Page 160 - Getting the Picture Modeling and Simulation in Secondary Computer Science Education
P. 160

158
Chapter 7
students’ difficulties related to the algorithmic approach and to count the errors related to each of the three procedural areas related to common student mistakes. Similarly, Basu et al. (2018) developed assessment tasks for the integration of CT in physics and again, the students were asked to complete a partially completed program. The rubrics used for the assessment assessed two aspects: expressing physics relations in a computational model and using programming concepts to model physics phenomena. In both of these examples, physics provides the modeling context. The aspects under scrutiny were labeled either as correct or incorrect. As to the modeling cycle, in comparison to our assessment instrument, only the design and the implementation of the models were of interest, rather that the whole modeling cycle. Louca et al. (2011) analyzed the computational models of a number of physics phenomena constructed by their students by using categories for the representation of objects, entities, behaviors and interaction, and additionally, for the accuracy of the phenomenon description. Each of these categories contains a number of subcategories specific for the context. Taken together, these categories are reminiscent of the description of the attainment levels specified in our rubrics. In further comparison to our assessment instrument, we see that in addition to design and the implementation of the models, their validity is of interest too — albeit to a limited extent, as discussed in the section 7.2.2 on students understanding and difficulties.
Other researchers have constructed assessment for models in their own right. For science teaching and learning, Papaevripidou et al. (2014) describe modeling competence in terms of modeling practices and meta-knowledge. They identify four modeling practices: construction, use, comparison and revision of models; they distinguish different levels of increasing sophistication for each of the modeling practices and they observe that these levels are independent of the modeling tool. While they are not concerned with computational models, we see similarities in the modeling practices they mention: construction and use are represented in our modeling cycle too, and their revision of models is represented in our framework in the fact that our framework considers modeling to be a cyclic process. Furthermore, for each of the modeling practices, they distinguish several levels of sophistication — an approach seemingly similar to our five ordered categories of SOLO taxonomy. However, they assess each of these practices as a whole. For example, for the practice of model use, they assess efficient use of the model without specifying details that characterize efficient use, while we provide detailed characterization of each of the levels in our assessment instrument.































































































   158   159   160   161   162