Research

When we know the problem, learning analytics can help us solve it

APOA-hanke, Annika Latva ja Hanna Teräs
Oppimisanalytiikan käyttöön tarvitaan strategia, joka huomioi sekä pedagogiset että eettiset näkökulmat, painottavat APOA-hankkeen projektipäälliköt Annika Latva ja Hanna Teräs.
Digitalisation enables collection of large amount of data and data analytics its utilisation in almost all sectors of the society, including learning. As digital learning environments become more and more common, a lot of data accumulates on students, learning and learning environments. Learning analytics uses the data to develop learning and teaching.

“Data will accumulate but data as such is not important. The point is what is done with it. We first have to ask what or whose is the problem which we try to solve with learning analytics”, says TAMK’s Principal Lecturer Hanna Teräs, who has developed digital learning environments for 20 years. 

Hanna Teräs leads an ongoing APOA project of one university and ten universities of applied sciences to develop learning analytics. The project started in 2018 and is coordinated by TAMK. It focuses on using analytics in development of pedagogy and learning. The project first implemented an extensive need mapping.

“We interviewed almost 300 students and staff members from the participating universities of applied sciences in small groups of 4-5 persons”, tells Project Manager Annika Latva, who is in charge of the project administration.  

“We wanted to avoid data-based thinking in which thinking is only directed by data and its collection possibilities. Instead we first wanted to establish the actual pains students and teachers face in the studies. When we know the problem, we can think how to solve it by means of learning analytics”, Teräs explains.  

Students do not need extras but good-quality pedagogy

Teachers most of all worried about how to know if learning goes well during a study unit. Students above all hoped for pedagogical quality – regardless of whether the studies take place in a classroom or in a digital learning environment.   

“Students do not need digital extras. They want well-planned study units, clear instructions, feedback and good-quality and easily available material. In other words, basic pedagogy”, Teräs emphasises.  

If students have difficulties, learning analytics applications can show the problem easier without the teacher having to dig the depths of the learning environment data. Larger groups’ data can also be made more visible.  

The learning system can also send an automatic message to the student. Interviews made in the APOA project confirmed our conception of students having a positive attitude to such automatic interventions if they are positive and supporting.  

”The system can send caring messages, such as ’Hello, how are you?’ There is never too much caring”, Teräs and Latva state in unison.  

Legal is not always ethical

Data protection and ethical issues also raise many questions in use of learning analytics.    

“GDPR and other legislation set some boundaries for using data but we have to remember that what is legal is not always ethical”, Teräs emphasises.  

Important ethical issues include who has access to the data and if the data is conveyed to other parties. Teräs emphasises the importance of transparency in data use and raises new challenges related to machine learning and artificial intelligence. 

“We can use learning analytics to create a model which predicts dropouts with the probability of for example 90 percent. But what do we do with it? Do we stop investing in some students? Or do we create an automatic intervention which aims at keeping students in their studies?”  

The error margins of the models are also alarming.  

“90-percent probability is never enough. It would mean that 10 percent of our students and more than 3,000 students in our universities community would receive a wrong diagnosis”, Teräs says.  

Pilots form the basis for national recommendations  

The problems raised by students and teachers are being solved in the APOA project by means of several pilots. They go into development of pedagogy, promotion and demonstration of students’ learning and identification of problems in integrating contact and online teaching.   

The APOA project aims at creating national recommendations on using learning analytics in universities of applied sciences. The pilot results will be revealed in the APOA blog in the spring. Thereafter the national recommendations will be published. 

“It is important for us to implement the project results and recommendations in Finnish universities of applied sciences. It is not enough that they are just listed on the project website”, emphasise the project managers.  

It is worth keeping your ears open.  

  

APOA is a leading project of the Finnish Ministry of Education and Culture. It pilots and studies use of learning analytics in universities of applied sciences. The project is coordinated by TAMK and the project participants are comprised of ten universities of applied sciences and the University of Turku.  

 

Further information:  

Hanna Teräs, Principal Lecturer, Pedagogical Innovations, Professional Teacher Education, hanna.teras [at] tuni.fi ()  

Annika Latva, Project Specialist, RDI Services, annika.latva [at] tuni.fi ()  

  

  

Text: Helena Pekkarinen 

Photo: Jaakko Saarilampi