Type of Publication: Article in Collected Edition
Evaluating Responses in Source Code Highlighting Tasks - Preliminary Considerations for Automatic Assessment
- Kramer, M.; Barkmin, M.; Brinda, T.
- Mühling, A.; Cutts, Q.; Schwill, A.
- Title of Anthology:
- Proceedings of the 13th Workshop in Primary and Secondary Computing Education (WIPSCE 2018), Potsdam, Germany, 4.-6. Oktober 2018
- ACM Press
- New York
- Publication Date:
- Object-Oriented Programming, Highlighting, Assessment
- Digital Object Identifier (DOI):
- Link to complete version:
- Download BibTeX
The reading and understanding of source code is a necessary prerequisite for prospective programmers to write their own code. Thus, a reasonable task is to determine certain concepts like class identifiers in a given source code. In order to automatically evaluate the responses, they must be quantified. Hence, it is vital to have reliable and valid methods of measurement. In this paper we apply different means of measurement on constructed examples and compare them. Cohen's κ seems to be an adequate measure.