Data Warehousing: Improved Analysis Through Data Integration

1
Data Warehousing Improved Analysis Through Data
Integration
James Roche, Director, Institutional Research,
Washington State University Fran Hermanson,
Director of Student Affairs Research, Washington
State University Janet Danley, Executive Director
of Enrollment Services, Washington State
University

• April 20, 2004 / AACRAO 2004 Annual Meeting
• Session ID 290

2
Data Warehouse Defined

• A collection of data designed to facilitate
  forecasting and planning and to support decision
  making
• Integration of many different databases related
  to the same topic
• (Data marts are subsets of larger data
  warehouses)

3
Components of anEffective Data Warehouse

• For the user
• Single source for data (same data for all users)
• Reliable
• Accessible (via the web where possible)
• Flexible
• Easy to extract and use in other programs
• Printable reports

4
Components of anEffective Data Warehouse

• For the manager
• Easy to update
• Automated wherever possible
• Needs little management once established
• Clearly defined meta-data (data about the data)
  and table/query relationships
• Quality inquality out

5
Automation of Monday Report

• Windows Scheduled Tasks triggers batch files
• DOS batch files
• One to copy the ASCII file from the mainframe
• One to open Access admissions database
• Access AutoExec macro
• Runs automatically when Access database is opened
• Runs module containing Visual Basic script, which
  converts ASCII file to database table
• Active Server Page (ASP) delivers data

6
Admissions, Recruitment, and Marketing Data
Integration
7
Access QueryShowing Relationships
8
Data Delivery Options

• Standard report delivered on schedule

• Data-on-demand for sophisticated users

• Hybrid approach, web-based with simple sort and
  filter functions

9
Advanced Analysis

• Once the data are in a data warehouse, linking
  them to other analytical software, such as SPSS,
  is a simple task.

10
Lost Market Data Mart

• Improve Your Ability to Target, Select, and
  Retain Students

11
A SuccessfulLost Market Data Mart

• Identify students offered admission but not
  enrolled on census date
• Upload data to the National Student Clearinghouse
  Enrollment Search Service
• Download data from NSC and link existing
  demographic information on students from
  institutional databases
• Link to other outside data to enrich knowledge of
  competition
• Build a longitudinal data mart to analyze your
  market position

12
Top Competitors by EnrolledFreshmen and Transfers
13
School Attended by StudentsState of Residence
14
School Attended ByState Location of School
15
Residency Status and Ethnicity
16
Academic Profileby Residency Status
17
Academic Profile of High-Ability Students by
School Attended
High-ability students are those with test
scores of 1200 or greater.
18
Top Five Competing Schoolsfor Students of Color
19
Cost of Attendanceat Competing Schools
20
Scholarships Waivers at Competing Schools vs.
WSU
21
Lost Market Data Mart Next Steps

• Integrate the lost market data mart with existing
  data
• Further study needed Data mining (automated
  extraction of hidden predictive information from
  large databases)
• Get the information into the hands of decision
  makers
• Set up session to discuss findings
• Brainstorm on strategies for
• Effective recruitment and retention strategies
• Targeting, selecting, and retaining students
• Competitive edge for highly desired students

22
Using the Data

• From Data to Knowledge

23
Using Data forOutreach and Recruitment

• Planning data
• Demographic trends
• Numbers of high school graduates by region
• Numbers of college-bound students by region
• Race and ethnicity
• Gender/age
• Income per capita

24
Using Data for Management

• Weekly admissions reports
• Information
• Management tool
• Assessment and guidance for focused efforts

25
Monitoring Race and Ethnicity Trends
26
Data-Based Targeting
27
Using Datafor Assessment

• Weekly Admissions reports
• Strength of applicant pool
• Selecting applicants to offer admission
• Managing quality of the admitted student profile

28
Strength ofApplicant Pool GPA
29
Using Data to Plan Housing and Dining Capacity

• A freshman live-on rule requires careful planning
  as the size of incoming freshman classes grows
• Housing uses data to determine the number of
  single rooms that will be available to students
• Dining determines capacity levels for each of the
  dining centers based on admission and enrollment
  projections

30
Housing and Dining Data
31
Managing Course Sections

• Using data to plan course sections
• Enrollment patterns from previous years
• Availability of faculty, TAs, and GAs
• Availability of facilities such as labs and large
  classrooms

32
Using Data to Retain Students

• Set tuition
• Establish academic progress rules
• Project majors by college
• Calculate financial aid needs
• Monitor student progress and time to degree
• Plan commencement ceremonies
• Determine faculty hiring needs

33
Using Data for Other Needs

• Curriculum development
• Infrastructure, such as IT
• Securing grants and contracts

34
For More Information

• Janet Danley danleyj_at_wsu.edu
• Fran Hermanson franherm_at_wsu.edu
• James Roche jroche_at_wsu.edu
• Presentation available online at
• www.ir.wsu.edu
• (Click on Presentations)