Using a Data Warehouse to Drive Decisions

1
Using a Data Warehouse to Drive Decisions

• Tom Raitzer Senior Systems Analyst
• Dr. Dena Dossett Evaluation Specialist

Jefferson County Public Schools
2
About Jefferson County Public Schools

• Serves Louisville, KY and surrounding county
• 26th largest in US
• 95,000 students 1/6 of Kentuckys public school
  children
• 152 school locations
• 5,600 teachers/18,000 total employees

3
JCPS Technology Environment

• InHouse IT department
• Most applications developed by JCPS
• Based on needs of user departments and
  state/federal reporting mandates
• Operational systems collect data,data warehouse
  reflects it for reporting and analysis

4
Data Driven Decisions at the LEA Level

• Research Evaluation Department
• Program Evaluation
• Institutional Research
• Longitudinal Studies
• District
• Planning
• Systems
• Dialogues

5
Uses of Data for Formative and Summative Program
Evaluation

• Conducting needs assessment
• Selecting program participants
• Selecting comparison groups
• Determining available data for program
  
  outcomes
• Analyzing program outcomes

6
(No Transcript)
7
(No Transcript)
8
Conduct Program Needs Assessment

• Plotting student participants addresses for
• identifying possible program sites

9
Selecting Program Participants

• Using the student database, program participants
  are targeted based on specified criteria
• Academic Information (i.e. test scores)
• Non-academic information (i.e. suspensions,
  attendance, tardiness)

10
(No Transcript)
11
Assessing Program Outcomes

• Comparison groups are matched on targeted
• criteria and demographic information
• Grade, socioeconomic status, race, gender, zip
  code
• Students participating in the program are
• compared to non-participants or to themselves
• longitudinally on measures of the program
• objectives (academic and/or non-academic
• areas)

12
(No Transcript)
13
(No Transcript)
14
(No Transcript)
15
Using Data for Institutional Research

• To examine the relationship between
• teacher characteristics and student
• achievement, the teacher database
• provides information about
• Professional development credit
• Certification
• Educational background
• Years of teaching experience

16
Use of Data at the District Level

• Planning used to provide trend data used for
  needs assessment and measurable goals in
  consolidated planning process for trend data
• Systems used to provide data for quality
  indicators of inter-departmental goals
  activities
• Dialogues used to provide central office with
  school and student level data for school visits

17
Sources for JCPS Data

• Operational Systems
• Student (Demographics, Attendance, Grades)
• Business (Financial, Human Resources)
• External
• Test Scores
• GIS

18
Technology Infrastructure

• Data Warehouse
• NCR Teradata 4700 WorldMark? 4-node system, 1
  terabyte of storage
• Populated mainly via flush fill method
• Transactional update in future
• Operational Systems
• Bull DPS 9000 online and batch
• All schools online
• Client/Server attendance system

19
Network

• Moving from microwave LAN to fiber optic backbone
  IP network
• About 60 complete, finished by 2003
• Computers in each classroom

20
History

• Operational Systems
• Online since 1982
• Was microwave LAN, moving to fiber optic backbone
  IP networking about 60 complete, done by
  2003
• 2700 online users, ¼M avg daily transactions, ½M
  peak
• Data Warehouse
• Since 1988
• Now over 200 users

21
Whats in Our Data Warehouse

• 300 Gbytes
• 13 years of student attendance
• 7 years of daily student records in 100M rows of
  data
• Enterprise-wide

22
Data Security

• Ownership/Need-to-know model
• Schools have access to detail for their students
  only

23
Data Warehouse Staffing

• 1 Database Administrator
• 4 Systems Programming
• 35 Support Development

24
Current Data-Gathering Projects

• KidTraxx Relating outside activities to school
  performance
• Test Scores
• CATS (Kentucky Standardized Tests)
• ACT SAT
• Graduation Requirements
• Combination of factors

25
Access to Data Warehouse

• Mostly via Windows PCs using ODBC
• Tools
• Microsoft Access
• Microsoft Excel
• Crystal Reports
• SPSS
• Also do traditional SQL reports
• Use Teradata enhanced SQL

26
Characteristics of Our Tools

• Limited metadata support
• Need to be familiar with data
• Cheap
• Need training

27
Future Data Warehouse Tools

• Web-based Data Access
• Why Web?
• Cross platform, low requirements
• Zero installation
• Can offer to general public
• Web Drawbacks
• Printing necessary, not well supported
• Software can be expensive and complex

28
Web-based Technologies

• WebView
• Developed in-house
• Subset of data, primarily student
• Contains features especially for schools
• Purchase of commercial web-based reporting tools
• Initially for central office

29
Public Access

• Structured, not general-purpose
• Example SchoolFinder
• On Web
• Allows parents, realtors to determine resides
  school
• Uses Microsoft Active Server Pages

30
For more Info

• JCPS web site
• http//www.jefferson.k12.ky.us
• Articles
• Teradata Customer Success Stories
• http//www3.ncr.com/product/brochures/pdf/eb3010.p
  df
• Bull
• Data Warehouse Makes the Honor Roll at Jefferson
  County
• Email
• ddosset1_at_jefferson.k12.ky.us
• traitze1_at_jefferson.k12.ky.us

31
Questions?