# AnalyticBridge

A Data Science Central Community

Subscribe to DSC Newsletter

# How can social scientists make elementary mistakes with probability yet continue to thrive professionally?

SUB-TITLE:   MAYBE "Standard Deviation", as a scientific term, should be 'thrown out of our vocabulary" as a goal for 2014?

Nassim Nicholas Taleb  -   Distinguished Professor of Risk Engineering, New York University School of Engineering ; Author:  THE BLACK SWAN, and many other books......

.....................that the term STANDARD DEVIATION is very misleading and not a term that should be used by commentators, newspaper writers, and the general public  but ONLY by MATHEMATICAL STATISTICIANS (if at all !!!!).   Why?   Nassim writes "......as it does more harm than good—particularly with the growing class of people in social science mechanistically applying statistical tools to scientific problems."

Nassim continues on:   "...It is all due to a historical accident: in 1893, the great Karl Pearson introduced the term "standard deviation" for what had been known as "root mean square error"."

If you want to see what other "colorful words" are used to describe "social scientists" you will have to read his blog:  http://www.edge.org/response-detail/25401

Views: 971

### Replies to This Discussion

Standard deviation is dangerous with outliers, as is the mean.

It simply measures the SPREAD of the data, but so do other methods and GRAPHS ARE BEST (The outlier boxplot for example).

So most of my work involves robust statistics, like MEDIAN and QUARTILES or IQR (the inter-quartile range which is simply the 75% - 25% range of the data...as seen in the BOX of the boxplot, which works well with data that has outliers, but shows the spread of the "middle half" of the data.

So yes, I could survive without the stdev for most work, and I am an engineer, but perhaps in same position as "social scientists" as far as lacking a PhD in Statistics, but slightly saner than much of the media in use of data.

## On Data Science Central

© 2020   AnalyticBridge.com is a subsidiary and dedicated channel of Data Science Central LLC   Powered by