As a sample of these techniques, please review NIST document "De-identifying government datasets. http://csrc.nist.gov/publications/drafts/800-188/sp800_188_draft2.pdf
Also, please review the U.S. Department of Education process for data release. Note pages 13 and 10. Note the step that considers the "Disclosure Avoidance Methodology" which are outlined in the above referenced NIST document. https://s3.amazonaws.com/sitesusa/wp-content/uploads/sites/1151/2016/10/The-Data-Disclosure-Decision-Department-of-Education-Case-Study_Mar-2015.pdf
Therefore, please employ your own disclosure avoidance methodology and de-identify the dataset because the public wants the data for the purposes of transparency and accountability. Please provide in one or more Excel files and database (.dbf) files containing individual test score performance on standardized tests (PARC, SAT, AP) taken in 2014-2016 school years. And for each observation provide: test offered, year offered, numerical and categorical data reflecting test grades, grade level of test taker, race, whether or not the student was eligible or receiving Free and Reduced Meals, and any other relevant demographic information.
As stated by ed.gov, "Educational agencies and institutions are permitted to release, without consent, education records, or information from education records that have been de-identified through the removal of all personally identifiable information.” . Reference: Dear Colleague Letter about Family Educational Rights and Privacy Act (FERPA) Final Regulations, U.S. Department of Education, December 17, 2008. http://www2.ed.gov/policy/gen/guid/fpco/hottopics/ht12-17-08.html
All MPIA requests are purposes of research and analysis.
At the individual student-level, the specific items requested, including PARCC Score, Year Taken, and Categorical Data (pass/fail), SAT Score and Year Taken, and Categorical Data (pass/fail), AP Score and Year Taken, and Categorical Data (pass/fail), Grade Level of Test Taker, Race and, FARMS, even with personally identifiable information (PII) removed, have the potential to reveal educational records under FERPA. According to the U.S. Department of Education “The simple removal of nominal or direct identifiers, such as name and SSN (or other ID number), does not necessarily avoid the release of personally identifiable information” (FERPA Federal Register Document – attached for reference, see page 74831). In order for education agencies to avoid disclosure, “educational agencies and institutions releasing public reports derived from students’ education records must perform an analysis of the data and apply statistical disclosure limitation methods to remove any PII from those reports prior to release” (ED OCPO Letter – attached for reference).
As you indicated in your request and the referenced documents above state, those fields at the individual student-level would require us to apply disclosure avoidance techniques in order to be preserve PII. The application of such techniques, much like the technical skills required to program a database to pull new reports/records, constitutes the creation of an entirely new record. Under the MPIA process, and as advised in the Fourteenth Edition of the Maryland Public Information Act Manual created by Maryland’s Office of the Attorney General, an agency is only obligated to produce existing records and is not required to program/generate new reports that are not already used in the normal course of business.
Alternately, existing HCPSS reports, such as the examples below, show the intersection of testing, FARMs and demographic data at an aggregated-level which eliminates the need to apply disclosure avoidance techniques and, with suppression, can be released under the MPIA:
- 2016 AP Board memo: http://www.hcpss.org/f/academics/2016-ap-exam-participation-performance-results.pdf
- 2016 SAT ACT Board memo: http://www.hcpss.org/f/academics/2016-sat-act-results.pdf
- 2016 – 2017 PARCC Performance on the Maryland Report Card: http://reportcard.msde.maryland.gov/downloadindex.aspx?K=99AAAA