Creswell follow-up responses to Commission in opposition to a comprehensive federal student database

Following Cassie Creswell’s testimony on behalf of our Coalition on January 5, 2017 to the Commission on Evidence-Based Policy, the Commission sent her some follow-up questions.  Here are her responses.

10 February 2017

Commission on Evidence-Based Policymaking

Washington, DC

 

Dear Commission Members:

Thank you for requesting additional information about the Parent Coalition on Student Privacy’s position on unit record systems. Below we address all three of the Commission’s follow-up questions:

  • Can you please clarify whether your objection is to unit record systems for elementary and secondary school students or if you also object to unit record systems for postsecondary students?

See the discussion of this issue below.

  • You argue for local control of student data and specifically for parental and teacher data stewardship. At what point should adult students in postsecondary education become the stewards of their own data?

The control of their data should pass to post-secondary students when they become adults, as occurs currently in Federal law.

  • Would your concerns about the intrusiveness of a student-unit record system be mitigated if the data were only maintained without personal identifiers that could be used to track an individual student, and if there were statutory protections that guaranteed that the data could only be used for aggregate, statistical analysis?

No, because data can easily be re-identified. If only aggregate data is used, only aggregate data should be collected.

We have significant reservations about the creation of any universal unit record system for students, whether for elementary, secondary or post-secondary students.

Elementary and secondary student unit record systems present a particular set of risks because the majority of information in a child’s K12 educational record should not be made “permanent.” Childhood is a time of growth, experimentation and development; and mistakes and challenges should not be part of a record that could follow one into adulthood and hamper a child’s chance of future success.

A unit record system for post-secondary education does not present an identical set of concerns. Students in post-secondary institutions do expect that some aspects of their transcript, including grades and credits, will persist into adulthood with the expectation of being  shared with employers and other educational institutions—with their consent. Yet other contents of their education records should never be made public.

Records from these years may also contain sensitive information about immigration status, counseling records, mental and physical health and disabilities, etc. At the age of 18, control of the record is transferred from a parent or guardian to the students themselves, but, crucially, privacy controls are still maintained.

We have several concerns about the need for and use of any universal post-secondary unit record system:

  1. The efficacy of methods to de-identify or anonymize personally-identifiable data is questionable. De-identified data can often be re-identified and exposed.[1]
  2. The government should not have access to a comprehensive database for all post-secondary students as this information could be easily abused for political or immigration reasons.  This is especially of concern given the current political climate. The Home Office in Great Britain has now requested access to a similar government student database for the purposes of “immigration control” that was promised to only be used for research.[2]
  3. Once the federal government starts collecting post-secondary data, this could easily lead to a creeping expansion of data collection from K12 institutions and districts.
  4. The quality of research based on large-scale correlational studies is of greatly varying quality[3] and does not justify the risks of universal tracking.
  5. Large amounts of data used for the purposes of evaluating post-secondary institutions’ effectiveness are already available, including the Department of Education’s College Scorecard, the Mobility Report Card Project—a collaboration of the US Treasury and the Department of Education, and the privately-run National Student Clearinghouse.
  6. Extensive regulations have already been implemented to ensure that post-secondary institutions are protecting student’s long-term financial interests, e.g. Negotiated Rulemaking on Gainful Employment implemented in Fall 2014.

We acknowledge that given the investment of taxpayer funds that support institutions of higher education, the federal government has a strong practical interest to make certain that those funds are being used efficiently and effectively.

We do not, however, think that a universal student record system created and administered by the federal government is a necessary component of fulfilling that interest and duty.

The federal government spends billions of dollars in medical research and health care as well, and yet there has been no proposal that we know of for the federal government to collect the personal health data for every person in the United States.

We support only the use of aggregated student data for the evaluation of postsecondary institutions. The collection of such data must include asking for consent for participation from either a parent/guardian or the students themselves if over the age of 18. Clear, transparent information about how any data is to be used and who it may be shared with must be presented before asking for consent. And, there should be no financial benefit or loss contingent on granting the consent.

We continue to urge the Commission to recommend against the creation of any universal federal student unit record system.

On behalf of the Parent Coalition for Student Privacy,

Sincerely,

Cassandre Creswell, PhD

Co-executive director

Raise Your Hand Action

Chicago IL

_____

[1] See, for example, the research of Latanya Sweeney on identifiability.

[2]The Home Office are turning teachers into immigration officers” G. Bhattacharyya. Politics.co.uk.

[3] See discussion from a variety of fields (medicine, psychology and linguistics) in “Data dredging, bias, or confounding.” Smith, G.D. and S. Ebrahim. BMJ. 2002; “Why Most Published Research Findings Are False.” Ioannidis, J. P. A. PLOS Medicine. 2005; “False-Positive Psychology: Undisclosed Flexibility in Data Collection and Analysis Allows Presenting Anything as Significant.” Simmons, J.P., L.D. Nelson and U. Simonsohn. Psychological Science. 2011; and “Linguistic Diversity and Traffic Accidents: Lessons from Statistical Studies of Cultural Traits.” PLoS ONE. 2013. Roberts S. and J. Winters.