Federico -

Well, you do need to be careful with hypothesis tests.  An isolated p-value is often misleading.  It is a particularly bad idea to use the same level in all cases.

 https://www.researchgate.net/publication/262971440_Practical_Interpretation_of_Hypothesis_Tests_-_letter_to_the_editor_-_TAS

Cheers - Jim

Article Practical Interpretation of Hypothesis Tests - letter to the...

My comment is that the paper is over ten years old and I haven't seen a rush, even in psychology journals, to ban p-values. I suggest that an editor banning a scientific procedure outright on the basis of one (his own) paper only is not a very good idea. I am awaiting a flood of papers supporting Trafimow's position with cogent arguments before I stop using p-values. Best wishes.

Federico -

Well, you do need to be careful with hypothesis tests.  An isolated p-value is often misleading.  It is a particularly bad idea to use the same level in all cases.

 https://www.researchgate.net/publication/262971440_Practical_Interpretation_of_Hypothesis_Tests_-_letter_to_the_editor_-_TAS

Cheers - Jim

Article Practical Interpretation of Hypothesis Tests - letter to the...

Having hypothesis and p-value, depending on the case you can use the corresponding one to each case.

I think it's not a totally right desicion . There are many researchers who have abused what P values , but would have been more appropriate to make an analysis on the appropriate use of statistical inference before make a desicion. I attach more info about the banning and context.

Hypothesis testing employs inferential statistics - there's no implicit statement of absolute truth. The best you can gain from a p-value is an inference of what MIGHT be true.

It's because of the lack of precision and certainty that we have set up a replicative system of scientific inquiry. We don't make important changes on the basis of a single study - regardless of how well structured, executed and explained they may be. We wait until there is confirmation of the findings in the hope that each confirmation of the findings allows us to become a little more confident that what the statistics have demonstrated is, in fact, the truth.

I think it would be a wise decision to stop reading the BASP as the Editor is obviously a fool. Just my opinion.

Thank you all for your answers, especially Jim and Oneida for the references. I think like mostly of the responses, the statistic approach of a research must be encouraged from the start in the research design and calculus of the sample size, then in the correct interpretation of statistical results that could be more than a p-value (effect size, confidence intervals). In respect to BASP only the time will give them the reason or not, it necessarily will be reflected in their ranking position and H index. Anyhow the exigency or not for correct use of statistical analysis vary from one journal to other, it could be time that statisticians and epidemiologists play a major role in the review of manuscripts and that the editorial boards of the journals consider their inclusion in the process. I mean: not only peer review, specific statistical review too.

Best regards.

Federico - 

I generally agree with your statement, except that, partially analogous to the p-value, journal 'rankings' may not be very meaningful either.

Cheers - Jim