Extracting Useful and Targeted StateLevel Data from IPEDS

1
Extracting Useful and Targeted State-Level Data
from IPEDS

• Experiences from the Land of 10,000 Lakes

2
Minnesota Measures

• First of a planned annual series of reports
  related to accountability
• Timeline
• May 2005 (initial charge to work on project)
• August 2005 (appropriations, NCHEMS contract,
  initial meetings)
• November 2005 (discussion of goals and
  indicators)
• January 2006 (statewide meetings)
• March 2006 (review of goals and indicators)
• June 2006 (NCHEMS final report)
• September 2006 (meetings with systems)

3
Indicators Goal 1Improve success of all
students, particularly students from groups
traditionally underrepresented in higher
education.

• College participation rates
• IPEDS data
• First to second year retention
• 3-, 4-, and 6-year graduation rates
• Degrees awarded as a proportion of total
  headcount enrollment
• Degrees awarded in critical fields (STEM and
  healthcare), disaggregated by race/ethnicity
• Proportion of young adults (25-34) in the state
  holding a postsecondary degree

4
Indicators Goal 2Create a responsive system
that produces graduates at all levels who meet
the demands of the economy.

• Credentials awarded at each level (IPEDS), per
  1000 people 20 and older in the states
  population (ACS).
• Proportion of credentials awarded at each level
  in STEM fields and number awarded (IPEDS) per
  1000 people 20 and older (ACS).
• Proportion of credentials awarded at each level
  in healthcare fields and number awarded (IPEDS)
  per 1000 people 20 and older (ACS).

5
Indicators Goal 3Increase student learning and
improve skill levels of students so they can
compete effectively in the global marketplace.

• Did not gather data in this area for the initial
  report.
• Currently looking into a variety of assessment
  instruments.
• Aware that IPEDS COOL will incorporate assessment
  results in the future.

6
Indicators Goal 4Contribute to the development
of a state economy that is competitive in the
global market through research, workforce
training and other appropriate means.

• The share of national academic research being
  done in Minnesota
• The ranking of the University of Minnesota on
  various studies of research activity (University
  of Florida report, London Times, Shanghai study,
  and Newsweek)
• Total research expenditures in the state as a
  proportion of gross state profit

7
Indicators Goal 5Provide access, affordability
and choice for all students.

• Proportion of residents aged 18-24 and 25-44
  participating in postsecondary education (ACS)
• Assigned family expectation (OHE data)
• Using NPSAS data
• Net tuition (after grants and scholarships)
• Average borrowing and rate at which students
  borrowed

8
Some Findings

• In general, Minnesota does not consistently rank
  among the top states. More often, were near the
  national average.
• The degree attainment of our citizens is high,
  but that is due in part to in-migration of
  college-educated citizens from other states.
• Native American, Black, and Hispanic students in
  Minnesota do not do well in college compared to
  their white and Asian counterparts.

9
Using the Dataset Cutting Tool (DCT) to get
State-Level Data

• Create a custom dataset
• Advantages
• Web-based interface
• Very customizable, can get data from multiple
  files over multiple years
• Can create a file that you can download
• Disadvantages
• Interface can be cumbersome
• Time-out issues
• Limited to 1,000 institutions in creating the file

10
Using the Dataset Cutting Tool (DCT) to get
State-Level Data

• Download entire data file
• Advantages
• Very straightforward
• You get all of the data for all institutions
• Can be imported into a program like SPSS, SAS or
  Access for report generation
• Disadvantages
• You get all of the data for all institutions,
  which includes a lot of imputation fields
• Data dictionaries are cumbersome

11
What We Did

• Download entire data files
• Import into an Access table, to provide control
  over
• Which fields were brought in
• The data type of those fields
• The names of the fields
• Consider a sample
• Building queries helps a great deal, as results
  can be copied/pasted into Excel for easy
  manipulation
• Why not just take data directly to Excel?
• Limits on Excel table size

12
Successes

• Getting state-level data is reasonably easy
• Crosstab queries
• Reports
• State and national averages
• Beware of averages of averages
• Actual averages reasonably easy

13
What We Will Do Differently (regarding our use of
IPEDS data)

• Degrees awarded as a proportion of total
  headcount enrollment will be rethought
• Goal incorporate part-time students into a
  degree completion measure
• Needs more context
• May talk more about and give more detail about
  transfers out
• No more mixing of IPEDS and ACS data
• May look more deeply into the use of DAS

14
Observations

• Data dictionaries (e.g., imputation variables)
• Variable names in general
• Lack of retention data by race
• Which is being addressed in part by new IPEDS
  data collection procedures
• But only for SMART grant fields
• Loss of transfer students and part-time students
  in computation of graduation rates

15
For More Information

• On Minnesota Measures
• The on-line version is available at
• http//www.ohe.state.mn.us/mPg.cfm?pageID1733
• You can also download a .pdf version of the
  entire report from this page.
• On the Office of Higher Education
• http//www.ohe.state.mn.us
• jim.bohy_at_state.mn.us