data science vs. statistician


A recent news report shows that “data scientist” is being searched more than “statistician“. A few notes: what this suggests to me is that traditional academic disciplines can no longer contain the skill set needed to manipulate and analyze big data. Too applied for math. Too CS for stats. Social science programs are too focused on topic and substance. We already have “informatics,” but now it will further split into a group that does big data handling vs. other tasks. I hope they maintain an applied focus and don’t retreat into algorithms for algorithms sake.

$3 is cheap for top notch career advice: From Black Power/Grad Skool Rulz 

Written by fabiorojas

December 21, 2013 at 4:58 am

Posted in fabio, mere empirics

2 Responses

Subscribe to comments with RSS.

  1. A bunch of groups of people (both proprietary and open source) will build the next Excel, Stata, Matlab, and R that can handle large data sets. Then all that stats and science will have to do is remake their course materials.


  2. I think this is also a big cultural shift in the way knowledge is produced, or “epistemic practices”. Economics and other disciplines that used statistics were long based on authoritative arguments or “good stories”. There would be no policy interventions based solely on descriptive stats because “correlation is not causation”.
    Big data is quite different, to my knowledge. Mostly the argument is all about patterns and no theory. This is not even the old engineering mindset, but a new engineering mindset. Big part of this mindset is experimentation. Causal theories have little use if one can simply run A/B test and figure out if the pattern provides useful guideline for action or not.
    It will be interesting to see how the social scientists who embrace big data type of analytics will do their theories.



    December 21, 2013 at 7:41 pm

Comments are closed.

%d bloggers like this: