|
SuperTECH
NEWS is the bi-monthly
newsletter of the BLE GROUP, which provides small- and medium-size
school systems with supplementary technology management to
produce high-quality educational results and efficient management.
The
purpose of SuperTECH NEWS is to provide education decision
makers with concise information that allows them to make informed
technology decisions to impact instruction, management and
communication. This is information you can use
on Monday morning.
Editor,
Susan DeMark
Web Design, Charlene
Polanosky
Publisher, Eliot Levinson
|
Our
February issue theme is Web-based assessment
products for high-stakes tests. Choose from the following
articles.
|
|
Note
from EliotLaunching the new No Child
Left Behind (NCLB) Assessment and Management Service.
The BLE Group is launching a new service for small and
medium-size school systems. We assess where school districts
are currently and exactly what they have to do to implement
NCLB. The service also provides school systems the supplementary
management support they need to purchase and implement
the new technology-delivered programs. |
|
|
Theme
of the MonthOnline high-stakes testing.
Online testing is the first line of products many school
systems will buy to address NCLB. These products pinpoint
student performance, deliver test results instantaneously,
and inform instruction for every student. It is not simple
to purchase these new products. This section gives you
information on how to intelligently purchase and implement
these products. |
|
|
ProductsWithin
the next two years, many school districts will adopt
the use of online assessment products aligned with state
standards and high-stakes state tests. We provide
you with in-depth, concise, and vendor-neutral descriptions
of a number of leading Web-based and Web-enabled assessment
products.
|
|
|
Best
Practices Noni Miller, who heads
educational services in Ypsilanti, Michigan, Public Schools,
shares lessons on how her district is using a leading
online standards-based assessment program to drive instruction
throughout the school system and to improve students'
reading and math proficiency. |
|
|
ConferencesCheck
out the relevant conferences coming in the next several
months. |
We
want to hear from you. What do you agree and disagree with
on this issue (we will post comments from readers in the next
issue). Please write us at eliot@blegroup.com.
DID
YOU MISS AN ISSUE?
You
can read past issues of SuperTECH NEWS online:
- December
2002
- Focus: Student Information Systems
- August
2002 - Focus: Purchasing Hardware 2002
- June
2002
- Focus: Web-based Applications for Early Reading
- May
2002 - Focus: Web-based Professional Development
- March
2002 - Focus: Technology of Accountability
|
THE
BLE GROUP AND NCLB ASSESSMENT AND MANAGEMENT SERVICE
WHO
IS THE BLE GROUP? We're a group of 25 CIOs and curriculum
directors of school systems who use technology to improve instruction
and management. The BLE Group has three lines of business:
- We
develop technology assessments and plans, and we provide management
services in more than 40 school systems.
- We
publish a newsletter, Super TECH NEWS, which offers senior administrators
easy-to-understand information on making technology decisions.
- We
conduct market research for technology firms on the appropriateness
of technology products for K-12 school systems.
Eliot
Levinson is the CEO of the BLE Group. Levinson founded the BLE
Group (www.blegroup.com)
in 1998. Levinson is known nationally for his work in technology
planning and management for school districts. He co-authors "Tech
from the Top," a monthly column that appears in Converge
Magazine. Levinson has experience in education and technology
as a teacher in California and Pennsylvania, a middle school principal
in Massachusetts, and an assistant to the chancellor of schools
in New York City. He has held research positions in educational
change at the RAND Corporation and MIT's Sloan School of Management.
Levinson holds master's degrees in Education and Anthropology and
a Ph.D. in Organizational Studies from Stanford University. Levinson
works as a strategic technology advisor to large school systems
and consults with several firms in the education technology market.
THE
BLE Group's principals, our leadership team, consists of:
- Eliot
LevinsonCEO, BLE Group
- Rick
RozzelleFormer CIO, Charlotte-Mecklenberg Schools, North
Carolina
- Charles
GartenExecutive Director, Educational Technology and
Information Services, Poway Unified School District, California
- Kenneth
EastwoodSuperintendent, Oswego City School District,
New York
- Ann
BoyleAssistant Superintendent of Curriculum, Instruction,
Assessment, and Technology, Scottsdale Unified School District,
Arizona
- Steve
FinchFormer CIO, Oklahoma City Public Schools, Oklahoma
|
THE
BLE Group's NCLB Assessment and Management Service
Why
are we launching the NCLB Assessment and Management Service?
Technology is necessary to implement the No Child Left
Behind Act. Technology is a central component to the solution
of every facet of NCLB, whether it is teacher quality, the
delivery of standards-based instruction, assessment, the monitoring
of student progress, school-parent communication, or reading
proficiency.
The
BLE Group knows that small- and medium-size school systemsthe
86 percent of school districts in the U.S. with fewer than
5,000 studentsoften lack the extensive resources and
knowledge base to implement NCLB. They lack sufficient technologists
and technology-savvy educators to plan and manage the Web-based
instructional programs and assessment systems that are the
solutions for NCLB. We make available reasonably priced services
that can provide the expertise school systems need to address
NCLB on a time-shared basis.
Excellent
technology staff is expensive and hard to find. We've created
the NCLB Assessment and Management Service as a means of supplementing
the staffs of small school systems with our own team of skilled
technologists and technology-savvy educators. We will help
you plan and execute an effective NCLB program.
The
NCLB Assessment and Management Service supplements
the instructional and evaluation capability of small and mid-sized
school systems so that they can effectively address No Child
Left Behind. There are two tiers to the NCLB Assessment and
Management Service. Tier 1the NCLB assessment
and plancreates an assessment for districts on how
effectively they are currently addressing the multiple requirements
of NCLB, such as teacher quality, assessment, reading achievement,
etc., and devises a specific plan to address NCLB. The plan
includes new technology-based solutions, a schedule, and a
timeline for addressing NCLB. Tier 2the management
servicesupplies ongoing management support to districts.
We furnish districts with ongoing service from the BLE Group
to purchase products and provide supplementary management
as districts implement their NCLB programs.
The
NCLB Assessment and Management Service evaluates how well
are you currently addressing NCLB and delivering on its mandates,
and it centers on exactly what you should do over the next
year to implement NCLB effectively so that your district's
performance improves.
Why
are the specific areas of the BLE Group's NCLB assessment
and plan?
The
BLE Group provides an assessment and solution for the following
NCLB requirements:
- ReadingIncludes
benchmarks, diagnostic testing
- Teacher
QualityCertification, paraprofessional certification,
online training
- TestingState
standards, diagnostic testing
- Staff
developmentWhat is needed to meet certification,
improve standards-based teaching, address technology skills
linked to teaching
- ParaprofessionalsTracking
certification
- Management
of NCLBPlanning for low-performing schools
- Information
AnalysisThe know-how to aggregate and disaggregate
scores
- Grant
proposalsWhat information is needed for the annual
district proposal to include all students
- State
accountabilityWhat does the state have to do to
improve accountability
The
NCLB assessment and plan focuses on instruction, assessment,
management systems, and technology. It includes:
- An
annual implementation plan. Quarter by quarter, the plan
lays out what has to be done in each of the 4 areas described
above.
- A
budget.
- An
assessment of the current state of NCLB linked with specific
recommendations on the items listed above.
NCLB
Supplementary Management Service
Following
are the supplementary management services that districts can
make use of after the BLE Group assessment and plan. BLE Group
CIOs and curriculum directors will supplement the district's
staff with the following services:
- RFPs.
For strategic systems purchases such as instructional management
and on line assessment systems.
- Review
of contracts.
BLE will review district technology contracts and write
effective contracts for the district.
- Monthly
phone consultations and quarterly visits to address NCLB
management.
- Vendor
Management. BLE Group will oversee your NCLB vendors.
- Access
to databases on instructional and administrative systems.
BLE Group maintains confidential databases on management
and instructional software for its' clients.
- Discounts
from collaborative buying of hardware and instructional,
assessment, and management software.
- SuperTECH
NEWS newsletter. The newsletter delivers information
to administrators on NCLB-related technology issues such
as assessment, data warehousing, and instructional management.
If
you are interested in the NCLB Assessment and Management Service,
please contact us to discuss the matter further. The cost
is reasonable.
Eliot
Levinson <eliot@blegroup.com>,
CEO,
THE BLE GROUP
202.281.1763
|
Back to top
The
State of Online Assessments for High-Stakes Testing
NOTE:
This issue addresses high-stakes diagnostic and prescriptive tests
- those tests that directly relate to year-end state standards testing.
Low-stakes tests, or those related to textbook-based curriculum,
will be addressed in a future issue of SuperTECH News.
In
our February issue, we look at Web-delivered assessment products
for high-stakes testing. The convergence of No Child Left Behind's
rigorous accountability demands and powerful technology have driven
the development of ambitious assessment products that pinpoint student
performance, deliver test data instantaneously, and inform instruction
for every student. There is a Gold Rush on toward helping students
with these high-stakes tests and supporting districts in meeting
NCLB mandates, through online assessments. These products may all
seem equal in quality and thoroughness, but they are not. The
assessment products are very early on, and they differ with regard
to thoroughness, reliability, and alignment with your state standards
and tests.
Some
of the assessment products out there have evolved from old print
tests used for test prep and are pretenders, not solid, well-tested
assessments to rapidly and accurately assess, with meaningful data,
a student's achievement relative to state standards and skills.
Some of the newest and most evolved are, however, diagnostic tools.
They are reliably predictive, and they can tell you in-depth where
each student is weak and strong relative to standards - and will
drive instruction by showing how teachers and administrators can
produce personalized learning plans for each student. These are
not day-to-day tests that check whether students are learning what
is in the textbooks, but rather the makers of the tests intend that
they be linked to state high-stakes tests and the NCLB.
Among
the other key developments, companies are perfecting computer-adaptive
online assessments aligned to state standards that adjust as each
student takes a test, altering the level of difficulty based on
each answer. Adaptive tests are those that address the specific
performance level of each student. Those who are promoting adaptive
tests say, for one, that conventional tests fail to stack up when
it comes to testing low and high achievers. These diagnostic assessments
help prepare you for the end of the year and manage your instruction
throughout the year.
These
products are definitely worthy of consideration, but you need to
be careful in your selections. While many of the products sound
alike, there are distinct variations in technology delivery, product
quality, test reliability and validity, and vendor service and support.
Following is a description of the factors to be considered in purchasing
online assessments geared to high-stakes testing.
Technology
How does the assessment product fit with your district's technology
infrastructure? Do you have sufficient workstations and the other
system needs for multiple administrations of assessments and among
a number of classrooms? Are there definite advantages that online
administration provides to your district? Do you want to host the
data on your server, or do you want to outsource hosting services
to the company from whom you purchase the assessments? How much
support does the company provide in the beginning as you develop
the technology infrastructure? Is the product easy to use? It is
critically important to know whether the assessment will meet your
district where the technology is, and give you enough flexibility
to meet your district's testing needs.
Quality
There is little worse than giving an assessment using poor quality
items or items that do not truly align with your state standards.
In the rush to get into the market, products are out there with
claims that they totally align with standards. But how thoroughly
do they? Companies now have databanks with thousands of items, yet
item quality is an issue. Take time to find out how educators and
other professionals are involved in the entire process of creating,
reviewing, testing, and revising assessments. Many companies will
say that educators are involved, but what does this really mean?
What are the backgrounds and educational credentials of those involved
in assessment creation? Take the time to find out whether teachers
are involved and how; whether there is an advisory panel for the
assessments; whether the items go through a well-established, thorough
item and bias review.
Reliability
and Validity
How do the companies whom you are considering talk about and ensure
the reliability of their assessments? This is extremely important,
given today's realities of high-stakes tests and NCLB.
- How
reliable is the assessment? An effective assessment instrument
gives a consistent and trustworthy measurement of what students
know. To what degree does the assessment provide a consistent
measure?
- How
does the company ensure the validity of its assessment products?
- Are
the results replicable? If you administered a test item three
times in a row, are you likely to get the same result?
- How
does the company do calibration for its measurements?
Training,
Service, and Support
The
companies developing online assessments offer a vast range of service
and support. Be sure to pay a lot of attention, because this could
shape how successful your district will be and whether you improve
academic performance with their assessment tools. Key among the
support is face-to-face workshops that show how to interpret data.
Other companies provide help in creating action plans arising out
of the assessment results, devising personalized learning plans,
and working with parents. Some companies incorporate such training
in their products, while others charge extra.
Now
let's examine some of the specific online assessment products in
the Products section.
Back to top
The
No Child Left Behind Act's accountability, high-stakes tests, and
the need for data on student performance multiple times per year
are prompting a rush of evolving online assessment products. These
standards-based products promise to pinpoint individual student
performance relative to standards and benchmarks with laser-like
precision. They make individual results available immediately to
help shape each student's instruction. Administrators and teachers
can aggregate data to see class, school, and district progress in
meeting specific standards and benchmarks. Within the next two years,
many more school districts will adopt the use of these products.
We
are providing you with concise and vendor-neutral descriptions of
a number of leading Web-based and Web-enabled assessment products.
This list is not a complete one, but it provides a look at the assessment
products from some of the well recognized, leading companies and
organizations. It's intended to guide you as you consider any purchase
in this arena, not only by providing some baseline information about
a number of the products, but also by helping you understand what
you should keep in mind. We compiled these summaries from interviews
with representatives of each company or organization and from information
they make available on their products.
We
examine the products through a number of different categories:
- Product
Summary and Features
- Item
Pool
- Reliability
- Training
- Cost
Structure
The
products evaluated include:
|
|
| Web
site address |
http://www.levingsco.com/ |
| Product
Summary and Features |
Levings
Learning, based in Oklahoma City, has developed Web-based
assessments for grades 3-12 in mathematics, English, social
science, science, and fine arts. The company has designed
the assessments to be fully aligned with state standards in
49 states (Iowa does not have separate state standards). The
company's package combines assessment, evaluation, and real-time
results. With the company's PASS Plan, teachers, parents,
and administrators can determine how students are progressing
using real-time data; assess whether students understand and
grasp the concepts required by standards; and gauge whether
the questions really align to standards and assess an individual
concept.
The
Levings Learning Internet-based program gives districts the
ability to create their own local assessments benchmarked
to state standards. Teachers can dynamically develop assessments
for their classes, using the company's pool of assessment
items. Levings Learning's tools also let districts import
their own tests into the program for use. The assessments
can be administered in a variety of ways: taken online on
regular computers; using handheld computers; or taken off-line,
scored with a scanner, and imported back into the online system.
Teachers can score them using online rubrics or they may generate
their own rubrics. The reporting tool built into the assessment
program permits teachers and administrators to aggregate and
disaggregate the data by varied categories.
Levings
Learning has districts in 19 states using the program, and
is aiming to be in all 50 states by mid-year.
|
| Item
Pool |
The
company's in-house team contains a large number of teachers
and curriculum developers, and an external group of consultants,
who include teachers and curriculum experts, continually adds
items to the assessment pool. Also, the company has licensed
approximately one-third of its items from publishers and other
sources, such as the Colorado-based Center for Performance
Assessment. Currently, about 25 percent of the items in the
program are ones developed by Levings, and 75 percent are
items added in by districts and states. There are approximately
26,000 items currently in the item database.
|
| Reliability |
To
assure reliability of its assessments, the company has two
strategies: It depends on the credibility and reliability
of the publishers with whom it partners. Also, the company
continually analyzes how items are performing relative to
state standards and revises items, based on feedback from
the districts and its analysis.
|
| Cost
Structure |
The
cost of Levings Learning program is on a per-student or per-site
basis. The price is all-inclusive, and there are no
additional costs for added modules.
|
|
|
| Web
site address |
http://www.edutest.com/products/ |
| Product
Summary and Features |
Lightspan
eduTest is a customizable, online benchmark assessment program
that assesses students relative to state and national standards.
The company's Assessment Builder offers a template-based tool
that guides teachers and educators in designing assessments
specific to district needs and state objectives. End users
can address the standards, assessments, and test parameters
that the district has established, and the program permits
them to have control over the length, subject matter, and
content of assessments. Essentially, the Lightspan products
couple test content with a tool that allows customized test
creation, and everything is Web-based with passwords and secure
technology.
EduTest
provides fixed benchmark tests that can be administered multiple
times during a school year, at the beginning, middle, and
end of the year. Teachers can obtain real-time student performance
data throughout the school year that enables them to target
instruction. Comprehensive reports allow them to identify
the needs of each student by specific standards, strands,
and objectives. Administrators can also access the data at
the district, school, and classroom level. How it works: Once
a district purchases the program, those in the district create
assessments using the Lightspan Assessment Builder; deliver
assessments through the eduTest assessment; and receive the
results. Teachers and students obtain feedback immediately,
and district administrators receive district and school-aggregated
scores within hours. Reports can be generated and provided
to parents in order to see how their child is progressing
against various objectives and standards.
Using
the Lightspan assessment, one can compare assessment results
to high-stakes tests and previous performance reports, and
analyze your district's data by gender, age, socioeconomic
status, and other criteria to comply with No Child Left Behind.
Student data can be disaggregated by these varied categories,
including gender, Title I, etc. Districts can upload and disaggregate
the data through the district's Student Information System.
The
assessments are criterion-referenced tests that report how
well students are performing relative to pre-determined levels
on specific goals or objectives. EduTest assessments cover
grades 2-8 in math, reading, and language arts.
|
| Item
Pool |
Users
can access a search engine with more than 60,000 test items,
and they may also develop their own test items for use.
Currently, the items can address state and national standards.
The item pool covers seven states with complete alignment;
Lightspan plans to launch complete coverage for all states
with state standards in the near future. The content writers
for Lightspan have educational backgrounds, and include teachers,
content specialists, and administrators.
|
| Reliability |
Lightspan
employs internal and external measures of reliability and
validity. All Lightspan tests and test items are reviewed
and developed following accepted standards for low-stakes,
criterion-referenced tests. A panel of educators and assessment
experts are employed to review and validate Lightspan's test
items. As a measure of test validity, correlation studies
have been conducted to examine the relationship between performance
on specific Lightspan eduTest assessment tests and state-mandated
criterion-referenced tests in multiple states.
To
examine test reliability, internal consistency estimates of
eduTest assessment test instruments have been calculated based
on test results from samples of students in various states.
In reliability studies, high internal consistency values were
observed for the Lightspan eduTest assessment benchmark tests
evaluated. In addition, Lightspan conducts alignment studies
to determine the validity of its assessments with state standards.
The company recently announced a partnership was with WestEd,
a California nonprofit education research agency, to review
a subset of Lightspan's assessment content.
|
| Training |
An
in-depth training component, with both in-person training and
ongoing support, is offered in the online assessment program.
It provides one day of on-site professional development
with subsequent consultation. Every school that participates
has access to an account service representative. The Lightspan
professional development team aims to help teachers and administrators
focus on implementation planning, student data analysis, ongoing
assessment, integration of resources, and family involvement.
In addition, Lightspan is developing a comprehensive online
professional development component that will be launched within
a few months, a company official said. |
| Cost
Structure |
The
Lightspan assessment products are sold by subscription by
the school, district, or state levels. Prices vary considerably
depending on the way it is purchased and what users choose
as part of the assessment suite.
|
Northwest
Evaluation Association (NWEA)
|
| Web
site address |
http://www.nwea.org/ |
| Product
Summary and Features |
The
Northwest Evaluation Association (NWEA), a Portland, Oregon-based
nonprofit organization, has created and offers a computerized
adaptive test known as Measures of Academic Progress (MAP).
The assessment aims to meet the goals of the No Child Left
Behind Act by testing and measuring a student's academic growth
in specific subjects over time. MAP is electronically delivered
and is Web-enabled.
The
NWEA assessments cover grades 2-10 in reading, math, and language
arts. They currently are aligned to standards in 48 states.
The organization customizes tests in math, reading, and language
usage using the NWEA item pools to create tests that align
with local curriculum and state standards. NWEA works with
districts across the country, and also has the capability
of working with large state contracts. For example, NWEA is
the primary contractor on a statewide Internet-based adaptive
assessment in Idaho for students in grades 2-9. The organization
also currently has partnerships with 716 school districts.
While
the student is taking the computerized adaptive test, it adjusts
and customizes to the student's achievement level. NWEA delivers
the assessment to a district through a Web-connected machine
and it is then distributed to students' computers that are
connected to the district's server. Once the assessment is
completed, the MAP data is uploaded to NWEA through the Web-enabled
computer and processed. The day after the assessment, teachers
and schools have access to the performance results online.
Teachers
receive a special report of the test results, and a suite
of reports is available on the assessment that can be aggregated
or disaggregated, according to various filters, including
Special Education students, Title I students, gender, etc.
|
| Item
Pool |
NWEA
develops all of the items for its MAP, and train item writers,
utilizing the services of teachers from all over the country.
Submitted items go through a content and bias review, and
then they are field-tested with hundreds of students. Once
they have gone through this process and are approved, they
are placed in the item pool. NWEA currently has 25,000
unique items in its item pool.
|
| Reliability |
A
research team from NWEA monitors the assessments and conducts
validity and reliability tests. NWEA recently completed
research concerning the stability of its measurement scale,
and found that it has been stable for the past 20 years. The
measurement scale crosses grade levels, and it is a continuous
scale across all levels. This allows NWEA to track individual
growth, in other words to compare individuals to themselves
over time. If you look at how a student is performing with
regard to the scale throughout the year, educators can predict
how a student will do on a state test later in the year. Because
thousands of students take NWEA's assessments, the organization
has the ability to report normative data, too, a company official
said.
Training
for the staff begins with initial face-to-face meetings that
explain how a test system will work within a district's plans
and regarding educational goals, prior to any setup of the
assessments. Prior to the administration of assessments, teachers
and administrators receive training in the technology, in
the setup of assessments, in administration procedures. Once
assessments are given and reports are first made, the group
provides training in interpreting the data to make maximum
use of the assessments to inform instruction. After two seasons
of testing, NEWA does another round of training, focusing
on how to interpret a student's academic growth. Training
is done on-site, working with staff representatives who in
turn train others.
|
| Cost
Structure |
The
cost of the NWEA computer adaptive assessments is based on
enrollment. Most common is a district license. NWEA, for
instance, computes the total enrollment of a district in K-12,
and multiplies it by $5.25 per student, to arrive at the fee.
For this amount, the district gets tests in reading, math,
and language arts, and can test students in grades 2-10 up
to four times a year. This amount does not include training.
The training is in addition to the basic cost and is sold
by module.
|
Princeton
Review - Homeroom.com
|
| Web
site address |
http://www.homeroom.com/homeroom_index.asp?track=01 |
| Product
Summary and Features |
Through
Homeroom.com, Princeton Review offers an ongoing and online
formative assessment and benchmarking tool aligned to state
standards and high-stakes state tests. Teachers can create
customized practice assessments, using the Homeroom.com pool
of thousands of items, for grades 3-12, and assess and diagnose
students' performance specific to the skills measured by a
state's tests. The product is completely Web-based and delivered
through a Web browser. It was launched in 1999. Homeroom is
sold on a statewide as well as district-wide basis. The company
has a statewide contract with Massachusetts, and district
wide contracts across the country including Philadelphia and
Memphis, to name a few.
The
math and reading content of Homeroom.com is aligned to multi-state
and state tests, all state standards, and state textbooks.
Students can take the tests either online or off-line. In
real time, Homeroom.com provides skills-based assessment results
at the class and student level, and the results can be drilled
down to objective and item. The up-to-the-minute results are
available to students, teachers, parents, principals, and
district administrators, and allow the tracking of students'
strengths and weaknesses on state-mandated skills, according
to the guidelines of No Child Left Behind.
Homeroom.com
assessment segment is complemented by a remediation component
through targeted online resources. Homeroom.com makes available
targeted skills practice with activities, lessons, and tips.
The product has more than 10,000 resources, ranging from games
to academic worksheets, each geared to work with individual
learning styles. Using Homeroom.com, for instance, parents
can not only monitor skill performance and see a child's latest
assignments, but also find resources that are intended to
help children improve their skills.
Homeroom.com
content currently covers nearly all of the 50 states in the
U.S. Such coverage varies from some states in which Homeroom.com
has the exact idiom and format of a state test to other states
in which the content basically matches the skills covered
in the state high-stakes assessment.
In-depth
reports allow administrators to access up-to-the-minute results
at the district, school, class, grade, and student levels.
The reporting capability of Homeroom.com lets school administrators
and teachers disaggregate the data by varied parameters according
to NCLB mandates, such as gender. The reports available to
teachers include the overall performance summary chart; student
performance summary; topic and skill analysis by student;
and item analysis by student. The data auto-import services
of Homeroom.com allow student, teacher, and administrator
data to be imported into Homeroom automatically so that teachers
and administrators do not have to spend time setting up each
student's account.
|
| Item
Pool |
In
creating an item pool, Princeton Review is actively involved
in understanding and reviewing all kinds of students assessments,
and applies that knowledge to item-development efforts to
present students with a wide variety of items to look like,
feel like, and test the same content as any norm- or criterion-
referenced test. Test items are written by The Princeton Review's
staff of educators based on analysis of all standards and
tests. There are 160,000 items in the item pool.
As
items are created, content editors, text editors, and copyeditors
review them for accuracy and to ensure that they are error-free.
Content reviewers also evaluate the questions and passages
for any possible bias. Once items have been through several
reviews, they are ready for inclusion in the item bank database.
|
| Reliability |
Each
item is associated with skills that serve as the backbone
of the Homeroom system, allowing users to cross-reference
their curricular goals or standards with their testing requirements.
Items are not initially field-tested. Instead, they immediately
become available to users. Princeton Review makes some assumptions
about the quality of the items because the company considers
the source items to be reliable, valid, and, at the very least,
of significant difficulty.
Princeton
Review, however, does not rely solely on such assumptions
to guarantee a valid and reliable item bank. The items are
evaluated using traditional item statistics. Items are modified
and/or removed if they are determined to be too difficult,
or if a single distracter is impacting student performance.
As the assessment product develops, the company is continually
looking at ways to improve our item bank, a company official
said. While the company does not perform calibration studies
or employ item response theory, Princeton review believes
that the item bank is strong and able to guide instruction
when used appropriately.
|
| Training |
Educators
learn how to use the online formative assessment tool through
in-school Homeroom.com training sessions. Princeton Review
provides different types of training on Homeroom.com, such as
one-on-one training with teachers or, more commonly, a workshop
for groups of teachers. Following the initial training, there
is ongoing support, such as an online training module; customer
service available through e-mail; and live customer service
help through online chat access, at selected times Monday through
Friday. |
| Cost
Structure |
Access
to Homeroom.com is sold on a subscription basis. The cost
is based on the number of students within the state/district/school
who will be using the product.
|
Renaissance
Learning - StandardsMaster
|
| Web
site address |
http://www.renlearn.com/standardsmaster/default.htm |
| Product
Summary and Features |
Renaissance
Learning offers StandardsMaster, an instant assessment coupled
with Web-based reporting software that covers math, reading,
and language arts in grades 3-10. StandardsMaster software
connects the district office and schools in a network that
makes state standards assessment data available to teachers,
students, parents, principals, and district administrators.
StandardsMaster was launched on April 1, 2002, and last September,
the company added online testing capability.
With
the StandardsMaster program, schools have the capability to
administer either a paper-and-pencil or an online test - in
math, reading, and language arts - that monitors every student's
progress toward state standards. The company is developing
content aligned to all 50 states, and has produced content
aligned to about half of those states thus far. The company
expects to complete content aligned for all states by the
next school year, 2003-2004, a company official said. Through
the criterion-referenced assessments, educators can analyze
specific areas in which students need aid in mastering state
standards.
The
assessments are done in several steps: Teachers can print
out an assessment, using the Web-based software, and administer
it to students using the paper-and-pencil method, similar
to most standardized tests. Using the AccelScan Intelligent
Mark Recognition Reader, teachers score the assessments in
their own classrooms, a process that can be done within minutes
of the test completion. Once the bubble card from the test
is fed through a scanner, you can access the test results
from any Web-connected computer in the network.
Customizable
reports on the test results can be created at the district,
school, classroom, and student level. Reports can be disaggregated
using filters such as gender, ESL students, Title I students,
etc. Teachers and students obtain data down to the specific
item level. Item analysis reports allow teachers to see the
standard assessed by each item and the percentage of students
that chose each answer option for each assessment item. Reports
can also be created to view students' progress over a defined
period of time.
Districts
can decide whether to run the program on a locally based Web-accessible
network or Renaissance Learning can host the network and make
it available to the district. Most user districts are currently
hosting the program on their own networks.
|
| Item
Pool |
In
terms of the product's content and item reliability, Renaissance
Learning employs experienced items writers, who have backgrounds
in education. Many have been teachers. Also, many have written
content for textbook publishers and similar educationally
oriented companies. In addition, the company had test-preparation
products that contain content, and most of that content has
been repurposed for use in the StandardsMaster product, a
company official said.
|
| Reliability |
The
company is currently engaged in research evaluating the reliability
and validity of StandardsMaster assessments and collecting
data from users regarding their performance on high-stakes
tests, and is planning to make such results available in the
future.
|
| Training |
Training
and consultation are provided upfront in training sessions for
teachers and administrators. In the post-assessment phase,
Renaissance Learning consultants are available to work with
schools to analyze results and come up with action plans. |
| Cost
Structure |
StandardsMaster
is sold on a per-student per-year subscription basis.
There is one fee for the software, and a fee for each additional
content area selected by a district. Total cost is dependent
on a number of factors, such as size of the district. Renaissance
Learning also offers data conversion services to get all current
student information into the system.
|
Riverside
Publishing - Assess2Learn
|
| Web
site address |
http://www.riverpub.com/products/online/assess2learn/index.htm |
| Product
Summary and Features |
Assess2Learn
is an assessment product that delivers comprehensive exams
to schools via the Internet. Geared to grades 1-8,
it is the online assessment product of Illinois-based Riverside
Publishing, a division of Houghton Mifflin Company and a major
test publisher that offers various educational testing products
and services. Assess2Learn assesses students' skills in
reading, language arts, and math relative to state and national
standards and provides immediate, real-time test results.
As
a Web-based diagnostic tool, Assess2Learn is both an assessment
content product aligned to state and local standards and the
engine for delivering the assessment. The company has developed
a test bank, and has created three test "forms"
for each subject area that can be accessed by teachers. Using
the Assess2Learn tool, teachers and educators can determine
where students are strong and weak on skills defined by state
and local standards; tailor in-depth instruction specifically
for each student; and identify at-risk students and populations
on the class, grade, and school level. With Assess2Learn,
users such as teachers, administrators, or curriculum specialists
also have the ability to create tests. The test author would
devise the tests, the sections, the questions, the answer
choices, and the grading standards.
Riverside
works with customers in exactly how they want to use Assess2Learn
and its content. The company will sell schools its own content
if they already have something they're using; it will sell
them usage of the Assess2Learn engine if they have their own
content to place within the engine; or Riverside will sell
them both the content and the usage of the engine. The third
option does not limit schools from loading their own content
to supplement the company's.
Assess2Learn is sold primarily to schools and districts. It
is currently used in 86 districts, and the Assess2Learn tests
are available to more than 800,000 students. The company has
also been awarded the contract by the state of Georgia to
develop a statewide assessment tool for grades 1-8.
The
assessment product is done on an ASP model in which a school's
student information data is first imported into the system.
Riverside does all of the data hosting, although if districts
want to import test-results data on to their servers, they
can. Students log on, take the appropriate tests, and click
a button for the test to be graded as soon as they are finished;
the results are immediately scored. The data is saved on the
ASP server. Test results are scored and displayed according
to the requirements specified by schools during the assignment
process. Results are available immediately to students, teachers,
and school and district administrators.
Assess2Learn
provides various comprehensive test result reports to teachers
and administrators. These include the ability to drill down
data and see skill proficiency reports by student, by class,
by building, and by school system. Results can also be used
to break out each state standard and measure how students
did against the standard. Item analysis reports permit educators
to see how students performed on each item.
Currently,
the assessment content is aligned to standards in 14 states,
with another state due to be added in the next year. There
is also a national edition available to schools. The company's
goal is to add content aligned to an additional 12 states
each year, until the entire country is covered, a company
official said.
Riverside
is currently adding in the features that would allow data
to be viewed based on No Child Left Behind requirements. Data
can currently be imported through these filters, but the system
does not yet generate reports along such categories. This
functionality is expected to be ready by August, a company
official noted.
|
| Item
Pool |
Riverside
Publishing develops and publishes criterion-referenced, norm-referenced,
and ability tests, as well as alternative assessments that
enable school personnel to measure the educational progress
of students. This test-development experience is brought to
bear in the creation and maintenance of the Assess2Learn item
pool. There are 25,000 test items available to be used
in the Assess2Learn assessments.
The
company has a staff of test development specialists who oversee
item writers and who have experience and training in many
fields of education, numerous content areas, psychometrics,
and management. The model for the electronic testing development
process includes seven steps: review the state's guiding documents
and develop content domain specifications; develop test blueprint
specs; develop item specifications; select and train item
writers; develop test items; conduct several stages of internal
item review for content, bias, and sensitivity; and produce
online test forms.
The
company's test development specialists are experienced item
writers. They also use a team of consultant item writers who
have composed items for other Riverside products and who are
teachers or former teachers with specialized knowledge concerning
their content areas of expertise.
There
are several stages of internal item review. In the review
process, the items are reviewed and verified for their alignment
to state standards; accuracy; appropriateness for difficulty
pertaining to the purpose of the test; relevance; and fairness.
They are reviewed to be sure that they are free of bias and
sensitivity issues. A content specialist and a senior reviewer
review all test items for potential bias.
|
| Reliability |
Riverside
focuses on content validity through the detailed process outlined
above (under "Item Pool") in the creation of its
online assessments. Once the test is created and put online,
a thorough QA process is utilized to ensure content validity;
that each item measures the standard it is purported to measure;
and that answer keys are correct, etc.
In
addition, the company is gathering both quantitative and qualitative
data to ensure items perform in an acceptable manner, and
the company makes ongoing changes as appropriate. Quantitative
data is gathered as the tests are being administered to students,
while qualitative data is obtained based on feedback from
educators in the various states where the company publishes
content. The company also examines other outcome measures.
The diagnostic nature of the tests provides detailed information
down to the granular level of the standards, a company official
said.
|
| Training |
Training
is available to schools at an extra charge, one that is figured
per day of training. Such training is provided on how to
use the delivery system; how to negotiate the technology infrastructure
(how to log on, how to assign tests, etc.); how to interpret
test results data; and how to custom-create tests. |
| Cost
Structure |
Assess2Learn
is sold on a per student per year basis. This cost is
divided into two components. There is one segment of the cost
that is for the technology - for the Assess2Learn engine itself.
And, there is a one-time fee for the Assess2Learn content.
The technology subscription cost typically runs $6 per student
per year. The content cost is a one-time charge of $3 per
student per content form.
|
Scantron-EdVision
- Performance Series
|
| Web
site address |
http://www.edperformance.com/ |
| Product
Summary and Features |
The
Performance Series is Scantron's Internet-delivered standards-based
adaptive measurement - a computer adaptive assessment created
to measure the different academic strands of individual state
standards. It aims to give immediate diagnostic information
to each teacher, including learning objectives a student has
not completed, groupings of students by ability, and the academic
gains made by individual and groups of students. Scanton's
Performance Series assessments are for grades 2-12. The company
has been offering Reading, Math, and Learning Styles assessments
for the past two years. Science and Language Arts modules
have also been developed and will be tested beginning this
spring. Since the product's launch about two years ago,
some 1.5 million students have been tested on these assessments.
The
EdVision-developed computer adaptive assessment targets the
instructional level of each student by altering question difficulty
based on previous answers. It is done on an ASP model,
which is delivered via the Web, with all data hosted by Scantron.
Schools can test students as long as they have a computer
with Internet access for the student. Educators can then view
the results of an assessment on the Web immediately.
This
diagnostic assessment places a student at his or her appropriate
instructional level immediately. It is geared toward driving
instruction and helping teachers, through its findings, to
create a personalized learning plan for each student, based
on his or her performance on the adaptive assessment. Such
adaptive testing is different than levels testing in that
a whole class will not receive the same assessment, and it
is based on the recognition that students in one class actually
are situated in many different grade levels, skills-wise.
The
Scantron standards-based adaptive measurement tracks students
using a consistent scale as they move from one school to the
next, and measures academic growth over one year or across
multiple years. Teachers get very detailed reports of a student's
results that indicate what would be the next step to address
specific skills weaknesses. Administrators can access scores
for the whole district, as well as for each school. The data
can be aggregated and disaggregated, so that educators can
group data and obtain reports in a variety of filters, such
as gender, students with disabilities, etc., which help districts
fulfill objectives of NCLB.
|
| Item
Pool |
The
Scanton assessments are research-based. The items are
internally created and go through peer review and bias testing.
When the company's content team creates an item, the item
goes out for external review by a team of consultants, made
up of educators (ex-teachers, teachers, and others with educational
background.) Once the item is checked, if it needs to be revised,
the item is sent out for review again and to go through an
approval process. Items are tested on a pilot site, and the
company's statisticians review the data with regard to the
performance of the item.
|
| Reliability |
The
concept of reliability for an assessment instrument can be
thought of as the degree to which an assessment instrument
provides consistent measures. Scantron utilizes two methods
of illustrating this concept statistically. At the beginning
of the project, test items were examined in a non-adaptive
manner. To do this, items were grouped into non-adaptive tests,
where, for example all 5th grade math questions only are grouped
together and administered to groups of 5th graders only. Under
this non-adaptive method, there are three considerations:
an item's difficulty, its ability to discriminate between
students with high ability and low ability, and how all of
the items on the non-adaptive test work together.
This
third concept is sometimes referred to as internal consistency
reliability, and is given as a coefficient value that is most
commonly referred to as a KR-20 or Cronbach Alpha value. This
statistic indicates how all items on a test relate to each
other and to the test as a whole. The theoretical maximum
is +1, and a higher score has an interpretation of a higher
degree of reliability for the test.
The
second method used accommodates the Computer Adaptive format
of the Performance Series. A statistic called the Standard
Error of Measurement is calculated in this case, as there
is no "set" of test items seen by a large group
of people for which a KR-20 or Cronbach's Alpha value can
be calculated. This Standard Error of Measurement value is
calculated for an individual and can be thought of as a confidence
interval within which an individual's measure will fall with
repeated assessments. A small Standard Error of Measurement
implies that repeated assessments for an individual would
provide measures of ability for the individual that are consistent.
|
| Training |
Scantron
employs a group of some 70 persons involved in the training
that is incorporated with using the Performance series. The
company has a basic training structure and makes available several
modules which districts receive depending on the type of program
they want. Besides training on use of the core product, Scanton's
training includes other modules, such as a data interpretation
class. Teachers learn how to teach and effect remediation based
on the Performance series assessments -- how to make changes
in curriculum and in individual learning plans based on a student's
performance. There is also an online help area with resources,
and statisticians are available online for consultation. |
| Cost
Structure |
The
cost of the Performance series assessments is on a per-student
per-year basis, and can differ somewhat due to the size of
the district and other options. Typically, it runs about
$15 per student, which includes unlimited testing and all
reports. Scantron also sends out service technicians to
work with schools' IT departments, to validate that the school's
technical infrastructure - firewall, content filters, browser
levels - are suitable and helping to complete the data importing.
|
|
|
| Web
site address |
http://www.testu.com |
| Product
Summary and Features |
TestU,
a company that concentrates most of its efforts on test prep
programs for grades 9-12, recently launched a new product,
Middle School Math Pathways, which is a standards-based math
assessment program. The Math Pathways Internet-based assessments
are aligned with state content standards. The program is a
"pure assessment product": designed to help school
districts assess their students' progress on state content
standards before they reach high school, says a company official.
It provides real-time data.
For
each year in grades 5-8, students are administered five required
tests to assess performance on core skills. Three types of
assessments are in the tests: readiness assessment, which
assesses "back to school" preparedness at the beginning
of the year; formative assessment, administered at the end
of the first quarter, which assesses concepts covered in that
time period and provides a snapshot of student progress; and
summative assessment, which covers topics from the entire
year or semester and reveals student mastery of the curriculum.
TestU
also offers a concentration of online tools in grades 9-12
geared toward practice for states' high school exit and high
school proficiency exams. The programs combine diagnostic
assessments, preparatory courses, and practice tests. High
school exit exams covered range from the New York Regents
Math and English and Texas TEKS to Florida's FCAT and the
California High School Exit Exam. The company's Skill Navigator
product is designed to aid students in other states meet requirements.
|
| Item
Pool |
For
their products, the company both generates its own test items,
working with consultants who have written content for particular
states, and it also licenses content from sources such as
Barron's. There are more than 15,000 test items in the
TestU pool.
|
| Reliability |
TestU
is not a predictor of student scores, according to company information.
What the company does assess is how well students are progressing
toward skill mastery. |
| Training |
TestU
is one of the companies responding to the need that teachers
have to get help in interpreting results. The company has developed
a separate tool, TestSCOPE, which measures student usage and
performance at the aggregate level and allows educators to disaggregate
data by student, class, student, and district.
Training
sessions are incorporated in these TestU products. The
packages sold to school contain one or two days of training,
depending on what is purchased. In addition, the company has
a team of coaches - who are current or former teachers - working
with schools on an ongoing basis.
|
| Cost
Structure |
TestU
formerly sold its tools on a per-student basis, but now bases
prices per school site or per classroom. There is a range
of prices, but typically, a package can be purchased for a
class with 30 students for $4,500, including two days of training.
The cost for a single high school site is typically $7,500,
including two days of training.
|
Back to top
YPSILANTI,
MICHIGAN, PUBLIC SCHOOLS
Noni Miller, executive director for educational
services
In
Ypsilanti, Michigan's Adams Elementary School, students' reading
scores on state tests in 2001 showed that only 39.4 percent of pupils
were reading proficiently. This figure showed the type of challenge
that school officials and teachers faced. Adams, a school in which
some 65 percent of the students come from poorer families, as defined
by qualifying for Free or Reduced-cost Lunch, has many struggling
readers. One year later, however, in 2002, reading scores increased
dramatically: 62.1 percent of the Adams Elementary students were
reading proficiently, according to state test results. In the
same one-year period, reading test scores for elementary students
statewide actually showed a decline in proficiency. The local story
in Ypsilanti was one of improvement and some success. Why? In
the view of Noni Miller, executive director of educational services
of the Ypsilanti schools, the significant improvement shows the
direct impact of the district's use of online standards-based assessments
that provide real-time results and that drive individual instruction
plans developed for each Ypsilanti student.
This
school year is the third one that Ypsilanti schools are using Scantron's
Performance Series online math and reading assessments, which are
computer-adaptive measurements administered in grades 3-12 of the
district. Such adaptive tests alter the difficulty of questions
based on the previous answers of a student and address the specific
performance level of each student. (See Products
section for more detailed description of this product.)
Ypsilanti schools first started using the Performance Series as
a pilot three years ago. The district also uses Scantron's Curriculum
Designer, a computerized tool to custom-design an education plan
aligned to district, state, and national standards, and Skills Connection,
a test development tool, says Miller.
Ypsilanti
Public Schools serves a small city and two surrounding townships
in eastern Washtenaw County, Michigan, and has a total enrollment
of 4,650 students. Fifty-six percent are minority students and 53
percent qualify for Free or Reduced-cost Lunch. The K-12 district
has 12 school buildings and has 250 teachers on staff. The district
decided to use the Performances Series computer-adaptive assessments
as a top-down initiative, from the district and building levels,
because too many students were not succeeding at state tests and
were achieving far below level, Miller explains. The district has
excellent teachers and suitable materials, she says, but officials
sought a new type of assessment to help address the problem of underachievement
on state tests and to drive instruction in a way that the district
could be sure that it was "teaching what it was testing, and
testing what it was teaching," as she notes.
Students
in grades 3-12 take Scantron's computer-adaptive math and reading
assessments in the fall and spring. By administering in both fall
and spring, the district approximates the "seed time"
for learning that students have in the classroom during the year,
says Miller. The assessments take 32 minutes of classroom time to
finish. Once the tests are completed, the test results are available
online to teachers and administrators immediately. The assessments'
detailed reports show which skills and objectives students have
mastered and which areas they are weak on, and they indicate the
next objectives for each student to complete.
With
the adaptive tests, a fourth grader will log on as a fourth grader
and begin answering items. Depending on whether a student succeeds
in answering an item correctly, the test subsequently and continually
adjusts the items administered in the assessment and locates that
student at the appropriate level, for example, going down to 3.9
or 3.8, or up to 4.3 or 4.4. When students take a standard test
that does not adapt in this way, Miller notes, they may only correctly
answer 25 percent of items on a fourth-grade assessment, and simply
conclude, "I'm dumb." A computer-adaptive test that adjusts
for each child allows a student to locate him or herself at the
exact level where he or she is comfortable, as Miller explains it.
Based
on the assessment results for each student, Ypsilanti teachers and
administrators then develop an "individualized plan for success,"
according to Miller. The individual plan dictates the teaching
strategy, remedial steps, materials, and curriculum emphasis that
will be used to address each student's weakness on skills relative
to state standards. Such plans incorporate different approaches,
with after-school tutoring planned for some students, in-school
tutoring for others, etc.
With
its mission of making every student successful, the district has
the ultimate goal of creating and maintaining an individualized
learning plan for each student. "We expect all of our buildings
to do this," Miller says. Using the online assessments and
the other tools the district has purchased, the district is phasing
in this plan. The first focus is on children who are most at-risk
to have individualized learning plans.
The
planning and execution of the online assessments is a system-wide
initiative, continually driven from the top down. The central administration
developed a common vision of how to use the Performance assessments
to inform the instruction, so that parents, teachers, administrators,
and the school board all understood the plan and how it is being
carried out. This work is facilitated primarily through the district
School Improvement Team and the Improvement Teams for each individual
school building. The improvement teams are composed of the building
principal, teachers, parents, and community members. Using the
Performance Series assessment results, the teams focus on "big-picture"
questions such as how should the results influence curriculum adoption;
has the district addressed students' competencies with changes in
teaching strategies; on which exact curriculum areas are students
performing poorly.
The
district has developed a district-wide calendar that lays down a
specific timetable of assessments for all district schools; this
allows administrators to monitor whether all classrooms are keeping
up with administering the assessments. As stated, the online assessments
are given in fall and spring. During winter, many students take
the official state tests (the Michigan Educational Assessment Program
tests), and those who don't will do the Performance assessment at
that time, too.
As
the head of educational services, Miller is not only what she calls
the "sexy cheerleader" for the entire online assessment
program, but also helps spearhead and oversee its implementation.
This means everything from making sure that people are on their
calendars and getting the assessments completed, to identifying
key skills and areas in which students are performing weakly relative
to state standards. For example, Miller will examine school-level
results and find that a certain segment of students are performing
poorly on a particular skill, such as learning fractions, and work
with building principals to find out exactly how that area is being
addressed with those students. The assessments make for a much
tighter accountability.
In
Ypsilanti, teachers are the ones who develop the individualized
learning plans, under the aegis of the building principals, who
are ultimately responsible for the progress of all students. Teachers
and administrators can disaggregate the Performance test results
data by such categories as gender; ethnic origin; socioeconomic
level; children who are Limited English Proficiency; Children who
are From Limited English Proficiency backgrounds; special education,
etc.
The
assessment results force principals to look more critically at the
strengths and weaknesses of their teachers relative to achievement
on state standards and benchmarks, according to Miller, and
to get teachers the professional development needed to address such
weaknesses. On an as-needed basis, principals meet with their teachers
on to make sure they follow up with their educational plans based
on assessment results. These are some of the ways that the assessment
program is driven through the entire organization.
Are
the assessments causing any different emphasis on curriculum? Miller
says that teachers and administrators are using the assessment results
to identify, district-wide, particular skills that students are
not mastering relative to state standards. One example is fractions.
Using the assessment results, administrators and teachers are examining
the teaching strategies and materials being used to teach fractions
and changing the instruction in order to emphasize mastery of the
skill, to evaluate whether age-appropriate instruction is occurring,
and to make sure that students have a strong conceptual foundation
to learn fractions. That is just one example of a targeted skill
or curricular area being modified district-wide based on assessment
results.
Miller
explains that there are two other especially important ways that
teachers and administrators find the computer-adaptive diagnostic
tool helpful. Teachers find the results extremely useful in shaping
their parent-teacher conferences. The assessment provides very
fine reports on students' mastery and weaknesses, relative to standards
and benchmarks, and teachers can thus discuss with parents the test
results and their importance in a more meaningful way. "It
is a different way to approach a parent than just saying, 'Susie
can't do math.' Instead, a teacher can say, 'Susie is very good
at adding numbers, and these are the next two objectives that we
will focus on,' " Miller says. "It's a whole different
slant on the partnership with parents." For every student,
but especially for those who are struggling, it is geared more toward
mastery and success than simply a focus on failure in the state
assessments.
|