Im Artikel wird ein fächerübergreifendes Service Learning Projekt der Friedrich-Schiller-Universität (FSU) Jena vorgestellt. Zunächst werden die Zielgruppe, die inhaltliche Gestaltung und die Rahmenbedingungen skizziert und im Anschluss die Struktur und Organisation aufgezeigt. Abschließend werden einige Erfolgsfaktoren für eine gelingende Implementation an der FSU genannt.

Die europaweite Einführung von Bachelor- und Masterstudiengängen im Zuge des Bologna-Prozesses ist unumkehrbar. Die damit verbundenen Hauptziele der größeren internationalen Mobilität und flexibleren Studienorganisation bleiben richtig. Dennoch sind die Probleme unübersehbar. Studierendenproteste gegen die Umsetzung der Studienreform waren und sind notwendig, denn Bundes- und Landesregierung sowie Hochschulen schieben sich die Verantwortung für Fehlentwicklungen gegenseitig in die Schuhe. Viele Hochschulen bauen vor dem Zugang zum Master hohe Hürden auf. Zum Teil ist nur für ein Drittel der BachelorabsolventInnen ein Master vorgesehen. Sogar bei fachlich aufeinander aufbauenden (konsekutiven) Studiengängen werden Zugangsbedingungen geschaffen, die über den notwendigen Erwerb eines Bachelor-Abschlusses hinausgehen. Die Studiengebührenfreiheit an sächsischen Hochschulen gilt nur für den konsekutiven Master.

Philosophiae Doctor - PhD / Micro-teaching provides a framework for teachers’ professional development with emphasis on the teach and re-teach cycle as a practical procedure for the development of micro-teaching skills. One fundamental element in this teach and re-teach cycle is the use of digital technology tools as a recording device which may help aid the understanding and application of micro-teaching skills. This study explores micro-teaching skills with the use of digital technology (cell phone) in B.Ed. programmes at a Western Cape university. The study aims to merge the past and present strategies to initiate the development of a simple model to improve micro-teaching methodology using a simple and mostly available digital technology tool which can allow for self-evaluation and personal reflection. The study seeks to investigate how a digital technology, through a cell phone, can guide practical learning to support pre-service teachers in the micro-teaching presentation. The underpinning theories that frame this study were Vygotsky’s Socio-cultural Theory, Bandura’s Social Learning Theory and Mishra & Koehler’s Technological Pedagogical Content Knowledge (TPACK) Theory. These theories emphasize social interactions, modelling and the use of technological tools as a significant aspect of skills development. The combination of these theories informs the conceptually developed knowledge acquisition, knowledge construction and the implementation with technology (KACIT) model with the aim to encourage the construction and implementation of micro-teaching knowledge with available digital technology tools. The study employs a sequential explanatory design, using a case study approach, which draws on both quantitative and qualitative methods as sources of data collection. The quantitative procedure adopted video recording as a form of data collection, using a conceptually developed teaching skills rubric, with a three level Likert-scale rating. SPSS version 25 was used to get the aggregated descriptive statistical values of each teaching skill. The qualitative procedure employed unstructured interviews, lesson plan observation, and thematic and content analysis was used to interpret and analyse the interviews and lesson plans respectively. The sample size of this study is drawn from the B.Ed. students in the Faculty of Education at the University of the Western Cape (UWC). Two hundred and sixteen (216) participants were involved in the quantitative data collection phase, while the qualitative data collection phase consisted of twelve (12) participants for the focus group discussion, five (5) school-based supervisors, five (5) pre-service teachers and (5) five lesson plans from the students-portfolio documents. The findings attest to the general context of micro-teaching as a strategy for developing pre-service teachers‟ teaching skills. It further highlights the high theoretical knowledge levels of micro-teaching of the B.Ed. students, and the relative lack of practical knowledge of micro-teaching of the students. In addition, the findings reveal the high technological knowledge levels of the participants and the need to deepen the professional knowledge of the B.Ed. students along the lines of authentic teaching experiences within the university environment. As part of the recommendations, the study emphasises the use of a digital video platform as a complementary strategy and as a form of reflective practice in micro-teaching presentations. Importantly, the study further recommends that the Minimum Requirements for Teacher Education Qualifications (MRTEQ) policy document, on the knowledge mixes for teaching competences in terms of the observation outcome, should be re-considered with a clear guideline for South African universities to have a uniform framework and a clear picture of its implementation.

Student attrition from universities carries high costs for individuals, universities, and society. Despite these costs, there has been limited research on the problem from the students' perspectives, specifically the perceptions of university graduates about what factors may have influenced their own retention at historically Black colleges and universities (HBCUs). African American students complete college at the lowest rate compared to other ethnic sub groups. Guided by Tinto's theory of student departure, this qualitative consensual research study focused on the perceptions of 15 bachelor of social work graduates regarding (a) the factors that helped them to persist to graduation at an HBCU, and (b) the internal and external factors that influenced college retention. Data were gathered from semi-structured interviews, a demographic survey, and a follow-up questionnaire. Data were transcribed; member checked for enhanced trustworthiness; and then analyzed inductively using a team to develop and code domains by consensus, construct core ideas, and develop categories. Findings indicated that, among these 15 graduates, internal factors such as loss of scholarships, lack of faculty support, and lack of academic preparation influenced retention. External issues such as family strain and lack of university/community partnerships were also reported as influencing retention. In addition, findings suggested that these 15 students required increased academic, financial, mental health, and social support services to persist to graduation. This study contributes to social change by affecting improvement in retention and ensuring support services equal student needs. Improvements in retention and support could help to grow an educated and skilled work force.

This dissertation includes three essays that examine the potential of policies, interventions, and curricula in improving transitions to four-year colleges and universities in the U.S. The first essay provides a novel investigation of New York City’s adoption of SAT School Day (SSD), which provides universal access to college entrance exams for high school juniors. I examine the causal impact of the policy on SAT taking and four-year college enrollment patterns for the first three post-policy cohorts. Using a two-way fixed effects differences-in-differences (DID) model, I find that the SSD policy increases SAT-taking rates by 20% but has little impact on four-year enrollment. These findings suggest that increased college entrance test participation alone is not sufficient to increase enrollment to four-year colleges. The second essay provides a systematic review and meta-analysis of 19 studies on the causal impact of late pre-college advising interventions, which I refer to as LPCAIs. LPCAIs aim to connect juniors and seniors with pre-college advisors to guide them through the complex college application process. The studies use experimental and quasi-experimental designs to identify the causal effect of LPCAIs on postsecondary enrollment outcomes for historically underrepresented students. I find that the offer of an LPCAI, on average, increases postsecondary enrollment and improves institutional choices. However, the in-person advising format accounts for nearly all of the positive impact derived from LPCAIs. Several implications of these findings are discussed, particularly in the context of COVID-19. The third chapter presents a correlational study that uses transcript data from a statewide community college system to estimate how well GPA, credit accumulation, and course taking at community college predict upward transfer and baccalaureate completion. While GPAs and credit accumulation rates in all academic branches are strong, positive predictors of transfer, GPAs in humanities are significantly more predictive of transfer compared to GPAs in non-humanities branches. GPAs and successful credit accumulation rates in humanities courses are also associated with greater likelihoods of baccalaureate completion, conditional on transfer. However, humanities courses are equally valid predictors of baccalaureate completion as courses in other academic branches. In contrast to strong academic performance, taking more courses in humanities, although associated with greater likelihoods of vertical transfer, is associated with lower likelihoods of baccalaureate completion. Taken together, these essays contribute to our understanding of approaches to improve transitions to four-year colleges and universities. They also highlight the importance of academic preparation and advising within transitional pathways.

The purpose of this study is to see what the students include in the area of police work and police science A summary of bachelor thesis hasn’t been done before. This thesis contribute to understand how police science in Sweden is interpreted by students at undergraduate level. The police science discipline is emerging and we have to make knowledge of what it is. All the studies that are included in this study are made by sworn police officers or police academy students that are qualified. That means that all studied bachelor thesis are made by persons within the police domain. By using grounded theory by Barney Glaser in the analysis, the author of this study has discovered themes of working method, leadership, working environment, law, equality, information and other. The result of this study shows that the law is the biggest theme and also working method and working environment are most of interest. Although even sports and foreign service is included as police work. This study has also been related to three Nordic overwievs from the early 90´s to 2011, to see similarities and differences of the author´s results. In the future there will be interesting to see if the themes that´s been found in this study are the same or if it´s changing over time.

The music industry has seen a great influx of new channels to browse and distribute music. This does not come without drawbacks. As the data rapidly increases, manual curation becomes a much more difficult task. Audio files have a plethora of features that could be used to make parts of this process a lot easier. It is possible to extract these features, but the best way to handle these for different tasks is not always known. This thesis compares the two deep learning models, convolutional neural network (CNN) and long short-term memory (LSTM), for music genre classification when trained using mel-frequency cepstral coefficients (MFCCs) in hopes of making audio data as useful as possible for future usage. These models were tested on two different datasets, GTZAN and FMA, and the results show that the CNN had a 56.0% and 50.5% prediction accuracy, respectively. This outperformed the LSTM model that instead achieved a 42.0% and 33.5% prediction accuracy. / Musikindustrin har sett en stor ökning i antalet sätt att hitta och distribuera musik. Det kommer däremot med sina nackdelar, då mängden data ökar fort så blir det svårare att hantera den på ett bra sätt. Ljudfiler har mängder av information man kan extrahera och därmed göra den här processen enklare. Det är möjligt att använda sig av de olika typer av information som finns i filen, men bästa sättet att hantera dessa är inte alltid känt. Den här rapporten jämför två olika djupinlärningsmetoder, convolutional neural network (CNN) och long short-term memory (LSTM), tränade med mel-frequency cepstral coefficients (MFCCs) för klassificering av musikgenre i hopp om att göra ljuddata lättare att hantera inför framtida användning. Modellerna testades på två olika dataset, GTZAN och FMA, där resultaten visade att CNN:et fick en träffsäkerhet på 56.0% och 50.5% tränat på respektive dataset. Denna utpresterade LSTM modellen som istället uppnådde en träffsäkerhet på 42.0% och 33.5%.

The aim with this project was to investigate how the amount and composition of resins varied during the process producing bleached birch pulp at the mill SCA Packaging Munksund. A literature study about how the resin removal can be improved has also been included. Problems with resins in the process are common at pulp and paper mills, especially when birch is used as a raw material. The resin can cause deposits on the equipment leading to process stops, but also lowered mechanical properties and spots on the paper products. The addition of tall oil to the digester is one way of improving the removal of resins, seasoning of wood, and a good debarking are other ones. Also the different washing and bleaching steps can affect the amount of resin remaining in the pulp. In this study pulp samples from eight different positions in the process were analyzed. To extract the samples a Soxtec device was used. Results showed that the most effective resin removal happened during the washing in their first washing step after the digester, a DD-washer. Here 77 % of the resin was removed, of totally 88 % during the whole process. Another step which was effective was the final washing step, the PO-press. About 36 % of the remaining resins in the pulp which entered the PO-press were washed out here. The extracts were analyzed with GC-FID and GC-MS to identify and quantify the substances, and determine how the composition varied over the manufacturing process. Twelve different compounds were identified, and the birch bark resin, betulinol, turned out to be the hardest component to remove. Over all, the sterols and triterpenols were hard to deresinate, while the removal of resin acids, fatty acids, and fatty alcohols was more effective. The PO-reactors showed a positive effect on the fragmentation of components consisting of long carbon chains and double bonds, like squalene and betulaprenols. / Målet med projektet var att utreda hur mängden och sammansättningen av hartser varierade i den blekta björkmassan under tillverkningsprocessen på pappersbruket SCA Packaging Munksund. Att göra en litteraturstudie för att ta reda på vilka metoder som kan användas för att förbättra avhartsningen var en annan del av målet. Att pappers- och massatillverkare har problem med hartser är vanligt, speciellt när björkved används som råmaterial. Hartserna kan orsaka beläggningar på utrustningen som i sin tur kan leda till processtop, men även prickar och försämrade mekaniska egenskaper hos pappersprodukten. Att tillsätta tallolja till kokaren är en vanlig metod som används för att förbättra hartsreningen, att lagra träet innan koket, och lägga fokus på att förbättra barkningen är andra metoder. Även de olika blek- och tvättstegen som används under tillverkningen av massan kan påverka hartshalten. I denna studie analyserades massaprover från åtta olika positioner i processen. En Soxtec-utrustning användes för att extrahera massaproverna. Resultatet visade att den mest effektiva avhartsningen sker under tvätten i det första tvättsteget efter kokaren, DD-filtret. Här tvättades 77 % av hartserna ut, och totalt minskade mängden hartser med 88 % under hela processen. PO-pressen var ett annat tvättsteg som var effektivt ur avhartsningssynpunkt. Här togs 36 % av den kvarvarande mängden hartser bort från massan. Extrakten som erhållits från extraktionen analyserades vidare med GC-MS och GC-FID. Dessa metoder användes för att identifiera och kvantifiera extraktivämnena, och bestämma hur sammansättningen varierade under tillverkningsprocessen. Tolv olika ämnen identifierades och den substans som visade sig vara svårast att tvätta bort var björkbark hartset, betulinol. Stort sett så var steroler och triterpenoler svårast att bli av med, medans avhartsingen av hartssyror, fettsyror och fettalkoholer var mer effektiv. Det visade sig även att PO-reaktorerna hade en sönderdelande effekt på substanser såsom squalen och betulaprenoler, dessa består av långa kolkedjor med många dubbelbindningar mellan kolatomerna.

Christiane is a 7 meter sailing vessel and a 1:30 scale model of Oceanbird, a wind powered car carrier planned to be launched by the year 2024. The model is used for design validation and to research control algorithms. However, the development of new algorithms is limited to the microcontrollers used, which are not able to run resource intense applications. Furthermore, there is no easy way for new computational nodes to interact with the vessels system. In this thesis, a high-level, easy to use, interface was proposed, designed and implemented. This was done by identifying the the requirements and limitations of the project and examining the system to find a suitable solution. The two biggest limitations found were that the solution had to include a single-board computer and that the implementation could not drastically alter the system architecture of the model. The implementation included a single-board computer being added on top of the models architecture which publishes all gathered sensor data and actuator commands on a robotics operating system interface. The interface was an immediate improvement to the usability of the vessel since it allows for graphical presentation of gathered data in real time. / Christiane är en 7 meter lång segellfarkost och en modell i skala 1:30 av Oceanbird, ett vinddrivet lastfartyg för fordon som är planerad att sjösättas år 2024. Modellen används för att validera designen samt för forskning kring kontrollalgoritmer. Utvecklingen av nya algoritmer är dessvärre begränsad till systemets mikrokroller, vilka inte klarar av att köra resursintensiva applikationer. Det finns inte heller något enkelt sätt för nya beräkningsnoder att interagera med båtens system. Denna rapport föreslår ett högnivå-gränssnitt till modellen som är enkelt att använda och presenterar en implementation av det som underlättar framtida forskning. Detta gjordes genom att först identifiera projektets krav och begränsningar samt genom att undersöka modellens system för att hitta en lämplig lösning. De två största funna begränsningarna var att lösningen var tvungen att inkludera en enkortsdator och att implementationen inte fick innebära några större förändringar av modellens system-arkitektur. Implementationen bestod av att en enkortsdator lades till ovanpå modellens arkitektur som publicerar all samlad sensordata och alla kommandon för ställdon på ett robotics operating system gränssnitt. Gränssnittet gav en direkt förbättring av fartygets användarvänlighet då det möjliggjorde grafisk presentation av den samlade datan i realtid.

One of the most globally common types of cancer is skin cancer, where melanoma is the most deadly form. An important and promising tool for diagnosing diseases such as skin cancer is computer aided diagnostics, a tool which utilizes machine learning to predict and classify cancer. Limiting the complexity of the data, known as feature selection, can potentially improve classification accuracy. This report evaluates the accuracy of four different classifiers - Support Vector Machine, Naive Bayes, Decision Tree and Artificial Neural Network - with four different feature selection methods - Sequantial Forward Selection, Sequantial Backward Selection, Entropy and Principal Component Analysis - on the PH2 skin cancer dataset, containing dermoscopic images of skin lesions and their respective metadata. The findings reveal that all feature selection methods led to an improved accuracy rate on at least one classifier compared to not using feature selection. Furthermore, certain feature selection methods resulted in a significant gain in accuracy, indicating the potential value of feature selection techniques in improving the accuracy and efficiency of machine learning classifiers in computer-aided diagnosis systems for melanoma skin cancer detection. However, the results also underscore the importance of careful selection of the number of features to avoid adverse effects on model performance. This research contributes to the field by demonstrating the impact of feature selection methods on melanoma skin cancer detection and highlighting considerations for their application. / En av de globalt vanligaste typerna av cancer är hudcancer, där melanom är den mest dödliga typen. Ett viktigt och effektivt verktyg för att diagnostisera sjukdomar som hudcancer är datorstödd diagnostik, ett verktyg som använder maskininlärning för att förutse och klassificera cancer. Att begränsa komplexiteten i data, känt som attributurval, kan potentiellt förbättra klassificeringsnoggrannheten. Denna rapport utvärderar noggrannheten hos fyra olika klassificerare - ”Support Vector Machine”, ”Naive Bayes”, ”Decision Tree” och ”Artificial Neural Network” - med fyra olika attributurvalsmetoder - ”Sequantial Forward Selection”, ”Sequantial Backward Selection”, ”Entropy” and ”Principal Component Analysis” - på PH2 hudcancerdatasetet, som innehåller dermoskopiska bilder av hudlesioner och deras respektive metadata. Resultaten visar att alla attributurvalsmetoder ledde till en förbättrad noggrannhetsgrad på minst en klassificerare jämfört med att inte använda attributurval. Dessutom resulterade vissa attributurvalsmetoder i en betydande ökning i noggrannhet, vilket indikerar det potentiella värdet av attributurvalstekniker för att förbättra noggrannheten och effektiviteten hos maskininlärningsklassificerare i datorstödda diagnossystem för detektering av melanom hudcancer. Däremot understryker resultaten också vikten av noggrant urval av antalet attribut för att undvika negativa effekter på modellens prestanda. Denna forskning bidrar till fältet genom att demonstrera inverkan av attributurvalsmetoder på detektering av melanom hudcancer och belysa överväganden för deras tillämpning.