Data reasoning is a relatively automatic process of summarizing large quantities of information, write the authors, but can be susceptible to things like cognitive biases and fallacies. So the tools of scientific reasoning, including "external representations, scientific hypothesis testing, and drawing probabilistic conclusions," are needed to reduce the liklihood of such errors. To that same end, Nicholas Friedman and Stephen Esser argue elsewhere that the philosophy of science should be taught in high school. I wouldn't disagree. Anyhow, this article (19 page PDF) proposes a model of data reasoning that includes elements such as data sensemaking, "a process that includes declarative (e.g., scientiï¬c facts) and procedural (e.g.,conducting unconfounded experiments) elements," which is then supplemented with the tools of scientific data reasoning.
Today: 1 Total: 16 [Direct link] [Share]
