Musings and Rants:
A Musing
First in a series

In statistics, the term "statistically significant" is unwieldy. Let's shorten it and make official the usage of its portmanteau, "statsig."

When reporting statistical results, it is common to use the parlance, "statistically significant" when the effect that a factor exerts is unlikely to be due to chance alone. However, in practice, this is an unwieldy phrase.

Accordingly, it is common for journal articles to drop the adverb, leaving only "significant" to describe these results. However, this is poor usage of the language, as "significant" by itself implies that the result may be important. Those familiar with statistics understand that some results may be statistically significant, but are hardly significant (i.e., not due to chance, but not important either).

My prescription is simple. Let's compress the phrase into the portmanteau, "statsig." That way, we can retain the proper meaning while inducing brevity.

How about it, America? Let's make "statsig" common parlance.

Back to Musings and Rants home

Page last updated: 18 June 2003
This page and all its contents are Copyright © 2003 Jerry S. Tsai