Because the current symbols, H0: p=0, gave rise to a strong anxiety among Psychologists disguised in Statisticians, we advise to change it . . . For example H0: p not distinguishable to 0, say ______________H0: p=0 n.d. With this latter they stop (perhaps) to believe that the proposal is to state the arithmetic nullity of the parameter p. Conversely Ha: p=/0 would be _____________Ha: p=/0 a.s. (a.s. read as, almost surely). We can improve this way to set NHST: ______H0: p=0 n.d. at 5% level and n=40 ______Ha: p=/0 a.s. at 5% level and n=40 This caution is advisable because a lot of Psychologists say that ´increasing the sample size, or the level, an insignificant value turns up significant´, the undisputable proof that the procedure is untenable. Furthermore, they say, the Null Hypothesis is surely always wrong . . . By the way . . . Every time you find a significant value you do not fail the advise; THIS VALUE DOES NOT. AT ALL, MEANS THAT YOU HAD FOUND SOMETHING PRACTICALLY RELEVANT.
(Am I, or not, a true guardian angel towards all Scientific People excluding, of course, Psychologists that, them, no one is for that . . .).