers enter scores into customized
spreadsheet and email to their Elementary
Curriculum Specialists (ECS) within one week of
test administration.
Day 34 Teachers score test.
10Cycle of Assessment and Instruction
Philadelphia, PA
Days 26-30 Teachers are encouraged to review or
extend math content.
Days 1-25 Teachers follow EDM curriculum.
Day 31-35 Teachers either receive and administer
paper-and-pencil tests or students take benchmark
assessments online (Princeton Review/SchoolNet).
Days 37-43 Based on item analysis, teachers
re-teach as needed. Teachers are encouraged to
collaborate on data interpretation. The pacing
guide allows for six days of review before
teachers must proceed with the next instructional
unit.
Day 36 Princeton Review scores benchmark
assessments and scores are posted online
(SchoolNet).
Day 37 Teachers and principals complete their
respective Benchmark Data Analysis Protocols.
Teachers should turn in worksheet to their
principal.
Day 36 Parents can access childrens scores
through SchoolNet.
Day 36 Principals and coaches can access scores
through SchoolNet.
Day 36 Teachers can access their students
scores through SchoolNet. They can perform item
analyses, group by content area or can access
individual student performance.
11Overview of findings
- What are the expectations for benchmark
assessment use? - How are the assessments aligned with the
curriculum and with other assessments? - What are the current supports for assessment use?
- How are data used to improve instruction?
- What constraints affect optimal implementation of
the benchmark assessment system?
12Expectations for assessment use
- Expectations for district
- Create instructionally actionable assessments
- Identify district-wide PD needs.
- Expectations for principals
- Identify children needing remediation
- Identify school PD needs
- Follow up on teachers data-driven instruction
- Inform School Improvement Plan
- Collaborate with fellow principals on how to
raise student achievement.
13Expectations for assessment use
- Expectations for teachers
- Identify children needing remediation and provide
appropriate instruction - Identify children qualifying for enrichment and
provide appropriate instruction - Review test with class to provide all students
with feedback - Form flex groups and provide groups with
appropriate instruction - Re-teach to whole group
- Expand instructional repertoire
- Use data as a vehicle for collaboration with
fellow teachers - Ask school leaders for instructional support
- Plan or modify curriculum
- Communicate with parents about assessment results
(if child not doing well) - Reflect on instruction.
14Alignment
- District curriculum and assessment directors take
responsibility for aligning the benchmark
assessments with state standards and with
Everyday Math. Test construction occurs over the
summer (Abington) or in weeks prior to
administration (Philadelphia). - In Abington, math coaches contribute to test
construction, while there is little participation
from school-based coaches or teachers in
Philadelphia. - In Philadelphia, the assessment goes through
several iterations between the district and
Princeton Review.
15Alignment
- Concepts covered on the benchmark call attention
to areas that teachers need to emphasize - Math Coach Even the kindergarten and first
grade teacher didnt realize, Oh, I never
realized what we do with symmetry was so
important. Just because its not covered in
third grade, well thats where theyre getting
their background for the PSSA, from you. - Principal They were more aware. They just
have to really reflect what they do. And I think
they got to see that by giving the test.
16Alignment
- Standards-based benchmark assessments have lead
some teachers to shift from norm-referenced
evaluation to criterion-referenced feedback. - Principal Typically here the teachers have been
pretty, I would say, rigorous graders. But it was
easier for them to be rigorous graders with A, B,
C. I think the shift is making it a little bit
more challenging, because its standard-based.
You know, A, B, C-- What does a D in English or
reading or math mean? What does that mean? What
does that tell me that they know or they dont
know? You know, its just like-- Like, its good
as an average, which doesnt give you much. Its
just like taking all the colors and making it
grey.
17Alignment
- Districts have had to find ways to address format
non-alignment between the benchmark assessment
and the state test. The Abington curriculum team
creates additional open-ended items. Philadelphia
has contracted with a local university to provide
access to open-ended problems. - District leaders are looking for open-ended
supplements to benchmark assessments, but have
not found a provider who can score items in an
adequate amount of time.
18Supports
- In the large district, SchoolStat meetings
provide support for principal data use, but
contribute to accountability pressure. - Principal Benchmarks, when they first came
out, it was something that was going to be used
just by the schools. Its grown. Theyve almost
became lower PSSAs, because everybody looks at
it. And so, the pressure is on with these
benchmarks. Whereas before, it was kind of like
we were going to use it, and nobodys going to
know about it but us.
19Supports
- Professional development and support for data use
is seen as part of the everyday work of the
school math coach or computer support personnel. - Coaches, however, have many other
responsibilities - Conducting school and district PD
- Teaching (own class, pull-out, covering)
- Collaborate with school psychology team
- Discipline
- Lunchroom duty
- Tech support
- Whatever the principal needs
jack-of-all-trades.
20Supports
- Perhaps as a result, professional development for
data use remains undefined and focused on
gathering evidence and interpreting data. - Still, we found some examples of support and
planning for improved instruction. -
- Math coach Sometimes were able to talk it
through, because Ill say, You know, what do you
think is happening here? And a lot of times, it
requires some talking it through. I might say,
Well, how much practice did they get in this?
Was it presented in homework or in a daily-- You
know, in your daily math message in the same way?
You know, how much practice did they have with it
in the same way it was presented on the test?
...And Ill say, How many times throughout your
instruction, throughout this story, did these
kids have to answer, in writing, an open-ended
question? laughter And thats the response I
get. So, and it sort of opens their eyes, because
theyre doing what theyre told, you know, and
theyre doing it the way the book is telling
them. And thats sometimes a problem. And when
you say, you know, when they look at me and they
dont understand because they did exactly what
the book told them to do. And sometimes teaching
requires you to go beyond that and pull from a
bag of tricks.
21Data Use
- District and principal reaction to the benchmark
assessment use tends to be positive overall. - Philadelphia credits assessments with a large
part of rising student achievement. - Seen as a lens into the classroom.
- However, principals in Philadelphia note that
SchoolStat meetings have changed the low-stakes
benchmark assessments into high-stakes tests. - In addition, principals often view benchmark
assessment data with an eye toward AYP.
22Data Use
- Data used primarily for remediation, which is
addressed largely through flex grouping or
re-teaching. - Benchmark assessment data used to construct flex
groups may be based on only a few items. - Flex groups may actually remain stable over long
periods of time, even years. -
-
23Data Use
- Data also used for accountability.
- In some schools, accountability pressures may be
the driving force behind data analysis. -
- Principal Well, we have to be mindful of it
because-- It is monitored. It is monitored.
Everything is monitored now. Because everything
is online. It is monitored .
24Data Use
- Data also used for accountability.
- But, accountability not an intended use
- Curriculum Director I think benchmark tests are
being misused at the classroom level. If were
misusing it and were comparing schools and were
comparing teachers and were comparing classrooms
then theres pressure on teachers to perform. And
so as soon as theres pressure on teachers to
perform, that frequently is transferred to
students which is not what this is intended to
be
25Data Use
- Principals who are instructional leaders as well
as data users contribute to a culture of data
use. Data not seen as an end in itself, but
rather as a means to an end (e.g., discussing
item analyses at grade-group meetings). - But there is variation in the degree to which
principals directly engage with data. Some
principals use a designate, while others may
spend hours looking at data. -
-
26Data Use
- Three common interpretation mistakes were
mentioned - Common practice of using benchmarks as indicator
of growth over time, in spite of validity and
reliability problems with doing so. - Widespread belief that better benchmark
assessment scores lead to better state-test
performance in spite of lack of predictive
validity evidence. - Errors in aggregation and disaggregation of
scores.
27Implementation Issues
- Teachers lack of time restricts attention
devoted to data. -
- Math Coach We dont have enough time for
collaboration. It either ends up being first
thing in the morning, on the fly after school,
trying to catch them on their prep or their
lunch. And thats not always fair either. You
know, we have that one collaboration time a
month.
28Implementation Issues
- Pacing is seen to have a great impact on the
cycle of assessment. - Disagreements have emerged over administration of
testing (i.e., pencil and paper versus online
administration). Curriculum leaders and
principals argue for value of keeping student
work, while IT argues for expediency of online
administration. - District and school leaders noted aspects of the
benchmark assessment system that have attracted
the attention of the Philadelphia Federation of
Teachers.
29Implementation Issues
- The pedagogical content knowledge of teachers may
be limited, affecting their ability to interpret
data and adjust instruction. In Philadelphia,
math content knowledge of elementary school
teachers has been classified anywhere from a
disaster to good. - Funding for professional development has been cut
in Philadelphia. - The regional math coach position has been
eliminated, to be replaced with school-based
growth coaches assigned to school not making
AYP. - Summer 2006 PD has been cancelled.
30Summary
Interpret evidence
Gather evidence about standard achievement and
student learning
Use evidence to improve instruction.
Cycle of Improvement
Carry out improved instruction
School and district leaders realize that
data-driven instruction is here to stay. Do
teachers feel the same way?
31Summary
Interpret evidence
Gather evidence about standard achievement and
student learning
Use evidence to improve instruction.
Cycle of Improvement
Carry out improved instruction
Districts and schools have numerous expectations
for data use, but tend to use data in more
limited ways
32Summary
Interpret evidence
Gather evidence about standard achievement and
student learning
Use evidence to improve instruction.
Cycle of Improvement
Carry out improved instruction
Support for data-driven instruction seems to wane
during the last two or three steps of the cycle.
33Summary
Interpret evidence
Gather evidence about standard achievement and
student learning
Use evidence to improve instruction.
Cycle of Improvement
Carry out improved instruction
Perhaps leading to common interpretation errors.
34Summary
Interpret evidence
Gather evidence about standard achievement and
student learning
Use evidence to improve instruction
Cycle of Improvement
Carry out improved instruction
Still, district and school leaders report
instructional change as a result of benchmark
assessment data use. We will follow this up next
year.
35The end?