Data Warehousing Making A Difference In Improving Schools

Data Warehousing Making A Difference In Improving Schools Bill Flaherty Hanover County Public Schools

How would I show SOL results in deciles by school for multiple years?

How would I provide each teacher with past SOL results for each student in his/her classes?

How would I provide each student with a realistic prediction of what GPA and SAT is needed for admission to different colleges?

Today’s Presentation ¨ ¨ ¨ Three Questions Data Driven Decision Making An Overview of Data Warehouse Technology Survey What is Hanover’s Instructional Decision Support System? The Road We Traveled Challenges Advantages Cost Questions Demonstration of IDSS – Answers to the Three Questions + More

Data-Driven Decision Making High Stakes Testing

What is Data-Driven Decision Making? ¨ Mining the Data ¨ Analyzing the Data ¨ Communicating the Data ¨ Using the Data

Current Systems On-line Transaction Processing System OLTP

Building The Data Warehouse -Bill Inmon, 1992

Data Warehouse A subject-oriented, integrated, time variant, non-volatile collection of data in support of management’s decisionmaking process. -William Inmon

Data Warehouse A Data Source Data B Data Source C Data Source Warehouse User Community

Fundamental Characteristics of a Data Warehouse: Ø Separate Decision Support System database from OLTP systems Ø Storage of data only; no data is created, but it may be derived Ø Integrated data Ø Scrubbed data Ø Historical data

Fundamental Characteristics of a Data Warehouse: Ø Read only Ø Various levels of summarization Ø Subject oriented Ø Easily accessible

“Data warehousing is a journey, not a destination” -J. D. Welch

The Survey June, 2000

Survey Participants: ¨ Superintendent ¨ Assistant Superintendents ¨ Directors ¨ Principals ¨ Assistant Principals ¨ Senior Teachers ¨ Lead Teachers

Student test data is readily available. Summer, 2000 Summer, 2001

There are tools available to analyze student data. Summer, 2000 Summer, 2001

Training is available to analyze student data. Summer, 2000 Summer, 2001

Hanover County Public Schools Instructional Decision Support System (IDSS)

Our Current System CIMS On-line Transaction Processing System OLTP

IDSS CIMS Database Data Warehouse Test Scores (SQL 7. 0) User Community (Business Objects) www. businessobjects. com In-house Data, Surveys

1999 -2000 School Board Goals 4. Promote instructional programming in the following areas, among others. 1. Professional development 2. Curriculum development 3. Implications of longitudinal assessment of student achievement 4. Graduation requirements staffing needs 5. Vocational / technical / alternative education study recommendations 6. Standards of learning implementation / implications

Support From The Top ¨ Longitudinal Assessment Committee Chaired By The Superintendent – First Meeting May, 1999 – Committee members • • Board member Assistant Superintendent of Instruction Assistant Superintendent of Finance & Technology Director of Guidance & Testing Director of Technology Services School Principal Supervisor Information Technology

The Road We Traveled ¨ Evaluation of database for warehouse ¨ Training of staff in Microsoft SQL 7. 0 data warehousing tools and techniques ¨ Evaluation of user interface to the data warehouse

The Road We Traveled ¨ Focus group with principals ¨ Training of technical staff in Business Objects ¨ Survey of users ¨ Kickoff of Instructional Decision Support System at Administrative Retreat ¨ Aggressive and ongoing training of end users in Business Objects ¨ Website & List. Serv for IDSS Users ¨ IDSS Users Group

ID Student

Challenges Past ¨ Connectivity ¨ Wide-area network ¨ Hardware ¨ Compaq Proliant Server ¨ Pentium 466 mhz workstations (or better) ¨ Software ¨ Microsoft SQL Server 7. 0 ¨ Business Objects ¨ Client ¨ Designer ¨ Supervisor

The Network Frame Relay Workstation School Board Office School Network School Server

Challenges Past ¨ Identifying data elements ¨ Data Migration • Conversion • Different sources ¨ Learning new software • Technical Staff • End-users

Challenges Past ¨ Staying focused • Overwhelming amounts of data ¨ Meshing styles • Instruction vs. Technical ¨ Training • Customized manual • Initial – 2 day training – Over 80 administrators • Follow-up – 2 one day options

Challenges Future ¨ Training • Follow-up • User Group – Meets every other month – Receives tips from central office – Group members share knowledge learned and valuable reports ¨ Maintaining accurate data (Data Integrity) • Multiple sources • User input

Challenges Future ¨ Keeping data current ¨ Limiting data ¨ Have stakeholders continue using the IDSS ¨ Provide stakeholders with “live” data ¨ Evaluation

Advantage of IDSS ¨ Ease of use ¨ Speed in acquiring data ¨ Ability to create custom reports ¨ Make decisions based on information ¨ Improve student performance

Key Elements of Success ¨ Board and Superintendent Support – Board Goal – Personnel to support the goal ¨ Cost effective ¨ Infrastructure in place ¨ Training ¨ Core reports

Hardware Cost: ¨ Compaq Proliant Server – 600 mhz – Dual Processor – 1 G RAM – 18 G Hard Drive – Tape backup ¨ Total Hardware $7, 493. 00

Software Costs (30 Users): ¨ Microsoft SQL 7. 0 – License $390. 88 – 30 Users ($27. 46 each) 823. 80 ¨ Business Objects – Starter Pack 6, 996. 50 • 8 End Users • 1 IS Bundle – 1 Designer – 1 Supervisor – 22 Additional Users ($829. 50 each) ¨ Total Software $26, 460. 18 18, 249. 00

Training Costs: ¨ Microsoft SQL 7. 0 – Basic (1 person-1 wk. ) $800. 00 – Using Data Warehousing (2 people-1 wk. ) 3, 200. 00 ¨ Business Objects – End User (3 people – 1 day) – Designer (2 people – 3 days) – Supervisor (2 people – 1 day) ¨ Total Training 1, 211. 25 2, 422. 50 807. 50 $8, 441. 25

Cost Summary: ¨ Total Hardware ¨ Total Software ¨ Total Training 7, 493. 00 $26, 460. 18 8, 441. 25 ¨ Grand Total $42, 394. 43

The Dynamic Warehouse! What’s Next? ? ?

Questions http: //hanover. k 12. va. us/presentations/dw http: //www. nea. org/issues/high-stakes/dddm. html

Today’s Presentation ¨ Three Questions ¨ Survey ¨ What is Hanover’s Instructional Decision Support System? ¨ The Road We Traveled ¨ Challenges ¨ Advantages ¨ Cost ¨ Questions ¨ Demonstration of IDSS – Answers to the Three Questions + More

Data Warehousing Nuts and Bolts

In the Beginning… ¨ We met, discussed, and did not speak each other’s language ¨ BUT we had a common goal

Data Driven Decision Making

Building the Framework Data Elements

Data Compatibility/Cleansing ¨ How is it all going to connect? ¨ Is the data congruent? ¨ Is the data correct?

Where Do We put this Data?

Nuts and Bolts Daily Maintenance ¨ Constant Data Downloading ¨ Constant Compatibility Issues ¨ Daily Transaction Processing ¨ Server Growth ¨ Security Issues

What Teamwork Accomplished ¨ Data Accessibility ¨ Core Reporting ¨ Data Driven Decision Making

Questions