IPEDS Data Retrieval

1

• IPEDS Data Retrieval
• Center Design

SHEEO PresentationApril 17, 2008
2
Project questions Methods

• Questions
• What are the needs of IPEDS data users,
  especially novices and those outside the
  traditional institutional research (IR)
  community?
• How well are the current IPEDS online tools
  meeting those needs?
• What additional tools or features of current
  tools are needed for a complete suite of tools
  and IPEDS Data Center?

• Methods
• Unstructured interviews/focus groups
• Extensive review of current tools including flow
  and functions
• Review of tools in comparison to other available
  tools
• Documentation of different audiences for tools
• Analysis of web server logs

3
Server log analysis
4
Server log analysis
5
Server log analysis cont.
6
Server log analysis cont.

• Only 2007 data available, w/ no discernible trend
  across months
• Almost 10K users of the IPEDS tools each month
• Peak use in February, slight decline in May
• DAS gets little attention w/ 228 average users
  per month. Slightly more than half of all traffic
  is to the PAS (4,848), followed by the DCT
  (2,600) and the ExPT (1,666)
• Only half of users (54.3) get to the point of
  creating a product, with DAS 25.9, ExPT 33.6,
  PAS 56.3, and DCT 76.7

7
Sample feature comparison
8
Types of tools

• Data comparison tools (17, including 5 at ED)
  IPEDS DCTIPEDS PASIPEDS ExPTNCES Compare
  Public LibrariesNCES Compare Academic Libraries
• Institution Finder/Profile tools (27, including 8
  at ED) NCES College NavigatorIPEDS COOLOPE
  Campus Security StatisticsEquity in
  AthleticsNCES Public School District SearchNCES
  Public Private School SearchNCES Public
  Library SearchNCES Search for Schools, Colleges
  Libraries
• Table Generator tools (12, including 2 at ED)
  DASOLNCES State Education Data Profiles

9
Types of tools

• The 54 tools included 112 different features 85
  sub-features
• The availability of features varies with the
  tools
• Common basic features include - viewing results
  online in HTML format (52 tools) - providing
  contact information (51) - selecting schools
  based on criteria (41) - presenting search
  results (39)
• Evolution of the tools is still at an early
  stage, subject to the vagaries of funding
  organizations unique vision for data sharing
• While there are models to follow, these are more
  in the use of specific features, such as
  crosstabs, dynamic reports, widgets

10
Study conclusions

• IPEDS data are being used more than ever in a
  wide range of publication types
• IPEDS is viewed as an important source of data
  for dissertations peer-refereed journals
• The IPEDS tools are rich in features meet a
  variety of needs
• Many models exist for how to share data, but none
  are any more compelling than those of IPEDS,
  which are as good if not better than other
  available tools

11
Key study recommendations

• Improve the time it takes to retrieve data
• Improve displays
• Facilitate simple requests
• Simplify the process to decrease abandonment
• Break apart the tools into features that could
  function individually, with interoperability
• Step-by-step process
• Integrate data from other sources
• Provide different levels of aggregation, with
  drill-down capability
• Provide a single login that works across all four
  tools

12
Key study recommendations cont.

• Include a set of unique audience pages
• Create a My IPEDS page
• Convert all of the tools to .NET make more use
  of widgets
• Modernize and increase navigation capabilities,
  with a clear road map
• Create a Report Wizard
• Incorporate web-based, dynamic, data-driven
  reports
• Allow three types of access (1) final data (2)
  preliminary data and (3) collection level
• Keep all user IDs, passwords, peer comparison
  groups, master variable lists, table parameter
  files online, stored in a special NCES database
  table which allows users to manage share these
  files

13
Recommendationsother features

• Alpha browse of school names
• Select/de-select institutions by type (Title IV,
  tribal, other school types)
• Automatically incorporate allocated parent/child
  data
• Provide additional selection typologies
• Toggle back forth between yearly trend data
  sector, control, level, Carnegie classifications,
  state, location, etc.
• Provide more single-record and merged data
• Make qualifying and categorical variables easier
  to use
• Clarify simplify key data
• Increase training for all levels of users,
  especially for analysis
• Acquire needs data from audience groups

14
User-specific recommendations

• Diverse Users
• Make the tools question-driven
• Dont make the user learn all about IPEDS just to
  get the data
• Provide lists of schools compiled using different
  selection criteria
• Provide reports and datasets related to specific
  research/ policy analysis topics

• New Users
• Include help manuals that are searchable
  printable, including material written especially
  for new users, audience-specific manuals, a
  glossary of terms, and contextual help
• Explain confidentiality and perturbation
• Tell new users what is not in IPEDS