The TOHTOS project, supported by the European Social Fund, is about developing the working life relevance of doctoral training. As part of the project actions, SLAM project’s Ph.D. candidate Héctor J. Pijeira-Díaz had the opportunity to briefly introduce himself on a short video. You can watch the video here. Héctor’s public doctoral defense is scheduled on May 3rd at noon at the L10 lecture hall of the University of Oulu. The defense is to be streamed live (link and more details to be informed closer to the D-day).
The data collection at Ritaharju before Christmas produced a huge amount of data. We have now 48 gigabytes of physiological data, different types of self-reports and over 200 hundred hours of video from students collaborative working to play with.
If you think about the amount of data we now have in our hands, you probably understand that we want to make sure we will have a solid and shared understanding on regulated learning in collaboration. That is why we engaged in a very stimulating discussion in which we built our conceptual clarity on what we mean with time, granularity and contextuality in relation to studying socially shared regulation from process data. As you can imagine, this required a lot of coffee and sweets. Now, we are conceptually ready to get our hands into the amazing data we have collected.
Our intensive data asks for organized and systematic approach. Therefore, we are setting up a MySQL database to simplify the synchronization, preprocessing and filtering of the data. For this we were happy to get a new member in our team to help us. Our new intern Andy has presented his progress on developing a database for storing and query-based retrieval of the project data.
In addition, we are looking at possibilities to use AI to analyse the huge amount of video data we have. For example, how do we recognise the features of group interactions relevant for exploring regulated learning processes. Stay tuned for exciting news! :-)
As a final note, a lot more candies and many many cups of coffee to be consumed in the following months :D
Tenure Track Positions – Co-evolution of Human Capabilities and Intelligent Technologies at the University of Oulu
Do you want to contribute to ambitious research on how human skills and capabilities can be strengthened in the emerging digital age? Are you expert in learning and interaction processes, multimodal big and thick data?
Tenure track positions open:
See more: www.oulu.fi/university/genz
Our research team is happy to finish our big data collection in Ritaharju secondary school in Oulu, Finland. We have had excellent collaboration with teachers planning, designing and carrying out collaborative science tasks in a physics topic of light and sound. For the first time we have been able to follow more than 100 students working in small groups for several months as a part of their real school work. This has given us an opportunity for long-term temporal multimodal data collection. A special case is that we work in authentic school context capturing all the intervening factors present in real school life.
Yes, we know this will be challenging when we start analyzing the data - it will be messy and complex, but interesting for sure! It will be definitely worth it though. We genuinely believe that the data will help us to take next steps in our understanding of socially shared regulation in collaborative learning. Thank you for the great collaboration teachers’ and students’ in Ritaharju school! Thank you also our great research team! You all deserve a warm thank you for the huge efforts!
We are eagerly waiting for New Year and to start working with the data. Before that it’s time to relax. Happy holidays and Merry Christmas from the LET team!
The academic year in LET team typically starts with conferences. This year, LET team members participated for two conferences, namely Motivation SIG which was held in Denmark and Metacognition SIG that was held in Zurich, which is a beautiful capital of Switzerland.
Metacognition SIG is not the biggest SIG in EARLI, and this year, there was 114 participants, who attended to the conference. From SLAM team, there was 6 participants, namely Jonna Malmberg, Marta Sobocinski, Sara Ahola, Eetu Haataja, Muhterem Dindar and Sanna Järvelä who attended. Each of them had a presentation about their studies, which all dealt with thematic areas such as Metacognition, Adaptation in regulation, and also combining multiple multiple data modalities to better understand how different data modalities such as electrodermal activity (EDA) and heart rate (HR) could be used to better understand invisible learning processes that are difficult to capture otherwise.
E-CIR network presented a panel discussion that focused on complexity of multichannel trace data for learning scientists and learning analytics, namely how to use the data, how to hypothesize trajectories in individual and social regulation of learning. The challenge is not only to tackle with multiple modalities of the data, but also to understand how these data modalities can reflect motivation, emotion, cognition and metacognition. Roger Azevedo noted that “We need a whole village for this type of research”.
Sanna also participated and gave a keynote at the Nordic Learning Analytics Summer Institute on the 29. and 30. of August 2018, in Copenhagen (Aalborg University’s Campus). LASI-Nordic 2018 is a part of the International Society for Learning Analytics Research (SOLAR) series of local learning analytics summer institutes. LASI was a great opportunity to discuss and share our progress in the SLAM.
Conferences offer a good kick start for academic year, since they often stimulate your thinking processes, but there is also possibility to get understanding of what are the questions that researchers from the same field are dialing and working with currently.
It is almost gone, but the summer of 2018 has been the hottest in the last decade in many spots of the Northern hemisphere. In the northern Finnish city of Oulu, not only summer left scorching days, but it was also “hot” with research. Three pieces of SLAM research were published in two high-quality outlets: the Journal of Computer Assisted Learning and Computers in Human Behavior.
In May, “Profiling sympathetic arousal in a physics course: How active are students?”, a study of arousal in the classroom as a concomitant of cognitive-affective processes, and its relation to academic achievement, was accepted for publication in the Special Issue on Multimodal Learning Analytics of the Journal of Computer Assisted Learning. The paper, with “implications for a need to focus more on addressing low arousal states in learning” and “potential applications for biofeedback, teacher intervention, and instructional design”, found that “low arousal was the level with the highest incidence (60% of the lesson on average) and longest persistence, lasting on average three times longer than medium arousal and two times longer than high arousal level occurrences. During the course exam, arousal was positively and highly correlated (r = .66) with achievement as measured by the students' grades.”
In early June, “Monitoring in collaborative learning: Co-occurrence of observed behavior and physiological synchrony explored”, a study of “how students in a group monitor their cognitive, affective and behavioral processes during their collaboration, as well as how observed monitoring co-occurs with their physiological synchrony during the collaborative learning session”, became available online in Computers in Human Behavior. The paper reports that the main targets of monitoring for the case groups analyzed were cognition and behavior, while monitoring of affect occurred the least. “Most of the student pairs inside the groups showed significant amounts of physiological synchrony” and “high values of physiological synchrony occurred when monitoring was frequent”. In this direction, “physiological synchrony could potentially shine a light on the joint regulation processes of collaborative learning groups”.
Later in June, “Going beyond what is visible: What multichannel data can reveal about interaction in the context of collaborative learning?”, a study utilizing “multichannel data, namely physiological data, video observations, and facial recognition data, to explore what they can reveal about types of interaction and regulation of learning during different phases of collaborative learning progress”, was also published online in Computers in Human Behavior. “The results show that simultaneous arousal episodes occurred throughout phases of collaborative learning and the learners presented the most negative facial expressions during the simultaneous arousal episodes. Most of the collaborative interaction during simultaneous arousal was low-level, and regulated learning was not observable.” The study “represents an advance in testing new methods for the objective measurement of social interaction and regulated learning in collaborative contexts.”
Other SLAM manuscripts are under review or close to submission, which will continue to report on the exploration of novel data sources such as physiological data, in the quest of increasing our understanding of learning, learning regulation, and collaborative learning processes.
We are planning a new longitudinal study to start in August 2018 in secondary school. We will join our forces and led by three PIs Sanna Järvelä, Hanna Järvenoja and Jonna Malmberg together with a great group of post-docs and PhD students we aim to tackle how to make invisible complex cognitive, motivational, and emotional learning processes visible for people to help and support them in regulating themselves to learn and function more effectively. Our international collaboration with Professors Paul Kirschner, Allyson Hadwin and Roger Azevedo and their research teams continues. More exciting news to report soon.
I have spent the whole February visiting professor Roger Azevedo and his team at North Carolina State University. Roger Azevedo is a professor at the Department of Psychology, and he and his team are studying cognition, metacognition, motivation and affect in the context of self-regulated learning in computer-based learning environments. The SMART lab and LET share many similarities in the focus of research, and there has been collaboration between the two teams for years now, as both teams aim to understand learning processes using advanced technologies (e.g. virtual learning environments, physiological sensors).
The focus of my PhD studies is related to exploring patterns of regulation in collaborative learning situations and in my recent work I have been struggling with integrating heart rate measures collected during collaborative learning with video data coding, so it seemed like the perfect destination to deepen my understanding of using multimodal methods in learning research. The visit was also an exciting opportunity to gain an insight into the daily life at SMART lab.
During my stay there I had the chance to discuss best practices and current problems in analyzing multimodal data with the SMART team members (who were indeed smart!), and I also had a chance to present my work and get valuable feedback.
At the time of my visit there were several on-going data collections. One was related to Metatutor IVH, which uses an Intelligent Virtual Human in order to help students determine whether or not the content they are reading is useful for them or not. Another data collection was in Cary Acadamy, a local private school, where high school students were asked to complete a gamified STEM task in virtual reality developed by Lucid dream. It was really interesting to see how intuitively the students oriented themselves in the environment and how excited both the teachers and students were (not to mention the researchers). For more information about ongoing projects at SMARTLAB see here: https://psychology.chass.ncsu.edu/smartlab/#projects
As VR equipment is becoming more readily available and affordable, it is a very topical question how we can use it to better understand learning. One of the unique aspects about collecting data in a vr environment is that it is possible to easily collect data about where the users looked, what they “touched” where they turned, providing an insight into what the learner does, and how do the strategies used change over time. I also started wondering how it would be possible to set up a collaborative VR environment.
Besides visiting the smart lab, I also had the chance to meet Jeff Greene and his research team at Chapel Hill, where I could present my research and engage in interesting theoretical discussions. Thank you for the warm welcome!
I am very grateful to Roger and the whole team for making me feel part of the SMART family and for this learning opportunity! I am looking forward to possible future collaborations with the SMART team.
Marta Sobocinski, doctoral student at LET
Jonna Malmberg, a Post-Doctoral researcher from the SLAM project is visiting at the School of Informatics, University of Edinburgh, Scotland. The research visit will strengthen already existing networks with Prof. Dragan Gasevic and his research team.
Dragan Gasevic is a Professor and the Chair in Learning Analytics and Informatics in the Schools of Education and Informatics at the University of Edinburgh since February 2015.
Since SLAM research project builds in for interdisciplinary research, Dragan Gasevic and his team can provide and add valuable methodological knowledge in terms of how to process multichannel data that consists of physiological reactions and their associations for self-regulated learning. Specifically, they focus of investigating how machine learning methods can self-regulated in the context of collaboration.
During the research visit, Jonna has been working intensely with Postgraduate Student Oliver Fincham, who is a member of Institute for Adaptive and Neural Computation and Markus Hörmann. He is a research fellow at the chair of Teaching and Learning with Digital Media at the University of Technology Munich and works closely with Prof Maria Bannert.
The research visit at the beautiful and old city of Edinburgh provides Jonna a great opportunity to learn about state of the art research projects dealing with, for example Natural Language Processing (NLP) and machine-learning methods that can definitely be useful and valuable for the progress of the SLAM research project.
The SLAM research will continue!
Oulu University EUDAIMONIA strategic funding was given to our research proposal for the years 2018-2021:
Making Complex Learning processes Visible for Enabling Regulation:
Change human behavior for learning success (CLEVER change)
The PI is Sanna Järvelä and co-PI Paul Kirschner, and the SLAM team will contribute.
Our collaborators are Ass. Prof. Allyson Hadwin, University of Victoria Canada and Prof. Roger Azevedo from North Carolina State University.
"SLAM project harnesses advanced technologies to enhance strategic regulation of individual and collaborative learning"