Laird: Remote Proctoring of Exams Is an Invasive Tool Without Clear Security Protections. States & Districts Should Avoid Rushing In
- Laird: Remote proctoring of exams is an invasive tool without clear security protections. States and districts should avoid rushing in @CenDemTech
- Laird: There is still time to avoid yet another negative and unintended consequence of the pandemic in which we trade students’ privacy to administer assessments, for which they will pay for years to come @CenDemTech
Get essential education news and commentary delivered straight to your inbox. Sign up here for The 74’s daily newsletter.
The education sector knows all too well the harm that can occur when assessments aren’t given securely. There’s the Atlanta Public Schools cheating scandal that resulted in the indictment of 34 teachers and principals, along with the district superintendent, and the eventual racketeering conviction of 11 teachers. Inadequate testing security made even bigger news when the FBI found celebrities and other wealthy families buying their children better scores on college readiness exams.
When the U.S. Department of Education announced its waiver process for statewide assessments and accountability in light of the ongoing effects of the pandemic, it allowed state exams to be administered “remotely, where feasible.”
Given the importance of test security, what guidance or practices are available for the remote administration of standardized assessments?
Without any guidance, school districts and states are considering remote proctoring software. In fact, two companies that support the administration of statewide assessments, Pearson and McGraw-Hill, already have partnerships with companies that offer remote proctoring software.
From my own experience as a test security auditor in Washington, D.C., I know that maintaining trust in assessment results is hard, nuanced and resource-intensive. Remote proctoring software is attempting to take the place of years of honed practices, experienced proctors and testing experts. Unfortunately, they are a bad idea for students because they undermine privacy and can actually harm students.
Remote proctoring tools are different from the quiz or exam administration practices that have been happening in K-12 schools during remote learning. During the past school year, teachers have observed students just as they would if they took a test in a classroom, using no additional technology beyond video conferencing software, like Zoom.
Conversely, remote proctoring software is a tech-based solution that excludes teachers and uses either computers or strangers to track what students are doing. There are two main types of remote proctoring software:
- AI-based: Certain companies sell software programs that utilize artificial intelligence capabilities like facial recognition and motion detection, and even attempt to track where students’ eyes are looking to identify cheating or other violations of test security protocols.
- People-based: Some companies attempt to replace in-person proctors with those who do the job remotely. These adults — who, unlike a teacher, are unknown to the students — will watch a test be taken in real time or require that students show the room in which they are testing to ensure there is no material that could help them. The difference between this and in-person proctoring is that many students are not taking the test in a classroom, so they have to share their home environment with a stranger. This is invasive and raises privacy concerns.
These remote proctoring tools first rose to infamy in higher education circles, when universities began using them during the pandemic and students experienced their limitations first-hand. Some software couldn’t identify test-takers’ faces because of their skin color; children with disabilities were flagged for anomalous movements while taking exams. As a result, several universities stopped using remote proctoring software.
This software is now poised to enter K-12 communities to fill the demand for state testing. In addition to the general creepiness factor, there are several concerns:
- There is no independently verified evidence that remote proctoring software works; however, there is evidence that it discriminates against students who might fall outside the norm, like students with disabilities and children of color. Moreover, during an assessment, the remote proctoring tool can itself become a barrier to completing the test if it malfunctions and increases students’ stress, making for poorer performance.
- Remote proctoring tools, by their nature, collect sensitive information about a student’s home environment (some even require a student ID or full administrative access to a student’s computer). In some cases, such as with children experiencing homelessness, this may include information that students do not want to disclose. Given that this information is collected by third parties, it raises important questions about how that information will be stored, who will have access to it and what steps will be taken to ensure it will not be breached.
Although remote proctoring tools may (arguably) meet the minimum criteria for legal compliance, they do not build trust or promote responsible data use. As a result, their use will undermine the confidence that schools have earned thus far with student data.
In our research of parent and teacher views toward privacy, we found that both trust schools with protecting sensitive student information: 72 percent of parents trust their child’s school with the information being collected about their child, and 89 percent of teachers trust that their school will not inappropriately use information collected about its students.
In looking ahead to the testing that will not only happen this school year but in the years ahead, it is important to consider these risks.
To avoid normalizing an invasive tool that has the potential to directly harm students, the education field needs guidance and best practices on how to administer exams remotely, ideally from the U.S. Department of Education, as it has provided such guidelines in the past for in-person testing. Additionally, families and communities should be consulted before proceeding with this unproven technology to maintain trust. And lastly, schools should consider the needs of all students, and how technology negatively affect them, when making decisions about adopting new software.
Many states are considering these issues now as they formulate waivers that will chart the path for remote testing. There is still time to avoid yet another negative and unintended consequence of the pandemic in which we trade students’ privacy to administer assessments, for which they will pay for years to come.Submit a Letter to the Editor