Science Literacy and media Literacy: a missing link?

The study of science in the media is increasingly highlighted within science programmes and represents an authentic context for interdisciplinary collaboration. Yet the literature on ‘media across the curriculum’ makes surprisingly little mention of links to science and cross-curricular approaches to teaching about and with science-based media resources is an area that is under-explored. This research study focuses on science in the news. The project involved 28 teachers from seven schools and brought together science and English teachers to explore collaborative working with the aim of promoting critical engagement with media reports with a science component. Teachers planned, developed and implemented a school-based activity with an emphasis on ‘connected learning’ rather than the compartmentalised learning that tends to accompany the discrete treatment of science matters in science class and media matters in English class. Not only did the project raise teachers’ awareness of science in the media as a potential, purposeful and profitable area for collaborative working, but it demonstrated how the synergy of the different experiences and expertise of science and English teachers produced very varied approaches to a programme of activities with an enhanced capacity to promote criticality in relation to science literacy and media literacy.

Learning Extended Tree Augmented Naive Structures

This work proposes an extended version of the well-known tree-augmented naive Bayes (TAN) classifier where the structure learning step is performed without requiring features to be connected to the class. Based on a modification of Edmonds' algorithm, our structure learning procedure explores a superset of the structures that are considered by TAN, yet achieves global optimality of the learning score function in a very efficient way (quadratic in the number of features, the same complexity as learning TANs). We enhance our procedure with a new score function that only takes into account arcs that are relevant to predict the class, as well as an optimization over the equivalent sample size during learning. These ideas may be useful for structure learning of Bayesian networks in general. A range of experiments shows that we obtain models with better prediction accuracy than naive Bayes and TAN, and comparable to the accuracy of the state-of-the-art classifier averaged one-dependence estimator (AODE). We release our implementation of ETAN so that it can be easily installed and run within Weka.