Data Collection

It can be easy to toss our noses up at people who complain about data collection on kids, saying that they are being paranoid. But then an essay like this comes along, and instead of shaking our heads, we have to deal with the fear the reality brings. Excerpts:

A particularly troubling aspect of the Common Core scheme is the emphasis on massive data-collection on students, and the sharing of that data for various purposes essentially unrelated to genuine education. U. S. Secretary of Education Arne Duncan has said:

Hopefully, some day, we can track children from preschool to high school and from high school to college and college to career... We want to see more states build comprehensive systems that track students from pre-K through college and then link school data to workforce data. We want to know know whether Johnny participated in an early learning program and then completed college on time and whether those things have any bearing on his earnings as an adult. 

To know all this, of course, we have to know pretty much everything Johnny does, throughout his lifetime.

What kinds of data are we talking about? The National Education Data Model includes over 400 data points, including health history, disciplinary history, family income range, voting status, religious affiliation, and on and on.

Pursuant to this enthusiasm for sharing student data, where might that data end up? One illustrative example (and an obvious one, in this workforce-development model) is departments of labor. In fact, USED [US Education Department] and the U. S. Department of Labor have a joint venture called the Workforce Data Quality Initiative, the purpose of which is “developing or improving state workforce data systems with individual-level information and enabling workforce data to be matched with education data.” Student education data are to be shared, to the extent possible, with labor agencies to promote the goal of workforce development.

Of course, under the new regulations that gutted federal student-privacy law, data can now be shared with literally any agency if the correct enabling language is used: the Department of Health and Human Services, Homeland Security... the IRS?

Both USED and state education officials further insist that privacy concerns are overblown, because student-level data will be anonymized. In the first place, this is simply not true – the data coming from the Common Core assessment consortia, and the workforce tracking data showing which students participated in which education programs and then earned which salaries, are necessarily student-specific.

In the second place, in the era of Big Data, there really is no such thing as anonymization. When there are multiple, perhaps hundreds, of items in the database, the absence of a name or a Social Security number becomes a mere inconvenience, not an obstacle to identifying the student.

There are many examples of data re-identification.  For instance in Kentucky in 1999, a researcher was able to match over 2,300 students who appeared on anonymized lists of test-takers – and the match had 100% accuracy.  And this was almost 15 years ago – long before education bureaucracies were collecting the myriad data they are now.

Where are we headed with all this? It is instructive to look at what USED itself is working on and writing about.

One report that appeared on the Department’s website in February of 2013 is called Promoting Grit, Tenacity, and Perseverance. The thesis is that education must inculcate these qualities in students, and that their presence or absence must be measured in some way. How? The report suggests assessment of physiological reactions that a student exhibits to stimuli such as stress, anxiety, or frustration. These reactions could be measured through posture analysis, skin-conductance sensors, EEG brain-wave patterns, and eye-tracking, (pg. 41-45).And the report barely mentions the appalling invasion of privacy this kind of physiological measurement would entail; rather, it focuses on the “problem” that this isn’t practical for the classroom – yet, (pg. 45).

I haven’t even mentioned the ever-present problem of data-security. Hacking into student databases will occur, and in fact has already occurred. The wealth of data collected on students and their families is a hugely tempting target for people with malicious motives. But as serious as this problem is, the deeper problem is that the government has deemed our children little machines to be programmed, “human capital” to be exploited...

This is real.