Nowadays there is a noticeable lack of validated evaluation instruments that can measure the different stages of initial programming learning. For this reason, we developed an instrument from Bloom's Taxonomy that allows to compare the students' performance regardless of the programming language with which they started their learning. The instrument presents multiple choice questions for the first four levels and open-ended questions for the last two levels. After its construction, it was validated by 12 specialists and submitted to a pilotgroup (N=29) to analyze the items' reliability and difficulty. We obtained an alpha of 0,837 and positive correlation between al the test's items and the total of items with a significance level of p < 0,05. The differences between the frequency of correct answers in the items are statistically significative (Q(12)=76,767; n=29; p=0,000). These results demonstrate that the test is valid, reliable and discriminative.
CITATION STYLE
Costa, J. M., & Miranda, G. L. (2017). Desenvolvimento e validação de uma prova de avaliação das competências iniciais de programação. RISTI - Revista Iberica de Sistemas e Tecnologias de Informacao, 2017(25), 66–81. https://doi.org/10.17013/risti.25.66-81
Mendeley helps you to discover research relevant for your work.