Observations as a Method to Evaluate a Computer-Based Approach to Learning Algorithms

This paper has two contributions. Firstly, we present the results of using observations as a complementary method to evaluate an active approach to learning greedy algorithms based on the interactive assistant GreedEx. Two kinds of observations were conducted: questions to the teacher, and students´ activity. Observations were useful in two ways: they reinforced findings of two evaluation methods -usability questionnaires and analysis of students´ reports-, and they allowed identifying new, interesting data to further enhance our approach. A second contribution is the methodological lessons learnt about how to use observations as a research method. In particular, we found human observation useful for some purposes but not with respect to software usage.