鶹Ƶ

Skip to main content

Where Does the Data Go? A Behind-the-Scenes Look at iSAT’s Security Measures for Classroom Data Collection and Handling

By Charis Clevenger

With a Master's in Family and Human Development, 󲹰’s personal research interests include AI in education, relationship building, and learning through collaboration, equity in public schools, and viewing learning through the biopsychosocial model.

Do you ever wonder what happens to student data once the microphones and cameras are out of the classroom? With AI in education, there can be a lot of questions and concerns about how CU Boulder is protecting students’ information, whether it be their name, voice, image, or even the work they submit in class. It is challenging enough to navigate the school age years – worrying about how data remains secure shouldn’t be one of the contributing factors.

My name is Charis Clevenger, and I am the data manager for the Institute of Cognitive Sciences and iSAT. As a mother and former educator, the protection of vulnerable populations including our children is a critical motivating force in my role as data manager. Having been with iSAT since its founding (we are now in year 5), I make it a priority to ensure that we keep up to date with the latest best practices and safest measures for securing the data we collect.

iSAT, as a whole, is committed to following the  where we protect the future from harm by emphasizing a stewardship of science and innovation in the present. Below are some ways how we apply this framework for our research policies on collecting data in classrooms. 

Anonymizing personally identifying information at every stage

The first step after we collect data involves removing any information from the data that can identify a student participant. For this, we use study IDs instead of students’ real names. We also anonymize any information about their context, whether it’s who their teacher is, which school they attend, and what district they are in. Additional measures we take are:

  1. Using untraceable identification numbers,
  2. Blurring videos used for general analysis,
  3. Transcribing speech to minimize the need for additional video use.

Ensuring raw data is secure once collected

Data is kept on secure servers that are password protected. Data collectors follow rigorous cyber security protocols and safeguards such as never “staying logged in” to any data networks.

Additionally, iSAT has put into place the careful curation of datasets based on specific needs from our in-house expert research teams. This happens only after the collected data has been rigorously checked and rechecked for any issue that could reveal identifying information. For example, suppose there is a school announcement made over the intercom during data collection and it may contain identifying information about the school; if this ends up being audible on the recording, we remove it. In doing so, our team ensures that collected data has to pass several levels of inspection and cleaning as well as move through various access control channels before it ever gets forwarded to research teams. And then we also track what data is being used and by whom. This minimizes the access to data that is not necessary to complete research by any given team.

In summary, it is imperative to update and refine security measures that protect the privacy of student participants. That is why iSAT has created a system that runs all collected data through various pre-processing and cleaning stages, limits access to data for research purposes only, and securely stores data for the lifetime of its use.