Cameras in Class: Insight vs. Privacy

Having previously written about using analytics to improve education, I know it's a topic that's important to many readers here on All Analytics. But I don't believe facial recognition analytics has ever entered our conversation… until now.

Today I'm bringing you news about SensoStar Labs, which believes that, yes, facial recognition can be used to improve education. It's created the technology, called EngageSense, to make that happen. EngageSense works by analyzing the facial expressions and eye movements of individual students to see if they're engaged in the lesson or have their minds elsewhere. An analysis of that would let teachers know what's working and what isn't in their lessons.

In an email interview, I chatted with SensorStar co-founder Sean Montgomery, who is also an engineer and adjunct instructor at New York University, about the benefits he sees for his technology and the concerns over student privacy and the Orwellian feeling suggested by cameras in the classroom. With respect to benefit, Montgomery said it's a matter not just of recording but of efficiently extracting the key data through technology:

The idea of using cameras in classrooms to improve teaching is not a new one, and in fact many classrooms now record lesson plans for teachers to review after the fact and improve their lessons. But this means doing double time, so that for every hour of teaching, there is an hour video to review.

EngageSense makes it possible to extract the essential points without spending the same amount of time in reviewing the reaction to the class. The algorithms point to the "portions of a lesson in which students were particularly engaged or losing interest." It's not limited to a single class but shows "changes over class periods, weeks, or even whole semesters," giving the teacher feedback on progress.

The same kind of feedback can be used "as a tool for students to learn how to learn better," Montgomery said. As EngageSense can make students aware of their own reactions, students with attention and affect disorders can benefit from its insight to work on "how they engage and review material in the context of their own learning style to make the most of their education."

On the question of privacy, he said EngageSense can be anonymized for more general feedback that's not linked to particular students. He suggested individuals consider the potential benefits weighed against their own privacy concerns, the same way they have to consider what they hope to gain out of using an app that picks up on their data.

Some individuals may decide to let the teacher use anonymous engagement statistics to improve their teaching. Others may see the value of giving the teacher full data access so that they can get more personalized assistance. Others still may opt out entirely.

Speaking for himself, Montgomery said he'd consider the benefit of a full personal data review worthwhile for his own child. Plus, anything EngageSense would record would already have been done in front of that teacher live.

As for those who bring up 1984, he acknowledged that people do have valid privacy concerns but said that dropping terms like that does not really help clarify the real issues at hand:

The 1984-type comments about EngageSense in particular are possibly a tad ironic, given that these comments are often made on corporate websites whose purpose is to track as much as possible about you and make a profit selling your data to advertisers. That aside, I think people are justifiably nervous as we enter into a future where privacy and data ownership is likely to be understood very differently than it is today. It's my hope that people will take the public discourse about privacy seriously and move the beyond 1984 doomsday comments into a more nuanced discussion about controlling access to your own data and giving the individual the right to trade data for service.
What do you think? Would you support cameras in the classroom toward the end of improving the educational experience of students?

Related posts:

Ariella Brown,

Ariella Brown is a social media consultant, editor, and freelance writer who frequently writes about the application of technology to business. She holds a PhD in English from the City University of New York. Her Twitter handle is @AriellaBrown.

Hiring Trend: Machine Learning Reduces Bias, Increases Applicant Pool

Unilever is using machine learning tools to increase the pool of entry level applicants and then test and evaluates candidates in the pool to cull those who are the best fits. Adding machine learning tools reduces bias, increases diversity, and makes the whole process more efficient, according to Unilever.

Analytics and School Attendance: A Laundry Story

Appliance maker Whirlpool suspected there was a correlation between student attendance and access to clean clothes. Here's the story of how the company placed washers and dryers in schools and tracked the difference it made in student performance.

Re: Cameras in class
  • 12/12/2013 7:34:57 PM

Having cameras in the office of a school is okay in my book, but I wouldn't want them in my child's classroom. I could not imagine being a teenager trying to learn-- knowing that there is a camera watching me. 

Re: Cameras in class
  • 12/12/2013 2:18:33 PM

@Srs1 these cameras are not for security, though, I assure you, schools already have those. My own kids' school has several with monitors on them in the office.

Cameras in class
  • 12/12/2013 2:06:27 PM

Cameras in class sounds like an invasion of privacy, unless the school has a crime problem it is not needed. First we let the government invade our privacy at the airport then at home, and now at school?

Re: Mandatory attention for the classroom
  • 12/8/2013 8:55:20 AM

@magneticnorth perhaps, there are always going to be some teachers who are not truly competent or motivated. But they are answerable to administrators and others within the school hierarchy.

 I do get the impression that the teachers would only be given the anonymized results to see how their lessons fare unless they are given express permission to see individual student reactions. Individual assessment would only be done if parents agree, and the teacher would not get away with blaming indiviudal students if a signficant percentage of the class is shown to not pay attention at the same points in the lesson. They can't pretend it's the one kid's fault if 10 out of 20 kids all show the same reaction. And they can't erase that part of the data.

Re: Sharing a stigma?
  • 12/8/2013 4:32:19 AM

In my experience, test results from guidance counselors would be handed to students (after interpretation), and the student will decide whether or not to give the results to teachers. That's respect for privacy right there. But in the case of EngageSense, data is gathered in a teacher's classroom and will be used by the teachers. It'll be hard for teachers not to get influenced by the data.

Re: Mandatory attention for the classroom
  • 12/8/2013 4:21:50 AM

@Ariella I'd have to agree with @Lyndon_Henry on this one. There are many teachers out there who give tests just to have grades to submit and not to accurately gauge performance against objectives. These are the same teachers who will misuse data from EngageSense.

Re: Sharing a stigma?
  • 12/3/2013 4:22:02 PM

Ariella, regarding that comprehensive record based on hours of class interaction -- I wonder if even better would be a comprehensive record based on years of class interaction. One interesting option for schools would be to start this type of analytics program with each consecutive incoming kindergarten or first-grade classes (vs full-school rollout from the get-go) to it can establish year-over-year tracking from the start. This would be kind of like the tablet rollouts I've seen some schools undertake -- incoming freshman first, say.


Re: Sharing a stigma?
  • 12/3/2013 11:41:30 AM

@Beth I think it does have that potential, yes. But I think that it has quite a bit of upside going for it. As I mentioned, schools tend to administer IQ test and put a lot of faith in those even though they are only a single snapshot of a child's performance in a single half an hour. If the child would have a more comprehensive record based on hours of class interaction, it would likely be more accurate.

Re: Mandatory attention for the classroom
  • 12/3/2013 11:39:40 AM

@kicheko unfortunately, I think that today's attention span is down to 20 seconds. Actually, that's the figure I was told for attention on the phone when making cold calls. As for in class, I don't know, I'd imagine it's whatever span TV programs run now between commercials.

Re: Evaluating the Lesson
  • 12/3/2013 11:26:56 AM

I have never liked early morning classes because i have to wake up so early. But when i think of it carefully, this is the best way to make myself maximumly productive on any day. Because my main job is in the research lab with just a little teaching work, its better when i can jump into the deep end with teaching first because then i have no option to waste 30-40 mins adjusting to the work day. By the time i leave class i am fully adjusted. As for keeping time, i realized it earns you a lot of credibility with students.

Page 1 / 4   >   >>