Targeting Each Student

1
Targeting Each Student

• David Chiszar
• Director of Assessment and Quality Analysis
• Naperville Community Unit School District 203
• Leo Bohman
• VP Application Development, K-12
• SPSS Inc.

2
K-12 Analytics System Components

• SPSS Dimensions
• Surveys
• Assessment deployment
• Paper based
• Hand held
• On line
• Item bank development

• SPSS 15.0
• Descriptive statistics
• Significance tests
• Segmentation models
• Prediction models
• Value added / growth models
• Data visualization

Statistics
LongitudinalData Mart
Visualization
Data collection

• Microsoft
• SQL Server 2005
• Analysis Services
• Integration Services
• Reporting Services
• Office 2007
• Sharepoint 2.0
• ASP .Net 2.0

• Professional Services
• Project management
• Technical support
• Application hosting
• Staff development througheducation partners

OLAP/Reporting
3
Longitudinal Data Mart

• Data Model
• Improvement Planning

4
Students

• Demographics
• Gender
• Ethnicity
• IEP Special education needs
• FRL Socio-economic status
• LEP English is a second language
• Enrollment
• Grade
• Year
• Courses
• Teachers
• Schools
• Interventions

5
Assessment Instruments

• ISAT Illinois Standards Achievement Test
• ITBS Iowa Test of Basic Skills
• IAA Illinois Alternative Assessment
• IMAGE Illinois Measure of Annual Growth in
  English
• CogAT Cognitive Abilities Test
• ACT
• PLAN
• PSAE Prairie State Achievement Examination

6
Metrics

• Scale score
• correct
• Normal curve equivalent
• Percentile rank
• National, State, District, School, Grade Age
• Standard score
• Raw score
• Standard age score

7
Formative assessments

• Dec 2006 Finals
• Courses
• Algebra
• French
• Accounting
• Keyboarding
• Honors pre calculus
• Patterns
• Item detail
• Questions
• Responses
• Statistics
• State standards
• Curriculum objectives

8
Staff

• Role
• Teacher
• School administrator
• District administrator
• System administrator
• Security rules
• Data access
• System function

9
Benchmarks

• ISAT goals
• District
• School
• Grade
• Subject
• Standards cut scores
• Exceeds
• Meets
• Watch
• Probation

10
Interventions

• LEAP
• Kindergarten early intervention program
• Students between special education and district
  expectations
• Monitor as a cohort group

11
Use in Improvement Planning

• Web based for staff
• Finite number of ad hoc views
• Analytic presentations
• Student reports

12
Introduction to Naperville 203

• Suburban Chicago
• 18,000 students
• 21 buildings
• 2 high schools
• 5 middle schools
• 14 elementary schools
• 1200 teachers

13
Questions???
14
SPSS DimensionsAssessment Deployment
Hand held
Paper
Web
15
Paper Process

• Scan operation
• Pick students test
• Create test documents
• Scan answer sheets
• Results reporting
• Class summary receipts
• Data to data mart
• Edit results
• Item analysis

16
Authoring with Question Librarian

• New paradigm for authoring and inventory
  management of items
• User defined item attributes
• Subject, grade, class, strand, etc.
• Statistics P-value, Chi square test, Point bi
  serial correlation
• Primary secondary state standard
• Curriculum objectives
• Sophisticated search engine
• Easy assembly of items into assessments
• Supports authoring/approval work flow process
• Teacher public/private district public/private
  repositories
• Facilitates sharing the development of item banks
• Beta for early adopters April full release
  October

17
Questions???
18
SPSS Advanced Analytics

• Tests of Significance
• Visualization with Predicted Score

19
Predicted Score Methodology Today
MULTIVARIATE MODEL for each grade
2007 ISAT Math prediction model f(district
history with that grade on ISAT Math score
history ISAT Reading score history ITBS Math
score history ITBS Reading history . Any
other standardized tests history available)
Predicted student ISAT math score f(2007 ISAT
Prediction model for the grade applied to the
students history on the standardized tests
Method typically explains 60 - 70 of variation
in test scores
Standardized tests administered annually
Predictions are static for a year!
20
Predicted Score - 2007
MULTIVARIATE GROWTH MODEL for each grade
2007 ISAT Math prediction model f(district
standardized test history with that grade
District Math assessment history District Math
intervention history District Math staff
development history . Any other relevant
quantifiable district activity)
PARADIGM SHIFT needed to address a richer data
set and an interest in measuring the effect of
each local variable. Growth model development is
a function of experimentation with sampling
theory and probability analysis within a
simulation model. Stochastic vs. deterministic
process
Increase accuracy of prediction Refinement the
measurement of the inherent error in the
prediction.
Relevant district data is collected, organized
and used in the model! Predictions are updated on
demand! Contributions of variables to progress
can be measured!
21
Growth Model Development Partners

• Dr. Yeow Meng ThumCollege of Education, Michigan
  State University
• Arie van der Ploeg, Senior Researcher
• Apply SPSS Software to Growth Modeling
  Methodologies Adapting value-added approaches
  for use by schools and districts to predict
  patterns of improvement using local and state
  data

22
Questions???
23
Webinar follow up

• Web site resources
• Effective Use of Electronic Data Systems A
  Readiness Guide for School and District
  LeadersLearning Point Associates and Educational
  Service Agency Alliance of the Midwest
• Predictive Analytics Extending the Value of Your
  Data Warehousing InvestmentThe Data Warehousing
  Institute, Jan 2007
• Finding Additional Value in New Accountability
  Systems Learning Point Associates Dr. Thum
• Recording of this session on www.spss.com
• Survey
• Feedback on session and resources
• Ideas for future sessions

24
Contact Information
David Chiszar Naperville Community Unit School
District 203 203 West Hillside Road Naperville,
IL 60540 P. 630.420.6963 e-mail
dchiszar_at_naperville203.org website
www.naperville203.org
Leo Bohman SPSS Inc. 233 S. Wacker Drive, 11th
Floor Chicago, IL 60606 P. 262-569-8922 e-mail
lbohman_at_spss.com website www.spss.com