Politics Foreign Affairs Culture Fellows Program

Common Core and Data Mining: Fact and Fiction Part II

Thoughtful commenters on last week’s “Fact and Fiction at CPAC Common Core Panel” expressed some disappointment with the piece’s lack of in-depth analysis. I hope to respond to these concerns, and provide greater information on the “fact and fiction” surrounding Common Core. Unfortunately, it will take (a lot) more than 500 words. This article looks […]

Thoughtful commenters on last week’s “Fact and Fiction at CPAC Common Core Panel” expressed some disappointment with the piece’s lack of in-depth analysis. I hope to respond to these concerns, and provide greater information on the “fact and fiction” surrounding Common Core. Unfortunately, it will take (a lot) more than 500 words. This article looks specifically at questions and concerns over student privacy with the new standards.

Question: Do the Common Core State Standards encourage or require government data mining of students’ personal information?

Answer: Not directly. The real culprit, it seems, lies with federal dollars given to states—take Obama’s Race to the Top: an initiative that gave federal dollars to states in exchange for their commitment to “four key areas of reform.” Though the reforms never reference the Common Core by name, Race to the Top called for “development of rigorous standards and better assessments.” It also required “Adoption of better data systems to provide schools, teachers, and parents with information about student progress” (this is where concern over data mining comes in).

Also, the State Fiscal Stabilization Fund (SFSF) offered stimulus money to governors in 2009 for state education. SFSF money came with a very specific set of qualifications for states: it doesn’t reference Common Core by name, either, but required that states “progress toward rigorous college- and career-ready standards and high-quality assessments that are valid and reliable for all students.” Its second qualification, though, has many parents up in arms: “The state must assure that it will take actions to … establish and use pre-K-through-college and career data systems to track progress and foster continuous improvement.”

In sum, SFSF told states it has to track the students’ educational progression from preschool through college, and even beyond.

The Council of Chief State School Officers (CCSSO) is leading the actual data-collection effort on the ground. This is where things get interesting, and why many people are grouping Common Core with the data mining effort: CCSSO is co-holder of the Common Core Standards copyright. They’re developing a system of “statewide longitudinal data systems (SLDSs),” that will track students from pre-K through college (and career).

What data does CCSSO intend to gather via SLDSs? In 2010, the National Center for Education Statistics released a technical brief titled “Guidance for Statewide Longitudinal Data Systems (SLDS),” with a little information on the sort of data they’re likely to include: student’s name, name of parents or family members, address of his/her family, a personal identifier (like a SSN, student number, or biometric record), and “indirect identifiers” such as a date of birth, place of birth, or mother’s maiden name.

In a New American article on student data mining, Mallory Sauer claimed that the Education Department brief called for extraction of sensitive information “which delves into the intimate details of students’ lives.” But the list of “sensitive information” she provides from the report is provided with a clear disclaimer: the collection of such information, it says, would require “written parental consent” before any minor could be required to participate in “any survey, analysis, or evaluation” funded by the Department of Education: “In the event any data elements under consideration for inclusion in a student record system involve any of these eight topics, those data elements should be included on the inventory of PII and should be identified on the list as PPRA-related variables.” In other words, the report warns such information cannot be involuntarily collected. Sauer continues:

Parents might reasonably assume that the “personally identifiable information” collected for the database will include students’ test scores and perhaps other measures of academic proficiency. But they would be much less likely to imagine that the federal government envisions something far more extensive and invasive than merely tracking academic performance. According to the Department of Education’s February 2013 report Promoting Grit, Tenacity, and Perseverance: Critical Factors for Success in the 21st Century, “Researchers are exploring how to gather complex affective data [emphasis added] and generate meaningful and usable information to feed back to learners, teachers, researchers, and the technology itself. Connections to neuroscience are also beginning to emerge.”

However, the passage Sauer pulls from the Department of Education’s report refers to “new opportunities” in behavioral task performance measures—not already-established measures. Neither does the report ever explicitly outline what “complex affective data” is—the passage before Sauer’s pulled quote refers to data collection regarding students’ response to specific educational tasks:

Dependent behavioral variables associated with a challenge at hand may include responses to failure (e.g., time on task, help-seeking, revisiting a problem, gaming the system, number of attempts to solve a problem, use of hints), robustness of strategy use (e.g., planning, monitoring, tools used, number of solutions tried, use of time), level of challenge of self-selected tasks, or delay of gratification or impulse control in the face of an enticing off-task stimulus. Such data can be examined for discrete tasks or aggregated over many tasks.

This sort of psychological analysis may be alarming to many parents, who want their students to learn (or not learn) without a wealth of education officials tracking and analyzing their behavioral patterns. But the amorphous and ominous reference to “gathering of students’ personal information” may create false apprehensions for many parents. There is reason to be frustrated with federal efforts to grow data mining—but we should be very clear what those frustrations are about.

Additionally, CCSSO’s site explicitly states that adoption of SLDSs is strictly voluntary. As an Education Week article on the data collection effort put it, “The new federal guidelines are non-binding and contain no new regulations, reflecting a desire to encourage ‘self-policing’ by industry and better policies and practices by school systems as first steps towards shoring up students’ privacy protections.”

There is some serious concern that data could be leaked to third parties, for non-educational purposes. Although the Family Educational Rights and Privacy Act (FERPA) was created to protect children’s personally identifiable information from such use, a Department of Education document released in February, titled “Protecting Student Privacy While Using Online Educational Services: Requirements and Best Practices,” shows that such laws are “somewhat limited in their power to prevent such outcomes in the new age of ‘big data’ and ubiquitous digital learning tools,” says Education Week:

Take, for example, the ‘metadata’ collected on students via digital devices and online learning programs, which can include keystroke information, the time and place at which a device or program is being used, the type of device on which the service is being accessed, and more. Under some circumstances, such metadata are not protected under FERPA and may thus eligible to be used for data-mining and other non-educational purposes.

Most adults are well aware of such “big data” collection issues, considering the recent NSA scandal and targeted advertising of such online giants like Google and Facebook. The frustration here, of course, is that the subjects of this data collection are children—and the fear that such educational data collection may be mandatory. Whereas an adult can refuse to use the Internet or engage in social media, children are required to go to school. Concerned parents should read the full Education Week story: it offers an unbiased look at data collection’s benefits and detriments.

Some states have already started implementing the new information gathering practices, and are using a private database managed by inBloom, Inc., a private organization funded by the Bill and Melinda Gates Foundation (the foundation has also given a considerable amount to support Common Core). inBloom does work with some third party affiliates: its website says it “enables a wide range of applications to use student data more effectively and efficiently, with the ultimate goal of unleashing innovation in the marketplace and making personalized learning a reality.” They provide a list of such partners on their website—people who will share access to their “Data Store.” They note on their FAQ page that “Districts that use inBloom … may choose to disclose certain student information to trusted third-party providers to power the learning applications that are implemented in their classrooms. Those disclosures are controlled by the school district, not inBloom.” Additional information from the page: “States and districts determine what student information is stored, how this information is used, and who gets access to the records, just as they always have.”

Now that we’ve looked at all the other state and federal groups calling for data mining, let’s look at the Common Core standards themselves: do they ever “require children’s personal information to be provided to a database,” as Sauer put it? After perusing their website at length, I found this information on the FAQ page:

Are there data collection requirements associated with the Common Core State Standards?
No. Implementing the Common Core State Standards does not require data collection. Standards define expectations for what students should know and be able to do by the end of each grade. The means of assessing students and the data that result from those assessments are up to the discretion of each state and are separate and unique from the Common Core.

Yes: the Race to the Top, SFSF, and CCSSO are calling for more data collection. And there are most definitely ties between Common Core and all three of those initiatives. But Common Core itself does not “require” data mining. Neither is the new SLDS system mandatory for school districts. It’s a purely voluntary effort. There may be pressure to adopt it in exchange for federal funds, but states can choose to accept or reject such incentives.

Stay tuned for later articles on Common Core fact and fiction, including (but not limited to):

– Does Common Core replace literature with informational texts in English classrooms? If so, what ratio will be replaced? Will this become an opportunity for leftist propaganda to infiltrate our schools?
– Will the new Math standards utilize Soviet-era math curricula?
– Is the Common Core initiative truly state-led?
– Could the implementation of Common Core lead to the adoption of a mandatory nationwide curriculum?


**Correction: While the CCSSO, as part of the CEDS Consortium, is “enthusiastically supporting” the development of statewide longitudinal data systems, they are not actually developing the systems, nor do they collect any student data.