If you have to use statistical significance testing then it is not good science
With good science you would control & understand the factors enough that the results consistently fit the results predicted by the model & if they do not consistently fit then you would make a new model instead of saying that they kind of do fit it with an alpha value of whatever so now I do not have to understand & control the factors but merely say it is or is not statistically significant
Social "scientists" like sociologists who promote totalitarian types of socialism usually prefer statistics
Have you noticed how rarely social scientists choose algebraic, trigonometric or calculus models involving equations that predict an output for a given input
Instead they usually choose statistical significance testing that does not predict a specific output for a given input but only gives a answer of statistically significant or not statistically significant
@Phynster
Real Science like what Isaac Newton allegedly discovered did not rely on Statistical Signifcance Testing, they did not even have that nonsense when he was discovering the important stuff
Real Science uses Algebra, Trigonometry and Calculus instead of Statistical Significance testing