School Facebook posts could threaten student privacy
Joshua Rosenberg, University of Tennessee
Like many of us, schools in the United States are active on social media. They use their accounts to share timely information, build community, and highlight staff and students. However, our research has shown that school social media activity can interfere with student privacy.
As a researcher specializing in data science in education, my colleagues and I have inadvertently brought up the topic of student privacy. We were exploring how schools were using social media during the early days of the COVID-19 pandemic, particularly in March and April 2020. During this research, we noticed something surprising about how Facebook worked: We could view school posts – including pictures of teachers and students, even when not logged in to our personal Facebook accounts.
The ability to access the pages and images even when we were not logged in revealed that not only can anyone access the school posts, but they can also be systematically viewed using methods of communication. ” exploration of data or new research methods involving the use of computers and statistical techniques. to discover patterns in large datasets – often publicly available.
Since virtually all schools in the United States report their websites to the National Center for Education Statistics, and many schools link to their Facebook pages from their websites, these posts can be viewed comprehensively. In other words, not only researchers but also advertisers and hackers could use data mining methods to access all posts from any school with a Facebook account. This full access has allowed us to study phenomena such as violations of student privacy on a massive scale.
The risks are present
The easy access to photos of students we met comes despite broader concerns about individual privacy on social media. Parents, for example, have expressed concerns about teachers posting about their children on social media.
Fortunately, our research of media coverage and academic publications did not reveal any harm suffered by students because their schools published articles about them. However, there are a number of potential risks that identifiable student positions could pose. For example, potential bullies and bullies might use the messages to identify individual students.
In addition, there are new threats that students may face. For example, facial recognition company Clearview collects internet data – and social media data – from the World Wide Web. Clearview then sells access to that data to law enforcement agencies, who can upload photos of a potential suspect or person of interest to display a list of potential names of the person depicted on. the uploaded photo. Clearview is already accessing identifiable photos of minors in the United States from public posts on Facebook. Student photos from school Facebook pages may be viewed and used by companies such as Clearview.
Even if we are not aware that these things are actually happening, there is no reason not to be concerned. In times when our privacy is often surprisingly threatened, as tech journalist Kara Swisher writes, “only the paranoid survive.” My fellow researchers and I believe that this cautious view – even a paranoid one – is especially justified when it comes to students as minors who may not give their explicit permission to be included. in messages.
Millions of student photos available
In our study, we used federal data and an analytics tool provided by Facebook to access posts from schools and school districts. We use the term “schools” to refer to both schools and school districts in our study. From this collection of 17.9 million posts published by approximately 16,000 schools from 2005 to 2020, we randomly selected – sampled – 100 and then coded these publicly available posts. We determined whether the students were named in the post with their first and last names, and whether their faces were clearly shown in a photo. If these two elements were present, we considered that a student was identified by name and school.
For example, a student in a Facebook post whose photo includes a name in the caption, such as Jane Doe, would be considered identified.
We determined that 9.3 million of the 17.9 million messages we analyzed contained images. Of these 9.3 million messages, we estimated that approximately 467,000 students were identified. In other words, we found nearly half a million students on publicly accessible schools Facebook pages who are photographed and identified by their first and last names and the location of their school.
Assess the risks
While many of us already post photos of ourselves, our friends and family – and sometimes our children – on social media, school posts are different in an important sense. As individuals, we can control who can see our posts. If we want to limit it to only friends and family, we can change our own privacy settings. But people don’t necessarily control how schools share their posts and images, and all of the posts we analyzed were strictly accessible to the public. Anyone in the world can access it.
Even considering the potential damage from this situation to be minimal, schools can take small steps that could make a noticeable difference in whether or not this potential is present:
1. Refrain from publishing the full names of students
Not displaying the full names of students would make it much more difficult for companies to target individual students and sell student data and link it to other data sources.
2. Make school pages private
Making school pages private means that data mining approaches similar to ours would be much more difficult, if not impossible, to implement. This single step would greatly reduce the risk to student privacy.
3. Use press release activation policies
Press release acceptance policies require that parents explicitly agree to their child’s photos being shared through communication platforms and the media. These can be more informative for parents – especially if they mention that communication and media platforms include social media – and more protective of student privacy than opt-out policies, which require parents to contact. their child’s school if they don’t want their child’s photo or information to be shared.
In summary, school Facebook pages are different from our personal social media accounts, and posts on these pages can threaten student privacy. But using social media doesn’t have to be a proposition for schools. That is, it doesn’t necessarily come down to a choice between using social media without considering privacy threats or not using social media at all. On the contrary, our research suggests that educators can and should take small steps to protect student privacy when posting from school accounts.
[Understand new developments in science, health and technology, each week. Subscribe to The Conversation’s science newsletter.]
Joshua Rosenberg, Assistant Professor of STEM Education, University of Tennessee
This article is republished from The Conversation under a Creative Commons license. Read the original article.