Theoretical math based on gambling was formalized by the French mathematician Pierre Fermat in the 1650s, from his exploration of probability theory came a theorem from Reverend Thomas Bayes. Bayes work was well reasoned from the axioms of mathematics, however, it was computationally unfeasible, and therefore lived only in theory until the 1990s, when algorithms made Bayes’ theorem computationally feasible and improved dramatically the world of statistics and machine learning.
Charles Darwin’s cousin, Francis Galton, worked with a mathematical model made by Johann Gauss to develop the idea of statistical regression, using statistics to analyze observational data. Observational data is abundant in our digital age. Observational data is automatically being collected and could be used to inform research experiments. Today everyone who plans on researching using the scientific method ought to understand collecting and interpreting observational data. To see things through the new lens of online data we need to pull from the world of the theoretical, arrived to on the ship of deduction, to build data finding and processing algorithms.
The past called for specialization of a field of knowledge like sociology, math, computer programming, economics or statistics. Those who specialized typically were the more successful in their research as they could focus on something so specific that no one had attempted it before. Today is the age in which flexibility is the dominantly trait of a world changing researcher. At present we have so much information recorded. It would inform focused research so well in social and organizational sciences. However, there is a division between mathematicians, statisticians and social scientists that needs to be breached.
We need to change how the classes are taught so that students are ready to explore beyond the forte of their field. Statisticians need to reason and derive their formulas deductively, they need to explore being a mathematician until they are comfortable coming up with models and formulas on their own. Mathematicians need to estimate results before learning how to derive them exactly and reason through them totally soundly. Social scientists need to become technology savvy and learn how to interpret and embrace new types of numerical data. Doing this will open all three to the possibility of understanding the social and organizational world around them and performing more useful and significant research.
Great post. As an actuarial science major, (basically a stat major at BYU) I couldn't agree more with your post. I think that your thought are valid especially because (like you pointed out) the world is changing. Data is everywhere. This calls for a change in authorities and frames of references of all people you mentioned. I think that was the only thing I was hoping for in your post, that didn't happen at the end. What do we need to do to make this change happen?
ReplyDelete