I blame the use of statistical signifance tests instead of a return to models using, algebra, trigonometry and calculus as the source of this problem
I also blame the lack of raw data being presented when these statistical significance tests are done
& the choice to assign an alpha value prior to testing and get a answer of significant or not significant based on that alpha instead of only giving the P value and not choosing an alpha value to give a false sernse of objectivity
There are also many social problems resulting in the replicability crisis but I believe even if some of these social problems are solved the replicability crisis will not be solved until we additionally abandon statistical significance tests
If the worldview involved in statistical significance tests is true then their will be a certain percent of experiments that actually get wrong results and changing the alpha value will not fix it but only change the kind of wrong results
