Automatic Evaluation of Tasks for Instantaneous Diagnostics in Computer Science Lessons

Decanato - Facoltà di scienze informatiche

Data: 26 Febbraio 2020 / 09:30 - 10:30

USI Lugano Campus, room SI-013, Informatics building (Via G. Buffi 13)


Mike Barkmin, University of Duisburg-Essen, Germany


For researching learning processes researchers of all subjects are challenged to find new ways for extracting substantial information from solutions when dealing with big groups. Using complex and authentic tasks in Online-Assessment-Systems seems to be a promising approach to overcome this challenge on the item-level. Another potential of Online-Assessment-Systems on the test-level is to use dynamic, highly individualized sets of tasks to maximize the diagnostic value for each individual learner. 

In the talk conceptual considerations for an Online-Assessment-System, which is currently under development at the Chair of Computer Science Education at the University of Duisburg-Essen are presented. The considerations have their starting point in the question of which types of tasks in the field of computer science can be offered in an Online- Assessment-System with a focus on diagnosis. For this purpose several research studies were carried out, the results of which are presented in the talk. Based on this, it is shown which diagnostic information the Online-Assessment-System can provide - compared to paper-based conventional diagnostic formats. In particular, the diagnostic potential of recorded processes and automatically categorized key moments will be presented. In this context, research is used to show how the recordings can be used to understand cognitive processes while programming. 

Finally, potential extensions of the Online-Assessment-System with regard to privacy, automatic evaluation of the recorded processes and adaptivity of the selection of tasks are discussed. 


Mike Barkmin is a research assistant and doctorial student at the Chair of Computer Science Education at the University of Duisburg-Essen. His main research interest is the transferabilty of programming skills from secondary education to university, especially the transfer between programming languages and paradigms. He holds a Master of Education for Mathematics and Computer Science. 

Host: Prof. Matthias Hauswirth