Since the advent of the Internet, we have been observing a fast-paced development within the computing world. One of the major innovations in recent years is the “Internet of Things”, which brings interconnectedness between devices and humans to unprecedented heights. This technological breakthrough enabled the emergence of a new sub-field within Learning Analytics, Multimodal Learning Analytics, which makes use of several types of data sources to study learning-related processes. As computers and sensors become increasingly cheaper and more accessible,  research within this new sub-field grows, yet some gaps remain unexplored. Additionally, there is a research bias toward computer-assisted learning environments, rather than physical ones. At the same time, the current labor market is highly competitive, and possessing profession-related skills is not sufficient to land a job. Besides these skills, there is an increasing demand for social skills, such as communication, teamwork, and collaboration. However, there is a gap between the skills that are trained in an academic setting and the ones that are required by the labor market. Having this background in mind, this work aims at designing and evaluating an IoT sensor system capable of tracking patterns observed under social interactions within a group, and more specifically, in terms of the distance between group members while solving a task. Another important aspect of this study is the system's cost-effectiveness so that it can be employed in a scalable and sustainable manner. To achieve this goal, a multimethodological approach for Design Science Research was adopted, which implied the combination of several methods such as sketching, prototyping, and testing. As a result, this study contributes both to the research area of Multimodal Learning Analytics, and to educational practices.

Learning Analytics plays an increasing role in the analysis of virtual learning activities. This article addresses the gap between educational needs and technical supply. By means of a Systematic Literature Review of the LAK conferences the authors extracted observations, methods and tools which represent potential solutions for a given eCollaboration scenario. Based on three prioritised examples of an observation sheet, methods are derived and recommendations for the use of Learning Analytics tools are given. The result is a catalogue that enable users to select suitable methods and tools for an implementation. The (semi-) automation can increase the efficiency of Community Managers in monitoring the participants and hence make real-time intervention feasible.

Wissen bekommt in unserer Gesellschaft eine immer stärkere Bedeutung und stellt Individuen und Organisationen vor verschiedenste Herausforderungen. Der gesamtwirtschaftliche Anteil materieller Güter wird gegenüber wissensintensiven Dienstleistungen zurückgehen, d.h. es bedarf zukünftig immer stärker gut aus- und weitergebildeter Experten, die miteinander die Innovationskraft von Organisationen steigern und kundenspezifische Lösungen entwickeln. Ausgehend von den für Experten benötigten Kompetenzen sogenannter Wissensarbeiter fokussiert diese Arbeit die Begleitung und Bewertung von kollaborativer Teamarbeit. Als Forschungsobjekt dienen Virtual Collaborative Learning-Veranstaltungen, in denen die zukünftigen Experten als Lernende teilnehmen. Sie werden in virtuellen Klassenräumen von Lernbegleitern beobachtet und bewertet, um sie bei der Entwicklung neuer Kompetenzen zu unterstützen und so die Erreichung der Lernziele zu gewährleisten. Da die Lernbegleitung bislang manuell durchgeführt wird, ist eine zeitnahe Beurteilung und Intervention nicht möglich. Mit Learning Analytics sollen Lösungen vorgeschlagen werden, die die Lernbegleiter in ihrer Arbeit unterstützen und den Aufwand reduzieren. Hierzu thematisiert diese Dissertation in sechs kumulativen Beiträgen, wie die zugrundeliegenden Daten verfügbar gemacht werden können und welche Beobachtungen anhand welcher Kriterien mittels ausgewählter Methoden der Learning Analytics durchgeführt werden sollten. Im Rahmen des Design Science Paradigmas werden verschiedene qualitative und quantitative Forschungsmethoden zur Datenerhebung und -auswertung angewendet. Im Ergebnis entsteht ein beispielhafter Katalog von Learning Analytics Methoden, die sich auf spezifische Erkenntnisziele der Beobachtungen von kollaborativer Gruppenarbeit beziehen. Ebenso wird der Einsatz einer ausgewählten Methode evaluiert. Die gewonnenen Erkenntnisse können zukünftig auf vergleichbare kollaborative Lehrangebote in der Ausbildung von Wissensarbeitern, aber auch zur Analyse und Unterstützung der virtuellen Zusammenarbeit im geschäftlichen Umfeld transferiert werden.

Den för datalogin grundläggande datastrukturen hashtabell är krävande att tillägna sig. En undersökning gjordes på en kurs i algoritmer och datastrukturer med 200 deltagare. Lärandemålet ”implementera hashtabell och hashfunktion” bröts ner i grundläggande färdigheter som testades i en automaträttad programmeringsuppgift med riktad återkoppling till studenterna. 86 studenter gjorde 334 försök att lösa uppgiften. Undersökningen visade att testerna som ingår i den automaträttade uppgiften svarar mot de fel studenterna gör. Studenternas fel kategoriserades efter de grundläggande färdigheter som tagit fram. Kategoriseringen kan användas för att identifiera svaga områden hos studenterna och modifiera undervisningen därefter. Försöken visar också att när uppgiften kräver samtidig tillämpning av två begrepp leder detta ofta till fel i implementationen av algoritmen eller ineffektiva lösningar. / The hash table data structure, which is fundamental to computer science, is demanding to learn. A survey was conducted in a course on algorithms and data structures with 200 participants. The learning outcome of implementing a hash table was broken down into basic skills that were tested in an automated programming task with targeted feedback to the students. 86 students made 334 attempts to solve the task. The study showed that the tests included in the automated task correspond to the errors made by the students. The students' errors were categorized according to the basic skills developed. The categorisation can be used to identify weak areas in the students and modify the teaching accordingly. The experiments also show that when the task requires the simultaneous application of two concepts, this often leads to errors in the implementation of the algorithm or inefficient solutions.

Inom den högre utbildningen är det fortfarande vanligt att använda traditionella undervisningsmetoder även om det finns andra metoder som går att tillämpa. En faktor som påverkar införandet av nya metoder är digitaliseringen. Många lärosäten använder sig idag av en digital plattform för undervisningen, ibland också flera stycken olika plattformar och där dessa då kräver en teknisk kompetens för att kunna användas på ett bra sätt i undervisningen. Speciellt är kurser inom datalogi som har goda förutsättningar för att kunna nyttja nyare metoder. Dock kräver detta kunskap i metoden, hur de tekniska hjälpmedlen går att tillämpa för kursens ändamål och att man vågar pröva nya metoder i kurser. Målet med denna studie är hitta nya sätt som kan mäta en undervisningsmodells effektivitet. Lösningen kommer att bestå i att undersöka existerande forskning inom området, specifikt implementation av metoden, vad gav metod för effekt och hur denna effekt har mätts. Av den existerande forskningen pekar på att detta är en svår fråga att analysera, det är många faktorer som vi behöver ta i anspråk för att kunna komma till konsensus i hur den potentiella lösningen bör skapas. Den gemensamma faktorn är att det finns en positiv inverkan på studenternas resultat baserat på att det har införts en ny undervisningsmodell. Det vi kom fram till är att skapa en mätmetod som använder studenternas lösningar för befintliga uppgifterna i kursen, genom att vi fick göra ett antal antaganden för de problem som ansågs kunna påverka mätmetoden. Av detta kunde vi sedan dra slutsatser av metodens effektivitet. / In higher education, it is still common to use traditional teaching methods, although there are other methods that can be applied. One factor influencing the introduction of new methods is digitalization. Many higher education institutions today use a digital platform for teaching, sometimes also several different platforms and where these then require a technical competence to be used in a good way in teaching. Especially are courses in computer science that have good conditions for being able to use newer methods. However, this requires knowledge of the method, how the technical aids can be applied for the purpose of the course and that you dare to try new methods in courses. The goal of this study is to find new ways that can measure the effectiveness of a teaching model. The solution will consist of investigating existing research in the area, specifically implementation of the method, what gave method for effect and how this effect has been measured. From the existing research indicates that this is a difficult issue to analyze, there are many factors that we need to take into account in order to be able to come to a consensus on how the potential solution should be created. The common factor is that there is a positive impact on student outcomes based on the introduction of a new teaching model. What we came up with is to create a measurement method that uses the students’ solutions for existing tasks in the course, by making several assumptions for the problems that were considered to affect the measurement method. From this we were then able to draw conclusions from the effectiveness of the method

Актуальность работы состоит в необходимости персонального онлайн обучения, включая высокое качество передачи знаний одновременно с автоматизацией и уменьшением затрат со стороны модераторов курса. Невозможность существующих систем обеспечить этим многих пользователей ведет к поиску новых подходов. Научная новизна основана на внедрении технологии интеллектуального чат бота с функциями обработки естественного языка в среду обучения и новом подходе к выполнению заданий, исключающем лимиты времени и количества попыток проверить ответ, а также декомпозиция сложных задач на шаги. Это позволит ученикам достичь правильный ответ самостоятельно, тем самым повышая вовлечение и мотивацию понять трудную тему. Работа содержит следующие стадии: литературный анализ, моделирование, проектирование, реализация, результаты и прогноз. Благодаря предложенному решению больше пользователей получат возможность внедрить приложение и распространить знания с улучшенной средой онлайн обучения. Студенты смогут иметь поддержку в течение всего образовательного процесса в любое время, а их прогресс будет обеспечен анализом данных и методами машинного обучения. / The relevance of the work is grounded by the necessity of the personalized online learning, including the high-quality education providing with better automation and less efforts from moderators` side. Inability of the current systems to open these opportunities for many users leads to the exploration of new approaches. The scientific novelty is based on the embedment of the intellectual chatbot technology with Natural Language Processing features into the learning environment and the new approach of the assignments` accomplishment, which inferences the elimination of the time and attempt limits as well as complex tasks on steps segmentation. It gives the possibility for learners to reach the correct answer in more independent way, thereby increasing the engagement and motivation to perceive the difficult topics. The work contains stages: Literature analysis, Modeling, Designing, Implementation, Results and Forecast. Due to the solution more users are able to employ the application and distribute the knowledge with improved eLearning environment. The students get the opportunity to have the support during the educational process any time and their progress is maintained by the data analysis and advanced technologies.

Learning Analytics beschreibt das Messen, Sammeln, Analysieren und Berichten von Daten, um Lernprozesse verstehen und verbessern zu können (Siemens und Long 2011). Eng verwandt mit Learning Analytics sind die Bereiche Academic Analytics und Educational Data Mining, die mit jeweils unterschiedlicher Ausrichtung ebenso die datenbasierte Überprüfung, Vorhersage und Änderung von akademischen Prozessen untersuchen (Baepler und Murdoch 2010). In diesem Beitrag fokussieren wir entsprechende Benutzungsschnittstellen, welche die gesammelten Daten visualisieren und verfügbar machen. ... [Aus: Einleitung]

Det som ska undersökas i denna fallstudie är ifall visualisering av studerad tid kan öka motivationen hos en högskolestudent att studera på högskola. I denna uppsats definieras studietillfällen som schemalagda föreläsningar och övningar, eller självorganiserade studiepass. Kopplingen mellan motivation och visualisering av tid lagt på träning i gym har tidigare undersökts. Vi bygger vidare på det och undersöker ifall det finns ett liknande samband med mängden tid nedlagd på studier och motivationen att studera vid högre utbildning. Frågan är av intresse för främst utvecklare av lärplattformar, då ifall att fallstudiens resultat skulle tyda på att visualisering av tid nedlagt på studier orsakar en ökad motivation skulle det vara värdefullt att implementera i lärplattformar. Vidare, är frågan av intresse för studenter eftersom det skulle kunna öka deras studiemotivation. Fallstudien inleddes med att 28 studenters motivation mättes, för att få fram ett referensvärde. Därefter fick studenterna under en tvåveckorsperiod dagligen logga den tid de lagt på studier. Utöver det fick studenterna dagliga uppdateringar med stapeldiagram som visuellt representerade den loggade tid de dagligen lade ner på sina studier. Därefter mättes deras motivation återigen, som sedan jämfördes med den inledande mätningen. Resultatet visade att det fanns en signifikant skillnad i två av de sju motivationsskalorna i “Academic Motivational Scale”, den inre motivationen att prestera och den yttre introjicerade motivationen. Detta styrker tidigare studier som visat en koppling mellan visualisering av en elevs nedlagda tid på avklarade moment och elevens prestation i skolan. / This case study aims to examine if a visualization of undergraduate students’ learning sessions can increase their motivation to study. Previous studies have shown that there is a correlation between motivation and visualization of time spent on, for example, training in gyms. Consequently, this case study aims to expand on this subject and examine if there is a similar correlation in time spent on studying and the motivation to study in higher educational settings. This question is mainly of interest for developers of learning management systems, such as online course platforms, since if the case study results show that there is an increase in motivation caused by a visualization of time spent studying it would be of value to implement in their learning management systems. Furthermore, this question is of interest for students since it could increase their motivation to study. The case study was initialised by measuring and evaluating 28 undergraduate students’ motivation, to calculate a reference value. The students were then asked to log their study sessions each day and they were given daily updates based on their loggings of their time spent studying during two weeks, with visual bar charts. The case study finished by re-evaluating the motivation of the students and comparing it to the initial measurement. The results showed that there was a significant difference in 2 of the 7 motivational categories of the employed “Academic Motivational Scale”, namely the intrinsic motivation towards accomplishment, and the extrinsic introjected motivation. This further validates previous studies which shows a connection between visualising a students total time spent on accomplished course modules.

When students use an online eTextbook with content and interactive graded exercises, they often display aspects of two types of behavior: credit-seeking, and knowledge-seeking. Any given student might behave to some degree in either way in a given assignment. In this work, we look at multiple aspects of detecting the degree to which either behavior is taking place, and investigate relationships to student performance. In particular, we focus on an eTextbook used for teaching Formal Languages, an advanced computer science course. This eTextbook is using Programmed Instruction (PI) framesets to deliver the material. We take two approaches to analyze session interactions in order to detect credit-seeking incidents. We first start with a coarse-grained approach by presenting an unsupervised model that clusters the behavior in the work sessions based on the sequence of different interactions that happens during them. Then we perform a fine-grained analysis where we consider the type of each question in the frameset, which can be a multi-choice, single-choice, or T/F question. We show that credit-seeking behavior is negatively affecting the learning outcome of the students. We also find that the type of the PI frame is a key factor in drawing students more into the credit-seeking behavior to finish the PI framesets quickly. We implement three machine learning models that predict students' midterm and overall semester grades based on their amount of credit-seeking behavior on the PI framesets. Finally, we provide a semisupervised learning model to aid in the work session labeling process. / Master of Science / Students frequently exhibit features of two types of behavior when using an online eTextbook with content and interactive graded exercises: credit-seeking and knowledge-seeking. When solving homework or studying a material, students can behave in either manner to some extent. In this research, we study links between student performance and different elements of recognizing the degree to which either behavior is occurring. We concentrate on an eTextbook used to teach an advanced computer science course, Formal Languages and Automata, using a teaching paradigm called Programmed Instruction (PI). In order to detect credit-seeking instances, we use two ways to study students' behavior in the Programmed Instruction sessions. We begin with a coarse-grained approach by building a model that can categorize work sessions into two groups based on the interactions that occur throughout them. Then we do a fine-grained analysis in which we analyze the question types in the framesets and their effect on the students' behavior. We show that credit-seeking behavior has a negative effect on students' learning outcomes. We discovered that the PI frame type is an important factor in enticing students to engage in credit-seeking behavior in an attempt to finish PI framesets fast. Finally, we present three predictive models that can forecast the students' midterm and total semester grades based on their credit-seeking behavior on the Programmed Instruction framesets.

Providing a sufficient level of personalized feedback on students' current level of strategic knowledge within the context of the natural programming environment through IDE-based learning analytics would transform learning outcomes for introductory programming students. However, providing sufficient insight into the programming process was previously inaccessible due to the need for more complex and scalable data collection methods and metrics with a wider variety for understanding programming metacognition and the full programming process. This research developed a custom-built web-based IDE and event compression system to investigate two of the five components of a five-dimensional model of cognition for programming skill estimation (1) Design Cohesion and (2) Development Path over Time. The IDE captured students' programming process data for 25 participants, where each participated in two programming sessions that required both a design and code phase. For Design Cohesion, the alignment between flowchart design and source code implementation was investigated and manually classified. The classification process produced three Design Cohesion metrics: Design Cohesion Level, Granularity Level, and Granularity Score. The relationship between programming skill and Design Cohesion was explored using the newly developed metrics and a case-study approach. For the Development Path over Time, the compressed programming events were used to create a Timeline of Events for each participant, which was manually examined for distinct clusters of programming patterns and behavior such as execution behavior and debugging patterns. Custom visualizations were developed to display the timelines. Then, the timelines were used to compare programming behaviors for participants with different programming skill levels. The results of the investigation into Design Cohesion and Development Path Over Time contribute to the fundamental understanding of differences between beginner, intermediate, and advanced programmers and the context in which specific programming difficulties arise. This work produced insight into students' programming processes that can be used to advance the model of cognition for programming skill estimation and provide personalized feedback to support the development of programming skills and expertise. Additionally, this research produced tools and metrics that can be used in future studies examining programming metacognition.