P-20W Data Models

1

• P-20W Data Models
• Wednesday, November 16th, 2011
• 245 345 pm
• Irene Koffink, Mike Schwartz (NH)
• Cathy Wagner (MN)

2
NHs Dimensional Data Warehouse

• Overview of NH P-20 Data Model
• Dimensional Model
• Non Proprietary
• Supports K-12 and Post- Secondary
• Star Schema

3
Who was involved in P20 data model creation?

• NH DOE SLDS Team
• Project Director, Project Manager, DBA, Business
  Analyst, Education Consultant, DoIT, program
  staff for business requirements and business
  rules.
• Vendor (Including partners)
• Project Manager, Data Model Team (including DW
  architect and systems architect), ETL Specialist,
  EdFacts Specialist, various system developers.
• Comprehensive involvement and support of the SLDS
  
• Support from various stakeholders
• (see next slide)

4
Comprehensive Involvement
Institutes of Higher Education
I4see Advisory Group
Data collection and use -- district
advisors PerformancePLUS vendor and DOE
development team Collaborative workgroup with
DHHS Pilot districts for data sharing and
collection Meetings of every state
Superintendent Key representative from every
LEA CIA, RTI, SINI/DINI and other support
groups DOE Training Team LEA Support
P20 collaboration P20 Multi Agency Defined
goals Multi State High School Reform common
metrics NH DOE Leadership Postsecondary and
policy research organizations DOE and vendor
development team Postsecondary and DHHS
legislative team
Governors P16 Council
PDevelopment
NESSC
DHHS
Cabinet
DistrictPilot Group
Research Group
Superintendents Meeting
Data Warehouse Development
I4see Coordinators
Legislative Team
Program Partners
P Trainers
5
Pros and Lessons Learned

• Pros
• Learned a lot about the process of developing a
  dimensional data model, ETL, cube development
• Flexibility in slice and dice of data
• Non-proprietary -- No ongoing maintenance/licensin
  g costs
• Using BI tools that can be used with other
  applications staff is familiar with these tools
• Eden Reporting
• Lessons Learned
• Complexity, Time, Staff, In-house knowledge
• Consider how you will use the DW real time vs
  snapshot (how often to run ETL)
• Dimensional model may be resource intensive --
  performance
• Important to define role and responsibilities of
  vendor -- clear outcomes
• Consider current effort to connect data across
  systems -- if all systems are proprietary then
  DW needed, if not need may be less.

6
P-20W Data Models

• Contact Info and Resources
• Baron Rodriguez, [phone],
  Baron.Rodriguez_at_slds-sst.org
• Irene Koffink, [phone], Irene.Koffink_at_doe.n
  h.gov
• Mike Schwartz, [phone],
  MSchwartz_at_ed.state.nh.us
• Cathy Wagner, [phone], cathy.wagner_at_state.m
  n.us
• To Request SST Support
• Fill out a TA Request
  
  
• Form on GRADS