Whether we should abandon "statistically significant" is one of the hottest ongoing arguments in science (as of Apr. 2019). Shall we? If we do not rely on this magical word, what technique should we use alternatively? Here I briefly summarize the recent argument about Null Hypothesis Significance Testing (NHST). 

In short,

DO NOT SAY "STATISTICALLY SIGNIFICANT" ANY MORE; SHOW EFFECT SIZE & CONFIDENCE INTERVAL.

Yet, reporting p-value per se is fine).  It is increasingly important to show both actual data values w/ its distribution and the difference in mean from control (effect size w/ confidence interval). Some easy-use tools for doing it are available more and more.  Hereafter, I just show some of the most important sentences taken from the key articles published in the recent few years.