A new report released by CAUSE is well worth reading: _Connecting Research to Practice in a Culture of Assessment for Introductory College-level Statistics, _www.causeweb.org/research/guidelines/ResearchReport_Dec_2012.pdf

Read it.  We’ll discuss later.  Pop quiz.

I haven’t yet read it myself (in my eagerness to publicize it as quickly as possible),  but of particular interest to this blog is the role that data science plays, or does not play.  For instance, Question 1 under Research Priority 1 is “What core learning outcomes  employed in a particular profession do individuals  need to develop in order to perform well in that profession (e.g., the outcomes that are common and those that are unique to disciplines such as psychology, biology, and economics?)”

I recently had a discussion with someone in a data-heavy business, and was struck by how core statistical concepts were seen as just one of many necessary core skills—the rest of the skills requiring computing, psychology, and communication.  It is fashionable in statistical circles to be somewhat dismissive of claims that computation take precedence over statistics, but in at least this case, I think that this paints an unfair portrait.  The data scientist in question held statistics in high esteem, and was well aware of the pitfalls of being lured by transitory patterns, as compelling as they might at first glance seem.  His use of statistics came at the ‘high end’, employing very modern data smoothing techniques, multivariate models, and a need for sophisticated understanding of model evaluation.  But he, like many in his field I suspect, came to statistics in a round-about way, after becoming successful in computer science and then studying statistics to close the gap.  I doubt he considered himself a statistician, but instead one who frequently found statistical tools and concepts to be useful for getting things done.

We’re in a very exciting position, as educators, to dream about how to develop future data scientists who incorporate statistics with computation from the very beginning of their conception of statistics.  But one thing to keep in mind, is that part of the excitement of this new age of statistics is that many of the careers we’re preparing our students for don’t yet exist.  It seems so many of the data challenges that are raised in a general realm of endeavor such as marketing,  the arts, genetics,  law, have solutions that don’t live purely in one field.  And so when we ask ourselves about skills needed in particular professions, let’s do so with our eyes open to the fact that the profession that many of us have in mind —data science—doesn’t really yet exist.

Data science is, or will be, a specialist’s field.  But this blog is devoted to considering the data science skills needed by all students.  I think, therefore, that the issues raised by this report concerning the ‘core’ skills are very important.  A data scientist may have a specialists' collection of skills, in the aggregate, but many of this skills and understandings, in isolation, will need to be part of our core education.  This report encourages us all to think seriously about precisely which skills and understandings those should be.