Is it ok for colleges to use smartphones to follow students everywhere they go?

As the new year arrives, I will be writing about some topics that are likely to enter the national conversation in 2020.

First up is establishing limits around what data science businesses are allowed to do. The dam really broke on this issue after the Cambridge Analytica scandal. (See my coverage here.) Since then, the mainstream press has done a commendable job speaking on behalf of the silent plurality (?) who have concerns about how far the big data industry has pushed. These journalists continue to uncover new examples of businesses trying to hide or confuse users about what data they are collecting on them and for what purpose.

In these discussions, it is important to distinguish between what is already implemented, what can be implemented, where industry insiders say they draw the line, whether we can trust what they tell us, and whether and why government regulators are silent or slow to act.

***

The Washington Post published a well-researched article describing how some colleges and universities in the U.S. have started using smartphones to amass databases of student locations. I’m hopeful that the youngsters may finally re-assess their relationship with “smart” technologies. The double-edged sword raises many difficult questions, and not just for students, since the same surveillance technology affects the rest of us.

For those who want to skip reading the Post article, here is a summary:

Some college administrators have purchased technology that tracks students everywhere they go through their smartphones. Several uses of the location data are described. Some professors use the data to take class attendance. The software reveals who walks in to class two minutes late (in reality, location tracking is precise down to fractions of a second.) Athletic coaches use the data to ensure student-athletes remain in compliance. Vendors analyze the data to generate early warnings of “troubled” students [my added quotation marks]. Data privacy advocates, some educators and some students are voicing concerns about the negative implications of such technologies.

There is so much to discuss here that I am breaking it up into multiple posts.

In Part 1, I spotlight a few myths around the topic of data privacy that this Washington Post article forcefully debunks. 

Bluetooth and Wifi expose your whereabouts.

Smart phones and smart devices collect our precise locations, not merely through GPS or cell towers. The preferred location tracking technologies today are Bluetooth and Wifi. Google engineers realized this long time ago – remember when they were caught pulling down Wifi network names and passwords using the vans that take photos for Google Maps? (You might not remember this episode so here is a link to get started. Also, I made a video about this.)

Typically, apps are necessary to operationalize location tracking. Many apps are free only because they are quietly collecting location data from users that are then sold to the highest bidders. (Most famous among these are the weather apps, many of them now owned by IBM. Here is pre-woke mainstream reporting on the IBM acquisition, and here is post-woke reporting, three years late but still welcome.)

We are far from the world of “data exhaust” in which our location is revealed because our cell phone talks to nearby cell towers as a basic function. What’s happening today is that specific technologies (e.g. the Bluetooth beacons mentioned in the Post) are created to harness location data, upon which entire businesses are built.

See my previous post on data sleaze for more on this topic.

***

Location tracking is not anonymous.

Consider the following excerpt: “the system can email a professor or advisor automatically if a student skips class or walks in more than two minutes late.” Later, the journalist noted: “Advisors text some freshmen athletes if they don’t show up within five minutes of class.”

These fragments show that the buyers of location data want to effect specific behavioral changes: the coach calls out individual athletes to get them to class pronto! You should always assume that data collection is neither anonymous nor anonymized. Why? If the student’s identity were not known, the software couldn’t have traced it to the coach, nor could the coach have texted the specific student!

In the same way, marketers who pay for location data want to send promotional offers to specific shoppers deemed likely to respond to them. Political parties target “persuadables” ahead of elections. Personalizing services or actions requires the individuals’ identities as well as their contact information. They simply can’t be anonymous or anonymized.

In an otherwise excellent article, the journalist failed to realize that the vendors implicitly admitted they collect personally identifiable information (PII) and that the data are not anonymous. The next time you hear a tech executive justifying data collection with anonymity, you know s/he isn’t serious about data privacy.

***

Location tracking is not aggregated.

What else isn't true? The claim that businesses only sell “aggregated” data.

You can’t aggregate the data when your goal is to monitor individual students. Aggregating data eliminates privacy concerns but the reality is that many buyers of location data attempt to change individual behavior, which requires individual-level data.

Here’s the bottom line: data are collected at the individual level. Aggregating data is extra work. Buyers of data spend higher sums for individual-level data. So data aggregation is higher cost and lower revenue. If your job is to grow profits, what do you think you’ll be selling?

The decision to throw away personal identifiers and only sell anonymous or aggregated data is a conspicuous choice to pursue a less profitable path.

It's not that no one sells aggregated data. Anyone who is "personalizing" anything for real isn't aggregating the data.

P.S. Note that I said location tracking is not aggregated. Location databases can be aggregated, if the vendor so chooses to, but the raw data feeding such databases are definitely not aggregated.

***

Trade-offs.

And yes, you can use a “burner” phone that cannot be traced back to you. Then, you’ll feel the double-edged sword of “smart” technologies. If you are looking for a restaurant, and the phone doesn’t know your location, you’d have to tell the software where you are. It’s a few more keystrokes, or clicks – for some, convenience is worth the price of giving up their location privacy; for others, it’s not.

For the students, having the smartphones on campus clearly enriches their lives in many ways. Now, they learn that these devices are informing school administrators of their daily activities and movements. How will they react? It's good that the conversation is happening now.

 

P.S. [1/9/20] Part 2 of this series is now live. See why totes of smartphones may invade the lecture hall!

P.S.[1/23/20] Part 3 is now up. In this new post, I examine the other use case of student surveillance data - predicting student behavior. 

Comments

[Ken]

I assume that what college administrators do is to require certain apps or websites are used on the phone so that they can work out who it is. Google, etc need to do something so that they can link your name to the phone. Google has become an expert at this, through supplying apps.

The need to force students to attend courses is also becoming more important. It can be part of the metrics and also can be used to justify increased pass rates, even if it is more because of lower standards. Pass rate is now a metric because it indicates greater engagement and better teaching, if you believe the hype. I much preferred the system where students decided how much effort they would put in, and if it wasn't sufficient they failed.

[Kaiser]

Ken: So many interesting issues in one problem! Using pass rate as a measure of quality is highly questionable. Making attendance a key determinant of pass rate - except for discussion-based courses - is also dubious. And all of this is happening on top of rampant grade inflation in which you have to work hard to fail! On the one hand, I like to treat students as adults; on the other hand, those who disrupt classes using their phones aren't behaving like adults.