Thesis offers

Start: possible immediately

Description: This project investigates how the quality of learner-generated tasks can be reliably assessed. The starting point is a project in which the categorization by experts and AI systems did not match. The initial work will therefore critically examine how reliable and selective the existing categorization scheme for evaluating learner-generated tasks actually is: How consistent are assessments by experts? Where is there room for interpretation or inconsistencies in the category system? Based on these findings, the category system is to be optimized. In a second final thesis, the extent to which AI-supported categorization can deliver comparable results will be investigated. The aim is to develop methods that can improve both the human and algorithmic evaluation of task quality and make it more transparent. The work combines qualitative analysis approaches with quantitative comparison methods and thus shows ways in which (AI-supported) qualitative categorizations can be improved in order to reliably assess task quality in education.

In cooperation with Aline Mangold (Prof. Technical Design, )

Supervision: Dr. Antje Proske (antje.proske@tu-dresden.de)

Degree program: MA HPSTS

Start: possible immediately

Description: This experimental thesis examines how learners perceive and use AI-generated feedback compared to peer feedback. Particular attention is paid to whether it makes a difference if the AI feedback is labeled as such or if it is designed to look like peer feedback. The aim is to find out how constructively learners can work with the different types of feedback and which form of feedback they find particularly supportive. The results should help to better understand the potential of AI feedback for the learning process and derive recommendations for its design.

Supervision: Dr. Antje Proske (antje.proske@tu-dresden.de)

Degree program: MA HPSTS

Start: possible immediately

Description: This experimental thesis investigates why generating own tasks does not lead to a higher learning performance compared to answering given tasks, although this would theoretically be expected. The focus is on the role of the tutor's perspective and the granularity of the task content: It is examined whether taking a tutor perspective and using tasks created by learners themselves make a difference in knowledge transfer. The results should help to identify conditions under which the generation of tasks is a particularly effective learning strategy.

Supervision: Dr. Antje Proske (antje.proske@tu-dresden.de)

Degree program: BA Psychology

Start: Immediately

Study Program: M.Sc. Human Performance in Socio-Technical Systems (HPSTS)

Description: Every new technology-based measurement tool needs to be validated with another reliable and valid method, under normal circumstances. This research project is designed to validate the knee sleeve´s performance in monitoring knee movement during squats and lunges. It will involve a thorough literature review to gain insights into existing knee monitoring technologies and video analysis techniques. The research design will focus on defining the target population, selecting relevant participants, and utilizing the Kinovea program for video analysis during squat and lunge exercises. The data analysis process will involve the measurement and comparison of knee angles obtained from both the knee sleeve and the recording analyzed via Kinovea program. Subsequently, statistical methods will be employed to evaluate and establish the accuracy of the knee sleeve in monitoring knee movement. The study aims to provide comprehensive insights into the knee sleeve´s effectiveness, with conclusions and recommendations for technology improvement, alongside detailed documentation, and presentation of findings.

Data: Data has to be collected.

The project will be in English.

Contact: ismail.bayram@tu-dresden.de

Start: Immediately

Study Program: M.Sc. Human Performance in Socio-Technical Systems (HPSTS)

Description: People with flatfoot can suffer from pain, ankle instability, rapid fatigability. On the long term, they can develop various musculoskeletal pathologies in the lower limb joints. While foot orthoses are commonly prescribed for people with flatfoot, their efficiency is left to be proven. Indeed, various systematic reviews questioned the routine prescription of foot insoles (1–3). Particularly, Hoang et al. (4) concluded that active interventions had better efficacy in reducing pain. Indeed, active interventions improve joint stability through muscle strengthening, while also enhancing proprioceptive feedback. Accordingly, augmenting sensory feedback during gait to indicate the propre muscle activation scheme and plantar foot distribution could reduce pain associated with flatfeet during gait, and improve the arch definition on the long run.

Thus, within this project, we want to develop a real-time vibrotactile feedback based on kinematic and plantar pressure data to improve foot kinematics during gait. Modalities of feedback will be investigated, as well as the effect of the intervention on the kinematics and kinetics gait parameters to infer if such intervention could decrease the risk of pathologies of the lower limb joints.

The project will be in English.

1. Herchenröder et al. J Foot Ankle Res. 2021;14(1):57.
2. MacKenzie et al. J Pediatr Orthop. 2012;32(8):830‑4.
3. Banwell et al. J Foot Ankle Res. 2014;7(1):23.
4. Hoang et al. Int J Environ Res Public Health. 2021;18(15):8063.

Contact: