Collaboration is one of the four important 21st-century skills. With the pervasive use of sensors, interest on co-located collaboration (CC) has increased lately. Most related literature used the audio modality to detect indicators of collaboration (such as total speaking time and turn taking). CC takes place in physical spaces where group members share their social (i.e., non-verbal audio indicators like speaking time, gestures) and epistemic space (i.e., verbal audio indicators like the content of the conversation). Past literature has mostly focused on the social space to detect the quality of collaboration. In this study, we focus on both social and epistemic space with an emphasis on the epistemic space to understand different evolving collaboration patterns and collaborative convergence and quantify collaboration quality. We conduct field trials by collecting audio recordings in 14 different sessions in a university setting while the university staff and students collaborate over playing a board game to design a learning activity. This collaboration task consists of different phases with each collaborating member having been assigned a pre-fixed role. We analyze the collected group speech data to do role-based profiling and visualize it with the help of a dashboard.

Collaboration is an important 21st Century skill. Co-located (or face-to-face) collaboration (CC) analytics gained momentum with the advent of sensor technology. Most of these works have used the audio modality to detect the quality of CC. The CC quality can be detected from simple indicators of collaboration such as total speaking time or complex indicators like synchrony in the rise and fall of the average pitch. Most studies in the past focused on “how group members talk” (i.e., spectral, temporal features of audio like pitch) and not “what they talk”. The “what” of the conversations is more overt contrary to the “how” of the conversations. Very few studies studied “what” group members talk about, and these studies were lab based showing a representative overview of specific words as topic clusters instead of analysing the richness of the content of the conversations by understanding the linkage between these words. To overcome this, we made a starting step in this technical paper based on field trials to prototype a tool to move towards automatic collaboration analytics. We designed a technical setup to collect, process and visualize audio data automatically. The data collection took place while a board game was played among the university staff with pre-assigned roles to create awareness of the connection between learning analytics and learning design. We not only did a word-level analysis of the conversations, but also analysed the richness of these conversations by visualizing the strength of the linkage between these words and phrases interactively. In this visualization, we used a network graph to visualize turn taking exchange between different roles along with the word-level and phrase-level analysis. We also used centrality measures to understand the network graph further based on how much words have hold over the network of words and how influential are certain words. Finally, we found that this approach had certain limitations in terms of automation in speaker diarization (i.e., who spoke when) and text data pre-processing. Therefore, we concluded that even though the technical setup was partially automated, it is a way forward to understand the richness of the conversations between different roles and makes a significant step towards automatic collaboration analytics.

Collaboration is one of the important 21st-century skills. It can take place in remote or co-located settings. Co-located collaboration (CC) is a very complex process that involves subtle human interactions that can be described with indicators like eye gaze, speaking time, pitch, and social skills from different modalities. With the advent of sensors, multimodal learning analytics has gained momentum to detect CC quality. Indicators (or low-level events) can be used to detect CC quality with the help of measurable markers (i.e., indexes composed of one or more indicators) which give the high-level collaboration process definition. However, this understanding is incomplete without considering the scenarios (such as problem solving or meetings) of CC. The scenario of CC affects the set of indicators considered: For instance, in collaborative programming, grabbing the mouse from the partner is an indicator of collaboration; whereas in collaborative meetings, eye gaze, and audio level are indicators of collaboration. This can be a result of the differing goals and fundamental parameters (such as group behavior, interaction, or composition) in each scenario. In this article, we present our work on profiles of indicators on the basis of a scenario-driven prioritization, the parameters in different CC scenarios are mapped onto the indicators and the available indexes. This defines the conceptual model to support the design of a CC quality detection and prediction system.

This paper presents results from our design and evaluation studies of the Learning Analytics Cockpit (LA Cockpit) for a quiz app, which aims to provide lecturers with important information about students’ knowledge levels. We define a LA Cockpit as a tool for instructors that enables them to steer students’ learning process by providing a LA Dashboard which visualizes students’ learning indicators and an intervention feature enabling instructors to give feedback based on students’ knowledge levels. To address the needs of lecturers we applied the Double Diamond (DD) design process model which consists of four stages: discover, define, develop & refine. Following the DD process, we first conducted a qualitative study by interviewing four lecturers and student teachers to discover their needs. Results from the interviews allowed us to define requirements of the lecturers. We used these results to develop the first version of the tool where we refined it through informal feedback by the interviewed teachers. In preparation for a larger effectiveness-study, we evaluated the LA Cockpit in terms of usefulness and usability in a preliminary study with 16 university lecturers. Results from this qualitative study indicate that the LA Cockpit can measure the students’ knowledge level and supports self-reflection for lecturers. Moreover, results show that the LA Cockpit enables lecturers to address knowledge gaps and provide interventions to students before the exams.

Learning analytics dashboards (LADs) are designed as feedback tools for learners, but until recently, learners rarely have had a say in how LADs are designed and what information they receive through LADs. To overcome this shortcoming, we have developed a customisable LAD for Coursera MOOCs on which learners can set goals and choose indicators to monitor. Following a mixed-methods approach, we analyse 401 learners' indicator selection behaviour in order to understand the decisions they make on the LAD and whether learner goals and self-regulated learning skills influence these decisions. We found that learners overwhelmingly chose indicators about completed activities. Goals are not associated with indicator selection behaviour, while help-seeking skills predict learners' choice of monitoring their engagement in discussions and time management skills predict learners' interest in procrastination indicators. The findings have implications for our understanding of learners' use of LADs and their design.

Unequal stakeholder engagement is a common pitfall of adoption approaches of learning analytics in higher education leading to lower buy-in and flawed tools that fail to meet the needs of their target groups. With each design decision, we make assumptions on how learners will make sense of the visualisations, but we know very little about how students make sense of dashboard and which aspects influence their sense-making. We investigated how learner goals and self-regulated learning (SRL) skills influence dashboard sense-making following a mixed-methods research methodology: a qualitative pre-study followed-up with an extensive quantitative study with 247 university students. We uncovered three latent variables for sense-making: transparency of design, reference frames and support for action. SRL skills are predictors for how relevant students find these constructs. Learner goals have a significant effect only on the perceived relevance of reference frames. Knowing which factors influence students' sense-making will lead to more inclusive and flexible designs that will cater to the needs of both novice and expert learners.

We developed the CPR Tutor, a real-time multimodal feedback system for cardiopulmonary resuscitation (CPR) training. The CPR Tutor detects mistakes using recurrent neural networks for real-time time-series classification. From a multimodal data stream consisting of kinematic and electromyographic data, the CPR Tutor system automatically detects the chest compressions, which are then classified and assessed according to five performance indicators. Based on this assessment, the CPR Tutor provides audio feedback to correct the most critical mistakes and improve the CPR performance. To test the validity of the CPR Tutor, we first collected the data corpus from 10 experts used for model training. Hence, to test the impact of the feedback functionality, we ran a user study involving 10 participants. The CPR Tutor pushes forward the current state of the art of real-time multimodal tutors by providing: 1) an architecture design, 2) a methodological approach to design multimodal feedback and 3) a field study on real-time feedback for CPR training.

This paper introduces the Visual Inspection Tool (VIT) which supports researchers in the annotation of multimodal data as well as the processing and exploitation for learning purposes. While most of the existing Multimodal Learning Analytics (MMLA) solutions are tailor-made for specific learning tasks and sensors, the VIT addresses the data annotation for different types of learning tasks that can be captured with a customisable set of sensors in a flexible way. The VIT supports MMLA researchers in 1) triangulating multimodal data with video recordings; 2) segmenting the multimodal data into time-intervals and adding annotations to the time-intervals; 3) downloading the annotated dataset and using it for multimodal data analysis. The VIT is a crucial component that was so far missing in the available tools for MMLA research. By filling this gap we also identified an integrated workflow that characterises current MMLA research. We call this workflow the Multimodal Learning Analytics Pipeline, a toolkit for orchestration, the use and application of various MMLA tools.

Collaboration is an important 21st century skill; it can take place in a remote or co-located setting. Co-located collaboration (CC) gives rise to subtle human interactions that can be described with multimodal indicators like gaze, speech and social skills. In this demo paper, we first give a brief overview of related work that has identified indicators during CC. Then, we look briefly at the feedback mechanisms that have been designed based on these indicators to facilitate CC. Using these theoretical insights, we design a prototype to give automated real-time feedback to facilitate CC taking the help of the most abundant modality during CC i.e., audio cues.

Learning analytics can bridge the gap between learning sciences and data analytics, leveraging the expertise of both fields in exploring the vast amount of data generated in online learning environments. A typical learning analytics intervention is the learning dashboard, a visualisation tool built with the purpose of empowering teachers and learners to make informed decisions about the learning process. Related work has investigated learning dashboards, yet none have explored the theoretical foundation that should inform the design and evaluation of such interventions. In this systematic literature review, we analyse the extent to which theories and models from learning sciences have been integrated into the development of learning dashboards aimed at learners. Our analysis revealed that very few dashboard evaluations take into account the educational concepts that were used as a theoretical foundation for their design. Furthermore, we report findings suggesting that comparison with peers, a common reference frame for contextualising information on learning analytics dashboards, was not perceived positively by all learners. We summarise the insights gathered through our literature review in a set of recommendations for the design and evaluation of learning analytics dashboards for learners.

The analysis of multimodal data in learning is a growing field of research, which has led to the development of different analytics solutions. However, there is no standardised approach to handle multimodal data. In this paper, we describe and outline a solution for five recurrent challenges in the analysis of multimodal data: the data collection, storing, annotation, processing and exploitation. For each of these challenges, we envision possible solutions. The prototypes for some of the proposed solutions will be discussed during the Multimodal Challenge of the fourth Learning Analytics & Knowledge Hackathon, a two-day hands-on workshop in which the authors will open up the prototypes for trials, validation and feedback.

Recommender systems provide users with content they might be interested in. Conventionally, recommender systems are evaluated mostly by using prediction accuracy metrics only. But the ultimate goal of a recommender system is to increase user satisfaction. Therefore, evaluations that measure user satisfaction should also be performed before deploying a recommender system in a real target environment. Such evaluations are laborious and complicated compared to the traditional, data-centric evaluations, though. In this study, we carried out a user-centric evaluation of state-of-the-art recommender systems as well as a graph-based approach in the ecologically valid setting of an authentic social learning platform. We also conducted a data-centric evaluation on the same data to investigate the added value of user-centric evaluations and how user satisfaction of a recommender system is related to its performance in terms of accuracy metrics. Our findings suggest that user-centric evaluation results are not necessarily in line with data-centric evaluation results. We conclude that the traditional evaluation of recommender systems in terms of prediction accuracy only does not suffice to judge performance of recommender systems on the user side. Moreover, the user-centric evaluation provides valuable insights in how candidate algorithms perform on each of the five quality metrics for recommendations: usefulness, accuracy, novelty, diversity, and serendipity.

Collaboration is an important 21st century skill; it can take place in a remote or co-located setting. Co-located collaboration (CC) is a very complex process which involves subtle human interactions that can be described with multimodal indicators (MI) like gaze, speech and social skills. In this paper, we first give an overview of related work that has identified indicators during CC. Then, we look into the state-of-the-art studies on feedback during CC which also make use of MI. Finally, we describe a Wizard of Oz (WOz) study where we design a privacy-preserving research prototype with the aim to facilitate real-time collaboration in-the-wild during three co-located group PhD meetings (of 3–7 members). Here, human observers stationed in another room act as a substitute for sensors to track different speech-based cues (like speaking time and turn taking); this drives a real-time visualization dashboard on a public shared display. With this research prototype, we want to pave way for design-based research to track other multimodal indicators of CC by extending this prototype design using both humans and sensors.

Almost all studies on course recommenders in online platforms target closed online platforms that belong to a University or other provider. Recently, a demand has developed that targets open platforms. Such platforms lack rich user profiles with content metadata. Instead they log user interactions. We report on how user interactions and activities tracked in open online learning platforms may generate recommendations. We use data from the OpenU open online learning platform in use by the Open University of the Netherlands to investigate the application of several state-of-the-art recommender algorithms, including a graph-based recommender approach. It appears that user-based and memory-based methods perform better than model-based and factorization methods. Particularly, the graph-based recommender system outperforms the classical approaches on prediction accuracy of recommendations in terms of recall.