Viktor has been working with the the subject of privacy in learning apps of Danish primary schools. In the following writings, Viktor reflects on the challenges of running a research project, where stakeholders interests and would-be simple practicalities are sometimes the biggest obstructions to be faced by a researcher.

Privacy in Primary Schools – Takeaways from a Brief Collaboration with the “LoL” Project

Being a Global Business Informatics student with an interest in data science, digital methods, privacy, and security, I had been participating in public ETHOS events as a volunteer at the Lab for an extended period. I gained an interest in a research-oriented career very early on in my education, and have taken every opportunity throughout my studies at ITU to bring me closer to this goal. During my 5th semester, I decided that becoming more actively involved in the Lab and taking part in a 7,5 ECTS elective research project under its guise would be a very valuable opportunity in my aforementioned pursuits. In this essay, I will summarize the findings of the research project I embarked on, and reflect on the challenges I encountered, as well as on my learning and further perspectives.

A relevant case emerged shortly after I approached the management of the Lab, which appeared to be connected to several of my interests, including privacy aspects of data use, education, and mobile apps, and seemed to allow for an application of my experience with ethnographic fieldwork and (computer) ethics. This case was the “Leg og Læring” or “LoL” App, an iPad application currently under development for use in Danish primary schools. The app is intended to aid teachers and pedagogues in fulfilling the Danish School Law’s requirement for incorporating physical activity in teaching, primarily by providing a list of activities and exercises that can be performed with no preparation needed. In the future, the developers plan to extend the applications and functionality of the app, e.g. by recording activities and providing ways of sharing information with parents.

Based on this premise, I was especially eager to focus on the privacy aspects of the (planned) use of data in this school context. The team behind the app had contacted ETHOS Lab previously, and were open to collaborating with me on a 7.5 ECTS elective project. I set out to investigate the case by applying qualitative methods like surveys, interviews and/or participant observation to either explore the thoughts and feelings of the stakeholders involved in the development and use of the app, or map decision making in the development process from an ethical perspective.

Findings

While the members of the “LoL” project were relatively quick to accept my proposal for collaboration, and I also had access to some data from their prototyping stage, I routinely encountered difficulties in terms of their availability when communicating with them (most of which was through short messages), and especially when I tried to schedule face-to-face meetings. The core team of the project consists of two pedagogues, who have been inspired by their Albertslund colleagues’ challenges, as well as their own ones in fulfilling the nation-wide requirements for involving physical activities as part of teaching. As mentioned previously, they seek to offer a solution to this issue by assembling a collection of physical activities for pedagogues and teachers that would tie in with their respective subject areas, and require no additional preparation by staff. This would be made accessible as an iPad app developed by an external software engineer/consultant, which would first be tested and implemented at the school where the pedagogues heading the project work. The first version of the app, planned for release in March, would provide teachers and pedagogues with guidelines and access to online, printable assets for exercises specific to the age of students – the current focus is on kindergarten to 3rd grade, or 5 to 9 years of age – and the subject being taught (as well as a few generally applicable games or relaxing activities).

In the future, however, they not only plan to increase the targeted range of users in terms of the age of students being addressed and deployment in additional schools (this, of course, would be affected by factors such as responses from initial users and financing), but also to add several new features and functions to the app in forthcoming versions, all of which would capture or use data in various ways: these include introducing accounts for the teachers, students, and parents (through UNI-Login), sharing the students’ schedules with parents, as well as what activities they did, and how much physical activity they performed. They also want to increase the extent to which the exercises are mediated by the app itself (through installing and using it on the students’ devices), track the movement of students (when they do exercises with their iPads in hand), potentially record data about the students’ performance, get ratings or other feedback about the exercises, and provide a channel for teachers to communicate with parents about the behavior and well-being of individual students. Extending the functionality of the app in such ways would not only introduce technical and infrastructural challenges, but ethical implications regarding the handling of potentially sensitive student data.

Perhaps owing to the fact that these features were in quite early stages of ideation or planning (even in comparison to other aspects of the app, with no definitive timeline of implementation), these implications had not been explored. While the pedagogues were aware of the sensitivity of information about the wellbeing of individual students, they had not considered the ethical perspectives of implementing these prospective features, the nature of the data produced by them, various privacy and security risks involved, or approaches to mitigating these. Moreover, they expressed a lack of awareness in terms of the specific policies that govern the use of data in this educational context, and the authorities involved in the making and enforcing of these policies.

Challenges, Risks and Possible Mitigation

The main challenges of my research project had so far been the availability and responsiveness of the project members, and adjusting the scope of my project to the fact that all of the data-intensive features would only be implemented in future versions of the app, and with no clear timetable. In light of this, as well as the still somewhat unclear motivation for the project members’ engagement in my inquiry, and their scheduled beta testing in December, I approached them about the possibility of observing the testing process. I was curious whether it would be valuable for them if I would trace and document the upcoming period of deployment and development from multiple perspectives using ethnographic fieldwork, focusing on the expectations and reactions of the teachers and kids testing the app, and how this could be reflected in the development of the project, especially in terms of their iterations of the planned data-intensive features. While they first responded with uncertainty about how they could benefit from this, they rejected my proposal immediately afterwards, following a meeting with the external developer. They argued for postponing the collaboration, citing a lack of progress in development, and later stated that they will only be interested in analysis of their as-of-yet hypothetical data with digital methods, which effectively ended our communication.

The thorough understanding of the disparate, sometimes conflicting interests of stakeholders in a collaborative partnership can often be critical to maintaining such a relationship. While keeping an eye on such matters can undoubtedly help us succeed in our academic collaborations, analyzing the possible reasons leading to a failure can provide equally valuable insights for future projects. My work with “LoL” has been characterized by some asymmetry in access to information, motivations, and engagement from the start, circumstances that often affect the work of researchers to varying degrees. Based on the information available to me, perhaps the most significant risk factor in this particular collaboration stemmed from the nature of the case organization. Startups, due to forces both internal and external, are liable to change, grow, or simply change direction rapidly, which can easily result in shifting balances in outside partnerships. The environment that an education startup is situated in may also hold surprises, e.g. due to the different conditions for success in a politicized, largely publicly funded market. The relatively small number of personnel in startups can be an enabler for rapid changes, and it can also introduce complications of its own in extreme cases. When only 2-3 people have active roles in a project, they may simply be too busy to deal with ancillary matters, and the introduction of an additional, external observer could feel like a violation of a relatively intimate space.

Although this was not prioritized at the time due to the slow progress in communication and the case organization’s apparent interest in working with ETHOS Lab, in hindsight, I believe that pushing to create a collaboration contract as early as possible could have been the best approach to minimize the risks involved in this case. Such a document could not only be a valuable token when negotiating access and other practical matters of a collaboration, but it could serve an important purpose in establishing the various stakeholders’ expectations regarding the partnership, as well as whether the conditions for a viable collaboration are actually present.

Concluding Remarks and Further Perspectives

Along with playing an active role in other activities at the Lab, I have found my participation in this research project tremendously helpful. Experiencing the possible challenges and failures that can be encountered when conducting collaborative projects first-hand can be invaluable for learning how to manage such projects successfully. Additionally, I have gathered a lot of impulses and inspiration, and synthesized a clearer, more detailed picture about how a project aligned with both my own interests and those of the Lab could be designed and carried out in practice, and I hope to apply these insights very soon.

My main takeaways from the topics this project explored and the conversations it sparked would be the trends present in education and primary schools: a great amount of data collection and usage is happening in this context right now, and this will only increase in the years to come. Yet these particular areas have been largely left unexplored, and awareness of the implications of these practices with regards to privacy and security is often lacking among key stakeholders – including both drivers and subjects, such as developers, pedagogues, teachers, parents, and last but not least the students, who are probably most affected, but have the least control. By introducing some critical questions about these matters of data usage and privacy during the development of an app project that is part of this phenomenon, I hope to have made a contribution, however small, to raising awareness about these increasingly significant themes.

/Viktor Hargitai