|
Abstract:
|
As universities and colleges rush to offer courses and even degree programs in data science, it's fair to wonder whether data science is genuinely new or is merely a rebranding of statistics. This round-table will discuss important and substantial ways that a statistics course that genuinely engages data science differs from traditional statistics. These include an emphasis on prediction, classification and causality rather than the traditional focus on estimation and significance. As a background for the discussion, you may want to refer to a new book, *Stats for Data Science* (available at https://dtkaplan.github.io/SDS/preface.html) which provides ideas for topics and pedagogy. The book illustrates some of the ways that opening up the intro statistics curriculum to data science can make statistics courses more useful and exciting and better correspond to the precepts of the ASA GAISE report and the ASA statement on p-values.
|
