How can we help students develop an understanding of chemistry that integrates conceptual knowledge with the experimental and computational procedures needed to apply chemistry in authentic contexts? The current work describes ChemVLab+, a set of online chemistry activities that were developed using promising design principles from chemistry education and learning science research: setting instruction in authentic contexts, connecting concepts with science practices, linking multiple representations, and using formative assessment with feedback. A study with more than 1400 high school students found that students using the online activities demonstrated increased learning as evidenced by improved problem solving and inquiry over the course of the activities and by statistically significant improvements from pre- to posttest. Further, exploratory analyses suggest that students may learn most effectively from these materials when the activities are used after initial exposure to the content and when they work individually rather than in pairs.

Three experiments investigate how self-generated explanation influences children's causal learning. Five-yearolds (N = 114) observed data consistent with two hypotheses and were prompted to explain or to report each observation. In Study 1, when making novel generalizations, explainers were more likely to favor the hypothesis that accounted for more observations. In Study 2, explainers favored a hypothesis that was consistent with prior knowledge. Study 3 pitted a hypothesis that accounted for more observations against a hypothesis consistent with prior knowledge. Explainers were more likely to base generalizations on prior knowledge. Findings suggest that attempts to explain drive children to evaluate hypotheses using features of "good" explanations, or those supporting generalizations with broad scope, as informed by children's prior knowledge and observations.

No abstract

Automated guidance on essays and drawings can improve learning in precollege and college courses.

Many tasks in NLP stand to benefit from robust measures of semantic similarity for units above the level of individual words. Rich semantic resources such as WordNet provide local semantic information at the lexical level. However, effectively combining this information to compute scores for phrases or sentences is an open problem. Our algorithm aggregates local relatedness information via a random walk over a graph constructed from an underlying lexical resource. The stationary distribution of the graph walk forms a "semantic signature" that can be compared to another such distribution to get a relatedness score for texts. On a paraphrase recognition task, the algorithm achieves an 18.5% relative reduction in error rate over a vector-space baseline. We also show that the graph walk similarity between texts has complementary value as a feature for recognizing textual entailment, improving on a competitive baseline system.